Campus-Wide Information Systems Volume 21, Number 5, 2004
ISSN 1065-0741
Selected papers from the Fourth International Network Conference 2004, Plymouth, UK Guest Editor: Steven Furnell
Contents 174 Access this journal online 175 Abstracts & keywords 177 Guest editorial 179 An evaluative methodology for virtual communities using web analytics A.D. Phippen 185 Knowledge broker network based on communication between humans Robert Loew, Udo Bleimann and Paul Walsh 191 Atlantis University: a new pedagogical approach beyond e-learning Udo Bleimann
196 Text appeal: the psychology of SMS texting and its implications for the design of mobile phone interfaces Fraser J.M. Reid and Donna J. Reid 201 The future of mobile technology and mobile wireless computing Jim Hart and Mike Hannan 205 Translating scalable video streams from wide-area to access networks Carsten Griwodz, Steffen Fiksdal and PaÊl Halvorsen 211 An audio stream redirector for the Ethernet Speaker Ishan Mandrekar, Vassilis Prevelakis and David Michael Turner 217 Note from the publisher 219 Call for papers 220 Awards for Excellence
Access this journal electronically The current and past volumes of this journal are available at:
www.emeraldinsight.com/1065-0741.htm You can also search over 100 additional Emerald journals in Emerald Fulltext:
www.emeraldinsight.com/ft See page following contents for full details of what your access includes.
authors mean by man-man communication. The basic idea of the knowledge broker is that of a hybrid man-machine system that enables knowledge transfer within companies not only theoretically, but also in practice.
Abstracts & keywords
Atlantis University: a new pedagogical approach beyond e-learning Udo Bleimann Keywords Learning methods, Virtual organizations, Universities
An evaluative methodology for virtual communities using web analytics A.D. Phippen Keywords Ethnography, User studies, Worldwide web The evaluation of virtual community usage and user behaviour has its roots in social science approaches such as interview, document analysis and survey. Little evaluation is carried out using traffic or protocol analysis. Business approaches to evaluating customer/business web site usage are more advanced, in particular using advanced web analytics to develop greater understanding of their customers’ use of their sites. The application of such techniques to virtual communities is discussed and experimentation of such techniques on a specific virtual community project demonstrates the potential for such techniques in the evaluation of social and culture web usage.
Knowledge broker network based on communication between humans Robert Loew, Udo Bleimann and Paul Walsh Keywords Knowledge management, Information brokers, Man-machine systems This paper proposes a new paradigm to overcome many problems in the area of information technology involving communication between humans. The paradigm will be exemplified using the example of knowledge transfer within a company. The solution is not interaction between a user and a highly intelligent system, but communication between people, supported by intelligent systems. The paper’s concept of the “knowledge broker” will clarify what the
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · Abstracts & keywords q Emerald Group Publishing Limited · ISSN 1065-0741
Atlantis University is an ambitious international project in the area of learning and is currently being developed by a group of universities and companies. It combines three different types of learning and teaching to form a single package offered to students and people in the workplace alike: face-to-face learning, e-learning and project-based learning. The paper gives an overview of the advantages and disadvantages of the different learning methodologies, and describes the new Atlantis approach. The first practical solutions Atlantis University has developed, namely the Virtual Classroom, ELAT learning environment and Project Service Center, are likewise briefly introduced.
Text appeal: the psychology of SMS texting and its implications for the design of mobile phone interfaces Fraser J.M. Reid and Donna J. Reid Keywords Mobile communication systems, Psychology, Gender Argues that understanding the psychological drivers behind SMS uptake among key user groups could open the door to a range of user-centred applications capable of transforming handset usability – and hence operator revenues – for this inexpensive form of text messaging. Combines the findings of our own web-based survey of SMS users with psychological evidence and research on related text-based conversational systems to draw out lessons for a userbased approach to the design of mobile phone handset displays that capitalise on the social affordances of SMS texting.
The future of mobile technology and mobile wireless computing Jim Hart and Mike Hannan Keywords Mobile communication systems, Computers, Forecasting It is often stated that mobile wireless computing is going to be the next big technology revolution that will grip the world in the same way mobile telephones did in the 1990s. However, while the technology is
175
Abstracts & keywords
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · 175-176
rapidly improving, the rate of uptake has been lower than expected. This paper describes some of the reasons for this, and discusses some of the proposed solutions.
basic functions and help in discovering appropriate modifications. For the investigation, an experimental approach was used where network behavior and content are emulated.
Translating scalable video streams from wide-area to access networks
An audio stream redirector for the Ethernet Speaker
Carsten Griwodz, Steffen Fiksdal and Pa˚l Halvorsen
Ishan Mandrekar, Vassilis Prevelakis and David Michael Turner
Keywords Video, Internet, Transmission Control Protocol/Internet Protocol, Data handling
Keywords Audio media, Internet, Protocols
Transmitting video over user datagram protocol (UDP) has been considered advantageous because it allows for discarding of packets in favor of retransmissions, and sender-controlled timing. Using UDP has been criticized because it allows video streams to consume more than their fair share of bandwidth, which is typically associated with the back-off behavior of transmission control protocol (TCP). TCP-friendly algorithms are meant as a middle path. However, UDP delivery to end systems may still be prevented by firewalls or for other reasons, and TCP must be used. This in turn suffers from bandwidth fluctuations. Investigates an architecture that separates the transfer of a video stream over long distances. Considers a proxy server to translate the traffic and two straightforward approaches for the translation of a layered video stream transmission from the TCP-friendly transport protocol to TCP. Does not expect that one of these two approaches is by itself suited for the task, but investigating them will provide insights into their
The authors have developed the “Ethernet Speaker” (ES), a network-enabled single board computer embedded into a conventional audio speaker. Audio streams are transmitted in the local area network using multicast packets, and the ES can select any one of them and play it back. A key requirement for the ES is that it must be capable of playing any type of audio stream, independent of the streaming protocol, or the encoding used. The authors achieved this by providing a streaming audio server built using the kernel-based audio stream redirector (ASR) in the OpenBSD kernel. The ASR accepts input from any of the existing audio file players, or streaming audio clients. Since all audio applications have to be able to use the system audio driver, this system can accommodate any protocol or file format, provided that there exists some compatible player running under OpenBSD. This paper discusses the design and implementation of the server as an ASR, the streaming protocol developed for this application, and the implementation of the client.
176
Guest editorial
About the Guest Editor Steven Furnell is based at the Network Research Group, School of Computing, Communications and Electronics, University of Plymouth, Plymouth, UK.
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 177-178 q Emerald Group Publishing Limited · ISSN 1065-0741
The Fourth International Network Conference (INC 2004) The papers in this issue of Campus-Wide Information Systems are based on a selection of those that were presented at INC 2004, the Fourth International Network Conference, which was held in Plymouth, UK, from 6-9 July 2004. INC events have been running since 1998, and provide a forum for sharing the latest research in computer networks and related technologies. The main themes addressed by the 2004 conference were Web Technologies and Applications, Network Technologies, Security and Privacy, Mobility, and Applications and Impacts. The full conference proceedings include a total of 70 papers, with coverage ranging from discussion of applications and services, down to details of specific underlying technologies[1]. The papers selected for this issue have been chosen to be representative of the broad range of topics, while at the same time addressing areas of relevance to the journal readership. The papers address a combination of technology- and application-level issues. Beginning very much at the technology level, Griwodz, Fiksdal and Halvorsen consider some of the problems inherent in sending video streams from wide area networks into the access networks of the end systems, and examine two approaches for translating the streams between the two domains. Staying with streaming technologies, Mandrekar, Prevelakis and Turner describe the design and implementation of a flexible audio stream redirector, and its application within the “Ethernet Speaker” – an embedded device that is able to play audio streams received from the network. Looking at communication from a higher level of abstraction, Loew, Bleimann and Walsh identify the problem of managing and sharing knowledge within an organisation. As a means of overcoming the difficulties, they propose the concept of an intelligent ‘knowledge broker’, which aims to improve the process of knowledge transfer within a Thanks are due to Dr Paul Dowland, Jayne Garcia, Denise Horne, and the various members of the Network Research Group without whose support and involvement the conference would not have taken place. Thanks must also go to the numerous members of the INC 2004 International Programme Committee, who assisted with the original review process. Finally, the guest editor is very pleased to acknowledge the support of the British Computer Society, the Institution of Electrical Engineers, Orange, Symantec, and of course Emerald (the publishers of Campus-Wide Information Systems) as co-sponsors of the conference.
177
Guest editorial
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · 177-178
company. The topic of communication between people is further explored by Reid and Reid, albeit from a rather different perspective. Their paper examines the psychological issues behind one of the most popular forms of electronic communication to have emerged in recent years – the mobile phone text message. Based on their findings from a survey of end-users, the authors consider the implications for the design of future text-based messaging systems. The human issues of technology are also a theme in the paper from Phippen, who considers the concept of virtual communities and examines the applicability of web analytics as a means of drawing information regarding the behaviour and usage of the community. The approach is evaluated in the context of a specific virtual community. The mobility theme is resumed in the paper from Hart and Hannan, who consider the more general future of mobile wireless computing, and some of the barriers that still need to be overcome. The final paper, from Bleimann, describes a context in which many of the aforementioned
issues and technologies could come together namely a new university approach combining e-learning and virtual classroom environments, alongside face-to-face and project-based learning. The discussion outlines the advantages and disadvantages of the different learning methodologies, as well as their combination in the proposed Atlantis approach. Following the success of the 2004 conference, the next INC event has been scheduled for July 2005. However, in a departure from the previous events, INC 2005 will be held on the Greek island of Samos. Readers interested in attending, or submitting a paper, are asked to visit www.inc2005.org for further details. Steven Furnell
Note
178
1 Full copies of the conference proceedings can be obtained from the Network Research Group at the University of Plymouth, UK (
[email protected]).
Introduction
An evaluative methodology for virtual communities using web analytics A.D. Phippen
The author A.D. Phippen is a Senior Lecturer, Network Research Group, University of Plymouth, Plymouth, UK.
Keywords Ethnography, User studies, Worldwide web
Abstract The evaluation of virtual community usage and user behaviour has its roots in social science approaches such as interview, document analysis and survey. Little evaluation is carried out using traffic or protocol analysis. Business approaches to evaluating customer/business web site usage are more advanced, in particular using advanced web analytics to develop greater understanding of their customers’ use of their sites. The application of such techniques to virtual communities is discussed and experimentation of such techniques on a specific virtual community project demonstrates the potential for such techniques in the evaluation of social and culture web usage.
Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/1065-0741.htm
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 179-184 q Emerald Group Publishing Limited · ISSN 1065-0741 DOI 10.1108/10650740410567518
The concept of a virtual, or online, community has been credited to Rheingold (1993), denoting a collective of geographically distributed individuals bound by a common interest exploiting internet technologies to enable communication. In the last ten years this term has been accepted to relate to any collective of individuals using internet technologies for a common purpose conforming to defined policies (Preece et al., 2003). The vast majority of literature related to the evaluation of behaviour and usage of virtual communities is drawn from the social sciences, in particular social anthropology and ethnography. The much-cited work of Paccagnella (1997) examined how ethnographic techniques may be applied to the evaluation of virtual communities. Since then, the terms “electronic ethnography” (Pink, 2000) and “virtual ethnography” (Hine, 2000) have been used to describe the ethnographic approaches to the study of the internet phenomena, and these concepts have been developed further by other authors (for example, Warren and Skerratt, 2003). When considering these ethnographic approaches to studying and conducting research on these communities, virtually all of the techniques used in the evaluation of community behaviour and usage can be placed into two distinct groups: (1) Primary evidence from participants – interview, surveys, etc. of user groups within the community to satisfy specific research goals. (2) Document analysis – analysis of e-mails, newsgroups, discussion forums, etc. to identify behaviour, attitudes, etc. in the community, again to satisfy specific research goals. The techniques are all well established in the social sciences and are certainly effective in achieving research aims. However, this paper suggests that the origins of such evaluation in the social sciences also fails to develop the virtual aspect of community evaluation – while the network technologies upon which these communities function are exploited to enhance established evaluative techniques (for example, the power of web-based vs postal surveys is well researched (for example, Hewson et al., 2003) drawing information from the network traffic and protocols has very little discussion in literature. In this paper we propose examining another area of web site assessment generally tied to the evaluation of successful business web sites, namely advanced web analytics, and determining its
179
An evaluative methodology for virtual communities using web analytics
Campus-Wide Information Systems
A.D. Phippen
Volume 21 · Number 5 · 2004 · 179-184
interactions between web site visitor actions and web site offers, as well as leverage insight to optimise the site for increased customer loyalty and sales.
suitability as a non-intrusive means to the evaluate the behaviour and usage of virtual communities.
Web metrics and analytics – measuring web site success Web analytics is an evaluative technique originating from and driven by the business world in its need to get more value out of understanding the usage of its web sites, and strategies therein. A large organisation may invest significant resources in developing what they would like to be a strong web strategy and, as with any resource investment, the organisation also needs to be able to measure the success of such strategies. Basic web analytics takes easily obtained statistics, or metrics, in order to be able to assess web site usage. This basic information can be drawn from web logs – the raw data held on a web server relating to core statistics about each HTTP request it serves. While this information can vary, depending on the logging format, it will generally contain information such as: . clienthost IP address; . username; . log time and date; . method passed; and . resource requested. While raw log data provide an intimidating and overwhelming volume of technical information, with some fairly straightforward processing (generally provided by some type of reporting software) simple web metrics can be easily drawn out. The most fundamental metrics are those such as hits and page views. Arguably, one can also identify session information – the pages a single user views in a single web site visit (whether this represents a true session is something that is debated in the literature – for example, Fletcher et al. (2002)). However, basic metrics, while serving some small use in giving raw statistics, have been dismissed in some literature as unreliable and unrepresentative (for example, Kilpatrick (2002), Buresh (2003), Schmitt et al. (1999), Whitecross (2002)). As a result of the dissatisfaction with basic web metrics and log file analysis, the concept of “advanced web analytics”, or e-metrics (Sterne and Cutler, 2000), was developed. Advanced web analytics aims to measure and understand the relationship between the customer and the web site through a richer analysis of web traffic and related data. Aberdeen Group (2000) defines advanced web analytics as: Monitoring and reporting of web site usage so that enterprises can better understand the complex
The important issue to note with advanced analytics is that it is not just concerned with web site statistics, but the relationship and interaction between a web site and its customers. It does not just collect website information, but also it uses it in conjunction with other data, such as demographics, customer profiles and subscription information. For example, while basic clickstreams will show how a specific user interacts with a web site, they offer little information on who that user is. However, if a clickstream is associated with a specific user, and if that use has profile information stored by the organisation, the value of the information begins to become far greater. In general, the use of advanced analytics has been centred on large business web strategies – organisations that have the resources to be able either to develop their own analytic strategies or to pay for consultants to develop them. A study on the use of advanced analytics in a multi-national airline company (Sheppard et al., 2004) examines a specific organisation’s use of advanced analytics in more detail, and shows the value the organisation places in their use.
Applying advanced analytics to social settings The use of web logs to draw information regarding usage and behaviour at present made a small contribution in virtual community literature. A few studies (for example, McLaughlin et al., 1999; Nonnecke and Preece, 2000; Smith, 1999; Wellman et al., 2002) examine network logs to understand basic statistics about number of logins, time and date, audience numbers, referral, etc. Additionally, Preece (2000) discusses the value of web logs for evaluating community behaviour and identifies the need to define metrics that could be applied to social and community settings. It is unquestionable that web logs do hold a huge amount of raw data (see subsection entitled “Initial results from the study” for example). However, the techniques that could be used to gain richness from the data to inform on how a community is maturing over time are lacking. In order to develop the evaluation of community usage beyond basic metrics we would propose that the ethos behind advanced analytics can be equally applied to social and cultural settings. If we take the basic premise from advanced analytics to be understanding the complex interactions that take place on a web site and use them to understand
180
An evaluative methodology for virtual communities using web analytics
Campus-Wide Information Systems
A.D. Phippen
Volume 21 · Number 5 · 2004 · 179-184
one’s customers better, it is entirely acceptable to apply the same to social and community web sites. Within a community or social setting, while finance may not be the driving force, there are factors that contribute success or failure in a site. Obvious general figures might relate to the number of community members, the number of sessions each member carries out on the site, the freshness of information, etc. However, just as a commercial web site may have specific goals, the same could also be said of a social web site. Therefore, it is not only important to identify core measures for community sites, but also to develop evaluative methodologies in order to be able best to determine success using the data available to the researcher.
either be browsed or searched based on a researcher’s needs. While the locations were all separate with different researchers assigned to specific locales, one of the aims of the project was to encourage the investigation of cross-regional interests and cultures. The main mechanism for achieving crossregional searching was to classify resources according the themes. Seven overall themes (the person, living together, work, play, worship, objects, spaces) were defined, each being further decomposed into sub themes (for example, objects’ subthemes include religious objects, toys, gifts and tourist objects). Any given resource could then be assigned one or more subthemes to provide a further means of classification. The searching facilities are such that a site member could search generally, tied to a location, tied to a theme, or in any combination, to encourage people to search beyond their specific locale to identify shared cultural perspectives beyond geographic boundaries. The multi-lingual approach descriptors related to each resource and also the site in general aim to widen participation within the site, by providing both a local feeling of ownership (by providing the resource in the local language) while at the same time removing the potential for isolation/lack of interest by also providing the same details in English. The project is an ideal candidate for experimentation as it has some very clearly defined goals for its use. Scott (2004) defined its intention for use to be as a: . research resource; . pedagogic tool; and . source of information to the interested general public.
Applying the concepts to a suitable setting While this paper has proposed the concept of an evaluative strategy to be applied to virtual communities, it is necessary to test this theory. In assessing these beliefs, the evaluation of a large research project with a specific web strategy was carried out (and is continuing to be carried out). The chosen test subject involves the Mediterranean Voices project – a European Union (EU) funded project that aims to: . . . create a database of the oral histories and cultural practices of the Mediterranean’s cosmopolitan urban neighbourhoods (Scott, 2004).
The project as a whole draws together resource from 12 distinct locations within the Mediterranean to fulfil a number of anthropological aims in determining what is meant by culture to the people who dwell within the locations. The majority of investment in the project went into the development of resources by researchers at each of the locations. The researchers spent a great deal of time interviewing, filming and photographing their specific locale (working with senior citizen’s clubs, youth groups, music and cultural associations, etc.) to capture the cultural essence of the location. This media was then transcribed and archived so that it could be classified according to: . location (defined below); . media type (film, audio, slide shows, photographs, text, flash images); and . themes (see below). Additional information relating to the resource, in both English and the local language was then added to provide a wealth of descriptive information. Finally, the media was provided to wider audience through the MedVoices web site (www.med-voices.org), where it was catalogued to provide a searchable database of media that could
Its also has very clearly defined intentions to be cross-locational and cross-cultural in its distributing of information and resources. It aims to promote a culture of sharing knowledge about the studied locations to develop greater understanding and empathy among people in the Mediterranean. Therefore, we can initially evaluate “success” in the MedVoices site through understanding the behaviour of its audience related to how it views resources and draws cultural information from the site. Experimental approach This experiment took place during the development phase of the MedVoices web site. It was not accessible to the general public at this time, but was accessible to the MedVoices researchers across Europe who had the facilities to both add their own resources to the site, and also browse and use the site as a general user. All the researchers were aware of the overall aims of the
181
An evaluative methodology for virtual communities using web analytics
Campus-Wide Information Systems
A.D. Phippen
Volume 21 · Number 5 · 2004 · 179-184
project and came from a social sciences background. Owing to these constraints, we only considered the first intention of the web site for evaluative purposes. The aim of the experiment was defined as being to determine the use of the MedVoices site in terms of: . location; . language; and . theme.
Table I Basic site statistics
n Hits Number of page views (*.htm* or *.asp*) Unique client IP addresses
From this aim, we hoped to gain a greater appreciation of how the site was currently being used which would indicate the relative success of the project aims to date. The information resources used within the experiment are detailed in Figure 1. The majority of the information is drawn directly from the MedVoices server. The “session db” was a small additional table added to the core resource information to hold basic session information (session identifier, client IP, time, date) to make it possible to marry session details with log information. The only other additional resource was the IP location database. In order to reduce network traffic, the GeoIP database and API (Maxmind Ltd, 2004) were used to ease the resolution of IP addresses to locations. Types of information drawn from these resources were guided by the experimental aims. The complete data set was loaded into a suitable relational structure, and analysed using SQL querying. Initial results from the study Basic statistics In order to establish core data, this paper details some basic measures in Table I. While they do not help much in achieving our experimental aims, they do provide a few interesting figures when
105,604 34,132 318
considering the volume of data available to the analyst. From 150 Mbytes of web log data, the basic statistics shown in Table I were drawn. While the number of unique client IP addresses obviously does not provide too much useful information, it is interesting to note that they were drawn from 87 unique class B addresses, showing a breadth of locations in a site that has not yet been launched. A couple of other basic statistics drawn from the logs, coupled with the session database, are shown in Table II. Resource usage In order to understand how users of the site viewed resources, a matrix of user location vs resource location is presented in Table III. Table III demonstrates some very interesting results. Home locations are highlighted on the table in italics. While in almost every case the home locations show the largest proportion of resources viewed, the data also show the level of breadth of location for resources viewed. In the majority of locations, there are resources viewed from most location. The main distinction in the data comes from resources viewed from the UK. As the researchers in the UK are co-ordinating partners for the project, there is no surprise that they have interests in resources across the entire site.
Figure 1 Information resources in experiment
Language usage The second measure was drawn from the viewing of resources in either the local language or English version of the resource. As there were a lot of resources only using English descriptors, in order to get a representative data set, only those resources that could be viewed in either language were sampled. Table IV details the resource language viewing based on location. When considering the multi-national aims of the MedVoices project, these results are encouraging – there is a great deal of use of local language resources in the majority of locations. Table II Session details
n Number of sessions No resources viewed
182
1,231 4,672
An evaluative methodology for virtual communities using web analytics
Campus-Wide Information Systems
A.D. Phippen
Volume 21 · Number 5 · 2004 · 179-184
Table III Percentage of resources viewed based on client location Resource location Valletta (Malta) Alexandria (Egypt) Mallorca (Spain) London (UK) Chania (Greece) Nicosia (Cyprus) Ancona (Italy) Marseilles (France) Granada (Spain) Bethlehem (Palestine) Istanbul
Cyprus
Spain
France
UK
Greece
Italy
Malta
Palestine
Turkey
17.28 2.47 9.88 12.35 9.88 30.86
1.4
0.78 0.23 1.01 1.4 3.26 0.62 0.39 84.26 2.87 3.8 1.4
4.77 2.27 0.68 9.09 18.64 0.91 1.35 23.64 22.27 9.55 8.18
0 0 0 0 59.46 0 1.35 31.08 8.11 0 0
0 0 0 0 0 0 0 40 0 0 60
100 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 4 0 96 0
1.71 0 0 6.84 4.27 0.85 0.21 25.64 4.91 1.71 53.85
2.47 14.81 0 0
3.15 3.5 1.4 0 5.59 71 4.2 9.79
resources ! Marseilles (objects) three resources ! Marseilles (spaces) three resources.
Table IV Percentage resources view in local language and English
Cyprus Spain France UK Greece Lebanon Malta Turkey Palestine
Local (%)
English (%)
46.51 39.44 42.69 23.63 26.67 15.63 33.33 13.51 0.00
53.49 60.56 57.31 76.37 73.33 84.38 66.67 86.49 100.00
From this small sample, we cannot draw any firm conclusions. However, we can see some encouraging trends, in particular in sessions 2 and 3, where resource viewings move across both location and theme throughout the clickstream.
Conclusions and further work
Theme usage In the final part of the experiment, we wished to evaluate the use of theme usage within the resource database to draw people away from their specific locale to investigate resources of similar themes in different locations. For this study, we randomly sampled a number of sessions from the database to determine whether resources viewed through themes would cause the user to deviate from their locale. In each session, we show the clickstream in terms of resource location, theme and number of resources viewed: (1) Session 1 (location: Spain): Bethlehem (the person) two resources ! Granada (worship) three resources ! Bethlehem (worship) one resource ! Bethlehem (spaces). (2) Session 2 (location: Spain): Marseilles (Spaces) two resources ! Bethlehem (the person) two resources ! Granada (the person) three resources ! Marseilles (the person) one resource ! Marseilles (work) four resources ! Granada (work) one resource. (3) Session 3 (location: UK): Granada (worship) one resource ! Bethlehem (the person) two resources ! Granada (the person) one resource ! Granada (play) three resources ! Chania (play) one resource ! Chania (spaces) six resources. (4) Session 4 (location: France): Marseilles (spaces) four resources ! Marseilles (worship) five
The aims of the study presented in this paper were to investigate the use of advanced web analytics to virtual community settings through the study of applying such approaches to a specific virtual community. While this is only a pilot study run in the development phase of the community, the results achieved to date are very encouraging. The implications for this study are twofold: (1) The results demonstrate the value of the analysis related to the aims of the MedVoices project, and as such represents the start of ongoing work evaluating its usage. It is envisaged as the site matures, and the analytical approach gets more detailed, the understanding of the usage of the web site should contribute greatly to the strategic growth of the site. The project is also developing more complex ways of relating resources together further to remove the rigid locational structure initially imposed by the web site structure, and these will also contribute to further analysis. (2) The results demonstrate the effectiveness of an analytical methodology in the evaluation of a socially focussed web site. While this is a single study, the potential for such techniques is encouraging. As well as further study on the MedVoices site, it is envisaged that this work will be applied to other virtual communities with which the author is involved, in order to
183
An evaluative methodology for virtual communities using web analytics
Campus-Wide Information Systems
A.D. Phippen
Volume 21 · Number 5 · 2004 · 179-184
move closer to a more general approach to applying analytic approaches to virtual community evaluation. In developing the techniques, it is proposed that further analytics techniques are applies to the evaluative framework – in particular we are interested in applying OLAP techniques (Thomsen, 2002), to the ever-increasing volumes of data available to the evaluator. While the methods used to date have proved solid, it would be interesting to use more expressive approaches to the analysis of the data. In developing these approaches, it is hoped that such evaluative techniques will contribute to the growing need to understand virtual community usage and development in the future.
References Aberdeen Group (2000), Web Analytics: Translating Clicks into Business, Aberdeen Group, Boston, MA. Buresh, S. (2003), “Your web traffic and your bottom line”, available at: www.marketingprofs.com/2/buresh.asp (accessed 18 March). Fletcher, P., Poon, A., Pearce, B. and Comber, P. (2002), Practical Web Traffic Analysis: Standards, Privacy, Techniques, Results, GlassHaus, Birmingham. Hewson, C., Yule, P., Laurent, D. and Vogle, C. (2003), Internet Research Methods, Sage Publications, Beverly Hills, CA. Hine, C. (2000), Virtual Ethnography, Sage Publications, Beverly Hills, CA. Kilpatrick, I. (2002), “Too many hits are bad for your web site!”, available at: www.contractoruk.co.uk/tech-hits.html (accessed 1 May). McLaughlin, M., Goldberg, S., Ellison, N. and Lucas, J. (1999), Doing Internet Research, Sage Publications, Beverly Hills, CA. Maxmind Ltd (2004), “The GeoIP Country database”, available at: www.maxmind.com/ Nonnecke, B. and Preece, J. (2000), “Lurker demographics: counting the silent”, Proceedings of CHI’2000, The Hague.
Paccagnella, L. (1997), “Getting the seat of your pants dirty: strategies for ethnographic research on virtual communities”, Journal of Computer Mediated Communication, Vol. 3 No. 1, June. Pink, S. (2000), “’Informants’ who come ‘home’”, in Amit, V. (Ed.), Constructing the Field: Ethnographic Fieldwork in the Contemporary World, Routledge, New York, NY. Preece, J. (2000), Online Communities: Designing Usability, Supporting Sociability, John Wiley & Sons, New York, NY. Preece, J., Maloney-Krichmar, D. and Abras, C. (2003), “History and emergence of online communities”, in Wellman, B. (Ed.), Encyclopaedia of Community, Berkshire Publishing Group and Sage, Great Barrington, MA. Rheingold, H. (1993), The Virtual Community: Homesteading on the Electronic Frontier, Perseus Publishing, Philadelphia, PA. Schmitt, E., Manning, H., Yolanda, P. and Tong, J. (1999), Measuring Web Success, Forrester Research, Cambridge, MA. Scott, J. (2004), “Time, place and cyberspace: locating the field in an EU project”, paper presented at ASA Conference 2004, Durham. Sheppard, L., Phippen, A. and Furnell, S. (2004), “A practical evaluation of web analytics”, Internet Research Journal, (forthcoming). Smith, M.A. (1999), “Invisible crowds in cyberspace: mapping the social structure of the Usenet”, in Smith, M.A. and Kollack, P. (Eds), Communities in Cyberspace, Routledge, New York, NY. Sterne, J. and Cutler, M. (2000), “E-metrics: business metrics for the new economy”, available at: www.emetrics.org/ articiles/whitepaper.html Thomsen, E. (2002), OLAP Solutions: Building Multidimensional Information Systems, 2nd ed., John Wiley & Sons, New York, NY. Warren, M. and Skerratt, S. (2003), “The ‘virtual village’ – rural community web sites and webmasters as agents in rural development”, in Banks, R. (Ed.), Survey and Statistical Computing IV: The Impact of Technology on the Survey Process, Association for Survey Computing, Chesham. Wellman, B., Boase, J. and Chen, W. (2002), “The networked nature of community: online and offline”, IT & Society, Vol. 1 No. 1, Summer. Whitecross (2002), “From web logs to web loyalty: managing your customers throughout the customer lifecycle”, available at: www.whitecross.com (accessed June).
184
Knowledge broker network based on communication between humans Robert Loew Udo Bleimann and Paul Walsh The authors Robert Loew is PhD Student, Cork Institute of Technology, Bishopstown, Ireland and Researcher, Institute of Applied Informatics Darmstadt (aida), University of Applied Sciences Darmstadt, Darmstadt, Germany. Udo Bleimann is the Managing Director, Institute of Applied Informatics Darmstadt (aida), University of Applied Sciences Darmstadt, Darmstadt, Germany. Paul Walsh is Senior Lecturer, Cork Institute of Technology, Bishopstown, Ireland.
Keywords Knowledge management, Information brokers, Man-machine systems
Abstract This paper proposes a new paradigm to overcome many problems in the area of information technology involving communication between humans. The paradigm will be exemplified using the example of knowledge transfer within a company. The solution is not interaction between a user and a highly intelligent system, but communication between people, supported by intelligent systems. The paper’s concept of the “knowledge broker” will clarify what the authors mean by manman communication. The basic idea of the knowledge broker is that of a hybrid man-machine system that enables knowledge transfer within companies not only theoretically, but also in practice.
Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/1065-0741.htm
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 185-190 q Emerald Group Publishing Limited · ISSN 1065-0741 DOI 10.1108/10650740410567527
Introduction Today, not only the internet but also information technology (IT) systems used in companies generate a considerable amount of information and information services. These have reached such unmanageable proportions than employees find it impossible to handle the volume. While the half-life of information is constantly decreasing, new knowledge is often considered to be out of date after a relative short time. Even search engines cannot always solve this dilemma, since they typically generate a huge amount of unsorted information, meaning that the collected data are of little value. The usual approach to finding solutions is a structured search for information in documents. In our approach we assume the solution of complex problems cannot normally be found in documents. Rather it is the result of communication and collaboration between different people. To address this information overload problem we propose a new paradigm: the man-man communication paradigm. Instead of seeking to solve problems by having users interact with an intelligent system, people (supported by intelligent systems) will communicate with one another. The new paradigm is illustrated by the example of a knowledge broker (KB) network – presented in the section entitled “Our concept: the KB network” – which we have applied to internal knowledge management. An employee addresses a KB with a description of his/her problem. This broker tries to identify the person who is best able to help solve the problem most effectively. The broker will then bring together the solution provider with the employee seeking help.
What is knowledge? To our mind, knowledge is information in the context of a user. The knowledge pyramid (see Figure 1) shows that human comprehension is based on the extraction of information from data and on the allocation of information to a certain context. However, this produces the following problem: it is not sufficient to make information available in a document in an IT system. Only edited information, which meets the needs of the employee and which is specifically contextualised, can be a valuable resource. This gives rise to another conundrum: “How should knowledge be stored in an IT system?”. The first step is to reduce the knowledge to information; second, this information must be assigned to a context. This can be achieved in semantic networks (see the subsection entitled “How can we improve the usability of information?”). However, the chosen context is usually that of the author and not the context of the person
185
Knowledge broker network based on communication
Campus-Wide Information Systems
Robert Loew, Udo Bleimann and Paul Walsh
Volume 21 · Number 5 · 2004 · 185-190
Figure 1 Knowledge pyramid
systems, and from a technical viewpoint it would be. However, we understand knowledge as information in the context of a human being. In simple terms “context” here means the relationships to further information with type descriptions of these relationships. Ontologies offer very flexible options for representing relationships between real objects.
interested in the knowledge. This is why in our concept the interested person and an employee, identified as an expert, are brought together. Subsequently, they correlate their contexts through discussion. The resulting common context enables both to understand the topic the other is talking about. We conclude that every human basically learns by communicating with other people. The common use of capability When an employee is in the process of solving a problem an additional synergy can be achieved by bringing an expert employee in on the problem. Either both can cooperate in a joint project or the expert can act as mentor for his/her colleague’s project. Our concept addresses the above-mentioned concerns, and shows how relations between employees can be handled (see the section entitled “Our concept: the KB network”).
Applying a knowledge management system (KMS) in a company If an organisation grows beyond a critical size, the quality of knowledge availability becomes a crucial factor for its further success. As indicated by Senge’s (2001) definition: “knowledge is the capacity for effective action”, the use and increase of knowledge is very important for an organisation. Different systems are used for this, e.g. intranet portals and KMSs. Components of a KMS In most cases, a KMS consists of existing information systems such as the following information sources and IT systems: internal company documents in digital form, personnel data, enterprise resource planning (ERP) systems, customer relationship management (CRM) systems, workflow management systems, groupware, intranet, search and retrieval systems. How can we improve the usability of information? It might now seem feasible to simply store all the knowledge of a company in one of the above
Semantic networks and ontologies Semantic networks are directed graphs, consisting of nodes, links and link labels. A node describes an object of the real world, while links represent relationships between two nodes. A link label describes the type of the relationship and the direction of that link (Winston, 1993). An ontology is a formal, explicit specification of a commonly used abstract model of the world (Gruber, 1995). It allows the representation of knowledge structures (Maier, 2001). We recommend “definitional networks” for the representation of internal company information. They are suitable for the construction of ontologies, especially for class and generalisation hierarchies. Semantic networks allow the graphical visualisation of information and navigation through this graph. Implementing a KMS We recommend companies employing a KMS develop a lexicon based on a semantic network. This lexicon should contain all the reality concepts relevant for the organisation. Nodes of the knowledge network could refer to documents, to smaller entities or to other resources, e.g. to IT systems the company uses. This enables users to search in the KMS. Search results are a list of hits (like those produced by internet search engines), but it is also possible to navigate within the semantic network at a graphic level. Knowledge management today – goal-driven or not? The biggest problem facing knowledge management is not technical, but human. Who is willing to enter their “knowledge” into the KMS? Will employees who are busy with their normal work always find the time to identify relevant information and update the KMS? The general question is: “Is it possible and realistic to enter all information relevant to business into a system?”. Existing literature cites three reasons why existing KMS are not often used by employees (Reimer, 2002), and we would like to add a fourth one: (1) Employees do not find the required information.
186
Knowledge broker network based on communication
Campus-Wide Information Systems
Robert Loew, Udo Bleimann and Paul Walsh
Volume 21 · Number 5 · 2004 · 185-190
(2) Employees assume that the required information does not exist in the KMS. (3) Employees have to work under pressure so they do not start a search at all. (4) Employees assume that the required information does not exist in the KMS, because they themselves would never enter this kind of information.
The role of the information professional Taking as your starting point the knowledge spiral described by Nonaka and Takeuchi (1995), you examine the role of information professionals, i.e. how they handle knowledge, and execute a learning process. Typically, you reach the conclusion that information professionals should support the knowledge processing stage (Goodall, 2003).
Related research
Our concept: the KB network
Semantic web
Every employee working on a problem can approach the KB, who will identify experts, i.e. other employees who can help solve the problem submitted by the person requesting information. KBs should be available/accessible from all locations involved in creative processes. The KB should have a certain degree of knowledge of the areas of expertise necessary to solve the problems arising in his location. Moreover, KBs should have connections to other KBs within the organisation. This network will be used to enable them to find the experts necessary to solve problems outside their own area of expertise. If the KB is unable to locate an expert among his/ her resources he/she will pass the information request on to the KB network and will request help from the other KBs within the organisation. The other KBs will try to identify the right expert to solve the problem. If another KB can establish such a contact a handshake-like procedure will be initiated. The employee who is an expert in the required area will be asked if he/she agrees to be contacted by the person requesting information. If the expert agrees to be contacted the contact will be established. Provided the contact is successful, the KB will be available to provide assistance.
The Semantic Web is the abstract representation of data on the World Wide Web, based on the Resource Description Framework (RDF) standards and other standards to be defined (W3C, 2001).
The vision behind the semantic web is to define and link data in ways that go beyond mere display in a web browser, such as using it in machines for automated processes, integration and re-use. The processes involved should exchange and process data, although the programs have been designed and implemented independently. This vision could only be realised via standardisation. The RDF standard (Lassila and Swick, 1999) is a first step in this direction. As more standards are necessary (and expected) it will take some time before the semantic web is “up-and-running”. Community relationships A whole array of projects is based on forming communities, in order to create and support networks between people. The idea that you know someone who knows someone else etc. is reproduced using technology. In the network you can find a suitable person to solve your problem. Once that person has been identified, you look or the shortest route to this person and contact the person in your own circle of friends who is closest to the potential problem solver. This contact is then pursued until you reach the final destination i.e. the person you wish to request help from. One of the first projects of this nature was evolved by Friendster (2003) in the form of an online community that connects people through networks of friends. A list of some practical applications of this idea in a professional environment such as Spoke, ZeroDegrees and VisiblePath is given in Loew (2003). Multimodal human-computer interaction Another method is to simplify the use of computer interaction. To this end, projects such as SmartKom (Wahlster, 2002) are developing intelligent user interfaces that accept and support the natural communication style of humans.
Proactive knowledge brokering One of the KB’s tasks is to respond to information requests. In order to be able to respond to requests from other KBs (e.g. from other locations or areas of expertise) it is crucial that information is shared among the KBs within the organisation. This information exchange should not only happen when a request is sent into the KB network but on a regular and frequent basis. This could be done by regular meetings with a focus on an area of expertise or based on a region. In worldwide organisations it might be useful to sometimes have a global meeting of the KB network. A company’s personnel data are often stored in so-called directories, for example in X.509 directories. To this end, many organisations use the “human resources” modules of an ERP system or workflow management system. As an important prerequisite for the work of the KB it is necessary
187
Knowledge broker network based on communication
Campus-Wide Information Systems
Robert Loew, Udo Bleimann and Paul Walsh
Volume 21 · Number 5 · 2004 · 185-190
to maintain the accuracy of employee profiles. To ensure this the KB should be involved in the maintenance process. An employee should have the opportunity to maintain his/her own profile by drawing on set vocabulary to describe skills, expertise and knowledge. It would be the KB’s job to motivate and assist the employee in maintaining this profile on a regular basis.
has already recorded relationships to experts in this field. The system could offer the information requestor profiles of experts in this field. If the information-seeker wants to contact an expert, the system could establish this contact. This function is a part of the KB network IT system. If the answer generated by the system is not considered adequate the person seeking information would define his/her request in greater detail. The KB would then establish a personal contact between the information-seeker and expert. As the costs for the work of a human KB are higher than the use of a KB support system this would help to lower costs, as routine requests would be handled by the system. The KB can define the rules, which are applied by the system for the automatic answering of requests. As the KB has to possess a great deal of knowledge about the areas of expertise needed in his area of responsibility, an experienced employee would be the preferred choice for such a position. Other necessary abilities for a KB would be management and leadership skills.
Relationship management system (RMS) The RMS allows access to available employee information – based on directories – to manage, visualise and maintain relationships between employees. A semantic network is recommended for this. The nodes of the network represent the employees. The links of the network are usually used to represent the communication relationships. An important issue is that a history of the relationships can be maintained, i.e. to see how relationships among participants of an information exchange have developed. This could be achieved by having the system store a time marker every time a relationship is updated or entered into the system. By doing so it will be possible to see how/whether a former relationship between two employees regarding a topic has changed (i.e. has been deleted) because one of the employees has moved on to another area. Every link between the information-seeker and the experts will be under the guidance of the KB supported by the RMS. The RMS will enable the KB to identify personal networks as shown in the section “Community relationships”. Today, organisations often use external experts such as management consulting firms. The KBs should take these outside experts into account. Implementation of the KB Cooperation between man and machine The KB could be designed by using artificial intelligence methods as a system. Experience from different projects indicates this is not feasible because of the complexity of such a system. The other extreme would be a manual system that allows the greatest flexibility but results in high costs. Because of the complexity of such a system’s requirements, e.g. searching for information, establishing relationships or contacts, the best approach for such a system is that it should be implemented as a cooperative system between machine and human. This could be achieved using a hybrid solution, i.e. a machine-supported human mind. This solution would offer the best ratio between cost and performance. Such a system could answer requests from an employee about an area of expertise if the system
KB and employee communities Establishing communities of KBs, employees and external experts seems to have positive effects on the organisation. These communities can extend beyond the borders of the normal organisation and occur across hierarchies within the organisation. For example, a RMS system process may utilise data mining, the employee profiles and the semantic network of the RMS, to form such communities of interest. If the system identifies such a group of persons, a KB will be informed and the community can be established. If the area of interest does not match the interest of the KB, he/she can pass this on to other KBs who may share the interests of the identified group of experts. Based on the result of the system the KB will identify whether the group really shares the same interest. If a group of persons has been identified, the KB will form the community and send out invitations to participate it. Connecting KMS islands using the KB network Larger organisations commonly use different KMSs depending on the area of expertise, the location or for historic reasons (e.g. pre-merger systems). It is usually rather difficult to establish connections between such systems. One approach could be the use of a meta search engine that queries all KM systems and aggregates the results before returning them. By using the KB network an employee from location A could access the knowledge and the KMS from location B (see Figure 2).
188
Knowledge broker network based on communication
Campus-Wide Information Systems
Robert Loew, Udo Bleimann and Paul Walsh
Volume 21 · Number 5 · 2004 · 185-190
(see Figure 3 bottom right). The said KB will then either proceed to gather the requested information or will contact an expert whom he/she asks to contact the information-seeker.
Figure 2 Connecting KMS islands using the KB network
Summary and further outlook
Prototype for the University of Applied Sciences Darmstadt The results of a survey revealed that only 70 per cent of those polled managed to find answers to complex questions using the university web site. Consequently, we decided to develop a new portal based on a semantic network. Now the user can choose between searching via the type of search engine used on the internet (see Figure 3 top left) or to navigate the knowledge network graphically (see Figure 3 top right). Once the user has made a certain number of navigation steps, he/she is offered the services of a KB (see Figure 3 bottom left). A KB with expertise in that given area is identified and if the user wishes to contact him/her the relevant information is displayed on the screen
This paper presents a new paradigm in which manmachine communication and cooperation will deliver the desired result. This will be supported by the above-mentioned system. This hybrid system uses a systemic approach which delivers results and at the same time supports or enables the KB to answer and track more complex information requests. A first prototype was developed for the portal for the University of Applied Sciences Darmstadt web site and we are currently identifying KBs for some areas of interest. Currently our research project is a joint project run by the aida team in Darmstadt, the Cork Institute of Technology, Ireland and different industry partners, and plans envisage implementing such a KB network in one of the participating companies. We would like to focus more strongly on cooperation between man and machine. One aspect we are particularly interested in is how an information-seeker’s request can be automatically
Figure 3 Prototype for the University of Applied Sciences Darmstadt
189
Knowledge broker network based on communication
Campus-Wide Information Systems
Robert Loew, Udo Bleimann and Paul Walsh
Volume 21 · Number 5 · 2004 · 185-190
answered. One option might be to devise a FAQ system of specific questions, whose answers could be formed dynamically using the semantic network.
Maier, R. (2001), Knowledge Management Systems, SpringerVerlag, Berlin. Nonaka, I. and Takeuchi, H. (1995), The Knowledge-creating Company: How Japanese Companies Create the Dynamics of Innovation, Oxford University Press, New York, NY. Reimer, U. (2002), “Einbettung des Wissensbereitstellung in den Arbeitskontext”, Informatik Spektrum, No. 3, June, pp. 196-9. Senge, R. (2001), “Reflection on a leader’s new world: building learning organizations”, in Morey, D., Maybury, M. and Turaisingham, B. (Eds), Knowledge Management – Classic and Contemporary Works, MIT Press, Cambridge, MA, pp. 54-60. W3C (2001), “Semantic web”, available at: www.w3.org/2001/ sw Wahlster, W. (2002), “SmartKom: fusion and fission of speech, gestures, and facial expressions”, Proceedings of the 1st International Workshop on Man-Machine Symbiotic Systems, Kyoto. Winston, P.H. (1993), Artificial Intelligence, Addison-Wesley, Reading, MA.
References Friendster (2003), “Online community web site”, available at: www.friendster.com Goodall, G. (2003), “The profession of knowledge conversion: continued relevance for information professionals”, available at: www.deregulo.com/facetation/pdfs/ continuedRelevanceInfoProf.pdf Gruber, T. (1995), “Towards principles for the design of ontologies used for knowledge sharing”, International Journal of Human-Computer Studies, Vol. 43 No. 5/6, pp. 907-28. Lassila, O. and Swick, R.R. (1999), “Resource description framework (RDF) model and syntax specification”, 3C Recommendation, 22 February, available at: www.w3.org/ TR/1999/REC-rdf-syntax-19990222Vol.22 Loew, R. (2003), “Projects of community relationships”, available at: www.aida.fh-darmstadt.de/projects/ infobroker/knowledge_broker_network/ community_relationships
Further reading Davenport, T.H. and Prusak, L. (2000), Working Knowledge, Harvard Business School Press, Boston, MA.
190
1. Introduction
Atlantis University: a new pedagogical approach beyond e-learning
Education is increasingly decisive for our societies. Consequently, many people are looking for new ideas in the area of learning and teaching. Atlantis University is one such ambitious project; it combines three different types of learning and teaching to form a single package for offer to all customers (Figure 1), be they are full-time students or people in the workplace. The three pillars of this university form a new paradigm for higher education development, delivery, support and assessment using Advanced Technology for Learning in a Net based Information Society (Atlantis). Atlantis University is an international partnership involving nine universities and companies from the European Union (EU) and USA, and is committed to developing an innovative high-quality institution with common branding. The idea evolved from long-term experience in face-to-face (f2f) and project-based teaching as well as several e-learning (e-L) research projects (Furnell et al., 1998, 1999; Stengel et al., 2003). First practical solutions have already been developed within the last two years (see section 5).
Udo Bleimann
The author Udo Bleimann is Managing Director, Institute of Applied Informatics Darmstadt (aida), University of Applied Sciences Darmstadt, Darmstadt, Germany.
Keywords Learning methods, Virtual organizations, Universities
Abstract Atlantis University is an ambitious international project in the area of learning and is currently being developed by a group of universities and companies. It combines three different types of learning and teaching to form a single package offered to Students and people in the workplace alike: face-to-face learning, e-learning and project-based learning. The paper gives an overview of the advantages and disadvantages of the different learning methodologies, and describes the new Atlantis approach. The first practical solutions Atlantis University has developed, namely the Virtual Classroom, ELAT learning environment and Project Service Center, are likewise briefly introduced.
Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/1065-0741.htm
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 191-195 q Emerald Group Publishing Limited · ISSN 1065-0741 DOI 10.1108/10650740410567536
2. Learning process hierarchy The well-known knowledge pyramid can be used as a basis for clarifying the learning process hierarchy (Barnett, 1994). This hierarchy begins with data gathering/processing (Figure 2). Data in a context with a meaning are called information. Knowledge is defined as information in the context of humans. Applying knowledge to solve problems leads to capability. Capability goes beyond both knowledge and skills/competence in that it represents “an integration of knowledge, skills, personal qualities and understanding used appropriately and effectively” (Stephenson and Yorke, 1998). By contrast, competence has to do primarily with the ability to perform effectively in the here and now, in a known or familiar setting; capability goes beyond this, having to do with the realization of capacity in an unknown or unfamiliar future context as well as good performance in the known present. Therefore we have placed the term capability rather than competence at the apex of the pyramid of learning. The learning process in higher education usually starts from the bottom up and moves very slowly to the top of the pyramid. Against this background, many scientists and teachers are looking for a more efficient path to capability.
191
Atlantis University: a new pedagogical approach beyond e-learning
Campus-Wide Information Systems
Udo Bleimann
Volume 21 · Number 5 · 2004 · 191-195
Figure 1 Atlantis University portal: one face to the customer
Figure 2 Learning process hierarchy
3. Existing learning methodologies In this section we describe the three main learning methodologies with their respective advantages and disadvantages, and subsequently classify them in terms of the learning process hierarchy. 3.1 f2f learning f2f or classroom learning is the classical learning approach, which has always been popular and will continue to be so. The reason for this is that humans learn from humans, that we are modeled on human archetypes. Table I shows principal advantages and disadvantages. All professionals in higher education have a great deal of experience with groups in classrooms and know it takes a long time to get students to the top of the learning hierarchy. 3.2 e-L The hype about e-L started in the middle of the 1990s. Why was that? The technology (especially the Web) was available, and many people thought self-directed individual learning at home was the future – independent of everything and everybody. The whole e-L community had to learn a lot: For
instance, it is very hard to motivate an isolated student watching a tutorial video for a longer time. The production costs of e-L content are very high – a good estimate is 100 times the cost of content for a f2f lecture. Blended learning emerged as an approach that overcomes some of the disadvantages (see Table II) by combining e-L and f2f, and is now state-ofthe-art for a virtual university. The latest innovation in this area is mobile learning (Nyiri, 2002; Seppa¨la¨ and Alama¨ki, 2002), which might be taken to mean a version of e-L using a mobile device. But this underestimates the potential of mobile learning, which could take place at the relevant location, e.g. learning how to build bridges in front of a bridge or information on buildings while walking through a city. Several research projects are addressing this point, especially for museum applications (Igd, 2003). Regarding the learning process hierarchy, like classroom learning, e-L starts from the bottom and does not address capability.
3.3 Project-based learning (proL) Solving real-world problems involves complex tasks – most of which are executed as projects. As such, in many cases, practical learning means proL (see Table III). It is obvious from many research studies and our own experience that proL strongly addresses both knowledge and capability (Bruffee, 1999; Boud and Solomon, 2001; Batatia et al., 2002). The group members not only learn the relevant subjects/contents, but also skills for teamwork, communication and project management. This makes proL the ideal learning methodology. Table II e-L
Advantages
Disadvantages
Learning any time, any location Active learning at own speed Easier quality control Easy distribution
Resource-intensive (time, budget, tutoring) Content mistakes are more serious Technology problems Costs for students Not immediately applicable No immediate answers Isolation (no social contacts)
Table I f2f
Advantages
Disadvantages
Direct communication Feedback/questions possible Very flexible Not very dependent on technology
Uniform pace for all learners Variable teaching quality Same location, same time No repeat (no archives) Not immediately applicable No immediate answers Learners might be (and often are) passive
Table III ProL
192
Advantages
Disadvantages
Very intensive direct communication Very flexible Immediate response Feedback/questions possible Can be at the right location
Time-consuming Highly resource-intensive Unpredictable outcome How to choose group members Assessing group members
Atlantis University: a new pedagogical approach beyond e-learning
Campus-Wide Information Systems
Udo Bleimann
Volume 21 · Number 5 · 2004 · 191-195
4. The integrated pedagogic approach of Atlantis University
Figure 3 Interaction between the three pillars
Bearing in mind the various advantages and disadvantages of the different learning methodologies we propose an extended blended learning approach: Atlantis University. Atlantis University is characterized by: . The teaching methods are a carefully engineered mix of project-based learning, e-L and f2f learning – depending on the learner’s situation (life long learning). . International cooperative/collaborative and interdisciplinary learning and teaching leading to internationally accredited and recognized qualifications (global education for the global market). . Knowledge and technology transfer between participating universities and companies and the workplace – private public partnership (not only academics). . Projects incorporate real-world problems – resulting in qualifications (students) and solutions (workplace) – and eliminating the “practice shock”. The Atlantis approach addresses knowledge or even capability rather than simple gathering and evaluation of information, and is delivered at an early stage of the educational life-cycle. It results in student-centered, integrated and more efficient learning. Atlantis University seeks a standardized approach, curriculum and an interchangeable body of knowledge. Course contents and organizational issues are set up in a standardized way, among other things using ECTS, which makes the mutual recognition and portability of credits much easier. Implementing the principle “Write/develop once, use many times and locations” courses are developed in cooperation but made the responsibility of the strongest partners in each case. These courses could be used by all partners. Thanks to our innovative approach and the subject of the course itself, we address students of different cultures, languages and of different regions, seeking to bridge significant differences and thus achieve an enduring impact. Two example scenarios and Figure 3 show the possible combination and interaction between the three pillars of Atlantis University. In developing learner capability across a range of fields, Atlantis will look to find the optimal mix of: . f2f and e-L environments; . individual study and project-based group work; and . teacher- and student-directed learning.
A learning path through a typical Atlantis course could be represented in a simplified form as in Figure 4.
5. Towards a practical solution The theoretical approach used is complemented by three practical projects, with the aim of creating a combined, integrated infrastructure for the three pillars of our concept. The e-L project has been in place for over seven years, while the other projects started in 2003.
5.1 Virtual classroom as part of the learning infrastructure To support our international Atlantis partnership we installed a virtual classroom infrastructure. This enables “remote” students to participate in lectures held in a specially equipped classroom elsewhere. The remote students sit in a classroom that has the necessary technology. This is still f2f learning and teaching, even though the student only sees the teacher via video and vice versa. The technical infrastructure in the main classroom consists of an interactive whiteboard and two projectors – one for the video input from the remote classroom (Figure 5, see monitor on the left) and one focused on that board with the option to annotate any computer presentations Figure 4 Typical Atlantis learning path
193
Atlantis University: a new pedagogical approach beyond e-learning
Campus-Wide Information Systems
Udo Bleimann
Volume 21 · Number 5 · 2004 · 191-195
Figure 5 One lecture in two remote classrooms
using a special pen (see teacher in front of the interactive whiteboard). Two projectors (or big monitors) also have to be installed in the remote classroom – one for the video input (camera focused on teacher), and one for the input from the whiteboard. 5.2 The e-L Project ELAT The Environment for Learning and Teaching (ELAT) has been developed in the project “Modules for multimedia net-based higher education (2MN)” with several partners and coordination in Darmstadt. ELAT is a learning platform with the following functions: composition of course content, distribution of teaching materials, web-based communication, cooperative net-based learning, support in completing tutorial exercises, management of e-L courses and administration. ELAT enables self-managed learning using multimedia teaching materials, like audio-, videoclips and different XML-documents, which comply with the Learning Object Metadata Standard (LOM). The main principle of ELAT is to keep the handling as simple as possible (Figure 6). For example the authoring tool supports an easy way to produce courses by adding different objects, called knowledge units, to timelines, as shown in Figure 7. At the moment ELAT integrates the following software packages: . The “Digital Communications Simulation System (DCSS)” which supports simulations in telecommunication, including fibre-optic transmission systems and components. . NetSim, a virtual lab environment that simulates TCP/IP networks. . Hydrotrainer and TWL-Trainer, training software for civil engineering and architecture.
Figure 6 ELAT student user interface
Figure 7 ELAT authoring tool
Under development is a user modelling module for tailoring a course program which meets the users’ needs in an optimal way and a QTI compliant module for online tests. The 2MN project developed lots of e-L courses for teaching in
194
Atlantis University: a new pedagogical approach beyond e-learning
Campus-Wide Information Systems
Udo Bleimann
Volume 21 · Number 5 · 2004 · 191-195
engineering and IT. The first lectures were given with success in 2002/2003 at universities in Darmstadt, Cologne and Munich. 5.3 The Project Service Center supporting ProL For projects in industry you usually need ample resources and a sophisticated infrastructure. To support student projects and their special needs we have developed a Project Service Center (PSC) as part of the Atlantis University concept. The PSC offers students communications and technological support for the duration of an entire project. Since students only participate in projects for a relatively short period of time (typically five months), the system must be quick and simple if it is to gain acceptance. Consequently, it is essential that the PSC has a high level of usability and permits decentralized administration of the student projects. The PSC provides support for project management, collaboration management as well as document and content management. Since project management assists students in defining tasks and milestones it also generate project plans. The other two areas address communication between project members and document management. Consequently, the PSC covers the three most crucial areas involved in project execution. The PSC is integrating all three areas into a single, user-friendly platform by April 2004. It is possible to access all PSC functions via the web, in order to ensure maximum accessibility, but also to promote virtual teams. All that users require is a standard browser.
6. Conclusion and outlook Atlantis University seeks to address many disadvantages of well-known pedagogical concepts for higher education. The proposed extended blended learning with its three pillars f2f, e-L and proL is an integrated student-centered approach with great potential, which differs significantly from traditional approaches.
The project is ongoing with further technical and organizational developments on a fairly large scale. To demonstrate the basic idea we will examine in detail a project management module used for different courses within a wide variety of subjects in all partner locations. This module will be delivered to different user groups the “Atlantis way” and traditionally, allowing us to conduct a comparison of learning results.
References Barnett, R. (1994), The Limits of Competence: Knowledge, Higher Education and Society, Open University Press, Maidenhead. Batatia, H., Ayache, A. and Markkanen, H. (2002), “Netpro: an innovative approach to network project based learning”, Proceedings of the International Conference on Computers in Education (ICCE’02). Boud, D. and Solomon, N. (Eds) (2001), Work-based Learning: A New Higher Education?, Open University Press, Maidenhead. Bruffee, K.A. (1999), Collaborative Learning: Higher Education, Interdependence, and the Authority of Knowledge, Johns Hopkins University Press, Baltimore, MD. Furnell, S., Bleimann, U., Girsang, J., Roeder, H., Sanders, P. and Stengel, I. (1999), “Security considerations in online distance learning”, paper presented at the 4th Euromedia Conference, Munich. Furnell, S., Onions, P., Bleimann, U., Gojny, U., Knahl, M., Roeder, H. and Sanders, P. (1998), “A security framework for online distance learning and training”, Internet Research, Vol. 8 No. 3, pp. 236-42. Igd (2003), “Project Geist”, available at: www.igd.fhg.de/ projects/index.html (accessed December). Nyiri, K. (2002), “Towards a philosophy of m-learning”, Proceedings of the IEEE International Workshop on Wireless and Mobile Technologies in Education (WMTE’02). Seppa¨la¨, P. and Alama¨ki, H. (2002), “Mobile learning and mobility in teacher training”, Proceedings of the IEEE International Workshop on Wireless and Mobile Technologies in Education (WMTE’02). Stengel, I., Bleimann, U. and Stynes, J. (2003), “Social insects and mobile agents in a virtual university”, Campus-Wide Information Systems, Vol. 20 No. 3, pp. 84-9. Stephenson, J. and Yorke, M. (1998), Capability and Quality in Higher Education, Kogan Page, London.
195
Introduction
Text appeal: the psychology of SMS texting and its implications for the design of mobile phone interfaces Fraser J.M. Reid and Donna J. Reid
The authors Fraser J.M. Reid is Associate Head and Donna J. Reid is a PhD Student, both at the Centre for Thinking and Language, School of Psychology, University of Plymouth, Plymouth, UK.
Keywords Mobile communication systems, Psychology, Gender
Abstract Argues that understanding the psychological drivers behind SMS uptake among key user groups could open the door to a range of user-centred applications capable of transforming handset usability – and hence operator revenues – for this inexpensive form of text messaging. Combines the findings of our own webbased survey of SMS users with psychological evidence and research on related text-based conversational systems to draw out lessons for a user-based approach to the design of mobile phone handset displays that capitalise on the social affordances of SMS texting.
Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/1065-0741.htm
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 196-200 q Emerald Group Publishing Limited · ISSN 1065-0741 DOI 10.1108/10650740410567545
This paper addresses important mobile phone interface design issues that arise out of the phenomenal growth in the volume SMS text messaging over the last five years. According to the Mobile Data Association (MDA (www.mdamobiledata.org)), the total number of chargeable person-to-person text messages sent across the four UK GSM networks reached 1.73 billion during September 2003, marking the high point of a steady and continuous rise since the MDA began collating data on behalf of the UK network operators in 1998. Although industry analyses forecast stagnating SMS traffic volumes as new enhanced messaging alternatives become increasingly available, consumer behaviour is not expected to change rapidly, and SMS is expected to continue to dominate mobile messaging by both traffic and revenue well into the last quarter of the decade (www.forester.com). Why is SMS so resilient? Evidence now beginning to emerge – both from our own research (Reid and Reid, 2003) and from commercial surveys (e.g. www.cpp.co.uk) – is that for some users and some contacts, SMS may continue to be preferred to multimedia and instant messaging, and even to voice calls. It would therefore be unwise to overlook the simple user benefits and unique affordances of SMS. In this paper, we argue that understanding the psychological drivers behind SMS uptake among key user groups could open the door to a range of user-centred applications capable of transforming handset usability – and hence operator revenues – for text messaging. To do this, we combine the findings of our own web-based survey of SMS users with other psychological evidence and studies of textbased conversational systems to draw out lessons for a user-based approach to the design of mobile phone handset displays.
Up close and hyperpersonal Our interest in this area has been triggered by obvious similarities in the social affordances of the new mobile communications technologies with those of longer-established desktop varieties, such as e-mail, online chat, and instant messaging. In particular, we now know from psychological studies that asynchronous text-based internet services are valued because they allow users time to select, craft, and edit the personality they present online (O’Sullivan, 2000). As a result, the impressions users create and receive of each other are deeper, richer, and more intense than those available through direct personal contact (Hancock and
196
Text appeal: the psychology of SMS texting
Campus-Wide Information Systems
Fraser J.M. Reid and Donna J. Reid
Volume 21 · Number 5 · 2004 · 196-200
Dunham, 2001). For some, this provides greater freedom of self-expression, and leads to online relationships that are stronger and more enduring than those in the real world (Parks and Floyd, 1996). For others – particularly the lonely, insecure and socially anxious – the internet can nurture budding relationships that would otherwise be too vulnerable to thrive in face-to-face contact (McKenna et al., 2002). Thus the appeal of the Internet is not just that it opens up personal communication, but that it creates the conditions for hyperpersonal communication (Walther, 1996), a form of contact that surpasses real world relationships in both intimacy and intensity. Our own findings (Reid and Reid, 2003) show this to be equally true of SMS. We found that the texters in our survey – heavy text users who preferred texting to talking on their mobiles – reported that SMS helped them develop new and deeper relationships with “text mates”, and altered the way they expressed themselves – not only were they more comfortable saying things in texts that they would hesitate to say face-to-face, but they also felt more able to express the “real me”. Texters also committed more time and effort to the process of texting – they were more likely than those who preferred making voice calls to edit and rewrite their text messages, and their messages were more likely to make use of the full character limit available on their mobile phones. Even more revealing is the fact that most texters in our survey preferred to text their closest friends than speak to them on their mobiles, and a significant proportion even preferred to text them rather than meet up face-to-face. Nevertheless, certain features of mobile telephony set SMS apart from conventional internet use, as well as resembling it from the user’s point of view. In particular, the immediacy, mobility, and perpetual contact afforded by the mobile phone allows near-conversational levels of texting, so that an exchange of text messages can resemble online chat in turn taking and discourse structure (Kasesniemi and Rautiainen, 2002). We suspect that it is the combination of these two features – the spontaneous sociability of the chat room coupled with the artful editability of e-mail – that lends texting a special, but paradoxical, appeal to key user groups. Put simply, text messaging seems to provide an opportunity for intimate personal contact while at the same time offering the detachment necessary to manage selfpresentation and involvement. What are the implications of these findings for the future of text messaging? Let’s examine three highly publicised alternatives to SMS: (1) MMS. By the end of 2003, handsets capable of sending multimedia messages are likely to account for only 10 per cent of the western
European market, while mobile phones with built-in cameras will have a penetration rate of just 4 per cent (www.textually.org). Even in Japan and Asia, where explosive growth in multimedia messaging has been forecasted, SMS is expected to continue to generate the highest volume of messages of any format (www.instat.com). Clearly, sending photos and downloading video clips through mobile phones has failed to emulate the success of SMS. If only a fraction of the appeal of SMS is that users can manufacture the personal image they present of themselves to their contacts, then it is easy to understand why. (2) P2T. Despite fears that they might prove too intrusive, new GPRS push-to-talk services that turn mobile phones into walkie-talkies have also been tipped soon to make text messaging obsolete (www.textually.org). The push-to-talk feature can be used quickly to connect the user to a single person, or a group of people, by using a key mounted on the side of the handset. However, for the texters in our survey, P2T is likely to be regarded as simply another type of real-time voice call, offering limited scope for selective self-presentation and no substitute for the hyperpersonal affordances of SMS. (3) IMPS. Instant messaging and presence services (e.g. Nokia’s Presence Server, www.nokia.com; OZ IMPS Server, www.oz.com; Ericsson’s EIMPS, www.ericsson.com) promise to offer virtually synchronous text messaging. An IMPS user will be able to update his or her current status on his or her mobile phone, allowing authorised contacts to check whether they are available, busy, or free for a private or group chat. IMPS is therefore expected to appeal to users already sold on the merits of instant messaging (www.textually.org). But will it appeal to SMS texters? Again, the question is whether IMPS can support hyperpersonal communication. SMS provides asynchronous but “always there” contact, allowing users to thread a conversation together with sufficient time to exchange carefully crafted messages without the expectation of an immediate reply. In contrast, IMPS offers real-time involvement in which users are able – indeed expected – to enter into an attentional contract with each other to engage more or less continuously in chat (Nardi et al., 2000). These are precisely the real-time conditions that lead texters not to prefer face-to-face and voice contact with their communication partners. Ironically, we forecast that when presented with the greater functionality of IMPS, user groups such as the heavy texters in our survey will prefer to
197
Text appeal: the psychology of SMS texting
Campus-Wide Information Systems
Fraser J.M. Reid and Donna J. Reid
Volume 21 · Number 5 · 2004 · 196-200
manage their presence information so as to downshift handset functionality and mimic lower-grade SMS availability.
“Texting – that’s more of a girl thing” This comment, made by a male respondent in a survey carried out by Ling (2003) in Norway, hints at an emerging fact – that it is women, particularly teenage and young adult women, that have taken the lead in SMS use. In Ling’s survey, young females sent an average of seven texts per day, 40 per cent more than males in the same age group. Even more striking are the results of a survey by the MDA in May 2003, which reported that 74 per cent of women had texted in the two minutes prior to being surveyed, compared with just 26 per cent of men (www.mda-mobiledata.org). In our own survey (Reid and Reid, 2003), we found that heavy texters – specifically those that preferred texting to talking on their mobiles – were far more likely to be younger and female. In fact, only 39.5 per cent of the 396 male respondents in our survey reported a preference for texting, compared with 49.1 per cent of the 676 female respondents. Furthermore, 51.3 per cent of females classified as texters in this sample were 21 years old or younger, compared to 43.7 per cent of males in this category. What is it that makes texting “more of a girl thing”? The first indication is that texting is predominantly concerned with friendship work. In Thurlow’s (2003) study of undergraduate text messages, only about one-third of messages accomplished functional or practical goals – the remainder fulfilled a combination of friendship maintenance, romantic, and social functions associated with highly intimate and relational concerns. Our own survey also showed that heavy texters establish small networks of text mates, with whom they exchange messages more or less continuously, engaging in extended text conversations consisting of multiple partners and multiple turns, even preferring this kind of contact over voice calls with their partners. As Thurlow (2003, p. 12) puts it, texting creates “a steady flow of banter . . . used . . . to maintain an atmosphere of intimacy and perpetual social contact. In this sense, text messaging is small talk par excellence – none of which is to say that it is either peripheral or unimportant”. SMS is therefore an important vehicle for establishing a sense of social connection to others, creating awareness moments in which people feel connected to each other, with or without the need to convey specific items of information (Nardi et al., 2000). The fact that young women value this more than men is consistent with what is known about the kinds of relationships girls develop with same-sex friends from school age onwards –
generally girls lean towards more intimate, personto-person bonds than boys (Markovits et al., 2001). Paradoxically, girls’ same-sex friendships are also more unstable, shorter-lived, and generally more fragile than those of boys (Benenson and Christakos, 2003), and this creates intense pressures among girls to monitor and maintain their close friendships through regular personal contact. Among Thurlow’s corpus of messages, texts that simply expressed positive feelings, establish a mood of sociability, or that offer apologies, support, thanks – all appeared to fulfil these important friendship maintenance functions. How might the mobile phone support this essential activity? At first glance, IMPS services offer precisely this kind information. With Nokia’s Presence Server, for example, users can publish and update their presence status as well as control who have access to this information, while contacts can subscribe to this information and have it dynamically updated on their mobile handsets. However, availability is not the same as sociability. To communicate sociability, the server needs first to convey a recognisably positive sentiment, and then must address this sentiment to one or more specific contacts. Kuwabara et al. (2002) describe two simple, intuitive and non-intrusive desktop visualisations that suggest potential handset analogues: (1) Gleams of People. This is an application that allows a user to click on a screen icon and transmit a “mood signal” to a specified receiver’s screen, where a gleaming coloured sphere provides a visual representation of the user’s presence and status information. (2) FaintPop. This application extends the logic of Gleams of People, allowing a group of contacts to maintain a sense of connectedness among its members. Implemented as a touch screen “group photograph” analogue, members communicate with each other simply by touching screen images of their friends. This is encoded and broadcast to all members of the group, where it is represented as an onscreen animation effect. These two applications capitalise on what Kuwabara refers to as “connectedness-oriented communication”, or mood signals targeted at maintaining and enhancing human social relationships. We now know that this kind of communication accounts for about two-thirds of SMS traffic.
Text mates and text circles Perhaps the most striking finding to emerge from our survey was the existence of text circles –
198
Text appeal: the psychology of SMS texting
Campus-Wide Information Systems
Fraser J.M. Reid and Donna J. Reid
Volume 21 · Number 5 · 2004 · 196-200
well-defined and close-knit groups of contacts with whom texters regularly, sometimes continuously, exchange messages. Text circles did not appear to be particularly extensive – texters messaged about the same number of people on a regular basis as did “talkers” – but they had fewer contacts in their phonebooks, and proportionally more of these were mobile rather than landline numbers. They also engaged more frequently in extended text conversations, sending nearly twice as many messages in these conversations than talkers. Furthermore, texters were more likely to text a specific, well-defined group rather than many more diffuse groups of contacts, and more frequently participated in several simultaneous text conversations, findings which taken together reinforce the idea that texters share interconnections within a close group of friends in perpetual SMS contact with one another. Here again, strong gender differences are likely to lie behind these circles of contacts. We know from research with children and young adults that boy’s same-sex friendships are more interconnected and group-oriented than those of girls, who tend to form pairwise attachments with individual friends (Markovits et al., 2001). Boys also tend also to focus their communications on more tribal, group-based activities, while girl’s communications are more relationally-focused (Seeley et al., 2003). The fact that boys’ attachments are more group- and activitybased does not mean they are less close or intimate – instead it implies that while girls’ social networks are networks of friendship pairs, boys’ networks form coherent units with a stronger sense of group identity and fairly clear group boundaries and memberships. These differences have profound implications for supporting SMS text circles. For boys, a presence service offering group-wide SMS would be capable of maintaining the clubby atmosphere of their “all for one, and one for all” network of attachments – a single, jocular text addressed to the whole group will be sufficient to maintain it as a coherent social unit. However, this will not do for girls’ social networks. Girls typically build up highmaintenance networks of relatively independent one-to-one attachments, each of which needs to be separately monitored and cultivated through “friendship work” and regular social contact (Benenson and Christakos, 2003). Put simply, girls need to know more about the network context of their attachments than do boys, and this network-wide information is unlikely to be supported by IMPS services in their present usercentred configuration. Instead, what female users require is the capability to visualise the network from a server-centred rather than a user-centred perspective – put simply, they need to see an aerial view of all messaging activity across the network.
This visualisation goal happens to be at the heart of the online visualisation projects of the Sociable Media Group based at MIT and headed by Judith Donath (http://smg.media.mit.edu/). These projects include several desktop client-server visualisations that have the potential to translate into handset analogues: . The Loom. This is a usenet visualisation tool, creating a visual representation of participants and conversations in a threaded newsgroup (Donath, 2002). One form this visualisation takes is to represent unrelated posts as a scattered field of clickable icons – threaded posts are depicted as dense clusters of active conversational topics. . Coterie. This is a similar visualisation of the structure of chat activity. Coterie’s display presents an environment that looks like a conversation – coherent discussions form a solid, central core, while scattered chat contributions are dispersed randomly over the screen (Donath, 2002). . Chatscape. The Chatscape interface and clientserver architecture (Lee, 2001) supports a virtual chat environment populated by avatars, each representing a different user, able to post onscreen messages. Clusters of avatars represent local conversations. Avatars can be moved physically around the environment so as to join or leave conversational clusters, as well as follow or avoid other avatars. Avatars can also display selected identities as well as be assigned “reputation” attributes by other users. Chatscape is the most recent in a series of chat systems based on Chat Circles (Donath et al., 1999), and comes closest to mimicking the physical and social characteristics of a real informal gathering, depicting visually both the dynamics of chat groupings as well as the emergence and propagation of socially shared topics and sentiments. In short, Chatscape – together with related visualisations projects from the Sociable Media Group – point to important visual display functionalities that could enable SMS users to monitor and participate more readily in the social networks of which they are a part. Simplified analogues of these functionalities could readily be incorporated into mobile phone handset displays.
Summing up SMS is a resilient messaging medium, which, despite the emergence of new enhanced messaging services, is likely to remain popular with mobile phone owners for the foreseeable
199
Text appeal: the psychology of SMS texting
Campus-Wide Information Systems
Fraser J.M. Reid and Donna J. Reid
Volume 21 · Number 5 · 2004 · 196-200
future. We have reviewed psychological evidence for this popularity, identified the key social affordances that SMS provides, and examined a range of desktop applications that show potential for supporting SMS functionality. We believe that it would be foolhardy for network operators to ignore the lessons that have been learned about the psychology of SMS use, its simple user benefits, and unique social affordances. Our research on the psychological appeal of SMS is only just beginning – however, we can already draw out three key recommendations at this stage: (1) New enhanced messaging services must continue to support the hyperpersonal affordances of asynchronous, editable message composition and transmission. Network operators will need to ensure that where enhanced presence, instant messaging and voice services are provided, users can downshift to lower-grade SMS functionality. (2) The greater proportion of SMS traffic – particularly among younger female users – fulfils friendship maintenance, mood signalling and sociability functions, often with minimal informational content. Network operators may wish to consider enhancing instant messaging and presence services to support simple, one-click personto-person or person-to-group sociability messaging. (3) Future development of client-server IM and presence services must recognise the psychological characteristics of the text circles with which key user groups regularly, sometimes continuously, exchange messages. The evidence suggests that male and female texters require different kinds of network information for optimal participation. We forecast that while males will be satisfied with group-wide SMS messaging, female users will prefer server-centred network information, allowing them an aerial view of SMS activity across their social networks. Important lessons can be learned from chatbased client-server applications aimed at the desktop environment – especially those developed by the Sociable Media Group at MIT – for the design of future handset displays.
Donath, J. (2002), “A semantic approach to visualizing online conversations”, Communications of the ACM, Vol. 45 No. 4, pp. 45-9. Donath, J., Karahalios, K. and Viegas, F. (1999), “Visualizing conversation”, Journal of Computer-mediated Communication, Vol. 4 No. 4, available at: www.ascusc.org/jcmc/vol4/issue4/donath.html Hancock, J.T. and Dunham, P.J. (2001), “Impression formation in computer-mediated communication revisited”, Communication Research, Vol. 28 No. 3, pp. 325-47. Kasesniemi, E.L. and Rautiainen, P. (2002), “Mobile culture of children and teenagers in Finland”, in Katz, J.E. and Aakhus, M. (Eds), Perpetual Contact: Mobile Communication, Private Talk, and Public Performance, Cambridge University Press, Cambridge, pp. 170-92. Kuwabara, K., Watanabe, T., Ohguro, T., Itoh, Y. and Maeda, Y. (2002), “Connectedness oriented communication: fostering a sense of connectedness to augment social relationships”, IPSJ Journal, Vol. 43 No. 11, pp. 3270-8. Lee, M.W. (2001), “Chatscape: a behavior-enhanced graphical chat built on a versatile client-server architecture”, Master of Engineering in Electrical Engineering and Computer Science Thesis, Massachusetts Institute of Technology, Cambridge, MA. Ling, R. (2003), “The socio-linguistics of SMS: an analysis of SMS use by a random sample of Norwegians”, in Ling, R. and Pedersen, P. (Eds), Front Stage – Back Stage: Mobile Communication and the Renegotiation of the Social Sphere, Conference Proceedings, Grimstad, 22-24 June. McKenna, K.Y.A., Green, A.S. and Gleason, M.E.J. (2002), “Relationship formation on the internet: what’s the big attraction?”, Journal of Social Issues, Vol. 58 No. 1, pp. 9-31. Markovits, H., Benenson, J. and Dolenszky, E. (2001), “Evidence that children and adolescents have internal models of peer interactions that are gender differentiated”, Child Development, Vol. 72 No. 3, pp. 879-86. Nardi, B., Whittaker, S. and Bradner, E. (2000), “Interaction and outeraction: instant messaging in action”, CSCW’00, ACM Press, New York, NY, pp. 79-88. O’Sullivan, P.B. (2000), “What you don’t know won’t hurt me: impression management functions of communication channels in relationships”, Human Communication Research, Vol. 26 No. 3, pp. 403-31. Parks, M.R. and Floyd, K. (1996), “Making friends in cyberspace”, Journal of Communication, Vol. 46 No. 1, pp. 80-97. Reid, D. and Reid, F.J.M. (2003), “Text mates and text circles: insights into the social ecology of SMS text messaging”, in Lasen, A. (Ed.), The Mobile Revolution: A RetrospectiveLesson on Social Shaping, Proceedings of the 4th Wireless World Conference, 17-18 July, Digital World Research Centre, University of Surrey, Guildford. Seeley, E.A., Gardner, W.L., Pennington, G. and Gabriel, S. (2003), “Circle of friends or members of a group? Sex differences in relational and collective attachment to groups”, Group Processes & Intergroup Relations, Vol. 6 No. 3, pp. 251-63. Thurlow, C. (2003), “Generation txt? Exposing the sociolinguistics of young people’s text messaging”, Discourse Analysis Online, (online pre-print). Walther, J.B. (1996), “Computer-mediated communication: impersonal, interpersonal, and hyperpersonal communication”, Communication Research, Vol. 23 No. 1, pp. 3-43.
References Benenson, J. and Christakos, A. (2003), “The greater fragility of females’ versus males’ closest same-sex friendships”, Child Development, Vol. 74 No. 4, pp. 1123-9.
200
1. Introduction
The future of mobile technology and mobile wireless computing Jim Hart and Mike Hannan
The authors Jim Hart is Security Program Manager and Mike Hannan is Security Engineer, both at Symantec, Maidenhead, UK.
Keywords Mobile communication systems, Computers, Forecasting
Abstract It is often stated that mobile wireless computing is going to be the next big technology revolution that will grip the world in the same way mobile telephones did in the 1990s. However, while the technology is rapidly improving, the rate of uptake has been lower than expected. This paper describes some of the reasons for this, and discusses some of the proposed solutions.
Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/1065-0741.htm
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 201-204 q Emerald Group Publishing Limited · ISSN 1065-0741 DOI 10.1108/10650740410567554
Popular mobile computing technology began in the 1960s and 1970s with digital watches (Pulsar in 1971) and calculators (Texas Instruments in 1967). It was readily apparent that the two combined would be a useful tool and the two devices were indeed merged to some success when miniaturisation technology permitted (Pulsar in 1975). This trend for miniaturisation of popular technologies is still very much one of the largest driving forces for technological progress that is enabling ever more complex computing devices to be ported to smaller and smaller “hand-held” formats. The most popular mobile computing device by far is the mobile phone. Arguably this is because of its inherent usefulness as a communication device, with the emphasis on “communication”. The success is due to the simple fact that everybody can communicate primarily through voice and can therefore find some obvious relevance in their own domain of experience. As long as that technology is available and useful to (almost) everybody, is easy to use and understand, and is not inconvenienced by some other limitation such as an annoyingly short battery life, then it has the necessary ingredients of success. However, while mobile phones are still enjoying massive demand, other related technologies such as personal digital assistants (PDAs) and mobile phones combined with extra computing functionality and fast data network access have not met the expectations of either the consumer or the manufacturers.
2. Main 2.1 Screen technology When looking at the current batch of PDAs, one of the main limiting factors in terms of size reduction of the unit (and subsequently its appeal against the mobile phone) is the display size. Also the mobile phone, with the releases of popular picture phones, is looking for an advanced screen as possible (according to Nokia, half of all mobile handsets could incorporate cameras in the near future). The manufacturers are caught in a situation where they want to present as much information as possible on the displays, without compromising the size and mobility of the unit. Therefore screen technology is constantly pushing the technical limits of both screen size, and resolution technology. The ideal to this solution would be to have a screen that is big enough to contain all that the consumer realistically would want on display at one time, coupled with it being small enough to be
201
The future of mobile technology and mobile wireless computing
Campus-Wide Information Systems
Jim Hart and Mike Hannan
Volume 21 · Number 5 · 2004 · 201-204
appealing. (For the sake of this discussion, we will say that the desirable size is roughly the size of an average pocket). When addressing screen technology, the following should be considered. People viewing hypertexts can only presently view them electronically. If the consumer wishes a paper format, the hypertext looses most of its appeal (i.e. its ease of access, links, and updatability). It is not yet possible for paper to display hypertext, but it maybe possible for electrical displays to take the form of paper. E Ink with Philips has said that they expect to have a “broad commercial distribution” by 2004 of a “paper” screen. This display, utilising electronic ink technology, will produce a resolution of 160 pixels per inch. The paper has a display containing millions of microcapsules, each smaller than a human hair in diameter. Every capsule contains positively charged white particles and negatively charged black particles suspended in a clear fluid. Last year E Ink demonstrated a display half the thickness of a credit card. Engineers at Xerox created Gyricon, which uses tiny beads and charges to present one of two colours on screen. Xerox has produced a screen that retracts and rolls up like projector screen that is the size of a pen, with a screen that retracts. While Gyricon is aiming at the shop sign market, E Ink is pressing ahead with displays for mobile phones, and other technologies. Not only does this suggest power consumption reduction for the handheld user, but also the screen itself would no longer have to be standard. Screens could be part of clothing, part of your car window, in the gym, and also could be rolled/folded/crumpled up and transported with you anywhere that you choose. Coupled with a local area network (LAN) wireless technology like Bluetooth, then the possibilities are increased, as you would just need to be in the vicinity of your device to be able to use it. Flexible organic light emitting device (FOLED) technology already permits the use of a cheap substrate that can be made to display a highresolution moving image with its own inbuilt light source (see Plate 1 and Bulovic et al. (1996)). This low power consumption design may soon enable the many applications discussed for flexible screens and liberate the hand-held market from the constraints of their small viewable areas. The possibility of colours in these “paper” screens is yet to be hugely discussed, but is a natural progression. The technology needs to be developed, to enable more than two colour options linked to the positive and negative charge, but this will happen. The technology will lead to devices being smaller, but displays being larger and more
Plate 1 FOLED screen
versatile. That way the manufacturers will have reached their goal of large colour screens for small hand-held and pocket devices.
2.2 Wireless connectivity In an age where speed is allegedly the goal for us all, the spectrum of wireless devices continues to grow, with the devices always getting “faster”. However, when considering wireless speeds of mobile devices one must first consider the speed of the network that the wireless device is functioning from. The primary wireless communication devices that appeared in the 1980s (1G) were based on the Advanced Mobile Phone service (AMPS) standard. 1G provided the foremost voice service-using analogue, but no data service. Technology advanced onto 2G networks, which used a variety of standards (TDMA, GSM, CDMA, PDC) to provide maximum data rates of up to 14.4 kb/second, as well as advancement of digital voice. The 2.5G network utilises General Packet Radio Service (GPRS), which improved the performance of GSM and TDMA by converting them to packet-based networks. This provided 2.5G networks with . 144 kb/second, and the “always on” technology that today is well known. Add to 2.5G Enhanced Data GSM Environment (EDGE), and the bandwidth of GPRS is increased to 384 kbs. With the emergence of 3G networks, the development of wireless bandwidth has increased again. 3G can boast . 2 Mbps of data transfer, coupled with ever increasing coverage. The future technology holds 4G. 4GMF estimates that by 2008 the 4G mobile market will be worth over $400 billion, and from 2005-2010 4G will dominate wireless communications. 4G
202
The future of mobile technology and mobile wireless computing
Campus-Wide Information Systems
Jim Hart and Mike Hannan
Volume 21 · Number 5 · 2004 · 201-204
provisionally has the capability to reach speeds of up to 20 Mbps. The simultaneous use of a 4G network by a number of users has been shown to have little effect on an individual’s bandwidth, and will also be very stable for use in moving vehicles (Wei Wang et al., 2003). Japanese mobile communication company DoCoMo is presently aiming to provide 100 Mbps for downloads, coupled with 20 Mbps for uploads on 4G. DoCoMo aim to get 4G at this speed commercialised in 2010. Early testing and development has progressed well. During October 2002, a successful test was completed in the laboratory – DoCoMo are now attempting field trails of the new system. It may be worth mentioning that some companies have already attempted to cater for a higher data connection speed as the grail for mobile users. By offering local wireless datanetworking using the 802.11b standard users can achieve speeds of around 11 Mbps. Recently many operators of 802.11b networking accesspoints have dramatically expanded their deployment of “hot-spots” to cafe´s, hotels and airports. Reports continue to cast doubt on whether this technology will ever provide sufficient coverage or appeal to users, and most analysts continue to predict that the vast majority of the hot-spot owners will fail to bring in the number of users per month to make their services even break even on costs. Many reasons exist for the slow uptake of these services, but the main ones are lack of roaming capability due to different charging structures by different operators and short ranges. Compared to the promise of fast almost global coverage of 4G by mobile network services, excellent roaming capability and lower charges, 802.11b wireless networking may only be a partial success while users wait for the 4G service deployment. 4G will provide an amount of bandwidth that would provide adequate speed and power to control most of today’s modern applications, and would dramatically bring down the costs associated with data network usage. Using the 4G mobile network wireless bandwidth mentioned above, any device would have real-time access to the internet, always on, and with very little delay. This would free the mobile data user from the current constraints of the existing low bandwidth and unreliable GSM and 2.5G networks, and the restrictions imposed by 802.11b. Speed and cost is the largest factor contributing to the low demand for the new multi-media services offered by the service providers, but with much faster and relatively cheaper access the mobile network operators can expect the long anticipated boom in mobile data connectivity.
2.3 Size, battery and computing power Another major factor contributing to the true mobilisation of the connected user is the significant inconvenience of short battery life. While mobile phones have made impressive improvements on standby and talktime, this is in fact mostly due to the gains made by the efficiency of the electronic components. Laptop computers and PDA devices have not seen any significant increase in battery life for over five years, and while this seems odd considering the advances made in battery technology, the reason is that the demands on the battery has increased with faster processors (Buchmann, 2001). As the processor speeds increase, so has the power consumption. The result is that while batteries have a higher capacity than five years ago, the actual runtime of portable computing devices has remained roughly the same. Until there is a quantum leap in battery technology, the run time for portable computing devices is not expected to change. This leap may by just around the corner with fuel cell technology. The latest lithium-ion batteries can offer a high energy density of around 150 Wh/kg, but fuel-cells which can run on easily produced methanol can push this energy density to 6,000 Wh/kg. Standard battery technology relies on the chemical energy stored in an electrochemical cell and the subsequent release of electrical energy during a controlled chemical reaction. Fuel cells use methanol as a base to carry hydrogen, which is the key to the energy densities possible with this type of battery. The limits of energy density are governed by the chemicals and materials used in the battery, and even the best modern lithium-ion batteries will never approach that possible in a fuel cell. Hydrogen can be used to combine with oxygen in a reaction that is many times more energetic than using conventional battery cell technology. Early commercial interest with fuel cells originally focussed around the clean energy applications for automotive transport functions, as the only by-product of a fuel cell is water and carbon dioxide (see Figure 1). Development for this type of application has increased the efficiency of the cells to a point where a miniature version (Gottesfeld, 2002) is capable of providing the currents necessary to run modern small electronic devices (see Plate 2). Methanol fuel cells typically use hydrogen liberated from an aqueous methanol solution to create electricity using a polymer electrolyte membrane to act as a catalyst for a chemical reaction to separate electrons from hydrogen which is now positively charged and combines with negatively charged oxygen atoms to form water. Unlike conventional battery technology, fuel cells
203
The future of mobile technology and mobile wireless computing
Campus-Wide Information Systems
Jim Hart and Mike Hannan
Volume 21 · Number 5 · 2004 · 201-204
Figure 1 Direct methanol fuel conversion (DMFC)
go days without recharging or replacing batteries. The potential for a leap in mobile computing power will also be made possible because one of the main restrictions imposed on processor chip manufacturers was to provide small processors with low power consumption. Increasing the allowable power consumption has an immediate effect on the processor chips that can be used for mobile devices.
3. Conclusion Faster processors already exist that have not been used for mobile computing devices simply because of the power required to use them. Mobile computing technology will receive the much needed boost in both performance and battery life that is required to make them more desirable when fuel cells are used in place of conventional batteries. Coupled with the possibility of large fold-away and more versatile screen technology and the ability to deliver faster media and data over 4G and 802.11b wireless technologies, mobile computing will become not only more accessible to everyone, but also easier to use. All of these ingredients are just around the corner, and when they are mixed together we should finally experience the revolution that has been waiting to happen.
Plate 2 MTI micro demonstrating fuel cell powering a PDA phone
References actually consume the “fuel” that drives them and require the input of the raw fuel material methanol when depleted. Luckily methanol is very cheap and many designs exist to allow quick and easy refuelling of commercially available cells. The combination of expensive catalysts and a large surface area required to produce sufficient current has meant fuel cells are not yet commercially viable. Aggressive development (Narayanan, 2003) does mean by 2004 this technology will begin to be seen but cost and restrictions for use on airplanes must be overcome before they will replace the conventional battery pack for laptops, mobile phones and PDAs. Increasing the battery life by orders of magnitude will enable mobile computing users to
Buchmann, I. (2001), “Choosing the right battery for portable computing”, Batteries in a Portable World, 2nd ed., Cadex Electronics, Richmond. Bulovic, V., Gu, G., Burrows, P.E., Thompson, M.E. and Forrest, S.R. (1996), “Flexible organic light emitting devices”, Nature, Vol. 380, p. 29. Gottesfeld, S. (2002), “Fuel cells for portable power applications: an opportunity for early market entry”, paper presented at the 2nd Ohio Fuel Cell Symposium. Narayanan, S.R. (2003), “Development of advanced catalysts for direct methanol fuel cells”, Hydrogen, Fuel Cells, and Infrastructure Technologies FY 2003 Progress Report, California Institute of Technology, Pasadena, CA. Wang, W., Ottosson, T., Sternad, M., Ahl’en, A. and Svensson, A. (2003), “Impact of multiuser diversity and channel variability on adaptive OFDM”, paper presented at the IEEE Vehicular Technology Conference VTC2003-Fall, Orlando, FL.
204
1. Introduction
Translating scalable video streams from wide-area to access networks Carsten Griwodz Steffen Fiksdal and Pa˚l Halvorsen The authors Carsten Griwodz is Associate Professor, Steffen Fiksdal is Chief Consultant and Pa˚l Halvorsen is Associate Professor, all at IFI, University of Oslo, Oslo, Norway.
Keywords Video, Internet, Transmission Control Protocol/Internet Protocol, Data handling
Abstract Transmitting video over user datagram protocol (UDP) has been considered advantageous because it allows for discarding of packets in favor of retransmissions, and sender-controlled timing. Using UDP has been criticized because it allows video streams to consume more than their fair share of bandwidth, which is typically associated with the back-off behavior of transmission control protocol (TCP). TCP-friendly algorithms are meant as a middle path. However, UDP delivery to end systems may still be prevented by firewalls or for other reasons, and TCP must be used. This in turn suffers from bandwidth fluctuations. Investigates an architecture that separates the transfer of a video stream over long distances. Considers a proxy server to translate the traffic and two straightforward approaches for the translation of a layered video stream transmission from the TCPfriendly transport protocol to TCP. Does not expect that one of these two approaches is by itself suited for the task, but investigating them will provide insights into their basic functions and help in discovering appropriate modifications. For the investigation, an experimental approach was used where network behavior and content are emulated.
Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/1065-0741.htm
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 205-210 q Emerald Group Publishing Limited · ISSN 1065-0741 DOI 10.1108/10650740410567563
The amount of streaming media services in the internet has been increasing over the last few years, and on-demand access to stored content makes up a major share of this. To reduce the increase in bandwidth demand on their up-link connection to the general internet, many internet service providers (ISPs) make use of proxy caching at the edge of their network. Since these proxy caches cannot store all content, there will still be delivery of streams from original servers, but we propose that the proxy servers should be used to improve this delivery too. Typically, streaming services in the internet are envisioned to use user datagram protocol (UDP) for data transfer because multimedia applications have other demands on bandwidth, reliability and jitter than offered by TCP. This is in reality impeded by firewalls installations in access networks and on end systems allowing TCP traffic only. Therefore, a solution that uses streaming of TCP in access networks is desirable. The transport of streaming media data over several backbone networks, however, would be hindered by the use of TCP because of its probing behavior, which leads to rapid reduction and slow recovery of its packet rate. Applications using UDP do not have to follow this approach and have therefore been criticized because they do not leave other traffic its fair share of the available bandwidth. Since TCP is considered an appropriate reference for fairness, protocols and mechanisms that react to congestion in a way that is similar to TCP are classified TCPfriendly and considered more acceptable. Widmer et al. (2001) provide a survey of such protocols and mechanisms. One definition of TCP-friendliness is that the packet rate of a TCP-friendly application should be proportional to q the inverse square root ffiffiffiffiffiffiffi of the packet loss rate (1= p ). TCP-friendly transport achieves the same throughput as TCP on average but with less throughput variations. We show considerations for combining the use of TFRC in the backbone with the use of TCP in access networks. In particular, we identify buffer use on the proxy and rate fluctuations in the network. We relay on scalable video, which makes it possible for real-time streaming services to adapt to variations in the packet rate. A variety of approaches exists, and in this paper, we consider an approach to course-grained scalable video that achieves high quality, scalable MPEG (SPEG) (Krasic, 2004). Even though Krasic (2004) aims at an approach for streaming scalable video over TCP, it works with TCP-friendly approaches that have fewer variations in the packet rate than TCP and lead to reduced buffer use on server and clients. The rest of this paper is organized as follows: section 2 looks at some related work. In section 3, we describe our experiments and our
205
Translating scalable video streams from wide-area to access networks
Campus-Wide Information Systems
Carsten Griwodz, Steffen Fiksdal and Pa˚l Halvorsen
Volume 21 · Number 5 · 2004 · 205-210
results. Based on our results, we outline different mechanism combinations in section 4, and finally, we conclude in section 5.
used emulation. Since the throughput of the TCPfriendly protocol and TCP will be different, a proxy implementation has four means of handling congestion in the access network: (1) It can buffer data that it receives from the server until the access network recovers from congestion. (2) It can create backpressure on the link between proxy and server by refusing to accept packets. (3) It can adapt the bandwidth between proxy and client by dropping packets from the server in a controlled manner. (4) A combination of the above.
2. Related work Proxies have been used for improved delivery of streaming media in several earlier works. Prefix caching has addressed issues of latency and backbone throughput (Sen et al., 1999). In multicast works such as receiver-driven layered multicast (McCanne et al., 1996), multicast routers which may also be proxy servers are meant to forward packets only to those subtrees of the multicast tree that have subscribed to those packets. In more recent work, proxies have filtered out layers of a scalable format to adapt to endsystem capabilities (Zink, 2003). While the idea of using scalable codecs for adaptivity is rather old, scalable codecs for streaming video that are able to maintain a high quality over a wide range of bandwidths have been developed only in the last few years. Horn and Girod (1994) and Krasic (2004) have worked on course-grained scalability, more recently fine-grained scalability and multiple description coding have attracted interest. In this paper, we want to draw attention to issues that occur when a proxy is used to translate transport protocols in such a way that TCP-friendly transports mechanisms can be used in backbone networks and TCP can be used in access networks to deliver streaming video through firewalls. Krasic (2004) argues that the most natural choice for TCP-friendly traffic is using TCP itself. While we agree in principle, their priority progress streaming approach requires a large amount of buffering to hide TCP throughput variations. In particular, this smoothing buffer is required to hide the ratehalving and recovery time in TCP’s normal approach of probing for bandwidth which grows proportionally with the round-trip time. To avoid this large buffering requirement at the proxy, we would prefer an approach that maintains a more stable packet rate at the original sender. The survey of Widmer et al. (2001) shows that TFRC is a reasonably good representative of the TCP-friendly mechanisms for unicast communication. Therefore, we have chosen this mechanism for the following investigation.
3. Experiments To investigate the effects of translating a video stream from a TCP-friendly protocol in the backbone to TCP in the access network, we have
In this investigation, we have considered the first two options. We expect that the third option would yield results similar to our investigation for TCPfriendly transport in backbone and access networks in Zink et al. (2003). The fourth option requires understanding the first three and is future work. 3.1 Experimental set-up In our experimental investigation, we have emulated a real system using the server, proxy and client implementations of the KOMs Streaming System (komssys, Zink et al. (2001)), the NistNET network emulator (Carson and Santay, 2003) to emulate delay in backbone and access network links as well as packet loss in the access network, and the tg traffic generator to model cross traffic in the backbone[1]. For the investigation, an implementation of RTP over TCP was added to komssys. This option was preferred over the inband delivery of video-data in the RTSP control channel in order not to break the separation of control and data flow in komssys. To support TCP-friendly behavior between server and proxy, we also use the TFRC implementation in komssys, which is an RTP variation in which we use application-specific RTCP messages that are used to acknowledge each received packet. Note that TFRC is modelled after TCP Reno, but as TCP implementation we used the default of Linux 2.4, which is TCP FACK. The content used consists of dummy videos that are modeled after the SPEG encoding that is used in priority progress streaming. The dummy videos have four layers of constant equal bandwidth of 0.25 Mbps. Only the server in our scenario discards packets in a controlled manner according to their priority. In particular, and in contrast to Zink (2003), the proxy in this scenario uses priority-aware filtering. By considering layered video, we achieve a situation where controlled packet loss can allow the reduction of bandwidth that is required for transmission of the video and maintain the videos original frame rate. We intend to maintain this frame rate. We ignore badput at
206
Translating scalable video streams from wide-area to access networks
Campus-Wide Information Systems
Carsten Griwodz, Steffen Fiksdal and Pa˚l Halvorsen
Volume 21 · Number 5 · 2004 · 205-210
the proxy, i.e. we do not discard high layer packets at the proxy when their corresponding lower layer packets have become lost. In the following two sections, we show the observations that we made in the case of using buffering and backpressure, respectively. We use the term packet rate when we refer to the amount of data that are sent from the server to the proxy. This is appropriate because the transmission is performed in independent UDP packets, which is an important detail for the interpretation of section 3.3. We use a constant MTU size of 1,500 bytes, and therefore, the packet rate translates directly to a bandwidth. In each of the tests shown in Figure 1 we investigate nine different scenarios. These are combinations of packet loss in the access network and cross-traffic in the backbone. We consider 0 percent, 2 percent and 5 percent packet loss in the access network, and no, medium or high cross traffic in the backbone. Both medium and high cross-traffic are modeled by one Pareto source with a Hurst parameter of 0.5 for the average selfsimilarity and two greedy TCP sources.
proxy grows infinitely. This condition must still be fulfilled on average when packet loss between proxy and client leads to a bandwidth reduction. In these scenarios, two resources are critical, the buffer size in the proxy and the jitter experienced at the client: (1) The buffer size can become large at the proxy even under the condition that the average data rate that is sent from the server is lower than the average TCP throughput between proxy and client. Since we perform neither packet dropping or reordering at the proxy, the latency of packet arrivals at the client grows with the buffer size at the proxy, and it must be accommodated by pre-buffering the beginning of earlier data at the client. The user will experience a large startup latency, unless the client implementation ignores these buffering delays and chooses to break the continuous playback of the video. (2) The jitter experienced by the client is near exclusively due to the delays introduced by the TCP congestion control between proxy and client. In addition to the buffering at the client that is necessary to overcome the delay that is introduced by the variation in average throughput on the TCP link, buffering is also necessary to hide jitter from the end-user.
3.2 Buffering In this investigation, we look at scenarios where the proxy buffers all data received from the server until they can be forwarded to the client. No data are discarded at the proxy. This implies that the packet rate between server and proxy is decoupled from the transmission of data from the proxy to the server. So, in case of congestion that is experienced between proxy and client, buffer growth at the proxy cannot be avoided. However, the bandwidth between proxy and client must be sufficient to accommodate the full data rate of the video that is sent from the server to achieve a stable scenario. If this is not the case, the buffer consumption at the Figure 1 Latency comparison for 0 percent, 2 percent, 5 percent access net packet loss, and no, medium, high backbone cross-traffic
The graph in Figure 1 compares the latency under various network conditions. It shows that the cross-traffic between server and proxy, which leads to loss and packet rate reduction in TFRC, affects the end-to-end latency even though the latency between server and proxy remains nearly constant. Thus, the change is only due to the development of buffer space at the proxy. It is noteworthy that the affect on the minimal and average latency is limited, while the standard deviation and 95percentile is considerably smaller when the TFRC algorithm reduces the bandwidth between server and proxy. This implies that the server reduces the bandwidth of the video at the source by sending only the lower layers of the layered video. The reason for the limited effect on the average delay is that the buffer at the proxy must be empty most of the time. Since the average bandwidth between proxy and client is limited to a bandwidth that can be sustained even in the case of high packet loss between proxy and client, we have guaranteed that the buffer on the proxy is frequently empty even if no packets are lost between server and proxy (see Figure 2). Thus, the buffer is small in the average case and grows only under temporary bandwidth reduction due to the TCP congestion control algorithm. Then, the rate of packet arrival from the server determines how quickly the buffer grows.
207
Translating scalable video streams from wide-area to access networks
Campus-Wide Information Systems
Carsten Griwodz, Steffen Fiksdal and Pa˚l Halvorsen
Volume 21 · Number 5 · 2004 · 205-210
Figure 2 Buffer development on a proxy TCP link
3.3 Backpressure In this section, we consider a single-threaded implementation that uses blocking TCP write for transmission from the proxy to the client. This variation is advantageous because of its straightforward implementation. We can also expect that congestion in the access network affects the connection between server and proxy. For this scenario, we increase the TFRC bandwidth limit in such a way that the access network cannot support the maximum data rate that can be sent from the server. The sending rate of the server must be limited by backpressure from the proxy. This means for TFRC that the proxy must report loss. In our experiments, we observe huge fluctuations in the TFRC transmission rate that far exceed the bandwidth variations of TCP even in the best case, without packet loss in the access network and no cross traffic in the backbone. Obviously, this means that the server changes the number of layers that it sends for video frequently. Even though all packets are delivered to the client, it experiences frequent quality variations. We found that the reason for this is the TCP implementation, in particular the way in which the kernel notifies the application about free space in the TCP send buffer. The reason is found in the Linux implementation of Clark’s algorithm: a sender that blocks in a TCP write operation is woken up only when the send buffer becomes one third empty. This behavior affects the TFRC algorithm indirectly. While the proxy blocks in a TCP write operation, UDP packets continue to arrive from the server. However, UDP packets are not queued in the kernel and since the receiving thread blocks in the TCP write, the UDP packets are dropped. Consequently, they are reported as lost packets to the TFRC algorithm, which reduces that data rate. Subsequently the lower data rate can be handled by the TCP send buffer without running full and blocking the proxy thread. No more misses are reported to the server
and TFRC algorithm increases the bandwidth gradually. An interesting detail of this problem is that it worsens when the average bandwidth is reduced. Thus, when the bandwidth of the access network is lower, the server will, on average, send no more packets than can be supported by the access network, but the intensity of quality changes that become visible at the client is larger. Figure 3 shows this effect. The difference in the figure appear because it takes longer to transfer one third of the constant-sized TCP send buffer on a 128 kbps line than on a 512 kbps line. In the time it takes on a 128 kbps line, more packets from the server are lost and TFRC reduces the bandwidth much further than in the 512 kbps case. Subsequently, the same number of packets are successfully copied into the TCP send buffer (even though of lower video layers), the same low packet loss numbers are reported to TFRC as in the 512 kbps case, and the sending rate of the server rises in a similar manner. Another detail is that the situation improves somewhat when some loss occurs in the access network.
4. Evaluation We conclude from the buffering experiments in section 3.2 that it would be very beneficial for the latency at the client to buffer only the lower layers of the video in case of temporary congestion on the proxy-client link, and to discard the higher layers in a controlled manner. We can imagine two approaches for reducing the effects of the TCP congestion control mechanism: (1) We can send feedback to server to reduce bandwidth at the server side. However, this Figure 3 Latency variation over access networks with 512 kbps and 128 kbps and 0 percent, 2 percent, 5 percent packet loss
208
Translating scalable video streams from wide-area to access networks
Campus-Wide Information Systems
Carsten Griwodz, Steffen Fiksdal and Pa˚l Halvorsen
Volume 21 · Number 5 · 2004 · 205-210
may take too long because the time for sending an entire TCP send buffer of 64 kbytes which contains between 62.5 ms and 250 ms of video data, depending on the number of video layers transmitted. We assume big round-trip times between server and proxy (200 ms in our experiments), and we have seen in Figure 2 that even in the worst case, the proxy buffer does not exceed the size of one additional maximum TCP send buffer. Thus, this would be counterproductive and only prevent higher layers from reaching the client after the temporary congestion has disappeared. However, reporting to the server the average rate that is supported between proxy and client would be a good alternative to the fixed bandwidth limit of our experiments. (2) We can filter layers at the proxy. Filtering at the proxy is an alternative for this. By examining the timestamps and priorities that are associated with the packets that arrive from the server, packets of the higher layers can be discarded. In this way, lower layers packets are buffered that have later timestamps, and thus, the effective latency at the time of arrival at the client is reduced.
send buffer space to the application. One such means is the TCP Minbuf patch cited in Krasic (2004). Their approach reduces the TCP send buffer to the congestion window size and applies Clark’s algorithm only to that buffer. Thereby, applications that perform TCP write are awaken more frequently when the available bandwidth is low. In our backpressure approach, this would lead to more frequent, but less severe TFRC loss reports to the server. The server would then determine a less varying packet rate and send more lower layer and less higher layer packets. The priority-aware dropping and jitter threshold approaches have the disadvantage that they decouple backbone and access network like the buffering approach, even though the effect quality is reduced. But to some extent, packets that are dropped at the proxy could have been dropped from the transmission of the server in the first place. One possibilty would be the use of early congestion notification (ECN) extensions to TFRC that makes it possible to reduce the sending rate without actual losses.
As an alternative to this approach, which decouples the backbone and access network traffic completely, we can look at a strict coupling without blocking as well. In Priority-aware dropping, we drop higher layers from the queue in a controlled manner to prevent jitter when the application level buffer grows. Approaches that could be adapted to this have been presented by Zink (2003) and by Krasic (2004). Furthermore, the proxy could filter packets based on a jitter threshold. It requires that the client informs the proxy about the amount of artificial latency that it introduces to reduce jitter, and the amount of buffer space that it uses for this. The proxy can use this information to make decisions about packet forwarding. It can extract the expected playout time of packets arriving from the server, and discard packets from its send queue if they would have to be discarded at the client. As shown by Zink (2003), the best solution is not to discard the oldest packet, but the number of layers that are forwarded should be maintained constant for as long as possible. This prevents foremost the jitter but also smoothes out the layers so that the quality variations can be kept at a minimum. From the backpressure experiments in section 3.3 we can conclude that an implicit limitation of the bandwidth between server and proxy is not at all a viable solution without a better means of reporting availability of TCP
Conclusion and future work In this paper, we have investigated two straightforward approaches for the translation of a layered video stream transmission from a TCP-friendly transport protocol in the backbone to TCP in the access network. As expected, we found that neither of these unmodified approaches are particularly well suited for the task, but we have gained valuable insight into the reasons for the failures of these two approaches. Specifically, we found for the case of strict decoupling of backbone and access network that unpredictable amounts of buffering are necessary at the proxy. These would force clients to prefetch large amounts of data before starting playback or to accept interruptions in playout. For the case of strict coupling of backbone and access network, we traced the problems to the typical implementation of TCP. We looked at forwarding in a single thread using blocking TCP write. Typical TCP implementations allow applications to copy data into the TCP send buffer in burst, which leads to long phases of blocking and long phases of non-blocking in the proxy thread, which results in inconclusive feedback to the server and frequently quality variations at the client. We have therefore discussed a set of options that lead to a middle way and can improve performance. We plan to investigate alternative options like those described in section 4 in the future.
209
Translating scalable video streams from wide-area to access networks
Campus-Wide Information Systems
Carsten Griwodz, Steffen Fiksdal and Pa˚l Halvorsen
Volume 21 · Number 5 · 2004 · 205-210
Note 1 The packages used are available from http://komssys. sourceforge.net, http://snad.ncsl.nist.gov/nistnet and http://www.kom.tu-darmstadt.de/rsvp, respectively.
References Carson, M. and Santay, D. (2003), “NIST Net: a Linux-based network emulation tool”, ACM Computer Communication Review, Vol. 33 No. 3, July, pp. 111-26. Horn, U. and Girod, B. (1994), “Pyramid coding using lattice vector quantization for scalable video applications”, Proceedings of the International Picture Coding Symposium (PCS), Sacramento, CA, September, pp. 183-5. Krasic, C. (2004), “A framework for quality adaptive media streaming”, PhD thesis, OGI School of Science & Engineering at OHSU, Beaverton, OR, February. McCanne, S., Jacobson, V. and Vetterli, M. (1996), “Receiver-driven layered multicast”, ACM Computer
Communication Review, Vol. 26 No. 4, August, pp. 117-30. Sen, S., Rexford, J. and Towsley, D. (1999), “Proxy prefix caching for multimedia streams”, Proceedings of the Joint Conference of the IEEE Computer and Communications Societies (INFOCOM), New York, NY, March, IEEE Press, Piscataway, NJ, pp. 1310-19. Widmer, J., Denda, R. and Mauve, M. (2001), “A survey on TCP-friendly congestion control”, IEEE Network Magazine, Vol. 15, February, special issue: Control of best effort traffic, pp. 28-37. Zink, M. (2003), “Scalable internet video-on-demand systems”, PhD thesis, Darmstadt University of Technology, Darmstadt, November. Zink, M., Griwodz, C. and Steinmetz, R. (2001), “KOM player – a platform for experimental VOD research”, IEEE Symposium on Computers and Communications (ISCC), July, pp. 370-5. Zink, M., Griwodz, C., Schmitt, J. and Steinmetz, R. (2003), “Scalable TCP-friendly video distribution for heterogeneous clients”, Proceedings of SPIE/ACM Conference on Multimedia Computing and Networking (MMCN), San Jose, CA, January, SPIE, Bellingham, WA, pp. 102-13.
210
Introduction
An audio stream redirector for the Ethernet Speaker Ishan Mandrekar Vassilis Prevelakis and David Michael Turner
The authors Ishan Mandrekar is a Graduate Student, Vassilis Prevelakis is an Assistant Professor and David Michael Turner is an Undergraduate Student, all in the Department of Computer Science, Drexel University, Philadelphia, Pennsylvania, USA.
Keywords Audio media, Internet, Protocols
Abstract The authors have developed the “Ethernet Speaker” (ES), a network-enabled single board computer embedded into a conventional audio speaker. Audio streams are transmitted in the local area network using multicast packets, and the ES can select any one of them and play it back. A key requirement for the ES is that it must be capable of playing any type of audio stream, independent of the streaming protocol, or the encoding used. The authors achieved this by providing a streaming audio server built using the kernel-based audio stream redirector (ASR) in the OpenBSD kernel. The ASR accepts input from any of the existing audio file players, or streaming audio clients. Since all audio applications have to be able to use the system audio driver, this system can accommodate any protocol or file format, provided that there exists some compatible player running under OpenBSD. This paper discusses the design and implementation of the server as an ASR, the streaming protocol developed for this application, and the implementation of the client.
Electronic access The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at www.emeraldinsight.com/1065-0741.htm
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 211-216 q Emerald Group Publishing Limited · ISSN 1065-0741 DOI 10.1108/10650740410567572
Audio hardware is becoming increasingly common on desktop computers. With the advent of easy to install and cost-effective networking solutions such as ethernet and more recently, wireless local area networks (LAN), there has been an increasing interest in allowing audio sources to be shared by multiple devices. One central server can stream audio data to a number of clients who are all part of the same network. The clients just need to capture the streamed data and play it on their audio devices (Figure 1). This concept can work at home, as well as a campus or office LAN where it can be used as the basis for wireless surround audio systems, background music, public announcements, etc. However, the popularity of audio services in the internet has resulted in the deployment of a large number of protocols and data formats producing an online Babel of mismatched applications and protocols. Moreover, most of the popular streaming services have adopted incompatible and proprietary formats. It is particularly noteworthy that even though Real Networks released the source code for Helix Client, they did not provide source for the Re-alAudio G2 and RealAudio 8 decoders (Helix, 2002). In addition to the streaming audio formats, there are also a large number of encoding schemes for file-based audio players (MP3, wav, and so on). In order to be able to cope with all these formats, users have to use multiple audio playback applications, each with its own user interface and peculiarities. Moreover, most commercially available streaming servers cannot stream audio files in their native formats. These files need to be “hinted” with metadata which aids the streamer in generating real time protocol (RTF) packets (Apple, 1998). This mandates that the audio files undergo a format conversion before they become eligible for streaming. Our goal was to produce an audio server that would be independent of the format of the audio source. We also endeavored to create a single client that would be able to accept music from any of the above players without the need for specialized plug-ins or updates. In the next section we discuss the issues surrounding the design of the audio client and the streaming server. We, then, describe our prototype implementation, which allows the redirection of the output of off-the-shelf audio player applications using a kernel-based redirector. Finally, we discuss how our work compares with related projects and systems, and our plans for future enhancements. This work was supported by NSF under Contract ANI-0133537.
211
An audio stream redirector for the Ethernet Speaker
Campus-Wide Information Systems
Ishan Mandrekar, Vassilis Prevelakis and David Michael Turner
Volume 21 · Number 5 · 2004 · 211-216
Figure 1 Re-broadcasting the audio from a remote server in the LAN
Design The server A typical audio device (Lowe and Rizzio, 1998), as shown in Figure 2, consists of two logically independent systems, one for recording, and another for playback. In the recording section, analog input is first passed through a mixer before being fed to an analog to digital converter (ADC), whose output is subsequently stored in a buffer. The audio hardware dictates various configuration parameters such as resolution (8 bits or 16 bits), data formats (linear or m-law), channels (mono/ stereo) and the sampling rate. The playback section makes use of a similar buffer to hold data that is to be passed to the digital to analog converter (DAC). The output of the DAC is channeled through a mixer before being
sent to the speakers. The operating system also provides format conversions to supplement the native features supplied by the hardware. An audio playback application like mpg321 (1), works by continuously reading blocks of audio information from the source (file, or network connection), performing some processing and sending the result to the audio device for playback. This is done by writing blocks of data to the audio special device (/dev/audio). These data are placed in the playback buffer. The device driver further divides the playback buffer into smaller size blocks before passing the data to the audio hardware. I/O control (ioctl (2)) calls are used by applications to ask the device driver about the capabilities of the underlying hardware, or to set various parameters for the audio device (e.g. sampling rate). The audio device driver in the OpenBSD 3.1 kernel is divided into a high level hardware independent layer, and a low level hardware dependent layer. The high-level driver interacts with user-level applications. It provides a uniform application programming interface to the underlying hardware dependent driver modules. Low-level drivers provide a set of function calls (NetBSD, n.d.) which allow the hardware to be integrated into the runtime environment of the operating system. A device driver does not need to implement all the calls, but at a minimum a useful driver needs to support opening and closing the device, device control and configuration, and I/O calls. Our technique involves intercepting the audio stream as it flows through the upper half of the audio driver and sending the audio information to the network (grey box in Figure 2).
Figure 2 Typical audio device with ASR extension
212
An audio stream redirector for the Ethernet Speaker
Campus-Wide Information Systems
Ishan Mandrekar, Vassilis Prevelakis and David Michael Turner
Volume 21 · Number 5 · 2004 · 211-216
The audio stream redirector Instead of writing a user level application using streaming protocols (e.g. RTF) and CODECs for complex encodings such as MP3 (MPEG Layer 3), we have added a new audio stream redirector (ASR) to the OpenBSD operating system. The ASR packets and multicasts the audio play buffers onto the network. The presence of the ASR should not be detectable by the application, thus an audio subsystem with an ASR should be indistinguishable from a real audio device to the application program. This is achieved because the ASR does not affect the normal operation of the audio driver so the application is totally unaware of the redirection of the audio information to the network (Figure 3). By positioning the audio redirector in the kernel, we provide an audio streaming facility that does not depend on the audio application, or the encoding format. Moreover, the music file does not need to be in a special “hinted” format unlike the Darwin Streaming Server (http://developer. apple.com/darwin/projects/streaming). This enables the ASR-based server to accommodate new audio encoding formats as long as a player has been ported to OpenBSD. The redirected audio stream is bound to a specific multicast address (and port). Once the program opens the audio device and starts writing to it, the ASR will begin sending packets to the network.
version of OpenBSD, a network connection (preferably wireless), an audio card, and a pair of speakers. The ES receives music from the network and plays it on its speakers. Since Drexel has a campus-wide network, the ES can receive music anywhere within our campus. Being an embedded device, the ES has limited resources, so it was considered important to standardize the format of the incoming streams. Another constraint is synchronization; if more than one ES is playing the same channel within earshot, any delay between the two outputs creates cacophony. Finally, since we expect a large number of ESs to be deployed in our campus, we had to minimize the impact of the ESs on the network resources. These three constraints made the use of an offthe-shelf player infeasible. We briefly experimented with the Real Time Audio Tool (Kouvelas and Hardman, 1997) but it did not allow us to convert, for example, real audio streams into something the ES could play. In the end we decided to use multicast packets for the streaming audio with one address per channel. This achieves synchronization between playing ESs and reduces the load on the network. We then redirect the output of existing player applications into the network. In this way we do not need to have the players in the ESs, but in one or more audio servers. By removing the players we can also standardize the user interface of the ESs. Finally, we redirect the packets received at the ESs to their corresponding audio devices without any post-processing. This reduces the processing power requirements at the ESs and achieves format standardization of the incoming streams. To avoid burdening the server, we also require that
The client Our client, which we call “The Ethernet Speaker” (ES) is a rather simple device, comprising a single board computer (SBC) running an embedded
Figure 3 Application plays audio to the normal audio device, while the ASR redirects the audio data to a remote machine via the network
213
An audio stream redirector for the Ethernet Speaker
Campus-Wide Information Systems
Ishan Mandrekar, Vassilis Prevelakis and David Michael Turner
Volume 21 · Number 5 · 2004 · 211-216
the ES selects which channel to play, without having to make arrangements with the server. Protocol design The communication protocol between the audio streaming server and the ESs has to be lightweight and easy to implement. On the server side all the protocol is running in kernel space, so a simple design decreases debugging (and kernel reboots). Also, a simple protocol imposes a lesser burden on the resources on the client side, allowing lightweight clients to be used. Given the real-time nature of the transmissions, and the continuous nature of the audio stream, there is no need for retransmissions or error correction. Incorrect packets are simply discarded. This is acceptable in our target environment which, being a LAN, exhibits low packet loss rates. Clients may join a transmission at any time and are able to recover from lost packets or transient network problems. This allows the server to be oblivious to the number or type of the clients operating in the network. Finally, we use multicast packets to reduce network load and synchronize the clients.
Implementation The streaming server When a client application accesses the audio device (via the open (2) call), data flow to the network via the ASR. This is done by having the ASR open a kernel UDP socket. The destination address in the socket’s address structure is set to a predefined multicast address and destination port. Every time data get written into the play buffer by a user-level application, they is packetized and sent to the network in the following manner: a special control packet containing information to configure the audio device at the client side precedes the transmission of the actual data. The contents of the play buffer are then sent as one or more 1,024 byte packets. The socket is finally closed when the user process terminates the connection with the audio driver. A key decision was whether to send the audio stream directly to the network, or pass it to a userlevel process, which would then decide what to do with the audio stream. Although the latter technique is more flexible and allows better handling of the audio stream, we decided not to adopt it as it is particularly wasteful in terms of resources. Since the main application of the ASR is the redirection of audio streams to the network, directing the stream to a user-level process, which then transmits it to the network, would involve superfluous data copying and context switching.
We, therefore, decided to inject the audio byte stream directly into the network as a series of multicast packets. This allows the ASR to construct the packets and forward them to the networking code entirely within the kernel. Synchronizing the audio output of two devices is not as simple as ensuring that we send the audio data to both devices at the same time. Variations in the routing topology (e.g. different number of hops between the server and the various receivers), Operating system actions (e.g. interrupt processing, scheduling the client process, etc.) may result in loss of synchronization. The human ear is sensitive to the resulting phase differences, so we were careful to create an operational environment where such conditions are unlikely to occur. We have standardized the configuration of the clients (hardware, operating system, applications), and our network topology follows a logical design where machines in the same area are normally connected to the same ethernet segment. The only case where this is violated is where we have wireless clients next to a client connected to the wired LAN. The client-server protocol The client must select a multicast stream and then start playing the audio. The design of the protocol is such that the client never contacts the server. We achieve this by interspersing control packets between the normal data packets. The control packets contain sufficient information (sampling rate, bit precision, encoding, channels, etc.) for the client to configure its own audio hardware so that it can process the data sent by the server. The data in the ordinary packets is sent directly to the audio device via the standard interface (/dev/audio). Multimedia streaming servers make use of the RTF control protocol (RFC 3605) to obtain feedback on the quality of data distribution. The protocol consists of periodically transmitting control packets (sender and receiver reports) to all participants in the session, using the same distribution mechanism as the data packets. This feedback may be directly useful to the server for control of adaptive encoding schemes, or alternatively may be used to diagnose faults in the distribution. Sending reception feedback reports to all participants helps in analyzing whether problems are local or global. The ASR-based server transmits the audio data in its raw (uncompressed) form over a relatively benign LAN environment, where network resources are plentiful. This does not justify incurring the extra overhead that feedback entails, given that the probability of congestion, or other network problems is low in a LAN. Also, the feedback traffic imposes a burden on the server. Clients can
214
An audio stream redirector for the Ethernet Speaker
Campus-Wide Information Systems
Ishan Mandrekar, Vassilis Prevelakis and David Michael Turner
Volume 21 · Number 5 · 2004 · 211-216
join or leave the multicast groups anytime without notifying the server. The server remains oblivious of the number of clients “tuned in”, and its performance does not degrade as the number of the client increases. This makes the application very scalable and it can be used across campus or office networks.
to augment it by advertising available programs on a separate multicast channel. This information can include the name of the song being played and the multicast address of the channel. The user will then be able to select the desired audio track by touching the name of the channel on a touchsensitive screen.
The audio client We have constructed a sample client that receives packets from the network and plays them back via the local audio device. The client configures its audio hardware using the control packets that are sent periodically. In this way the client can start playing an audio stream by simply waiting for the first control packet. To differentiate between audio streams with varying parameters, the client compares the configuration parameters in every control packet to the parameters that have been previously used to set-up the audio device. If there is a mismatch, it invokes the configuration set-up routine before sending data to the device. Since the client receives raw data using the native encoding of the audio driver from the network, there is no decompression or decoding overhead incurred, thereby minimizing the load on the client. The client is a simple program that is designed to run on a variety of platforms (our current implementation runs under OpenBSD 3.4 and RedHat Linux 7.3). The low computing requirements make future ports of the client to hand-held devices a viable option. The client application runs in the ES, an embedded platform for audio playback over the LAN. The ES is a single board computer with a Pentium class CPU and 64 Mb of RAM. The computer also includes a 8 Mb flash memory as a boot medium, a PCMCIA slot for the network interface, and an audio card. The device boots a kernel that contains a RAM disk with the root partition and downloads its configuration from the network. The design of the ES borrows a lot from development carried out as part of the embedded VPN gateway (Prevelakis and Keromytis, 2002) project, which used a similar embedded design for a combined Firewall and Virtual Private Network gateway. Channel selection is currently handled by a number of pushbuttons on the front panel of the machine. The interface attempts to mimic the preset buttons on car radios. The ES monitors a number of “well known” multicast addresses for audio information and identifies the pushbuttons that are associated with active audio streams. The user presses one of these “active” buttons and the ES begins to playback the appropriate audio stream. This interface is rather limited and we plan
Related work SHOUTcast (http://www.shoutcast.com) is a MPEG Layer 3-based streaming server technology. It permits anyone to broadcast audio content from their PC to listeners across the internet or any other IP-based network. It is capable of streaming live audio as well as ondemand archived broadcasts. Listeners tune in to SHOUTcast broadcasts by using a player capable of streaming MP3 audio, e.g. Winamp for Windows, XMMS for Linux etc. Broadcasters use Winamp along with a special plug-in called SHOUTcast source to redirect Winamp’s output to the SHOUTcast server. The streaming is done by the SHOUTcast Distributed Network Audio Server (DNAS). All MP3 files inside the content folder are streamable. The server can maintain a web interface for users to selectively play its streamable content. Since the broadcasters need to use Winamp with the SHOUTcast source plug-in, they are limited to the formats supported by Winamp. Also it is not clear whether the system supports multiple concurrent Winamp server sessions on the same machine. Moreover, the server is designed to support outbound audio transitions (from the LAN to the outside) and thus does not support multicasting. Finally, the server is tied to the Windows platform. The Helix Universal Server from RealNetworks is a universal platform server with support for live and on-demand delivery of all major file formats including Real Media, Windows Media, QuickTime, MPEG4, MP3 and more. It is both scalable and bandwidth conserving as it comes integrated with a content networking system, specifically designed to provision live and ondemand content. It also includes server fail-over capabilities which route client requests to backup servers in the event of failure or unexpected outages. A similar application to the ASR is the multicast file transfer protocol (MFTP) from Star-Burst Communications. MFTP is designed to provide efficient and reliable file delivery from a single sender to multiple receivers. The concern that messages sent by clients participating in a multicast, can flood the server is mentioned in the Multicast FTP draft RFC (Miller, 1998). Under the MFTP protocol, after a
215
An audio stream redirector for the Ethernet Speaker
Campus-Wide Information Systems
Ishan Mandrekar, Vassilis Prevelakis and David Michael Turner
Volume 21 · Number 5 · 2004 · 211-216
file is multicast, clients contact the server to get missing or corrupted blocks of the file. MFTP aggregates these requests (NAKs) from each recipient, so that one NAK can represent multiple bad or dropped packets. The ASR design acknowledging that a small number of discarded packets is acceptable for audio transmissions, allows the clients to ignore bad, or lost packets. MFTP also uses a separate multicast group to announce the availability of data sets on other multicast groups. This gives the clients a chance to choose whether to participate in an MFTP transfer. This is a very interesting idea in that the client does not need to listen-in on channels that are of no interest to it. We plan to adopt this approach in the next release of our streaming audio server, for the announcement of information about the audio streams that are being transmitted via the network. In this way the user can see which programs are being multicast, rather than having to switch channels to monitor the audio transmissions. Another benefit from the use of this out-of-band catalog, is that it enables the server to suspend transmission of a particular channel, if it notices that there are no listeners. This notification may be handled through the use of the proposed MSNIP standard (Fenner et al., 2002). MSNIP allows the audio server to contact the first hop routers asking whether there are listeners on the other side. This allows the server to receive overall status information without running the risk of suffering the “NAK implosion” problem mentioned earlier. Unfortunately, we have to wait until the MSNIP appears in the software distributions running on our campus routers.
firewalls, routers etc. Finally, the large number of special purpose audio players (each compatible with a different subset of available formats), alienates users and creates support and administrative headaches. By implementing the audio streaming server as an extension to the audio driver on the system running the decoding applications, we have bypassed the compatibility issues that haunt any general-purpose audio player. Our system confines the special-purpose audio players to a few servers that multicast the audio data always using the same common format. The existence of a single internal protocol without special cases or the need for additional development to support new formats, allowed the creation of the ES an embedded device that plays audio streams received from the network. The communications protocol also allows any client to “tune” in or out of a transmission, without requiring the knowledge or co-operation of the server. The protocol also provides synchronization between clients broadcasting the same audio stream.
Conclusions Our motivation for doing this work was that we felt that existing audio players presented an inferior compromise between interoperability and reliability. In a LAN environment, there is no need for elaborate mechanisms for adapting to network problems. On the other hand the need for multiple connections to remote audio servers increases the load on the external connection points of the network and the work that has to be performed by
References Apple (1998), available at: http://developer.apple.com/techpubs/ quick-time/qtdev-docs/RM/frame-set2.htm Fenner, B., Haberman, B., Holbrook, H. and Kouvelas, I. (2002), “Multicast source notification of interest protocol (MSNIP)”, November, available at: draft-ietf-magmamsnip-01.txt Helix (2002), available at: www.helixcommunity.org/2002/intro/ client Kouvelas, I. and Hardman, V. (1997), “Overcoming workstation problems in a real-time audio tool”, paper presented at the USENIX Annual Technical Conference, January. Lowe, J.S. and Rizzio, L. (1998), “Multimedia driver support in the FreeBSD Operating System 5”, paper presented at the USENIX Annual Technical Conference, FREENIX track, June. Miller, K. (1998), “StarBurst multicast file transfer protocol (MFTP) specification”, internet draft, Internet Engineering Task Force, April. NetBSD (n.d.), “Writing a pseudo device”, NetBSD Documentation, available at: www.netbsd.org/ Documentation/ker-nel/pseudo Prevelakis, V. and Keromytis, A. (2002), Proceedings of the 3rd International Network Conference, Plymouth, July.
216
teachers. An online, moderated information exchange will also allow interaction with other researchers internationally.
Note from the publisher
Helping academic institutions and business schools perform better
Management journals and more: Emerald Management Xtra launches in 2005 For 2005, Emerald Group Publishing Limited will be launching the world’s largest collection of peerreviewed business and management journals, Emerald Management Xtra. As well as more than 100 research journals, the Emerald Management Xtra collection also features independent reviews of papers published in 300 of the world’s major English language business journals, plus a newly developed range of online support services for researchers, teaching faculty, academic authors and others. Emerald Management Xtra will be sold on subscription, primarily to libraries serving business schools and management departments, with online access to research papers and faculty services networked throughout an institution. Designed in consultation with an international working group of business academics, Emerald Management Xtra is intended to consolidate the library’s position at the heart of a university’s research and information needs. In addition to nearly 40,000 research papers published by Emerald journals over the past ten years, Emerald Management Xtra features help for aspiring authors of scholarly papers, advice on research funding, reading lists and lecture plans for academic teaching faculty, and an online management magazine for business students and
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · pp. 217-218 q Emerald Group Publishing Limited · ISSN 1065-0741
Input and advice from heads of research and deans and directors of business schools have helped focus Emerald Management Xtra on academic productivity improvement and time-saving, which is seen as particularly useful given the increasing pressures on academic time today. Building on the success of the Emerald FullText collection, available in more than 1,000 business schools and universities worldwide, Emerald Management Xtra puts the latest thinking, research and ideas at users’ fingertips. “Already, 95 of the FT Top 100 business schools in the world subscribe to Emerald journals or the Emerald Fulltext collection,” said John Peters, editorial director of Emerald. “We will find over the coming years that Emerald Management Xtra will be an essential part of any credible business school or management faculty. You can’t get all Emerald management and library information journals in full and in their current volumes from anywhere else; you can’t get the support services we offer to faculty and the academic community anywhere else. We can even help you plan your conference travel through the world’s largest independent register of academic conferences in business and management, Conference Central. We developed Emerald Management Xtra by asking academics, librarians and directors around the world what their biggest challenges were, we then put solutions in place to address them.”
Integrated search and publishing innovation An important part of the Emerald Management Xtra collection is an integrated search facility of Emerald-published papers with Emerald Management Reviews, which features nearly 300,000 independently written summary abstracts from a further 300 of the world’s top business and management journals. This allows quick and easy summary of the relevant management literature. From 2005, both Emerald journals and Emerald Management Reviews will feature Structured Abstracts. A first in management publishing, structured abstracts provide a consistent structured summary of a paper’s aims, methodology and content. Used successfully in
217
Note from the publisher
Campus-Wide Information Systems Volume 21 · Number 5 · 2004 · 217-218
many medical and scientific journals, structured abstracts allow a quicker assessment of an article’s content and relevance, thus saving researchers time and effort.
Value Emerald is already recognised as a world leader in offering value for money from its online collections. Research undertaken by a large US library purchasing consortium showed that the Emerald portfolio was “best value” of any publisher collection, based on the ratio of usage and downloads to subscription price. Unlike most other publishers, the Emerald Management Xtra portfolio is entirely subjectspecific – focusing just on business and management (whereas other publisher collections contain a diverse range of subjects in sciences, social sciences, arts and so on). This means that Emerald Management Xtra gives superb value to business schools and management departments, with no subject redundancy.
Emerald Group Publishing Limited Established for more than 30 years, Emerald Group Publishing has its headquarters in the north
of England, with sales offices in the USA, Malaysia, Australia and Japan. Emerald was founded as Management Consultants Bradford, later MCB University Press, as a spin-off from the University of Bradford business school, by faculty and alumni, and has maintained a tradition of having academics, former and current, as key decision-makers and advisers. Emerald journals are almost exclusively international in scope, with around 75 per cent of all authors coming from outside the UK. More than 100 of the Emerald Management Xtra journals collection are traditionally peerreviewed. The Emerald Management Xtra journals collection is completed by a small number of secondary digest titles and practitioner-focused magazine titles. Emerald Management Xtra includes long-established titles such as the European Journal of Marketing, Journal of Documentation, International Journal of Operations & Production Management, Management Decision, Personnel Review, Accounting, Auditing & Accountability Journal, and the Journal of Consumer Marketing. Emerald Management Xtra is available now. Contact Emerald for more information. E-mail:
[email protected] Tel: þ 44 1274 777700. Mail: Emerald Group Publishing Limited, Toller Lane, Bradford BD8 9BY, UK. Rebecca Marsh Head of Editorial
218
Call for papers Campus-Wide Information Systems ± the journal of technology on campus Publish your research papers or case studies in this international journal. Campus-Wide Information Systems provides comprehensive and independent coverage on the management, use and integration of information resources and educational technologies on campus. All articles are reviewed by the editor and at least one other subject expert in the field, and a swift and fair reviewing process is guaranteed. The journal would be pleased to receive submissions on: Innovations in teaching and learning with technology ± e.g. the changing roles of teacher and student, technology and student learning, teaching in new technology environments, interactive teaching modules, standards and assessment. New technologies ± e.g. wireless communication and instruction, voice recognition, mobile computing, advanced networking applications.
IT implementation and support ± e.g. designing and developing computer networks, technology support services, managing web-based resources, digital libraries, information repositories. Planning and administration ± e.g. developing and maintaining networks effectively and efficiently, managing a networked environment, infrastructure, course management systems, accreditation of distance learning programs, campus portals, computer security, intellectual property.
Submission Articles should be between 2,000 and 4,000 words in length. Case studies should be issue- or results-focused examples of technology applications in higher education. Please contact Managing Editor, Vicky Williams at
[email protected] in the first instance. www.emeraldinsight.com/cwis.htm
219
Literati Club Awards for Excellence Dirk H.R. Spenneman and
John S. Atkinson
Charles Sturt University, Albany, Australia
are the recipients of the Journal's Outstanding Paper Award for Excellence for their paper
``A longitudinal study of the uptake of and confidence in using e-mail among parks management students'' which appeared in Campus-Wide Information Systems, Vol. 20 No. 2, 2003 Dirk H.R. Spenneman is Associate Professor in Cultural Heritage Management at Charles Sturt University. He is the author or editor of 20 books and over 140 academic papers, mainly in the field of historic preservation and Pacific Islands history and heritage. In addition to this research focus, he has a strong interest in the use of IT for tertiary education and the implications of IT for small Pacific Island nations. John S. Atkinson is a Senior Lecturer at Charles Sturt University. His research interests are varied; however, they basically relate to how information technology impacts on areas such as education and the community. His current research areas include cluster computing, investigating the digital divide in a regional community and the effect of plagiarism in a university environment.
220
