Browsing Research Papers (Computer Science) by Title
-
Tucker, William D.; Keats, Derek (Telkom, 1998)[more][less]
Abstract: Grafting an ATM network onto an existing Ethernet network is a non-trivial exercise in network configuration. The process of mating ATM technology to an Ethernet network without tampering the underlying backbone and network configuration gives rise to a variety of networking issues including addressing, compatibility, security, efficiency, administrative effort, and scalability. Examples of actual situations serve to illustrate these issues. We have designed several scenarios to tackle the topological problems of an ATM graft. A dual IP approach works but suffers from a lack of scalability. We expect an edge routing solution, with subnetting, will be the most elegant and scalable. URI: http://hdl.handle.net/10566/448 Files in this item: 1
TuckerKeatsATM1998.pdf (612.2Kb) -
Julius, Elroy P.; Tucker, William D. (Telkom, 2004)[more][less]
Abstract: This research aims to find an empirical solution for guaranteeing the delivery of synchronous and asynchronous messages within a semi-synchronous IP-based communication domain. The communication infrastructure that is needed between the Deaf and hearing communities forms the application domain within which the research is situated. SoftBridging is a framework for multi-modal bridging as well as multiuser, multi-modal conversation sessions. An implementation of this concept called SoftBridge, is a communication platform that allows a hearing and Deaf person to communicate inside a single uniform space. The system is based on an asynchronous transport mechanism that makes use of various web services to do the actual data conversions such as voice to text and text to voice. Publish-subscribe systems are an emerging paradigm for building a range of distributed applications. The architecture of publish-subscribe systems make use of Message Oriented Middleware (MOM) to guarantee reliable delivery of messages within a communication domain. We will choose one of these systems and incorporate it within the overall architecture of a SoftBridge system. We will modify the existing architecture of a SoftBridge system to reliably transport synchronous as well as asynchronous data over a synchronous established session. We will use the Session Initiation Protocol (SIP) to establish a synchronous session between various users and a SoftBridge system. The system will then be used as a basis for developing a Deaf telephony application that guarantees the delivery of messages no matter the synchrony. URI: http://hdl.handle.net/10566/463 Files in this item: 1
-
Julius, Elroy P.; Tucker, William D. (Telkom, 2005)[more][less]
Abstract: This paper presents an empirical solution for guaranteeing the delivery of synchronous and asynchronous messages within a semi-synchronous IPbased communication domain. The communication infrastructure that is needed between Deaf and hearing communities forms the application domain within which the research is situated. SoftBridging is a framework for multi-modal bridging as well as multi-user, multi-modal conversation sessions. An implementation of this concept called SoftBridge for Instant Messaging Bridging Application (SIMBA), is a communication platform that allows a hearing and Deaf person to communicate inside a single uniform space. The system makes use of various web services to do the actual data conversions such as voice to text and text to voice. Publish-subscribe systems are an emerging paradigm for building a range of distributed applications. The architecture of publishsubscribe systems make use of Message Oriented Middleware (MOM) to guarantee reliable delivery of messages within a communication domain. We have incorporated a publish-subscribe system within the overall architecture of SIMBA. We have thus modified the existing architecture of SIMBA to reliably transport semi-synchronous data over a synchronous established session. We have chosen to use the Session Initiation Protocol (SIP) for the establishment of a synchronous session between various users and SIMBA. The system is currently being used as a basis for developing a Deaf telephony application that guarantees the delivery of messages no matter the synchrony. URI: http://hdl.handle.net/10566/509 Files in this item: 1
JuliusTuckerSoftBridging2005.pdf (589.9Kb) -
Adesemowo, A. Kayode; Tucker, William D. (Telkom, 2003)[more][less]
Abstract: As mobile devices and networks become more data-centric, text messaging requires an extension of co-presence capabilities for better interactive discussion between input/output handicapped handheld users and their free form desktop counterparts. Given some degree of co-presence (online presence and awareness) in text-based Instant Messaging, we hope to increase this co-presence on a handheld with fast-track feedback (FF) mechanisms. These are user-defined gestures, such as emoticons, and prioritised messages. We are building a system that facilitates gesture input within an interactive text-based chat over a wireless network running Session Initiation Protocol (SIP). The chat bridges a robust desktop environment and a small number of handhelds. This work in progress looks into the SIP Instant Messaging and Presence Leveraging Extensions as a platform to provide fast-track interactive and co-presence feedback. The aim is to rubberise a hitherto plastic and limited instant messaging conversation. We shall measure co-presence levels between FF and standard IM environments in and without desktop and handhelds by engaging two groups of users in chats session. We expect to demonstrate that there exists high co-presence notion in open (work by Sun Research Awarenex Group and AT&T Hubbubme in similar domain were in closed systems) FF text messaging open environment as we would like to show that this higher presence brings a much more enhanced interactive discourse. The envisaged results are intended to extrapolate to online aware co-presence in the up and coming smart phones on 3G packet networks. URI: http://hdl.handle.net/10566/539 Files in this item: 1
AdesemowoTuckerMobileDevices2003.pdf (180.9Kb) -
Adesemowo, A. Kayode; Tucker, William D. (Telkom, 2004)[more][less]
Abstract: A text only Instant Messaging (IM) built on the IETF open standard SIP/SIMPLE has been developed in line with our proposed introduction of a user-defined text Hotkey feature. These act as an on-click Affective Gesture (AG): in similitude to Face-to-Face (F2F) expressive gesture-like abilities. Given that text communication possesses expressive discourse with some presence level, we seek to show that one-click text-gesture fast-tracking enhances text communication further. For this study, we are taking a hybrid quantitative and qualitative approach. Initial Pre-trial results have shown that an AG approach is more likely to improve IM chat spontaneity/response rate. Further experimental trials are being undertaken. Mobile devices and networks are becoming more data-centric (evident in Japanese I-mode) even as mobile network voice Average Revenue Per User (ARPU) are declining, new stream of data services are required which must take cognisance of handhelds features albeit their small screen estate and input/output limitation. Given that IM is entrenched in the social space, especially among teenagers and gaining wide adoption in the business place, we believe extensions are required for IM steep uptake in the mobile world, much as SMS has gained prominence. Enhanced input mechanisms for handheld IM system are expected to increase co-presence between handheld users and their desktop-based counterparts while in a synchronous discussion. URI: http://hdl.handle.net/10566/548 Files in this item: 1
AdesemowoTuckerInstantMessaging2004.pdf (165.2Kb) -
Adesemowo, A. Kayode; Tucker, William D. (SAICSIT in association with ACM, 2005)[more][less]
Abstract: Text communication can be perceived as lacking in chat spontaneity, or plastic, due to medium limitations during interaction. A form of text messaging, Instant Messaging (IM), is now on the uptake, even on mobile handhelds. This paper presents results of using affective gesture to rubberise IM chat in order to improve synchronous communication spontaneity. The experimental design makes use of a text-only IM tool, running on handhelds, built with the Session Initiation Protocol (SIP) and the SIP Instant Messaging and Presence Leveraging Extensions (SIMPLE). The tool was developed with a novel user-defined hotkey – a one-click context menu that fast-tracks the creation and transmission of text-gestures and emoticons. A hybrid quantitative and qualitative approach was taken in order to enable data triangulation. Data collected from user trials affirms that the affective gesture hotkey facility improves chat responsiveness, thus enhancing chat spontaneity. URI: http://hdl.handle.net/10566/464 Files in this item: 1
AdesemowoTuckerInstantMessaging2005.pdf (364.4Kb) -
Naidoo, Vaughn; Tucker, William D. (Telkom, 2000)[more][less]
Abstract: With QoS available on IP-dominant NGNs, Policy-based Network Management (PBNM) is an effective mechanism for managing services as opposed to actual devices. IP is ubiquitous, and any NGN must contend with legacy devices that do not support emerging PBNM protocols. This Work in Progress uses the common Internet notion of a proxy to enforce policies on legacy equipment. We use VoIP as a mechanism to test the validity of our intended solution. URI: http://hdl.handle.net/10566/449 Files in this item: 1
NaidooTuckerVoIP2000.pdf (53.16Kb) -
Kobo, Hlabishi Isaac; Tucker, William D.; Norman, Michael (Telkom, 2010)[more][less]
Abstract: This paper discusses a way of offering voice instant messaging based on Internet Protocol using Session Initiation Protocol. The purpose of this investigation is to enhance the modern social communication amongst the people of South Africa who are already accustomed to text-based instant messaging. The proposed application aims to implement the traditional Push-to-Talk technology using Internet Protocol. Thus the proposed IP-based Push-to-Talk is a new approach to voice communication which emulates a walkie-talkie system. On the mobile phone IP-Push-to-Talk herein referred to as Push-to-Talk over a cell phone can be viewed as a voice SMS. The adoption of a Push-to-Talk service was inspired by the fact that it applies half-duplex communication. This enhances the primary objective of offering a cheap voice instant messaging. In half-duplex communication, only one person can talk at a time, thereby avoiding bidirectional charging. The project was implemented on two platforms, a PC and a mobile phone. The PC Push-to-Talk was implemented through client server approach whilst the mobile Push-to-Talk through a peer-to-peer approach. Several software engineering strategies were used for user requirements gathering as well for testing. Six users participated in the test and the results were gathered through questionnaires. The results showed that, half-duplex communication is efficient and yet very economical as it makes less usage of system resources. URI: http://hdl.handle.net/10566/623 Files in this item: 1
KoboNormanTucker-SATNAC2010.pdf (532.1Kb) -
Jeffries, Michael; Tucker, William D. (Telkom, 2001)[more][less]
Abstract: The Session Initiation Protocol and H.323 dominate the Voice over IP world. These signaling protocols are responsible for call setup and call tear down by IP telephony solutions. This project analyses and discusses the underlying H.323 and SIP interoperability issues and intends to build a scalable solution for the interoperability of these protocols. The gateway maps the call sequences between the two protocols. A test scenario establishes a barrage of call setup and tear down requests to test the scalability of any H.323/SIP interoperability solution. URI: http://hdl.handle.net/10566/465 Files in this item: 1
JeffriesTuckerSIP2001.pdf (65.32Kb) -
Jeffries, Michael; Tucker, William D. (Telkom, 2000)[more][less]
Abstract: Current VoIP signaling protocols are interoperable with the PSTN SS7 via H.323-to-SS7 or SIP-to-SS7 gateways. As we move toward the Next Generation Network, the PSTN falls away and we work toward carrier-grade VoIP interoperability between H.323, SIP, and other future VoIP signaling protocols. This paper present a Work in Progress to design and implement a solution based on the Internet Protocol itself. URI: http://hdl.handle.net/10566/425 Files in this item: 1
-
Vuza, Xolisa; Tucker, William D. (Telkom, 2004)[more][less]
Abstract: We intend adding video messaging and conferencing to Multimodal Telemedicine Intecommunicator (MuTI), a previous system that has already been in place in the target environment. This is to allow a semi-synchronous communication to occur over the Internet Protocol. Videoconferencing can be used to communicate synchronously and video messaging in a store and forward fashion can be used to communicate asynchronously. MuTI supports store and forward of still images, voicemail and text. The system also supports real time communication by means of audio. The aim of this research is to learn how to bridge the digital divide by building applications that are useful and relevant to the users we build for. We are developing this application together with the users of the application in an iterative fashion where we build the prototypes and show the users at each cycle. The feedback that we get from the users plays an important role in building a prototype for the next cycle. The software will be instrumented to provide statistics about the system usage. We are also going to get data from the users by using questionnaires and compare this data with the statistics about software usage to see how much the system has been used and which features in the system do the users like. We are using a rural South African context in conducting this research. We hope to bring about guidelines and recommendations on how to develop applications to bridge the digital divide in a developing world context. URI: http://hdl.handle.net/10566/435 Files in this item: 1
-
Yi, Long; Connan, James (Association for Computing Machinery, 2007)[more][less]
Abstract: Self-tuning has been an elusive goal for operating systems and is becoming a pressing issue for modern operating systems. Well-trained system administrators are able to tune an operating system to achieve better system performance for a specific system class. Unfortunately, the system class can change when the running applications change. Our model for self-tuning operating system is based on a monitor-classify-adjust loop. The idea of this loop is to continuously monitor certain performance metrics, and whenever these change, the system determines the new system class and dynamically adjusts tuning parameters for this new class. This paper describes KernTune, a prototype tool that identifies the system class and improves system performance automatically. A key aspect of KernTune is the notion of Artificial Intelligence (AI) oriented performance tuning. It uses a support vector machine (SVM) to identify the system class, and tunes the operating system for that specific system class. This paper presents design and implementation details for KernTune. It shows how KernTune identifies a system class and tunes the operating system for improved performance. URI: http://hdl.handle.net/10566/53 Files in this item: 1
Yi_KernTune(2007).pdf (262.3Kb) -
Yi, Long; Tucker, William D. (Telkom, 2008)[more][less]
Abstract: This paper describes Kiara, an open source SIPbased communication system that provides the building blocks to enable Deaf relay services. We have implemented a prototype that provides real-time text, voice and video to a variety of end user devices over a variety of networks. The work-in-progress concerns the addition of relay services for the Deaf. URI: http://hdl.handle.net/10566/523 Files in this item: 1
YiTuckerKiara2008.pdf (173.6Kb) -
Tucker, William D.; Blake, Edwin H.; Marsden, Gary (Telkom, 2004)[more][less]
Abstract: We looked for and found two situations within the South African Digital Divide where constant and severe macro-scale latencies would most likely interfere with the usage of Information and Communication Technology (ICT) solutions. We describe how these systems tend to exhibit both social and technical macro-scale delays. Our experience building bridges for these communities has inspired several innovations with respect to the design, development and measurement of IP communications systems. The main contribution is the Open User Interconnect (OUI) stack, a communications system model that explicitly includes the user in relation to the underlying network. The OUI stack is orthogonal to the Open System Interconnect stack. There are two outcomes from including the user in the model. First, the OUI approach necessitates the adoption of usercentred methods such as Participatory Design and in the case of the South African Digital Divide, the adoption of Action Research principles. The second outcome is a user-centred measurement notion, called Quality of Communication (QoC), that is a macro-scale spin on Quality of Service. QoC explicitly includes the measurement of user behaviour in addition to system metrics. In conclusion, we suggest that these Digital Divide-inspired contributions may be applicable to IP communications system design, development and measurement in general. The paper concludes with a brief dissertation completion plan. URI: http://hdl.handle.net/10566/485 Files in this item: 1
TuckerDigitalDivide2004.pdf (131.4Kb) -
Naidoo, Vaughn; Tucker, William D. (Telkom, 2001)[more][less]
Abstract: Next Generation Networks use policy-based network management and QoS protocols to provide voice services on IP networks. Large numbers of older equipment still push IP packets. We present several solutions that include legacy equipment into a PBNM and QoS system. A static design uses traffic shaping at the legacy edge. A dynamic solution uses a Quasi-PEP implemented over RSVP and COPS. Preliminary test results are presented for the static solution. Collected data shows that the solution decreases packet loss on the legacy side. URI: http://hdl.handle.net/10566/519 Files in this item: 1
NaidooTuckerNextGenNetworks2001.pdf (59.43Kb) -
Glaser, Meryl; Tucker, William D.; Mashao, Daniel (Telkom, 2004)[more][less]
Abstract: Following on the development of several prototypes, we have built a semi-automated Deaf Telephony prototype on the SoftBridge platform. This prototype relays text and speech between Deaf users on the Internet and hearing users on the telephone system. Previous work with a pilot trial in the laboratory revealed several opportunities for enhancement. We added a Wizard of Oz (WoOz) to replace the poorly performing automatic speech recognition functionality as well as H.323 breakout, more extensive logging and advanced call initiation functionality. In order to trial the current prototype, we initiated an Information and Communication Technology (ICT) training programme with the Deaf Community of Cape Town. Twenty Deaf users participated in the training. In addition to the training, much baseline user data was collected to give an indication of how Deaf users communicate with hearing users as well as how familiar they are with ICT devices and services. The work for the rest of this year requires us to recruit and train a WoOz operator. Subsequent trials will essentially consist of monthly cycles of prototype introduction, training, automated metric and log collection, user feedback and then feature enhancement. Linguistic analyses of the text output of the Deaf users will be analyzed. We hope to refine the SoftBridge prototype to fit the needs of the Deaf and hearing users, from both technical and social viewpoints. We expect that these iterative cycles will continue for some time and will teach us many lessons regarding multi-modal semi-synchronous IP-based communications systems. URI: http://hdl.handle.net/10566/447 Files in this item: 1
GlaserTuckerMashaoDeafTelephony2004.pdf (133.0Kb) -
Segers, Vaughn; Connan, James (2009)[more][less]
Abstract: This paper discusses an implementation for gesture recognition using eigenvectors under controlled conditions. This application of eigenvector recognition is trained on a set of defined hand images. Training images are processed using eigen techniques from the OpenCV image processing library. Test images are then compared in real-time. These techniques are outlined below. URI: http://hdl.handle.net/10566/63 Files in this item: 1
Segers_Real2009.pdf (179.9Kb) -
Tucker, William D.; Blake, Edwin H.; Marsden, Gary; Pearson, Murray; Westerveld, Rudi (Telkom, 2007)[more][less]
Abstract: This paper reports on three years of research and fieldwork with a rural wireless Internet Protocol communication project. We built a long-range WiFi network and custom communication software to support a rural telehealth project in the remote Eastern Cape. We report on our work using cellular networks, devices and applications as reference technologies because our users and beneficiaries are very comfortable with them. Of most concern are the technological and contextual issues affecting take up of the systems we designed. The paper intends to provide a summary analysis of our experience so that others in the field can learn from our successes and mistakes with respect to rural Information and Communication Technology for Development (ICT4D) in a South African context. URI: http://hdl.handle.net/10566/487 Files in this item: 1
TuckerIP-Communication2007.pdf (1.427Mb) -
Daniels, Francois; Tucker, William D. (Telkom, 2008)[more][less]
Abstract: This paper presents a fully decentralised peer-topeer voice communication tool intended for use across mobile ad hoc networks (MANET) by distributed groups who desired collaboration. We examined the synergy between MANETs and peer-to-peer virtual overlay networks which allowed the creation of ad hoc applications. One style of communication considered suitable for task oriented distributed group collaboration was push-to-talk. This research was focused on providing a push-to-talk communication platform suitable for deployment across MANETs. The research methodology employed was a proof of concept approach within a classical experimental computer science paradigm. We developed a prototype which used JXTA, a peer-to-peer virtual overlay network, to provide push-to-talk functionality across MANETs. Guaranteed delivery of messages was provided via a peer-to-peer voicemail delivery system. While the system did what intended we show that JXTA had a problem with the efficient delivery of voice samples. URI: http://hdl.handle.net/10566/508 Files in this item: 1
DanielsTuckerMANET2008.pdf (216.0Kb) -
Tucker, William D.; Blake, Edwin H. (International Information Management Corporation, 2008)[more][less]
Abstract: We describe the use of Outcome Mapping to guide the design of a rural telemedicine consultation system in South Africa. While Outcome Mapping was not primarily intended to guide design, we show that it tied in well with a cyclical participatory design method for an Information and Communication Technology for Development project. URI: http://hdl.handle.net/10566/525 Files in this item: 1