Current Research Projects
Master's Projects
Mobility Management in Heterogeneous Wireless Networks
Francis Chai (Supervisor: Harsha Sirisena)
Different types of wireless packet networks have evolved to serve different needs: cellular wide area networks (WANs) such as GPRS and local area networks (LANs) such as 802.11. With the growing desire to be always on-line, there is a need for the interoperability and mobility of mobile terminals between heterogeneous systems, e.g. the wireless LAN at the office and a cellular WAN when on the move. The aim of this project is to develop architectures for achieving this goal. It is envisaged that extensions to Mobile IPv6 will be required to facilitate inter-system handover. Initial
emphasis will be on 802.11b/GPRS handover, and the goal will be to develop a (simple) working prototype system.
Quality of Service (QoS) over Wireless LAN
Abdulla Firag (Supervisor: Harsha Sirisena and Krys Pawlikoski)
Wireless LAN (WLAN) is rapidly growing field in computer networks today. WLAN standards such as IEEE802.11 are being accepted widely and rapidly for many different environments. However at the present time bein these WLANs support only best-effort services. Very recently, with the growing demand for multimedia applications many researchers have focused their interest in wireless networks supporting Quality of Service (QoS). This research project aims at developing a fully distributed MAC scheme that supports service differentiation and hence is suitable for transporting voice and video as well as data across WLANs.
Packet switching in Ad Hoc networks
Sun Yen-Rong (Supervised by Krys Pawlikowski and Harsha Sirisena)
An ad hoc network is a collection of wireless nodes that form a temporary network without using an existing network infrastructure or centralized administration. Routing protocols has been widely studied for ad hoc networking, but they are not ideal for communications requiring low latency. Switching is a solution used in infrastructure networks to provide time sensitive services, however rarely mention in ad hoc networking field. The goal of this research is to explore the possibilities of applying switching concept in ad hoc networking.
Dimensioning and Configuring Next Generation Cellular Networks to Support Quality Multimedia Services
Prasanna Venkatesh
(Supervised by Harsha Sirisena and Dr Alan Murch, Vodafone New Zealand)
The basic architectures of emerging next generation such as GPRS (2.5G) and WCDMA ( 3G) cellular wireless networks are being standardised. Implementation specific details such as channel allocation strategies, admission control, and quality of service provisioning are left open. Network dimensioning and configuration are impacted upon by these important issues and are the suject of this project. The research entails the analytical modelling of GSM/GPRS network using Markov modelling and matrix analysis methods.
PhD Projects
Reactive Traffic Control Mechanisms for Networks with Self-Similar Traffic
Sven Ostring (Supervisors: Harsha Sirisena and Irene Hudson)
This project has been investigating the interaction between self-similar network traffic and reactive traffic management algorithms, particularly when these algorithms are designed specifically to accommodate the stochastic structure of this type of traffic. The global thirst for information and the need to communicate in more diverse ways has driven designers of networks to build a differentiated service framework into the architecture of broadband networks that can support the quality that various applications require. Ensuring this quality necessitates a very good understanding of the nature of the traffic.
One important type of service is the transport of elastic traffic, that is traffic that can adapt to network conditions. This service is provided for traffic that can respond to feedback from the network. This avoids congestion situations, where there would be significant data loss for the connection, and utilises any available bandwidth, as the resources become free. In this work, the situation where the network is also carrying non-adaptable traffic that is self-similar is being addressed, and feedback mechanisms are being designed based on predicting the available bandwidth. These rate controllers use the long-range dependence of the traffic to improve the accuracy in predicting the explicit rates for the sources. Results show that significant improvements can be made by recognising the self-similar nature of a network's background traffic.
Distributed Cache Architecture for QoS Routing in Communication Networks
R. Mohammad (Supervisors: Krys Pawlikowski and Harsha Sirisena)
The rapid advances in Internet technologies is the driving force behind the emergence of a new generation of multimedia applications such as Internet Telephony and real- time video applications. The Quality-of-Service (QoS) requirements of these applications are one or more quantities such as bandwidth or delay. To successfully deploy these new applications, the networks should be QoS capable so that resources such as bandwidth can be reserved during the lifetime of an application. A fundamental element of a QoS-capable network is QoS routing. Current architectures for QoS routing employ complicated algorithms to compute routes that can satisfy a set of QoS constraints.
This project is developing a distributed cache architecture to reduce the route computing load caused by on-demand execution of the QoS routing algorithms. The architecture has been designed to easily scale to large networks. We have shown that this architecture increases the robustness of QoS routing in the presence of inaccurate network state information caused by long network state update intervals. This means that the overhead traffic caused by frequent distribution of the network state information is reduced, while achieving a good performance. To further utilise the advantages offered by the architecture, we are investigating several novel techniques to improve the performance of the distributed cache architecture in different ways.