Lists
Lists
El Maameri, Said David Soud (2004) A fuzzy logic based approach to quality of service in 802.11b wireless networks. Doctoral thesis, London Metropolitan University.
A generally held view is that the bulk of the data transmitted across packet switched networks in the future will be video and audio traffic. In recent years public telecommunications carriers have replaced circuit switched networks with integrated service packet networks that support both multimedia and data. One challenge with sending multimedia data is that it is delay-sensitive. To encourage the widespread use of the INTERNET as a multimedia platform it is necessary to have a minimum level of service. Guarantees are required that ensure satisfactory end to end transmission of delay-sensitive multimedia data. In wired systems the mechanisms used to provide Quality of Service (QoS) are Connection Admission Control (CAC), Usage Parameter Control (UPC) and Multi-Priority Control (MPC). An increasing number of users are connecting to the Internet through Wireless Local Area Networks (WLANs). The most widespread standard for WLAN s is that defined in the Institute of Electrical and Electronic Engineers (IEEE) 802.11b. Wireless hotspots and local area networks using equipment working to this standard are now ubiquitous. Regrettably the 802.11b standard has no mechanisms for providing QoS across wireless channels. Users of 802.11b networks cannot rely on the same support for QoS that would be available in a fixed network.
This thesis proposes the redesign of 802.11b gateways using fuzzy logic to include mechanisms that deliver Quality of Service. A fuzzy controller has been developed that implements QoS with CAC, UPC and MPC. This controller was optimised for 802.11b WLANs. Fuzzy logic has been used as a control mechanism for each QoS component. It converts imprecise human observations to crisp outputs for computer processing. With fuzzy logic it is possible to enhance performance at various points in the packet transmission process. Fuzzy feedback control regulates admissions to the network (CAC) preventing a link from becoming saturated. Once an admission is made, a fuzzy controller smoothes individual flows (UPC). Traffic flows are classified and prioritised by class and then sent to the designated queue for that given class (MPC). A fuzzy packet scheduler regulates packet transfer through each wireless gateway. In this research, experiments have been carried out in real-time with CAC, UPC and MPC using fuzzy controllers for multimedia transmission over 802.11b. The results have been successful, with faster response times and throughput despite having lower processing requirements.
 |
View Item |