| Summary: | 碩士 === 國立臺灣師範大學 === 應用電子科技學系 === 99 === The IEEE 802.16e a.k.a. mobile WiMAX (Worldwide Interoperability for Microwave Access) network has been implemented as one of the options for 4th generation mobile communication system. According to the criteria of IMT-A, the 4G communication in the future must achieve: high-speed uplink / downlink transmission, better bandwidth utilization, better transmission capacity, and higher mobility. The mobile WiMAX network technology can provide above requirements, thus it comes to be a major competitor in the 4G network technology standard. Follow the technique specification of the IEEE 802.16e, real and non-real time traffics are classified by different transmission parameter to enhance the resource efficiency and satisfaction of clients. To implement QoS (Quality of Service) in the packet switch transmission, the MAC layer of IEEE 802.16e could distinguish the type of service flow and operate the CAC (connection admission control) in the base station (BS), which could be used to estimate the capacity of the whole system. By the way, the BS would apply an accurate QoS for admitted connections. The IEEE 802.16e-based network system has adopted various ways to execute the demands of different service flow. It is provided with a traditional high- and low-priority methods: unsolicited granted service (UGS) and BE (Best Effort). Beside that there are more complex methods such as polling services: ertPS (Extended Real-Time Polling Service), rtPS (Real-Time Polling Service), nrtPS (Non Real-Time Polling Service) for diversity.
In these days, the studies on the collation of service are between new incoming and existing users, but lack for the description of handoff users. The current method applied bandwidth reservation for handoff connection in the radio coverage. The advantages of this way are simple and practical. On the other hand, this is an inflexible method for dynamic network scenario. The total efficiency of the system maybe over or under estimated, no matter which condition happened it would harm the performance of the whole network. In this research, we analyzed the associations among handoff, real and non-real time connections. After that we derived some integrated variables from bandwidth requests for diverse handoff connections. Besides that we proposed a QoS-aware utility function for handoff users. A QoS-aware utility function can help the handoff user to indicate its requirements accurately. At last, we build up a new algorithm for resource allocation by the Nash bargaining solution (NBS). The proposed algorithm estimates the load of the system by the minimum utility rather than the minimum bandwidth. The simulation results show that the algorithm could fully use the system by pulling up the system capacity.
|
|---|
