Research On IEEE802.15.6MAC Protocol And Spectrum Allocation

With the development of wireless communication technology and the appearance ofwireless portable devices, wireless networks have been the focus. Because of the rapidexpansion of the global aging populationand the scarce health care resources, WirelessBody Area Networks (WBANs) employed in the medical field abstract more and moreattention in society. What’s more, the limited spectrum sources can’t satisfy the need ofthe increasing wireless network traffic. Cognitive Radio technology provides analternative scheme for the spectrum sources issues and has promoted the developmentof spectrum resource management and allocation. Therefore, it is significant to analyzethe WBAN performance and to study the spectrum allocation in the cognitive radionetwork. The main work includes two parts, the contributions are as following:First, a three dimensional discrete Mrakov model and a backoff model are used toanalyze the channel access behavior of the nodes for the MAC protocol of IEEE802.15.6. In saturation condition, we take into account the parameters of the nodes’number, the length of frames, the different informations rate and the user priority, anddrive the mathematical expressions of the normalized throughput and the average accessdelay. By comparing with IEEE802.15.4, the simulation results prove that IEEE802.15.6can provide a reliable and timely delivery for high user priority and satisfy therequirements of medical application.Further, when considering the flexible spectrum requirement of Secondary users, thedynamic spectrum allocation of the cognitive radio network is studied. A two dimensionspectrum model is proposed to satisfy the users’ flexible requirement, which is dividedby the time and frequency. Cooperating with the combinatorial auction, the dynamicallocation of idle spectrum is realized among the secondary users in the cognitive radionetwork. The dynamic spectrum allocation is based on greedy algorithm and mixedalgorithm, and the simulation results are given in the large-scale network and thesmall-scale network, respectively. Finally, we can show that the mixed algorithm canachieve a better performance with the large-scale network.