Research Of MAC Level Power Management For Wireless Network

With the rapid development of modern wireless communication and Internet, people have become more and more dependent on mobile information services. This trend is driving the technology of wireless network, which requires low cost and little dependency on infrastructure, to become one of the hottest focuses in academic and professional research. Energy conservation is one of the most important objectives in the design of wireless network protocols and MAC power management is an effective way to achieve this goal. Focusing on the popular MAC standards and problems of power management, this thesis makes contribution as described in the following paragraphs. The thesis gives a review of popular IEEE 802 wireless network standards, comparing them in performance and prospective industrial applications. With numerical simulation, the thesis compares throughput of 802.11e and 802.15.3 over high physical rates and compares bandwidth efficiency of 802.11 and 802.15.4 over low physical rates. The related analysis and evaluations should be helpful for research of wireless network and product development.With the extensive analysis of MAC level power management protocols in the state of art, this thesis induces the universal policies for MAC level power management: less contention, less overhear, fewer state transitions, adaptive power control and multi-rate modulation, less control overhead, adaptive ARQ and FEC and channel allocation based on energy priority. At the same time, the thesis proposes an analytical model for power management.Based on 802.11PSM, the thesis points out two problems existing in power management for CSMA/CA—energy loss due to overhear and delay of response time for stations in power saving mode. For coordination functions of DCF, EDCA, PCF and HCF, the thesis proposes corresponding methods to reduce overhear, which is the greatest source of energy loss for 802.11PSM. The thesis also proposes a RTTA-PSM protocol to make a tradeoff between response time and energy consumption.With an extensive analysis of 802.15.3MAC, this thesis points out that reducing transition times is a primary problem of power management in TDMA systems. Flow schedule algorithms can be used to reduce the overall transition times in such a system. Thanks to the graph theory, the schedule problem can be changed into a Hamilton Graph problem for theoretical analysis. At the same time the lower bound and upper bound of the minimum transition times are deduced. However, since there is no optimal solution for Hamilton Graph problem, the thesis proposes two sub-optimal schedule algorithms—LSA and MDS. Simulation results prove that the proposed algorithms perform much better than existing algorithms solving similar problems and MDS can actually achieve optimal solution in most scenarios.Studied with both theoretical and practical consideration, the mechanisms and algorithms proposed in this thesis can be easily implemented within current wireless network products.