Cross-layer Protocol Based On The Multiple Transmit Power Of The Wireless Network Key Technology Research

Power control is a cross-layer problem which can affect the connectivity, throughput, energy consumption and network life and many other network performance indicators. Therefore, in this thesis, we research into the power control and link scheduling for wireless networks. The main contributions of this thesis are as follows:1. As sending nodes which use different power to transmit data have different conflict range, they affect different throughput of their neighbors. Using the characteristic, we present a gain evaluating based power allocation algorithm. The scheme first estimates transmit and loss data of each node under different power using probability method. Then, it summarizes power allocation problem as active node selection problem and uses dynamic programming strategy to get the power assign scheme. Finally, considering scheduling fairness among nodes and links, we get the final power allocation and link scheduling scheme. Theoretical and experimental results show that our proposed method can effectively improve the network throughput and minimize network latency.2. To take full advantage of multi-channel resources and improve the network throughput, we study protocol conflict model based power selection problem in multi-channel wireless networks. First, we present the minimum number of needed channels which guarantees that all send-receive node pairs can transmit data in parallel under minimum power level. Then, we analyze the throughput upper bound using the same power level and hybrid power levels under certain number of radios and channels. Finally, using integer linear programming model we get the power allocation algorithm in single-radio multi-channel wireless networks.3. Considering the characteristic that signals cumulate under actual physical environment, this paper presents a physical conflict model based power allocation and link scheduling algorithm in multi-radio multi-channel wireless networks. The method first obtains concurrent link set of each channel using hexagonal graph coloring theory. and then computes the minimum power vector using the knowledge of linear algebra which satisfies SINR limit. Experimental results show that in the networks whose link length is symmetrical, the proposed algorithm can improves the network spatial reuse ratio and effectively increases the network throughput.