Transmission power control strategy based on partially observable Markov processes for IEEE802.11 WLAN

Yanpeng Feng, Hongyan Zheng, Bo Wu

Education Technology and Information Center, Liuxian Road, Shenzhen Polytechnic, Shenzhen, China

With the limitation of independent channels in IEEE802.11 WLAN, co-channel AP interferes with each other in repeat coverage area. As traditional AP launched a fixed power in sending data packages, which ignores the differences and mobility of STA. According to the mobility of STA and the partially observable feature of AP-STA link, this paper analyses the internal relations among the link state, transmission power and channel interference of co-channel AP, and proposes a single-link transmission power control (TPC) algorithm based on the Partially Observable Markov Decision Processes (POMDP). Firstly, the single AP-STA link POMDP-TPC model is constructed, and the neural network learning model is established to describe the observation function of POMDP-TPC. Secondly, the algorithm constructs reachable belief searching trees to obtain the approximate optimization, which implements the dynamic creating and on-line updating of the power consumption policy. Finally, under the experiment environment in OPNET and IEEE802.11b, the results demonstrate the algorithm can reduce the AP power consumption efficiently and improve the network throughput greatly.