Mathematical system modeling and dynamic resource allocation through Kalman Filter based prediction in IEEE 802.11 PSM

Abstract

In the IEEE 802.11 power saving mode (PSM) which is specified for distributed coordinate function (DCF), time is divided into beacon intervals. Beacon interval is composed of ATIM window and data transmit duration. When data transmission is needed, hosts must send and receive ATIM window frames during ATIM window. According to ATIM frames, each host decides to sleep or awake. Clearly, this method is not optimized one since it can't manage dynamic transmitting situation. To solve this problem, dynamic power saving mechanism (DPSM) is proposed. But in this method changing ATIM window size is linearly progressed and considering situation is highly limited. So we have developed new dynamic traffic assignment (DTA) systems for optimizing PSM in wireless networks. For this purpose, we bring in optimal state estimation which is formalized by mathematical equations that utilize a proven error calculation method, Kalman Filter. This method makes it possible to predict the best suitable ATIM window size to increase possible doze time. And it also improves power efficiency of wireless devices more than the one achieved through DPSM. In this way, we have developed basic mathematical system modeling in wireless networks and the usage of Kalman Filter. Simulation results show that proposed approach outperforms previous mechanisms PSM, DPSM in terms of throughput and the amount of energy consumed.