Dynamic contention window selection scheme to achieve a theoretical throughput limit in wireless networks: a fuzzy reasoning approach

In WLANs, the IEEE 802.11 MAC protocol has been the main element that determines the efficiency of sharing the limited communication bandwidth. The performance of the protocol could be improved by using an appropriate tuning of the backoff algorithm to approach a theoretical throughput limit. However, in real cases, a mobile node would not exactly know the network and load configuration information (e.g., number of active nodes). Hence, in this paper, we propose a novel and efficient selection of backoff window size mechanism using a fuzzy reasoning approach, named the fuzzy backoff controller (FBC), to achieve the system throughput limit. This mechanism is developed on the following ideas: by observing the busy degree of the medium and the number of neighbors, the proposed FBC could generate a proper backoff window thus reduce the probability of collisions and enhance the throughput. The obtained simulation results show that the proposed FBC not only significantly improves the performance of the protocol but also achieves good fairness of accessing the medium between nodes.