An improved algorithm based on fuzzy theory for 802.11 DCF

Abstract

As the IEEE 802.11 DCF has the disadvantage of not adjusting itself to the network load and the probability distribution of slot selection is extremely uneven, an improved algorithm based on fuzzy theory is presented. With the use of local history regarding successes or collisions in packets transmission to characterize the current network condition, a proper membership function is established to reflect the relation between network load level and the possibility of packet collision and select an appropriate slot when the node will access the shared channel. The main advantage of the proposed algorithm is that there is no need of additional overheads which means simple design and low complexity. Simulation results confirm the effectiveness of the new algorithm in elevating the network performance on normalized saturation throughput, average delay, and congestion recovery.