A Novel Estimation-Based Backoff Algorithm in the IEEE 802.11 Based Wireless Network

Abstract

This paper proposes a new backoff algorithm to enhance both the delay and the throughput of the Distributed Coordination Function (DCF) in the IEEE 802.11 based wireless networks. The proposed algorithm, which is named as the Estimation-based Backoff Algorithm (EBA), observes the number of the idle slots during the backoff period in order to estimate the number of active nodes in the network. Especially, when the number of nodes or the amount of traffic dramatically varies, the proposed algorithm determines a more appropriate contention window based on the estimation algorithm. This paper evaluates the performance of the proposed EBA by using simulation, and it compares the EBA's performance with other backoff algorithms such as the binary exponential back-off (BEB), the exponential increase exponential decrease (EIED), the exponential increase linear decrease (EILD), the pause count backoff (PCB) and the history based adaptive backoff (HBAB). The simulation results show that the EBA outperforms the other backoff algorithms because it has better adaptability to the network load variation. By comparing the performance of the EBA to that of the BEB, which is defined in the IEEE 802.11, the EBA increases the network throughput by around 25%, and it decreases the mean packet delay by about 50% when the number of nodes is 70.