Improving Goodput of the IEEE 802.11 MAC Protocol by Using Congestion Control Methods

Abstract

Fragmentation is proposed for mitigating the bad affects resulted from intermittent interferences and noises. However, the current fragmentation algorithm used in the IEEE 802.11 protocol only supports fixed fragment size. Since the wireless channel and the network load are time-varying, this fixed algorithm is obviously not optimal for the goodput performance. In this paper, we present and analyze two simple yet effective adaptive fragmentation algorithms for goodput enhancement. First, we study the analogies between the TCP protocol and MAC protocols in terms of their resource control methods. Next, we propose an adaptive approach based on the heuristics of congestion control schemes by dynamically varying the fragment size according to changes of wireless channel quality. Simulations have been conducted by using real channel measurements in a typical indoor environment, and results show that the proposed algorithms outperform the current fixed fragment scheme and can significantly improve the goodput performance up to seven times more than the fixed fragment scheme. Moreover, the proposed adaptive fragmentation algorithms introduce insignificant overhead and the implementation is very practical.