Improvement of WLAN Contention Resolution by Loss Differentiation

In a realistic WLAN environment, frame losses may be caused by collisions or channel noise. The existence of noise-induced losses reduces the effectiveness of the standard WLAN backoff algorithm for contention resolution, which assumes that all losses are caused by collisions and always doubles the contention window to reduce contention upon a frame loss. In this paper, we propose new backoff algorithms that take advantage of a new capability to differentiate the losses, and thereby sharpen the accuracy of the contention resolution process. Analytical models are developed to analyze the performance of these algorithms under heterogeneous link conditions in a WLAN. Both analysis and simulation results show that significant improvement of throughput and fairness can be obtained for WLANs in which contention resolution is. enhanced by the loss differentiation ability