Adaptive Access Mechanism with Optimal Contention Window Based on Node Number Estimation Using Multiple Thresholds

How to improve the short-term fairness and increase the aggregated throughput simultaneously is a major problem in WLANs, yet has not been solved satisfactorily in previous work, especially in the situation that network loads vary significantly. To improve the performance in saturation state, we propose a novel access mechanism that sets an optimal Contention Window (CW) for all nodes in the network. By introducing a new parameter of CW Index (CWI), we present a linear CW adjustment rule based on the active node number and then develop an estimation algorithm of node number with three thresholds. Compared to previous algorithms, the new algorithm predicts the future node number based on the current network status, thus minimizes the fluctuating effect of idle slot intervals observed at each process. Moreover, it can identify the dense degree of idle slot intervals and track the changes of the accurate node number quickly. By selecting the optimal CWI adaptively, the maximum aggregated throughput keeps basically a constant despite the variation of the number of nodes. Meanwhile, by eliminating multiple transmission attempts of the same node, the short-term fairness is improved significantly. To evaluate the mechanism, we derive closed-form expressions for the network performance. The simulation results demonstrate the validity and good scalability of the proposed access mechanism.