Fairness and performance enhancement in CSMA-based random access networks

Abstract

In IEEE 802.11 wireless local area networks (WLANs) have been widely deployed in many places such as homes, schools, and other pubic places. Since the IEEE 802.11 systems operate based on the carrier sensing multiple access (CSMA)-based random access, it is important to maximize the system performance while maintaining fairness between nodes or links. In this dissertation, we examine the fairness problems caused by factors in PHY and MAC layers in CSMA-based random access networks. In the first part of dissertation, we address the fairness problem caused by physical (PHY) layer factors. We introduce the hybrid automatic repeat request (HARQ) schemes in WLAN and propose HARQ-based fairness schemes to resolve the fairness among nodes or links having different frame error rates (FERs). We find out that the proposed schemes not only provide the near-perfect fairness in error-prone channels but also improves the whole system performance. In the second half of the dissertation, we focus on the UL/DL fairness problem caused by media access control (MAC) layer factors. We propose mechanisms for estimating the number of backlogged nodes which contend for the channel access when UL and DL traffic coexists in unsaturated CSMA systems. Based on the backlog estimation, we propose optimized MAC protocols which ensure the near-perfect UL/DL fairness while improving the overall system performance.