Throughput Measurement-Based Access Point Selection for Multi-rate Wireless LANs

Abstract

In this paper, we propose a novel access point (AP) selection algorithm to maximize the system throughput while considering user fairness. The main idea is that when a new-coming user enters an overlapping area of a wireless local area network (WLAN), it first estimates the system throughput as if it were associated with each of the APs involved. Then it chooses the AP that can achieve the highest system throughput. For existing users that locate in the overlapping area, they may also need to change the AP association due to the dynamic nature of traffic load. To enable the fairness among users, each user is guaranteed the minimum transmission opportunity. Another significant contribution of this paper is that we find that load-balancing based approaches could not achieve the maximum throughput for multi-rate WLANs, although load-balancing has been considered as an effective approach to improve the network throughput for single-rate WLANs. In-depth theoretical analysis and extensive simulations are performed to verify the throughput optimization and user fairness.