An Optimal Distributed Algorithm for Best AP Selection and Load Balancing in WiFi

The ubiquitous use of IEEE 802.11 has aggravated the need to make efficient use of available bandwidth. Currently handoff decision in IEEE 802.11 is made based on the received signal strength but these results in poor connectivity specifically when an access point is overloaded. Overlapping regions where users can be connected to multiple access points, switching to less loaded access point can improve overall network capacity. In this article, we propose a decentralized approach for best access point selection which also prevents an access point to get overloaded. We propose an algorithm for handover strategy to improve network capacity via load balancing and it also minimizes switching overhead. We perform detail analysis on publically available dataset which consists of millions of Wi-Fi sessions with multiple access points.