Decentralized AP Selection in Large-Scale Wireless LANs Considering Multi-AP Interference

TL;DR

This paper introduces a decentralized AP selection method for large-scale WLANs that considers interference, significantly improving throughput by selecting APs based on SINR estimates rather than just signal strength.

Contribution

It proposes a novel distributed AP selection algorithm that accounts for interference at candidate APs, outperforming existing schemes in large-scale WLAN simulations.

Findings

Achieves 99% throughput gain over SSF scheme.

Attains 43% higher throughput compared to MPD algorithm.

Effective in dense STA environments, enhancing overall network performance.

Abstract

Densification of access points (APs) in wireless local area networks (WLANs) increases the interference and the contention domains of each AP due to multiple overlapped basic service sets (BSSs). Consequently, high interference from multiple co-channel BSS at the target AP impairs system performance. To improve system performance in the presence of multi-BSSs interference, we propose a decentralized AP selection scheme that takes interference at the candidate APs into account and selects AP that offers best signal-interference-plus noise ratio (SINR). In the proposed algorithm, the AP selection process is distributed at the user stations (STAs) and is based on the estimated SINR in the downlink. Estimating SINR in the downlink helps capture the effect of interference from neighboring BSSs or APs. Based on a simulated large-scale 802.11 network, the proposed scheme outperforms the strongest signal first (SSF) AP selection scheme used in current 802.11 standards as well as the mean probe delay (MPD) AP selection algorithm in [3]; it achieves 99% and 43% gains in aggregate throughput over SSF and MPD, respectively. While increasing STA densification, the proposed scheme is shown to increase aggregate network performance.