A Fast Collision Detection Algorithm in IEEE 802.11 through Physical Layer SINR Monitoring

Adaptive transmission in IEEE 802.11 requires rapid rate adaptation according to the variation of wireless environments. Existing rate adaptation schemes such as Automatic Rate Fallback (ARF) have been commercially implemented due to its simplicity. However, collision detection, which is an essential requirement to avoid network congestion and maintain the quality of service (QoS), is not well investigated in these traditional IEEE 802.11 link adaptation schemes. In this paper, we propose a fast collision detection scheme based on tracking the changes of the signal to interference and noise ratio (SINR). With the help of a nonparametric order-based cumulative sum (CUSUM) algorithm, collisions can be detected within a delay of a few OFDM symbols. Simulation results show the effectiveness of our proposed scheme and demonstrate its flexible implementation in existing IEEE 802.11 systems to assist current rate adaptation.