How to find an optimal threshold for SSB algorithm

Abstract

Slow Start Backoff (SSB) Algorithm is a simple, efficient, distributed, and well performing medium access control (MAC) layer protocol designed for ad-hoc wireless networks to share the medium. SSB is able to alleviate intensive collisions in high load environments, and SSB can also minimize the average channel access delay in light load environments. Compare with the traditional binary exponential backoff (BEB) algorithm, SSB provides better performance over a wider range of network condition. In this paper, we introduce an analytical model to analysis the saturation throughput of SSB algorithm, and then we use this analytical model to find out an optimal value of threshold for SSB algorithm that maximizes the network capacity based on the current channel status. Simulations are conducted to evaluate the performance of SSB protocol. The results confirm that the performance of SSB can be greatly improved by proper choice of the threshold according to the current channel status. Besides, the results also show the SSB algorithm outperforms the traditional BEB algorithm and its variants in both high load and light load environments when its threshold is optimal.