Delay and Energy Improvement for CSMA Networks Using Alternating Optimization

This paper proposes a carrier sense multiple access (CSMA) algorithm for interference networks of conflict graph model. The developed algorithm employs sleeping, waking-up, back-off, and channel-holding modes and transits randomly among them. We pose an optimization problem to find the average durations of these operation modes in that queuing delay and consumed power for all nodes are minimized. We tackle the formulated problem using the alternating optimization technique and reach a stationary solution for the design parameters. The achieved solution guarantees that the average service time is sufficiently large to propel stacked packets at the node to have very small lengths. It also assures that the average awake time of the node is reduced to increase the power efficiency. The numerical results show that our CSMA scheme attains significant reduction in both queue length and energy usage.