Analyzing and Minimizing Random Access Delay for Delay-Sensitive Machine-to-Machine Communications: A New Perspective on Adaptive Persistence Control

Many machine-to-machine (M2M) applications are characterized by the requirement to support a large number of machines for reporting data back to the aggregator for further data processing. While related work has investigated various access barring and de-prioritization approaches for machine-type communications, they fail to work for delay-sensitive M2M applications, where the random access delay needs to be properly constrained or minimized. In this paper, we start from a multi-channel random access network and investigate how its performance can be optimized for minimizing the overall access delay of all machines with bursting traffic. We first propose a novel and simpler approach for analyzing the random access delay involved in such a network, and then propose an approach to minimize the random access delay through intelligent control of the persistence probability used by contending machines. Without resorting to dynamic programming, the proposed approach can effectively determine the optimal value of the persistence probability in constant computation time. Simulation results substantiate that compared to baseline approaches the proposed approach incurs significantly lower computation overheads without noticeable degradation in optimality.