A Blind Adaptive Tuning Algorithm for Reliable and Energy-Efficient Communication in IEEE 802.15.4 Networks

Abstract

Previous studies have shown that the IEEE 802.15.4 CSMA/CA with default parameter values suffers from a severe unreliability problem, which is undesirable in many wireless sensor network applications. For time-varying and unknown operating conditions, how to autonomously adjust parameters for reliability guarantees with low energy consumption is a significant challenge, especially when the acknowledgement (ACK) mechanism is disabled in order to reduce extra overhead and waiting time in control or alarm applications. In this paper, we propose a lightweight distributed algorithm called Blind ADaptive Access Parameter Tuning (BADAPT) for reliable and energy-efficient communication in IEEE 802.15.4 networks without using the ACK mechanism. Unlike previous adaptive algorithms that require a costly ACK to notify a sensor node whether a packet is successfully transmitted or not, the BADAPT allows sensor nodes to merely use the locally measured busy channel probabilities to estimate individual reliability under unknown operating conditions. Simulation results show that the BADAPT is effective both in stationary and dynamic scenarios.