Energy Efficient Learning Automata Based QLRACH (EELA-RACH) Access Scheme for Cellular M2M Communications

Abstract

This paper introduces an Energy Efficient Learning Automata Q-Learning Random Access Channel (EELA-RACH) Access Scheme to improve energy efficiency. The proposed EELA-RACH scheme employs a Distributed Learning Automata (DLA) technique based on Learning Automata (LA) feedback to minimise the energy consumed during updating Q-value and storing transmission history. The scheme also utilizes an adaptive duty cycle assignment to control the energy consumption of the Machine-to-Machine (M2M) devices within the cellular M2M communication cycle. The results show that the proposed EELA-RACH scheme achieves better performance compared to the Prioritized Learning Automata Q-Learning RACH (PLA-QL-RACH) and an Enhanced Learning Automata QL-RACH (ELA-QL-RACH) schemes with 9.41% and 65.72% decrease in energy consumption and increase in device lifetime, respectively.