Random access for machine-to-machine communication in LTE-advanced networks: issues and approaches

TLDR

Machine‑to‑machine communication enables ubiquitous connectivity among autonomous devices, underpinning smart‑city applications such as metering, surveillance, infrastructure management, and eHealth, while the challenge of supporting massive MTC devices lies in efficient multiple access and minimizing network overload. This article reviews LTE‑Advanced M2M communication techniques, identifies major research issues, and proposes a reinforcement‑learning eNB selection algorithm for self‑organizing MTC device access. The authors analyze random‑access overload control mechanisms and develop a reinforcement‑learning algorithm that lets MTC devices autonomously choose eNBs to reduce congestion.

Abstract

Machine-to-machine communication, a promising technology for the smart city concept, enables ubiquitous connectivity between one or more autonomous devices without or with minimal human interaction. M2M communication is the key technology to support data transfer among sensors and actuators to facilitate various smart city applications (e.g., smart metering, surveillance and security, infrastructure management, city automation, and eHealth). To support massive numbers of machine type communication (MTC) devices, one of the challenging issues is to provide an efficient way for multiple access in the network and to minimize network overload. In this article, we review the M2M communication techniques in Long Term Evolution- Advanced cellular networks and outline the major research issues. Also, we review the different random access overload control mechanisms to avoid congestion caused by random channel access of MTC devices. To this end, we propose a reinforcement learning-based eNB selection algorithm that allows the MTC devices to choose the eNBs (or base stations) to transmit packets in a self-organizing fashion.