Optimal Backoff Tuning for Massive Access of Machine-type Devices in LTE Systems

Abstract

In this project, we aim to address a key issue, which is of great interest in Ultra-Reliable Low-Latency Communication (URLLC) for Machine-to-Machine (M2M) communications over long-term evolution (LTE) networks, that is, how to minimize the outage probability of access delay, and maximize the network throughput with a certain bound on the outage probability. Specifically, the outage probability is characterized as a function of system parameters and further minimized through optimally tuning the Access Class Barring (ACB) factor. Basing on this, we derive the outage probability-bounded maximum throughput by first finding the region of ACB factor within which the outage probability constraint for given delay bound can be guaranteed, and then deriving the outage probability-bounded maximum throughput and optimal settings of ACB factor within the given outage probability constraint. Simulation results demonstrate that with proper tuning of ACB factor, the outage probability-bounded throughput can achieve its highest value when keeping the outage probability lower than the given constraint.