Reinforcement learning for Admission Control in 5G Wireless Networks

TL;DR

This paper compares traditional threshold-based admission control with a novel reinforcement learning approach using neural networks in 5G wireless networks, demonstrating RL's superior performance in complex, realistic scenarios.

Contribution

The paper introduces a reinforcement learning policy for admission control in 5G networks, outperforming traditional threshold policies in dynamic, heterogeneous environments.

Findings

RL policy reduces blocking probability more effectively.

RL outperforms threshold policies in diverse scenarios.

Reinforcement learning proves practical for real-world wireless networks.

Abstract

The key challenge in admission control in wireless networks is to strike an optimal trade-off between the blocking probability for new requests while minimizing the dropping probability of ongoing requests. We consider two approaches for solving the admission control problem: i) the typically adopted threshold policy and ii) our proposed policy relying on reinforcement learning with neural networks. Extensive simulation experiments are conducted to analyze the performance of both policies. The results show that the reinforcement learning policy outperforms the threshold-based policies in the scenario with heterogeneous time-varying arrival rates and multiple user equipment types, proving its applicability in realistic wireless network scenarios.