Power Allocation in Multi-user Cellular Networks With Deep Q Learning Approach

TL;DR

This paper introduces a deep Q learning-based power allocation method for multi-user cellular networks, combining offline simulation training with online fine-tuning, resulting in improved sum-rate performance and generalization over benchmarks.

Contribution

It proposes a novel two-step training framework for deep Q networks in power allocation, integrating offline simulation and online real-data fine-tuning.

Findings

DQN achieves higher average sum-rate than existing DQL methods.

The approach outperforms benchmark algorithms across various user densities.

The method demonstrates strong generalization ability in different network scenarios.

Abstract

The model-driven power allocation (PA) algorithms in the wireless cellular networks with interfering multiple-access channel (IMAC) have been investigated for decades. Nowadays, the data-driven model-free machine learning-based approaches are rapidly developed in this field, and among them the deep reinforcement learning (DRL) is proved to be of great promising potential. Different from supervised learning, the DRL takes advantages of exploration and exploitation to maximize the objective function under certain constraints. In our paper, we propose a two-step training framework. First, with the off-line learning in simulated environment, a deep Q network (DQN) is trained with deep Q learning (DQL) algorithm, which is well-designed to be in consistent with this PA issue. Second, the DQN will be further fine-tuned with real data in on-line training procedure. The simulation results show that the proposed DQN achieves the highest averaged sum-rate, comparing to the ones with present DQL training. With different user densities, our DQN outperforms benchmark algorithms and thus a good generalization ability is verified.