Optimal Wireless Resource Allocation with Random Edge Graph Neural Networks

TL;DR

This paper introduces the random edge graph neural network (REGNN) for optimal wireless resource allocation, leveraging permutation-equivariant convolutional architectures trained via unsupervised primal-dual learning, demonstrating superior performance and transferability.

Contribution

The paper proposes the REGNN architecture for resource allocation in wireless networks, enabling model-free, permutation-equivariant learning adaptable to different network sizes.

Findings

REGNN outperforms heuristic benchmarks in simulations.

REGNN maintains permutation equivariance for transferability.

Unsupervised primal-dual training effectively learns resource policies.

Abstract

We consider the problem of optimally allocating resources across a set of transmitters and receivers in a wireless network. The resulting optimization problem takes the form of constrained statistical learning, in which solutions can be found in a model-free manner by parameterizing the resource allocation policy. Convolutional neural networks architectures are an attractive option for parameterization, as their dimensionality is small and does not scale with network size. We introduce the random edge graph neural network (REGNN), which performs convolutions over random graphs formed by the fading interference patterns in the wireless network. The REGNN-based allocation policies are shown to retain an important permutation equivariance property that makes them amenable to transference to different networks. We further present an unsupervised model-free primal-dual learning algorithm to train the weights of the REGNN. Through numerical simulations, we demonstrate the strong performance REGNNs obtain relative to heuristic benchmarks and their transference capabilities.