GFlowNets for Active Learning Based Resource Allocation in Next Generation Wireless Networks

TL;DR

This paper introduces a novel GFlowNet-based active learning framework for resource allocation in next-generation wireless networks, achieving significant performance improvements with fewer iterations.

Contribution

It presents a new GFlowNet-based approach for sequential resource allocation, effectively handling high-dimensional, discrete problems in wireless networks.

Findings

Achieves 20% performance gain over benchmarks

Requires less than half the acquisition rounds

Generates diverse high-quality solutions

Abstract

In this work, we consider the radio resource allocation problem in a wireless system with various integrated functionalities, such as communication, sensing and computing. We design suitable resource management techniques that can simultaneously cater to those heterogeneous requirements, and scale appropriately with the high-dimensional and discrete nature of the problem. We propose a novel active learning framework where resource allocation patterns are drawn sequentially, evaluated in the environment, and then used to iteratively update a surrogate model of the environment. Our method leverages a generative flow network (GFlowNet) to sample favorable solutions, as such models are trained to generate compositional objects proportionally to their training reward, hence providing an appropriate coverage of its modes. As such, GFlowNet generates diverse and high return resource management designs that update the surrogate model and swiftly discover suitable solutions. We provide simulation results showing that our method can allocate radio resources achieving 20% performance gains against benchmarks, while requiring less than half of the number of acquisition rounds.