ISAC-Based UAV-Relay Network Optimization With Multiagent Q-Learning Approach

Ji Min Park

doi:10.1109/jsyst.2026.3687248

ISAC-Based UAV-Relay Network Optimization With Multiagent Q-Learning Approach

2026-05-20

0

PRE

AI

DOI：10.1109/jsyst.2026.3687248

Original

Original request for help

Save

Abstract

En 中文

In the sixth-generation wireless communication networks, uncrewed aerial vehicles (UAVs) have been considered as one of the most significant network entities to realize ubiquitous connectivity. When using UAVs operating as relays, i.e., UAV-relays, a base station (BS) tracks and controls the position of UAVs to provide connectivity to ground users efficiently. To this end, an integrated sensing and communication (ISAC) technology, which utilizes a single waveform, e.g., a packet with pilot and data parts, for both communication and radar sensing functions, can be adopted. By controlling power allocation to pilot and data parts as well as the position of UAVs, the performance of ISAC can be optimized. In this article, we propose a distributed reinforcement learning approach that optimizes both communication and radar performance in aerial networks with multiple ground BSs and UAV-relays. Through intensive simulations, it is shown that the proposed approach can achieve better performance compared to benchmark models.

Keywords:

A2G network

ISAC

optimization

Q-learning

UAV

ISAC-Based UAV-Relay Network Optimization With Multiagent Q-Learning Approach

Abstract

Journal

Organization