Deep reinforcement learning cuts delay for multi-user wireless updates
DRL-Based AoI Minimization for RSMA in Finite-Blocklength MU-MISO
Information Theory
Summary
Wireless networks need to send fresh information quickly, especially for short updates in devices like sensors. The authors look at a complex system called RSMA that helps manage wireless signals to many users but is hard to optimize by hand. They designed a smart method using deep reinforcement learning to automatically figure out the best way to send data so information stays fresh. Their approach works better than existing methods, especially when signal quality is low and data packets are short, and it is faster to run once trained.
What this means in practice
- •For wireless network engineers: Optimize resource allocation for multi-antenna wireless systems to minimize data freshness delay in low-latency updates.
- •For industrial iot system designers: Implement learned scheduling and power control policies to improve timely state updates from multiple devices with limited transmission time.
Authors
Enes Kaya, Mehmet Alp Demircioğlu, Elif Tugce Ceran, Melda Yuksel
Abstract
This letter investigates age-of-information (AoI) minimization in multi-user wireless networks operating in the finite-blocklength (FBL) regime, which is critical for low-latency transmission of short state-update packets. While rate-splitting multiple access (RSMA) provides a powerful and flexible framework for interference management in multi-user FBL systems, the joint optimization of its parameters, such as precoding vectors, power allocation, and rate-splitting ratios, to guarantee information freshness results in analytically intractable complexity. To address this challenge, we propose an actor--critic deep reinforcement learning (DRL) framework to learn dynamic resource-allocation policies in multi-user multiple-input single-output (MU-MISO) broadcast channels. Simulation results show that the proposed RSMA-RL framework achieves consistently lower AoI than the state-of-the-art benchmarks, with substantial gains observed at low signal-to-noise ratio (SNR) and short blocklengths, while matching benchmark performance at high SNR with significantly lower online complexity via a single neural-network forward pass at execution.