AI improves communication and computing for virtual reality devices
Uncertainty-Aware Federated Learning for Joint Communication and Computation in XR Systems
Information TheoryMachine Learning
Summary
Extended reality (XR) devices need both internet connection and computing power to work well, but uncertainties in network and computer performance can cause delays. The authors created a way to use AI to manage wireless signals and computing resources smarter, even when there is uncertainty in information. They use a special method called federated learning so that multiple base stations can work together without sharing all their data. This method helps reduce delays for XR devices while keeping user data private and systems adaptable.
What this means in practice
- •For network schedulers: Optimize wireless and computing resources jointly in XR systems to reduce worst-case delays with uncertain channel and computing states.
- •For cloud computing operators: Improve distributed computation allocation for latency-sensitive XR services by integrating uncertainty-aware learning without central data pooling.
Authors
Robert-Jeron Reifert, Hayssam Dahrouj, Alaa Alameer Ahmad, Haris Gacanin, Aydin Sezgin
Abstract
The convergence of communication and computation, empowered by artificial intelligence (AI) enhanced capabilities, emerges as a key enabler for realizing the promising functionalities of the sixth-generation of communication systems (6G). To this end, this paper investigates an AI-native, robust network, where multiple base stations (BSs) simultaneously serve several extended reality (XR) devices. The paper then presents an innovative distributed deep learning-based approach to simultaneously manage the transmit and computing powers, and computation allocation, amidst uncertainties in both channel and computing states information. Under such an interference network scenario, the paper proposes a robust approach for minimizing the worst-case delay across the served XR devices, subject to computation and power constraints. The paper leverages a deep neural network (DNN) that can be trained either centrally or in a federated manner to map estimated channels and computation requirements to optimized resource allocations per BS. During training, uncertainty samples, reflecting both channel and computing estimation errors, are injected after the DNN output to account for such joint uncertainties. Our results validate the enhanced robust delay performance and scalability of joint uncertainty injection as compared to uncertainty-agnostic methods and state-of-the-art benchmarks, with federated learning-based solutions achieving near-centralized performance while enhancing efficiency, privacy, and adaptability.