XR Offloading Across Multiple Time Scales: The Roles of Power, Temperature, and Energy

2025-06-23

AI Summary

This work addresses the multi-timescale computation offloading optimization problem for XR wearable devices, subject to instantaneous power constraints, short-term thermal rise, and long-term battery degradation. We propose Temperature-Aware Three-Dimensional Offloading (TAO), a novel strategy that jointly models power, temperature, and energy across disparate timescales. TAO integrates stochastic steady-state offloading formulation with physics-based thermal simulation (via COMSOL) to enable system-level co-optimization within hardware safety bounds. Compared to state-of-the-art approaches, TAO reduces total offloading cost by over 35% while strictly respecting real-time power limits, thermal thresholds, and battery health constraints. It significantly improves both energy efficiency and thermal reliability. The framework provides a scalable theoretical foundation and practical solution for multi-timescale edge offloading in thermally sensitive wearable systems.

Technology Category

• Machine Learning: Hardware-aware ML
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Extended reality (XR) devices, commonly known as wearables, must handle significant computational loads under tight latency constraints. To meet these demands, they rely on a combination of on-device processing and edge offloading. This letter focuses on offloading strategies for wearables by considering their impact across three time scales: instantaneous power consumption, short-term temperature fluctuations, and long-term battery duration. We introduce a comprehensive system model that captures these temporal dynamics, and propose a stochastic and stationary offloading strategy, called TAO (for temperature-aware offloading), designed to minimize the offloading cost while adhering to power, thermal, and energy constraints. Our performance evaluation, leveraging COMSOL models of real-world wearables, confirms that TAO reduces offloading cost by over 35% compared to state-of-the-art approaches, without violating the wearable operational limits.

Problem

• Optimize XR offloading across power, temperature, and energy constraints
• Develop strategy balancing instantaneous and long-term wearable performance
• Reduce offloading cost while adhering to operational limits

Innovation

• Stochastic stationary offloading strategy TAO
• Considers power thermal energy constraints
• Reduces cost by 35% versus benchmarks