DRL-Based Federated Self-Supervised Learning for Task Offloading and Resource Allocation in ISAC-Enabled Vehicle Edge Computing

2024-08-27

Digital Communications and Networks

AI Summary

In ISAC-enhanced vehicle-infrastructure cooperative scenarios, resource-constrained vehicles hinder convergence of Federated Self-Supervised Learning (FSSL) and impede simultaneous optimization of real-time performance and energy efficiency. To address this, this paper proposes a joint optimization framework based on Proximal Policy Optimization (PPO). It is the first to deeply integrate PPO with FSSL—specifically the SimCLR architecture—to jointly optimize task offloading decisions, transmission power, CPU frequency, and computation allocation ratio. A novel dynamic offloading threshold mechanism is introduced to enable energy- and time-efficient local-RSU collaborative training without compromising model accuracy. Simulation results demonstrate that, compared to baseline algorithms, the proposed framework reduces system energy consumption by 32.7%, improves offloading efficiency by 28.4%, and increases FSSL model accuracy by 5.2 percentage points.

Technology Category

• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Intelligent Transportation Systems (ITS) leverage Integrated Sensing and Communications (ISAC) to enhance data exchange between vehicles and infrastructure in the Internet of Vehicles (IoV). This integration inevitably increases computing demands, risking real-time system stability. Vehicle Edge Computing (VEC) addresses this by offloading tasks to Road Side Unit (RSU), ensuring timely services. Our previous work FLSimCo algorithm, which uses local resources for Federated Self-Supervised Learning (SSL), though vehicles often can't complete all iterations task. Our improved algorithm offloads partial task to RSU and optimizes energy consumption by adjusting transmission power, CPU frequency, and task assignment ratios, balancing local and RSU-based training. Meanwhile, setting an offloading threshold further prevents inefficiencies. Simulation results show that the enhanced algorithm reduces energy consumption, improves offloading efficiency and the accuracy of Federated SSL.

Problem

• Optimize task offloading and resource allocation in ISAC-enabled VEC
• Balance local and RSU-based training to reduce energy consumption
• Improve efficiency and accuracy of Federated Self-Supervised Learning

Innovation

• DRL-based federated self-supervised learning
• Partial task offloading to RSU
• Energy optimization via resource adjustment