Adaptive Privacy-Preserving Task Scheduling with User Device Collaboration in High-Density Edge Computing Environments

The rapid advancement of 5G technology has led to high-density IoT ecosystems, creating challenges in maintaining service quality, especially during large-scale events. This research addresses service degradation in such environments by integrating privacy protection techniques with user device assistance for adaptive service capacity adjustments. We propose a novel approach using deep reinforcement learning (DRL) for VNF offloading and task scheduling, incorporating differential privacy to protect data. Experimental results show that the Advantage Actor-Critic (A2C) model, despite a 9% longer response time than the DQN model, significantly outperforms privacy rewards and service quality, especially in high-density environments. These findings highlight the effectiveness of integrating privacy-preserving methods into edge computing and offer a foundation for future optimization in high-demand scenarios.