In this paper, we propose a load balancing algorithm based on Reinforcement Learning (RL) to optimize the performance of Fog Computing for real-time IoT applications. The algorithm aims to minimize the waiting delay of IoT workloads in dynamic environments with unpredictable traffic demands, using intelligent workload distribution. Unlike previous studies, our solution does not require load and resource information from Fog nodes to preserve the privacy of service providers, who may wish to hide such information to prevent competitors from calculating better pricing strategies. The proposed algorithm is evaluated on a Discrete-event Simulator (DES) to mimic practical deployment in real environments, and its generalization ability is tested on simulations longer than what it was trained on. Our results show that our proposed approach outperforms baseline load balancing methods under different workload generation rates, while ensuring the privacy of Fog service providers. Furthermore, the environment representation we proposed for the RL agent demonstrates better performance compared to the commonly used representations for RL solutions in the literature, which compromise privacy.

翻译：本文提出一种基于强化学习的负载均衡算法，用于优化面向实时物联网应用的雾计算性能。该算法通过智能工作负载分配，旨在动态环境中不可预测的流量需求下最小化物联网工作负载的等待延迟。与现有研究不同，我们的解决方案无需从雾节点获取负载和资源信息，从而保护服务提供商的隐私——此类信息可能被竞争对手利用以制定更优定价策略。所提算法在离散事件模拟器上进行了评估以模拟实际环境部署，并在超过训练时长的仿真中测试了其泛化能力。结果表明，在不同工作负载生成速率下，我们的方法在确保雾服务提供商隐私的同时优于基准负载均衡方法。此外，相较于现有文献中会泄露隐私的常用强化学习环境表示方法，我们为强化学习智能体提出的环境表示具有更优性能。