Adaptive Hormesis-Based Optimization (AHBO) for Efficient Task Offloading in Industrial IoT Edge Environments

Adaptive Hormesis-Based Optimization (AHBO) is an adaptation of the Hormesis-Based Optimization (HBO) framework for task offloading in edge computing, targeting Industrial IoT (IIoT) environments. AHBO enhances the original HBO model by introducing an adaptive tuning method. The algorithm operates online, does not require any training and maintains almost a linear time complexity, making it suitable for edge scenarios in IIoT with a frequently varying environment. AHBO is evaluated against Reinforcement Learning Q-learning (RLQ), Harris Hawks Optimization (HHO), and Slime Mould Algorithm (SMA) across 12 different simulation configurations. It consistently outperforms RLQ, SMA and HHO algorithms in most configurations, offering up to 200–300% latency reduction in high-load, low-resource conditions. While SMA shows slight latency advantages in a few low-load cases, its decision time is still significantly higher than AHBO, making AHBO a compelling solution for real-time IIoT task scheduling under variable system stress.