Hybrid Metaheuristics-Driven Distributed Task Scheduling for Latency-Sensitive Edge Data Processing

The rapid advancement of Industry 4.0 and Internet of Things (IoT) has led to an increasing demand for efficient scheduling of latency-sensitive data tasks in distributed edge computing. Real-time applications, such as smart manufacturing and intelligent transportation, demand fast and seamless task execution for operational efficiency. However, scheduling in these dynamic environments is challenging due to resource constraints, task dependencies, and strict realtime requirements. To deal with this challenge, in this paper, we model the scheduling problem as a Distributed No-Wait Flow Shop Scheduling Problem (DNWFSP), with the objective of minimizing total tardiness (TT) while enforcing no-wait constraints to ensure uninterrupted task transitions. For that, we propose two advanced hybrid metaheuristic algorithms: an Iterated Greedy (IG) approach and a Genetic Algorithm (GA), both enhanced with Simulated Annealing (SA) to escape local optima and improve solution quality. Experimental results demonstrate that our proposed algorithms significantly reduce tardiness compared to traditional scheduling methods, making them well-suited for latency-critical applications.