Parallel Double‐Loop Particle Swarm Optimisation for Offloading Strategy in Blockchain Based Cloud‐Edge‐End Collaborative Computation
Junjie Cao, Zhiyong Yu, Jian Yang, Jinjin Chai, Zheng Liang, Yuanfeng Yang · IET Communications · 2025
ABSTRACT Computing offloading strategies have emerged as a pivotal topic in IoT research, serving as a strategic approach to minimising latency and energy consumption in edge computing. Nevertheless, these advancements have also introduced new challenges, especially in the context of computationally intensive tasks and increasing data privacy requirements in IoT applications. Addressing these challenges requires innovative solutions, with blockchain technology emerging as a highly promising tool. Our work explores the integration of blockchain into IoT edge computing, specifically focusing on computing offloading. We have devised a system architecture that seamlessly integrates edge computing with blockchain technologies, aiming to tackle both optimisation and security concerns. By framing the offloading problem as a multi‐objective optimisation issue, we developed a Parallel Double‐Loop Particle Swarm Optimisation algorithm. This algorithm introduces a dual‐loop subpopulation structure based on broadcast feedback, which enhances information exchange between subpopulations and increases population diversity, thereby mitigating the risk of premature convergence to local optima. Moreover, an innovative collaborative mechanism is introduced, allowing multiple generations of updated particles to iterate simultaneously, significantly boosting the search efficiency of the population. Simulation results demonstrate that the PDPSO algorithm not only achieves significantly faster update speeds but also excels at identifying optimal solutions for the objective function. When compared to other existing algorithms, it offers superior overall performance and holds significant potential for addressing multidimensional constraints such as system delay and local energy consumption. Our findings suggest that this approach could be transformative for future IoT edge computing systems, providing a robust framework for efficient and secure task offloading.