A Distributed Maximum Offloading Gain Algorithm With Fairness Consideration for Dependent Task Offloading in Multiuser/Multifog Scenario

Existing dependent task offloading approaches mainly rely on centralized algorithms and are typically constrained to single-objective optimization. In addition, offloading fairness among fog servers remains insufficiently addressed. To tackle these challenges, this paper proposes a novel distributed method that integrates Gale–Shapley Matching (GSM) with Improved Particle Swarm Optimization (IPSO), whose goal is to maximize the offloading gain on the user side and the fairness on the fog server side. Specifically, we firstly formulate an offloading gain maximization problem for dependent task offloading in multi-user and multi-fog computing scenarios, where the offloading gain is defined by jointly optimizing energy consumption and latency of task offloading. Then, we propose a GSM preference list construction method that considers user offloading gain, the number of offloading acceptances by fog servers and channel quality. Finally, GSM is employed to achieve stable user–fog associations, and IPSO is applied to optimize subtask offloading decisions. Experimental results show that, compared with existing algorithms, this method increases the average offloading gain by 9%; and improves the average fairness by 16.6% with minimal loss of user offloading gain. Therefore, the proposed algorithm can jointly optimize the latency and energy consumption of user tasks while ensuring excellent long-term fairness among fog servers.