Optimization of Data Upload Based on Cooperative Game Theory in Cloud Edge Scenarios

In the existing mobile crowd intelligence sensing scenarios based on edge computing, the collaboration among edge servers is little considered. Collaboration among multiple edge servers faces challenges such as edge service caching, edge device resource constraints, and large-scale edge user task offloading. Addressing the above challenges, this paper formulates the optimization problem for service latency and service cost based on collaborative multi-edge server cooperation, and proposes a computational offloading algorithm based on multi edge server cooperative game and discrete particle swarm optimization (MCG+DPSO) optimization (MCG+DPSO). Firstly, an initial relay scheme for edge users’ tasks among multiple edge servers is derived through Multi-Edge Server Cooperation Game (MCG). Then, the obtained scheme is taken as an initial solution for the Discrete Particle Swarm Optimization (DPSO) algorithm. Finally, the DPSO algorithm is used to obtain the optimal solution, achieving the matching between user tasks and edge servers, thereby maximizing the reduction of service latency and costs for data uploading on mobile sensing platforms. Through extensive comparative experiments on real datasets, it is shown that, compared with cloud strategy, edge non communication strategy, random strategy, cooperative game strategy, DPSO algorithm, and differential evolution algorithm, the proposed MCG+DPSO algorithm can reduce service costs by up to 3.2% to 56.0% and service latency by 3.6% to 24.5%.