Tensor-based task allocation using multi-objective optimization in GECC environment

Green Edge-Cloud Computing (GECC) has emerged as a promising paradigm to meet the diverse requirements of modern applications by integrating edge and cloud resources. Existing task allocation strategies in GECC environment often fail to adequately address the problems of low resource utilization and high economic cost in multi-objective conflicts. Therefore, this paper proposes a tensor-based task allocation scheme using multi-objective optimization in GECC environment. We first extend the task allocation problem in GECC environment to a multi-objective optimization problem and construct five optimization models, i.e., energy, system reliability, quality of experience, economic cost, and latency. Then, to address the complex relationship among these objectives, we develop a tensor-based representation and calculation model for task allocation across cloud, edge service, and edge device platforms. Furthermore, we propose a tensor-based multi-objective beetle swarm optimization algorithm combined speed limiting and dynamic step strategies (TMOBSO-SLDS) that dynamically adjusts the step size and limit speed to improve the global search efficiency and the diversity of solution set. Extensive experimental results in various task allocation scenarios demonstrate that our proposed TMOBSO-SLDS algorithm outperforms existing approaches, as measured by the HV value. It can significantly enhance the diversity of the solution set and improve resource utilization.