Delay Minimization Offloading for Interdependent Tasks in Energy-Aware Cooperative MEC Networks
Yao Lin Zhu, Yulin Hu, Anke Schmeink · 2019
The partial offloading technologies in the cooperative mobile edge computing (MEC) networks are considered as promising solutions to enable the emerging latency-sensitive and compute-intensive applications. In this paper, we characterize the performance model of an energy-aware MEC networks with multiple servers cooperatively computing a set of interdependent tasks. To minimize the total delay of the whole process of the set of tasks, an optimal offloading design is provided under given energy constraints. In particular, we provide an optimal solution to the offloading problem, which makes a 3-dimensional decision (matrix) representing at which time instant to offload which task to which server. Via simulation, we investigate the performance of the proposed design for the tasks with different interdependency structures. In particular, the impacts of the number of MEC servers, CPU frequencies and in the proposed algorithm on the system performance are studied. In addition, the tradeoff between the delay performance and computation complexity is addressed.