Task Similarity-Aware Cooperative Computation Offloading and Resource Allocation for Reusable Tasks in Dense MEC Systems

As an emerging paradigm for supporting computation-intensive and latency-sensitive services, mobile edge computing (MEC) faces significant challenges in terms of efficient resource utilization and intelligent task coordination among heterogeneous user equipment (UE), especially in dense MEC scenarios with severe interference. Generally, task similarity and cooperation opportunities among UE are usually ignored in existing studies when dealing with reusable tasks. In this paper, we investigate the problem of cooperative computation offloading and resource allocation for reusable tasks, with a focus on minimizing the energy consumption of UE while ensuring delay limits. The problem is formulated as an intractable mixed-integer nonlinear programming (MINLP) problem, and we design a similarity-based cooperative offloading and resource allocation (SCORA) algorithm to obtain a solution. Specifically, the proposed SCORA algorithm decomposes the original problem into three subproblems, i.e., task offloading, resource allocation, and power allocation, which are solved using a similarity-based matching offloading algorithm, a cooperative-based resources allocation algorithm, and a concave-convex procedure (CCCP)-based power allocation algorithm, respectively. Simulation results show that compared to the benchmark schemes, the SCORA scheme can reduce energy consumption by up to 51.52% while maintaining low latency. Moreover, the energy of UE with low remaining energy levels is largely saved.