Task-Oriented Computation Offloading and Resource Management in D2D-Assisted Heterogeneous Networks

With the emergence of various newly developing applications, the computational tasks generated in the internet of things become more diverse. It is vital to match the diverse tasks with network resources according to the task differential de-mands (e.g., latency, energy, security). In device-to-device (D2D) assisted heterogeneous networks (HetNets), tasks can be offloaded to idle devices or base stations to improve the offloading efficiency and reduce the burden from the core network. Exploiting these benefits, we investigate task-oriented computation offloading and resource management in D2D-assisted HetNets. We aim to minimize the weighted sum of latency and energy consumption, by jointly optimizing the offloading decision, caching strategy and computation resource allocation, while guaranteeing differential task latency requirements and the energy budget of all devices. We first employ K-means algorithm to cluster tasks into several categories based on their characteristics. Further, as the formu-lated problem is a large-scale mixed-integer nonlinear optimizing problem, which is difficult to solve within a reasonable time, a distributed iterative algorithm based on alternating direction method of multipliers (ADMM) is proposed to find a near-optimal solution. Simulation results reveal that our proposed approach can achieve higher performance than other benchmark approaches in reducing latency and energy consumption.