Joint Computing Resource, Power, and Channel Allocations for D2D-Assisted and NOMA-Based Mobile Edge Computing

TLDR: By exploiting device-to-device (D2D) communication for enabling user collaboration and reducing the edge server’s load, this paper investigates the D2D-assisted and NOMA-based MEC system and proposes a scheduling-based joint computing resource, power, and channel allocations algorithm to achieve the joint optimization.

Abstract: Mobile edge computing (MEC) and non-orthogonal multiple access (NOMA) have been considered as the promising techniques to address the explosively growing computation-intensive applications and accomplish the requirement of massive connectivity in the fifth-generation networks. Moreover, since the computing resources of the edge server are limited, the computing load of the edge server needs to be effectively alleviated. In this paper, by exploiting device-to-device (D2D) communication for enabling user collaboration and reducing the edge server's load, we investigate the D2D-assisted and NOMA-based MEC system. In order to minimize the weighted sum of the energy consumption and delay of all users, we jointly optimize the computing resource, power, and channel allocations. Regarding the computing resource allocation, we propose an adaptive algorithm to find the optimal solution. Regarding the power allocation, we present a novel power allocation algorithm based on the particle swarm optimization (PSO) for the single NOMA group comprised of multiple cellular users. Then, for the matching group comprised of a NOMA group and D2D pairs, we theoretically derive the interval of optimal power allocation and propose a PSO-based algorithm to solve it. Regarding the channel allocation, we propose a one-to-one matching algorithm based on the Pareto improvement and swapping operations and extend the one-to-one matching algorithm to a many-to-one matching scenario. Finally, we propose a scheduling-based joint computing resource, power, and channel allocations algorithm to achieve the joint optimization. The simulation results show that the proposed solution can effectively reduce the weighted sum of the energy consumption and delay of all users.