Joint Optimization of User Grouping and Mode Selection in NOMA-MEC System

With the development of network technology in the new era, Non-orthogonal Multiple Access (NOMA) and Mobile Edge Computing (MEC) are regarded as the promising network technologies for facilitating the fifth generation network technology (5G). This technology can significantly improve the system throughput and spectrum efficiency, and can also reduce the delay of system data transmission. In this paper, users are divided into several NOMA groups, and two different offloading methods MEC and cloud are discussed. This paper also considers the delay caused by the link layer queuing effect and packet retransmission caused by packet error. In order to solve the optimization problem of joint user grouping and mode selection aiming at the minimum total delay, an improved genetic algorithm is proposed. Simulation results show that the proposed algorithm significantly reduces the total delay and achieved better optimization performance.