Improving Secrecy Rate and Social Welfare by NOMA Technique in D2D Communications Network

We investigate the advantage of employing the non-orthogonal multiple access (NOMA) in device-to-device (D2D) communications, where D2D pairs are clustered with content demanders. We show that the physical layer security of D2D users who require protection against eavesdropping is improved by employing this NOMA D2D scheme. Moreover, the additional demanders obtain their desired content by participating in a suitable cluster. Hence, we solve this win-win collaborative problem by maximizing the social welfare of the network through in order to optimize user clusters and to assign powers. This optimization turns out to be a mixed-integer non-convex problem. As remedy, we split the original problem into two subproblems. We solve the user clustering subproblem first by using two-sided one-to-one matching theory. Then using the user clustering results, we propose a suboptimal efficient algorithm to solve the power assignment subproblem, which is a non-convex polynomial fractional programming. Our extensive simulations confirm significant advantages our proposed approach.