An Analytical Framework for Multi-Tier NOMA Networks With Underlay D2D Communications

This paper develops a new spatial model for non-orthogonal multiple access (NOMA)-based two tier-cellular network underlying inband device-to-device (D2D) communication with randomly deployed base stations and users. In particular, we model the locations of users by considering two independent Poisson hole process and Poisson cluster process. Using the developed model, we first derive the interference distributions at typical user (cellular/D2D receiver) by considering contribution from dominant interferers. In addition, the performance of NOMA with successive interference cancelation (SIC) is impacted by intra-user interference. Hence, for cellular users, we propose an SIC with intra-user interference estimation (SIE) receiver which attempts to locally estimate and remove intra-user interference for better decoding. Based on the interference distributions results, we analyze the performance of a typical user in terms of outage probability. In particular, the performance of typical cellular user is evaluated under conventional SIC and proposed SIE techniques with perfect and imperfect intra-user estimation. Using these outage probability results, we also characterize the average link throughput experienced by a typical cellular user or D2D receiver (D2D-Rx). Moreover, simulation results are also presented to validate the accuracy of the derived results.