LAUSR

In this paper, we investigate the sum-rate maximization of cellular users (CUs) and device-to-device (D2D) pairs for the D2D communications underlaying cellular networks over the Nakagami- $m$ fading channel. We consider a resource-abundant scenario where the number of CUs is more than that of D2D pairs and discuss the joint optimization of signal-to-interference-plus-noise ratio (SINR) thresholds and CU-D2D pairing. We formulate the problem as mixed-integer non-linear programming, which is NP-hard in general. To solve the problem, we propose a two-step approach. First, we add virtual D2D pairs to make the number of D2D pairs equal to that of CUs and derive the near-optimal SINR threshold for each possible CU-D2D pairing by the simulated annealing algorithm. Second, the optimal CU-D2D pairing can be obtained by the Hungarian (HG) algorithm. The simulation results show that: 1) the proposed SINR policy outperforms the other policies by about 11.3%–63.4% and 2) compared with several state-of-the-art resource allocation policies, HG allocation can achieve an average sum-rate improvement of 10.9%–72.8%.