Integration of Device-to-Device communication in 5G Heterogeneous networks (HetNets) promises to increase resource utilization and spectrum efficiency (SE). However, the broadcast nature of wireless communication makes it vulnerable to an eavesdropper, who can steal useful information from the transmission channel. Hence, ensuring secure communication for both cellular users (CUs) and D2D pairs is the need of the time. In this paper, we study the optimization problem of resource allocation in 5G and beyond wireless networks to maximize the secrecy rate of both CUs and D2D links from the physical layer security perspective. We consider multiple eavesdroppers overhearing the CUs and D2D pairs. Two optimization problems are formulated: one considering separate channels for CUs and D2D user (DU), and second, considering the shared channel between CU and DU. The nature of the formulated problems is mixed-integer non-linear programming (MINLP) problem and to solve it, we propose an $ϵ$-optimal based successive linearization algorithm. This algorithm decomposes our original problem into finite sub-problem branches and solves each separately. Simulation results show that the proposed technique can significantly improve the secrecy rate, throughput, and energy efficiency of the network especially in the case of separate channels for CUs and DUs.

5G异构网络（HetNets）中设备到设备通信的集成有望提高资源利用率和频谱效率（SE）。但是，无线通信的广播性质使其容易受到窃听者的窃听，窃听者可以从传输通道中窃取有用的信息。因此，确保蜂窝用户（CU）和D2D对的安全通信是时间的需要。在本文中，我们将从物理层安全性的角度研究5G和无线网络以外资源的优化问题，以最大程度地提高CU和D2D链路的保密率。我们考虑了多个窃听者窃听CU和D2D对。提出了两个优化问题：一个问题考虑了CU和D2D用户（DU）的单独渠道，第二个问题考虑了 考虑CU和DU之间的共享信道。提出的问题的性质是混合整数非线性规划（MINLP）问题，为解决该问题，我们提出了$ϵ$-基于最优的连续线性化算法。该算法将我们原来的问题分解为有限的子问题分支，并分别求解每个问题。仿真结果表明，所提出的技术可以显着提高网络的保密率，吞吐量和能效，特别是在CU和DU分别使用信道的情况下。