Adaptive Interference Hypergraph-Based Secure Resource Allocation for Multi-Cell Multi-Carrier MISO-NOMA IoT Networks With Imperfect CSI

With the proliferation of sensitive information transmitted in Internet of Things (IoT), physical layer security (PLS) has emerged as a key technique to ensure the data confidentiality in complex cellular IoT environments. However, the massive IoT devices (IoTDs) organized in multiple cells pose challenges to secure IoT such as severe inter-cell interference and limited spectrum resources. In this article, we focus on the secure communication and interference management in multi-cell multi-carrier multiple-input single-output non-orthogonal multiple access (MISO-NOMA) IoT networks with imperfect channel state information (CSI). To maximize the secrecy sum rate (SSR), a resource optimization problem is formulated by jointly designing subchannel assignment, secure beamforming and artificial noise (AN) injection, while guaranteeing the quality-of-service (QoS) requirements, power budget, and subchannel assignment constraints. To solve the non-convex problem, an interference-aware secure resource allocation scheme is proposed, which decomposes the problem into two joint optimization subproblems. For the first subproblem regarding subchannel assignment, we develop an improved matching algorithm based on adaptive interference hypergraph (AIHG). The second involves secure beamforming and AN injection, which is designed by using accurate surrogate functions and alternating optimization (AO) algorithm. Simulation results validate the framework’s robustness and convergence, demonstrating a significant improvement in secrecy performance over benchmarks.