Graph Theory Based Cooperative Transmission for Physical-Layer Security in 5G Large-Scale Wireless Relay Networks

In this paper, multi-hop cooperative techniques are adopted to improve the physical-layer security in 5G large-scale decode-and-forward relay networks with massive relays and eavesdroppers. The utilization of graph theory is investigated to relieve the burden of massive nodes and ease the cooperative anti-eavesdropping transmission designs. In particular, a secrecy weighted graph is first established according to the network topology. Three scenarios associated with different levels of wiretapping capability are taken into account. Accordingly, secrecy measurements are converted into the weight of each edge and three efficient cooperative anti-eavesdropping strategies are then proposed for physical-layer security enhancement based on the shortest path algorithm, respectively. It is verified that the proposed cooperative anti-eavesdropping strategies have the property of low complexity and are more attractive for large-scale networks. Simulation results highlight the efficiency and effectiveness of our designs. It has been shown that two-hop transmission does not always promise performance gain in terms of secrecy gain. On the contrary, the proposed strategies are able to provide considerable improvement for different cases, emphasizing the necessity of adopting multi-hop cooperative anti-eavesdropping techniques to improve the physical-layer security.