Distributed Internet of things (IoT) is becoming an important technology with the popularization of intelligent devices, which builds a bridge between physical networks and social networks. Therefore, how to ensure the security of IoT is an urgent problem. As an effective measure to improve physical layer security, cooperative jamming has been extensively adopted to reduce the eavesdropper's ability to listen to the transmitted signals. Over the past decades, there has been lots of research on cooperative jamming in the single physical layer network. Meanwhile, some works use game theory to analyze the strategies of jammers and legitimate channels. However, these studies assume that the system is ideal and static, which does not conform to the real-world environment. On the other hand, they often ignore the social relationships that exist in the network, nor do they consider the impact of social relationships on the game. In this paper, we first establish a two-layer network model, which includes a physical layer and a social layer. Then we combine a two-layer network with game theory. In the single jamming node scenario, the jamming node is the leader and the legitimate channel is the follower, so we establish a Stackelberg game. Noting the influence of social strength on the game, we rewrite the benefit function of the jamming node. Our experiment results show the relationship between a legitimate channel's benefit and its purchase power quantity. At the same time, we can find those good social relationships can significantly improve the security rate of the communication system.