With the rapid development of artificial intelligence, the Internet of Things (IoT) is moving towards a new stage of intelligence. IoT technology has been widely used in many emerging fields such as intelligent transportation and cities, which brings new challenges and opportunities for data security protection technology. Although complex networks are recognized as powerful tools that can solve problems in various domains, existing research is still limited to single, interconnected, and interdependent networks, ignoring the numerous complex phenomena in the IoT. Therefore, to fill the gap of complex networks in this part of the IoT domain. We aim to develop a new three-layer network model based on IoT and research the robustness of the network under different failures. In contrast to previous work, our model adopts a bottom-up approach, dividing the IoT into “sensing-transmission-application” layers. This model employs two classical algorithms for generating intra-layer connections and a random connection algorithm for inter-layer connections. The overall properties of the model are researched by analyzing the structural characterization metrics. We also adopt two attack strategies and compare the changes of key indicators under different degree distributions to assess the robustness of the network. The experimental results show that the robustness of the network is closely related to the peak of the degree distribution. Our research provides new ideas and methods for modeling and data security governance techniques in IoT. It lays the foundation for building more reliable and secure IoT models.