IoT Security Performance Intelligent Prediction Based on Lightweight M-SqueezeNet
Xinpeng Zhou, Lingwei Xu, Shubo Cao, Yufang Li, Yong Ling Fu · 2023
Emerging technologies such as artificial intelligence and big data are driving the progression of pluralistic Intelligent Internet of Things (IoT) applications. However, due to the openness and diversity of IoT channels, secrecy information is vulnerable to illegal theft, resulting in IoT communication network interruptions and serious information security problems, which limits the use of IoT in the field of transmitting sensitive data. To handle complex IoT security incidents in real time, accurate secrecy performance predictions are critical to support mobile IoT networks. In this paper, a model of multi-antenna secure communication system based on Decoding Forwarding (DF) relay is proposed. Combining the characteristics of SquezeNet and MobileNet networks, an improved lightweight M-SqueezeNet model is designed which consists of depth separable convolution blocks and fire modules in parallel. This method is suitable for the nonlinear of IoT secrecy data. Experiments have proved that the proposed algorithm has better IoT secrecy performance than others.