Cross-Technology Communication between WiFi and ZigBee

Abstract

The coexistence of heterogeneous devices in the Internet of Things wireless network brings a new topic of cross-technology communication in order to improve the communication efficiency between heterogeneous devices and promote the cooperation between heterogeneous devices. WiFi and ZigBee are the two most commonly used IoT devices in IoT environments. They have the advantages of low communication delay, high speed, low cost, and easy deployment. Since the spectrum of WiFi and ZigBee overlap, ZigBee signal will affect the transmission of CSI signal. We propose a CTC technique based on machine learning and deep learning, which uses only WiFi channel state information (CSI), and can distinguish whether there is ZigBee signal transmission in WiFi signals by classifying WiFi channel state information. In this paper, machine learning classification methods and deep learning methods are used to classify CSI sequences respectively. We extract the features of 8 CSI sequences to improve the accuracy of machine learning classifier and LSTM classifier, and adopt transformer network classifier to further improve the classification accuracy. In our experimental dataset, the highest accuracy rate can reach 94.9%. The evaluation results show that the accuracy of this method is significantly higher than the existing classification methods.