Empowering Heterogeneous Wireless Networks through Efficient Signal Identification

With rapid development of Internet of Things (IoT), the number of deployed IoT devices increases dramatically to meet the requirements of various applications, such as smart homes, smart cities, and environmental monitoring. WiFi, ZigBee, and LoRa are well-known wireless technologies used to connect IoT devices. The trend of rapid growth shows that these devices will coexist in large quantities in the 6G era, leading to severe cross-technology coexistence problems. Our observation on the heterogeneous wireless networks reveals that the sensing of signal type, as well as signal channel, can assist wireless devices with more efficient transmission strategies. In this article, we propose a general signal identification method based on deep learning to predict the signal type and channel for the three kinds of signals. Specifically, we adopt object detection on the spectrogram images to achieve this task. We collect real signals based on a hardware testbed and construct a dataset for training and testing the model. We also make a preliminary investigation for this method, and the results demonstrate its high performance. Finally, we discuss a series of open challenges and directions for further research on this topic, and hope this article can inspire more researchers to conduct studies in this area.