Research on Wireless Sensing Method of Human-Vehicle Recognition for Intelligent Transportation
The wireless sensing-based human-vehicle recognition(HVR) method has the advantages of low deployment cost and non-in-vasiveness,which is the key means to improve the intelligent level of intelligent traffic system.However,most of the existing wireless sensing methods are based on 2.4 GHz wireless energy data,ignoring the application of sub-GHz in this field,and there are few works that explore the intrinsic relationship between antenna height and algorithm performance.Therefore,to explore the relationship between different wireless carrier frequencies and antenna heights and algorithm performance,and a wireless sensing-based HVR method with convolutional neural network (WsHVR-CNN) is proposed to clarify the design criteria and multi-parameter influence mechanism of WsHVR-CNN method.Experimental results show that both sub-GHz and 2.4 GHz wireless energy data can lay the foundation for the construction of wireless sensing-based human-vehicle recognition method,and the WsHVR-CNN based on the 2.4 GHz band with the antenna height of 0.8 m has the best performance,reaching 98.15％.Additionally,a dataset developed is publicly available at https://github.com/TZ-mx/WiParam to provide the basis for in-depth study of related work.
万方数据
维普
