Identification of NLOS Condition in Different Scenarios Based on Wireless Propagation Features

Safety driving is an essential requirement in the intelligent transportation system on expressways and urban roads. Particularly, identifying the presence of the visual connection between vehicle and vehicle, i.e., the non-line-of-sight (NLOS) identification, is important for many sensor systems on board such as optical radar, wireless positioning system and so forth. Usually NLOS identification can be done by checking the presence of the line-of-sight propagation path in wireless radio systems. In this paper, we present a study on the non-line-of-sight identification based on measurement data collected on expressways and urban intersections. Features of channel impulse response including skewness, kurtosis, root-mean-square delay spread, peak-to-average ratio and Rician-K factor are considered, and the identification accuracy of different features in different scenarios is presented. The results reveal that the skewness, kurtosis and peak-to-average ratio features are less affected by the environment, while the root-mean-square delay spread feature is obviously affected by the environment. Furthermore, it is found that the Rician-K factor is not suitable for identifying non-line-of-sight propagation in these two scenarios.