Skip to Main Content
In this paper, we study the radar sensor wireless channel modeling in foliage environment, a rich scattering and time-varying environment, based on extensive data collected using ultra-wideband (UWB) and narrowband (200 and 400 MHz) radar sensors. We apply two approaches to the wireless channel modeling: Saleh and Valenzuela (S-V) method for UWB channel modeling and CLEAN method for narrowband and UWB channel modeling. We validated that UWB echo signals (within a burst) do not hold self-similarity, which means the future signals cannot be forecasted based on the received signals and channel modeling is necessary from statistical point of view. Based on the S-V method for UWB channel modeling, in foliage UWB channel, the multipath contributions arriving at the receiver are grouped into clusters. The time of arrival of clusters can be modeled as a Poisson arrival process, while within each cluster, subsequent multipath contributions or rays also arrive according to a Poisson process. At different distance (near distance, medium distance, and far distance), we observe that the Poisson process parameters are quite different. We also observe that the amplitude of channel coefficient at each path follows Rician distribution for medium and far distance, and it is non-stationary for paths from short distance (one of two Rician distributions), and these observations are quite different with the IEEE indoor UWB channel model and S-V indoor channel model. Based on the CLEAN method, the narrowband (200 and 400 MHz) and UWB channel impulse responses have many similarities: both can be modeled as linear time-variant filter channel. We also studied the large-scale fading using path-loss and log-normal shadowing model for foliage enviroment, and observed that the path-loss exponent is very high because it has rich scattering.