A New Method for Enhancing the Robustness of Close-Range Targets Imaging Based on UWB

The ultrawideband (UWB) device-free localization (DFL) technique is promising due to its ability to localize targets without needing additional equipment. In particular, the radio tomographic imaging (RTI) method is of interest due to its low cost and ease of deployment. Traditional RTI methods, which depend on line-of-sight (LOS) path and analyze the variation of received signal strength (RSS), often struggle with multipath interference, especially when the distance between targets is close. This can lead to degraded image quality due to the increased presence of multipath components (MPCs). To overcome this challenge, this study adopts a single moving anchor deployment method, utilizing communication between a moving anchor and fixed tags to obtain channel impulse response (CIR) information. Preprocessing such as interpolation and alignment of CIR data is performed, followed by the computation of correlation coefficients under on-load and no-load conditions to serve as inputs for the RTI model. Furthermore, to solve the problem of close-range target overlap, the K-means clustering algorithm is applied to post-process the correlation coefficients. The threshold detection algorithm is utilized to correct the correlation coefficients. Experiments show that the algorithm based on CIR can achieve effective differentiation and localization of close-range targets in typical indoor scenes.