A Machine Learning-Based Algorithm for Through-Wall Target Tracking by Doppler TWR
Jiaxuan Cao,Yipeng Ding,Yiqun Peng,Yuxin Chen,Fangping Ouyang
DOI: https://doi.org/10.1109/tim.2024.3369133
IF: 5.6
2024-03-06
IEEE Transactions on Instrumentation and Measurement
Abstract:Doppler through-wall radar (TWR) enables noncontact behind-the-wall target trajectory tracking, which has a wide range of application scenarios in the field of detection. However, when facing unknown wall parameters, the detection accuracy of Doppler TWR becomes severely limited. Hence, in this work, we propose a machine learning-based target tracking algorithm for through-wall sensing applications. First, using the peak search method based on the short-time Fourier transform (STFT) to obtain a roughly predicted trajectory under the free-space assumption. Then, a classifier based on support vector machine (SVM) is used to estimate the wall thickness from the predicted target trajectory. Finally, a backpropagation neural network (BPNN) is constructed to obtain the corrected target trajectory, whose inputs are the estimated wall thickness and the predicted target trajectory. Experimental results demonstrate that the proposed algorithm significantly improves target tracking accuracy in through-wall detection applications, achieving up to an 80% improvement compared to traditional methods.
engineering, electrical & electronic,instruments & instrumentation