A Real-time Location Method Under Unknown Wall Characteristics in Through-wall Imaging

To solve the real-time through-wall location problem under unknown wall characteristics, a method based on extreme learning machine (ELM) is proposed in this paper. Through the training process of ELM, the nonlinear and sick relationship between the scattered signals and the target location is established, then the location will be estimated according to the relationship. The number of random hidden neurons, the activation function, and normalization of data in ELM are discussed. Numerical results demonstrate that all these factors have an impact on the accuracy of the location, and the ELM method is suitable for real-time locating under unknown wall characteristics both for noiseless and noisy circumstances.