Synthesis of Through-Wall Micro-Doppler Signatures of Human Motions Using Generative Adversarial Networks

Kainat Yasmeen Shobha Sundar Ram

2024-04-13

Abstract:Narrowband radar micro-Doppler signatures are heavily used to identify and classify human activities. When the radar is operated in through-wall environments, the complex electromagnetic propagation phenomenology introduces considerable distortions in the micro-Doppler signatures through attenuation and multipath. The problem is particularly severe in inhomogeneous wall scenarios involving multiple wall layers, air gaps, or metal reinforcements. Through-wall radar data collection using simulations and measurements involves significant time and effort. In this paper, we propose an alternative method of synthesizing through-wall radar micro-Doppler signatures from their free space counterparts using the generative adversarial network (GAN). We train the GAN using radar micro-Doppler signatures generated from electromagnetic simulations. We generate the radar data for different human motions, along different orientations, and under diverse through-wall conditions. The synthetic radar micro-Dopplers generated from the neural networks are then evaluated for their realism using a denoising autoencoder, which shows an excellent realism score.

Signal Processing

What problem does this paper attempt to address?

The paper attempts to address the problem of synthesizing radar micro-Doppler spectrograms in complex non-uniform wall environments using Generative Adversarial Networks (GANs). Specifically, the paper focuses on how to simulate radar signal characteristics after passing through walls using a GAN model, based on radar micro-Doppler spectrograms of human motion obtained under free-space conditions. The presence of walls leads to complex electromagnetic phenomena such as signal attenuation, multipath effects, and refraction, which cause significant distortion in traditional micro-Doppler spectrograms when passing through walls. This, in turn, affects the accuracy of human activity recognition and classification based on micro-Doppler spectrograms.

To overcome this challenge, the researchers propose a new method that involves training a GAN model to learn and simulate the impact of walls on radar signals. This approach can avoid the time and labor costs required to collect a large amount of radar data through walls in real environments, while also generating realistic through-wall micro-Doppler spectrograms for subsequent classifier training and testing. Additionally, this method can enhance the diversity and richness of the dataset, which helps improve the generalization ability of the classifier.