Improving Passive Radar Target Recognition Using a Convolution Composite WiFi Preamble

This paper studies indoor passive radar target recognition and presents a convolution composite scheme involving a WiFi long preamble and radar chirp signal to improve target recognition resolution. In addition, we present trials of different-sized ball recognition and human standing/lying recognition by using the convolution preamble based on finite-difference-time-domain (FDTD) calculations. The results show that the channel estimation anti-fading of the convolution composite preamble is much better than that of the time-domain composite based on receiver cancellation. Hence, more radar chirp constituents can be convolved, the recognition resolution is notably improved, and the convolved signal has the same or even a slightly better resolution than the chirp signal which is in the conventional multiplexing mode. The given method may be applied to device-free people counting to distinguish adults and children and to detect elderly people falling in bathrooms, among other health monitoring applications.