Smart Radar System, Inc.

Jong Il Kim

One-dimensional data obtained by performing a Fourier transform on a digitally converted frequency modulated continuous wave (FMCW) radar signal at every pulse repetition interval is input to a recurrent neural network (RNN) to find the probability of the existence of a target in each range index. The range indices, i.e., bit frequencies are selected based on the probability. In order to reduce the size of the RNN, windowing may be applied. In addition, a speed-index-specific coefficient value may be calculated by accumulating and performing a Fourier transform on the selected bit frequencies in which the target exists over a plurality of pulse repetition intervals. Thus, it is possible to calculate the speed of the target.

Method and apparatus for radar signal processing using recurrent neural network

Two-dimensional data obtained by performing a Fourier transform on a digitally converted FMCW radar signal at every pulse repetition interval over N pulse repetition intervals is input to a convolutional neural network (CNN) to find the probabilities of the existence of a target in range indices. The range indices, i.e., bit frequencies are selected based on the probabilities of the existence of the target. In order to reduce the size of the CNN, windowing is applied to the two-dimensional data. A speed-index-specific coefficient value may be calculated by re-performing a Fourier transform on range data for the selected bit frequencies. Thus, the range and speed of the target may be calculated.

Method and apparatus for radar signal processing using convolutional neural network

Disclosed is a technique for processing signals received from a radar and, in particular, a technique for tracking a target on the basis of detected target candidate signals. The proposed invention introduces a recurrent neural network with a memory function in order to find a target signal from signals with noise and fake signals mixed therein. This recurrent neural network is trained to have a maximum of Q tracking buffers therein. According to an additional aspect, it is possible to increase tracking accuracy through a serial connection of the recurrent neural network. According to an additional aspect, it is possible to track multiple targets through a parallel connection of the recurrent neural network.

Growing community of inventors

Yongin-si, South Korea

Jong Il Kim