Estimation of Spectral Components Parameters of the Time Series of Raw FMCW Radar Data to Determine the Range and Speed of Location Objects

• Main Authors: Fedotov, Alexander A., Badenko, Vladimir L., Kuptsov, Vladimir D., Ivanov, Sergei I., Eremenko, Danila Yu
• Format: Conference Proceeding
• Language: English
• Subjects: Artificial Intelligence; Convolutional Neural Networks (CNN); FFT; filtering; FMCW; multi-target; object detection; quantization noise; Radar; Radar detection; range; Signal representation; speed; Time series analysis; Time-frequency analysis; Traffic control; Training
• ISSN: 2771-697X
• DOI: 10.1109/EExPolytech56308.2022.9950902

Radars with frequency modulated continuous wave (FMCW) have significant advantages in determining such parameters of sounded objects as range and speed. FMCW radars are widely used in autonomous vehicle control systems, traffic control systems, smart city technologies and many other areas of human life. However, the main gaps which we try to bridge in the manuscript lie in the processing of the FMCW radar signal to detect range and speed of sounded object. The development of data analysis and knowledge extraction methods based on artificial neural networks (ANN) has recently demonstrated significant progress in various fields of science and technology. The hypothesis of this research is that the application of ANN methods to the analysis of radar data should solve above mentioned gap. This article discusses the results of processing raw FMCW radar data using artificial neural network technologies to solve the problems of extracting information about the speed and range of an object.