@article{Маленчик_Жук_2024, title={Adaptive Detection of Signal of Moving Target in FMCW Radar with Unknown Noise Power}, url={https://radap.kpi.ua/radiotechnique/article/view/1998}, DOI={10.20535/RADAP.2024.96.32-41}, abstractNote={<p>One of the most promising ways of detecting moving targets at short distances is a FMCW radar. It provides: high-precision of range and radial velocity measurement and low power consumption. The source information for a target detection algorithm in the FMCW radar is the range-Doppler map. It is formed by two-dimensional discrete Fourier transformation (DFT) over the demodulated signals of the corresponding modulation periods obtained during the interval of coherent accumulation. In the case of homogeneous noise with unknown power, the usage of CFAR (constant false alarm rate) algorithms leads to excessive computational costs due to the sliding estimation of the noise power. In addition, the dimensions of the sliding window are limited, which does not allow obtain an estimate of the noise power with necessary accuracy. A harmonic signal with unknown amplitude, frequency and initial phase can be used as a mathematical model of the useful signal from the target. The algorithm for adaptive detection of a harmonic signal with unknown parameters, received at the interval of coherent accumulation of FMCW radar with known noise power, is considered. The detection device is built according to the periodogram scheme. An analysis of the FMCW radar signal detection characteristics at known noise power, which can act as a lower limit at unknown noise dispersion, was performed. Based on the maximum likelihood method, an algorithm for estimating the unknown power of noise based on a test sample obtained from a range-Doppler map is proposed. The estimate of the unknown noise power is a sample mean. Based on the method of interval estimation, confidence intervals will be determined regarding the probabilities of false alarm and target detection depending on the volume of the test sample. The limits of the probability of a false alarm do not depend on the estimation of the noise power. Based on the obtained dependencies, it is possible to determine the volume of the sample, which provides an acceptable value of the length of the confidence interval of the probabilities of a false alarm and target detection.</p>}, number={96}, journal={Visnyk NTUU KPI Seriia - Radiotekhnika Radioaparatobuduvannia}, author={Маленчик, Т. В. and Жук, С. Я.}, year={2024}, month={Jun.}, pages={32-41} }