The Chinese Remainder Theorem is usually used to resolve ambiguities for both high-prf and medium-prf pulseDoppler waveforms. In a previous paper it was shown that superior results for a single target can be achieved by using a clustering algorithm. In thls paper the problem of multiple targets is considered. A maximum likelihood technique which incorporates the clustered algorithm is developed for the multiple target problem. Simulation results are presented for a high-prf waveform. I INTRODUCTIONA major classification of waveforms deals with ambiguity resolution in range and Doppler. Low-prf waveforms are unambiguous in range but ambiguous in Doppler, medium-prf waveforms are ambiguous in both range and Doppler, and high-prf waveforms are ambiguous in range but unambiguous in Doppler. In a previous paper [I], the ability of two techniques, the Chinese Remainder Theorem [2].and a clustering algorithm, for resolving the range and Doppler ambiguties of a single target for both medium-prf and high-prf waveforms was considered. It was found that the clustering algorithm provides a significant improvement in performance. In tlus paper, the ability of the clustering algorithm to resolve the range and Doppler ambipties of multiple targets is quantified. The specific problem of interest is N targets at the same azimuth (i.e. in the same beam of a phased array) moving at the same speed but separated by a considerable distance in the range dimension. I1 CLUSTERING ALGORITHMFor high-prf waveforms, the most common algorithm for resolving the range ambiguities is the Chinese Remainder Theorem. The problem with the Chinese Remainder Theorem approach is that a small range error on a single prf can cause a large error in the resolved range and there is no indication that this has happened. To avoid this problem, a clustering algorithm was developed and shown [2] to be superior in performance to the Chinese Remainder Theorem.This clustering algorithm does not require a specific relationship between the multiple prf's and produces a cost function which indicates the goodness of the resolution process.The clustering algorithm can be used to resolve either range or velocity ambiguities. In this paper we will be mainly be concerned with resolving range ambiguties. First for each of the n ambiguous range measurements Ri (note, Ri can be a fraction of a range cell), all possible ranges Rfi are generated by where R i is the ambiguous range associated with the i-th detection and the integer k runs fiom and R m a is the maximum range of interest. All the possible ranges generated by the n ambiguous measurements are ordered fiom smallest to largest and denoted by Q i . The first average squared error CR( 1) is found by selecting the m smallest ranges with each range coming fiom a different prf. Then, the average square error is found for these m ranges. The next set of m ranges is found from the previous set of m ranges by adding the next smallest range which has not been previously used and removing a range if it came fiom the same prf a...
Application of neural networks to radar target detection in non-gaussian noise environments is investigated. More specifically two new probabilistic neural networks, namely Gram-marlier Neural Network (GCNN), and Gram-marlier Porbabilistic Neural Network (GPNN) are applied to radar detection problem. The performance of these detectors is evaluated and compared with Backpropagation and Bayesian classifiers by simulation for Gaussian, Weibull, and Lognormal noise environments. Specht [2]. Two new probabilistic neural networks, Gram-marlier Neural Network (GCNN), and Gram-Charlier Probabilistic Neural Network (GPNN) [4] are proposed and applied here for implementation of a maximum likelihood detector described by equation (3). Performances of these detectors for radar target detection are evaluated by simulation. Also performance of these detectors are compared with those of Bayesian classifier and Backpropagation (BP) classifier [l] for Gaussian, Weibull, and Lognormal noise backgrounds. US. Government work not protected by U.S. copyright.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.