Studies On Signal Processing Technology Of Doppler-Only Passive Radar

Passive radar using the non-cooperative illuminators of opportunity, is an important developing aspect of modern radar technology and has attracted more and more attention from the international radar community. Containing no transmitter, the passive radar can offer numerous benefits. Most importantly, passive radar is virtually undetectable to surveillance receivers and there is also no constraint in spectrum allocation. In most case, passive radar is smaller, more portable and is of lower cost compared to conventional active radar. In the passive radar operation, one, two or all of the time-delay, Doppler frequency and bearing parameters are exploited to locate and track the moving target. The convenient and common used illuminator is FM signal or television broadcast signal. In this dissertation, we will focus on the signal processing algorithm of Doppler frequency-only passive radar that uses the television signal as illuminator. First the signal model is builted. Then some high effective methods are presented for estimating TV carried frequency and Doppler frequency. Finally, the observability requirements and the expressions for the Cramer-Rao low bound for two or more transmitters are derived, and some interesting conclusions are obtained. The main contributions of this dissertation are summarized as follows:1) A general receiver signal is modeled and the TV signal is studied. And then the illuminator, the target echo signal and the receiving signal are modeled respectively. This makes the basic step in studing the basied TV signal passive radar system.2) For the problem of high effective estimation and tracking of TV siginal carried frequency, a novel, fast and accurate single frequency estimator is proposed, which can realtime track the TV siginal carried frequency and whose performance can approach the Cramer-Rao low bound. Moreover two new effective frequency estimators based on phase difference and autocorrelation are presented, which can estimate the TV siginal carried frequency fast and effectively and whose performance can both reach the Cramer-Rao low bound.3) For the problem of high effective Doppler frequency estimation, the proposed discrete Fourier transform-based algorithm performs a frequency interpolation on the results of an N point complex Fourier transform very well with a novel residue frequency discriminator, whose performance can approach the Cramer-Rao low bound. The algorithm can effectively estimate the Doppler frequency when the background noise is comparative weak. Moreover, a new approach, in which truncated Taylor polynomials of the comple exponentials are combined with FFT to compute the nonuniform DFT is presented, which, performs much more accurately than the min-max and other interpolation NUFFTs. The new approach works better when the background noise is comparative strong.4) The fundamental properties and performance of the locating a moving target from corrupted Doppler frequency measurements obtained by a single receiver with two or more cooperative or non-cooperative transmitters are analyzed. Both the observability requirements in two dimensions and the expressions for the Cramer-Rao low bound for two or more transmitters are derived. This builds the fundamention in studing this kind of passive radar system.5) Further study reveals the unusual relationship between the location performance and the number of transmitters, and also shows that the radars can theoretically exploit the Doppler information to provide a resolution comparable to conventional radars.The system simulation model is builted and the conclutions are demonstrated by two classic scenario simulations.