| The jamming of multiple noise and the polynomial phase modulation caused by non-uniform relative motion are the main problems in MMW Doppler guidance radar signal processing. To Research and solve this problem, from the viewpoint of non-stationary signal processing, has been one of the important direction and hotspot in the signal processing domain. This paper carries out a thorough research on signal detection and parameter estimation of moving target under the co-jamming of additive and multiple noise, with the tools of cyclostationary signal processing.In this paper, the concept of Acceleration Ambiguity Function (AAF) and Acceleration Resolution is put forward for the first time. Using AAF, the effects of target's relative acceleration on several outputs of a linear-phase matched filter are analyzed, such as the output signal-noise-ratio loss, the Doppler resolution, the constraint on optimal accumulative time (OPT) and their tolerable limits. The representation of AAF and its resolution not only enriches relative theory of radar signal processing and is instructive in radar parameter design and performance analysis, but also provides a practical method to quantitatively evaluate the acceleration effects on signal processing. Through the detailed analysis of the AAF, the defection of using Chirp-Fourier Transform in guidance radar signal processing is pointed out.Based on the characters of signal processing in Doppler guidance radar and the background of actual engineering application, conditions including the complex observation signal, the conjugate cyclic moment and the narrowband gauss color noise are adopted to characterize the observation signal model. Signal detection and parameter estimation of radar maneuvering target under the co-jamming of additive and multiply noise are researched profoundly.First of all, by establishing the observation model of complex harmonic signal and complex linear chirp signal respectively under the co-jamming of additive and multiply noise, the paper constructed cyclic moment detectors of different orders, deduced SNR expression of the detector through theoretic analysis, discussed the relation of detector performance and noise statistical characteristic. The curve of detection performance is obtained through Monte Carlo simulation test, proving the related conclusion derived.And then, adopting the same model, cyclic moment estimation statistic of different orders are constructed respectively, large sample variance expressions of the estimatorthrough theoretic analysis are derived, as well as the constraint relation of estimator performance and noise statistical characteristic. The curve of estimation performance is obtained through Monte Carlo simulation test, which proves the related conclusion derived.Although the work is based on the nonstationary signal of quadratic phase, the methods can also be extended to higher-order polynomial phase signal.
|
PDF Full Text Request