The Technology Research Of Multi Signal Recognition And Suppress Interference Direction Finding

The widespread application of active decoy and radar interference technology claim tohigher requirement in signal separation and detection.Active decoy technology makes thedecoy signals and rader signal are overlapping in time domain, frequency domain andmodulation domain,leading the existing signal sorting technology can not meet therequirement.Blanket jamming is a common way for radar interference technology, whichcover the opposing echo signal via noise signal to hinder or stop the opposing side detecttarget through the echo signal.This paper focus on multi signal recognition and suppressinterference direction finding.The paper submit a array signal processing based scheme to thefollowing solutions to multi signal identification problem and an improved widebanddirection finding algorithm based on cross-spectra is also presented.Array signal processing technology consists of two aspects: spatial spectrum estimationand adaptive beamforming.In the multi signal recognition scheme， spatial spectrumestimation technology is used for angle estimation of multi signals, and adaptivebeamforming technique is used to extract a single signal. These two techniques combinedwith fingerprint analysis and recognition technology are used in multi signal recognition.The principle and the decorrelation algorithm of MUSIC algorithm,a classical spatialspectrum estimation algorithm,is discussed. And a new two-dimensional MUSIC algorithm isproposed.This method utilizes four times one-dimensional MUSIC to replacetwo-dimensional MUSIC, thereby realizing the one-dimensional coherent algorithm fortwo-dimensional signals coherent angle estimation. In addition, the MUSIC algorithm is verysensitive to the channel amplitude and phase error, this paper presents the narrowband signalarray channel amplitude and phase error model, deduces the narrowband signal amplitude andphase error correction method.The principle and a detailed performance analysis in all kinds of error cases and thecommonly used classical robust algorithms are discussed. According to the structure of multisignal identification scheme,an improved LCMV algorithm is proposed in this paper,theweight vector is calculated via virtual interferences constructed with the aid of angle information,which is very suitable for extracting the single signal.The actual data is used to verify the proposed scheme.Processing results show theeffectiveness of the scheme.Noise frequency modulation signal is the main form of blanket jamming.The noisefrequency modulation signal model is proposed in the last part.The principle and defect ofcross-spectra algorithm is discussed. Moreover, a new method is presented, consideringcontradictions between direction finding precision and phase ambiguity in using cross-spectraalgorithm to estimate the DOA. The new algorithm uses frequency-difference andcorresponding phase difference of Cross-spectra to solve phase ambiguity with only twoantennas, and the constrain that the length of baseline must be less than the shortesthalf-wavelength of the signal is not need, thus accuracy of direction finding is effectivelyimproved.Simulation results show the effectiveness of the algorithm.