Research On Wideband Signal Array Direction Finding Technology And Application

Array direction finding(DF)technology has been an active research topic with its wide application prospects in civil and military fields,while most of the existing DF algorithms focus on narrow band signals.With the rapid development of modern communication,wideband signal has been widely used in communication system with the advantage of easily target detection,large information carrying capacity and robust anti-jamming capability.Although the DF technology of wideband signal array has developed rapidly recently,it still has some problems,such as low estimation accuracy under low signal-to-noise ratio(SNR),high algorithm complexity,weak timeliness,over-reliance on prior information and etc.Especially in the non-cooperative scenario,the receiving end often does not have good receiving conditions and more prior information.In order to better meet the needs of actual engineering,this article relies on the project topics,and the DF technology and application of wideband signals are studied,The main work and innovation point are as follows:1.For the test of orthogonality of projected subspace(TOPS)algorithm,the reference frequency signal subspace estimation may be inaccurate which causes the error to spread to the entire frequency point,and the overall algorithm performance degradation.We present an improved algorithm based on the discrimination of eigenvalues.The proposed method uses the eigenvalue discrimination degree corresponding to the subspace at each frequency point,and selects the frequency point with the largest mean value difference as the reference frequency point,so as to construct the signal subspace at the reference frequency point.The proposed algorithm reduces the error of the entire frequency band caused by the inaccurate selection of the reference frequency.Numerical simulation results show that our approach performs better than the original algorithm with the estimation accuracy DOA the selected reference frequency points improved by about 0.2,the success probability increased by about 10%,and the threshold of 100% resolution probability SNR diminished by 8 d B.2.To resolve the problem that the test of orthogonality of frequency subspace(TOFS)algorithm cannot deal with coherent signals.An improved decoherence algorithm for virtual array is proposed by using specific elements of the covariance matrix construct virtual array data,and exploiting the spatial smoothing to achieve decoherence,And simulation experiments indicate that the proposed algorithm has no array freedom loss,the estimation accuracy is improved by 2 at low SNR,and the SNR threshold of 100% success probability is diminished by 8d B.3.An enhanced TOFS algorithm based on joint estimation of sub-band energy and sources number is recommended for the DF technology of wide and narrow band mixed signals under aliasing conditions.The traditional engineering method relies too much on the prior information with large possibility of ‘missing estimation'.Firstly,using energy searches,pattern division,source number estimation,and data updates,the array received data is divided into three modes:independent narrowband data,wideband data,and aliasing narrowband data.Then,a judgment matrix is constructed using the orthogonality of the direction vector with the noise subspace in different modes.Finally,the DOA is estimated using spectral peak analysis.The simulation results show that the proposed algorithm is feasible to a certain extent,avoiding the influence of lack of prior information on the algorithm performance,and improving the estimation accuracy.It is an extended application of TOFS algorithm in practical engineering.4.In view of the problem that the harmonic information of time modulated array wideband signal direction measurement technology is not fully utilized and the estimation accuracy is not high under the condition of low SNR,a wideband signal DF algorithm based on multi-harmonic analysis is studied.The method comprehensively utilizes the multiple harmonic components and obtains the best estimated value through generalized least squares estimation,which reduces the sub-band estimation error caused by noise.The simulation results reveal that the algorithm is more robust under low SNR environment.5.Relying on the project,we complete the implementation of the wideband signal DF system engineering.The system realize the functions of source number estimation,wideband coherent signal and wide narrowband source DOA estimation in the bandwidth range of 200 MHz? 3 GHz.The performance of the system was evaluated through field experiment tests,and according to the results of the measured data.The estimation accuracy of the system was about 0.7,and it was able to successfully distinguish the wideband coherent signals and the different types of sources under aliasing conditions.System performance indicators meet the subject requirements.