Source Enumeration Based On A Uniform Circular Array In A Determined Case

In array signal processing,the most estimation algorithms of high resolution direction of arrival implement is premised on knowing precisely the number of sources,the inaccurate sources enumeration causes the direction finding algorithms performance being greatly deteriorated.Thus,The accurate sources enumeration of the array signal has great significance to the subsequent array signal processing.The most of existing algorithm of sources enumeration is implemented under the background of white noise and the number of array elements is far more than the number of sources.However,when the background noise is space colored noise and the total number of sources is close to the number of array elements,especially in the number of sources is only one less than the number of array elements,the estimation performance of the existing algorithm will be invaild,To address this problem,this paper proposes a source enumeration modeling based on a UCA.The main works of the paper include:1.By looking up a large number of excellent domestic and foreign literatures,The paper detailed presentation of the research background of array signal processing and the research status of algorithm of sources enumeration by scholars at home and abroad.2.Review the basic knowledge of array signal.It include the classical array model,the model of array signal,the model of additive noise and The influence factors of the source enumeration.3.The paper briefly introduces the classical algorithm of narrow band signal sources enumeration,such as eigenvalue method,information theory criterion method and gerschgorin radii method,The influence of SNR,the number of snapshots,background noise and number of signal sources on the estimation performance of algorithm is analyzed through theoretical data experiments.4.The modeling method of the paper is based on Hilbert-huang Transform and support vector machine,the paper provides a detailed introduction of the two classical theories and discusses their advantages in many fields.5.The paper heavily present the design process and experimental validation of the algorithm.In order to address the problem of performance of the source enumeration isdeteriorated considerably by the space colored niose circumstance,when the total number of the signal sources is closed to the number of array elements,especially in the number of sources is only one less than the number of array elements,the paper proposes a source enumeration modeling based on a UCA having M antennas estimate the M-1 or less signal source number.It is based on the phase difference of each array sensor of UCA to receive signal source,HHT is appropriate the extraction of the instantaneous phases of original array signal,the extract the eigenvalues for instantaneous phase as trainning sample train in SVM and obtain a mathematical model,in the finally,Estimating the number of sources of unknown array signals using a trained mathematical model.Both the computer numerical simulations and the Radio Frequency(RF)anechoic chamber experiments validate that mathematical model based on our proposed modeling can accurately estimate the source number of array signals in determined case.It also has a high robustness to the array signals under the conditions of incident angles variations,the SNR variations and the snapshot number variations.