Research On Single Channel Source Number Estimation Algorithm

The number of sources and the direction of arrival estimation are the two main research directions of array signal processing technology.They occupy an extremely important position in many fields such as electronic reconnaissance,medical diagnosis,and image processing.The accurate estimation of the number of sources is the precondition of the direction of arrival(DOA)estimation.If the number of sources is unknown or the estimation is not accurate,the performance of subsequent algorithms will be greatly reduced.Many existing estimation algorithms for the number of sources and the direction of arrival are for array signals.In practical engineering applications,only one single channel is acquired due to limitations of device requirements,configuration costs,and application environments signal.The existing source number estimation and direction-of-arrival estimation algorithms are for array signals.Therefore,the number of single-channel source and direction of arrival estimation algorithms must be solved.In this paper,the existing source number estimation algorithm can not be directly applied to the single-channel received signal and the problem of poor noise suppression ability is studied,and the existing algorithm is improved.The major accomplishments include:The existing source number estimation algorithm is only applicable to the array processing signals and cannot satisfy the problem of single-channel acquisition of signals.In order to achieve accurate estimation of the number of single-channel signal sources,a space sampling method is adopted to convert a single channel into a single channel.Multi-channel,multi-dimensional conversion.The Akaike Information Criteria(AIC)method,the Minimum Description Length(MDL)method,and the Spatial Smoothing Rank(SSR)are analyzed and compared in different sources,different SNRs,and different speeds.The number of shots and the detection performance in the case of a coherent signal source;For single-channel AIC and MDL methods in which the number of sources is distorted in a colored noise environment,the diagonal loading technique is adopted,and the eigenvalue of the covariance matrix is loaded with a constant term to characterize the divergent interference signals.The value is optimized to reduce the coherence,and the value of the loading constant term is analyzed and given.From theory and experiment,it is proved that the single-channel source number estimation algorithm combined with loading technology and information theory has consistency;The single-channel Gay-Cyll circle criterion estimation algorithm under color noise is studied.Based on the estimation of the Gaiguille circle criterion,a single-channel source number estimation algorithm using knife-cut method is proposed.The vectorization space reconstructs multiple covariance matrices,and the results are averaged after the transformation.Finally,the loop iteration is used to obtain the optimal values.The simulation verified its feasibility and effectiveness;The MUSIC algorithm is used to estimate the signal direction of arrival(DOA).Briefly outline the principles and implementation steps of the MUSIC algorithm.And proposed to adopt the knife cut method to improve the MUSIC algorithm.The simulation accuracy or the minimum uniform error is simulated under the condition of a uniform linear array and a nested array,respectively.