Study On Several Improved Direction Of Arrival Estimation Algorithms

With the rapid development of array signal processing,its theoretical connotations have been continuously enriched.The DOA estimation is an important research branch and hotspot.The reason that the DOA estimation can receive such great attention is because it can be applied to fields such as medicine,geological exploration and radar.Thus its practical needs are very urgent.The traditional DOA estimation algorithms' angular resolution and measurement accuracy need to be improved.Therefore,it is necessary to study and improve the traditional DOA estimation algorithms.This thesis focuses on the DOA estimation of far-field narrowband signal sources.The specific research work is as follows:1.Introduce the relevant basic theory of the DOA estimation.First introduce the basic concepts of the DOA estimation,such as far field,narrowband,incoherent,etc.Then introduce several common array forms and array models,and analyse the DOA estimation algorithms with three incoherent signal sources: DBF algorithm,Capon algorithm,and MUSIC algorithm.Finally,the DOA estimation algorithms with two coherent signal sources are studied: IMUSIC algorithm and ES_MUSIC algorithm.2.Apply the idea of image sharpening in imagery to the above several DOA estimation algorithms.All of the above algorithms use the spectral peak search to perform the DOA estimation.The spatial spectrum can also be viewed as an one-dimensional image.In the image science,the image is sharpened by second-order difference.In the DOA estimation,the second-order difference can also be used to sharpen the spatial spectrum,so as to improve the DOA estimation performance of the algorithms.In this thesis,the idea is applied to DBF,Capon,MUSIC,IMUSIC,and ES_MUSIC algorithms.First,their spatial spectrums are second-ordered,and then their sharpened spatial spectrums are obtained.Computer simulation results show that the new algorithm is superior to the original algorithm in angular resolution,angle accuracy,and robustness.3.The DOA estimation using the sharp peaks of some above classical DOA estimation algorithms.The spatial spectrums of the Capon,MUSIC,IMUSIC,and ES_MUSIC algorithms are not derivable in the direction of arrival,that is,their spatial spectrum is sharp in the direction of arrival.Therefore,the spectral function has a weak correlation in the direction of arrival and strong correlation in other directions.This paper makes use of this feature to the DOA estimation and applies this idea to Capon,MUSIC,IMUSIC and ES_MUSIC algorithms respectively,and obtains a new DOA estimation algorithm.The computer simulation shows that the new algorithm also has angular resolution,angle accuracy,and robustness.All are better than the original algorithm.Finally,the performance of the DOA estimation algorithm based on the new algorithm and the second-order difference are compared by computer simulation.The simulation results show that the former is slightly better than the latter.