Key Technology Research In Indoor Acoustic Localization Based On Microphone Arrays

Indoor acoustic localization technology based on microphone arrays is one of the important research content and the most popular applications.Some key technologies about this field are well-studied in this thesis.Some effective improvements and optimizations are also put forward.Based on these works, many theoretical analysis and simulation experiments are presented.Firstly, according to the significant subject in indoor acoustic localization technology, investigations about technology background and current research at home and abroad are taken. Works in this thesis supply a certain practice guidance.Then, some related research theory foundations and preconditions are shown including the assumption of research conditions, the physical and topology of arrays, modeling and analysis of array signals, signal preprocessing, the analysis of the signal characteristics and correlation.These works provide necessary groundwork and support in respect of estimation algorithms research.Based on the work above, this thesis discuss the fundamental theory,improvement and application about MUSIC algorithm. Some derivative algorithm such as Weighted MUSIC, the algorithm of cyclostationary theory, Toeplitz MUSIC, singula value vector, propagator method and ESPRIT.Meanwhile, another important content is the ROOT-MUSIC algorithm which studies the spatial smoothing theory and the improvement. By means of weighted reconstruction transformation of the signal covariance matrix, the estimated accuracy is enhanced, the estimation performance to the low SNR signal and coherent signal sources is improved.Moreover, the MUSIC algorithm based on minimum redundant array is also one of the important research content. By means of weighted reconstruction transformation of the signal covariance matrix and the reasonable configuration of array structure, combining the interference cancellation technique, the estimation performance to the signals which are coherent and disturbed by active jamming is improved effectively.Generally speaking, many algorithms researched and simulated in this thesis will be of certain value in terms of deeper insights to acoustic positioning, engineering application and be of great theoretic and practical significance.