Font Size: a A A

Acoustic Array Wideband Signal Processing On DOA Estimation Algorithm And Application Research

Posted on:2017-03-03Degree:MasterType:Thesis
Country:ChinaCandidate:Z X TanFull Text:PDF
GTID:2308330485463746Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Array signal processing is an important part of signal processing, as the change of environment and the development of technology, wideband signal processing in array signal processing system has become more and more necessary. Direction of arrival estimation as a part of array signal processing has attracted a lot of domestic researcher and foreign scholars to do a lot of studies which have made remarkable achievements. As wideband signal has larger bandwidth and carries more information than a narrowband signal, it is very necessary to study how to do with wideband signal. This thesis focuses on how to deal with wideband signal multi-source direction of arrival estimation. The main research work of this thesis can be summarized as follows:Firstly, this thesis gives an introduction of the wideband signal model and the direction of arrival estimation theory, and then some evaluation parameters are listed for better evaluate the estimated performance. Especially the Cramer-Rao Bound, which can be compared with the different algorithm’s results at the same conditions. And demonstrate how it influences the direction of arrival estimation when the array aperture is smaller or larger than half of the wavelength.Secondly, the traditional algorithms of estimating at each frequency point based on focus of wideband signal direction of arrival must get pre-focused matrix, which need pre-selected focus point. However, direction of arrival estimation’s error will have a significant impact on pre-focused effect, once the deviation appears, it may cause these algorithms ineffectively. This article will estimate the direction using some narrowband signal estimation theory algorithms to enhance the broadband signal processing, such as beamforming technology, the maximum likelihood estimator and the theory of spatial spectrum estimation of subspace algorithms, then we propose a multiple signal classification smoothing algorithm of the circular array, breaking the traditional theory which only can applied in line array. Simulation experiments on various algorithms both single-source and double-source estimation performance, and various types of algorithms resolution performance to do some analysis are given in different SNR estimation different snapshots and different aperture of arrays.Then, we will get the high-order array signal information of signal in estimation rather than the traditional second-order array signal information, which can extend the array aperture and estimate when signal source number is more than number of array elements. It improves the estimate performance, although it needs larger amount of data and higher computational complexity than before.Finally, wideband signal direction of arrival estimation algorithms which been proposed in the article have been applied into the actual system, different arrays can get different features, so it is need different algorithms to correspond. Both test in a laboratory which own low noise environment and other outside experiments can measure the direction of arrival estimation of sound of helicopter and tanks, and use the results of estimation to track trajectory which are very effective. In conclude, those experiments proved the helpful of proposed algorithms in article.
Keywords/Search Tags:Wideband signal, direction of arrival estimation, spatial spectrum, Cramer-Rao Bound, Higher order
PDF Full Text Request
Related items