Application of a New Subspace Updating Algorithm in DOA Estimation
Cai Can-hui · Journal of Huaqiao University · 2012
In order to improve the weakness of the traditional methods on tracking sudden parameter changing signals slowly,the approximation eigenvalue decompositionan method,an efficient subspace updating algorithm with variable forgetting factor,is presented.The novel algorithm adopts a new rank_1 updating equation with variable forgetting factors to update the covariance matrix of the array output signals.At the same time it proposes a new cost function which is based on the prior information and the eigenvalue decomposition(ED) of the updated covariance matrix.And then the signal subspace is updated real time by the least square solution of the new cost function.Finally,some examples are simulated and the results exhibit that the algorithm provides not only one fifth of the direction of arrival estimation error,but also five times of the tracking speed of sudden parameter changing signals to the original algorithm,which confirms the accuracy and effectiveness of this new algorithm.