Open AccessA Sequential Bayesian Algorithm for DOA Tracking in Time‐Varying Environments
Author(s) -
Gao Xunzhang,
Li Xiang,
Jason Filos,
Dai Wei
Publication year - 2015
Publication title -
chinese journal of electronics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.267
H-Index - 25
eISSN - 2075-5597
pISSN - 1022-4653
DOI - 10.1049/cje.2015.01.023
Subject(s) - computer science , robustness (evolution) , algorithm , kalman filter , maximum a posteriori estimation , direction of arrival , covariance , benchmark (surveying) , state vector , gaussian process , gaussian , mathematical optimization , artificial intelligence , mathematics , maximum likelihood , telecommunications , biochemistry , chemistry , statistics , physics , geodesy , classical mechanics , antenna (radio) , gene , geography , quantum mechanics
This paper focuses on the Direction of arrival (DOA) tracking problem in dynamic environments where each source signal is modeled as a Gaussian process with time‐varying mean and unknown covariance. In the presence of highly dynamic environment, benchmark algorithms usually have deteriorated performance. By treating the source signals as a function of the arrival angles, a sequential Bayesian tracking approach named Simultaneous angle‐source update (SASU) is proposed based on the Maximum a posteriori (MAP) principle. The key feature of the proposed approach is to simultaneously update the arrival angles and the source signals in the Kalman filter step by converting the update process of the state vector into a joint optimization problem. An iterative Newton method to efficiently solve the joint optimization problem is proposed. The accuracy and robustness of the proposed SASU algorithm is demonstrated via simulations.