Research on the Algorithm of Nonuniform Compressive Sensing in DOA

• Published in: 2021 IEEE 9th International Conference on Information, Communication and Networks (ICICN) pp. 46 - 51
• Main Authors: Cui, Xuerong, Guo, Bai, Chen, Haihua, Zhang, Yucheng, Li, Shibao, Lian, Xiaochen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 25.11.2021
• Subjects: Compressed Sensing (CS); Dictionaries; Direction-of-arrival estimation; Estimation; Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT); fewer snapshots; Genetic Algorithm (GA); Matching pursuit algorithms; Non-uniform Overcomplete Dictionary(NUOD); Orthogonal Match Pursuit (OMP); Radar; Signal processing algorithms; Sonar
• DOI: 10.1109/ICICN52636.2021.9673984

Direction of Arrival (DOA) estimation is one of the most fundamental problems in array signal processing, and it is widely used in many fields such as radar, sonar, and communications. This paper uses Compressed Sensing (CS) technology to focus on DOA estimation. Aiming at the problems caused by fewer snapshots in the practical application of DOA estimation, a Nonuniform Overcomplete Dictionary (NUOD) combining initialization space is proposed. In addition, the new DOA estimation algorithm of Orthogonal Match Pursuit (OMP) is improved by a Genetic Algorithm (GA). Furthermore, in order to improve the performance of DOA estimation of CS technology, the following three improvements are proposed. (1) Make full use of the Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) to determine the initialization point and then determine the scope of the initialization space according to the Crame'r-Rao bound (CRB). (2) Aiming at a large amount of calculation, the algorithm can realize the nonuniform over-complete dictionary design of the popular array matrix according to the sparse representation of the partition of the initialization space. (3) To further improve the algorithm's performance, the nonuniform over-complete dictionary design is combined with the processing flow of the GA improved OMP algorithm. The simulation results prove that the method in this paper has the advantages of less computing time, low estimation error, and wide application range.