Design of periodic window functions in filter window filter banks for harsh environments

• Published in: IEEE International Conference on Industrial Informatics (INDIN) pp. 986 - 991
• Main Authors: Yang, Chuqi, Xiao, Jun, Zeng, Yufeng, Deng, Baiwei, Ling, Wing-Kuen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2016
• Subjects: Distortion; filter bank; Filter banks; filter window filter structure; Floors; Matched filters; Matching pursuit algorithms; Nickel; Optimization; sparse optimization
• ISSN: 2378-363X
• DOI: 10.1109/INDIN.2016.7819305

This paper proposes the designs of the periodic window functions in the filter window filter banks for harsh environments. Here, two different cases are considered. The first case assumes that the second periodic window functions are constant. The second case assumes that the second periodic window functions are time varying. For the first case, the design of the periodic window functions is formulated as an L 0 norm optimization problem. In particular, the total number of the nonzero coefficients of the impulse responses of the periodic windows is minimized subject to the exact perfect reconstruction condition. The solution of this L 0 norm optimization problem is found via an orthogonal matching pursuit method. For the second case, the design of these two periodic window functions is also formulated as an L 0 norm optimization problem. In particular, the total number of the nonzero coefficients of the impulse responses of these two periodic window functions is minimized subject to the exact perfect reconstruction condition. Likewise, the orthogonal matching pursuit method is employed for finding a solution of the optimization problem. The designed filter window filter banks are employed for harsh environments.