Title :
Super directive sensor array with neural network structure
Author :
Kobatake, Hidefumi ; Morita, Wataru ; Yano, Yoshiharu
Author_Institution :
Fac. of Technol., Tokyo Univ. of Agric. & Technol., Japan
Abstract :
A sensor array system with a neural network structure that realizes superdirectivity characteristics is introduced. The fundamental array system adopted is a three-layered neural network. It is, of course, a nonlinear system. Generally the nonlinearity causes undesirable distortions. However the nonlinearity is used positively to realize superdirectivity, which is unrealizable by a linear array system. A novel network structure is proposed to prevent nonlinear distortions of the desired signal. Computer simulation to test the performance of the proposed system has been performed, and results show that the system works well and the directivity is much sharper than that of a linear system
Keywords :
array signal processing; neural nets; computer simulation; neural network structure; nonlinear system; sensor array system; superdirectivity characteristics; three-layered neural network; Computer simulation; Linear systems; Neural networks; Nonlinear distortion; Nonlinear systems; Performance evaluation; Sensor arrays; Sensor phenomena and characterization; Sensor systems; System testing;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on
Conference_Location :
San Francisco, CA
Print_ISBN :
0-7803-0532-9
DOI :
10.1109/ICASSP.1992.226055