Neural network match filter of chirp pulse compression

Author

Baojun, Zhao ; Caicheng, Shi ; Yueqiu, Han

Author_Institution

Beijing Inst. of Technol., China

Volume

3

fYear

2000

fDate

2000

Firstpage

1872

Abstract

Because of the large amplitude of the sidelobe of match filter output, the sidelobe suppression must be used in the match filter which reduce the sidelobe jamming and the sensitivity of the system. Because the backpropagation (BP) neural networks have the ability to approximate any nonlinear function, it can be used in a match filter. Training of synaptic weights of the BP neural networks adopts the traditional grads algorithm. The combination between BP and niche GA (NGA) can improve the training convergence and insure to reach at global minimum

Keywords

backpropagation; chirp modulation; convergence of numerical methods; filtering theory; genetic algorithms; matched filters; neural nets; radar computing; radar resolution; backpropagation neural networks; chirp pulse compression; genetic algorithm; global minimum; grads algorithm; match filter output; neural network match filter; niche GA; nonlinear function approximation; radar resolution; sidelobe amplitude; sidelobe jamming reduction; sidelobe suppression; synaptic weights training; system sensitivity; training convergence; Chirp; Convergence; Filtering theory; Frequency; Matched filters; Neural networks; Neurons; Nonlinear filters; Pulse compression methods; Signal resolution;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Proceedings, 2000. WCCC-ICSP 2000. 5th International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-5747-7

Type

conf

DOI

10.1109/ICOSP.2000.893469

Filename

893469