DocumentCode :
2962368
Title :
Analysis of inter-spike interval characteristics of piecewise constant chaotic spiking oscillators
Author :
Hasegawa, Tomonari ; Matsuoka, Yusuke ; Saito, Toshimichi
Author_Institution :
EECE Dept., Hosei Univ., Tokyo
fYear :
2008
fDate :
1-8 June 2008
Firstpage :
3435
Lastpage :
3440
Abstract :
This paper studies dynamics of a simple chaotic spiking oscillator having piecewise constant characteristics. The state variable can vibrate and is reset to the base level just after it reaches the threshold. Repeating this vibrate-and-fire behavior, rich chaotic spike-trains can be generated. Since the solution and return map are piecewise linear, precise analysis is possible. We have investigated characteristics of inter-spike intervals (ISIs) and have found interesting properties: rdquoThe system can output chaotic spike-trains characterized by line-like spectrums of ISIs. Such phenomena and chaos with continuous spectrum appear alternately and make window-like structure in the parameter space. The continuous spectrum of chaos can have wider-band than other types of spiking oscillators.rdquo Presenting a simple electric circuit, typical phenomena are confirmed experimentally.
Keywords :
chaos; neural nets; chaotic spike-trains; continuous spectrum; inter-spike interval characteristics; line-like spectrums; piecewise constant chaotic spiking oscillators; piecewise linear; return map; vibrate-and-fire behavior; Chaos; Chaotic communication; Circuits; Equations; Extraterrestrial phenomena; Intersymbol interference; Neurons; Oscillators; Piecewise linear techniques; Switches;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks, 2008. IJCNN 2008. (IEEE World Congress on Computational Intelligence). IEEE International Joint Conference on
Conference_Location :
Hong Kong
ISSN :
1098-7576
Print_ISBN :
978-1-4244-1820-6
Electronic_ISBN :
1098-7576
Type :
conf
DOI :
10.1109/IJCNN.2008.4634287
Filename :
4634287
Link To Document :
بازگشت