Title :
A neural-line like chain model generating 1/f noise
Author :
Yamada, Mitsuhiro ; Agu, Masahiro
Author_Institution :
Dept. of Electr. & Electron. Eng., Ibaraki Univ., Japan
Abstract :
A neural-line like chain model is proposed whose temporal evolution generates 1/f noise. This model is composed of neuron like threshold cells with internal states. The cells form one-dimensional chain and interact with nearest-neighbor ones. The dynamical behavior obeys simple rules such that random signal is applied at the first edge of the chain and nonlinear interaction with dissipation occur like avalanche breakdown. Simulation results shows that internal states of all the cells fluctuate following the 1/f power spectrum. This is interpreted as the superposition of the Lorentzian power spectra of the fluctuation of each cell. This simple model provides a measure for an explanation for 1/f phenomena in neural systems
Keywords :
1/f noise; neural nets; nonlinear dynamical systems; 1/f noise; Lorentzian power spectra; avalanche breakdown; dynamical behavior; internal states; neural-line like chain model; neuron like threshold cells; one-dimensional chain; temporal evolution; Avalanche breakdown; Boundary conditions; Cutoff frequency; Distribution functions; Electric breakdown; Nearest neighbor searches; Noise generators; Statistics; Tin;
Conference_Titel :
Evolutionary Computation, 1996., Proceedings of IEEE International Conference on
Conference_Location :
Nagoya
Print_ISBN :
0-7803-2902-3
DOI :
10.1109/ICEC.1996.542717