Title :
Switched-capacitor chaotic neuron for chaotic neural networks
Author :
Horio, Y. ; Suyama, K.
Author_Institution :
Dept. of Electr. Eng., Columbia Univ., New York, NY, USA
Abstract :
A switched capacitor implementation of a neuron model which exhibits chaotic behavior is presented. The model is based on the experimentally observed characteristics of a squid neuron. It is shown that the proposed circuit qualitatively replicates the response of the real neuron by extensive simulations using SWITCAP2 and simulated program with IC emphasis (SPICE). Various nonideal effects such as DC gain, input offset voltage, and parasitic capacitances associated with SC implementation are investigated and the operational amplifier offset voltage is identified as a major problem. An existing offset voltage compensation scheme is adopted to alleviate the problem
Keywords :
SPICE; chaos; circuit analysis computing; digital simulation; neural chips; operational amplifiers; switched capacitor networks; DC gain; SPICE; SWITCAP2; chaotic neural networks; input offset voltage; neuron model; nonideal effects; operational amplifier offset voltage; parasitic capacitances; squid neuron; switched capacitor implementation; Capacitors; Chaos; Circuit simulation; Integrated circuit modeling; Neural networks; Neurons; Operational amplifiers; Parasitic capacitance; SPICE; Voltage;
Conference_Titel :
Circuits and Systems, 1993., ISCAS '93, 1993 IEEE International Symposium on
Conference_Location :
Chicago, IL
Print_ISBN :
0-7803-1281-3
DOI :
10.1109/ISCAS.1993.393897
