MODELING AND SIMULATION OF A NEW VLSI ANALOG CIRCUIT FOR ARTIFICIAL NEURAL NETWORKS APPLICATIONS | ||||
International Conference on Aerospace Sciences and Aviation Technology | ||||
Article 51, Volume 12, ASAT Conference, 29-31 May 2007, May 2007, Page 1-10 PDF (172.32 K) | ||||
Document Type: Original Article | ||||
DOI: 10.21608/asat.2007.23969 | ||||
View on SCiNiTO | ||||
Author | ||||
MOHAMED ALAZAB | ||||
HIGHER TECHNOLOHICAL INSTITUE, DEPARTMENT OF ELECTRONICS AND COPMUTER ENGINEERING, RAMADAN TENGTH, EGYPT | ||||
Abstract | ||||
This paper presents modeling and simulation for a new VLSI –Artificial Neural Network (ANN) analog circuit designed in 0.34 μm CMOS technology. Differential voltage controlled current source (DVCCS) is used for making synaptic weights and activation functions and capacitors are used for dynamics . Multi output bidirectional neuron current mirror is achieved using super MOSFET which behaves as a single MOS transistor but with nearly zero channel modulation factor λ and intrinsic gain of more than 90 dB. Verification of successful operation of all circuit components were done using HSPICE program. Transfer characteristic of DVCCS with a given bias current which can be seen to have a trance conductance of about 160 μ S with band width reaches 400MHz. | ||||
Keywords | ||||
Differential Voltage current control (DVCCS); Super MOSFET; artificial neural network; Analog VLSI Neural network | ||||
Statistics Article View: 104 PDF Download: 161 |
||||