• DocumentCode
    3398208
  • Title

    VLSI implementation of neural-type cell with MOS linear resistor

  • Author

    Moon, G. ; Zaghloul, M.E. ; Newcomb, R.W.

  • Author_Institution
    Dept. Electr. Eng. & Comput. Sci., George Washington Univ., Washington, DC, USA
  • fYear
    1991
  • fDate
    14-17 May 1991
  • Firstpage
    784
  • Abstract
    A CMOS integrated circuit for the weighted synapse and the summation of the synaptic signals is presented. Neural-type cells (NTCs) are used as the processing elements along with a voltage-controlled linear MOS resistor. This variable resistor is used to control the synaptic weights as pulse densities, and thus the weights are controlled by the gate control voltage. By adding buffered inverter stages, the output signal of the NTC is converted into the normalized pulse stream of the 5 Vp-p signal for easy handling. The summation is executed by a capacitor integration circuit where the currents from different NTCs are accumulated. Simulation results are presented
  • Keywords
    CMOS integrated circuits; VLSI; neural chips; CMOS integrated circuit; MOS linear resistor; VLSI implementation; buffered inverter stages; capacitor integration circuit; neural-type cell; synaptic signals summation; variable resistor; voltage-controlled resistor; weighted synapse; Capacitors; Circuits; Frequency; Pulse inverters; Pulse width modulation inverters; Resistors; Space vector pulse width modulation; Very large scale integration; Voltage control; Weight control;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Circuits and Systems, 1991., Proceedings of the 34th Midwest Symposium on
  • Conference_Location
    Monterey, CA
  • Print_ISBN
    0-7803-0620-1
  • Type

    conf

  • DOI
    10.1109/MWSCAS.1991.251996
  • Filename
    251996