MOSFETの弱反転領域を用いた4象限・低電圧・低消費電力乗算器とCNNセルへの応用

概要

論文の詳細を見る
The thesis describes the technique of the LSI realization for a Cellular neural network's (CNN). One of the serious problems is the increase of power consumption. Because many multipliers are used in CNN cells, the relization of very low power multiplier is most effective solution for the problem of the power consumption described previously. In this thesis, a low-voltage and low-power multiplier using MOSFET's operated in the weakinversion region is proposed. In the weak inversion region, the drain-to-source current of MOSFET is very small as compared with a strong reversal region, and is known that the exponential characteristic is shown. I design the four-quadrant multiplier based on the characteristics and tranlinear principle. Finally, the CNN cell using the proposed circuits is designed and is evaluated using HSPICE simulation. The results of the evaluation are shown in this thesis.
2007-08-30