VLSI Implementation of Cellular Neural Networks
VLSI Implementation of Cellular Neural Networks
复制标题
细胞神经网络的 VLSI 实现
DOI：
10.1007/978-1-4757-2580-3_4
复制
发表时间：
1997
期刊：
影响因子：
--
通讯作者：
M. Steyaert
中科院分区：
文献类型：
--
作者：
P. Kinget;M. Steyaert
在VLSI中实现模拟并行计算电路需要将系统模型转换为执行计算的等效电路;电路结构必须设计为使得描述电路行为的微分方程等效于系统的微分方程。此外，期望高电路密度，使得设计者必须提出消耗尽可能少的功率并且具有高速操作的紧凑电路解决方案。为了能够使用高密度标准数字CMOS技术，模拟电路也应该与这些CMOS技术完全兼容。
The realization of analog parallel computation circuits in VLSI requires the translation of the system model into an equivalent electrical circuit that performs the computations; the circuit structure must be designed such that the differential equation describing the behavior of the circuit is equivalent to the differential equation of the system. Moreover, a high circuit density is desired so that the designer has to come up with compact circuit solutions which consume as less power as possible and have a high speed operation. In order to have access to the high density standard digital CMOS technologies, the analog circuitry should also be fully compatible with these CMOS technologies.