vlsi-compatible implementations for artificial neural networks
smith,fakhraie
Synopsis "vlsi-compatible implementations for artificial neural networks"
vlsi-compatible implementations for artificial neural networks introduces the basic premise of the authors approach to biologically-inspired and vlsi-compatible definition, simulation, and implementation of artificial neural networks. in addition, the book develops a set of guidelines for general hardware implementation of anns. these guidelines are then used to find solutions for the usual difficulties encountered in any potential work, and as guidelines by which to reach the best compromise when several options exist. furthermore, system-level consequences of using the proposed techniques in future submicron technologies with almost-linear mos devices are discussed. while the major emphasis in this book is to develop neural networks optimized for compatibility with their implementation media, the work has also been extended to the design and implementation of a fully-quadratic ann based on the desire to have network definitions epitomized for both efficient discrimination of closed-boundary circular areas and ease of implementation in a cmos technology. vlsi-compatible implementations for artificial neural networks implements a comprehensive approach which starts with an analytical evaluation of specific artificial networks. this provides a clear geometrical interpretation of the behavior of different variants of these networks. in combination with the guidelines developed towards a better final implementation, these concepts have allowed the authors to conquer various problems encountered and to make effective compromises. then, to facilitate the investigation of the models needed when more difficult problems must be faced, a custom simulating program for various cases is developed. finally, in order to demonstrate the authors findings and expectations, several vlsi integrated circuits have been designed, fabricated, and tested. vlsi-compatible implementations for artificial neural networks serves as an excellent reference source and may be used as a text for advanced courses on the subject.
