Evolvable Hardware (Genetic and Evolutionary Computation)

Evolvable Hardware (Genetic and Evolutionary Computation)

Xin Yao, Tetsuya Higuchi, Yong Liu

Language: English

Pages: 227

ISBN: 2:00282284

Format: PDF / Kindle (mobi) / ePub

Evolvable hardware (EHW) refers to hardware whose architecture/structure and functions change dynamically and autonomously in order to improve its performance in carrying out tasks. The emergence of this field has been profoundly influenced by the progress in reconfigurable hardware and evolutionary computation. Traditional hardware can be inflexible—the structure and its functions are often impossible to change once it is created. However, most real world problems are not fixed—they change with time. In order to deal with these problems efficiently and effectively, different hardware structures are necessary. EHW provides an ideal approach to make hardware "soft" by adapting the structure to a problem dynamically.

The contributions in this book provide the basics of reconfigurable devices so that readers will be fully prepared to understand what EHW is, why it is necessary and how it is designed. The book also discusses the leading research in digital, analog and mechanical EHW. Selections from leading international researchers offer examples of cutting-edge research and applications, placing particular emphasis on their practical usefulness.

Professionals and students in the field of evolutionary computation will find this a valuable comprehensive resource which provides both the fundamentals and the latest advances in evolvable hardware.

Operating Systems: A Spiral Approach

Wireshark 101: Essential Skills for Network Analysis

Introduction to High Performance Computing for Scientists and Engineers (Chapman & Hall/CRC Computational Science)

Linear Programming and Algorithms for Communication Networks: A Practical Guide to Network Design, Control, and Management

See MIPS Run (2nd Edition) (The Morgan Kaufmann Series in Computer Architecture and Design)

















these categories, placing particular emphasis on their practical usefiilness. Tetsuya Higuchi Yong Liu Xin Yao CONTENTS Preface v 1. Introduction to Evolvable Hardware Tetsuya Higuchi, YongLiu, Masaya Iwata andXin Yao 1 2. EHW Applied to Image Data Compression Hidenori Sakanashi, Masaya Iwata and Tetsuya Higuchi 19 3. A GA Hardware Engine and Its Applications Isamu Kajitani, Masaya Iwata and Tetsuya Higuchi 41 4. Post-Fabrication Clock-Timing Adjustment Using Genetic Algorithms

studies to specify the crossover and mutation operations and to decide population size, with a view to compact implementation. The target circuits are the memory test pattern generator (circuitl) and the multiplier (circuit2) (Takahashi, 2003). 4.2.1 Crossover and Mutation The first simulation was carried out to select a crossover operation. One of simplest crossover operations, called one-point crossover, was applied in the previous report (Takahashi, 1999), however, given the nonlinear

spectrum analyzer, which measured the amplitude of the output signal and transferred the values to the computer. The computer calculated the image-rejection ratio and ran the GA to determine new parameters. The new evolved parameters were sent as a control voltage for the mixer circuit by the DAC. We conducted an adjustment experiment to compare three approaches on improving the performance level of an image-rejection mixer: GA, IHC or manual adjustments by an experienced engineer. The results,

CMOS technologies and radiation tolerant design, Nuclear Science Symposium Conference Record, Volume: 1, 15-20 Oct. 2000, 2/2. Chen, Panxun, et al., 1991. Total dose radiation effects on the hardened CMOS^uIk and CMOS/SOS. In Radiation and its Effects on Devices and Systems. RADECS 91, First European Conference on, 9-12 Sept. 1991, 249-253. Guertin, S. M., G. M. Swift, and D. Nguyen, 1999. Single-event upset test results for the XilinxXQlTOlLPROM. In Radiation Effects Data Workshop, 1999,12-16

program tree composed of random choices of the primitive fixnctions and terminals. The initial individuals are usually generated subject to a preestablished maximum size (specified by the user as a minor parameter in the fourth preparatory step). In general, the programs in the population are of different size (number of fimctions and terminals) and of different shape (the particular graphical arrangement of fimctions and terminals in the program tree). Genetic programming iteratively transforms

Download sample