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.

Practical Analysis of Algorithms (Undergraduate Topics in Computer Science)

Introduction to the Theory of Computation (3rd Edition)

Database Systems: Design, Implementation and Management (11th Edition)

Computational Network Science: An Algorithmic Approach




















1,000 times). Clearly, a chromosome in a population of 10 has a much greater probability of being optimized by the GA search compared to a chromosome in a population of 1,000. However, because low populations tend to yield insufficient optimization, population size must be determined carefully, considering both acceptable search duration and hardware size. 3. A GA Hardware Engine and Its Applications 59 Figure 3-14 shows a t5^ical example of operating yields as a function of population size.

generator) and a mutation circuit. The mutation circuit performs the Gaussian mutation by adding a mutation mask (a Gaussian random number). In this application, the distribution range of the Gaussian random numbers is set from -A to +3 (three-bit signed binary numbers), and these can be generated by the summation of eight uniformly distributed random bit-strings (fixed-point random numbers). 4.3.4 Implemented Hardware In this chapter, the designed hardware was implemented on an FPGA, in order

(right diagram in Figure 6-2). Contrary to the dynamic routing, the configuration bits of the switchbox are loaded and fixed at configuration time, and can be changed only by a full or a partial reconfiguration of the circuit. 2.1.2 Partial Reconfiguration Partial reconfiguration is one of the main innovations in the molecular substrate. This feature not only allows any molecule to change the configuration of another molecule, but also allows the user to decide specifically which parts of the

Transconductance Amplifiers (OTA) operated at low temperatures (Terry, et al., 2004). Both of these hardening approaches are limited, in particular for analog electronics, by the fact that current designs are fixed and, as components are affected by EE, these drifts alter fimctionality. A recent approach pioneered by JPL is to mitigate drifts, degradation, or damage on electronic devices in EE by using reconfigurable devices and an adaptive self-reconfiguration of circuit topology. This new

JBIG2 standard. Key words: data compression, JBIG2 (Joint Bi-level Image experts Group, 2), ISO/IEC standard, EHW chip. 1. INTRODUCTION Since the emergence of Desk Top Publication (DTP) in the graphic (imaging, printing and publishing) industry, digital image data has been handled in many ways, such as with digital printers, on-demand printing/publishing (ODP), and so on. On the other hand, the large costs for storage and transfer of an enormous amount of huge images have become a serious

Download sample