Adaptive hardware evolution under unpredictable environmental changes

Abstract

This paper suggests new methods for hardware evolution (evolvable hardware, EHW) where the hardware structure is adaptable to the environment. The hardware structure is evolved in real-time by genetic algorithm to obtain a desired function of the digital circuit. In order to operate the EHW efficiently by using genetic algorithm processor (GAP), an encoding technique is employed. Proposed encoding technique shortens the chromosome length and increases the processing speed, as well. Some combinational circuits have been evolved successfully on XC6216 FPGA intrinsically.