Synthesis of analog filters on an evolvable hardware platform using a genetic algorithm
Proceedings of the 9th annual conference on Genetic and evolutionary computation
A fast and elitist multiobjective genetic algorithm: NSGA-II
IEEE Transactions on Evolutionary Computation
| Hi-index | 0.00 |
This work presents an approach based on Genetic Programming for the synthesis of continuous-time analog filters on a field-programmable analog array. A multi-objective algorithm is used to synthesize both the topology and parameter values of Gm-C filter structures to be instantiated on the FPAA based on a given filter specification. The presented algorithm is highly adapted to the underlying hardware platform, with the goal of making an efficient crossover of high-quality building blocks possible without biasing certain types of schemata. By manipulating of the program tree in the instantiation phase, it is assured that the resulting synthesized structure fits within the physical constraints of the underlying hardware platform.