Iterative improvement methodology for hardware/software co-synthesis of embedded systems using genetic programming

In this paper we present a novel developmental genetic programming methodology for co-synthesis of distributed embedded systems. This methodology is based on iterative improvement algorithms. Genotype is represented by a tree. Unlike existing genetic programming approaches, our methodology starts from the fastest architecture. Every node in the tree is a chromosome and describes system refinement option. Thus the co-synthesis process is evolving. Using genetic operators: crossover, mutation and selection new systems are created