Estimation of HW/SW Cost Parameters in Altera FPGA Design Environment

摘要

This paper explains the method for obtaining Hardware/Software (HW/SW) cost parameters such as hardware area, hardware time, software area and software time using Altera FPGA design environment. HW/SW partitioning of FFT and JPEG FDCT are derived using multi-objective optimization techniques Weighted Sum Genetic Algorithm (WSGA), Elitist Non-dominated sorting Genetic Algorithm (ENGA) and Multi-Objective Particle Swarm Optimization using Crowding Distance (MOPSO-CD) algorithms. Experimental results show that ENGA is effective in obtaining the HW/SW partition that obtains both minimum area and minimum time for both the applications.