Genetic Algorithms (GAs) are one of the most advanced optimization techniques. The main objective of this paper, is introducing an FPGA implementation based genetic algorithm, then applying it, as an adaptive algorithm, on a non-linear adaptive filters for the purpose of blind signals separation. In this case, the non-linear estimator has been used to predict the error filter and GA will be used to optimize the filter coefficients through the search for a near optimum solution. The proposed Hardware Genetic Algorithms (HGA) has been presented and tested, first, by different sine wave signals, then by audio wave signals to judge the design separation capability. The implementation results declare that HGA approach significantly enhances the system performance as a step toward real time performance.