Implementation of FIR Filter & Mac Unit by Using Neural Networks in FPGA

FIR stands for Finite Impulse Response which act as signal conditioners for digital signal processor. It takes an input signal, limits the frequencies to a desired frequency range and giving the output as the subtraction of the original signal to the desired limited frequencies. The FIR filters play a vital role in wide range of DSP applications like speech processing, data compression, signal processing, image processing etc. where the noise cancellation and redundant sound cancellation is required for achieving a desired signal. So we design a FIR filter which gives better performance with less power consumption considering the resources. Implementing the design of FIR filter using (Multiplier and Accumulation Control Unit)MAC where MAC unit gives the fundamental concept of the neural networks. The proposed design gives a improved performance while in comparison with the conventional FIR filter. FIR filter is synthesized and implemented in Xilinx FPGA environment using Xilinx ISE platform evaluating the performance in terms of LUT's, FF's & power. Therefore concluding with the lower power consumption of the proposed FIR filter than conventional FIR filter.