FIR filter design for advanced audio/video processing applications

Electronic signal processors that change amplitude and/or phase according to frequency are called filters. A filter should neither add new input signal frequencies, nor should it alter signal component frequencies; instead, it should change the relative amplitudes and/or interactions of the different frequency components. Frequently, electric systems include filters because they assist in communications in the frequency range that other signals are unable to communicate. A novel approximation DA structure is designed for an adaptive FIR filter. The LMS algorithm is used to update the filter's weights. We used the CSD number with the DA method in this design to improve the filter's size and power efficiency. Without the usage of a LUT, the DA calculation is performed. Instead, registers are used to do DA computations. The Wallace tree adder is used to combine the partial results, increasing the filter's speed while decreasing its complexity. The device usage statistics for the structure indicate that the performance and power efficiency are both higher than in previous designs. The suggested structure was written in Verilog and implemented on a Zynq 7000 FPGA.