ANALYSIS OF CASCADED MULTIRATE FIR FILTER STRUCTURES WITH CSE AND CSD

Abstract

Decimation and interpolation play vital role in multirate signal processing for sampling rate conversion(SRC). Area optimization is also very important to meet efficient filter structures in multirate signal processing systems. This paper proposes an area efficient multirate FIR filter structure for sampling rate conversion(SRC) from Digital Audio Tape (DAT) to Compact Disc (CD) by cascading three stages of linear phase multirate FIR filters with different sampling rates. The filter structures are implemented using two different approaches, namely, coefficient symmetry and polyphase approach. Area reduction in these structures is achieved by using Common Subexpression Elimination(CSE) technique with CSD (Canonic Signed Digit) and binary representations of filter coefficients. The proposed designs are implemented using MATLAB Simulink model and the Verilog code is generated using HDL coder. The performance of proposed structures is analyzed using the Altera Quartus tool and the results are compared with conventional polyphase and coefficient symmetry approaches in terms of area and delay.