Using OFB and LDPC Coding to Decrease MSE Produced by Gaussian and Impulse Noise

摘要

This paper presents method to reduce Mean Squared Error (MSE) produced by additive Gaussian noise and impulse noise, using Oversampled Filter Banks (OFB) and Low-Density Parity-Check (LDPC) coding. Two methods are used for the OFB design: Discrete Cosine Transform (DCT) modulation and the Discrete Fourier Transform (DFT) modulation. The same lowpass prototype Finite Impulse Response (FIR) filter is used in both approaches. The polyphase decomposition of the analysis and synthesis filters bank is performed to obtain more efficient structure. The input signal, in the transmitter, is splitted into subbands using the analysis filter bank and coded with LDPC coding. On the receiver side, the inverse process is performed. The proposed method is compared with the similar approaches from literature, using Matlab simulations.