A second order approximation to reduce the complexity of LDPC decoders based on Gallager’s approach

Abstract

A piece-wise second order approximation to the f (x) = −log [tanh (x/2)] function is proposed to reduce the computational complexity of LDPC decoder’s utilizing Log-Likelihood Ratio Belief Propagation (LLRBP) algorithm based on Gallager’s approach. Simulation results show that the proposed low complexity approximation doesn’t cause BER performance degradation.