An Improved Belief Propagation Algorithm Based on Exponential Model

This paper proposes a new belief propagation algorithm to reduce the computational complexity of Low Density Parity Check codes. In the iterative process of the proposed algorithm, an exponential model is adopted to segmental approximate the hyperbolic tangent function. In addition, two kinds of nonlinear operations are simplified into one nonlinear operation. Simulations reveal that this proposed decoding algorithm can reduce the computational complexity of decoding algorithm with a minimal loss of performance.