Degree-dependent decoding of LDPC codes

In this paper we present a new method to reduce the decoding complexity of LDPC codes. The variable nodes are grouped according to their degrees, and in each iteration only a subset of variable node groups is activated. The variable and check node message updates are associated with a certain cost, and the activation is then optimized such that the overall cost is minimized. The optimum activation order is determined by density evolution. The complexity savings are illustrated by some examples.