Efficient Ranking of Rate-Compatible Puncturing Patterns for a Given LDPC Code Matrix

In this paper, we introduce a novel criterion to rank puncturing patterns for rate-compatible LDPC codes. Specifically, based on Gaussian approximation density evolution, a cost function is devised to characterize the degree distribution of the punctured code matrices, which are derived from a mother code matrix by matrix transformation. This cost function allows us to effectively compare the expected performance of candidate puncturing patterns and to sort out good ones. Combined with well-designed search algorithms, the proposed criterion can be applied on both standardized block-LDPC codes and generic binary LDPC codes to get good puncturing patterns with manageable complexity. Numerical simulation results verify the effectiveness of the proposed ranking criterion, and demonstrate that a series of good rate-compatible LDPC codes can be obtained by the proposed ranking criterion.