On the Construction of Rate-Compatible Punctured Polar (RCPP) Codes Using Hierarchical Puncturing

In this paper, we present an efficient method to construct a rate-compatible punctured polar (RCPP) code. In the proposed method, an information set is simply optimized only considering the highest rate and then the information set is updated for other codes in the family, by satisfying the rate-compatibility constraint. This is enabled by presenting a novel hierarchical puncturing and information-copy technique. In fact, the updated information sets are represented by the original information set and an informationdependent frozen vector. Namely, the proposed code uses a non-zero frozen vector differently from conventional RCPP codes. Via simulation results, we demonstrate that the proposed RCPP code attains a significant performance gain (about 2dB) over a benchmark RCPP code, in which both codes use the identical puncturing patterns but the latter employs the conventional all-zero frozen vector. Thus, the proposed approach would be crucial to design a good RCPP code efficiently.