Telecommunications Science ›› 2022, Vol. 38 ›› Issue (7): 96-105.doi: 10.11959/j.issn.1000-0801.2022168

• Research and Development • Previous Articles     Next Articles

Design of concatenated Hash-Polar code based on segmented CRC code

Zhengjie LI, Shunlan LIU, Xu ZHANG   

  1. School of Electronic Information, Hangzhou Dianzi University, Hangzhou 310018, China
  • Revised:2022-07-09 Online:2022-07-20 Published:2022-07-01
  • Supported by:
    The National Natural Science Foundation of China(U1809201);The Zhejiang Provincial Natural Science Foundation of China(LY18F010013)

Abstract:

As a linear block code, the polarization code has a lower coding complexity and a definite structure, but when it has a short or medium code length, the performance will be reduced.A design method of concatenated HashPolar codes based on segmented cyclic redundancy check (CRC) codes was proposed, which was based on the original Hash-Polar code and used CRC segmented check for double check.The segmented CRC code could assist the path measurement in the decoding process, that was, modify the decoding path, so as to improve the reliability of path selection and the performance.In addition, the segmented check was to add the check code dispersedly to the input information sequence.If the CRC failed during decoding, the decoding path could be terminated in advance to save unnecessary decoding calculations.Finally, at the end of the decoding, Hash check code verified the modified L paths and selected the best decoding path.The simulation results show that the design method proposed has better bit-error-rate performance than CRC-assisted Hash-Polar (Hash-CRC-Polar).In the Gaussian channel, when the code length is 128 bit, the code rate is 1/2, and the bit error rate is 10-3, the Hash-Polar code based on the segmented CRC code proposed achieves a gain of about 0.25 dB over Hash-CRC-Polar.

Key words: Polar code, cyclic redundancy check, Hash check, successive cancellation list decoding algorithm

CLC Number: 

No Suggested Reading articles found!