English

Capacity-Achieving Rate-Compatible Polar Codes for General Channels

Information Theory 2017-01-24 v2 math.IT

Abstract

We present a rate-compatible polar coding scheme that achieves the capacity of any family of channels. Our solution generalizes the previous results [1], [2] that provide capacity-achieving rate-compatible polar codes for a degraded family of channels. The motivation for our extension comes from the fact that in many practical scenarios, e.g., MIMO systems and non-Gaussian interference, the channels cannot be ordered by degradation. The main technical contribution of this paper consists in removing the degradation condition. To do so, we exploit the ideas coming from the construction of universal polar codes. Our scheme possesses the usual attractive features of polar codes: low complexity code construction, encoding, and decoding; super-polynomial scaling of the error probability with the block length; and absence of error floors. On the negative side, the scaling of the gap to capacity with the block length is slower than in standard polar codes, and we prove an upper bound on the scaling exponent.

Keywords

Cite

@article{arxiv.1611.01199,
  title  = {Capacity-Achieving Rate-Compatible Polar Codes for General Channels},
  author = {Marco Mondelli and S. Hamed Hassani and Ivana Marić and Dennis Hui and Song-Nam Hong},
  journal= {arXiv preprint arXiv:1611.01199},
  year   = {2017}
}

Comments

7 pages, 2 figures, accepted at WCNC'17 workshop on polar coding

R2 v1 2026-06-22T16:41:37.699Z