On bit error rate performance of polar codes in finite regime
Authors:
- 1 September 2010
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 188-194
Abstract
Polar codes have been recently proposed as the first low complexity class of codes that can provably achieve the capacity of symmetric binary-input memoryless channels. Here, we study the bit error rate performance of finite-length polar codes under Belief Propagation (BP) decoding. We analyze the stopping sets of polar codes and the size of the minimal stopping set, called “stopping distance”. Stopping sets, as they contribute to the decoding failure, play an important role in bit error rate and error floor performance of the code. We show that the stopping distance for binary polar codes, if carefully designed, grows as O(√N) where N is the code-length. We provide bit error rate (BER) simulations for polar codes over binary erasure and gaussian channels, showing no sign of error floor down to the BERs of 10 -11 . Our simulations asserts that while finite-length polar codes do not perform as good as LDPC codes in terms of bit error rate, they show superior error floor performance. Motivated by good error floor performance, we introduce a modified version of BP decoding employing a guessing algorithm to improve the BER performance of polar codes. Our simulations for this guessing algorithm show two orders of magnitude improvement over simple BP decoding for the binary erasure channel (BEC), and up to 0.3 dB improvement for the gaussian channel at BERs of 10 -6 .Keywords
This publication has 21 references indexed in Scilit:
- Polar Codes are Optimal for Lossy Source CodingIEEE Transactions on Information Theory, 2010
- Geometrically-structured maximum-girth LDPC block and convolutional codesIEEE Journal on Selected Areas in Communications, 2009
- Channel Polarization: A Method for Constructing Capacity-Achieving Codes for Symmetric Binary-Input Memoryless ChannelsIEEE Transactions on Information Theory, 2009
- A performance comparison of polar codes and Reed-Muller codesIEEE Communications Letters, 2008
- Results on Punctured Low-Density Parity-Check Codes and Improved Iterative Decoding TechniquesIEEE Transactions on Information Theory, 2007
- Weight Distribution of Low-Density Parity-Check CodesIEEE Transactions on Information Theory, 2006
- Stopping Set Distribution of LDPC Code EnsemblesIEEE Transactions on Information Theory, 2005
- On Decoding of Low-Density Parity-Check Codes Over the Binary Erasure ChannelIEEE Transactions on Information Theory, 2004
- Stopping sets and the girth of Tanner graphs2003
- Finite-length analysis of low-density parity-check codes on the binary erasure channelIEEE Transactions on Information Theory, 2002