Mutual information based polar code construction

Patent metadata

Field	Value
Publication number	US-11057151-B2
Application number	US-201716487050-A
Country	US
Kind code	B2
Filing date	Dec 14, 2017
Priority date	Mar 14, 2017
Publication date	Jul 6, 2021
Grant date	Jul 6, 2021

Abstract

Methods, systems, and devices for wireless communication are described for mutual information based polar code construction. A wireless device may receive a codeword over a wireless channel, the codeword encoded using a polar code. The wireless device may identify a set of bit locations of the polar code corresponding to information bits of an encoded information bit vector. The set of bit locations may be determined based at least in part on a reliability order of the bit locations of the polar code, and the reliability order may be determined based at least in part on a recursive model that includes at least one mutual information transfer function that is applied at each polarization stage of multiple polarization stages of the polar code. The wireless device may decode the received codeword to obtain the information bit vector at the set of bit locations.

First claim

What is claimed is:

1. A method for wireless communication, comprising:
receiving a codeword over a wireless channel, the codeword encoded using a polar code;
identifying a set of bit locations of the polar code corresponding to information bits of an encoded information bit vector, wherein the set of bit locations is determined based at least in part on a reliability order of the bit locations of the polar code, and wherein the reliability order is determined based at least in part on a recursive model that includes at least one mutual information transfer function that is applied at each polarization stage of a plurality of polarization stages of the polar code; and
decoding the received codeword to obtain the information bit vector at the set of bit locations.

2. The method of claim 1 , further comprising:
calculating mutual information of a first subset of the bit locations for a current polarization stage of the plurality of polarization stages based at least in part on mutual information of a second subset of the bit locations from a preceding polarization stage of the plurality of polarization stages.

3. The method of claim 1 , wherein the reliability order of the bit locations of the polar code is constructed based at least in part on the recursive model and a plurality of different target mutual information values.

4. The method of claim 3 , further comprising:
performing a sweep over the plurality of different target mutual information values to determine at least portions of respective output mutual information value curves for each of the bit locations of the polar code, and
determining the reliability order of the bit locations of the polar code corresponding to an order in which each of the respective output mutual information value curves reaches a defined output mutual information value with respect to the plurality of different target mutual information values.

5. The method of claim 3 , wherein a first subset of the set of bit locations is determined based at least in part on a first reliability order determined from applying the recursive model for a first target mutual information value and a second subset of the set of bit locations is determined based at least in part on a second reliability order determined from applying the recursive model for a second target mutual information value.

6. The method of claim 5 , wherein the first target mutual information value is based at least in part on a first number of information bits to be encoded and a number of the bit locations, and the second target mutual information value is based at least in part on a second number of information bits to be encoded and the number of the bit locations.

7. The method of claim 1 , wherein a first bit location of the bit locations corresponds to a punctured bit.

8. The method of claim 7 , wherein mutual information of the first bit location at a first polarization stage is calculated based at least in part on the at least one mutual information transfer function and mutual information associated with the punctured bit.

9. The method of claim 8 , wherein the mutual information associated with the punctured bit is set to zero for non-shortening based puncturing or is set to a unity value for shortening-based puncturing.

10. The method of claim 7 , wherein a target mutual information for bit channels corresponding to bit locations other than the punctured bit is determined based at least in part on a number of information bits to be encoded and a number of the bit locations excluding a bit channel associated with the punctured bit.

11. The method of claim 1 , wherein the at least one mutual information transfer function is based at least in part on a binary erasure channel (BEC) function and a correction term.

12. The method of claim 11 , wherein the correction term is based at least in part on a bit channel capacity of the each polarization stage and a capacity imbalance factor.

13. The method of claim 12 , wherein the correction term comprises an offset factor applied to the bit channel capacity.

14. The method of claim 13 , wherein the correction term comprises a scaling factor applied to the offset bit channel capacity.

15. The method of claim 14 , wherein the correction term comprises an offset applied to the scaled and offset bit channel capacity.

16. The method of claim 1 , wherein the codeword comprises at least one repeated bit of a bit location of the polar code, and wherein a target mutual information for bit channels of the polar code is determined based at least in part on a number of information bits to be encoded and a number of bit locations in the codeword including the repeated bit location.

17. The method of claim 16 , wherein an offset is applied to the target mutual information for the bit channels.

18. The method of claim 1 , wherein the recursive model is a Gaussian approximation density-based evolution model, and wherein the reliability order of the bit locations of the polar code is constructed based at least in part on the Gaussian approximation density-based evolution model and a plurality of different target signal to noise ratio (SNR) values.

19. The method of claim 18 , further comprising:
performing a sweep over the plurality of different target SNR values to determine at least portions of respective error rate value curves for each of the bit locations of the polar code, and
determining the reliability order of the bit locations of the polar code corresponding to an order in which each of the respective error rate value curves reaches a defined error rate value with respect to the plurality of different target SNR values.

20. The method of claim 1 , wherein identifying the set of bit locations of the polar code corresponding to the information bits of the encoded information bit vector comprises:
retrieving the set of bit locations from a table that includes a plurality of sets of bit locations.

21. The method of claim 1 , further comprising:
storing a table that lists a plurality of sets of bit locations, wherein identifying the set of bit locations of the polar code corresponding to the information bits of the encoded information bit vector comprises:
determining that a first bit location of the bit locations corresponds to a punctured bit;
determining not to retrieve any of the plurality of sets of bit locations from the table; and
applying the recursive model by a decoder to identify the set of bit locations.

22. A method for wireless communication, comprising:
identifying an information bit vector for encoding using a polar code;
identifying a set of bit locations of the polar code to be used for information bits of the information bit vector, wherein the set of bit locations is determined based at least in part on a reliability order of the bit locations of the polar code, and wherein the reliability order is determined based at least in part on a recursive model that includes at least one mutual information transfer function that is applied at each polarization stage of a plurality of polarization stages of the polar code;
encoding the information bit vector mapped to the set of bit locations using the polar code to obtain a codeword; and
transmitting the codeword over a wireless channel.

23. The method of claim 22 , further comprising:
calculating mutual information of a first subset of the bit locations for a current polarization stage of the plurality of polarization stages based at least in part on mutu

Assignees

• Qualcomm Inc
• Yang Yang
• Jiang Jing
• Sarkis Gabi
• Wei Chao
• Li Jian
• Xu Changlong
• Sankar Hari
• Soriaga Joseph Binamira

Inventors

• Yang Yang
• Jiang Jing
• Sarkis Gabi
• Wei Chao
• Li Jian
• Xu Changlong
• Sankar Hari
• Soriaga Joseph Binamira

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• H03M13/6356

  by repetition or insertion of dummy data, i.e. rate reduction · CPC title

• H04L1/0041

  Arrangements at the transmitter end · CPC title

• H04L1/0057 Primary

  Block codes (H04L1/0061, H04L1/0064 take precedence) · CPC title

• H04L1/0068

  by puncturing · CPC title