Design and decoding of polar-like codes
Event details
| Date | 29.06.2018 |
| Hour | 10:00 › 12:00 |
| Speaker | Kirill Ivanov |
| Location | |
| Category | Conferences - Seminars |
EDIC candidacy exam
Exam president: Dr. Andreas Burg
Thesis advisor: Prof. Ruediger Urbanke
Co-examiner: Prof. Emre Telatar
Abstract
Polar codes are the first class of capacity-achieving codes with low-complexity encoding and decoding algorithms. However, their performance on finite lengths is quite poor due to the small minimal distance. Generalized Plotkin decomposition allows one to employ low-complexity polar code decoding techniques for decoding powerful Reed-Muller and eBCH codes. However, their performance under successive-cancellation (list) decoding is significantly worse for small and moderate list sizes.
In this proposal, we examine various approaches to improve the decoding performance of codes which can be defined via polar-like decomposition, with a special case of Reed-Muller codes. These techniques include both code construction and decoding methods. We point out their drawbacks and suggest some ideas, which can be used in order to get better performance.
Background papers
List Decoding of Polar Codes, by I. Tal, A. Vardy.
Polar subcodes, by P. Trifonov, V. Miloslavskaya.
Decoding Reed-Muller Codes Using Minimum-Weight Parity Checks, by E. Santi, et al.
Exam president: Dr. Andreas Burg
Thesis advisor: Prof. Ruediger Urbanke
Co-examiner: Prof. Emre Telatar
Abstract
Polar codes are the first class of capacity-achieving codes with low-complexity encoding and decoding algorithms. However, their performance on finite lengths is quite poor due to the small minimal distance. Generalized Plotkin decomposition allows one to employ low-complexity polar code decoding techniques for decoding powerful Reed-Muller and eBCH codes. However, their performance under successive-cancellation (list) decoding is significantly worse for small and moderate list sizes.
In this proposal, we examine various approaches to improve the decoding performance of codes which can be defined via polar-like decomposition, with a special case of Reed-Muller codes. These techniques include both code construction and decoding methods. We point out their drawbacks and suggest some ideas, which can be used in order to get better performance.
Background papers
List Decoding of Polar Codes, by I. Tal, A. Vardy.
Polar subcodes, by P. Trifonov, V. Miloslavskaya.
Decoding Reed-Muller Codes Using Minimum-Weight Parity Checks, by E. Santi, et al.
Practical information
- General public
- Free
Contact
- EDIC - [email protected]