Thesis
Two dimensional irregular repeat-accumulate codes
Washington State University
Master of Science (MS), Washington State University
2009
Handle:
https://hdl.handle.net/2376/101119
Abstract
Irregular Repeat-Accumulate (IRA) codes are a subclass of low-density paritycheck (LDPC) codes with linear encoding complexity. They can be serially concatenated to form two dimensional block codes, called Serially Concatenated IRA (SC-IRA) codes. SC-IRA codes are constructed to address the error floor problem of LDPC codes. Previously a serial decoder structure with two component decoders was used to decode SC-IRA codes. In this thesis, the serial decoder structure is analyzed and extended. Then a novel one-graph decoding method is proposed. Simulation results show that the one-graph decoding method is better than the serial decoding method in both low and high Signal-to-Noise Ratio (SNR) regions. There is about 0.7 dB gain in the waterfall region compared to the serial decoding method. The one graph decoding method also suggests a good way to construct very long LDPC-like codes with low error floors. The performance of IRA codes and SC-IRA codes is analyzed based on a tight Maximum-Likelihood (ML) decoding upper bound. The ML bounds of several IRA codes and SC-IRA codes are computed for comparison.
Metrics
4 File views/ downloads
18 Record Views
Details
- Title
- Two dimensional irregular repeat-accumulate codes
- Creators
- Qingwei Ge
- Contributors
- Benjamin Joseph Belzer (Degree Supervisor)Krishnamoorthy Sivakumar (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; Pullman, Wash. :
- Identifiers
- 99900525051301842
- Language
- English
- Resource Type
- Thesis