Decoding rate-compatible 5G-LDPC codes with coarse quantization using the information bottleneck method

Publikationstyp
Journal Article
Publikationsdatum
2020-05-12
Sprache
English
Author
Stark, Maximilian orcid-logo
Wang, Linfang 
Bauch, Gerhard 
Wesel, Richard D. 
Institut
Nachrichtentechnik E-8 
DOI
10.15480/882.3584
TORE-URI
http://hdl.handle.net/11420/9102
Enthalten in
IEEE Open Journal of the Communications Society 
Volume
1
Start Page
646
End Page
660
Citation
IEEE Open Journal of the Communications Society 1: 646-660 (2020)
Publisher DOI
10.1109/OJCOMS.2020.2994048
Publisher
IEEE
Increased data rates and very low-latency requirements place strict constraints on thecomputational complexity of channel decoders in the new 5G communications standard. Practicallow-density parity-check (LDPC) decoder implementations use message-passing decoding with finiteprecision, which becomes coarse as complexity is more severely constrained. In turn, performance degradesas the precision becomes more coarse. Recently, the information bottleneck (IB) method was used to designmutual-information-maximizing mappings that replace conventional finite-precision node computations. Asa result, the exchanged messages in the IB approach can be represented with a very small number of bits.5G LDPC codes have the so-called protograph-based raptor-like (PBRL) structure which offers inherentrate-compatibility and excellent performance. This paper extends the IB principle to the flexible class ofPBRL LDPC codes as standardized in 5G. The extensions include IB decoder design for puncturing andrate-compatibility. In contrast to existing IB decoder design techniques, the proposed decoder can be usedfor a large range of code rates with a static set of optimized mappings. The proposed construction approachis evaluated for a typical range of code rates and bit resolutions ranging from 3 bit to 5 bit. Frame errorrate simulations show that the proposed scheme always outperforms min-sum decoding algorithms andoperates close to double-precision sum-product belief propagation decoding. Furthermore, alternatives tothe lookup table implementations of the mutual-information-maximizing mappings are investigated.
Schlagworte
LDPC codes
5G, message-passing decoding
mutual-information based signal processing
information bottleneck method
machine learning
DDC Class
004: Informatik
600: Technik
Publication version
updatedVersion
Lizenz
https://creativecommons.org/licenses/by/4.0/
Thumbnail Image
Name

09091572.pdf

Size

2.04 MB

Format

Adobe PDF