Johnson graph codes

Array codes are the preferred codes for distributed storage, such that different rows in an array are stored at different nodes. Layered codes use a sparse format for stored arrays with a single parity check per column and no other parity checks. Remarkably, the simple structure of layered codes is...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Designs, codes, and cryptography Ročník 90; číslo 12; s. 2923 - 2941
Hlavní autori: Duursma, Iwan, Li, Xiao
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: New York Springer US 01.12.2022
Springer Nature B.V
Predmet:
ISSN:0925-1022, 1573-7586
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Array codes are the preferred codes for distributed storage, such that different rows in an array are stored at different nodes. Layered codes use a sparse format for stored arrays with a single parity check per column and no other parity checks. Remarkably, the simple structure of layered codes is optimal when data is collected from all but one node. Codes that collect data from fewer nodes include improved layered codes, determinant codes, cascade codes and moulin codes. As our main result we show that the concatenation of layered codes with suitable outer codes achieves the performance of cascade and moulin codes which is conjectured to be optimal for general regenerating codes. The codes that we use as outer codes are in a new class of codes that we call Johnson graph codes. The codes have properties similar to those of Reed–Muller codes. In both cases the topological structure of the set of coordinates can be used to identify information sets and codewords of small weight.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0925-1022
1573-7586
DOI:10.1007/s10623-021-01003-1