Coded Distributed Computing Design for Graph MapReduce System

Disclosed are coding distributed computing technique and computer device in a graph MapReduce system for large-scale graph algorithm operation. The coding distributed computing method includes a step of shuffling intermediate operation values through communication between nodes for a reduce operatio...

Full description

Saved in:
Bibliographic Details
Main Authors WAN CHOI, SONG HEEKANG, PARK HYUNCHEOL
Format Patent
LanguageEnglish
Korean
Published 09.05.2024
Subjects
BASIC ELECTRONIC CIRCUITRY
CALCULATING
CODE CONVERSION IN GENERAL
CODING
COMPUTING
COUNTING
DECODING
ELECTRIC COMMUNICATION TECHNIQUE
ELECTRIC DIGITAL DATA PROCESSING
ELECTRICITY
PHYSICS
TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHICCOMMUNICATION
Online AccessGet full text

Cover

More Information
Summary:Disclosed are coding distributed computing technique and computer device in a graph MapReduce system for large-scale graph algorithm operation. The coding distributed computing method includes a step of shuffling intermediate operation values through communication between nodes for a reduce operation after a map operation in a graph-based MapReduce. The shuffling step may include a step of performing subgraph allocation and reduce function allocation for each distributed node based on embedded index coding. 대규모 그래프 알고리즘 연산을 위한 그래프 맵리듀스 시스템에서의 부호화 분산 컴퓨팅 기법 및 컴퓨터 장치가 개시된다. 부호화 분산 컴퓨팅 방법은, 그래프 기반 맵리듀스(graph-based MapReduce)에서 맵(map) 연산 이후 리듀스(reduce) 연산을 위해 노드 간의 통신을 통해 중간 연산 값을 셔플(shuffle)하는 단계를 포함하고, 상기 셔플하는 단계는, 임베디드 인덱스 부호화(embedded index coding)를 기초로 각 분산 노드에 대한 서브그래프 할당(subgraph allocation) 및 리듀스 함수 할당(reduce allocation)을 수행하는 단계를 포함할 수 있다.
Bibliography:Application Number: KR20220144651