Title :
An extended least difference greedy clique-cover algorithm for index coding
Author :
Sangwoon Kwak ; Jungho So ; Youngchul Sung
Author_Institution :
Dept. of Electr. Eng., KAIST, Daejeon, South Korea
fDate :
June 29 2014-July 4 2014
Abstract :
In this paper, linear binary index coding is considered. It is shown that the minimum clique-cover heuristic algorithm can provide an efficient way to solving linear binary index coding problems. Based on the least difference greedy (LDG) clique-cover algorithm, an existing minimum clique-cover algorithm for index coding, proposed by Birk and Kol [1], [2], we develop an extended LDG algorithm by considering a transpose index coding model and cycle detection in the side information graph. Numerical results show that the proposed algorithm considerably outperforms the conventional LDG algorithm in terms of the number of transmissions.
Keywords :
binary codes; graph theory; linear codes; cycle detection; extended LDG clique-cover algorithm; extended least difference greedy clique-cover algorithm; linear binary index coding problem; minimum clique-cover heuristic algorithm; side information graph; transpose index coding model; Encoding; Heuristic algorithms; Indexes; Merging; Receivers; Servers;
Conference_Titel :
Information Theory (ISIT), 2014 IEEE International Symposium on
Conference_Location :
Honolulu, HI
DOI :
10.1109/ISIT.2014.6874883