We propose a novel deterministic method for source coding in peer-to-peer networks. The main advantage of our method compared to randomized methods is that the coding will be invertible with probability one. A further advantage is a much lower communication overhead due to special coding vectors. We apply the coding algorithm in a scalable and robust peer-to-peer system. We analyze the deterministic method theoretically and prove upper and lower bounds for the number coded pieces. The theoretical results are backed up by simulations.
Published in:
Computational Intelligence and Informatics (CINTI), 2011 IEEE 12th International Symposium on
Date of Conference: 21-22 Nov. 2011