TY - GEN
T1 - Smart content delivery on the Internet
AU - Shen, Hong
PY - 2008
Y1 - 2008
N2 - As the ever-growing Internet applications in the emerging e-society are centered in sharing of various types of digital contents, content delivery on the Internet as a hot topic has been attracting most attentions. Web caching is an important technology for improving the efficiency of content delivery. As an emerging technology en-route caching computes locations among caches on access paths to store copies of an object such that specified objectives (such as satisfying future access demands to the object) are achieved. In this talk, I will first give an overview on recent developments in efficient content delivery. I will then introduce our recent work in tackling this problem by applying the technique of en-route caching, and present efficient solutions to the problem in systems containing single server and multiple servers respectively. While the problem is NP-hard in the general case, our solutions apply dynamic programming technique and run in polynomial-time in the worst case for both unconstrained and constrained cases in the network topologies of trees (for broadcast delivery) and linear arrays (for point-to-point delivery) respectively. They are shown theoretically either optimal or convergent to optimal. Finally, I will show some possible extensions of our solutions to other system settings.
AB - As the ever-growing Internet applications in the emerging e-society are centered in sharing of various types of digital contents, content delivery on the Internet as a hot topic has been attracting most attentions. Web caching is an important technology for improving the efficiency of content delivery. As an emerging technology en-route caching computes locations among caches on access paths to store copies of an object such that specified objectives (such as satisfying future access demands to the object) are achieved. In this talk, I will first give an overview on recent developments in efficient content delivery. I will then introduce our recent work in tackling this problem by applying the technique of en-route caching, and present efficient solutions to the problem in systems containing single server and multiple servers respectively. While the problem is NP-hard in the general case, our solutions apply dynamic programming technique and run in polynomial-time in the worst case for both unconstrained and constrained cases in the network topologies of trees (for broadcast delivery) and linear arrays (for point-to-point delivery) respectively. They are shown theoretically either optimal or convergent to optimal. Finally, I will show some possible extensions of our solutions to other system settings.
UR - http://www.scopus.com/inward/record.url?scp=46049108971&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-69501-1_1
DO - 10.1007/978-3-540-69501-1_1
M3 - Conference contribution
AN - SCOPUS:46049108971
SN - 9783540695004
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 1
BT - Algorithms and Architectures for Parallel Processing - 8th International Conference, ICA3PP 2008, Proceedings
T2 - 8th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2008
Y2 - 9 June 2008 through 11 June 2008
ER -