Merging multiple data streams on common keys over high performance networks
Proceedings of the 2002 ACM/IEEE conference on Supercomputing
Future Generation Computer Systems - iGrid 2002
Experimental studies using photonic data services at IGrid 2002
Future Generation Computer Systems - iGrid 2002
Transport protocols for high performance
Communications of the ACM - Blueprint for the future of high-performance networking
Data webs for earth science data
Parallel Computing - Special issue: High performance computing with geographical data
An end-to-end approach for transparent mobility across heterogeneous wireless networks
Mobile Networks and Applications
Concurrent multipath transfer using SCTP multihoming over independent end-to-end paths
IEEE/ACM Transactions on Networking (TON)
Performance enhancement of TCP in high-speed networks
Information Sciences: an International Journal
Journal of Network and Computer Applications
Self-configured multiple-network-interface socket
Journal of Network and Computer Applications
Enhancing video-on-demand playout over multiple heterogeneous access networks
CCNC'10 Proceedings of the 7th IEEE conference on Consumer communications and networking conference
Hi-index | 0.00 |
Transmission Control Protocol (TCP) is used by variousapplications to achieve reliable data transfer. TCP was originallydesigned for unreliable networks. With the emergence of high-speedwide area networks various improvements have been applied to TCPto reduce latency and achieve improved bandwidth. Theimprovement is achieved by having system administrators tune thenetwork and can take a considerable amount of time. This paperintroduces PSockets (Parallel Sockets), a library that achieves anequivalent performance without manual tuning. The basic ideabehind PSockets is to exploit network striping. By network stripingwe mean striping partitioned data across several open sockets. Wedescribe experimental studies using PSockets over the Abilenenetwork. We show in particular that network striping usingPSockets is effective for high performance data intensive computingapplications using geographically distributed data.