Improving round-trip time estimates in reliable transport protocols
SIGCOMM '87 Proceedings of the ACM workshop on Frontiers in computer communications technology
pTCP: An End-to-End Transport Layer Protocol for Striped Connections
ICNP '02 Proceedings of the 10th IEEE International Conference on Network Protocols
A Transport Layer Load Sharing Mechanism for Mobile Wireless Hosts
PERCOMW '04 Proceedings of the Second IEEE Annual Conference on Pervasive Computing and Communications Workshops
Concurrent multipath transfer using SCTP multihoming over independent end-to-end paths
IEEE/ACM Transactions on Networking (TON)
A transport layer approach for improving end-to-end performance and robustness using redundant paths
ATEC '04 Proceedings of the annual conference on USENIX Annual Technical Conference
Striping within the network subsystem
IEEE Network: The Magazine of Global Internetworking
| Hi-index | 0.01 |
New-age mobile devices are often equipped with multiple network interfaces, which can be exploited simultaneously by sharing a single traffic load over more than one available path. However, parallel data transfers introduce a serious risk to in-order data delivery for transport layer primitives. This is principally due to the difference in characteristics of one path from the other. It becomes crucial that the data be received in-order to avoid fast-retransmissions and head-of-line blocking problem at the transport layer. In this paper, we propose an intelligent transport layer scheduling mechanism which stripes data from one flow to multiple distinct paths and aggregates available resources efficiently. Our mechanism first estimates the arrival times of each packet to be received at the receiver. It then schedules the outgoing packets in a way as to avoid data re-ordering at the receiver. Simulation results show that our mechanism efficiently reduces out-of-order data delivery at reception. Comparing our proposal with Concurrent Multipath Transfer (CMT), an SCTP based load-sharing technique; we show that it outperforms CMT in case of disparate path delays.