Abstract
Multi-Path TCP (MPTCP) has attracted much attention as a promising technology to improve throughput performance of wireless devices that support multi-homed heterogeneous networks. Although MPTCP provides significant increase in network capacity, it may suffer from poor delay performance since the delay tends to be aligned with the worst-performing path: packets delivered through a short-delay subflow have to wait in the reordering buffer for packets being transmitted over a long-delay subflow. In this paper, we investigate the application-level delay performance of streaming traffic over MPTCP, and develop an analytical framework to take into account non-negligible network queuing delay and the interplay of congestion control between multiple subflows. We design a simple threshold-based subflow traffic allocation scheme that aims to minimize user-level delay and develop a receiver-centric traffic splitting control (R-TSC) that can be tuned to user preferences. The client-side R-TSC solution facilitates incremental deployment of low-delay streaming service over MPTCP. Through simulation and testbed experiments using commercial LTE and WiFi networks, we demonstrate significant performance gains over the standard MPTCP protocol.
Original language | English |
---|---|
Title of host publication | Proceedings - 2015 IEEE 23rd International Conference on Network Protocols, ICNP 2015 |
Publisher | IEEE Computer Society |
Pages | 245-255 |
Number of pages | 11 |
ISBN (Electronic) | 9781467382953 |
DOIs | |
Publication status | Published - 2016 Mar 18 |
Externally published | Yes |
Event | 23rd IEEE International Conference on Network Protocols, ICNP 2015 - San Francisco, United States Duration: 2015 Nov 10 → 2015 Nov 13 |
Publication series
Name | Proceedings - International Conference on Network Protocols, ICNP |
---|---|
Volume | 2016-March |
ISSN (Print) | 1092-1648 |
Conference
Conference | 23rd IEEE International Conference on Network Protocols, ICNP 2015 |
---|---|
Country/Territory | United States |
City | San Francisco |
Period | 15/11/10 → 15/11/13 |
Bibliographical note
Funding Information:This work was supported by the Samsung Electronics DMC research center and IITP grant funded by the Korea government (MSIP) (No. B0126-15-1064, Research on Near- Zero Latency Network for 5G Immersive Service).
Publisher Copyright:
© 2015 IEEE.
ASJC Scopus subject areas
- Computer Networks and Communications
- Software