Early Rate Adaptation Protocol in DiffServ for Multimedia Applications
Jonghun Park, Yoo Myungsik
Abstract
Jonghun Park, Yoo Myungsik
Abstract
As the multimedia application traffic takes more portion in the internet traffic, it is necessary to control the network congestion through the congestion control protocol. In addition, the QoS-enabled networks such as DiffServ become an indispensable technology when running the multimedia applications. However, the previously proposed end-to-end congestion control algorithms take the round trip time to react the network congestion. Thus, as the RTT becomes larger, the reaction against the congestion gets delayed further, while the network congestion gets worse. In addition the performance of end-to-end congestion control algorithm is degraded if the QoS-enabled network runs the congestion control mechanism in the network level without any coordination between them. In this paper, we propose the early rate adaptation protocol for the DiffServ network which effectively linke the congestion control algorithm at the host and the congestion mechanism in the network together. By taking advantage of early congestion notification from the network it is possible to react the network congestion more quickly and effectively.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
As the multimedia application traffic takes more portion in the internet traffic, it is necessary to control the network congestion through the congestion control protocol. In addition, the QoS-enabled networks such as DiffServ become an indispensable technology when running the multimedia applications. However, the previously proposed end-to-end congestion control algorithms take the round trip time to react the network congestion. Thus, as the RTT becomes larger, the reaction against the congestion gets delayed further, while the network congestion gets worse. In addition the performance of end-to-end congestion control algorithm is degraded if the QoS-enabled network runs the congestion control mechanism in the network level without any coordination between them. In this paper, we propose the early rate adaptation protocol for the DiffServ network which effectively linke the congestion control algorithm at the host and the congestion mechanism in the network together. By taking advantage of early congestion notification from the network it is possible to react the network congestion more quickly and effectively.
Key concepts: Network congestion, Computer network, Explicit Congestion Notification, Network traffic control, Computer science, Quality of service, The Internet, Network simulation