Frame-counter scheduler: A novel QoS scheduler for real-time traffic

Authors:
S. Ece (Guran) Schmidt;Hyong S. Kim
Affiliations:
Electrical and Computer Engineering, Carnegie Mellon University, 5000 Forbes Avenue Pittsburgh, PA 15213-3890, USA;Electrical and Computer Engineering, Carnegie Mellon University, 5000 Forbes Avenue Pittsburgh, PA 15213-3890, USA
Venue:
Computer Communications
Year:
2006

Citing 6
Cited 1

Design and analysis of frame-based fair queueing: a new traffic scheduling algorithm for packet-switched networks

Proceedings of the 1996 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Providing deterministic delay guarantees in ATM networks

IEEE/ACM Transactions on Networking (TON)
End-to-End Scheduling in Real-Time Packet-Switched Networks

ICNP '96 Proceedings of the 1996 International Conference on Network Protocols (ICNP '96)
Scheduling and switching architecture for virtual synchronous frame (vsf)

Scheduling and switching architecture for virtual synchronous frame (vsf)
Continuous framing mechanism for congestion control in broadband networks

Computer Communications
A scheme for real-time channel establishment in wide-area networks

IEEE Journal on Selected Areas in Communications

A new scalable service discipline for real-time traffic: The framed-deadline scheduler

Computer Communications

Quantified Score

Hi-index	0.24

Visualization

Abstract

Real-time traffic communication has Quality of Service (QoS) requirements such as end-to-end bandwidth and delay guarantees. We propose a novel frame-based QoS Scheduler, the frame-counter scheduler, for connection oriented packet switching networks. The frame-counter scheduler significantly reduces the end-to-end delay bound and buffer requirements provided by other frame-based schedulers. A fixed amount of buffer is required per node for no packet-loss operation. There is no need for frame synchronization or inter-node communication. The scheduling complexity of frame-counter scheduler is O(1) which makes it possible to implement it for high-speed networks. The required input traffic shape is not more restrictive than the traffic shapes used by the other schedulers. In this paper, we present the proof for the end-to-end delay bound and the buffer requirement for the frame-counter scheduler. We also provide simulation results to demonstrate the average performance which show that the average end-to-end delay and delay variation (jitter) of the packets is much lower than the end-to-end bound.