Value of Supporting Class-of-Service in IP Backbones

@article{Yuksel2007ValueOS,
  title={Value of Supporting Class-of-Service in IP Backbones},
  author={Murat Yuksel and K. K. Ramakrishnan and Shivkumar Kalyanaraman and Joseph D. Houle and Rita Sadhvani},
  journal={2007 Fifteenth IEEE International Workshop on Quality of Service},
  year={2007},
  pages={109-112}
}
The desire or ability of an ISP to provide differentiated service is a current hotly debated topic. In this paper, we quantify the value of having differentiated service (i.e., class-of-service (CoS)) support in an IP backbone. We compare the capacity requirements of a Diffserv environment providing service for applications that require delay or loss assurances in comparison to a network that provides classless (i.e., best-effort) service and still has to meet the same performance assurances… 

Figures from this paper

Required Extra Capacity : Estimating Overprovisioning in a Single Class-of-Service IP Backbone
TLDR
This paper quantifies the extra capacit y requirement for an over-provisioned classless network compared to a CoS network, and shows that the impac t of CoS on best-effort traffic is relatively small while still providing the desired performance for premium traffic.
Class-of-service in ip backbones: informing the network neutrality debate
TLDR
This work quantifies the extra capacity requirement for an over-provisioned classless network compared to a CoS network providing the same delay or loss performance for premium traffic and shows that REC can be significant even when the proportion of premium traffic is small.
On flat-rate and usage-based pricing for tiered commodity internet services
TLDR
This note discusses issues pertaining to end-to-end quality- of-service management of commodity Internet applications and associated pricing incentive mechanisms, and suggests that service differentiation may be more beneficial in lower capacity access networks than in high capacity core networks.
Addressing Multiservice Classes and Hybrid Architecture in WiMax Networks
TLDR
It is argued that using only polling based priority scheduling at subscriber stations and demand based dynamic bandwidth allocation at the base station it is possible to serve various traffic types in WiMax systems with only three service classes rather than four as proposed in the standard.
Blocking Probability in Pipeline Forwarding Networks
TLDR
This work analyses blocking probability of reservation requests since it affects the capability of utilizing network resources to carry traffic with deterministic QoS and analytically derived on general topology PF networks as function of the traffic intensity given the traffic matrix and the network routing.
An analytical approach for traffic modeling in mixed service architectures
TLDR
This paper will look into traffic with variable intensity which can effectively be described using Markov Modular Poisson Processes (MMPP), and new analytical model is offered, and graphs will show losses function depending on network parameters.
Network neutrality and the evolution of the internet
In order to create incentives for internet traffic providers, not to discriminate with respect to certain applications on the basis of network capacity requirements, the concept of market driven
Costs of Neutral/Unmanaged IP Networks
TLDR
This paper examines the current costs of unmanaged/neutral network technology and demonstrates that as customer Internet usage patterns evolve and become both more bandwidth-intensive and real-time oriented, a model of pure neutrality would be extremely expensive for consumers.
...
...

References

SHOWING 1-10 OF 20 REFERENCES
Cadenus: creation and deployment of end-user services in premium IP networks
TLDR
The Cadenus project is to bring theoretical and practical contributions to this area by defining a framework for the provisioning of advanced communication services in premium IP networks by comprising both the design of the proposed framework and its actual implementation.
A quantitative study of differentiated services for the Internet
TLDR
This paper compares the loss and delay behaviors that can be provided using the services based on combinations of two router mechanisms, threshold dropping and priority scheduling and two packet marking mechanisms, edge-discarding and edge-marking and shows that a significant improvement in throughput can be achieved.
Does over-provisioning become more or less efficient as networks grow larger?
TLDR
This paper investigates the influence that increasing network size has on the efficacy of over-provisioning in absorbing a certain range of traffic variations and preserving performance guarantees and develops a general model that accounts for network topology, base offered traffic, and traffic variations.
A new direction for quality of service: flow-aware networking
TLDR
It is argued that FAN is a new direction for Internet quality of service that is arguably necessary as an alternative to flawed classical architectures and sufficient to meet user performance requirements in a cost-effective way.
A network service curve approach for the stochastic analysis of networks
TLDR
The presented network service curve permits the calculation of statistical end-to-end delay and backlog bounds for broad classes of arrival and service distributions and is shown to be bounded by O(Hlog H), where H is the number of nodes traversed by a flow.
An Architecture for Differentiated Service
TLDR
An architecture for implementing scalable service differentiation in the Internet achieves scalability by aggregating traffic classification state which is conveyed by means of IP-layer packet marking using the DS field [DSFIELD].
An Architecture for Differentiated Services
TLDR
An architecture for implementing scalable service differentiation in the Internet is defined, which achieves scalability by aggregating traffic classification state which is conveyed by means of IP-layer packet marking using the DS field [DSFIELD].
Measuring ISP topologies with Rocketfuel
TLDR
New Internet mapping techniques that have enabled us to measure router-level ISP topologies are presented, finding that these maps are substantially more complete than those of earlier Internet mapping efforts.
Traffic matrix estimation: existing techniques and new directions
TLDR
This work provides a proof of concept showing that the incorporation of knowledge of POP features can reduce estimation errors and proposes a new direction for TM estimation based on using choice models to model POP fanouts.
Why is the internet traffic bursty in short time scales?
TLDR
It is shown that TCP self-clocking, joint with queueing in the network, can shape the packet interarrivals of a TCP connection in a two-level ON-OFF pattern, which creates strong correlations and burstiness in time scales that extend up to the Round-Trip Time of the connection.
...
...