Problem
Statement and Motivation - People - Publications
The
user expectation of Internet use has moved from best-effort connectivity to an
expectation of reasonable performance & capacity for all types of
applications. QoS-sensitive applications like IPTV,
gaming, and VoIP could be offered over such a converged IP broadband end-to-end
network. Network service providers also would like to support such applications
effectively. They need to provision their networks to meet the service level
agreements (SLAs). Customer experiences need to be
protected and predictable despite network failures and changes in demand as
well as application mix.
Currently
there is a wide ranging debate on the issue of
“network neutrality” which involves both economic and technical aspects. One
key aspect of the net neutrality debate is whether best effort application
traffic should be carried along with other (so-called “premium”) traffic for
which SLA commitments have been made (or are expected, either explicitly or
implicitly) without differentiation. An assertion often made in this context is
that over-provisioning is an economically viable strategy due to the declining
cost of capacity, instead of incurring the complexity and operational costs of
running a differentiated-services network. Our study focuses on this specific
question within the larger debate. We compare a classless network which is
over-provisioned against an engineered network using per-class queuing to offer
Class-of-Service (CoS) (i.e., differentiated-service)
and meet user expectations and SLAs. In most
situations a differentiated network can save significantly over a classless
network.
- M. Yuksel, K. K. Ramakrishnan, S. Kalyanaraman, J. D. Houle,
and R. Sadhvani, Quantifying Overprovisioning vs. Class-of-Service: Informing the
Net Neutrality Debate, Proceedings
of IEEE International Conference on Computer Communication Networks
(ICCCN), pages 1-8, Zurich, Switzerland, August 2010. (slides)
Abstract:
The benefit of Class-of-Service (CoS) is an
important topic in the "Network Neutrality" debate. Proponents
of network neutrality suggest that over-provisioning is a viable
alternative to CoS. We quantify the extra
capacity requirement for an over-provisioned classless (i.e., best-effort)
network compared to a CoS network providing the
same delay or loss performance for premium traffic. We first develop a
link model that quantifies this Required Extra Capacity (REC). For
realistic traffic distributions (e.g., long-range dependent), we find the
REC using ns-2 simulations of the CoS and
classless links. Our primary contribution is in using these link models to
quantify the REC for realistic network topologies under various scenarios
including "closed loop" environments with traffic generated by
TCP sources that adapt to the available capacity. We show that REC can be
significant even when the proportion of premium traffic requiring
performance assurances is small, a situation often considered benign for
the over-provisioning alternative.
- M. Yuksel, K. K. Ramakrishnan, S. Kalyanaraman, J. D. Houle,
and R. Sadhvani, Class-of-Service
in IP Backbones: Informing the Network Neutrality Debate,
(short paper) Proceedings of ACM International Conference on Measurement and
Modeling of Computer Systems (SIGMETRICS), pages 465-466, Annapolis,
MD, June 2008. (poster)
Abstract:
The benefit of Class-of-Service (CoS) is a
heated topic in the “Network Neutrality” debate. Proponents of network
neutrality suggest that over-provisioning is a viable alternative to CoS. We quantify the extra capacity requirement for an
over-provisioned classless (i.e., best-effort) network compared to a CoS network providing the same delay or loss
performance for premium traffic. We first develop a link model that
quantifies this Required Extra Capacity (REC). To illustrate key parameters
involved in analytically quantifying REC, we start with simple traffic
distributions when delay or loss probability is the performance goal.
Then, for more bursty and realistic traffic
distributions (e.g., long-range dependent), we find the REC using ns-2
simulations of the CoS and classless link cases.
Our primary contribution is in using these link models to quantify the REC
for realistic network topologies (obtained from Rocketfuel)
under various scenarios including link and node failures. We show that REC
can be significant even when the proportion of premium traffic is small, a
situation often considered benign for the over-provisioning alternative.
We also show that the impact of CoS on best-effort
traffic is relatively small while still providing the desired performance
for premium traffic.
- J. D. Houle, K. K. Ramakrishnan,
R. Sadhvani, M. Yuksel, and S. Kalyanaraman, The Evolving
Internet - Traffic, Engineering, and Roles, Proceedings of Research Conference on
Communication, Information and Internet Policy (TPRC), Arlington, VA,
September 2007. (slides)
Abstract:
Entertainment and real-time applications like voice-over-IP, medical telemetry, network gaming and streaming video are
quickly becoming prevalent applications over packet-based communication
networks. These applications impose an extremely diverse set of network
performance requirements. Further, networks are also experiencing rapid
growth in the number of users and traffic per user. There have traditionally been two
canonical approaches to handling such needs: (i)
provide substantially increased bandwidth on the links to create
sufficient overall capacity in a best-effort network, or (ii) provide
class-based differentiated service to meet each application’s performance
requirements. The current
network-neutrality debate reflects the tension between providing
substantial additional capacity to meet the most demanding application
needs and search for a viable business model to fund this capacity growth.
We examine three dimensions to lend insight to this ongoing debate. First, we focus on broad
application performance requirements and discuss traffic growth
projections. Second, assuming
a mix of emerging as well as traditional applications, we develop queuing
models for an IP backbone to quantify the economies achievable with a
differentiated network compared to a best-effort network with enough
capacity to achieve the required performance. Finally, we examine a variety of media delivery models
to understand the flexibility needed to achieve the objectives of
sustainability and customer expectations.
- M. Yuksel, K. K. Ramakrishnan, S. Kalyanaraman, J. D. Houle,
and R. Sadhvani, Value of
Supporting Class-of-Service in IP Backbones, (short paper) Proceedings
of IEEE International Workshop on Quality of Service (IWQoS),
pages 109-112, Chicago, IL, June 2007. (slides)
Abstract:
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. Our modeling framework first
develops a link model that quantifies the Required Extra Capacity (REC) in
order for a classless link to provide the same level of performance as
experienced by premium class traffic passing through a fixed capacity CoS link. We develop the REC calculations for the
cases when average delay or the average loss probability is the target
performance goal with Poisson or Markov Modulated Poisson Process (MMPP)
input traffic. Our primary contribution is in quantifying the value of the
CoS support in a network setting.
Problem
Statement and Motivation - People - Publications
Last updated on
January 13, 2011