Cariden often presents at conferences (see the news/events page for a recent overview). On this page we provide a selection of talks and papers by Cariden and its customers:
Event: DENOG 2009, Frankfurt, Germany
Author(s): Thomas Telkamp
In today’s economic climate, efficiency has become a key aspect of building networks, and operators are looking for the minimum amount of over-provisioning to meet QoS requirements. In this presentation, we analyze backbone traffic traces at different timescales. After showing that the aggregated traffic is well behaved, we present a methodology and simple empirical rule for capacity allocation on backbone links. In this manner, we demonstrate that some rules of thumb, such as a maximum link load of 50%, might not be acceptable approaches any more.
how_full_is_full_denog_11_09.pdf
Event: RIPE 52, April 2006, Istanbul, Turkey
Author(s): Thomas Telkamp, Cariden
For most Internet Service Providers the majority of their traffic enters or leaves the network via BGP enabled peerings or upstream provider(s).
Not only do these links need to have enough capacity during normal operation, they also need to provide redundant capacity during link failures. For the egress traffic (service provider to remote peer) this can be easily verified by simulating the rerouting under failure, as the topology of the network is completely known. The return traffic (remote peer to service provider) however can not be simulated, as the behaviour of the remote network is not known. This creates a gap in the planning process for external peering links.
We present a simple methodology for creating ‘Failover Matrices’ that describe the traffic redistribution under peering link failure conditions. The matrices provide a useful mechanism for sharing information and improving the mutual planning process without disclosing any proprietary information. We will describe the principles behind the process as well as walk through a real scenario.
Event: Internet Measurement Conference, October 2004, Sicily
Author(s): Anders Gunnar (SICS), Mikael Johansson (KTH), Thomas Telkamp (Global Crossing)
This paper considers the problem of estimating the point-to-point traffic matrix in an operational IP backbone. Contrary to previous studies, that have used a partial traffic matrix or demands estimated from aggregated Netflow traces, we use a unique data set of complete traffic matrices from a global IP network measured over five-minute intervals. This allows us to do an accurate data analysis on the time-scale of typical link-load measurements and enables us to make a balanced evaluation of different traffic matrix estimation techniques. We describe the data collection in- frastructure, present spatial and temporal demand distributions, investigate the stability of fan-out factors, and analyze the mean-variance relationships between demands. We perform a critical evaluation of existing and novel methods for traffic matrix estimation, including recursive fanout estimation, worst-case bounds, regularized estimation techniques, and methods that rely on mean-variance relationships. We discuss the weaknesses and strengths of the various methods, and highlight differences in the results for the European and American subnetworks
Event: Preparing for the Future of the Internet, Network Information Center, Mexico, November 29, 2007
Author(s): Arman Maghbouleh, Cariden
Knowledge of the amount of traffic between source and destination pairs of a network is crucial to fundamental operational tasks such as capacity planning, traffic engineering, and peering management. Router vendors, third parties, academic researchers, and ingenious network engineers have devised multiple ways of collecting and estimating traffic matrices. This session presents an overview of applications of traffic matrices and operational experiences with the various approaches, including NetFlow-based methods, mathematical estimation models, and MPLS (both RSVP and LDP) methods. Emphasis will be on practical experiences with each method.
Event: Asia Pacific IP Experts Conference (Cisco), November 4th, 2003, Shanghai, P.R. China
Author(s): Thomas Telkamp (Global Crossing)
Two main approaches exist today for providing quality of service (QoS) in IP backbones. One approach relies on scheduling/queuing; the other relies on the presence of capacity/bandwidth. The two schools of thought arise from different understandings of traffic characteristics. In this talk we present the results of an empirical study of Internet traffic characteristics. We use packet traces from a Tier-1 IP backbone network and introduce a non-parametric approach to study latency characteristics at high utilization levels. This approach requires minimal assumptions and has broad applicability. The 2nd part of the presentation is focused on the Network Planning framework, applying the derived guidelines to large IP networks. Topics covered include failure simulations, the role of Diff-Serv in the backbone, and traffic matrix
© 2001-2010 Cariden Technologies, Inc. All rights reserved.