Tiziana Refice
Tiziana received a Master's Degree in computer engineering at Roma Tre University in 2005 and a Ph.D. in networking at the same university in 2008 with a thesis on root cause analysis in inter-domain routing. She has worked as research engineer at the RIPE NCC on analysis of registration and routing data. She is currently working on Internet performance measurement.
Research Areas
Authored Publications
Google Publications
Other Publications
Sort By
Access and Analyze Broadband Measurements Collected using M-Lab
Summer 2010 ESCC/Internet2 Joint Techs
Preview abstract
Measurement Lab (M-Lab) is an open, distributed server platform for researchers, to deploy Internet measurement tools. Everybody can use M-Lab's tools to measure their own broadband connection performance. The M-Lab servers collect logs of all the users' tests and make them publicly available. As of July 2010, users have run millions of tests that have generated many terabytes of measurement data.
This talk will present the public repositories of M-Lab data and will explain how to analyze M-Lab data using Google's BigQuery. BigQuery stores M-Lab's measurements logs in a table with more than 60 billions of rows. It takes less than 1 minute to run a query against the whole dataset.
View details
An Argument for Increasing TCP's Initial Congestion Window
Jerry Chu
Tom Herbert
Amit Agarwal
Arvind Jain
Natalia Sutin
ACM SIGCOMM Computer Communications Review, 40(2010), pp. 27-33
Preview abstract
TCP flows start with an initial congestion window of at most four segments or approximately 4KB of data. Because most Web transactions are short-lived, the initial congestion window is a critical TCP parameter in determining how quickly flows can finish. While the global network access speeds increased dramatically on average in the past decade, the standard value of TCP’s initial congestion window has remained unchanged.
In this paper, we propose to increase TCP’s initial congestion window to at least ten segments (about 15KB). Through large-scale Internet experiments, we quantify the latency benefits and costs of using a larger window, as functions of network bandwidth, round-trip time (RTT), bandwidthdelay product (BDP), and nature of applications. We show that the average latency of HTTP responses improved by approximately 10% with the largest benefits being demonstrated in high RTT and BDP networks. The latency of low bandwidth networks also improved by a significant amount in our experiments. The average retransmission rate increased by a modest 0.5%, with most of the increase coming from applications that effectively circumvent TCP’s slow start algorithm by using multiple concurrent connections. Based on the results from our experiments, we believe the initial congestion window should be at least ten segments and the same be investigated for standardization by the IETF.
View details
Preview abstract
As IPv4 address space approaches exhaustion, large networks are deploying IPv6 or preparing for deployment. However, there is little data available about the quantity and quality of IPv6 connectivity. We describe a methodology to measure IPv6 adoption from the perspective of a Web site operator and to evaluate the impact that adding IPv6 to a Web site will have on its users. We apply our methodology to the Google Web site and present results collected over the last year. Our data show that IPv6 adoption, while growing significantly, is still low, varies considerably by country, and is heavily influenced by a small number of
large deployments. We find that native IPv6 latency is comparable to IPv4 and provide statistics on IPv6 transition mechanisms used.
View details
Measuring Route Diversity in the Internet from Remote Vantage Points
Andrea Di Menna
Luca Cittadini
Giuseppe Di Battista
Proc. International Conference on Networks (ICN 2009)
Preview abstract
Recent works on modeling the Internet topology have highlighted how the complexity of relationships between Autonomous Systems (ASes) can not be oversimplified without sacrificing accuracy in capturing route selection. Such a shortcoming can mislead the understanding, hence the prediction, of the BGP (Border Gateway Protocol) behavior. In particular, models that assume an AS to be an atomic entity fail to account for route diversity, informally defined as the selection within a single AS of multiple routes to the same destination prefix. Internet topology models are usually built out of BGP data collected by remote vantage points. Thus, in this paper we aim at extracting and characterizing the route diversity that can be measured using such dataset. Towards this goal, we devise a methodology to compute route diversity from a continuous stream of collected BGP messages. The analysis of our results shows that (i) accounting for the BGP dynamics allows to extract much more diversity than from a static snapshot of the Internet routing configuration; (ii) route diversity observed for an AS is strongly related to its location in the customer-provider hierarchy; (in) the distribution of route diversity over ASes is unlikely to be biased by the specific choice of the collection system, while the number of prefixes exhibiting route diversity can depend on both number and location of the vantage points.
View details
Tracking Back the Root Cause of a Path Change in Interdomain Routing
Alessio Campisano
Luca Cittadini
Giuseppe Di Battista
Claudio Sasso
IEEE/IFIP Network Operations and Management Symposium (NOMS 2008)
Preview abstract
Interdomain routes change over time, and it is impressive to observe up to which extent. Routes may change many times in the same day and sometimes in the same hour or minute. Such changes are caused by several types of events, e.g., a routing policy variation in an ISP, a router reboot, or a link fault. In this paper we do a step towards the identification of the cause of route changes, a problem that is attracting increasing attention from both researchers and network administrators. Namely, we propose a methodology for analyzing a given BGP route change in order to, at least partially, locate the event that triggered the change. The methodology is supported by a publicly available on-line service.
View details
YouTube Hijacking (February 24th 2008) Analysis of BGP Routing Dynamics
Antony Antony
Daniel Karrenberg
Robert Kisteleki
Rene Wilhelm
Dept. of Computer Science and Automation, University of Roma Tre(2008)
Preview abstract
On Sunday, 24 February 2008, Pakistan Telecom (AS17557 ) started an unauthorized announcement of the prefix 208.65.153.0/24. One of Pakistan Telecom’s upstream providers, PCCW Global (AS3491 ) forwarded this announcement to the rest of the Internet, which resulted in the hijacking of YouTube traffic on a global scale.
In this report we show how this event was observed by about 300 vantage points (also called collector peers) spread over the Internet by the Routing Information Service (RIS) of RIPE NCC and, in general, how to obtain hard data on network events using public available tools developed by the RIS and by the Compunet Research Group of Rome Tre University.
View details
A Study of Internet Routing Stability Using Link Weight
Mohit Lad
Jonathan Park
Lixia Zhang
Computer Science Department, University of California Los Angeles(2008)
Preview abstract
The global Internet routing infrastructure is a large scale distributed system where routing changes occur all the time.
While prior work on Internet routing dynamics examined routing stability to individual destinations, in this paper we study the routing stability of the Internet as a whole. We use the observed changes in the number of routes over each AS-AS link as a metric and measure such changes from multiple vantage points over a period of one year. We then apply Principal Component Analysis to identify those AS links that were most involved in routing changes. Our work is the first to combine measurement data collected from multiple monitors to gauge the overall routing stability in the Internet. Our results show that very few routing events impact the entire Internet, and those events were due to announcement of new prefixes either in the form of route leakages or address space de-aggregation. We also find that the impact of most routing events is confined to a small scope, and the existence of unstable AS links over long periods of time. We believe our approach represents a new direction in routing stability measurement and our findings shed new insight into the routing system performance.
View details
Measuring and Visualizing Interdomain Routing Dynamics with BGPath
Luca Cittadini
Alessio Campisano
Giuseppe Di Battista
Claudio Sasso
IEEE Symposium on Computers and Communications (ISCC 2008)
Preview abstract
The policy-oriented nature ofBGP provides network operators with great flexibility and control over the interdomain routing, nevertheless researchers showed that these benefits come at the cost of stability and predictability. In particular, policy interactions often separate, both in time and space, the effects of network events from their causes, making it hard to assess and debug network configurations. To better understand the complexity of BGP's behavior, we developed BGPATH, a publicly available tool that analyzes user-specified prefixes and relates their routing dynamics to the global Internet activity. Namely, BGPATH records data collected by distributed vantage points, checks the reliability of data sources, and evaluates the usage of each inter-AS link, both from a single and cross-vantage point perspective. The algorithms BGPATH relies on are shown to efficiently process huge streams of BGP data. We discuss time and space performance of these algorithms, underlining the potential to evolve towards a real-time system for BGP data measurement. BGPATH also provides the user with an effective visualization of the BGP activity related to a user-specified routing change, including contextual information which makes the analysis of the change less biased by BGP path exploration and outages of data sources. We exemplify the usefulness of our tool through a real usage scenario.
View details
Mediterranean Fiber Cable Cut (January-February 2008) Analysis of Network Dynamics
Antony Antony
Luca Cittadini
Daniel Karrenberg
Robert Kisteleki
Tom Vest
Rene Wilhelm
Dept. of Computer Science and Automation, University of Roma Tre.(2008)
Preview abstract
On the morning of 30 January 2008, two submarine cables in the Mediterranean Sea were damaged near Alexandria, Egypt. The media reported significant disruptions of Internet and phone traffic in the Middle East and South Asia. About two days later, a third cable was cut, this time in the Persian Gulf, 56 kilometers off the coast of Dubai. In the days that followed,
more news on other cable outages came in. We looked at the impact these events had on Internet connectivity by analyzing the data collected by the Routing Information Service (RIS) of RIPE NCC and using publicly available tools developed by the Compunet Research Group of Rome Tre University and by the RIS.
View details
How to Extract BGP Peering Information from the Internet Routing Registry
Preview abstract
We describe an on-line service, and its underlying methodology, designed to extract BGP peerings from the Internet Routing Registry. Both the method and the service are based on: a consistency manager for integrating information across different registries, an RPSL analyzer that extracts peering specifications from RPSL objects, and a peering classifier that aims at understanding to what extent such peering specifications actually contribute to fully determine a peering. A peering graph is built with different levels of confidence. We compare the effectiveness of our method with the state of the art. The comparison puts in evidence the quality of the proposed method.
View details
