Bahador Yeganeh

About Me

I'm a PhD candidate working at Oregon Network Research Group (ONRG) in University of Oregon. My research interest is in networked systems relying on cloud service providers that are informed by Internet measurements.


How Cloud Traffic Goes Hiding: A Study of Amazon's Peering Fabric

Bahador Yeganeh, Ramakrishnan Durairajan, Reza Rejaie, and Walter Willinger
In Proceedings of the ACM Internet Measurement Conference (IMC) - 2019


The growing demand for an ever-increasing number of cloud services is profoundly transforming the Internet’s interconnection or peering ecosystem, and one example is the emergence of "virtual private interconnections (VPIs)". However, due to the underlying technologies, these VPIs are not publicly visible and traffic traversing them remains largely hidden as it bypasses the public Internet. In particular, existing techniques for inferring Internet interconnections are unable to detect these VPIs and are also incapable of mapping them to the physical facility or geographic region where they are established. In this paper, we present a third-party measurement study aimed at revealing all the peerings between Amazon and the rest of the Internet. We describe our technique for inferring these peering links and pay special attention to inferring the VPIs associated with this largest cloud provider. We also present and evaluate a new method for pinning (i.e. geo-locating) each end of the inferred interconnections or peering links. Our study provides a first look at Amazon's peering fabric. In particular, by grouping Amazon's peerings based on their key features, we illustrate the specific role that each group plays in how Amazon peers with other networks.

On Mapping the Interconnections in Today's Internet

Reza Motamedi, Bahador Yeganeh, Balakrishnan Chandrasekaran, Reza Rejaie, Bruce M. Maggs, and Walter Willinger
In IEEE/ACM Transactions on Networking (TON) - 2019


Internet interconnections are the means by which networks exchange traffic between one another. These interconnections are typically established in facilities that have known geographic locations, and are owned and operated by so-called colocation and interconnection services providers (e.g., Equinix, CoreSite, and EdgeConneX). These previously under-studied colocation facilities and the critical role they play in solving the notoriously difficult problem of obtaining a comprehensive view of the structure and evolution of the interconnections in today’s Internet are the focus of this paper. We present mi2, a new approach for mapping Internet interconnections inside a given colocation facility. We infer the existence of interconnections from localized traceroutes and use the Belief Propagation algorithm on a specially defined Markov Random Field graphical model to geolocate them to a target colocation facility. We evaluate mi2 by applying it initially to a small set of US-based colocation facilities. In the process, we compare our results against those obtained by two recently developed related techniques and discuss observed discrepancies that derive from how the different techniques determine the ownership of border routers. As part of our validation approach, we also identify drastic changes in today’s Internet interconnection ecosystem (e.g., new infrastructures in the form of “cloud exchanges” that offer new types of interconnections called “virtual private interconnections”), and discuss their wide-ranging implications for obtaining an accurate and comprehensive map of the Internet’s interconnection fabric.

A View From the Edge: A Stub-AS Perspective of Traffic Localization and its Implications

Bahador Yeganeh, Reza Rejaie, and Walter Willinger
In Proceedings of IEEE Network Traffic Measurement and Analysis Conference (TMA) - 2017


Serving user requests from near-by caches or servers has been a powerful technique for localizing Internet traffic with the intent of providing lower delay and higher throughput to end users while also lowering the cost for network operators. This basic concept has led to the deployment of different types of infrastructures of varying degrees of complexity that large CDNs, ISPs, and content providers operate to localize their user traffic. Prior measurement studies in this area have focused mainly on revealing these deployed infrastructures, reverse-engineering the techniques used by these companies to map end users to close-by caches or servers, or evaluating the performance benefits that“typical” end users experience from well-localized traffic.To our knowledge, there has been no empirical study that assesses the nature and implications of traffic localization as experienced by end users at an actual stub-AS. This paper reports on such a study for the stub-AS UOnet (AS3582), a Research & Education network operated by the University of Oregon. Based on a complete flow-level view of the delivered traffic from the Internet to UOnet, we characterize the stub-AS’ traffic footprint (i.e.a detailed assessment of the locality of the delivered traffic by all major content providers), examine how effective individual content providers utilize their built-out infrastructures for localizing their delivered traffic to UOnet, and investigate the impact of traffic localization on perceived throughput by end users served by UOnet. Our empirical findings offer valuable insights into important practical aspects of content delivery to real-world stub-ASes such as UOnet.