Improving the energy-efficiency of core routers is important for ISPs and equipment vendors alike. We tackle this problem by focusing on packet buffers in backbone router line-cards. We broadly classify the talk into two parts - an evolutionary approach and a clean-slate design. In the former, we propose a simple power saving mechanism that turns buffers on/off to save energy. Our scheme can be incrementally deployed today and requires minimal changes to existing line-card design.
Read more arrow_right_altThis talk will introduce P2PSIP (Peer-to-peer Session Initiation Protocol) technologies, including the most common protocols and algorithms. The talk will also discuss the performance of these technologies in different network settings and the tradeoffs associated with deploying P2PSIP systems. Additionally, the talk will cover issues related to security and NAT traversal.
Read more arrow_right_alt
We describe the design and implementation of DeepDive, a system for transparently identifying and managing performance interference between virtual machines (VMs) co-located on the same physical machine in Infrastructure-as-a-Service cloud environments.
DeepDive successfully addresses several important challenges, including the lack of performance information from applications, and the large overhead of detailed interference analysis. We first show that it is possible to use easily-obtainable, low-level metrics to clearly discern when interference is occurring and what resource is causing it. Next, using realistic workloads, we show that DeepDive quickly learns about interference across co-located VMs. Finally, we show DeepDive’s ability to deal efficiently with interference when it is detected, by using a low-overhead approach to identifying a VM placement that alleviates interference.
Read more arrow_right_altDelay is a key Internet performance metric and its stability, variation, and abrupt changes have been well studied.
However, little could have been said about the Internet-wide delay distribution. In order to build a representative sample set for the Internet-wide delay distribution, one needs to draw data from a random selection of source hosts to destination hosts and there is no measurement system with access to every AS and subnet of the Internet. In this work we propose to apply the path-stitching algorithm to archival measurement data and reconstruct the past history of Internet delay distribution. The two main advantages of path stitching are that data from existing measurement projects is sufficient to provide accurate estimates and it produces delay estimates between almost any two hosts in the Internet.
Logged messages are invaluable for debugging and diagnosing problems. Unfortunately, many execution logs are inscrutable in their raw form. For example, a production Google system may generate a billion-line log file in a single day. In my talk, I will detail two log-analysis tools that I developed to deal with this problem. These tools infer concise and precise models from large execution logs of sequential and distributed systems. Both tools enable new kinds of program analyses and make logs more useful to developers.
Read more arrow_right_altWireless communication for intelligent transportation systems (ITSs) is a promising technology to improve driving safety, reduce traffic congestion and support information services in vehicles. A new era of vehicular networks that include vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications is approaching.
Read more arrow_right_altBitTorrent is currently the dominant Peer-to-Peer (P2P) protocol for file-sharing applications. BitTorrent is also a nightmare for ISPs due to its network agnostic nature, which is responsible for high network transit costs. The research community has deployed a number of strategies for BitTorrent traffic localization, mostly relying on the communication between the peers and a central server called tracker. However, BitTorrent users have been abandoning the trackers in favor of distributed tracking based upon Distributed Hash Tables (DHTs).
Read more arrow_right_alt
Give the attendees an evolutionary view on the access networks and open perspectives on the potential developments upcoming. It will always keep an eye on the standardization trends and other ecosystem parameters (as regulation, cost and timeline constrains) in order to provide the attendees the background to further perceive some of the upcoming opted solutions and dynamics
Read more arrow_right_altComplex networks, including the Internet, wireless and cellular networks, and on-line social networks, are becoming indispensable parts of our daily lives. These networks arising from a wide range of applications can be represented and studied as graphs, and the underlying link patterns play an important role in understanding and solving problems in such applications. For example, many online social networks, such as Twitter and Google+, can be viewed as directed graphs with uni-directional "following'' relations among users, and the link directions contain crucial information about how users form social communities. In another application, online social networks such as Slashdot and Epinions represent relationships between users as links with positive or negative weights, which correspond to friend and foe relations. These networks are referred to as signed networks, where those signed links generate new challenges in understanding and studying the underlying network properties. In this talk, I present my work on developing theories for studying and characterizing various crucial graph properties, such as the edge directionality in directed graphs and the edge polarity in signed graphs. I do so by emphasizing on applications to detecting stable social communities and understanding social influence propagation patterns on online social networks.
Read more arrow_right_altWith an optimistic assumption that a WiFi Access Point (AP) is idle for 50% of the time, an approximate energy waste caused exclusively by WiFi APs worldwide is estimated to be a striking 4.7 Billion KWh/year. Furthermore, many other types of future wireless network devices are expected to spend substantial amounts of energy in their idle listening modes, including wireless sensor and actuator nodes, wireless smart meters, mobile phones, and cellular base stations.
Read more arrow_right_alt
Recent Comments