- talks - lists.cs.princeton.edu

P Jiang FPO
by Melissa M. Lawson 30 Jul '13

30 Jul '13

Peng Jiang will present his FPO on Monday August 5 at 11 AM in 101 Carl Icahn (Note Location!). The members of his committee are: Mona Singh, advisor; Olga Troyanskaya and Hilary Coller, readers; Gertrud Schupbach and John Storey, nonreaders. A copy of his thesis, "Combinatorial code analysis for understanding biological regulation", is on display in Room 310. Everyone is invited to attend his talk.

1 0

3:30pm Wed Jul 24 talk on cellular networks by Erran Li (Bell Labs) in CS 402
by Jennifer Rexford 24 Jul '13

24 Jul '13

Speaker: Dr. Erran Li, Bell Labs Title: Making Cellular Networks Scalable and Flexible through Design-in-the-Large Date/time: 3:30-4:30pm Wednesday July 24 Location: 402 in the CS building Abstract: The rapid penetration of smart phones has put tremendous stress on cellular data networks. To understand the fundamental limitations of current cellular network architecture, I will first summarize my work on scalable network monitoring primitives and root cause analysis of performance problems. This work became a key part of the Alcatel-Lucent Wireless Network Guardian 9900 product. Our measurement study reveals that our mobile user experiences are hampered by the lack of access bandwidth at radio access networks and the inflexibility of cellular core networks. I will then address the two key problems through the principle of design-in-the-large. (1) To scale access bandwidth at radio access networks, I will present the design and prototype of Argos, a base station architecture that employs an unprecedented number of antennas simultaneously to serve a smaller number of mobile devices in the same band of frequencies. Both analysis and early experimental results suggest this design can lead to orders of magnitude increase in both spectral and energy efficiency. (2) To enable flexible provisioning of fine-grained policies at cellular core networks, I will summarize the design and prototype of SoftCell, a scalable software-defined cellular core network architecture. SoftCell achieves scalability through asymmetric edge design and selective multi-dimensional aggregation in the data plane and switch offloading in the control plane. Evaluations based on real LTE network traces show that a SoftCell controller can handle a network of thousands of base stations. This is joint work with collaborators at Princeton, Rice, Yale, and Bell Labs.

1 0

Re: [talks] 2pm Thu Jul 25 data center networking talk by Kyriakos Zarifis (USC) in CS 402
by Srinivas Narayana 23 Jul '13

23 Jul '13

Speaker: Kyriakos Zarifis, University of Southern California (USC) Title: Don't Drop, Detour! Date/time: 2-2.30pm, Thursday July 25 Location: 402 in the CS building Abstract: Today's data centers must support a range of workloads with different demands. While existing approaches handle routine traffic smoothly, ephemeral but intense hotspots cause excessive packet loss and severely degrade performance. This loss occurs even though the congestion is typically highly localized, with spare buffer capacity available at nearby switches. In this talk, we argue that switches should share buffer capacity to effectively handle this spot congestion without the latency or monetary hit of deploying large buffers at individual switches. Specifically, we present DIBS, a mechanism that achieves a near lossless network without requiring additional buffers. Using DIBS, a congested switch detours excess packets to neighboring switches to avoid dropping them. We implement DIBS in hardware, on software routers in a testbed, and in simulation, and we demonstrate that it reduces the 99th percentile of query completion time by 85%, with very little impact on background traffic. Speaker bio (from his webpage): Kyriakos Zarifis is a 1st year PhD student in the Networked Systems Lab at the University of Southern California. Previously he worked as a networking researcher at the International Computer Science Institute at Berkeley, where he also dedicated part of his time maintaining NOX. He holds an M.Sc. in Internetworking from Kungliga Tekniska Högskolan and a Ptychion (B.Sc.) from the Department of Informatics and Telecommunications of the University of Athens.

1 0

3:30pm Wed Jul 24 talk on cellular networks by Erran Li (Bell Labs) in CS 402
by Jennifer Rexford 17 Jul '13

17 Jul '13

Speaker: Dr. Erran Li, Bell Labs Title: Making Cellular Networks Scalable and Flexible through Design-in-the-Large Date/time: 3:30-4:30pm Wednesday July 24 Location: 402 in the CS building Abstract: The rapid penetration of smart phones has put tremendous stress on cellular data networks. To understand the fundamental limitations of current cellular network architecture, I will first summarize my work on scalable network monitoring primitives and root cause analysis of performance problems. This work became a key part of the Alcatel-Lucent Wireless Network Guardian 9900 product. Our measurement study reveals that our mobile user experiences are hampered by the lack of access bandwidth at radio access networks and the inflexibility of cellular core networks. I will then address the two key problems through the principle of design-in-the-large. (1) To scale access bandwidth at radio access networks, I will present the design and prototype of Argos, a base station architecture that employs an unprecedented number of antennas simultaneously to serve a smaller number of mobile devices in the same band of frequencies. Both analysis and early experimental results suggest this design can lead to orders of magnitude increase in both spectral and energy efficiency. (2) To enable flexible provisioning of fine-grained policies at cellular core networks, I will summarize the design and prototype of SoftCell, a scalable software-defined cellular core network architecture. SoftCell achieves scalability through asymmetric edge design and selective multi-dimensional aggregation in the data plane and switch offloading in the control plane. Evaluations based on real LTE network traces show that a SoftCell controller can handle a network of thousands of base stations. This is joint work with collaborators at Princeton, Rice, Yale, and Bell Labs.

1 0

Linjie Luo FPO
by Melissa M. Lawson 04 Jun '13

04 Jun '13

Linjie Luo will present his FPO on Wednesday June 12 at 10AM in Room 402. The members of his committee are: Szymon Rusinkiewicz, advisor; Tom Funkhouser and Sylvain Paris (Adobe), readers; David Dobkin and Adam Finkelstein, nonreaders. A copy of his thesis, "Accurate, Robust, and Structure-Aware Hair Capture", is on display in Room 310. Everyone is invited to attend his talk.

1 0

Starting Now: Udi Weinsberg Talk, Sherrerd 306: Building a Practical Privacy-Preserving Recommender System
by Joshua A. Kroll 29 May '13

29 May '13

Udi Weinsberg of Technicolor will be speaking at 2:00pm on May 29, in Sherrerd 306. His title and abstract are below. Building a Practical Privacy-Preserving Recommender System Many online services, such as recommender systems, email, and social networks collect user data, which is then used for both personalization and monetization. Although the latter enables services to be free, users are realizing that these services come at a hidden cost of potentially exposing their private data. In this talk I will show that even the common 5-star item-rating recommender system leaks private demographic information. Then, I will discuss methods for helping users preserve their privacy while getting accurate recommendations. Finally, a building block of many recommender systems, and an important machine-learning algorithm on its own, is linear regression. I will present a system that learns a linear model without learning anything about the private input data other than the output model.

1 0

Two Talks on Privacy: Tuesday 5/28 and Wednesday 5/29
by Joshua A. Kroll 27 May '13

27 May '13

Hi Everyone, The Security Group is holding two great talks on new research in privacy. The first is Tuesday, 5/28 at 4pm and the second is Wednesday, 5/29 at 2pm. Both are in Sherrerd Hall 306 (the CITP conference room). Please see the notices below. Thanks, Josh -------------------- Matthew Wright of the University of Texas at Arlington will be speaking at 4pm on May 28th (tomorrow!) in Sherrerd 306. His title and abstract are below. Leveraging Social Networks for Improved Anonymity and P2P Systems Social networks are great for connecting with other people, but they can also be leveraged for enhanced security properties. In this talk, I will describe two systems -- Pisces and Persea -- that we have designed to take advantage of the information that is inherent in the social network structure. Pisces is a system for enhancing anonymity in peer-to-peer (P2P) anonymity system designs. An anonymity system, such as the popular Tor network, helps protect your privacy on the Internet and enables people in countries like Syria to get around Internet censorship. In Pisces, we route our anonymity paths through users' social connections using verifiable random paths. We show that this technique provides much better privacy than prior designs in the face of strong attackers. Persea addresses the reliability of looking up information and resources in a P2P system, such as Skype or Bittorrent. Existing systems are vulnerable to an attacker adding many malicious peer nodes, e.g. by using a botnet, and having them undermine the reliability of lookups. We propose a P2P system, Persea, based on a bootstrap tree -- essentially a social network that shows how each person entered the P2P system via a series of invitations. We embed the bootstrap tree into the identities that nodes use to locate themselves and perform lookups. We argue that this approach is more suitable to P2P systems than prior approaches and show that it provides lookup success rates at least as good as in prior work. -------------------- Udi Weinsberg of Technicolor will be speaking at 2:00pm on May 29, in Sherrerd 306. His title and abstract are below. Building a Practical Privacy-Preserving Recommender System Many online services, such as recommender systems, email, and social networks collect user data, which is then used for both personalization and monetization. Although the latter enables services to be free, users are realizing that these services come at a hidden cost of potentially exposing their private data. In this talk I will show that even the common 5-star item-rating recommender system leaks private demographic information. Then, I will discuss methods for helping users preserve their privacy while getting accurate recommendations. Finally, a building block of many recommender systems, and an important machine-learning algorithm on its own, is linear regression. I will present a system that learns a linear model without learning anything about the private input data other than the output model.

1 0

Jialu Huang FPO
by Melissa M. Lawson 24 May '13

24 May '13

Jialu Huang will present her FPO on Monday June 3 at Noon in room 402. The members of her committee are: David August, advisor; Kai Li and David Walker, readers; Doug Clark and JP Singh, nonreaders. A copy of her thesis, "Automatically Exploiting Cross-Invocation Parallelism Using Runtime Information", is on display in Room 310. Everyone is invited to attend her talk.

1 0

Fwd: Mehmet E Basbug General Exam
by Melissa M. Lawson 20 May '13

20 May '13

Mehmet is an ELE student who is advised by Rob Schapire. He's asked that we announce his generals in case anyone from CS is interested in attending. ELE generals do not post a reading list like CS does. ---------------------------------------------------------------------- Mehmet E Basbug will present his research seminar/general exam on Monday May 20 at 1:30pm in E-Quad Room J323. The members of his committee are: Robert Schapire (advisor), Kai Li , and Peter Ramadge. Everyone is invited to attend his talk and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract is given below: Abstract ======== The decoding of brain states using machine-learning and pattern-classification techniques has become a widely used approach in many areas of neuroscience research. Decoding methods are especially essential to the development of non-invasive brain-computer interfaces. One of the most exciting prospects of decoding in basic neuroscience is the potential to modify a person's sensory cortex in response to brain events as they happen or perhaps slightly before they happen. We used EEG and MEG signals to decode the intention of movement even before the movement happens. In doing that we took a big data approach to the problem and developed reliable and scalable models for boosting algorithm.

1 0

N Katta generals
by Melissa M. Lawson 17 May '13

17 May '13

Naga Katta will present his research seminar/general exam on Wednesday May 22 at 2PM in Room 402. The members of his committee are: Jennifer Rexford (advisor), David Walker, and Michael Freedman. Everyone is invited to attend his talk and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. -------------------- Abstract: The \emph{network update problem} requires that a network operator replace the current, global network configuration with a new, global network configuration while preserving key correctness invariants. One way to solve the problem is by ensuring \emph{per-packet} or \emph{per-flow consistency}. The former ensures that every packet traverses either the old network configuration, or the new configuration, but not some combination of the two. The latter generalizes the former by ensuring in addition that every packet in the same flow traverses the same configuration. A consequence of either consistency model is that if both old and new configuration share some property of their paths through the network, such as loop-freedom, access control or connectivity then that property is guaranteed to persist across an update. A disadvantage of past implementation mechanisms for these consistency models is that the transition from one configuration to the next can double the configuration rule-space requirements on a network switch. we present general-purpose algorithms for solving the network update problem, that are designed to trade the time required to perform an update against the additional rule-space needed. More specifically, we break a global consistent update in to $K$ rounds, with each round transferring some subset of the network traffic from old to new configuration. The more rounds used, the longer the update time, but the less rule space overhead is required. To ensure consistency, we perform an analysis of old and new configurations, determining the rules that must be added and may be removed from a switch in each round. To determine the optimal use of rule space, we demonstrate how to represent the optimization problem as a mixed integer linear program. We also show how to extend the mixed integer linear program to additionally optimize for rapid update of the largest flows. Finally, we present initial empirical results that illustrate the tradeoff between time and space. Reading List: Book [1] L. Peterson, and B. Davie, "Computer Networks: A Systems Approach", Morgan Kaufmann, 5e, 2011 Research Papers [2] D. Clark, "The Design Philosophy of the DARPA Internet Protocols", ACM SIGCOMM, August 1988 [3] J.Saltzer, D.Reed and D.Clark, "End-to-End Arguments in System Design", ACM Transactions on Computer Systems, Nov 1984 [4] Ethane: Taking Control of the Enterprise, Martìn Casado, Michael J. Freedman, Justin Pettit, Jianying Luo, Nick McKeown, Scott Shenker, Sigcomm 2007 [5] Onix: A Distributed Control Platform for Large-scale Production Networks, Teemu Koponen, Martìn Casado, Natasha Gude, Jeremy Stribling, Leon Poutievski, Min Zhu, Rajiv Ramanathan, Yuichiro Iwata, Hiroaki Inoue, Takayuki Hama, Scott Shenker, OSDI 2010 [6] Abstractions for Network Update, Mark Reitblatt, Nate Foster, Jennifer Rexford, Cole Schlesinger, David Walker, Sigcomm 2012 [7] Header space analysis: static checking for networks. Peyman Kazemian, George Varghese, and Nick McKeown, (NSDI'12). [8] A Safe, Efﬁcient Update Protocol for OpenFlow Networks, Rick McGeer, HotSDN 2012 [9] DevoFlow: Scaling Flow Management for High-performance Networks, Andrew R. Curtis, Jeffrey C. Mogul, Jean Tourrilhes, Praveen Yalagandula, Puneet Sharma, Sujata Banerjee, Sigcomm 2011 [10] TCAM Razor: a systematic approach towards minimizing packet classifiers in TCAMs, Alex X. Liu, Chad R. Meiners, and Eric Torng, IEEE/ACM Transactions on Networking (TON) [11] Using CPU as a Traffic Co-processing Unit in Commodity Switches, Guohan Lu, Rui Miao, Yongqiang Xiong, Chuanxiong Guo, HotSDN 2012 [12] OFLOPS: An Open Framework for OpenFlow Switch Evaluation, Charalampos Rotsos, Nadi Sarrar, Steve Uhlig, Rob Sherwood, Andrew W. Moore, PAM 2012 [13] Flyways To Decongest Data Center Networks., S. Kandula, J. Padhye, and P. Bahl. In Proc. of HotNets, 2005.

1 0