Ashish Vulimiri

Publications

• More is less: Reducing latency via redundancy
  Ashish Vulimiri, Oliver Michel, P. Brighten Godfrey, Scott Shenker
  ACM HotNets 2012
  [abstract] [pdf]
  Low latency is critical for interactive networked applications. But while we know how to scale systems to increase capacity, reducing latency --- especially the tail of the latency distribution --- can be much more difficult.
  
  We argue that the use of redundancy in the context of the wide-area Internet is an effective way to convert a small amount of extra capacity into reduced latency. By initiating redundant operations across diverse resources and using the first result which completes, redundancy improves a system's latency even under exceptional conditions. We demonstrate that redundancy can significantly reduce latency for small but critical tasks, and argue that it is an effective general-purpose strategy even on devices like cell phones where bandwidth is relatively constrained.
• How well can congestion pricing neutralize denial-of-service attacks?
  Ashish Vulimiri, Gul A. Agha, P. Brighten Godfrey, Karthik Lakshminarayanan
  ACM SIGMETRICS 2012
  [abstract] [pdf]
  Denial of service protection mechanisms usually require classifying malicious traffic, which can be difficult. Another approach is to price scarce resources. However, while congestion pricing has been suggested as a way to combat DoS attacks, it has not been shown quantitatively how much damage a malicious player could cause to the utility of benign participants. In this paper, we quantify the protection that congestion pricing affords against DoS attacks, even for powerful attackers that can control their packets' routes. Specifically, we model the limits on the resources available to the attackers in three different ways and, in each case, quantify the maximum amount of damage they can cause as a function of their resource bounds. In addition, we show that congestion pricing is provably superior to fair queueing in attack resilience.
• Application of secondary information for misbehavior detection in VANETs
  Ashish Vulimiri, Arobinda Gupta, Pramit Roy, Skanda N. Muthaiah and Arzad A. Kherani
  IFIP Networking 2010
  [abstract] [pdf]
  Safety applications designed for Vehicular Ad Hoc Networks (VANETs) can be compromised by participating vehicles transmitting false or inaccurate information. Design of mechanisms that detect such misbehaving nodes is an important problem in VANETs. In this paper, we investigate the use of correlated information, called "secondary alerts", generated in response to another alert, called as the "primary alert" to verify the truth or falsity of the primary alert received by a vehicle. We first propose a framework to model how such correlated secondary information observed from more than one source can be integrated to generate a "degree of belief" for the primary alert. We then show an instantiation of the model proposed for the specific case of Post-Crash Notification as the primary alert and Slow/Stopped Vehicle Advisory as the secondary alerts. Finally, we present the design and evaluation of a misbehavior detection scheme (MDS) for PCN application using such correlated information to illustrate that such information can be used efficiently for MDS design.
• Unsupervised and supervised classification of hyperspectral image data using projection pursuit and Markov random field segmentation
  Anjan Sarkar, Ashish Vulimiri, Suman Paul, Md. Jawaid Iqbal, Avishek Banerjee, Rahul Chatterjee, Shibendu S. Ray
  International Journal of Remote Sensing, Vol. 33 Issue 18, 2012
  [abstract] [link]
  This work presents a classification technique for hyperspectral image analysis when concurrent ground-truth is unavailable and available. The method adapts a principal component analysis based projection pursuit (PP) procedure with an entropy index to reduce the dimensionality followed by the Markov Random Field (MRF) model based segmentation. An ordinal optimization approach to PP determines a set of good enough projections with high probability, the best among which is chosen with the help of MRF model based segmentation. When ground-truth is absent, the segmented output obtained is labeled with the desired number of classes so that it resembles the natural scene closely. When the landcover classes are in detailed level, some special reectance characteristics based on the classes of the study area in question are determined. These are later incorporated in MRF model based segmentation stage while minimizing the energy function in the image space. Segments are evaluated with training samples so as to yield a classified image with respect to the type of ground-truth data. Two illustrations are presented with (i) EO-1 Hyperion sensor image with concurrent groundtruth at detailed level classes and (ii) AVIRIS-92AV3C image with concurrent groundtruth - for supervised cases. Comparison of classification accuracies and computational times of some nonparametric approaches with that of the proposed methodology are provided for the illustrations. Experimental results demonstrate that the method provides high classification accuracy and is computationally faster compared to other methods.
  DOI: 10.1080/01431161.2012.670959
• Unsupervised hyperspectral image analysis with projection pursuit and MRF segmentation approach
  Anjan Sarkar, Ashish Vulimiri, Shantanu Bose, Suman Paul, Shibendu S Ray
  2008 International Conference on Artificial Intelligence and Pattern Recognition (AIPR-08), pp. 120-127
  [abstract] [pdf]
  This work deals with hyperspectral image analysis in the absence of ground-truth. The method adopts a projection pursuit (PP) procedure with entropy index to reduce the dimensionality followed by Markov Random Field (MRF) model based segmentation. Ordinal optimization approach to PP determines a set of "good enough projections" with high probability the best among which is chosen with the help of MRF model based segmentation. The segmented output so obtained is labeled with desired number of landcover classes in the absence of ground-truth. While comparing with original hyperspectral image the methodology outperforms principal component analysis with respect to class separation as exhibited in the illustration of an archive EO-1 hyperspectral image. The technique is not computationally intensive as is usually the case in hyperspectral image analysis. When training samples are available, the segmented regions yield a classified image with any cluster validation technique.

Manuscripts

• Controlled Ratio: Scheduling multiresource jobs efficiently
  Ashish Vulimiri, Hitesh Ballani, P. Brighten Godfrey, Thomas Karagiannis
  [abstract] [draft]
  The need for big-data analytics has lead to significant interest in data-parallel frameworks such as MapReduce. Such workloads can be quite diverse with jobs bottlenecked across multiple resources, such as the network link capacity, disk speed, or CPU. This heterogeneity results in significant complexities in scheduling jobs. In particular, existing schedulers can result in significant wastage of resources.
  In this paper, we introduce a technique to optimize completion time and resource utilization in data centers for jobs with heterogeneous resource usage. Our \emph{controlled ratio} scheduler divides jobs into classes according to their resource utilization profile, uses offline profiling to approximate the optimal mix of job throughputs across classes, and steers the running system towards this optimal mix whenever possible. We simulate the scheduler's operation on both a large range of synthetic workloads as well as on a trace from Facebook's Hadoop cluster, and show that in both cases it can achieve a $20\%$ improvement in average completion time over the first-come first-served scheduling that frameworks like Hadoop use by default.
  Available on request.
• Adaptive routing with end-to-end feedback
  Ashish Vulimiri, Oliver Michel, P. Brighten Godfrey
  [abstract] [draft]
  Routing protocols typically rely on information measured within the network to select routes dynamically. This approach can limit sources' flexibility to choose routes and can result in slow reaction to dynamics. In this work, we explore what is possible using routing decisions at the sources based purely on observed end-to-end performance. We show using measurements of a deployed overlay network that very simple algorithms can enable sources' selections to approach the optimal policy with a small amount of overhead. These results indicate that source-based route selection is a promising and simple approach to achieve high performance and responsiveness.
  Not yet available.

Other

• Misbehavior detection in Vehicular Ad Hoc Networks
  Ashish Vulimiri
  B.Tech. thesis, Computer Science and Engineering, IIT Kharagpur
  [summary] [pdf]
  We considered the problem of designing mechanisms enabling vehicles in Vehicular Ad Hoc Networks to detect and work around the presence of malicious entities. My B.Tech. thesis made two primary contributions:

  • A Misbehavior Detection Scheme capable of building misbehavior detection models from experimental or simulation data (these models would then be used independently by each honest node).
  • An algorithm enabling honest nodes to collaborate with each other on the misbehavior detection problem, under the assumption that misbehaving nodes are capable of influencing the collaboration process. This essentially boils down to solving the approximate agreement variant of the Byzantine agreement problem in the VANET scenario.

Posters and Talks

• Adaptive routing with end-to-end feedback
  • Poster at the 8th GENI Engineering Conference (GEC8), Jul 2010 [pdf]
  • Talk + demo at the 9th GENI Engineering Conference (GEC9), Nov 2010 [slides]

Awards and Achievements

• Was offered the following fellowships in 2009: Gordon Y.S. Wu Fellowship (Princeton University), Cornell Fellowship (Cornell University)
• My B.Tech. thesis was nominated by the Computer Science department at IIT Kharagpur for the IIT Kharagpur Best Project award and the Indian National Academy of Engineering (INAE) Innovative Project award
• Scholarship from the university throughout my four years at the Indian Institute of Technology, Kharagpur

Professional Experience

• Intern, Microsoft Research, Cambridge (Summer 2011)
• Intern, Microsoft Research, Redmond (Summer 2008)
• Research Assistant, UIUC (Fall 2009 - present)
• Teaching Assistant, UIUC
  
  • CS 498LA, Undergraduate Research Lab (Spring 2012)
  • CS 473, Undergraduate Algorithms (Fall 2012)
• Java programmer (part-time), Minekey Inc. (2007-2008)

Coursework

• Communication network analysis [Fall 2012]
• Random processes [Spring 2011]
• Optimization by vector space methods [Spring 2011]
• Concurrent programming languages [Spring 2011]
• Graduate algorithms [Fall 2010]
• Applied cryptography [Fall 2010]
• Advanced distributed systems [Spring 2010]
• Advanced Internet [Fall 2009]
• Actor languages [Fall 2009]