Beruflich Dokumente
Kultur Dokumente
Editors Column
The sky today is azure, The sun warm and golden A filigree of light and shadowplay Through the gently swaying trees. I clean my brushes, Choose my palette
CSE Newsletter
Inside this issue:
Of vibrant, living colors, And begin to fill Todays blank canvas (~ John McLeod) Each new session is a blank canvas for us to fill with memories of moments of joy, success, growth and a hope for future. As we move ahead in this session, let us have a break and take an overview of the growth of the department as a whole. This issue starts with an article discussing the progress and growth of Computer Science in general over the past few decades. In the subsequent columns, we also discuss various departmental activities like Hack U!, Freshers Welcome as well as Open House. This year, being the centenary of Alan Turings birth is special for Computer Science community all over the world, and we discuss some of the events organized in honor of the legend. When we talk of Alan Turing, another name that comes to our mind is Alonzo Church (Church Turing Thesis). In his remembrances we also discuss the life of Alonzo Church. An interesting project article, poem and crossword are some of our regular features which continue as usual. All this and much more follows as you turn the pages of this issue.
Computer Science-a unifying science Project article Personality Profile Newsflashes ACM Turing Centenary ACM DSP Open House 2012
2,3 4
5
6 7 8
An interesting project article: Experience with Heterogeneous Clock -Skew based Device Fingerprinting
Swati Sharma
Student : Swati Sharma Supervisor : Prof. Huzur Saran Publication Details : S. Sharma, A. Hussain, H. Saran, Experience with Heterogenous Clock-Skew based Device Fingerprinting, LASER 2012 (Learning from Authoritative Security Experiment Results), Arlington, Virginia, July 2012. Introduction : Clock skew based fingerprinting allows uniquely identifying a physical device on the network. Device identification and tracking has numerous applications. For instance, some of the major applications amongst others include computer forensics and attack attribution, monitoring active hosts in the network and intrusion detection for all these active hosts once their normal behavior has been established. Host identification today can be done at many layers of the network protocol stack. Traditionally, unique host identification has been done by using the MAC addresses or static or dynamic IP addresses. Alternative strategies for host identification depend on difference in host's response to non-standard and standard packet types. Typically, these can be broadly classified as operating system fingerprints, browser fingerprints, MAC sequence numbers based fingerprints, active behavioral fingerprints and fingerprints from other iden-tification parameters from the packet headers at the Physical, Data Link, Network & Transport Layers. Clock skew based fingerprint identifies a device based on the negligible shift in system clocks and hence, provides a robust hardware dependent method of identification. Background : Clock skew based device identification was introduced by Kohno et. al. in 2005. This technique utilized a constructive exploitation of the shift in system clock, to identify devices uniquely. Prior to this ap- proach, Moon and Paxson showed that clock skews were only a source of distortions in network measurements and accuracy in network protocols and network measurements necessitated synchronization of clocks for all devices in the network. Contributions : The contributions of this work are: Validation of the clock-skew based fingerprinting approach put forth by Kohno in 2005. The maximum error threshold we observed in the clock skew under all investigation parameters was 0.3 ppm. Systematic evaluation of the stability of clock skew based device fingerprint and the factors influencing it in a heterogeneous environment. During the nine month experimentation period, we fingerprinted and tracked 162 devices by performing a multi-dimensional comparison of factors influencing clock skews. These dimensions were (a) measurement techniques, (b) target host measurement environments, and (c) target host configurations. Exploring the feasibility of heterogeneous target identification in modern networks. We found that the skew pattern for the desktops and laptops was very different than that of the handheld devices. We observed skew jumps in the handheld devices caused by changes in the ambient temperature and power state of the device, suggesting a thorough investigation into it. Identification of factors affecting device's clock skew. The deviations from Kohno's measurement techniques included modifications in ICMP timestamp collection process, variable length measurement intervals between timestamp collections, operating temperature variations at target devices, power source variations from batteryPage 2
An interesting project article: Experience with Heterogeneous Clock -Skew based Device Fingerprinting (continued...)
backed to AC power operated, variations in operating systems on same target device, variations in system state, device configurations, network conn ectivit y and vantage points while calculating the skew estimate of a target device. Methodology : A clock's skew is defined as the first-order derivative of its time offset with respect to the true time or the reported time of another clock. This can be interpreted as the time difference between two machines for every second of true time that elapses and is measured in ppm (parts per million). This is shown in the figure. X -axis shows the observed offset between the target device and measurement point while the y-axis shows the time at the measurement point. The precision of the skew calculation lies in the accurate extraction of system time generated by the oscillator frequency. TCP timestamps from the TCP header and ICMP timestamps from the ICMP timestamp response header provide the adequate granularity for this purpose. TCP timestamps were captured from TCP communication taking place between the main server and clients, while for ICMP timestamps, the server initiated communication in the form of ICMP Timestamp Request mesgerprinting technique in a moderate size, heterogeneous device network. Specifically, we explore the feasibility of device identification on a network comprised of 152 desktops and laptops, 48 virtual machines, and 10 handheld devices and demonstrate that to achieve an error-free and stable clock skew estimate at least 70 timestamped packets must be captured from a target device. The introduction of batch ICMP mode helped improve the accuracy of the skew estimation algorithm and reduced the maximum observed error threshold to 0.3 ppm. The clock-skew was evaluated across measurement methodologies, target device environments and target device configurations. Our findings indicate that the skew estimate is affected by power state of handhelds, regular NTP updates in desktops, and the capture duration.
sages and extracted the target machine timestamps from ICMP Timestamp Response messages. We then measure the effect of several investigation parameters under the dimensions mentioned above for our set of heterogenous devices. For each of these investigation parameters we calculate the clock skew of the device with respect to multiple measurement points in the network repetitively for a duration of 9 months. A summary of our observations is provided in the tables given below.
Future work involves exploring the effect of drastic ambient temperature changes on the target device clock skew and the reason for the 1 ppm skew jump with change in the handheld power source. There do exist countermeasures that an adversary can take to escape unique identification, by tampering with the timeConclusions : We present an exten- stamp values inserted in ICMP and TCP sive systematic evaluation of the sta- packets or by playing with the abovebility of clock skew based device fin- mentioned factors that may affect the clock skew estimate. The only possibility to thwart such a countermeasure is to club the clock skew based fingerprint with another parameter and combine their results to provide unique identification .
Page 3
curb and was hit by a trolley car coming from his blind side; Mary was a nurse-trainee at the hospital, where they first met.
Shweta Agrawal, UCLA, 16 April 2012 Prof. Satyajit Mayor , NCBS , 20th April 2012 Prof. Madhu Mutyam, IIT Madras, 26th April Dr. Julian M. Bass, Robert Gordon University, Aberdeen, UK, 8th June 2012 Sumeet Agarwal from Oxford University, 11th June 2012 Prof. Krishnendu Chakrabarty, Duke University , 5th July 2012 Prof. Inderjit Dhillon, UT Austin, 26th July 2012 Saikat Guha, MSRI, 7th August 2012 Dr. Ramesh Hariharan, Strand Genomics, 8th August 2012 Prof. Chandrajit Bajaj, The thUniversity of Texas at Austin, 9 August 2012 Dr. Steve Scott, CTO for Tesla business, NVIDIA, 9th August 2012 Prof. Kaleem Siddiqi, McGill University, 13th August 2012 Akshay Sundararaman (IRISA Rennes, France, 14th August 2012 Prof. Abhijit Chatterjee, School of Elect & Comp Engg, Georgia Tech, 16th August 2012 Prof. Richard Anderson, University of Washington, 27th August 2012 Dr. Balaji Raman (VERIMAG, France), 27th August 2012 Dr. Mausam, University of Washington, Seattle, 6th September 2012 II. Publications
th
ity of Prebisimulation for Timed Automata , International Conference on Computer Aided Verification (CAV) 2012, Berkeley, California, USA: July 07-13, 2012
Rajesh Kumar Pal, Kolin Paul and Sanjiva Prasad. ReKonf: A Reconfigurable Adaptive ManyCore Architecture. In Proceedings of the 10th IEEE International Symposium on Parallel and Distributed Processing with Applications (ISPA'12), Madrid, Spain, 10-13 July 2012 Adjacent Grid Block Selection (AGBS) kNN-Join Method. In: Proc. of 9th International Conference on Machine Learning and Data Mining (MLDM) 2012, 16-20 July, Berlin, Germany. Swati Sharma, Alefiya Hussain, Huzur Saran, Experience with Heterogenous Clock-Skew based Device Fingerprinting , LASER 2012 (Learning from Authoritative Security Experimentation and Results, Arlington, Virginia, US, 18-19 July 2012. Swati Sharma, Alefiya Hussain, Huzur Saran, Markov Model Based Experiment Comparison, Comsnets 2012 Shivendra Tiwari, Saroj Kaushik , Boundary Points Detection UsingAdjacent Grid Block Selection (AGBS) kNN-Join Method. In: Proc. of 9th International Conference on Machine Learning and Data Mining (MLDM) 2012,16-20 July, Berlin, Germany. Priti Jagwani and Saroj Kaushik, Defending Location Privacy using Zero Knowledge Proof Concept in Location Based Services, IEEE MDM 2012: 13th International Conference on Mobile Data Management held during July 23-26, 2012, Bengaluru, India Shivendra Tiwari and Saroj Kaushik, Extracting Region of Interest (ROI) Details using LBS Infrastructure and Web-databases, IEEE MDM 2012: 13th International Conference on Mobile Data Management held during July 23-26, 2012, Bengaluru, India
Vaibhav Jain, Anshul Kumar, Preeti Panda , Exploiting UML based validation for compliance checking of TLM 2 based models, Design Automation for Embedded Systems , Springer Netherlands, Vol 16, Issue 2, Pages 93 -113, 2012 Sonia Khetarpaul, S. K. Gupta, L. Venkat Subramaniam, Ullas Nambiar, Mining GPS traces to recommend common meeting points, IDEAS '12, 16th International Database Engineering & Applications Symposium, Prague, Czech Republic, August 0810, 2012 Piyus Kedia, Sorav Bansal, Deepak Deshpande and Sreekanth Iyer , Building Resilient Cloud Over Unreliable Commodity Infrastructure , IEEE Cloud Computing for Emerging Markets, Bangalore, Oct. 11-12, 2012 Chetan Arora, Subhashis Banerjee, Prem Kalra, and S.N. Maheshwari, Generic Cuts: An Efficient Algorithm for Optimal Inference in Higher , Accepted for Oral Presentation at ECCV 2012 to be held in Italy in October Smruti R. Sarangi, Partha Dutta, Komal Jalan, IT Infrastructure for Providing Energy-as-a-Service to Electric Vehicles , IEEE Trans. Smart Grid 3(2): 594-604 (2012)
Arun Kumar Parakh, M. Balakrishnan, Kolin Paul. Performance Estimation of GPUs with Cache. In Proceedings of the 26th IEEE International Parallel & Distributed Processing Symposium Workshop, IPDPSW12, pages 23782387, 2012
Shibashis Guha, Chinmay Narayan and S. Arun-Kumar , On Decidabil-
B. V. N. Silpa defended her thesis on 26th April 2012 Tanveer Afzal Faruqiue defended his thesis on 14th May 2012
IV. Awards
The team of Rudradev Basak, Nikhil Garg and Pradeep George Mathis, under the guidance of Prof. Naveen Garg, have secured 18th place in the ACM ICPC 2012 contest.
Page 5
invited talks from select speakers and focused on Alan Turing's contributions, as well as the history, and the future of computing. Thirty-three Turing award winners including John Hopcroft, Richard M. Karp, Donald E. Knuth, Ronald Rivest and Adi Sahmir participated in the event which was also webcast live worldwide. The ACM Turing centenary celebration was organized by ACM and ACM SIGs with the collaboration of Microsoft, Intel and Google as corporate sponsors. The panelists discussed a wide range of topics from Turing, the man, to his vision of human memory. Some panelists shared their personal experiences with Alan Turing while others discussed the contributions of Turing in the in various fields, particularly, Artificial Intelligence (AI). The panelists discussed the ways in which AI theories and methods have influenced research on human cognition in behavioral sciences and neuroscience as well as scientific research in general. Driverless cars, Watson computer are some of the practical applications of AI techniques bringing machines closer to being intelligent. Panel discussion on Systems Architecture, Design, Engineering, and Verifica-
A computer would deserve to be called intelligent if it could deceive a human into believing that it was human. ~Alan Turing
Page 6
Speaker: Prof. Krishnendu Chakrabarty (Electrical and Computer Engineering at Duke University ) Title: Digital Microfluidic Biochips and Cyberphysical Integration: A Vision for Functional Diversity and More than Moore Abstract: Advances in droplet-based digital microfluidics have led to the emergence of biochip devices for automating laboratory procedures in biochemistry and molecular biology. These devices enable the precise control of nanoliter-volume droplets of biochemical samples and reagents. Therefore, integrated circuit (IC) technology can be used to transport and transport chemical payload in the form of micro/nanofluidic droplets. As a result, non-traditional biomedical applications and markets (e.g., highthroughout DNA sequencing, portable and point-of-care clinical diagnostics, protein crystallization for drug discovery), and fundamentally new uses are opening up for ICs and systems. However, continued growth depends on advances in chip integration and design -automation tools. Design-automation tools are needed to ensure that biochips are as versatile as the macro-labs that they are intended to replace, and researchers can thereby envision an automated design flow for biochips, in the same way as design automation revolutionized IC design in the 80s and 90s.
Web and Mobile Development related to HTML 5 & CSS 3. Apart from the lecture sessions, the students also had a platform to interact with and have their questions answered by the industry people. The star lineup of speakers from industry was an added attraction to the event.
on themes given to them. The fun quotient was maintained high with the outstanding and hilarious performances. Finishing up the night with a dance party on the Bharti balcony, the energy which the refreshments and the loud music infused in everyone set the mood. Not even the moist weather could check the high spirits, in fact the rain and the absence of lighting added to the already groovy mood, setting this event firmly in the minds of everyone. ACMs Technical Welcome Session: The ACM Chapters Freshers wel-
After organizing two successful HackU's at IIT Delhi in the past, Yahoo! decided to see the IIT Delhi hackers in action again for the third time. Invariably, this time also, students showed a lot of enthusiasm and came up with
Page 9
It does not matter how slowly you go as long as you do not stop. ~Confucius
Page 10
All parts should go together without forcing. You must remember that the parts you are reassembling were disassembled by you. Therefore, if you cant get them together again, there must be a reason. By all means, do not use a hammer. ~IBM Manual, 1925
Page 11
B R O U G H T OU T B Y A C M STUDENT CHAPTER Sponsor ACM Student Chapter SponsorDr. S. Sarangi Newsletter Team inMagazine in-charge: Aditi Kapoor (aditi@cse.iitd.ernet.in) Assisted by ACM chapter members Aayush Goel Parul Pandey Shukla Pulkit Yadav (Photos in above order left to right)