CS Postdoc Profiles

Joanna Klukowska from Graduate Center, CUNY

tdomf_c5ea8 — Wed, 10 Apr 2013 23:28:36 +0000

In my research I want to work on problems that have practical applications, which often leads me to interdisciplinary work. Most of my research in graduate school would be classified as biomedical imaging, more specifically image reconstruction (both two and three dimensional) from projections. I have been involved in projects related to computerized tomography, electron microscopy and, most recently, soft x-ray microscopy. It has been both fascinating and challenging to work with biologists, physicists and mathematicians, all of whom come with their own approaches, notational conventions, and priorities. Most of the projects with which I am involved attempt to improve the quality of reconstructions computed from on biomedical data.

I am planning to continue my work in image reconstruction from projection data. All of the projects that I have been involved in are still sources of interesting and challenging questions and the solutions that we provided can be further generalized or improved. I am also interested in expanding my work into other areas of biomedical imaging. I think of my research to some extent as providing answers to problems that arise in the general area of biomedical imaging and as long as I can collaborate with experts in the areas from which these problems arise, I can work on the development of computational solutions. The collaborators with whom I worked during my years in graduate school will remain the source of new problems, but I am also looking forward to new collaborations and challenging problems.

Contact Information

E-Mail: EMAIL OBFUSCATED

Manchun Zheng from National University of Singapore

tdomf_c5ea8 — Wed, 13 Mar 2013 04:53:55 +0000

Contact Information

E-Mail: EMAIL OBFUSCATED

Phone: +6584280202

Cheuk Yiu Ip from University of Maryland

tdomf_c5ea8 — Tue, 12 Feb 2013 07:48:12 +0000

I develop intelligent systems to effectively discover and visualize salient patterns and anomalies in big data. My work automatically segments and visualizes structures from medical volume datasets. My saliency-based system locates and guides users to explore informative regions on very large images. My research on 3D shape acquisition and matching classifies mechanical parts. My papers on volume segmentation and image exploration have received the Best Paper Award at IEEE SciVis 2012, and an Honorable Mention Award at IEEE Vis 2011. The papers related to my shape matching work have received more than 200 citations according to Google Scholar. I have also delivered commercial 3D reconstruction software for a 3D laser scanner.

Contact Information

E-Mail: EMAIL OBFUSCATED

Dimitrios Diochnos from University of Illinois at Chicago

tdomf_c5ea8 — Mon, 11 Feb 2013 21:20:08 +0000

My work on my PhD has focused on the theoretical and practical analysis of randomized algorithms. Please find more information in the links below for my CV and my Research Statement.

CV
Research Statement

Contact Information

E-Mail: EMAIL OBFUSCATED

Gulustan Dogan from City University of New York

tdomf_c5ea8 — Sun, 03 Feb 2013 19:34:00 +0000

Gulustan Dogan is a PhD candidate at the City University of New York. Gulustan holds a B.Sc in Computer Science from Middle East Technical University. She has co-authored five international papers in the area of wireless sensor networks, provenance and trust. She has worked as a student researcher in Army Research Lab funded project called Network Science Collaborative Technology Alliance.

Contact Information

E-Mail: EMAIL OBFUSCATED

Marin Lujak from University of Rome “Tor Vergata”, Italy

tdomf_c5ea8 — Thu, 31 Jan 2013 22:30:59 +0000

Marin Lujak, PhD in Management Engineering, is a researcher at the Artificial Intelligence Research Group on the University Rey Juan Carlos in Madrid, Spain. He graduated Mechanical Engineering with the specialization in Automation and Robotics at the University of Zagreb, Croatia in 2006. He received Masters of Arts degree in Business Management Engineering in 2007 and the PhD degree in Management Engineering in 2010, both from the University of Rome “Tor Vergata”. He joined the Artificial Intelligence Group at the University Rey Juan Carlos in Madrid in February 2011.

His research interests are Distributed and Decentralized Computational Optimization, Multi-Agent Systems negotiation and coordination, and Multi-agent Planning and Scheduling with applications in real industrial environment.

Marin Lujak is a member of the COST Action IC0801 – Agreement Technologies and the IEEE IES Technical Committee on Industrial Agents.

At the present, he is working at the Agreement Technologies project at the application of MAS coordination and negotiation techniques and principles in e-Health and is a co-organizer of the Robotic International Summer-School RISS-WoW 2013 in Dubrovnik, Croatia.

Contact Information

E-Mail: EMAIL OBFUSCATED

Salman Toor from Uppsala University, Sweden

tdomf_c5ea8 — Mon, 28 Jan 2013 12:05:55 +0000

I have completed my doctoral studies under the title “Managing Applications and Data in Distributed Computing Infrastructures” in June 2012. The projects during my PhD covers computational solutions, efficient resource discovery in heterogeneous environments and data management based on large scale distributed infrastructures.

In future, I would like to continue the research and development track. My aims are to continue working on both storage and computational solutions based on distributed computing infrastructures, especially the unconventional solutions for highly challenging scientific disciplines. I am also interested in exploring upcoming technologies to evaluate their benefits for my research and scientific community.

Contact Information

E-Mail: EMAIL OBFUSCATED

Cristian Gherghina from National University of Singapore

tdomf_c5ea8 — Mon, 21 Jan 2013 09:17:05 +0000

My research interests include programming languages semantics, specification logics, program verification, static analysis. More specifically I am interested in the development of richer specification logics aimed at helping program verifiers handle richer programming languages. I am working on making use of separation logic variants in order to verify functional correctenss of java-like languages. As such I am one of the people engaged in developing the sleek separation logic entaliment checker and hip program verifier.

Contact Information

E-Mail: EMAIL OBFUSCATED

El Sayed Mahmoud from University of Guelph

tdomf_c5ea8 — Sat, 22 Dec 2012 09:30:41 +0000

My research interests include machine learning and multi-objective optimization using genetic algorithms. The demand for robust and fast machine learning systems is steadily growing. Applying machine learning to many research areas reveals data with complex patterns and large volumes that represent a challenge for the existing machine learning systems. There is a backlog of data waiting to be analyzed.

My current research focuses on investigating analytic models for machine learning algorithms for addressing challenging patterns in health, business and privacy applications. The purpose of developing these analytic models is to derive guidelines for promoting the robustness of these algorithms while reducing the computational cost.

I developed a novel analytic model for the goodness of fit of a multi-classifier system mathematically throughout my Ph.D. program. This model was verified by applying hierarchical multiple regression to the results of a Monte-Carlo-based simulation for three types of challenging problems. In addition, I used this model to develop a robust framework for challenging pattern recognition problems.

My research yielded two theses (M.Sc. and Ph.D.), six peer-reviewed papers, and four international talks in Machine Learning. This research introduced a novel model for goodness of fit that produced guidelines for developing robust machine learning systems. I also investigated the implementation and auto-calibration of various machine learning systems for syndromic survalance, privacy applications, and other applications. In addition, I developed a new data generator to produce data with a predetermined Bayes error. This generator contributes to standardizing benchmark datasets. I used Hierarchal multiple regression analysis, Monte-Carlo-simulation, ANOVA and non-parametric statistical tests to examine the statistical significance of the results produced by my research.

Contact Information

E-Mail: EMAIL OBFUSCATED

Sriram Nandha Premnath from University of Utah

tdomf_c5ea8 — Mon, 26 Nov 2012 03:56:01 +0000

SRIRAM N. PREMNATH

School of Computing
University of Utah
50 S. Central Campus Drive, MEB 3190
Salt Lake City, UT – 84112

EMAIL OBFUSCATED

http://www.cs.utah.edu/~nandha

Phone: 805–312–0760

Objective
Seeking a full-time position that will enable me to utilize my excellent skills in the areas of Computer Networks, Network Security, Wireless Networks and Systems.

Education
University of Utah, School of Computing, Salt Lake City, UT
Ph.D. in Computer Science, GPA: 3.98 / 4.0
Advisor: Sneha Kumar Kasera
December 2012 (expected graduation)
Thesis title: Exploiting Cross Layer Opportunities for Secrecy and Efficiency in Wireless Networks

College of Engineering – Guindy, Anna University, Chennai, India
B.E. in Computer Science and Engineering, May 2005
Project title: Probability Based Intrusion Detection for Mobile Ad Hoc Networks

Selected Graduate Courses
Computer Networks, Network Security, Algorithms, Distributed Systems, Random Processes, Computer Architecture, Artificial Intelligence

Software Skills
Programming: C, C++, Java, Matlab, UNIX shell scripting
Libraries/Tools/Packages: OpenSSL, PThreads, Wireshark, UNIX Sockets, LaTeX, NIST Randomness Test Suite, VanetMobiSim, Windows Phone SDK
Operating Systems: UNIX, Linux, Windows

Awards
- ACM SIGCOMM Travel Grant (00) for COMSNETS conference, 2012 – Awarded to Top 10 papers from outside India
- ACWR 2011 Travel Grant (0 approx.) for ACWR conference, 2011
- Finished in Top 3 at ACM MOBICOM Student Research Competition, 2009. Received 0 award
- Winner, Student Research Poster competition, University of Utah, School of Computing, 2008

Professional Experience
Research Assistant, University of Utah, School of Computing, Salt Lake City, UT, May 2008 – Till date
Summer Research Intern, Bell Labs, Alcatel–Lucent, Murray Hill, NJ, Jun 2011 – Aug 2011
Teaching Assistant, University of Utah, School of Computing, Salt Lake City, UT, Aug 2007 – May 2008
Assistant Systems Engineer, Tata Consultancy Services, Bangalore, India, Aug 2005 – Jul 2007

Publications
S.N. Premnath, P.L. Gowda, S.K. Kasera, N. Patwari, R. Ricci, “Secret Key Extraction using Bluetooth Wireless Signal Strength Measurements”, IEEE PerCom 2013. Under review.
S.N. Premnath, J. Croft, N. Patwari, S.K. Kasera, “Efficient High Rate Secret Key Extraction in Wireless Sensor Networks Using Collaboration”, ACM Trans. on Sensor Networks. Under review.
S.N. Premnath, D. Wasden, S. K. Kasera, B. Farhang-Boroujeny, N. Patwari, “Beyond OFDM: Best–Effort Dynamic Spectrum Access Using Filterbank Multicarrier”, Accepted to appear in IEEE/ACM Trans. on Networking. Available now in IEEE Xplore.
S.N. Premnath, S. Jana, J. Croft, P.L. Gowda, M. Clark, S.K. Kasera, N. Patwari, S. Krishnamurthy, “Secret Key Extraction from Wireless Signal Strength in Real Environments”, Accepted to appear in IEEE Trans. on Mobile Computing. Available now in IEEE Xplore.
S.N. Premnath, D. Wasden, S.K. Kasera, B. Farhang-Boroujeny, N. Patwari, “Beyond OFDM: Best–Effort Dynamic Spectrum Access Using Filterbank Multicarrier”, IEEE COMSNETS Conference, 2012.
S.N. Premnath, S.K. Kasera, B. Farhang-Boroujeny, N. Patwari, “Efficient Dynamic Spectrum Access in Vehicular Networks using Filterbank Multicarrier”. ACM/IEEE ACWR Conference, 2011.
S. Jana, S.N. Premnath, M. Clark, S.K. Kasera, N. Patwari, S. Krishnamurthy, “On the Effectiveness of Secret Key Extraction from Wireless Signal Strength in Real Environments”, ACM MOBICOM Conference 2009.
S.N. Premnath, S.K. Kasera, N. Patwari, “Secret Key Extraction in MIMO-Like Sensor Networks Using Wireless Signal Strength”, ACM SIGMOBILE MC2R, 2010. Also presented at ACM MOBICOM SRC, 2009, Finished in Top 3!

University of Utah, School of Computing, Salt Lake City, UT: Research Assistant, May 2008 – Till date

Secret Key Extraction (SKE) from Wireless Signal Strength in Real Environments
- Developed efficient secret key extraction (SKE) methods based on symmetric wireless channel measurements of received signal strength that can provide security against adversaries with unlimited computational power.
- Evaluated SKE extensively on a variety of devices (Laptops, TelosB sensor nodes, Android Smartphones) based on different standards (WiFi, ZigBee, Bluetooth)
- Experimentally showed that high entropy bits are obtained fairly quickly in dynamic environments
- Demonstrated a new type of attack on SKE called predictable key generation attack in static environments
- Developed environment-adaptive SKE scheme that includes techniques from Quantum Cryptography.
- Implemented with C language using raw sockets, packet injection in monitor mode, radiotap header, OpenSSL.

Efficient High Rate Secret Key Extraction in Wireless Sensor Networks Using Collaboration
- Explored the novel use of two groups of TelosB sensor nodes for SKE. Showed that two collaborating groups of nodes can enable extracting stronger secret keys in an efficient manner (over 300% improvement). Showed that the performance can be improved further when both space and frequency diversities are exploited.
- Developed an analytical method to significantly reduce the predictable effects of shadow fading in which the size of the running average window is chosen depending on the speed of nodes and the sampling rate.
- Implemented using NesC language, Matlab, OpenSSL

Beyond OFDM: Best–Effort Dynamic Spectrum Access Using Filterbank Multicarrier (FBMC)
- Developed analytical model to compute interference power across subcarriers used by different transmitters.
- Devised a distributed, adaptive MAC protocol for exchanging best–effort traffic that promises fair share of spectrum to different nodes
- Established that FBMC, which uses special pulses at the PHY layer, achieves order-of-magnitude improvement over the widely-used OFDM in terms of data rate available at the MAC layer.
- Our research expected to have profound impact on design of future dynamic spectrum access networks
- Implemented simulation models for both indoor and vehicular network environments using Matlab, VanetMobiSim

Secret Key Extraction for Smartphones using Bluetooth Signal Strength Measurements
- Mentoring: Ongoing work of a Masters student who I am co-advising along with Professor Sneha Kumar Kasera.
- Designed efficient SKE method based on Bluetooth that is robust to jamming adversaries and heavy WiFi traffic.
- Experimentally evaluating the effectiveness of this approach using Google Nexus One Android Smartphones.

Radio Tomographic Imaging (RTI) using Ultra Wide Band radios
- Ongoing work. RTI useful for tracking mobile targets such as people in emergency/rescue operations through walls.
- RTI is a device free localization method. Amplitude changes across different multipath components of received signal are observed to track mobile targets.
- Real-world experiments in typical indoor environments with ultra wide band radios show promising results.

Bell Labs, Alcatel–Lucent, Murray Hill, NJ: Summer Research Intern, Jun 2011 – Aug 2011
- Contributed to the building of a secure communication layer for Smart Grid networks. Designed and implemented proprietary network security protocols for secure exchange of meter readings, pricing information, demand/response data among household meters, utility company servers, charging stations, etc.
- Implemented using C++, OpenSSL, PThreads
- Designed a randomized hashing mechanism to probabilistically detect replay attacks using tiny state. With s bits of information shared between a sender and receiver pair, our detection mechanism guarantees that the adversary will succeed only with a miniscule probability of 2-s.
- US Patent pending along with Dr. Vladimir Kolesnikov, Bell Labs, Alcatel–Lucent, 2011.

University of Utah, School of Computing, Salt Lake City, UT: Teaching Assistant, Aug 2007 – May 2008
- Conducted office hours, graded home works and programming assignments for Computer Networks (CS 5480) in Fall 2007, Compilers (CS 5470) in Spring 2008
- Substituted Professor S.K. Kasera’s Network Security class (CS 6490) lecture on “Secret Key Extraction using Reciprocal Wireless Channel Measurements” in Spring 2009 and Spring 2011.

Tata Consultancy Services, Bangalore, India: Assistant Systems Engineer, Aug 2005 – Jul 2007
- Contributed to a variety of real-life projects in the field of databases.
- Developed proprietary middleware software for a number of data warehousing projects.
- Implemented using Informatica software, Oracle SQL, PL/SQL, UNIX shell scripting.

Graduate Level Course Projects

Efficient Search Engine for Distributed Hash Tables (DHT) based Peer-to-Peer systems
- Maintained a list of documents called inverted indices in the DHT for potential keywords. For each keyword in a query, inverted indices are fetched and set-intersected to generate the results. Used caching for faster response.
- Implemented using Bamboo DHT and demonstrated the performance and scalability of our system with up to 32 nodes in the Emulab Testbed.

Achieving location privacy in 802.11 networks
- Proposed anonymization scheme for 802.11 wireless networks based on the concepts of Lamport hash chain and anonymity sets.
- Clients change their identifiers (MAC addresses) on-the-fly based on precomputed hash chain. The access point verifies the clients and tracks their packets by just applying computationally inexpensive hash operations.
- Implemented with changes to HostAP software on a laptop with Atheros based WiFi card.

Battery-Draining-DoS Attack on Bluetooth Devices
- Demonstrated a new DoS attack on a MotoRazr V3 cellphone exploiting the open port of Bluetooth Service Discovery Protocol. Cellphone becomes effectively inoperable. Reduced battery life by as much as 97%.
- Winner in University of Utah School of Computing’s Research Poster competition held in March 2008.
- Implemented using UNIX shell scripting

Outdoor Movement Tracking App for Windows Smartphones
- Captured latitude and longitude traces using inbuilt GPS on a Smartphone. Calculated distance traveled and instantaneous speed of moving vehicle using the Haversine formula.
- App useful for tracking the movements of real-world vehicles, pedestrians, etc.
- Implemented using Windows Phone SDK on a Samsung Focus phone.

Undergraduate Level Final Project

Probability Based Intrusion Detection for Mobile Ad Hoc Networks (MANETs)
- Developed an intrusion detection system for MANETs to detect routing attacks such as cache-poisoning, fabricated control packets, etc. Detected anomalies using percentages of changed/added/broken routes, etc.
- Our approach to intrusion detection is based on the idea that simultaneous occurrence of anomalies at several neighboring nodes is rare. Showed that our approach can achieve negligible false alarm rates.
- Evaluated the effectiveness of our approach for location aided routing protocol using the GloMoSim simulator.

Conference and Seminar Presentations

- Beyond OFDM: Best–Effort Dynamic Spectrum Access Using Filterbank Multicarrier, IEEE COMSNETS conference, India, 2012.
- Cross Layer Research on Building Secure and Efficient Wireless Networks, L3 Communications visit to the University of Utah, 2012
- Efficient Dynamic Spectrum Access in Vehicular Networks using Filterbank Multicarrier, ACM/IEEE ACWR conference, India, 2011
- Security Issues in Smart Grid Networks, Bell Labs, Enabling Computing Technologies Interns Talk Session, Murray Hill, NJ, 2011
- Replay Attack Protection with Small State, Bell Labs, Enabling Computing Technologies Interns Talk Session, Murray Hill, NJ, 2011
- Efficient High Rate Secret Key Extraction in Wireless Sensor Networks Using Collaboration, Army Research Office / Multidisciplinary University Research Initiative Meeting, Salt Lake City, UT, 2010
- On the Effectiveness of Secret Key Extraction from Wireless Signal Strength in Real Environments, ACM MOBICOM conference, China, 2009
- Secret Key Extraction in MIMO-Like Sensor Networks Using Wireless Signal Strength, ACM MOBICOM conference, SRC Winners Talk Session, China, 2009
- Various poster presentations at the University of Utah, School of Computing during 2007 – 2012

Service
Reviewer for the following journals and conferences: IEEE/ACM Transactions on Networking, IEEE Transactions on Mobile Computing, ACM/Springer WINET Journal, IEEE Transactions on Parallel and Distributed Systems, IEEE INFOCOM 2010–11, IEEE ICNP 2009, IEEE SECON 2009–10, IEEE WoWMoM 2009–10, IEEE ICCCN 2009, IEEE MASS 2008, MILCOM 2008.

References
Professor Sneha Kasera
Phone: 801–581–4541
EMAIL OBFUSCATED

Professor Neal Patwari
Phone: 801–581–5917
EMAIL OBFUSCATED

Professor Behrouz Farhang-Boroujeny
Phone: 801–587–7959
EMAIL OBFUSCATED

Contact Information

E-Mail: EMAIL OBFUSCATED

Phone: 8053120760