|
|
Brano Kusy:: CURRICULUM VITAE
|
|
|
|
|
|
|
print version (pdf)
|
|
Research
|
My general research area is wireless embedded networks and systems. Specifically, I am interested in services that form basic building blocks of sensornet applications, innovative signal analysis and information processing techniques, and in the design, prototyping, and the deployment of sensornet systems and testbeds. I am a strong proponent of research that has both theoretical depth and is evaluated experimentally in practical deployments. I believe that the most useful abstractions of the system services, tools, and techniques are distilled through building real-world systems that are deployed and validated in multiple environments. I positively value open and collaborative research. I therefore strive to make my tools and code publicly available. |
|
Education
|
|
Ph.D., Computer Science
Department of Computer Science, Vanderbilt University, Nashville, TN.
Advisors: Janos Sztipanovits, Akos Ledeczi, Miklos Maroti
Dissertation: Spatiotemporal Coordination In Wireless Sensor Networks
|
Aug 2007 |
|
M. S., Computer Science
("Magister" in Computer Science)
Comenius University, Bratislava, Slovakia.
Specialization: Mathematical Methods and Computer Graphics
Advisor: Martin Skoviera
Thesis: An Effective Algorithm to Determine the Maximal Genus of Signed Graphs
|
Jun 2002 |
|
B. S., Computer Science
(State Exams in Mathematics and Computer Science)
Comenius University, Bratislava, Slovakia.
|
Sep 2000 |
|
|
Patents
|
- US 7,558,583 B2 (System and methods of radio interference based localization in sensor networks)
|
|
Honors & Awards
|
Comenius University, Bratislava, Slovakia:
- Dean's Award for academic excellence, May 2002.
- Rector's Award for outstanding research, May 2002.
|
|
Research Experience
|
Stanford University, Stanford, CA
Postdoctoral Scholar in Guibas Laboratory
Collaborating with Stanford Information Networks Group (SING)
I studied the problem of delivering real-time data to mobile users and the feasibility of
acquiring and processing multimedia data in sensornets:
- Routing to a mobile sink
- The Whirlpool Ad-hoc Routing Protocol (WARP) - a data collection protocol for mobile users with high reliability and low latency of data delivery. Techniques used in WARP include active
detection of user mobility, speculative routing of data around the previous
location of the user, and blurring the
boundary between the control and data planes of routing protocols.
- Predictive routing protocol - a routing protocol for when a node moves non-locally in the network connectivity graph, requiring global reconfiguration of the routing state. We
proposed a novel data structure, the mobility graph, that encodes likely mobility patterns within the network and can be used to predict future location of mobile nodes. This enables routing protocols to reconfigure their state proactively and maintain uninterrupted data streams.
- Camera sensor networks
- information rich images provide superior information over simpler sensors, but their interpretation is hard in resource constrained domains. Our multi-tier system combines camera sensors with more capable devices that provide sufficient bandwidth for delivering images to users. Following successful internet applications, which recommend music, images, and movies based on user feedback, the system encourages user collaboration in interpreting images.
- Indoor localization of moving objects using RF signals
- we proposed a novel localization method that maps a short history of RF signal strengths of a mobile node to a 2D position. The method provides accurate localization without requiring the node to be stationary during the measurements.
|
2007 - 2009
|
Vanderbilt University, Nashville, TN
Graduate Student Researcher in Network Embedded Systems Group
My work primarily focused on developing system and
network services for sensor networks and was driven by building and deploying wireless sensor
systems:
- PinPtr - a sensor network based shooter localization system that
analyzes acoustic signals of gunshots (shockwave and muzzle blast) to determine
the location of the shooter and the trajectory of the bullet. Utilizing up to a
hundred spatially distributed sensor nodes, PinPtr achieves superior fault
tolerance and resilience to acoustic echoes. The system was demonstrated in
realistic urban environments and achieved 1 meter accuracy in 3D.
- inTrack - a tracking system capable of
localizing single moving object with better than 1 meter position accuracy using low-cost hardware. inTrack was used in a
dirty bomb detection and localization system that tracked the
movement of a guard carrying a radiation detector in real time.
- Radio-interferometric Positioning Service (RIPS) - a novel ranging
and localization method for wireless sensor networks. The technique relies on a
pair of nodes emitting radio waves simultaneously at slightly different
frequencies. Measuring relative phase offsets of the resulting interference signal at two receivers,
our Mica2 implementation yields an average
localization error of 3 cm and a range of up to 160 meters.
- Time synchronization protocols:
- Flooding Time-synchronization Protocol (FTSP) - a robust proactive time-synchronization service that offers microsecond precision on resource limited
wireless platforms. FTSP was implemented on a number of platforms and became part of TinyOS2 core in 2008.
- Routing Integrated Time-synchronization (RITS) - a reactive time
synchronization protocol that establishes a common reference time base post
facto. RITS provides high accuracy, having virtually no
radio communication overhead.
|
2002 - 2007
|
Oak Ridge National Lab (ORNL), Oak Ridge, TN
Summer Internship in the SensorNet Group
- mTrack - a tracking system capable of monitoring positions of
multiple (theoretically arbitrarily many) mobile objects. The system
utilizes radio interference to measure Doppler frequency shifts of mobile
objects and calculates velocity vectors of these objects.
|
Summer 2006
|
Comenius University, Bratislava, Slovakia
- Master's thesis - I focused on the topological graph theory,
specifically, on the orientation embeddings of
unsigned graphs. I proved several new theorems that
allowed me to develop an efficient algorithm to determine the maximum constrained
genus of signed graphs.
|
2001 - 2002
|
|
|
Software Development
|
I have significantly contributed to the design and development of a number of
middleware services in TinyOS, implemented on multiple hardware platforms
(Berkeley or Intel motes: Mica, Mica2, Mica2dot, MicaZ, Telos, Tmote, iMote2):
Camera chip drivers:
Ranging and localization:
Time synchronization:
Routing:
|
|
Professional Affiliations
|
member of ACM, IEEE |
|
Service
|
TPC:
- IPSN'09
- ImageSense'08
- Wideploy'08
Local arrangements co-chair:
Reviewer:
- SPL-IEEE 2009
- Transactions on Sensor Networks 2008, 2007, 2006
- SenSys 2008,2007,2006
- EURASIP-JES 2008, JCST 2008
- IWASN 2007, ICDCS 2007
Member of TinyOS Core WG (2008-present)
Member of ISIS Graduate Leadership Council
(2005-2007) |
|
Publications
|
information can be found at my publications page
|
|
References
|
Prof. Leonidas Guibas
Computer Science Department
Stanford University, Stanford, CA
(650) 723-0304
guibas at cs.stanford.edu
Prof. Janos Sztipanovits
Department of Electrical Engineering and Computer Science
Vanderbilt University, Nashville, TN
(615) 322-3455
janos.sztipanovits at vanderbilt.edu
Dr. Akos Ledeczi
Institute for Software Integrated Systems
Vanderbilt University, Nashville, TN
(615) 343-8307
akos.ledeczi at vanderbilt.edu
Dr. Vladimir Protopopescu
Computational Sciences and Engineering Division
Oak Ridge National Laboratory, Oak Ridge, TN
(865) 574-4722
protopopesva at ornl.gov
Dr. Miklos Maroti
Department of Mathematics
University of Szeged, Szeged, Hungary
+36-62-546-378
mmaroti at math.u-szeged.hu
|
|
Conferences & Workshops Presented |
The Army Science Conference (ASC '08), Orlando, FL.
Workshop on App., Systems, and Algorithms for Image Sensing (ImageSense '08), Raleigh, NC.
Embedded Networked Sensor Systems (SenSys '07), Sydney, Australia.
European Workshop on Wireless Sensor Networks (EWSN '07), Delft, Netherlands.
Information Processing in Sensor Networks (IPSN '06), Nashville, TN.
TinyOS Technology Exchange (TTX '06), Palo Alto, CA.
TinyOS Technology Exchange (TTX '05), Berkeley, CA.
Information Processing in Sensor Networks (IPSN '04), Berkeley, CA.
Model-Based Development of Computer Based Systems (ECBS '04), Brno, Czech Republic.
|
|
|
|
Branislav Kusy
Geometric Computing Group
Stanford University
Stanford, California 94305
|