|
PhD student
Office in room B412
Phone: +47 735 94400 or direct line +47 735 90704
Fax: +47 735 91441
E-mail: Elena.Hammari@iet.ntnu.no
Date of birth: January 14th, 1982
Research Group
Circuit and Systems Group
Work address
NTNU, Department of Electronics and
Telecommunications
O.S. Bragstads Plass 2A
NO-7491 TRONDHEIM
NORWAY
Education
MSc in Electrical Engineering, Norwegian University of Science and Technology, September
2007.
Master thesis:
“Accurate Delay Testing
of FPGA Interconnects by Branched Test Paths”
Courses
|
FE8109
|
Design and Utilization
of Memory Hierarchies in Multi-Media Applications
|
|
FE8122
|
PhD Seminar in
Circuits and Systems Design
|
|
TDT4260
|
Computer Architecture
|
|
FE8119
|
Modeling of Embedded
Systems and Systems On Chip
|
Duties and Responsibilities
- Technical support at European Test Symposium
2011
- Lab demonstrations for high school students
- Teaching Assistant in courses TFE4175-Realization
and Test of Digital Components, TFE4170-System on Chip
- Alumni Reunion-2008 organizing group
Work
Experience
Supervisors
My main advisor is Associate Professor Per Gunnar
Kjeldsberg, Dept. of Electronics and Telecommunications.
Co-supervisors are Professor Lasse Natvig from Dept. of Computer and
Information Science, and Stylianos Mamagkakis, IMEC, Leuven, Belgium.
Research Stays Abroad
October 2008 – December 2008
Visiting researcher at IMEC in Leuven, Belgium
Research interests
- Biomedical embedded systems
- Scenario-based design methodologies
- Low-power design
Other
Member of St. Anna of Novgorod
congregation, Christian Orthodox Church, Trondheim
|
|
PhD Research
The focus of my PhD study is on the efficient
utilization of computational resources in dynamic embedded systems.
Thesis
title: “Identification of
data-variable based scenarios for design of dynamic embedded systems”
Description:
The resource requirements in dynamic systems vary
in run-time and depend on run-time values of the inputs. As a result, they
cannot be planned in advance, which makes it difficult to utilize the
resources efficiently. Traditionally, a suggestion would be made on the
worst-case resource requirement and the system would be designed to
constantly provide the worst-case amount of resources irrespectively of the
run-time inputs. While this approach solves the problem of unknown input
sequences in a simple way, it produces an over-dimensioned system, where many
resources are kept occupied without doing any actual work, increasing the
system cost. For an embedded system
having strict cost constraints this is often not a feasible solution.
Instead, a run-time resource management mechanism
can be employed that arbitrates the resources according to the run-time
resource requirements of the system. The resources that are not needed at a
particular time instance are released, minimizing the system cost. This
approach is efficient given the resource management mechanism itself requires
a small amount of resources.
The current work adopts system scenario based
design methodology representing one variation of run-time resource
management. This methodology tries to minimize its run-time overhead by
grouping similar resource requirements together in “system scenarios” and
defining a corresponding limited set of system configurations to accommodate
the scenarios at run-time. The run-time resource manager consists of a
scenario detection or prediction task and the system configuration switching
task. The small set of scenarios gives low cost overhead for these tasks.
The current work concentrates on developing
methodology for identification of system scenarios in dynamic embedded
systems and demonstrates its feasibility on a biomedical system for
calculation of the largest Lyapunov exponent from human EEG signals, used in
epilepsy prediction.
Started: October 2007
Expected to complete:
Fall 2012
The research is being
done in collaboration with Design Technologies Group at Inter-university Micro-Electronics Center
(IMEC) in Leuven, Belgium
and with IMEC/Holst center in Eindhoven, The
Netherlands.
Publications
[1] Elena Hammari, Francky
Catthoor, Jos Huisken and Per Gunnar Kjeldsberg, Application of Medium-Grain Multiprocessor Mapping Methodology to
Epileptic Seizure Predictor, In 28th Norchip
Conference, Tampere, Finland, November 15-16, 2010
[2] Narasinga Rao
Miniskar, Elena Hammari, Satyakiran Munaga, Stylianos Mamagkakis, Per Gunnar Kjeldsberg and
Francky Catthoor, Scenario Based
Mapping of Dynamic Applications on MPSoC: A 3D Graphics Case Study. In
International Symposium on Systems, Architectures, Modeling and Simulation
(SAMOS IX), pages 48-57, 2009.
[3] Einar J. Aas, Bjorn B.
Larsen, Elena Hammari, Cryptographic
Design Project, In European Workshop on Microelectronics Education (EWME),
Budapest, Hungary, May 28-30, 2008
[4] Elena Hammari, Michiko
Inoue, Einar J. Aas, Hideo Fujiwara, Bidirectional Delay Test of FPGA Routing Networks. In European Test Symposium, Verbania, Italy, May 25-29, 2008
|
