Last update: November 28th,
2017.
News:
·
2017-11-28: Allocation of topics and papers + 2
new papers added.
·
2017-11-21: (Minor) changes to the schedule.
·
2017-11-15: The topics have been updated.
·
2017-11-13: The web-page has been created.
Note
On this site
the current status of the course will be maintained in terms of the contents,
and expected reading. Assignments and presented slides will only be available
from CANVAS.
This seminar the SAN
group addresses the latest developments in the system and software architecture
and networking areas, which did not yet find a place in the regular curriculum.
We involve students in ongoing research, and give them practical training in
presenting material, scientific writing and studying the literature. This
course is especially, but not exclusively, intended for students that want to
graduate in the area of system architecture and networking.
Previous topics were transport protocols, wireless
networks, software
architecture, wireless sensor
networks, resource management in networked systems, and cloud-supported multimedia services for smart homes.
This year, we address the following topics:
1.
Applications of machine learning;
2.
Virtualized Decentralized IoT;
3.
Resource management in networked systems: Resource Access Protocols (RAPs) for Multi-processor
Systems.
For each individual topic there will be an assignment,
which needs to be made by a team of 2 to 3 students.
Course objectives (content-wise):
Knowledge: Students can explain
aspects (principles, techniques, and metrics) of cloud-based services,
multi-media services, smart context-aware service, and resource management.
Capabilities:
·
Applications of machine
learning: Students have an insight into the
implementation of a machine learning application on an example of face
recognition solution.
·
Virtualized Decentralized IoT: Students have insight into how to create
virtual IoT networks out of legacy systems and how to
realize fully decentralized applications on them, considering interoperability
with other IoT systems.
·
RAPs for Multi-processor Systems: Students
have an insight in the approaches for RAPs for Multi-processor Systems, their
relative strength and weaknesses, and their demands on a developer as well as
the platform.
Course objectives (field):
The students
have become acquainted with typical
issues in the System Architecture and Networking field. They can study
scientific literature and have improved presentation and writing skills.
Organization
There
will be a set of lectures accompanied by a practical assignment.
Registration
is via CANVAS.
Students
are expected to:
·
be
present at all lectures;
·
give one presentation
(i.e. a lecture of 1 hour) addressing a specific article from a given set
provided by the teachers;
·
read all articles that will
be presented and to hand-in (two or) three questions
(to the responsible lecturer) concerning every article prior to the lecture
addressing it;
·
do an assignment (as a team of 2 or 3 students), for which a research
paper has to be written.
The
assignment will be based on studying some given papers and writing about a
given research question, which may involve experiments.
The
research paper consists of 6-8 pages, normal margins, 11pt.
Course program (preliminary):
Quartile 2 (November 13th, 2017 - February 2nd, 2018):
Week 1: (15 - 11) Cancelled due to problems with the “MyTimetable”
: (17 - 11) Overview of the course.
Introductory lectures on
some other topics.
Distribution
of papers to read and assignments to make.
Week 2: (22 - 11) No classes!
: (24 - 11) Introductory lecture on “Doing research” by dr.
Tanir Ozcelebi.
Introduction on “Applications
of Machine Learning” by Mike Holenderski.
Allocation of papers to
present and discussion about the assignment to make
Week 3: (29 - 11) No classes! (Preparation for presentations)
: (01 - 12) No classes! (Preparation for presentations)
Week 4: (06 - 12) No classes! (Preparation for presentations)
: (08 - 12) No classes! (Preparation for presentations)
Week 5: (13 - 12) Student
presentations (Preliminary) Topic 1 (Applications of Machine Learning):
·
T. van Tien - [3] Otto, Klare & Jain, 2015;
·
J. Li - [2] Lin et al.,
2017.
: (15 - 12) Student presentations (Preliminary) Topic 1 (Applications of Machine Learning):
·
S. Sharma -
[1] Otto & Jain, 2017.
Week 6: (20 - 12)
Student presentations (Preliminary) Topic 2 (Virtualized
Decentralized IoT):
·
S. Ravi - [1] Rahman et al. 2016;
·
K. Bimpisidis - [2] Derhamy et al., 2015;
: (22 - 12) Student presentations (Preliminary) Topic 2 (Virtualized
Decentralized IoT):
·
J. Nagdeo - [3] Bui et al., 2011;
·
S. Sistu - [4] Mataric, 1993.
Nc-CH : (27 - 12) No classes (Christmas Holidays).
: (29 - 12) No classes (Christmas Holidays).
: (03 - 01) No classes (Christmas Holidays).
: (05 - 01) No classes (Christmas Holidays).
Week 7: (10 - 01) No classes! (Work on assignment)
: (12 – 01) <Draft version of the
report shall be delivered> No classes! (Work on assignment)
Week 8: (17 - 01) No classes! (Work on assignment)
: (19 – 01) Student presentations and
demos on assignments.
Week 9: (24 - 01) Examination week; no classes.
: (26 - 01) Final version of the
report to be delivered. Examination
week; no classes.
Week 10:(31 - 01) Examination week; no classes.
: (02 - 02) Examination week; no classes.
Time &
Location:
Original:
Quart \ Day |
Wednesday: 15:45
- 17:30 |
Friday: 10:45 -
12:30 |
2 |
PAV A12 b |
MF-09 |
Changes: see in
the schedule above!
Examination: There will be no (written!) exam for 2IMN00. Instead,
grading is based on the questions handed in (25%), presentation (35%), and the assignment (40%). All parts are
mandatory.
Assignments: Will become available via
CANVAS.
Lecturers:
Dr.ir. R.J. Bril
Bldg.: MF 6.068, tel.: 5412, homepage
Dr. D.S. Jarnikov
Bldg.: MF 6.122b, tel.: 8311, homepage
Dr. T. Ozcelebi
Bldg.: MF 6.066, tel.: 4426, homepage
Expected reading (Final):
All articles
listed below can be found by means of IEEE Xplore.
Topic 1: Application of machine learning
[1] Otto, C., & Jain, A. (2017). Clustering millions
of faces by identity. IEEE Transactions on Pattern Analysis and Machine
Intelligence.
[2] Lin, W. A., Chen, J. C., & Chellappa,
R. (2017). A Proximity-Aware Hierarchical Clustering of Faces. arXiv preprint
arXiv:1703.04835.
[3] Otto, C., Klare, B., & Jain,
A. K. (2015, May). An efficient approach for clustering face images. In
Biometrics (ICB), 2015 International Conference on (pp. 2 43 -250), IEEE.
Topic 2: Virtualized Decentralized IoT
[1] Hasan Derhamy,
Jens Eliasson, Jerker Delsing,
Peter Priller. “A survey of Commercial Frameworks for the Internet of Things”.
In Proc. IEEE 20th Conference on Emerging Technologies
& Factory Automation (ETFA), pages 1-8, Sept 2015.
[2] Leila F. Rahman, Tanir Ozcelebi and Johan J. Lukkien, "Choosing Your IoT
Programming Framework: Architectural Aspects," In Proc. IEEE 4th
International Conference on Future Internet of Things and Cloud (FiCloud), Vienna, 2016, pp. 293-300. doi: 10.1109/FiCloud.2016.49
[3] V. Bui, J.J. Lukkien, E. Frimout,
G. Broeksteeg, “Bridging
light applications to the IP domain”, IEEE International Conference on Consumer Electronics
(ICCE), 2011.
[4] M.J. Mataric, “Designing Emergent Behaviors:
From Local Interactions to Collective Intelligence,” Proc. 2nd Int’l Conf. Simulation of Adaptive Behavior, 1993, pp.
432–441.
Topic 3: Resource management in networked systems not covered in 2017/2018
[1]
P. Gai, M. Di Natale, G. Lipari, A. Ferrari, C. Gabellini,
and P. Marceca, A
comparison of MPCP and MSRP when sharing resources in the Janus
multiple-processor on a chip platform, In: Proc. 9th IEEE
Real-Time and Embedded Techn. and
Appl. Symp. (RTAS), pages 189–198, May 2003.
[2]
S. Afshar, M.
Behnam, R. Bril, and T. Nolte, Flexible
spin-lock model for resource sharing in multiprocessor real-time systems,
In: Proc. 9th IEEE Int. Symp. on Industrial Embedded Systems (SIES), pages 41–51, June 2014.
[3]
S.M.N. Balasubramanian, S. Afshar, P. Gai, M. Behnam, and R.J. Bril, Incorporating implementation overheads in
the analysis for the Flexible Spin-Lock Model, In: Proc. 43rd
Annual Conference of the IEEE Industrial Electronics Society (IECON), October
2017.