TELECOMMUNICATIONS
SEMINAR Tietoliikennetekniikan seminaari

Code: TLTE3090

This time topic is ”Evolutionary algorithms in
Communications and systems”

Credits: 3–10 ECTS (3–10 op)

Prerequisites:
related
subject studies on telecommunication engineering

Learning
Outcomes: the aim of
this course is to introduce research oriented topics in telecommunications and
systems, after completing this course successfully, the student will be able
to seek scientific information and to prepare and give seminar presentations,
moreover, they will be able to demonstrate the principles of the seminar topic

Content: this course has varying content, the
current content is always indicated by the course subtitle presented in the
course website

Study
Materials: 1. depend on
the topic

Teaching
Methods: depending on
the topic

Modes of
Study: **attending seminar sessions (most of them)**, quizzes, preparing scientific report and
**giving at least one presentation (some
guidelines in blow)**

Languages: english

Grading: 1–5 or failed, or passed/failed (depends
on the topic)

Responsible
Person: Mohammed Elmusrati and Reino Virrankoski

Teacher(s): Mohammed Elmusrati,
Reino Virrankoski

Responsible
Unit: Department
of Computer Science

Additional Information: annual course, website cs.uwasa.fi/courses/tlte3090

Timetable: http://weboodi.uwasa.fi/oodi/

Presentation lecture 3.3.2015
“What are Evolutionary
algorithms”

Presentation lecture 10.3.2015
“More about EAs”

Those who are interested to
learn more about EAs, can read the slides etc. Evolutionary algorithms course
homepage.

A lot of information about EAs
can be found from Wikipedia EA category: https://en.wikipedia.org/wiki/Category:Evolutionary_algorithms

Note, last year we originally
thought that the topic of this seminar would be “MIMO Broadband Communication”,
but we changed it later, but I will make a presentation about how EAs are used
with MIMO systems planning and resource allocation, and will present that in
later in the exercises.

EXERCISES
from 17.3.

You should prepare
presentation, about 20 min from some research paper related to the topic. You
will find suitable research papers with keywords eg.
“evolutionary algorithm” + Telecommunications e.g.
from here:

Most of the pdfs are
downloadable from the University accounts.

**Exercise, Week 13, 14:**

Matlabise these these java codes (do the same with Matlab)

http://lipas.uwasa.fi/~timan/AUTO3120/

Some programming examples of 1. exercise

FirstGA.java http://lipas.uwasa.fi/~timan/AUTO3120/FirstGA.java

BinaryGA.java other version of binary GA http://lipas.uwasa.fi/~timan/AUTO3120/BinaryGA.java

I will deliver example code later

Some programming
examples of 2. exercise

Some01.java http://lipas.uwasa.fi/~timan/AUTO3120/Some01.java

DE.java http://lipas.uwasa.fi/~timan/AUTO3120/DE.java

DE class source, place in the same directory as
Some01.java

Some programming examples
of 3. exercise

ACOsudo1.java version where numbers are drawn freely

ACOsudo2.java version that forces all lines to have all integers {1, 2, ..., 9} once on and only once all the time

**EXERCISE, Week 16: F4104
from 14:30 to 17:00**

Test these Matlab codes and try to understand
what they are doing:

Some programming examples
of 4. exercise

GAFIR1.m GA version of FIR designer. There were some problems in class, but **b** is what we want to find.

DEFIR1.m DE version of FIR designer.

CMA-ES code: http://en.wikipedia.org/wiki/CMA-ES#Example_code_in_MATLAB.2FOctave

BBO-algorithm: http://en.wikipedia.org/wiki/Biogeography-based_optimization#MATLAB

Pareto Front: http://se.mathworks.com/examples/global-optimization/3844-simple-emoo-application-finding-the-pareto-front

EMOO: http://se.mathworks.com/examples/optimization/3852-simple-emoo-problem

Other
evolutionary optimization examples:

https://se.mathworks.com/examples/search/?q=genetic

https://se.mathworks.com/examples/search/?q=particle

Note most of these codes in Mathworks
require Global optimization toolbox which our classroom Matlab
may not have.