bild
Skolan för
elektroteknik
och datavetenskap

Kursanalys: Internets protokoll och principer, protok08

Course analysis for DD2392 and EP2120, 2008

Course analysis for DD2392 and EP2120

Authors:
    Olof Hagsand, KTH/NADA.
    György Dan, KTH/EE.
    Date: 2008-05-02

On-line version of course evaluation is available at http://www.s3.kth.se/lcn/courses/ep2120

Course Data

Number of credits:
    7.5p

Written examination TEN1 : 4.5p

Laborations :
    2.5p

Course dates: period 3, 2007/2008

Course examiners:
    Olof Hagsand (DD2392)
    Rolf Stadler (EP2120)

Lectures:
    Olof Hagsand (NADA)
    György Dan (EE),
    Måns Nilsson (KTHNOC).

Recitations:
    Olof Hagsand (NADA)
    György Dan (EE),

Lab assistants:
    Personel at KTHNOC,
    Ilias Chatzidrossos (EE).
    Olafur Helgason (EE).

Guest lectures:
    Börje Josefsson(SUNET)
    Ignacio Mas (Ericsson Research)

Number of lectures: 13 (26 hours)

Number of recitations: 5 (10 hours)

Number of homeworks (optional): 4

Number of laborations:
    4 (20 hours)
Web page:
    http://www.s3.kth.se/lcn/courses/ep2120
Course literature:
    TCP/IP Protocol Suite, Second Edition (Behrouz A. Forouzan)
    Lecture slides
    Recitation material, Lab instructions, etc

Number of registered students: 69

Number of students taking first exam: 51

Number of students passing first exam: 23

Number of students passing first exam: 28 (including complemented and appeals)

Number of students passing lab course: 51

Number of students passing both labs and exams: 28

Degree of performance: 55%

Degree of examination: 55%

Current course: What was new?

  • Tests were introduced at the lab sessions. Students who did not pass the test were sent home and could not participate at the lab session.
  • New course codes and Bologna grades.
  • Minor modifications in schedule: added dns lecture but removed DNS recitation. Added extra lecture for misc subjects.
  • Labs 3 and 4 used a virtual lab environment.

General comments on this year's course

In general good feedback as can be seen from the student evaluation. There were many teachers and assistant involved in the course, much more than previous years.

The students have lots of comments about the labs both postive and negative which reflects the amount of work that is spent with them. The labs are central in the course and they ned to be developed from year to year to maintain the quality. It is interesting to see the difference from year to year in students evaluations of the labs, even though the same labs are given, their evaluation is very different.

This year the main focus seems to be on the DNS lab (lab 4). One reason could be that the regular assistant was sick so that two unexperienced assistants had to give the DNS lab, which several students experienced as negative. In any case, the DNS lab seems to be in great need of some clarification and stream-lining to next year.

The other main theme that can be seen in the students input is the fact that the course covers a very broad subject, the book is detailed and the lectures do not always follow the book. Several students thought that the exam was difficult and is "too specific", and that the lectures, book and exam do not always match.

The course instructors agree in this and are concerned with a lowering of the examination quality. In many instances it seemed that students could master some of the advanced subject quite well (eg NAT traversal) but did not master the basic elements (eg TCP). In general the students knowledge of the basic components of the course was low something clearly seen in the exam results.

One explanation of this can be the poor presence of students on lectures and recitations, especially at the end of the course. There seemed to be a clear relation between the students that attended the recitations for example, with the students passing the exam. From the course instructor side this is a major issue that needs to be addressed the next time the course is given.

Lectures

The lectures followed the course book and were given by Olof and György. Måns gave two DNS lectures.

There were also two guest lecture, Börje Josefsson from SUNET talked about networking infrastructure and some projects he has been involved in including the Internet land-speed record, and Ignacio Mas from Ericsson research gave an introduction to SIP, and to how SIP is planned to be used within Ericsson.

The lectures were given in English.

Recitations

The students were prepared wih a homework, and new assignments were solved in groups during the recitations.

The same trend as last year was present in this year as well, the number of students attending the recitations decreased as the course went on, and there were few students remaining in the last recitations. This is one explanation for the poor exam results.

Homework

The course had four homeworks, which awarded bonus points to the exam. The homeworks were due to the recitations and thus worked as a preparation for the recitations. One homework was used as a preparation for a LAB (routing/lab3).

The homework seems to have worked well in general, there is a clear correlation between the number of points obtained at the exam and the number of homework points obtained.

We removed the DNS homework due to illness in the last minute and replaced it with a relatively advanced homework with NAT traversal. This was somewhat unfortunate since the students were noticeably worse prepared for the DNS lab.

Labs

The two first labs (intro and TCP) were given by EE staff at the Q-house. The remaining two labs (DNS and routing) were made by KTHNOC staff at SAM (Teknikringen 14).

The routing and DNS lab were made using a virtual environment which means that the students can sit in any lab room (Before, they had to make their labs in SAM). This worked very well.

As has been discussed above, the DNS lab is in most need of a re-design to next year. We also wish to change the TCP lab somewhat, to make it clearer and reduce the amount of data being collected at the lab, and try to make more analysis at the lab occasion, instead of afterwards.

We introduced a mini-exam to make the students more prepared for the lab. The negative with this is that it takes time from the lab, but it works OK.

Examination

The exam was worth 60 points. The homework gave bonus points (max 4) (if the student had at least grade E on the exam). The limit to pass was set to 30, the limit to Fx was 29.

51 students took the first exam, and 23 initially failed. The limit for complementing a failed result was set to 29. Three students were offered to complement, and all three succedded. Two later complemented after appeals.

1 students acheived grade A, 3 grade B, 3 grade C, 8 grade D and 8 grade E (before complement).

Like last year the exam was experienced as more difficult by students. This trend is alarming and especially as we see a decline in the knowledge of the basic elements of the course. This has to be addressed.

There were a large group of students who were a few points from passing. These should be able to make it by just studying the basic topics more thoroughly.

Literature

The Forouzan text-book worked well. We removed an extensive RFC-literature for DNS this year, but then we saw that students were ill-prepared for the DNS-lab. The reason could also be that there was no DNS homework.

Administration

The course administration worked very well with a joint home-page.

In general communication of the requirements towards the students was more effective than last year. We might have to raise the students' attention at the beginning of the course to the fact that the course covers a big amount of material, and hence they have to start studying early.

Changes for next year

Some changes proposed for next year include:
  • Communicate to the students early in the course that they have to study, and directions on how to meet the goals of the course.
  • Review of DNS lab. The lab itself is good (ie what is actually done by the students) but the instructions and the preparations for the lab needs a major reworking. In particular, the instructions need to be focussed and clearer.
  • Revision of TCP lab. Try to reduce the manual collection of data and increase the analysis at the lab.
  • Reading instructions of Forouzan to focus on the parts that are central to the course.
  • Spend more time on lectures and recitations on the basic parts of the course.
  • Change the written exam to one basic level - which the students must pass in order to obtain higher grades. This removes the option to compensate low knowledge in advanced areas and reduces the motivation for "strategic" learning.

Course Evaluation

The course evaluation was completed by the students during the last two weeks of the course. The evaluation was made using the ACE web-based system at NADA. 25 students completed the survey which is appended to this evaluation.

Copyright © Sidansvarig: Olof Hagsand <olofh@nada.kth.se>
Uppdaterad 2008-05-05