OPNET in Advanced Networking Education

J. Theunis, B. Van den Broeck, P. Leys, J. Potemans1, E. Van Lil, A. Van de Capelle
Katholieke Universiteit Leuven, Division ESAT-Telemic
Kasteelpark Arenberg 10, 3001 Leuven-Heverlee
E-mail: [email protected]
Abstract
In this paper we discuss how OPNET can be applied in advanced
networking education, enhancing the practical skills of future
network engineers. At the K.U.Leuven Master students in
Electrical Engineering focusing on Telecommunications have to
complete a complex networking design about Voice over IP,
Wireless LANs or Video Streaming.

aim of the analytical modeling is performing some simple handmade calculations of some parameters (such as delay) on a
simplified network configuration. The outcomes of these
calculations serve as a first quick indication of the performance
of the new service, the Quality of Service (QoS) the network can
offer and the load imposed on the network by the service.
Accurate results are obtained from more detailed simulations.
For these simulations, students use OPNET Modeler. Actually,
OPNETs IT Guru will do the job as well, but because OPNET
is also used for more advanced PhD research purposes, ESATTelemic has licenses for the Modeler. First, students had to
create a representative topology and add the desired traffic on it,
always bearing the level of detail in mind, balancing accuracy
and simulation time. This resulted in existing traffic being
defined as background traffic and the traffic from the new
services being implemented as explicit traffic.

Network simulation is a major part of these assignments.

OPNET Modeler is used for this purpose. In this paper we will
describe these projects and our tutorial that introduces OPNET
to the students. Project results and supervision issues are
published as well as remarks and hints for those wanting to
introduce OPNET in their courses.
ESAT-Telemic participates in the OPNET University Program
for research [1, 2, 3] and educational projects [4, 5].
Introduction
Practical insight in modern networking is a must for every
network engineer. Too often Master students have to gain it
after graduation. Lab sessions can be introduced in the
curriculum to offer students a practical touch with reality.
However, lab session must focus on a small aspect of the broad
world of networking. Therefore, Master students in Electrical
Engineering focusing on Telecommunications are offered handson network design projects on Voice over IP (VoIP), Wireless
LAN (WLAN) and Video Streaming in their last year at
K.U.Leuven, apart from lab sessions.

Choosing the proper statistics, conclusions had to be drawn from

the results of the simulations. It was always stressed that these
conclusions are the major outcomes of the design projects.
Along with a description of the workflow, these results are
bundled into reports, published on-line [6].
First of all, we will discuss the subjects of the three different
design projects. Afterwards we will comment on the OPNET
Modeler tutorial students had to follow in order to get used to
OPNET. Then we will point out some issues about supervision
of the students. Afterwards results obtained by the students will
be published in this paper. Finally we will end the paper with
some remarks and hints on the use of OPNET in networking
education.

These design projects are considered as a third part in an

advanced course on telecommunication networks. A real life,
complex and up-to-date networking design problem is assigned
to each group of students. These problems are possible
extensions of K.U.Leuvens own intranetwork. This has the
main advantage that students have the feeling that their results
are relevant and that their efforts are not wasted.

More information on the design projects, along with all the

reports (including previous projects) can be found on the
homepage of the course [7].
The Design Projects
In this section we will describe the assignments given to the
Master students. 13 students were divided in three teams. Each
team worked on one of the following subjects: VoIP, WLAN or
Video Streaming.

Students have to spend 150 hours on the projects. Their first job
is analyzing the problem and requirements, which is not always
straightforward. In most cases, they do not have any experience
designing networks yet and normally it is their first encounter
with the unknown technology. Afterwards, they start analyzing
the existing networks and the current configuration and perform
measurements on them. Statistics gathered here can later be
introduced in OPNET Modeler.

The VoIP team had to examine the possibility of IP telephony in

student dorms. They had to find answers to questions as:
Can K.U.Leuven offer IP telephony at all its student
dorms?
What are the implications on the delay, packet loss,
of existing data traffic?

Another important part of the assignment is the analytical

modeling and simulation of the new service on the network. The
1

Research Assistant of the Fund of Scientific Research - Flanders (FWO-Vlaanderen)

1

The tutorial is not a step-by-step guide, where students do not

have to reflect on the steps to be taken. The tutorial encouraged
students to make their own decisions based on information and
documentation available. Making a mistake was not a crime at
all; actually, one can learn a lot from his/her mistakes.

What is the quality of the VoIP calls?

Can the existing networks cope with the additional
traffic? If not, what is the bottleneck and how to solve
the problem?
What is the optimal configuration of the VoIP
infrastructure?
How much will this all cost?

OPNET Modeler Supervision

Supervision is another important aspect in networking design
projects. In these projects students learn to manage themselves
in finding information, improving their problem-solving skills,
and working independently. However, there are two reasons
why a rather strict supervision is needed.

The second team had to investigate the implementation of

wireless network access in auditoria through WLAN. They had
to design an optimal wireless access network, suitable to connect
portable PCs of the students to the K.U.Leuven backbone.
Questions to be answered were:
Is WLAN possible in auditoria?
How many access points have to and can be installed?
Which is the best technology?
What is the performance of the access network?
What are the bottlenecks and how can you solve them?
How much will this all cost?

First of all, students sometimes need a push in the back,

especially at the beginning of the project. They are
overwhelmed by the extent of the assignment, and do not know
where and how to start. Once they have mastered and analyzed
the problem, an experts opinion is sometimes needed to
discover mistakes early and to point to relevant sources of
information.

The last team had to focus on Video conferencing on KOTnet2.

In the future, courses will be taped on video, digitized and made
available on the Internet (in this case: on KOTnet). Also lab
sessions and public meetings will be put on-line. Some
questions are:
Is video streaming possible on the current KOTnet
network infrastructure?
What is the impact of the video streaming on the quality
of existing traffic on the network?
What is the quality of such a video stream?
What is the optimal network configuration?
What is the optimal technology (multicast vs. unicast,
)?
How much will this all cost?

Second, not all students are hardworking ones. Some do not

realize they have to work starting from hour zero; otherwise
their results will not be satisfactory. Supervision is needed as a
kind of control and as a stimulation to exert themselves.
Simulation Results from Students
In what follows, we will describe results from the OPNET
Modeler simulations, performed by the students. Complete
results can be found in [9, 10, 11].
a. Voice over IP
The network topology in OPNET Modeler is given in Fig. 1.
You can see a central ring of five backbone switches in the
middle. Student dorms are represented by the LANs directly
connected to the backbone.

Students had to run through the process of network design:

analyze the problem, look for solutions, perform measurements
on existing networks, model the proposed solution analytically,
simulate the solution in OPNET Modeler, and finally draw
relevant conclusions.
An OPNET Modeler Tutorial
OPNET Modeler is a bulky simulation program with a vast
range of features. It takes some time to get used to the interface
and to learn creating your own scenario. OPNET provides some
easy to follow tutorials on the basics of the Modeler. However,
these tutorials are not suitable for educational usage. Students
need a more specific tutorial, adapted to their assignment.
Therefore, ESAT-Telemic created its own tutorial session,
available on-line at [8]. In this introduction to OPNET Modeler,
students learn to create a project, create and configure a network
topology, import traffic on the network in several ways, choose
statistics and view results. During the tutorial, students had to
answer twenty questions about OPNET Modeler, technological
aspects and simulation insight.

Figure 1: K.U.Leuven backbone with VoIP

Node subnet_VTK is a particular student dorm, where a more
elaborate testing was performed3. Nodes RA2 switch and
3

A VoIP test-case was run in cooperation with Siemens.

Students of this dorm could use either a softphone or an IP
phone in their rooms to call internally as well as externally.

KOTnet is the part of the K.U.Leuven intranet dedicated to all

students.
2

are provided. Nodes subnet 1 and subnet 2 each bundle 25

students. Figure 5 shows such a subnet.

Gateway belong to a centralized place from where the

K.U.Leuven network is managed. This will be the optimal place
to put the gatekeeper4 and the gateway5.

To integrate traffic in this topology, the students used some of

the predefined applications in OPNET Modeler, such as e-mail,
file transfer, web browsing, and Telnet. Other applications were
imported through the ACE module. For e.g. a news application
(students reading news messages from a news server on the
Internet), the students made a trace of a real life news session
and imported this trace in their simulations. Chatting and SSH
were imported in the same way.

Traffic already present on the backbone was added in

background using the Traffic Conversation Pair Browser. For
VoIP traffic, the students used OPNETs predefined Voice
Application. Measurements of the number of students connected
to each backbone switch and their calling behavior permitted to
configure this application properly.

Figure 2: Delay of a VoIP call (cumulative distribution)

Figure 2 depicts the cumulative distribution function of the delay
of a VoIP call from a user in subnet_VTK to node_gateway. In
Fig. 3 you can see the total VoIP traffic imposed on the
backbone by the gateway (and thus coming from and going to
external parties). From these and other results, the students
concluded that adding VoIP would not cause any problems on
the K.U.Leuven backbone, because all parameters were well
below critical bounds. However, additional investments in the
access networks seemed to be necessary.

Figure 4: The WLAN topology

Figure 3: Total VoIP traffic from and to the gateway

Figure 5: 25 wireless students in an auditorium
b. Wireless LAN
The OPNET Modeler topology of a WLAN in an auditorium is
given on Fig. 4. It is assumed that 50 students have to get access
in an auditorium at the same time. Therefore, two access points

4
5

Two different profiles were defined: one for good students,

who browse the web for information about the courses and
transfer files and presentations belonging to the course, and
another one for bad students, who read and send e-mails, chat,
read and post news messages and use Telnet and SSH as well.

Sometimes called call manager: the server of all VoIP calls.

Connects the IP network to the PSTN.
3

channel is below 3 Mbps. The delay on Fig. 8 never exceeds 20

milliseconds. Out of these results the students concluded that
two WiFi access points in an auditorium suffice for 50 students.
c. Classes through KOTnet
The topology of this groups simulations is presented in Fig. 9.
You can see a part of the backbone (nodes LUDIT and CW), and
three peripheral switches (nodes kotnet2, alma3 and VTK). The
LAN nodes are included to add background traffic to the
network. The video server is located at the upper right corner
and video clients are situated in a student dorm at the lower right
of the figure.

Figure 6: Data dropped in the WLAN

Figure 9: Topology of the video streaming

Figure 7: Load of the WLAN
This group of students had quite some difficulties obtaining a
running simulation and their results are not that satisfying.
Reasons were a lack of time and effort, and postponing the
simulations to the weeks before the deadline6. The load on some
links of the network can be seen on Fig. 10.
In this case, only six videos are streamed over the network at the
same time. In case all students are offered video streams,
unicast video streaming will definitely fall short. All links will
be saturated in no time.
When broadcast video streaming is used, when not all students
are addressed, and when these students do not cluster in one
dorm, then video streaming is possible. However, these
conditions are not realistic enough to encourage the
implementation of video streaming of on-line classes on the
current infrastructure of KOTnet.
Figure 8: Delay of the WLAN
The simulation results for a mixed class, where 30 percent of the
students adhere the bad profile, are given in Fig. 6, 7 and 8. As
you can see in Fig. 6 and 7, even in this worst case scenario,
practically no data is dropped and the load of the wireless

This proves that a strong supervision is sometimes needed.

is a superb piece of software. Perform some small tests

beforehand in order to avoid a lot of student frustration. Also
frequently install the latest patches on the software used by the
students.
Try to introduce some milestones in the simulation process.
This makes it for you as a supervisor easier to manage the
students. This also enables you to intercept difficulties and
problems in time and to discover where special efforts are
needed. Splitting things up also comforts the students: otherwise
they might consider the simulation assignment as a vast,
unachievable goal.
Last but not least, weighing and evaluating the outcomes of the
simulation can also be tricky. Apart from a written report, a
small oral examination reveals a lot of the students capacities
and insight in OPNET simulation.
Conclusion
OPNET Modeler and IT Guru are powerful tools that students
can use during networking education. OPNET software can
offer students a broader insight in networking technologies,
simulation techniques and the impact of applications on network
performance, and makes them feel as if they are real network
engineers. It can be a welcome addition to every advanced
networking course, if you are well aware of its limitations for
educational purposes.
We proposed a way to use OPNET Modeler through advanced
network design projects. Subjects for these projects have to be
well chosen in order to be of interest for students without scaring
them off by their complexity. Additional tools such as a
straightforward tutorial have to be available; otherwise some
students will drown in the deep lake of features of the software.

Figure 10: Load of the network caused by video streaming

Supervision of students during the project is of major

importance. Without supervision, we would not have been able
to publish any results from students in this paper.

Remarks on OPNET for Educational Purposes

We definitely want to share our knowledge, built up during the
past years, on large design projects and the usage of OPNET
Modeler for education. Former projects are described in [7] and
[8]. Here we will give some remarks and hints for those
planning to introduce OPNET in their networking courses.

Finally, the introduction of OPNET software in a networking

course cannot be done without concerns. We enumerated some
hints and points of attention for those planning to do so.

First of all, it is important to realize that OPNET Modeler (and

IT Guru) is not that straightforward to learn. Students must
spend a considerable amount of time before they get the basics.
OPNET Modeler has a vast range of features students have to
get acquainted with. The best thing to do is to provide a selfwritten tutorial, where students only learn what they have to
know.

References
[1] B. Van den Broeck, P. Leys, J. Potemans, J. Theunis, E. Van Lil and
A. Van de Capelle, Validation of Router Models in OPNET,
OPNETWORK 2002, Washington D.C., USA, 2002.
[2] P. Leys, J. Potemans, B. Van den Broeck, J. Theunis, E. Van Lil and
A. Van de Capelle, Use of the Raw Packet Generator in OPNET,
OPNETWORK 2002, Washington D.C., USA, 2002.

Another thing we want to remark is that in most cases a strong

supervision is needed. This is already explained above.

[3] J. Potemans, J. Theunis, B. Rodiers, B. Van den Broeck, P. Leys, E.

Van Lil and A. Van de Capelle, Simulation of a Campus Backbone
Network, a case-study, OPNETWORK 2002, Washington D.C., USA,
2002.

Documentation of whats inside OPNET Modeler, if available, is

not always adapted for student use. Therefore, provide a kind of
OPNET help desk that students with OPNET problems can
easily contact, especially when deadlines are approaching.

[4] J. Theunis, J Potemans, M. Teughels, A. Van de Capelle and E. Van

Lil, Project Driven Graduate Network Education, Proc. to the
International Conference on Networking ICN01, Colmar, France, pp.
790-802, 2001.

You also have to realize that OPNET Modeler can sometimes

cause unexpected errors, especially in new releases, although it
5

[5] J. Potemans, J Theunis, M. Teughels, E. Van Lil and A. Van de

Capelle, Student Network Design Projects Using OPNET,
OPNETWORK 2001, Washington D.C., USA, 2001.

[9] T. Brans, T. De Keyser, C. Peirs, S. Pollin, Voice over IP, internal

student report (can be downloaded from [6]), 2001.
[10] L. Peeters, B. Rodiers, R. Swennen, K. Van Caekenberghe,
Wireless LAN, internal student report (can be downloaded from [6]),
2001.

[6] Reports of the Network Design Projects Homepage:

http://www.esat.kuleuven.ac.be/~h239/reports.htm.
[7] Homepage of the Network Design Projects:
http://www.esat.kuleuven.ac.be/~h239.

[11] J. De Wachter, E. De Witte, D. Indesteege, G. Moernaut, J.

Verbiest, Classes through KOTnet, internal student report (can be
downloaded from [6]), 2001.

[8] OPNET Starting Session Homepage:

http://www.esat.kuleuven.ac.be/~h239/opnetstart.htm.