UNDERSTANDING THE BEHAVIOUR OF CDMA-BASED

CELLULAR NETWORKS WITH A USER-FRIENDLY

SIMULATION SOFTWARE

Xavier Lagrange and Boulbeba Karoui

Institut TELECOM, TELECOM Bretagne, Universite europeenne de Bretagne

2 rue de la Chataigneraie , 35576 Cesson Sevigne Cedex, France

Keywords:

Universal Mobile Telecommunications System (UMTS), Code Division Multiple Access (CDMA), Simulation

Software, Virtual Labs, e-learning in network engineering.

Abstract:

The Universal Mobile Telecommunications System (UMTS) has been developed in order to create a uniﬁed

telecommunications system with multimedia capabilities. UMTS uses the Code Division Multiple Access

(CDMA) technique. In this type of system, all users share the same frequency at the same time. Interference

then plays a major role in the behaviour of a CDMA network and causes the coverage to vary according to the

load. Furthermore many parameters have an impact on the coverage (spreading factor, user rate, transmission

power etc...). So motivating students to learn CDMA concepts is often difﬁcult because they ﬁnd it very

difﬁcult to understand the physical phenomena. To overcome this problem, students need to practise several

laboratory exercises. Then they need a suitable simulation tool. In this paper a simple cellular model is

introduced and a CDMA software based on that model is described. Its effectiveness as a teaching method is

explained. This simulator can help students to understand the CDMA network behaviour more accurately.

1 INTRODUCTION

The exciting pace of wireless telecommunications

system evolution has marked the last few years. This

fast evolution is the result of the huge success of digi-

tal mobile systems and the increasing demands of mo-

bile users for Internet oriented applications. Second

generation systems mainly offered telephony services

and data transmission with moderate rates. Third gen-

eration networks have multimedia capabilities, such

as the support for high bit rates and the introduction

of packet data/IP access.

Main third generationsystems are UMTS (Univer-

sal Mobile Telecommunication System) standardized

within 3GPP (Third Generation Partnership Project)

(Lagrange, 2005) and CDMA2000, which is an evo-

lution of the north-American system IS-95 and which

is standardized by 3GPP2.

Both air interfaces use the CDMA (Code Division

Multiple Access). CDMA is suited for data trans-

fer with bursty behaviour. However, CDMA is much

more complex than classical Time Division Multiple

Access (TDMA). All users in a cell are transmitting

on the same frequency and thus create interference to

each other. Power control has, therefore, a major im-

pact on a network based on CDMA. Each transmit-

ter has to adapt its transmission power to the optimal

level that allows both a good quality of service and

a high capacity. That transmission power level in-

creases with the number of the users. Hence coverage

and capacity are two antagonist notions. The higher

the load is the lower the coverage is.

In this paper, a user-friendly simulation software

that may ease understanding of the behaviour of

CDMA-based cellular networks is presented. This pa-

per is divided into seven sections. Section 2 is a re-

minder of spread spectrum and cellular CDMA sys-

tems. Section 3 describes the simulation tool con-

text and objectives. Section 4 introduces the model

used in the simulator software and the main func-

tions. Section 5 gives the simulator features. Section

6 presents some laboratory exercises for students that

can be achieved through the simulator tool. Section

7 gives some feedbacks on the use of the tool. Fi-

nally, section 8 proposes some possible evolutions of

the simulator.

Lagrange X. and Karoui B. (2009).

UNDERSTANDING THE BEHAVIOUR OF CDMA-BASED CELLULAR NETWORKS WITH A USER-FRIENDLY SIMULATION SOFTWARE.

In Proceedings of the First International Conference on Computer Supported Education, pages 31-38

DOI: 10.5220/0001851000310038

 SciTePress

2 CELLULAR AND CDMA

FUNDAMENTALS

2.1 Spread Spectrum Interest

In a transmission system, the correct reception of a

transmitted information bit requires a certain rela-

tionship between the energy received per information

bit E

and the noise spectral density N

. The ratio

must be higher than a threshold to guarantee

a maximum bit error rate (BER). That threshold can

be assumed in the ﬁrst approximation step to be inde-

pendent of the transmission and modulation scheme.

Consequently that threshold is given by :

≥ (E

)

(1)

where the threshold (E

)

is typically 6 dB for a

−3

BER (ratio 4 in a linear scale).

In direct-sequence spread spectrum systems, de-

vices transmit chips instead of bits at a higher rate

than necessary. For a user bit rate R, chips are trans-

mitted at a rate W : each bit is replaced by a sequence

of W/R chips. In order to keep good spectrum prop-

erties the sequence is pseudo-random and hence the

bandwidth of the transmitted signal is W for simple

modulation schemes as opposed to the bandwidth of

the user bit ﬂow that is R. As a conclusion, the spec-

trum is spread and ratioW/R is then called the spread-

ing factor or the spreading gain.

In the transmitter, each data bit is multiplied with

the spreading sequence and then sent on the air. In

the receiver, the spread signal is also multiplied by the

same sequence and integrated over the bit period (this

is the simplest receiver, for an exhaustive presenta-

tion see (Proakis, 2001)). This operation is called de-

spreading. The original user ﬂow is then recovered.

The signal is generally interfered by thermal noise

and interferences. However, de-spreading is neutral

for all noise and interferences thanks to the random

properties of the spreading sequence. Let C be the

received signal power at the receiver. Let I

tot

be the

power of all the interferences. The following formula

is admitted:

tot

(2)

Let β = (E

)

. Constraint 1 may be combined

with equation 2 and be re-written as follows:

tot

> β (3)

Equation 3 shows that β is the minimum required

C/I. In UMTS (Holma and Toskala, 2004), W = 3.84

Mchip/s and R depends on the user service but is typi-

cally 15000 bps for voice services. HenceW/R = 256

(spreading gain 24 dB). In all casesW/R is larger than

1 and consequently

tot

may be lower than 1. As a

conclusion, spread spectrum systems tolerate a level

of noise and interference much higher than the signal.

For instance with (E

)

= 6 dB and a spreading

gain of 24 dB then β = −18 dB. For the system to

work well C/I

tot

must be just higher than −18 dB.

2.2 CDMA Principle

Code division multiple access (CDMA) is a form of

multiplexing that does not divide up the channel by

time (as in TDMA:Time Division Multiple Access) or

frequency (as in FDMA:Frequency Division Multiple

Access) but by codes. CDMA uses direct-sequence

spread spectrum but the sequences are based on Walsh

codes constructed with Hadamar matrices (Viterbi,

1995). The spreading sequences of the different users

are orthogonal. However, the orthogonality is only

guaranteed if synchronization is possible between the

different spreading sequences. Such synchronization

is simple when there is a common transmitter : it

can then be used on the downlink of cellular sys-

tems as the same base station is transmitting to all

mobiles of the cell. However, it is not possible on

the uplink as mobiles cannot be synchronized to each

other at the chip level because of variable propaga-

tion delays. Different users then use different pseudo-

random spreading codes on the uplink (Viterbi, 1995).

To illustrate the access of the CDMA type, we can

make an analogy with a gathering of people from dif-

ferent nationalities where everyone is speaking at the

same time but is using different languages. A new-

comer who for example understands only French lan-

guage would be able to extract only the French sen-

tences stated of the ambient hubbub. Conversations

of the others seem to him like a noise deprived of

mean. However, if the general level of noise coming

from the other conversations is too important, it will

be more difﬁcult or even impossible for our visitor to

understand the speech of his compatriot.

2.3 Interference Analysis in a Regular

Network

In a cellular network mobiles do not receive a perfect

signal from their serving base station. The signal is af-

fected by noise and interference. The noise is mainly

due to thermic agitation of electrons (thermal noise)

and to imperfections of the ﬁrst ampliﬁer stage. It is

denoted by N

CSEDU 2009 - International Conference on Computer Supported Education

Figure 1: Interferences in the UpLink come from mobiles

in the other cells and in the same cell.

Interference is mainly due to the reuse of frequen-

cies in several cells and is called co-channel interfer-

ence. Other types of interference are negligible com-

pared to the co-channel interference. In CDMA net-

works the same frequency is used for all terminals in

all cells. The level of interference is then high com-

pared to TDMA or FDMA networks.

Due to the limitation of the text length, we develop

the impact of the interference on the uplink (terminal

to network). The same phenomenons are observed

on the downlink though there is some differences due

to the possibility to have orthogonal codes between

transmissions to different users of the same cell (La-

grange, 2000).

A user terminal T is transmitting on frequency f

to its serving base station B in one cell. All termi-

nals in the same cell are transmitting on f

to B and

generate internal interference. Furthermore f

is also

used by all terminals connected to other base stations.

Such transmissions generate external interference (for

T). Let I

int

be the internal interference and I

ext

be the

external interference (see ﬁg 1). TheC/I ratio may be

written as :

C/I

tot

= C/(I

int

+ I

ext

+ N

) (4)

2.4 Power Control

In a CDMA system power control is a very important

issue because all the users of the network use the same

frequency band at the same time.

The general objective of power control is to ensure

the same E

ratio for all users. In order to mini-

mize the power consumption the E

target must

be the lowest acceptable value. That is :

= (E

)

(5)

In order to reach the (E

)

, each mobile must

adapt its transmission power. As N

is the same for

all mobiles in the same cell, reaching an E

target

is equivalent to reaching a common signal target de-

noted as C

. The more mobile there are, the higher

is and hence the higher C

is.

The maximum capacity is obtained when all mo-

biles are received at the same power level at the base

station. However, that level is not the same in all base

stations. Determining the level target for each cell is

a complex task in an operational network. Iterative

algorithms that try to approach the optimal solutions

are used. In a simple regular network it is possible to

analytically determine that target (see (Zander, 1992)

for more details). Such an approach is used in this

contribution but is not detailed for the sake of sim-

plicity.

2.5 Notion of Pole Capacity

If only one cell is considered and if the thermal noise

is negligible then with a perfect power control the

C/I

tot

is given by:

C/I

tot

(M − 1)C

= 1/(M − 1) (6)

where M is the number of mobiles in the considered

cell. Equation 6 and constraint 3 demonstrate that the

maximum number of mobiles able to transmit in one

cell is limited. Even if mobiles were able to transmit

at very high power it is never possible to have more

than

+1 active mobiles in one cell. The pole capac-

ity is then deﬁned as the maximum number of mobiles

per cell in a network with inﬁnite power mobiles. An

upper bound of the pole capacity is

+ 1 but deter-

mining it is not simple when external interference is

taken into account.

2.6 Variation of the Coverage

According to the Load

The coverage in a CDMA system is directly related

to the load of the network. The load is deﬁned as the

number of active users.

The more the number of mobiles in the cell in-

creases, the more they must increase their emission

power. However, each mobile is characterized by a

maximum transmission power. Consequently, mo-

biles that are located at the edges of the cell are no

longer able to reach the base station (the required

power would be above the maximum value). This

UNDERSTANDING THE BEHAVIOUR OF CDMA-BASED CELLULAR NETWORKS WITH A USER-FRIENDLY

SIMULATION SOFTWARE

phenomenon is known as cell breathing : coverage

decreases as the load increases.

The exact coverage depends on several parameters

like the data rate allocated to each user, the level of

noise but especially the load on the cell. In a simple

regular network with perfect power control, the cov-

erage can be calculated on the uplink and downlink.

3 SIMULATION TOOL

OBJECTIVES

At TELECOM Bretagne (Ecole Nationale Superieure

des Telecommunications de Bretagne) cellular net-

work fundamentals are given in ﬁrst year of gradu-

ation studies (MSc level). The second year is rather

dedicated to more advanced systems like 3G and 4G

wireless networks. In this context, spread spectrum

concept and CDMA is taught. As it was introduced in

part 2 there is a lot of interaction in cellular networks

due to the impact of the interference. Understanding

how a network ”lives” is not simple. Sometimes stu-

dents are confused by the technical details and do not

understand the basic physical phenomena. It is then

necessary to have labs where the student can practice.

As network UMTS devices are very expensive, simu-

lation is the only way for laboratories.

3.1 Simulation Tool Overview

Indeed, major simulation tools like OPNET (OPNET,

2008) or COMNET (COMNET, 2008) are more pro-

tocol oriented and require a lot of time to be used cor-

rectly. Furthermore the multi-cell aspect is generally

not included. Other planning tools such as ATOLL

(FORSK, ) or PLANET (MARCONI, 2008) are more

adapted to predict the coverage in a given environ-

ment. They use sophisticated digital elevation models

to predict coverage and evaluate the quality of service

delivered in a zone. The behaviour of a network for

different loads is precisely modelled and may be dis-

played. However, a huge number of parameters have

to be manipulated. When some variations in the cov-

erage are apparent, it is very difﬁcult to identify the

reason for them. For example, the coverage level may

be deeply impacted by some speciﬁcities of the prop-

agation models rather than the basic physical CDMA

characteristics.

There is then a need for a user-friendly simula-

tion software of a CDMA network that reﬂects the

behaviour of a CDMA wireless network.

3.2 VICTOR Objectives

The objective at TELECOM Bretagne was to have a

tool that every student could launch and use within

5 minutes. Simplicity was the ﬁrst objective. The

simulation is thus not event-oriented but is closer to

a Monte-Carlo approach. The developed software is

based on a model which will be described in the fol-

lowing and the simulator is called VICTOR (Visuali-

sation Interactive du ConTrOle de puissance dans un

Reseau regulier) : Interactive Visualization of power

control in a regular network.

4 MODEL PRESENTATION

4.1 The Network Model

A regular hexagonal network with omnidirectional

base stations is considered in VICTOR. The number

of cells is set to 7 : one central cell is surrounded by

six neighboring cells. Terminals are randomly spread

over the 7 hexagonal cells. They are assumed to be

ﬁxed. However, they are called mobiles, as it is usu-

ally done in cellular networks.

The radius of the cells may be changed by the user

but the default value is 1 km. A regular propagation is

considered also in VICTOR. The effect of obstacles

is not simulated in details. Irregularities in the prop-

agation are modelled by a random variable. Such an

approach is very classical in much research work (see

(Gilhousen et al., 1991) for instance). The user has

access to a large number of parameters (propagation

law, spread spectrum factor, noise ﬁgure,... ). How-

ever, predeﬁned values that correspond to a typical

coherent system are set by default.

4.2 Main Functions of the Software

VICTOR propose two possibilities to ﬁx its conﬁgu-

ration parameters: i) at the beginning of the simula-

tion, ii) at any time in the simulation. Once the pa-

rameters are ﬁxed, the user can very easily change the

number of mobiles in the network through the slider.

For a given number of mobiles, VICTOR calculates

the target level of the reception power in the down-

link and in the uplink. It derives the required power

level for each mobile on the uplink and for each base

station on the downlink.

In some cases, it is not possible to serve all mo-

biles. Mobiles that cannot be served are called inac-

tive mobiles : on the uplink the transmission power

of inactive mobiles is set to 0. On the downlink, no

CSEDU 2009 - International Conference on Computer Supported Education

Choose

uplinkor

downlink

Changethe

numberofmobiles

Parametersofthe

clickedbase

Figure 2: All mobiles are covered in the network: 10 mobiles in the central cell and 70 mobiles in the whole network.

power is dedicated to inactive mobiles. The commu-

nication of such mobiles would be cut or not estab-

lished in an operational network.

Once the transmission powers of the mobiles and

the base stations are deﬁned, VICTOR computes the

internal and external interferences for each mobile

and for each base station, the path loss and many other

ﬁelds. All the computation is done both on the uplink

and the downlink. The user chooses either the down-

link display or the uplink display. He or she can very

easily switch from one view to another one thanks to

a radio button .

5 OVERVIEW OF THE

SIMULATOR

VICTOR is programmed in Java language and

can run on any operating system. There is

also an applet version of VICTOR which can be

used from any browser (http://formations.telecom-

bretagne.eu/ressources/rsm). The software is approx-

imately 10 000 lines.

The main principle window of VICTOR is divided

in 4 parts (see ﬁg. 2). The network is represented by

the cells, the different base stations and mobiles. In

the right of the window, some gauge indicators are

disposed : i) in the uplink case, the two gauges rep-

resent the target power and the total interference (in-

ternal, external and thermal noise) of the base station

of the central cell; ii) in the downlink case, the two

gauges indicate the target power and the total interfer-

ence for a test mobile. In addition to the two gauges,

there are two indicators of the number of mobiles and

the number of satisﬁed mobiles (well served mobiles)

on both the central cell and the network.

VICTOR offers the possibility to select the visu-

alization of the network either on the uplink or on the

downlink by a radio operator button. The position of

the mobiles in the different cells can be changed by

carrying out a new random choice (while pressing on

the new Snapshot button). Finally, the user may see

the information such as the received power, the inter-

nal interference,... by clicking on a mobile or a base

station.

Let us consider the case of 10 mobiles in each cell

of the network (see ﬁg. 2). In the central cell, the

interference level is weak (-103 dBm) and the target

power level (-116 dBm) is low enough to serve all the

mobiles. Consequently, all the mobiles are covered

in the central cell (10 mobiles) and in the whole net-

work (70 mobiles). Fig. 2 shows all the information

of the base station on the uplink such as the internal

interference, the external interference and so on.

Let us now consider 30 mobiles in each cell (see

UNDERSTANDING THE BEHAVIOUR OF CDMA-BASED CELLULAR NETWORKS WITH A USER-FRIENDLY

SIMULATION SOFTWARE

ﬁg. 3). Mobiles that do not have enough power to

reach the base station are barred with a cross. As it

is depicted in the latter ﬁgure, the number of covered

mobiles (110 mobiles) is lower than the number of

mobiles in the network (210 mobiles). For example,

if the user looks at the central cell there are only 16

active mobiles among the 30 mobiles because of the

high interference level (-87 dBm). Consequently, the

user deduces that the cell has attained its pole capacity

(16 mobiles). Fig. 3. shows all the information of the

mobile station on the uplink such as the transmission

power, the path loss... All these examples are given

for uplink. However, VICTOR can also be used to

visualize the pole capacity notion on the downlink.

Figure 3: Pole capacity on the uplink: 16 mobiles in the

central cell and 110 in the network.

6 LABORATORY EXERCISES

In lectures, the basic concept of a CDMA network and

related issues and challenges are introduced. These

concepts are reinforced when the students set up the

laboratory exercises. In the following we present sev-

eral labs.

Laboratory 1: The Cell Breathing Phenomenon

on Uplink. In this introductory lab, students are ﬁrst

asked to analytically determine the maximum radius

of a cell on the uplink without any load. By consid-

ering only thermal noise, they calculate the minimum

level received power C

min

by combining equations 3

and 4. Using propagation equations withC

min

and the

mobile transmission power they then ﬁnd the range of

mobile terminals. The obtained value is much higher

than the radius of a hexagonal cell. This shows the

coverage for a non loaded network is very good.

Students then use VICTOR and look at the num-

ber of mobiles that can reach the base station for dif-

ferent loads. They realize the total interference level

increases with the number of mobiles in the network.

They are asked to determine the maximum number of

mobiles served. They are then asked to increase the

transmission power of mobiles by typically 3 dB to

check that the capacity increases slightly. They repeat

the same process (power increase). Above a certain

power they note the capacity does not increase any-

more : the pole capacity has been reached. They note

this maximum is the same for different snapshots. In

completing this practical, students develop a sound

knowledge of the impact of interference in a CDMA

system and an understanding of the pole capacity.

Laboratory 2: The CellBreathing Phenomenon on

Downlink. As in lab 1, students are asked initially to

calculate the pole capacity on the downlink case and

to check it with VICTOR. They notice that beyond a

certain power of the base station the number of cov-

ered mobiles does not exceed the already theoretically

calculated value. The pole capacity has been reached.

In completing this practical, students develop a better

understanding about the pole capacity on the down-

link case.

Laboratory 3: Impact of the User Rate. CDMA is

interesting in multimedia cellular networks because it

provides ﬂexibility. A network is able to accommo-

date different user rates R. As the bandwidth W re-

mains the same, increasing the user rate decreases the

spreading factor. Hence, higher rates require higher

power. In this lab, students are asked to increase the

user rate and to determine the impact on the capacity

both on the uplink and the downlink. Let us note that

VICTOR is not able to mix different user rates in the

same simulation like a UMTS network. However, it

is possible to change that parameter and to see the im-

pact of a new value with exactly the same position of

mobiles. Students can then really understand the im-

pact of high rates and the corresponding cost in term

of power budget.

Laboratory 4: TDMA vs CDMA. Due to multipath,

the different transmissions of a base station are not

perfectly orthogonal. This is generally modeled by

integrating in the interference calculation a non or-

thogonality factor α. This factor may vary from one

mobile to another and depends on its position in the

network. However, for the sake of simplicity the same

value is considered for all mobiles in VICTOR.

Students ﬁx the number of mobiles per cell to the

pole capacity of the uplink and they notice that all

mobiles are covered on the downlink for α = 0.5. In-

CSEDU 2009 - International Conference on Computer Supported Education

deed, the internal interference value on the downlink

is much lower than on the uplink thanks to orthogonal

codes on the downlink. Students are shown that for

α equals to 0, the internal interference is eliminated.

Consequently, the maximum capacity increases be-

cause mobiles receive less internal interference and

need less power. For α = 1, the capacity is lower due

to the rising internal interference in the cell. In this

case, the base station must redistribute equitably its

power between the mobiles.

In the following, a complementary lab describ-

ing the behaviour of a CDMA system compared to a

TDMA system is presented. Let us consider the case

of a strong non-orthogonality on the downlink. Two

strategies are proposed: i) the base station transmits

simultaneously two user ﬂows on two different codes;

ii) the base station transmits one ﬂow, then the other,

by reducing the spreading factor and by increasing the

instantaneous rate (case of the TDMA). Students are

asked then to consider the two strategies. Students

show that in the ﬁrst strategy there is internal inter-

ference but in the second strategy internal interfer-

ence is eliminated. Consequently, the network perfor-

mance becomes much better. Students may then con-

clude that TDMA is better on the downlink of a given

base station when there are no strict delay constraints.

In completing this practical, students gain experience

in modelling orthogonality on the downlink and re-

ceived a good introduction to the basis of 3.5 systems

like HSDPA (High Speed Data Packet Access) that

rely on a time shared downlink channel.

7 ASSESSMENT

7.1 Evaluation by Students

VICTOR was used during autumn semester in 2006-

2007 and 2007-2008. In 2006-2007, there were 32

students who were registered for the course unit. As

it was a trial, they did only labs 1 and 2 after the the-

oretical courses.

VICTOR software was evaluated anonymously by

students. The aim of the evaluation is to check VIC-

TOR is really easy to use and to analyze the percep-

tion of students on VICTOR’s beneﬁts. Students were

asked 5 questions as shown in ﬁgure 4. One addi-

tional open question asks them to give their opinion

on VICTOR. More than 95% ﬁnd the software is easy

or very easy to use. They do not spend much time

to become familiar with VICTOR as more than 65%

needed less than 5 minutes to that. Some students who

have an above average ability to follow theoretical

courses found VICTOR unnecessary since the con-

Howlongtobecomefamiliarwith

Victor

lessthen5

minutes

between5

and10

minutes

morethan

10minutes

% of responses

EaseofuseofVictor

very

easyto

use

easyto

use

difficult

touse

very

difficult

touse

% of responses

Victoris

unncessary

becausethe

conceptsstudied

areverysimple

Victoris

unncessary

becausethe

conceptsstudied

aretoocomplicated

Victorisagood

tooltounderstand

thebehaviorofa

CDMA network

Victorisagood

toolbutitis

necessarytowork

onthelessons

before

% of responses

IthinkIhaveabetterunderstanding

ofCDMA networksthankstoVictor

No Yes

% of responses

BenefitsofVictor(score0to3)

0 1 2 3

% of responses

Figure 4: Subjective evaluation of VICTOR by students.

cepts covered were simple. However, more than 95%

found it useful. Note that 61% consider VICTOR is

useful but does not exempt them from working on the

theoretical lessons. Students were asked to write one

sentence for their evaluation of VICTOR. Here are a

few answers:

• VICTOR is a good way to have practical applica-

tions of the theoretical courses,

• VICTOR facilitates understanding of the course,

• VICTOR is a simple software that really helps to

ﬁx the basic concepts of CDMA networks,

• I won’t say that I learnt new things. However, I

have understood why some elements of the course

are useful and thanks to VICTOR I asked myself

good questions. It is then a very useful help to

understand the course.

7.2 Objective Evaluation with a Quiz

Before the labs, students were asked to make a techni-

cal multiple-choice test comprising 12 questions. The

same test was used after the labs to analyze the bene-

ﬁts of the labs.

UNDERSTANDING THE BEHAVIOUR OF CDMA-BASED CELLULAR NETWORKS WITH A USER-FRIENDLY

SIMULATION SOFTWARE

Score before the labs Sore after the labs

Figure 5: Scores of all students before and after the labs (sorted in ascending order for the ﬁrst test.)

The individual score of each student is given in

ﬁgure 5. Most students (78%) got a better or an equal

score after the labs. The average score was 6/10 be-

fore the labs and 7.3 after the labs. Strangely, 6 stu-

dents got a lower score after the labs. This is due to

two different factors : ﬁrstly, some students answered

randomly to some questions. Secondly, the questions

were short. Hence, a few questions were partly am-

biguous. Several students thought they include traps.

Instead of giving the simple direct answer, they mis-

interpreted the question and gave a wrong answer

though they had well understood the concept. One

can notice that the students who had a low score at

the ﬁrst test improved their score greatly in the sec-

ond test. This should suggest that these students have

the ability to understand the concepts but have difﬁ-

culties in classical lessons. For such students lab work

is very important.

8 CONCLUSIONS

VICTOR can be used either for class demonstra-

tions, in enhancement of the traditional lecture envi-

ronment, or in the computer laboratory for hands-on

practical work in a UMTS networking course. It is

easy to use and provides a suitable interface that helps

users to gain better understanding of UMTS network.

The possibility of VICTOR enhancement are nu-

merous. However,the challenge is to keep a very sim-

ple man-machine interface to have software which is

very easy to use. Students may focus their attention

on the radio and network aspects and not on how to

use the software.

ACKNOWLEDGEMENTS

The authors want to thank students of TELECOM

Bretagne who also contributed to the development

of VICTOR software : Raymond Coulibaly, Souley-

mane Faye, Youness Oulbacha, Jihad Sarsi.

REFERENCES

COMNET (2008). Comnet description. Technical report,

http://www.comnet.at/website/output.php.

FORSK. Atoll description. Technical re-

port, http://www.forsk.com/htm/products/atoll

overview.htm.

Gilhousen, K., Jacobs, I., Padovani, R., Viterbi, A., Weaver,

L., and Wheatley, C. (1991). On the capacity of a cel-

lular cdma system. IEEE Transactions on Vehicular

Technology, 40(2):303–311.

Holma, H. and Toskala, A. (2004). WCDMA for UMTS,

third Edition. Wiley and sons Ltd.

Lagrange, X., editor (2000). Les reseaux Radiomobiles.

Edition Hermes Science.

Lagrange, X., editor (2005). Principes et Evolutions de

l’UMTS. Edition Hermes Science.

MARCONI (2008). Planet description. Technical report,

http://www.marconi.com.

OPNET (2008). Opnet description. Technical re-

port, http://www.opnet.com/products/modeler/home-

1.html.

Proakis, J. (2001). Digital Communications. Mc Graw Hill.

Viterbi, A. (1995). CDMA, Principles of Spread Spectrum

Communication. Addison-Wesley.

Zander, J. (1992). Performance of optimum transmitter

power control in cellular radio systems. IEEE Trans-

actions on Vehicular Technology, 41:57–62.

CSEDU 2009 - International Conference on Computer Supported Education