IEEE TRANSACTIONS ON EDUCATION, VOL. 53, NO.

4, NOVEMBER 2010 653

A Wireless Communications Laboratory on Cellular

Network Planning
Zaher Dawy, Senior Member, IEEE, Ahmad Husseini, Elias Yaacoub, and Lina Al-Kanj

Abstract—The field of radio network planning and optimization enhancements to GSM networks that are capable of supporting
(RNPO) is central for wireless cellular network design, deploy- packet switched mobile services and providing peak data rates
ment, and enhancement. Wireless cellular operators invest huge of 128 kb/s with GPRS and 384 kb/s with EDGE. The enhanced
sums of capital on deploying, launching, and maintaining their net-
works in order to ensure competitive performance and high user data rates offered by GPRS and EDGE are based on various
satisfaction. This work presents a lab course composed of 10 exper- advanced techniques that include adaptive modulation and
iments that cover the different phases of RNPO for various state-of- coding.
the-art wireless technologies such as GSM, UMTS, and WiMAX. In order to meet the forecasted growth of cellular subscribers
Each experiment constitutes a complete entity including the nec-
and the need for faster and more reliable data services, the
essary theoretical background and references, the lab tasks based
on real-world scenarios, and a research component consisting of third-generation UMTS system was developed and standardized
general questions. The proposed lab course represents a novel ini- in 1999, based on a code division multiple accessing (CDMA)
tiative to increase interactive learning by integrating communica- air interface. In a UMTS network, the frequency reuse factor
tions theory fundamental knowledge with state-of-the-art wireless is one, and thus all users in the network share the same
communications software tools and measurement equipment. The
experiments are carefully designed to enhance the analytical skills frequency band at the same time by using different spreading
and to advance the academic and practical knowledge of the stu- codes in order to limit the level of interference. The wideband
dents. CDMA (WCDMA) mode of UMTS is currently being widely
Index Terms—Cellular network planning, drive testing, GSM, deployed by cellular operators all over the world (more than 400
UMTS, WiMAX, wireless communications. million subscribers in more than 130 countries) [3]. WCDMA
provides higher spectral efficiency than GSM/GPRS/EDGE,
with peak data rates of 2 Mb/s. HSPA is a direct enhancement
I. INTRODUCTION
to UMTS networks that can provide an uplink peak data rate
of 5.7 Mb/s and a downlink peak data rate of 14.4 Mb/s. The
IRELESS cellular technologies are continuously
W evolving to meet the increasing demands for high data
rate mobile services. GSM (Global System for Mobile Commu-
enhanced data rates offered by HSPA are based on various
advanced techniques that include intelligent scheduling, fast
link adaptation, and fast retransmission. Long-Term Evolution
nications) is currently the most widely deployed cellular system
(LTE) is currently being developed and standardized as the
with more than three billion subscribers in more than 200 coun-
next step after UMTS/HSPA on the evolution track [2]. LTE is
tries [1]. GSM is based on a hybrid time division/frequency
based on an orthogonal frequency division multiple accessing
division multiple accessing (TDMA/FDMA) air interface. The
(OFDMA) air interface and is capable of providing considerably
allocated spectrum for a given GSM network is divided into
high peak data rates of 50 Mb/s in the uplink and 100 Mb/s
multiple frequency channels that are distributed among cells
in the network based on a frequency plan in order to control in the downlink. At the same time, the WiMAX (Wireless
the level of interference. In each frequency channel, time is Interoperability for Microwave Access) wireless technology is
divided into eight time slots. The proliferation of the Internet being further developed as an all-IP network with an OFDMA
and the continuously escalating market demand for mobile data air interface to provide high data rate broadband wireless mobile
services necessitated the evolution of GSM by developing new access [4]. WiMAX evolved from a last-mile wireless access
technologies such as GPRS (General Packet Radio Service), technology to a broadband wireless access technology, with
EDGE (Enhanced Data rates for GSM Evolution), UMTS nomadic and mobile connectivity as standardized in IEEE
(Universal Mobile Telecommunication System), and HSPA 802.16d and IEEE 802.16e, respectively.
(High-Speed Packet Access) [2]. GPRS and EDGE are direct In academia, teaching wireless communications is normally
oriented toward providing students with the necessary theoret-
ical background. Usually students do not have the opportunity
Manuscript received February 22, 2009; revised December 03, 2009. First
published February 05, 2010; current version published November 03, 2010. to work with wireless communications tools and equipment re-
Z. Dawy, E. Yaacoub, and L. Al-Kanj are with the Department of lated to state-of-the-art wireless technologies. At the Electrical
Electrical and Computer Engineering, American University of Beirut, and Computer Engineering (ECE) Department at the American
Riad El-Solh/Beirut 1107 2020, Lebanon (e-mail: [email protected];
[email protected]; [email protected]). University of Beirut (AUB), Lebanon, there are several courses
A. Husseini is with Ericsson-GSDC Lebanon, Beirut, Lebanon (e-mail: that are offered in the area of wireless communications, but a
[email protected]). wireless communications lab to complement the course mate-
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. rial was lacking. Hence, the idea of developing a wireless com-
Digital Object Identifier 10.1109/TE.2009.2039935 munications lab course evolved in order to provide students with

0018-9359/$26.00 © 2010 IEEE

654 IEEE TRANSACTIONS ON EDUCATION, VOL. 53, NO. 4, NOVEMBER 2010

important practical experience as part of their university studies. (BS) site of one of the local cellular operators. At the end of the
The proposed lab course introduces students to the field of radio semester, students are required to take a comprehensive final
network planning and optimization (RNPO) through a set of 10 exam.
coherently designed lab experiments based on state-of-the-art The lab course is designed and developed around the theme
software tools and measurement equipment. The developed ex- of RNPO. The process of RNPO normally starts with a basic
periments strengthen the concepts and theories studied in lec- set of inputs from the cellular operator, such as the services to
ture-based courses and provide students with profound under- be offered, the expected number of users per service, the areas
standing of wireless technologies such as GSM, UMTS, and to be covered, and the maximum tolerated blocking rate. The
WiMAX. The proposed lab course is applicable to any cur- main output of RNPO is the overall network layout, i.e., the lo-
riculum or track in the area of communications engineering. It cations and configurations of all BSs to be deployed over the
complements fundamental material covered in basic courses on area of interest. The lab course covers the three phases of RNPO:
digital communications or wireless communications. preplanning or dimensioning, detailed planning, and post-plan-
This paper is organized as follows. In Section II, a survey of ning or optimization [19]. It covers RNPO for the three state-of-
wireless communications lab courses from other universities is the-art wireless technologies GSM [20], UMTS [21]–[24], and
presented. In Section III, the general characteristics, objectives, WiMAX [25]. It is based on the following professional network
and student learning outcomes of the proposed lab course planning tools and measurement equipment: Ericsson TEMS
are provided. The developed lab experiments are described CellPlanner for GSM and UMTS network planning [26], ATDI
in Section IV. Assessment data is presented and analyzed in ICS telecom for WiMAX network planning [27], and Rhode &
Section V. Finally, conclusions are drawn in Section VI. Schwarz drive testing equipment with ROMES 3NG software
for GSM and UMTS network optimization [28]. Moreover, it
II. LITERATURE SURVEY makes use of several customized software tools that were inter-
nally developed using Excel and MATLAB. It should be noted
Most wireless communications lab courses offered at other
that in the context of this work, the term “optimization” refers to
universities deal with basic wireless communications concepts
the process of enhancing the network performance after having
such as the design of certain components of wireless transmit-
analyzed the measurement results and identified problems re-
ters and receivers, the implementation of signal processing al-
lated to the operation of the network; see, for example, [19] and
gorithms, or the modeling of wireless channels. They are nor-
[21]–[24].
mally either based on hardware equipment such as signal gener-
The objectives of the lab course are to:
ators and spectrum analyzers or based on simulation tools using
1) provide students with hands-on experience in radio net-
MATLAB, Simulink, or LabVIEW. On the other hand, some
work planning tools and techniques;
universities have designed experiments related to RNPO as part
2) provide students with hands-on experience in wireless
of regular lecture-based courses. These experiments focus on a
drive testing measurement systems and analysis;
single wireless technology and might include the use of a pro-
3) train students to analyze, evaluate, and improve actual
fessional network planning tool. Table I provides examples of
wireless network plans and channel models;
wireless communications lab courses and RNPO-related lab ex-
4) provide students with an understanding of wireless com-
periments that are offered at various universities.
munication technologies such as GSM/GPRS/EDGE,
The proposed lab course is unique in terms of its comprehen-
UMTS, and WiMAX;
sive and coherent coverage of all aspects of RNPO for several
5) provide students with practical background in wireless
state-of-the-art wireless technologies such as GSM, UMTS, and
communications that will allow them to practice in this
WiMAX. Moreover, it provides students with hands-on experi-
field.
ence in multiple professional network planning and optimiza-
The following are the student learning outcomes of the lab
tion tools. To the best of the authors’ knowledge, there are no
course:
similar lab courses offered at any other university.
1) understand the basic principles of radio network planning
and optimization;
III. LAB OBJECTIVES AND LEARNING OUTCOMES 2) understand the basic design and operation of the following
The lab course consists of 10 experiments. Each experiment wireless technologies: GSM, UMTS, and WiMAX;
is composed of three main parts: a theoretical background sec- 3) know how to perform link budget analysis and network
tion, a detailed description of the lab tasks and deliverables, and dimensioning for GSM and UMTS cellular networks;
a set of general questions. The lab course is based on an open-lab 4) know how to use professional radio network planning
policy where students in groups of two or three access the lab in tools such as TEMS CellPlanner (GSM, UMTS) and
reserved time slots. Each experiment normally requires around ATDI (WiMAX);
3 h of lab work and is preceded by a 1-h lecture to provide the 5) know how to perform radio network planning and capacity/
necessary background information on the underlying theoret- coverage analysis for GSM networks;
ical concepts and the tools used. Students are required to submit 6) know how to perform radio network planning and capacity/
weekly post-lab reports describing and analyzing the results of coverage analysis for UMTS networks;
their work. During the semester, students are required to par- 7) understand the importance and basics of GSM/UMTS
ticipate in site visits that include a visit to a real base station co-siting and coexistence;
DAWY et al.: A WIRELESS COMMUNICATIONS LABORATORY ON CELLULAR NETWORK PLANNING 655

TABLE I
WIRELESS COMMUNICATIONS LAB COURSES OFFERED AT VARIOUS UNIVERSITIES

8) know how to perform radio network planning and capacity/ summary of the learning objectives and tools used for each lab
coverage analysis for GSM/UMTS seamless networks with experiment. The authors would be pleased to share the lab exper-
co-siting; iments with any interested reader. The experiments are posted
9) know how to perform radio network planning and capacity/ online at the following link: http://webfea.fea.aub.edu.lb/wire-
coverage analysis for WiMAX networks; lesslab/.
10) understand the importance of drive testing for network op-
timization; A. Experiment 1: Parameter Estimation for Link Budget
11) know how to analyze real GSM drive testing log files; Analysis
12) know how to analyze real UMTS drive testing log files.
In Experiment 1, students understand important parameters
IV. LAB EXPERIMENTS for the link budget analysis (LBA) and estimate their values
This section describes the 10 experiments that constitute the using internally developed MATLAB simulation tools. The sim-
RNPO wireless communications lab course. Table II provides a ulation tools model various wireless transmitter and receiver
656 IEEE TRANSACTIONS ON EDUCATION, VOL. 53, NO. 4, NOVEMBER 2010

TABLE II
LAB EXPERIMENTS, UTILIZED TOOLS, AND LEARNING OBJECTIVES

order to obtain values for LBA parameters such as the uplink

and downlink interference margins, slow and fast fading mar-
gins, soft handover gain, diversity gain, and required for
different modulation and coding schemes. This experiment is
highly related to communications theory concepts that students
learn in typical lecture-based wireless communications courses.
As an example, Fig. 1 shows the graphical user interface (GUI)
for the simulation tool used to estimate the uplink interference
margin. Students are required to analyze and comment on the
results and plots obtained.

Fig. 1. GUI for the MATLAB simulation tool used to estimate the uplink in- B. Experiment 2: Link Budget Analysis and Network
terference margin in Experiment 1. Dimensioning
After completing Experiment 1, the students are ready to
components. Students are required to run MATLAB simula- tackle the subject of LBA, which is the main tool used in
tions, some of which are relatively lengthy (around 30 min), in network dimensioning, the initial phase of RNPO. The output
DAWY et al.: A WIRELESS COMMUNICATIONS LABORATORY ON CELLULAR NETWORK PLANNING 657

required to plan a GSM900 network covering Beirut city. By

answering a set of preliminary questions, students relate what
they have learned from Experiment 2 to the work they have to
do in this experiment. After using a LBA to obtain the approx-
imate number of BSs required to cover the given area, students
start the implementation of the planned network using a digital
map of Beirut city. The implementation tasks are provided to
the students on a step-by-step basis in such a way as to simplify
the use of TEMS CellPlanner. After the various parameters are
explained, students are required to set the network equipment
parameters that include antennas, feeders, tower-mounted am-
plifiers, and tower types. Next, students are required to define a
BS template and to deploy BSs over the given map for Beirut
city. Furthermore, they are required to explain the criteria used
Fig. 2. GSM uplink LBA calculator used in Experiment 2.
for their BS deployment strategy. Fig. 3 shows the deployment
of BSs over the area of interest. This experiment introduces stu-
dents to the TEMS CellPlanner tool and provides them with the
of the dimensioning phase is a rough estimate of the number of necessary background in order to launch the analysis features of
BSs and network elements needed, which is used as an input the tool in Experiment 4.
to the detailed planning phase. The main outcome of the LBA
is the maximum allowed path loss (MAPL) in the uplink and D. Experiment 4: GSM Radio Network Planning—Part II
downlink directions. The direction with the smaller MAPL is After having defined the various network and equipment
selected as the limiting direction if no balancing techniques parameters and after having deployed the different BSs over the
are applied. The LBA takes into account many parameters and area of interest, students are required to perform a network per-
specific BS and mobile station (MS) configurations, such as formance analysis in order to optimize the GSM network plan.
the transmission power capability, antenna gains, losses at the This experiment includes coverage, capacity, and frequency
receiver and transmitter ends, and propagation model suitable planning and analysis. Students perform separate planning and
for the area of interest [22]. analysis for both coverage and capacity, due to their indepen-
Students are required to enter specific values for various pa- dence in GSM networks. Thus, students are required to run
rameters in internally developed Excel tools that perform the the path loss calculation feature of TEMS CellPlanner in order
necessary LBA calculations. The values entered correspond to to check the resulting radio coverage plot. This helps them
the ones obtained in Experiment 1 via MATLAB simulations. in assessing the coverage of the planned network in order to
The students are required to compare different results obtained decide whether to add, remove, or relocate some BSs. Next,
by varying certain parameters. They are also required to study students are required to perform traffic analysis to assess the
the effect of balancing techniques, such as tower-mounted am- network from a capacity point of view. Traffic analysis is
plifiers and receive diversity, on reducing the difference between directly related to the grade-of-service (GoS) parameter, which
the uplink and downlink path losses. As an example, Fig. 2 is the maximum allowable blocking rate. Additionally, students
shows one of the Excel tools used for GSM uplink LBA cal- are required to use a TEMS CellPlanner feature to optimize the
culations. number of transceivers (TRXs) allocated to each BS. Students
vary the user load in the network in order to investigate the
C. Experiment 3: GSM Radio Network Planning—Part I impact on the number of required TRXs. Finally, students
This detailed planning phase constitutes the major part of the are required to perform frequency planning and assess it by
RNPO process. It involves the use of a professional planning evaluating the level of interference in the network. Frequency
tool such as Ericsson TEMS CellPlanner, Aircom ASSET, or planning involves assigning groups of frequency channels to
Nokia Totem Vantage. These tools normally support multiple different cells according to a reuse factor in a way that mini-
cellular technologies, use accurate digital maps for the areas to mizes intercell interference.
be planned, and provide the necessary simulations required to
obtain plots and reports for performance evaluation prior to the E. Experiment 5: UMTS Radio Network Planning
actual deployment and launching of the cellular network. The In this experiment, students are required to plan a UMTS net-
main output of this phase is a cellular network plan that provides work covering Beirut city using the TEMS CellPlanner tool.
sufficient coverage and capacity, meets the quality of service Three UMTS services are considered: speech, circuit-switched
requirements, and allows for flexible future network growth. A video, and packet-switched data. Students are provided with the
better design is the one that requires less network equipment LBA output consisting of an approximate number of needed BSs
and, thus, lower cost while meeting all input requirements. to cover the area of interest. Students are required to plan the
The main blocks of GSM detailed planning are coverage anal- UMTS network over three main stages: The first stage deals with
ysis, capacity analysis, and frequency planning. This experi- the coverage area and area definitions, the second stage involves
ment starts by explaining to students the scenario to be consid- defining the main equipment and resources to be used in the de-
ered and planned using TEMS CellPlanner [26]. Students are sign, and the last stage involves deploying UMTS BSs over the
658 IEEE TRANSACTIONS ON EDUCATION, VOL. 53, NO. 4, NOVEMBER 2010

Fig. 3. TEMS CellPlanner snapshot showing the BSs layout over the area of Beirut city. Each BS is three-sectored. The color density relates to the level of
coverage based on signal strength levels.

area of interest and performing iterative coverage and capacity G. Experiment 7: WiMAX Radio Network Planning—Part I
analysis. Iterative analysis is needed due to the cell breathing In this experiment, the ICS telecom tool from ATDI [27] is
effect in UMTS networks [2]. In UMTS, CDMA is used with a used to plan fixed and nomadic WiMAX networks covering the
frequency reuse factor of one. Therefore, increasing the number area of Beirut city. The experiment starts by explaining the sce-
of users per cell leads to an increase in the level of interference, nario for planning an IEEE 802.16d WiMAX network with cus-
which in turn decreases the cell coverage area. After several it- tomer premises equipment (CPEs) placed on rooftops. Students
erations, the output obtained is the number of BSs that ensure a are required to begin the implementation of the planned net-
balanced network for the uplink and downlink directions [21]. work over multiple steps. First, they are required to define the
Finally, students are required to generate and analyze several appropriate cartography data-set for the area of interest. They
performance plots in order to assess the quality of the network are provided with an advanced digital map containing different
plan. land layers including a buildings layer. Then, they are required
to do a nominal planning by searching for the optimal BS loca-
F. Experiment 6: Radio Network Planning for a GSM/UMTS tions and configuring the BSs with the appropriate parameters.
Seamless Network Furthermore, they are required to generate performance plots in
This experiment covers the concept of GSM/UMTS co-siting order to be able to propose enhancements for the WiMAX cov-
and coexistence. Students are required to perform a three-year erage plan in the area of interest.
plan for the gradual deployment of a UMTS network that coex-
ists with an already-deployed GSM network in the area of Beirut H. Experiment 8: WiMAX Radio Network Planning—Part II
city. They are provided with TEMS CellPlanner files containing Having learned, in Experiment 7 the basics of coverage cal-
the already-existing GSM network and templates of GSM and culation and analysis for the fixed wireless access (FWA) type
UMTS BSs. The service to be offered by the GSM network is as- of WiMAX, where the receivers are located on the rooftops of
sumed to be voice only. The services to be offered by the UMTS buildings, in this experiment students perform coverage calcu-
network are assumed to be voice, streaming video, and Internet lation and analysis for the FWA type of WiMAX, where the re-
access. For each year, the students are required to reuse all the ceivers are placed inside or on the façades of buildings. Further-
GSM BSs in the new area where UMTS must be deployed, more, this experiment allows students to perform coverage anal-
check the coverage and capacity of the UMTS network after this ysis for the portable type of WiMAX where the receivers may be
reuse, and then add the needed UMTS BSs to ensure the cov- located outdoors or indoors. Hence, this experiment introduces
erage and capacity requirements. Finally, students are required students to frequency assignment and interference calculation
to perform a cost–revenue analysis of the three-year deployment for WiMAX networks.
plan obtained. The purpose of this analysis is to quantify the In this experiment, students are required to plan a WiMAX
savings obtained by implementing a GSM/UMTS seamless net- network covering the city of Niort in western France, where
work compared to implementing a UMTS network from scratch. CPEs are considered to be placed respectively on building walls,
A cost–revenue analysis Excel tool was internally developed for indoors, and outdoors (including road level). Students are re-
the students to use as part of this experiment. This experiment quired to generate and analyze coverage plots for all these cases.
is particularly important as it deals with the advanced topic of Next, students are required to perform traffic analysis, where
GSM/UMTS co-siting and it introduces students to the cost di- each CPE is associated with the appropriate BS, and the best
mension when designing and deploying wireless cellular net- achievable modulation and coding scheme for each CPE is de-
works. termined. The purpose is to perform load balancing in order to
DAWY et al.: A WIRELESS COMMUNICATIONS LABORATORY ON CELLULAR NETWORK PLANNING 659

meet the constraint of maximum rate achievable by each BS.

Finally, students are required to use the automatic frequency as-
signment functionality of ICS telecom and to compare the inter-
ference levels before and after frequency assignment.

I. Experiment 9: GSM Drive Testing and Radio Network

Optimization
In this experiment, students are introduced to the area of drive
testing and network optimization, using a drive testing setup
from Rhode & Schwarz (TS9951—Portable Coverage Measure-
ment System with ROMES 3NG software) [28]. During the
pre-lab lecture corresponding to this experiment, students are
introduced to the drive testing hardware and software. In addi-
tion, they are accompanied by the instructor to perform real-time
outdoor and indoor GSM measurements and record the results.
After the lecture, students are required to replay similar mea-
surement files and use the relevant views in the ROMES 3NG
software for analysis.
Experiment 9 consists of replaying measurement files for
three real scenarios: GSM outdoor measurements, GSM indoor
measurements, and GPRS measurements. Students are required
to stop the replay at specific events in order to analyze them, Fig. 4. The GPRS RLC/MAC view obtained from the ROMES 3NG software
identify certain problems in the network operation, and propose based on real GPRS measurements using drive testing equipment.
appropriate solutions for optimizing the network performance.
For example, students should identify areas of excessive and
unnecessary handovers (ping-pong phenomenon), areas of high are required to monitor and analyze the different UMTS key per-
intercell interference, and areas of low coverage. They are formance indicators (KPIs) that they learned during the pre-lab
required to propose suitable solutions to overcome these prob- lecture corresponding to this experiment. Some of the important
lems and optimize the network performance. The identification KPIs that they have to monitor include block error rate measure-
is done by monitoring the appropriate software views, for ments, radio-related measurements such as the signal-to-inter-
example, the GSM Measurement Report view displaying the ference ratio, and active set size distribution [23]. They are also
received signal strength (RxLEV) from the serving BS and the required to investigate issues such as soft and softer handover in
best six neighboring BSs, and the GSM Handover view keeping addition to blocked and dropped calls. To analyze the measure-
track of the handovers throughout the drive testing duration. In ments, they monitor the various views related to the appropriate
the GPRS measurement file, students can check and analyze UMTS KPIs.
the data rates of the measurement mobile station in the uplink Students are required to propose their own solutions to some
and downlink directions, the asymmetry of the traffic, and the problems that took place during the drive testing measurements
bit error rate, in addition to several other parameters. As an in order to optimize network performance. For example, they
example, Fig. 4 shows the GPRS RLC/MAC view that contains need to analyze several measured KPIs jointly in order to deter-
information on the uplink and downlink time slot allocation, bit mine the reason for the high level of blocked and dropped calls
rates, and interference levels. in certain areas.

J. Experiment 10: UMTS Drive Testing and Radio Network V. LAB ASSESSMENT
Optimization
In accordance with ABET continuous improvement pro-
In this experiment, students are required to use the Rhode cesses, course assessment and evaluation are performed at the
& Schwarz ROMES 3NG tool to replay UMTS measurement end of each semester for all courses offered at the ECE Depart-
files obtained from real measurements from an area within the ment at AUB. Students are required to complete an anonymous
city of Munich, Germany. They are provided with drive testing online survey for each course they take in order to assess
measurement files for three mobile stations connected to both the level of achievement of the student learning outcomes.
GSM and UMTS networks belonging to two different operators The answers are based on a scale from 1 (strongly disagree)
covering the area of interest. The measurements were collected to 5 (strongly agree). According to the ECE Department’s
with the Rhode & Schwarz drive testing setup. guidelines, all scores above 4.0/5.0 are judged to be highly sat-
In this experiment, after learning to load and calibrate maps in isfactory. The average results for the wireless communications
Experiment 9, students learn to load BS database files and dis- lab course from spring term 2008–2009 are presented in Fig. 5.
play the BSs on the map. Students are required to determine the Results indicate that all outcomes were successfully achieved.
operator to which each mobile station is connected. Then, they In fact, the lab reports submitted throughout the semester
660 IEEE TRANSACTIONS ON EDUCATION, VOL. 53, NO. 4, NOVEMBER 2010

5) Were you satisfied with the lab strategy (e.g., open lab time
slots, pre-lab lecture, and post-lab presentation)?
Results demonstrate a high level of student satisfaction. The
majority of the students’ comments on the strengths and weak-
nesses of the lab course were of the form “lab well organized,”
“experiments were well written,” “lab practical and interesting,”
and “more challenging questions should be included at the end
of each experiment.” In fact, many of the graduating students
who took the lab course found job opportunities with local and
regional wireless operators or with major international telecom-
munication companies. These students indicated that the knowl-
edge they acquired from the lab played a crucial role in the offers
they received.

Fig. 5. Student learning outcomes assessment results for the wireless commu-
nications lab course from Spring term 2008–2009. The y-axis represents the
VI. CONCLUSION
average score per outcome, whereas the x-axis represents the number of the stu- A novel wireless communications lab course in the area of
dent learning outcome. The student learning outcomes are listed in Section III.
radio network planning and optimization (RNPO) was pre-
sented. The lab course consists of 10 experiments and covers
all phases of RNPO for several wireless communications
technologies that include GSM, UMTS, and WiMAX. The
lab course complements communications theory concepts
learned in regular lecture-based courses and provides students
with hands-on experience in several state-of-the-art network
planning and optimization tools. The authors would be pleased
to share the lab material with any one interested in it.

ACKNOWLEDGMENT
The authors are grateful for the support of Ericsson Com-
munications Lebanon, with special thanks to R. Sukkar and
D. Chackour.

REFERENCES
[1] GSM World, “Market data summary,” Dec. 7, 2009 [Online]. Available:
http://www.gsmworld.com/newsroom/market-data/
[2] H. Holma and A. Toskala, WCDMA for UMTS—HSPA Evolution and
LTE. Chichester, England: Wiley, 2007.
[3] “Fast facts,” UMTS Forum, Dec. 7, 2009 [Online]. Available: http://
www.umts-forum.org/
Fig. 6. Customized assessment results for the wireless communications lab [4] J. G. Andrews, A. Ghosh, and R. Muhamed, Fundamentals of
course for spring terms 2006–2007 and 2007–2008. The y-axis represents the WiMAX. Upper Saddle River, NJ: Pearson Education, 2007.
average score per question, whereas the x-axis represents the number of the [5] National Instruments, “UC Berkeley develops communications lab
question. with NI tools,” Dec. 7, 2009 [Online]. Available: http://zone.ni.com/de-
vzone/cda/pub/p/id/40
[6] Wireless Networking and Communications Group, University of Texas
at Austin, “EE 371C/EE 381V: Wireless communications lab,” Dec. 7,
demonstrate that students were able to assimilate all the target 2009 [Online]. Available: http://users.ece.utexas.edu/~rheath/courses/
wirelesslab/index.php
concepts. [7] The MathWorks, “Johns Hopkins University establishes virtual lab for
In addition to the surveys that are conducted by the ECE de- wireless communications with Simulink,” Dec. 7, 2009 [Online]. Avail-
partment, customized survey forms are filled out by students able: http://www.mathworks.com/products/matlab/userstories.html
[8] F. A. Cassara, “Wireless communications laboratory,” IEEE Trans.
and returned to the instructor in order to collect further feed- Educ., vol. 49, no. 1, pp. 132–140, Feb. 2006.
back about the lab course. The answers are based on a scale [9] Stanford University, “EE144/EE245—Wireless electromagnetic de-
from 1 (very unsatisfied) to 5 (very satisfied). The average re- sign laboratory,” Dec. 7, 2009 [Online]. Available: https://ccnet.stan-
ford.edu/ee144/
sults are shown in Fig. 6 for the Spring terms 2006–2007 and [10] Center for Wireless and Microwave Information Systems, University
2007–2008, with around 15 students attending the lab per term. of South Florida, “EEL 4936/EEL 6936 wireless communication sys-
The five questions that students were required to answer were: tems lab,” Dec. 7, 2009 [Online]. Available: http://wami.eng.usf.edu/
Classes/index.asp
1) How interesting was the lab course? [11] W. Padgett, B. Black, and B. Ferguson, “Low-frequency wireless
2) How beneficial were the lab experiments? communications system-infrared laboratory experiments,” IEEE
3) Were the topics covered in enough detail? Trans. Educ., vol. 49, no. 1, pp. 49–57, Feb. 2006.
[12] C. Furse, R. Woodward, and M. Jensen, “Laboratory project in wireless
4) How much did the lab course contribute to your practical FSK receiver design,” IEEE Trans. Educ., vol. 47, no. 1, pp. 18–25,
wireless communications knowledge? Feb. 2004.
DAWY et al.: A WIRELESS COMMUNICATIONS LABORATORY ON CELLULAR NETWORK PLANNING 661

[13] N. I. Sarkar and T. M. Craig, “Teaching wireless communication and Zaher Dawy (S’98–M’04–SM’09) received the B.E. degree in computer and
networking fundamentals using Wi-Fi projects,” IEEE Trans. Educ., communications engineering from the American University of Beirut, Beirut,
vol. 49, no. 1, pp. 98–104, Feb. 2006. Lebanon, in 1998 and the M.Sc. and Dr.-Ing. degrees in electrical engineering
[14] Virginia Tech, “New course gives students hands-on experience with from Munich University of Technology (TUM), Munich, Germany, in 2000 and
network interface, wireless networking,” Dec. 7, 2009 [Online]. Avail- 2004, respectively.
able: http://www.ecpe.vt.edu/news/ar03/newcourse.html Between 1999 and 2000, he worked as a part-time Communications Engi-
[15] Helsinki University of Technology, “S-72.3270: Cellular radio neer at Siemens AG Research Labs, Munich, Germany, focusing on the devel-
network planning methods,” Dec. 7, 2009 [Online]. Available: opment of enhancement techniques for UMTS. At TUM, between 2000 and
http://www.comlab.hut.fi/studies/3270/ 2003, he managed and developed a research project with Siemens AG where
[16] Helsinki University of Technology, “S-72.3290: Cellular radio network he designed advanced multiuser receiver structures for UMTS base stations.
optimization,” Dec. 7, 2009 [Online]. Available: http://www.comlab. Since September 2004, he has been an Assistant Professor with the Electrical
hut.fi/studies/3290/ and Computer Engineering Department, American University of Beirut. His
[17] Tampere University of Technology, “TLT-6406 radio network plan- research interests include cooperative communications, cellular technologies
ning,” Dec. 7, 2009 [Online]. Available: http://www.cs.tut.fi/tlt/RNG/ (WCDMA, HSPA, LTE), radio network planning and optimization, multiuser
TLT-6406.shtml information theory, multimedia over IP networks, and computational biology.
[18] KTH Royal Institute of Technology, “2E1512—Wireless network
course,” Dec. 7, 2009 [Online]. Available: http://www.cos.ict.kth.se/
education/msc/radio/
[19] A. Mishra, Advanced Cellular Network Planning and Optimisation. Ahmad Husseini received the B.E. degree in electrical and electronic engi-
Chichester, England: Wiley, 2007. neering from the Lebanese University, Beirut, Lebanon, in 2004 and the M.E.
[20] T. Halonen, J. Romero, and J. Melero, GSM, GPRS and EDGE Perfor- degree in computer and communication engineering from the American Uni-
mance. Chichester, England: Wiley, 2002. versity of Beirut (AUB), Beirut, Lebanon, in 2007.
[21] J. Laiho, A. Wacker, and T. Novosad, Radio Network Planning and In 2006, he worked on the development of the wireless communications lab
Optimisation for UMTS. Chichester, England: Wiley, 2006. course at AUB. Since June 2007, he has been with Ericsson—GSDC Lebanon,
[22] J. Lempiainen and M. Manninen, UMTS Radio Network Planning, Op- Beirut, Lebanon, as a Radio Network Consultant.
timization and QoS Management. Boston, MA: Kluwer, 2003.
[23] C. Chevallier, C. Brunner, A. Garavaglia, K. Murray, and K. Baker,
WCDMA (UMTS) Deployment Handbook: Planning and Optimization
Aspects. Chichester, England: Wiley, 2006.
[24] M. J. Nawrocki, M. Dohler, and A. H. Aghvami, Understanding UMTS Elias Yaacoub received the B.E. degree in electrical engineering from the
Radio Network Modeling, Planning, and Automated Optimization. Lebanese University, Beirut, Lebanon, in 2002 and the M.E. degree in com-
Chichester, England: Wiley, 2006. puter and communications engineering from the American University of Beirut
[25] Y. Zhang, Wimax Network Planning and Optimization. New York: (AUB), Beirut, Lebanon, in 2005, where he is currently a Ph.D. student.
Auerbach, 2009. He worked as a Research Assistant with AUB from 2004 to 2005 and with
[26] Ascom, “TEMS CellPlanner,” Dec. 7, 2009 [Online]. Available: the Munich University of Technology, Munich, Germany, in Spring 2005. From
http://www.ascom.com/en/index/products-solutions/technology-plat- 2005 to 2007, he was a Telecommunications Engineer with Dar Al-Handasah,
forms/platform/tems/solution/network-planning/product/tems- Shair and Partners, Beirut, Lebanon. His research interests include wireless
cellplanner-2/solutionloader.htm communications and antenna theory.
[27] ATDI, “ICS Telecom nG,” Dec. 7, 2009 [Online]. Available: http://
www.atdi.com/icstelecom.php
[28] Rhode & Schwarz, “Drive test tools,” Dec. 7, 2009 [Online]. Avail-
able: http://www2.rohde-schwarz.com/en/products/test_and_measure- Lina Al-Kanj received the B.E. degree in electrical and communications engi-
ment/Drive_Test_Tools/ neering from the Lebanese University, Beirut, Lebanon, in 2005 and the M.E.
degree in computer and communications engineering from the American Uni-
versity of Beirut (AUB), Beirut, Lebanon, in 2007, where she is enrolled as a
Ph.D. student. Her research interests include cooperative communications and
radio network planning and optimization.