Simulation models of mobile communication networks

Scientific Study, 2007

136 Pages, Grade: none




2. INSTALLING AND USING [illustration not visible in this excerpt]
2.1 MAIN FEATURES OF [illustration not visible in this excerpt] SIMULATION PACKAGE
2.4 CONFIGURE THE VIETNAMESE LANGUAGE SUPPORT FOR[illustration not visible in this excerpt]
2.4.1 Reconfigure the Operationg System to support Vietnamese Keyboard
2.4.2 Install Vietnamese typing tool
2.5 USING [illustration not visible in this excerpt] SIMULATION TOOLS

3.2.1 Cellular Principles
3.2.2 Cell planning for mobile cellular networks
3.4 CELL PLANNING FOR MOBILE CELLULAR NETWORKS USING [illustration not visible in this excerpt]
3.6.1 System model
3.6.2 Simulation of Neural Network-based equalization
3.6.3 Flow charts of the simulation algorithm
3.6.4 Using the simulation program

4.1.1 Introduction
4.1.2 Generate spread spectrum sequences using MobileSim
4.2.1 Direct sequence spread spectrum (DS-SS)
4.2.2 Simulation of BPSK DS-SS CDMA System
4.2.3 Simulation of QPSK DS-SS CDMA System
4.3.1 Frequency Hopping Spread Spectrum
4.3.2 Simulation of Fast Frequency Hopping with MobileSim
4.3.3 Simulation of Slow Frequency Hopping in MobileSim
4.4.1 Decorrelation Multiuser Detection (DEC)
4.4.2 MMSE (Minimum Mean Square Error)
4.4.3 Succesive Inteference Cancellation (SIC)
4.4.4 Parallel Inteference Cancellation (PIC)
4.5 SIMULATION OF MULTIUSER DETECTION TECHNIQUES WITH[illustration not visible in this excerpt]
4.5.1 DEC Detector
4.5.2 MMSE Detector
4.5.3 SIC Detector
4.5.4 PIC Detector
4.7 SIMULATION OF POWER CONTROL IN CDMA USING[illustration not visible in this excerpt]
4.8.1 Fundamentals of MS Positioning
4.8.2 Simulation Algorithm
4.8.3 Running Simulation with MobileSim
4.9.1 Introduction
4.9.2 Simulation with MobileSim
4.10.1 Introduction
4.10.2 Simulation program
4.10.3 Simulation Procedure

5.1.1 MC-CDMA Transmitter
5.1.2 MC-CDMA Receiver
5.2 SIMULATION OF MC-CDMA TRANSMITTER AND RECEIVER USING[illustration not visible in this excerpt]
5.4.1 Parameters and Assumptions
5.4.2 Simulation Procedure

6.1.1 Multi-code CDMA System
6.1.2 Multi-code MC-CDMA System
6.1.3 Rate Adaptive Control for Multi-code MC-CDMA system
6.2 SIMULATION OF MTC-MC-CDMA TRANSMITTER AND RECEIVER USING[illustration not visible in this excerpt]
6.2.1 Simulation Program
6.2.2 Simulation Procedure



About this book and the authors

This book is resulted from an electrical engineering research project at Ton Due Thang University - Ho Chi Minh City in Vietnam. In this book, the authors introduce both fundamental theory and practical issues about mobile communication networks, including GSM networks, CDMA-based networks, MC-CDMA and MTC-MC-CDMA networks. Each issue is introduced together with simulation model written in MATLAB and/or SIMULINK. Most part of this book (chapters 2-5) can be used as a useful reference book for undergraduate students who are studying mobile communication, especially about simulation aspect. Other remaining parts is dedicated for graduate students, that is, some sections in chapter 4,5 and the whole chapter 6.

Both the authors are young faculties of Electrical and Electronic Engineering Department of Ton Duc Thang University. They both got their BEng and MEng degrees in Electrical Engineering from Ho Chi Minh University of Technology in 2002 and 2005, respectively. The main author, TRAN THANH PHUONG is currently working toward his PhD degree at Purdue University in US. His research interest are wireless communications (focus on mobile networks), digital signal processing and embedded systems. NGUYEN TUONG DUY is now working as Deputy Head of the Office of Academic Affairs of Ton Duc Thang University.


This book required a rather long and hard working, and many people deserve our thanks. Our first heartfelt thanks is dedicated to Prof. Pham Hong Lien, currently the Dean of Electrical and Electronic Engineering (EEE) Department of Ton Duc Thang University, who has given us lots of advices on our work and provided us with her latest research results as well as the results from her graduate students. We also would like to express our special thanks to Dr. Tran Van Thanh, former Dean of EEE Department, for his administrative support to our project.

Many of our friends and colleagues made contributions to this book, and in particular we would like acknowledge the contributions of Le Dung on chapter 3 and 4; Hoang Manh Ha, Duong Thi Kim Dung, Thai Huynh Nghia, Doan Thi Thu Hien, Nguyen Ba Phuc, Tran Xuan Truong on chapter 4; Ho Van Khuong on chapter 5.

We also would like to appreciate our undergraduate students of Ton Duc Thang University who helped us in testing and debugging our source codes, in revising and formatting the manuscript - to name just a few: Ly Ngoc Phuc, Nguyen Phuong Lan Chi, Nguyen Quoc Khanh, Nguyen Toan Thanh, Tran Hong Ha and Phu Vu Trung Duong. The indefatigable help of our department administrative assistants, Nguyen Thi Thanh Vinh and Phan Van Nam, must also be acknowledged.

Finally this book is dedicated to our families who has always been with us during difficult time and suffered too much from our absences to pursuit our goals.



Since 1980, the mobile communication systems has been developing continuously and strongly from year to year. While the first generation systems (so called 1G) in the 80s were mostly based on analog technologies, in the next decade (90s), the second generation systems (2G) were launched, consisting of GSM (Global Systems for Mobile Communications), PDC (Personal Digital Cellular) and IS-95 (Interim Standard), basically based on digital technologies. However, the 2G systems were used mainly for voice services. Currently the third generation systems (3G) are still developed based on digital technologies, but they have the capability of integrating multiple services, including voice, data and multimedia as well as using both circuit switching and packet switching. The main technologies that were used in 3G systems (and the upcoming 4Gs) are CDMA and MC- CDMA technologies.

With the rapidly emerging of mobile networks today, soughting for knowledge and updating the latest research about mobile communication systems is very essential and pressing demand of students and even researchers and scientists in Electronic and Telecommunication Engineering. Nevertheless, doing research on practical systems is pretty difficult for developing country due to the large expense requirement. The general trend in the world (even in developed countries) nowadays is building virtual laboratories rely on computer simulation software and the networking between computers. For the purpose of helping students have a thorough grasp of mobile network knowledge via the visualized experiment results, this project will provide them with a variety of simulation models, from simple models to complicated ones, of the present mobile networks. These models not only effectively support the teaching at higher education in Electronic and Telecommunication Engineering, but also provide the updating results on theoretical researches, both nationwide and international, which are represented by simulation programs. Therefore, this report is a very useful reference for researchers, professors working on Electronic and Telecommunication Engineering in specific and for researchers and scientists in general.


- Developing the simulation models to illustrate the basic techniques of mobile communications that are currently implemented or will be implemented in the near furture. The target of this project includes making simulation lectures for undergraduate students who are studying and doing research on mobile communication systems. More specifically, our group will provide a useful software for teaching of telecommunication subjects . Also, this is a good electronic reference documentation for students, professors and researchers in telecommunication.

- Applicability: this project once carried out successfully can be apllied in the teaching of Telecommunication Labs for undergraduates of Telecommunication Engineering. As known by everyone, the management of mobile communication networks is a broad and controversy issue, it requires a group of telecom experts with huge amounts of funding. For that reason, in the scope of this project, the authors cannot mention all of the current problems of mobile networks, instead we only select some essential techniques and most common and important problems in mobile networks to investigate.

This project focus on 4 areas:

- GSM mobile network.
- CDMA-based mobile communication networks.
- MC-CDMA-based mobile communication networks.
- Multicode MC-CDMA mobile communication system.


- Carefully collect and review the results of theoretical studies published by both domestic and international scientists and institutions on mobile communication networks. Some specific resources are:

1) Web sites of well-known international universities.
2) Master theses and Doctoral dissertations in recent years at Ho Chi Minh University of Technology.
3) Recent research projects of students and faculties of Ton DucThang University.

- Investigate some available simulation platforms and select a most proper platform for simulating these systems. Some available platforms are listed as follows:

1) MATLAB and SIMULINK, product of The Mathworks, Inc.
2) Network Simulator (NS2)
4) LabView, National Instruments
5) IT++ Library, developed by Chalmers University of Technology, Sweden.
6) Other platforms based on various programming languages such as C/C++, Visual C, Visual Basic, Delphi,...

- Write simulation program and design the user interface.


Basically, this research includes 4 following tasks:

- Developing GSM simulation model: this section consists of the following problems: cell planning of a mobile networks, given the number of users and the covering area; simulation of signal transmitting and receiving in GSM mobile system, including noise processing tecniques.
- Developing the simulation models for the fundamental techniques of CDMA mobile communication systems. This section focuses on the followings: simulation of spread spectrum technique, the basic technique in CDMA-based systems; simulation of the multi­user detection methods which can reduce the multiple access interference (MAI); simulation of the power control methods which combat against near-far effect; simulation of the mobile user location technique in CDMA networks; simulation of congestion control algorithms for CDMA networks.
- Developing the simulation model of MC-CDMA-based mobile communication system. This is a technique that can combat effectively against fading effects and multiple access interference. Moreover, it allows us to improve the spectrum efficiency of system. Nói dung mô phong trong phan nay bao gôm: mô phong quá trinh thu phát tin hiêu va tách sóng da truy cap trong mang MC-CDMA; mô phong các phuong pháp diêu khiên công suat trong mang.
- Developing the simulation model for Multicode MC-CDMA system, a novel scheme which exploits the advantages of MC-CDMA systems, and can also support the high and variable rate services. This is a potential candidate for the 4G mobile communication networks. The main part of this section is to simulate the transmitting and receiving of information signals in Multicode MC-CDMA system, hence, to evaluate the performance of this new system.

illustration not visible in this excerpt

Fig. 1.1. Main user interface of MobileSim 2.0


The outcome of this project is the integrated simulation package ,[illustration not visible in this excerpt]which runs on the MATLAB© platform.

For the brevity of this report, in this section, the authors only want to introduce the basic theory for constructing simulation modules to solve these problems mentioned above. The detailed description about the project outcome as well as the complete user's guide will be provided in part 2 of this report.

Chapter 2 INSTALLING AND USING [illustration not visible in this excerpt]


- A simulation package running on MATLAB environment.
- Including the tools to simulate the theoretical (from simple to complicated) problems: GSM (2G), CDMA (3G), MC-CDMA (3G - 4G) and MTC-MC-CDMA (a proposed candidate for 4G).
- Users can design custom parameters and custom scenarios for simulation.
- Simulation results are introduced via images, waveforms, spectral plots and BER plots.


- Operating system: Microsoft Windows XP (recommended) or Windows 2000 Service Pack 3 hay 4, Windows NT 4.0 with Service Pack 5 or 6a
- CPU: Pentium III or above
- Hard disk: at least 500MB free
- RAM: at least 256MB, recommend: 512MB or above
- MATLAB version 7.0 or later
- Vietnamese language support (for Vietnamese version)
- Unicode (Unikey or VietFlex)


To run the simulation programs in MobileSim, you need MATLAB and SIMULINK installed in your computer. We recommend using version 7.0 or above. About how to install MATLAB and SIMULINK, Please refer to the following link: http ://

2.4 CONFIGURE THE VIETNAMESE LANGUAGE SUPPORT FOR [illustration not visible in this excerpt]

2.4.1 Reconfigure the Operationg System to support Vietnamese Keyboard

To display Vietnamese characters exactly in MobileSim, need to reconfig the Operating System (OS) to support Vietnamese keyboard.

At first, reboot the computer, insert Windows OS installation CD to the CD disk drive. Click Start>Control Panel. Double click on the icon Regional and Language Options.

illustration not visible in this excerpt

Fig. 2.1. Control Panel Window

Select Language tab. Check the square beside the text "Install files for complex script and right-to-left languages (including Thai)". If that square has been checked, it means your OS has Vietnamese language support already.

illustration not visible in this excerpt

Fig. 2.2. Language Options

Click Advanced tab. On popup menu below the text "Select a language to match the language version of the non-Unicode programs you want to use:", select Vietnamese.

illustration not visible in this excerpt

Fig. 2.3. Language Options (Advanced)

Click OK.

2.4.2 Install Vietnamese typing tool

To display and type Vietnamese characters, you need a Vietnamese typing tool installed in your PC. Two tools that are compatible with the Vietnamese keyboard functions provided by Microsoft are Unikey and VietFlex.

For VietFlex, install the tool and select Unicode character set, input method can be any of the list.

For UniKey, install the tool and select va chon bô ma Vietnamese Local CP 1258.

illustration not visible in this excerpt

Fig. 2.4. UniKey 4.0Screenshot

2.5 USING [illustration not visible in this excerpt]SIMULATION TOOLS

After configuring the language support, you can start using MobileSim. Copy all MobileSim folder to your computer hard disk. Open MATLAB program. Go to menu File>Set Path.

illustration not visible in this excerpt

Fig. 2.5. Set path to MobileSim in MATLAB

Click on the button Add with Subfolders... Select the path to MobileSim folder in your PC. Click OK. Click on Save button. Click Close button to close the Set Path window.

illustration not visible in this excerpt

Fig. 2.6. Set path for MobileSim in MATLAB (cont)

On the MATLAB Command Window, type the command MobileSim. Press Enter, the main user interface will appear.

On the left hand side, there are the information about this program:

- Button "Updated Features": to provide the information about latest updates features of current version.
- Button "Help": to open the user's guide (this file).
- Button "Comments": to allow users to input their comments / feedbacks about the software.
- Button "FAQ": to provide solutions for some common errors of the program.
- Button "Contact us": to provide mailing and email addresses of the authors.
- Popupmenu "Web site Links" provides the hyperlink to reference web site which has more information on the documentation and standards about mobile communication networks such as GSM, CDMA, ...

The program consists of simulation tools for 4 mobile networks, including CDMA, MC- CDMA, MC-MC-CDMA. Click on the corresponding button to choose the simulation tools you need. The following figures demonstrate the user interfaces of each components.

illustration not visible in this excerpt

Fig. 2.7. Main user interface

GSM Networks:

illustration not visible in this excerpt

Fig. 2.8. GSM simulation tools - User Interface

CDMA Networks:

illustration not visible in this excerpt

Fig. 2.9. CDMA simulation tools - User Interface

MC-CDMA Networks:

illustration not visible in this excerpt

Fig. 2.10. MC-CDMA simulation tools - User Interface

illustration not visible in this excerpt

Fig. 2.11. Parameter Input for MC-MC-CDMA simulation tool



The history of GSM can be traced back to a proposal of Nordic Telecom at the CEPT (Conference of European Post and Telecommunication) in 1982, which purpose is developing a new digital cellular standard to satisfy the increasing demand of European mobile communication network.

The European Committee (EC) issued a recommendation which suggested the member nations who use GSM allowing communication in the frequency band 900MHz. The European Telecommunications Standards Institue (ETSI) then defined GSM after the digital cellular phone standard has been accepted worldwide.

This proposal was accepted on September 1987, when 13 managers and executive officers of CEPT GSM agree to sign the so-called GSM MoU "Club", which is valid from July 1st, 1991.

GSM stands for Global System for Mobile Communications, formerly is Groupe Spécial Mobile.

The Global System of Mobile Communication (GSM) is a integrated and complete digital cellular communication system, which is first developed in Europe, and rapidly becomes popular throughout the world. This network is designed in corresponding with ISDN systems, and the services provided by GSM are the sub-systems of standard ISDN services.

illustration not visible in this excerpt

Fig. 3.1. Wireless Cellular Network Architecture

The first GSM system was designed to operate on the frequency bands 890-915 MHz and 935-960 MHz. Several main improvements was proposed by this system:

- Really good quality of main voice service.
- Reduced subscriber fee.
- Support international mobile communications.
- Capability of terminal handheld devices support.
- Support some convenient tools and new services.
- Spectrum efficiency.
- ISDN compability.

The standard was officially issued on January 1990 and the first commercial systems were established in the middle of 1992. The MoU (Memorandum of Understanding) signed firstly by GSM manager and executive directors from 13 countries, and now there has been more than 212 members all over the world.


3.2.1 Cellular Principles

Cell is the basic unit of the mobile network. On the geographical map for network planning, each cell has a hexagonal shape. Located in each cell is a BTS (Base Transceiver Station). The function of BTS is to provide wireless communications to all mobile station (MS) in that cell. The cell principle is illustrated in fig. 3.2:

illustration not visible in this excerpt

Fig. 3.2. Cells in a cellular network

The main properties of cellular model is that the frequency reuse and the area of each cell in the network. Fig. 3.3 illustrates the concept of frequency reuse. Each cell is allocated with a group of wireless channel frequencies. The letters A, B, C, D, E, F, G are the names of cells and also represent the specific frequency group which is used in that corresponding cell. For instance, frequency group A is used for all of cells Ai, that means group A is reused in multiple cells Ai, which has sufficiently large inter-distance, sufficiently small transmission power, so that the co-channel interference is negligible. The important property of frequency reuse is that the interference is independent of the absolute distance Dbetween cells Ai, but depends on the ratio D/R (where R is cell radius) instead. Therefore, the desired R will be determined by the designing of antenna height and the transmission power ([1], [2]).

illustration not visible in this excerpt

Fig. 3.3. Frequency reuse at distance D

illustration not visible in this excerpt

Fig. 3.4. Enhance capacity by dividing into smaller cells

In practice, due to the incessant growing of capacity in one cell toward the threshold at which the service quality becomes unacceptable, we usually have to divide cells into smaller cells. With these cells, we use the smaller transmission power and so the frequency reuse is carried out at a smaller scale, fig. 3.4 illustrates this concept.

3.2.2 Cell planning for mobile cellular networks

On the basis of spectral efficiency, designing the wireless communication system in mobile cellular networks can be divided into following sub-problems, which are related to each other:

- Frequency reuse channel
- Reduce co-channel interference
- Guarantee of ratio [illustration not visible in this excerpt] (signal/interference )
- Handover mechanism
- Divide cells into smaller cells.

Due to the limitation of frequency resource, the most challenging problem in designing is how to serve the maximum number of users, while still satisfy the QoS constraints. We have to answer three questions:

- In rush hours, how much users can the system support?
- How much subscribers does the system accept?
- How many frequency channels does the system need?

Generally speaking, the signals which have the same frequency will not interfere each other if they are transmitted in different time slots, or they are from different location in space, or they are assigned different (and orthogonal to each other) codes. These are some examples of diversity, a very important concept in wireless communications.

Fig. 3.5 introduces some examples of frequency reuse scheme: К = 4; К = 7; К = 12; К = 19 ...

illustration not visible in this excerpt

Fig. 3.5. Frequency reuse distance

The minimum distance for frequency reuse depends on many factors:

- Number of reused cell around the considering cell
- Geographical characteristics of covering area.
- Antenna height.
- Transmission power

illustration not visible in this excerpt

Theoretically, D increases when К increases, hence, the co-channel interference decreases. However, in practice, the total number of available frequency channels is fixed. If К increases, then the number of available channels for each cells decreases, and so the spectral efficiency also decreases. The question for us is: For which value of К do both following conditions are satisfied: К is small enough to get the acceptable spectral efficiency, but large enough to prevent co-channel interference?

To determine cell size and the number of channels allocated to each cell, we need to know the load condition during rush hours in that geographical area.

After determine the maximum number of channels that can be allocated to each cell, we can compute the maximum number of calls that can be served during rush hours in that cell (denote this number as Qi ). From the assumption (derive from the statistical data) about the percentage of users need to communicate during rush hours, we can compute the number of subscribers M that the network can serves.


During rush hours, the number of calls Qi in 10 cells is 2000, 1500, 3000, 500, 1000, 1200, 1800, 2500, 2800, 900, respectively.

Suppose that the percentage of demanding users is 60% (qc = 0,6), and each user has one call during rush hours. Then the total number of users in 10 cells is:

illustration not visible in this excerpt

The estimated total number of subscribers 10 cells is:

illustration not visible in this excerpt


Typical GSM networks consist of three main components. Mobile Station (MS) hold by subscibers. Base Station Subsystem (BSS) controls and communicates with MS via BTS. Network Subsystem (NSS) whose main components is Mobile services Switching Center (MSC), carries out the switching between mobile users, and between mobile users with the public telephone network. MSC processes the mobile managing operations. MS and BSS communicate together based on Um interface, also called air interface or wireless link. BSS is connected to MSC by using A interface ([2][5]).

Mobile Station

Mobile Station (MS) includes mobile devices (terminal devices) and a smart card, so called Subscriber Identity Module (SIM). SIM provide the personal information of mobile users, so that users can access the subscribed services without dependence on the terminal devices. By attaching SIM to GSM terminals, users can received calls, make calls or send/receive messages from and to other subscribers on those terminal devices.

Mobile devices are identified uniquely by the so-called International Mobile Equipment Identity (IMEI). SIM card stores the International Mobile Subscriber Identity (IMSI), which is used to identify mobile user in system, to authenticate user right and other information. IMEI and IMSI are independent. SIM card can be secured from unauthorized actions by using password or PIN (Personal Identification Number)

illustration not visible in this excerpt

Fig. 3.6. Architecture of a mobile cellular network

Base Station Subsystem

The BSS contains two components: BTS and BCS. These two components communicate together by the standardized Abis interface, which allows them operate various equipments provided by different manufacturers.

Base Station is basically a wireless transceiver, which covers one cell of the network, and controls the wireless communication protocols between BS and MS. In a range of a big city, most likely we must use many BTS' S, so it's required that BTS must be accurate, reliable, movable and low-cost.

BCS controls the wireless resource of one or more BTS'. BCS controls the channel establishing, frequency hopping and handover. BSC is the intermediate unit to connect between MS and MSC.

Network Subsystem

The main component of Network Subsystem (NSS) is MSC, operates like a switching node in PSTN or ISDN, provides all necessary functions to control a mobile user di noäng, including registration, authentication, location updating, handover, and routing. These services, which are provided together with other functional components, form the NSS. MSC provides the connection with other fixed networks (such as PSTN or ISDN). The signaling methods between functional components in NSS is Signaling System No.7 (SS7), which is used for intermediate center signaling in ISDN networks and extended to used in current public telephone network.

Home Location Register (HLR) and Visitor Location Register (VLR), together with MSC, perform call routing and mobile capability for GSM. HLR stores all of management information of each registered user in the corresponding network, and the current location of that user. Position of mobile users is often represented in terms of the signaling address of the VLR which controls that MS.

VLR contains the management information selecting from HLR, which is necessary for controlling the call and providing subscribed serveices to each user, which is currently in the geographical area managed by that VLR. Although each functional component can operate independently but until now, all switching device manufacturers produce VLR in company with MSC, so the geographical area managed by MSC is corresponding to the VLR of that area, that reduces the redundant signaling. Note that MSC does not contain information about MS', these information are stored in Location Registers.

There are two other registers using for authentication and security pupose. Equipment Identification Register (EIR) is a database containing all valid mobile devices in the network, each of them is authenticated by its IMEI. IMEI will be marked as invalid if the system receives the stolen/lost notice or if that mobile device is not accepted in the network. Authentication Center (AuC) is a protected database containing a copy of deciphering key in user SIM card, this key is used to authenticate and to encrypt messages when sending them on wireless channel.

The wireless spectrum is a limited resource shared among all users, so it's required that we must find an effective method to allocate the bandwidth to as large number of users as possible. GSM selects the combining method of Time Division Multiple Access and Frequency Division Multiple Access (TDMA/FDMA). Using FDMA, the bandwidth (less than or equal 25 MHz) will be divided into 124 carrier frequency with 200 KHz apart from one to its successor. Using TDMA, the 200-KHz wireless channel is divided into 8 time slots (results in 8 logic channel). Therefore, each logic channel is defined by its frequency and the index number of its TDMA time slot. By applying 8 time slot mechanism, each channel transmits its digital data on "burst" basis: the GSM terminal only transmits 1 of 8 time slots.

FDMA assigns the channels to each specific user. These channel are assigned based on the user demand. TDMA partitions the wireless spectrum into time slots and each time slot is only available for one of two actions: transmitting or receiving.

Burst is the transmission unit of GSM. Transmission occurs in a time window of (576+12/13) μs, i.e the bit period (156 + %). Each burst usually consists of two 58-bit package (57 data bit and 1 stealing bit) and a training sequence of 26 bits. The training sequence is a predefined transmitted sequence used to compare with the receiced sequence to better recovering the original signal (multipath equalization).

Normal burst (NB) (Flag Is relevant for TCH only) Fig. 3.7. GSM data frame format

3.4 CELL PLANNING FOR MOBILE CELLULAR NETWORKS USING[illustration not visible in this excerpt]

To support the cell planning issue, MobileSim program provide computing tool to evaluate the necessary parameters for cell planning given the covered area and the estimated number of users in that area. The program can support the number of users up to 10000 users per 1km2.

Fisrtly, start MobileSim 2.0. Click on the GSM simulation button. A GSM user interface appears (fig. 3.8). Continue by clicking on the button "Cell Planning". Select urban area or rural area by clocking on the corresponding button.

Input the values of parameters you have, including considered area, number of users per squared kilometer, and number of available frequency channel. Note that the number of users must be between 100 and 9999.

illustration not visible in this excerpt

Fig. 3.8. GSM user interface

Click on the button "Investigation Results". The evaluated parameters will appear below this button, including (see fig. 3.9):

illustration not visible in this excerpt

Fig. 3.9. Cell planning parameters

- Radius of each cell (km)
- Number of cells required
- Frequency reuse distance D
- Channel loss
- SIR (Signal-to-Interference Ratio, best value and worst value)

illustration not visible in this excerpt

Fig. 3.10. Computing results

Simultaneously, on the screen come three new windows:

- Illustrating image of cell planning for Ho Chi Minh City area.
- A plot that represents the relation between C/I ratio with number of frequency channels.
- A plot that represents the relation between the frequency reuse distance and number of frequency channels.

To go to other simulations, click on the button "Back". The new figures are now automatically closed.


Equalization is a technique to remove the Intersymbol Interference (ISI) by using digital signal processing algorithms. In noisy wirelss transmission, adaptive equalizers are usually used to suppress the ISI ([3])

illustration not visible in this excerpt

Fig. 3.11. Block Diagram of a communication system with adaptive egualizer at receiver

General operation modes of adaptive equalizer are training and tracking. At first, a training sequence with fixed length is transmitted so that the equalizer at receiver can estimate the channel and set up their parameter accordingly. Training sequence is usually a pseudo random binary signal or a predefined, deterministic sequence. Immediately follows this sequence, user data (may or may not be encoded) are transmitted, and the adaptive equalizer at receiver will use the recursive algorithm to determine the channel gain, and estimate the filter coefficients corresponding to the channel gain. Training sequences are designed to allow the equalizer to achieve the appropriate filter coefficients even in the worst case of channel condition. After finishing training period, the filter coefficients will reach their optimal value to recover the user signal correctly. When the user data are received, the adaptive algorithm of equalizer will trace to the variation of channel condition. As a result, the adaptive equalizer varies their filtering characteristics continuously.

From time to time, the equalizer status converges to their steady state, depending on equalization algorithm function, structure of equalizer and the variation rate of multipath wireless channel. Equalizer needs to be trained periodically to maintain the efficiency of ISI cancellation. It's often used in digital communication system, where user data are time divided. TDMA wireless communication system is especially appropriate for applying a equalizer. TDMA systems send data frame by frame, each frame is corresponding to a fixed time interval , and the training sequence is send at the beginning of frame. At each time, when a new frame is received, the equalizer is re-trained by the same training sequence.

Equalizer is implemented at the baseband stage or IF stage of receiver. In this project, the demodulated signals and the adaptive equalizer are simulated at the baseband. The adaptive equalizer is implemented by using Neural Networks.

The block diagram of a digital communication system with Neural-Network-based equalizer to remove ISI is shown in fig. 3.12:

illustration not visible in this excerpt

Fig. 3.12. A digital communication system with Neural Network based egualizer


3.6.1 System model

A common communication system consists of transmitter, channel (transmission media) and receiver. At transmitter, signals go through BPSK (or QPSK) modulation then through the bandpass filter, and finally transmitted to the channel. The signals reach receiver after transmitted over the channel, where they suffers from several kinds of noise: AWGN noise, Rayleigh fading, co-channel interference and ISI. To recover the original data, the received signals are applied to the demodulator, lowpass filter, and the decision unit, respectively, at receiver ([3], [4]).

The system model used in this simulation is mainly based on the block diagram plotted in fig. 3.12.

3.6.2 Simulation of Neural Network-based equalization

Due to the various kinds of noise and interference on channel, the received data may have some errors. There are several techniques to cancel noise and interference which has been introduced rigourously in theory, but in this project we only focus on the equalization technique. In practive, a variety of equalizer might be applied to process communication signals, and in this project we simulate the equalization based on Neural Networks method. Other interesting equalizers will be updated to our package later on.

The following is a brief description about some kinds of Neural Networks ([7][10][11]):

- Perceptron network: transfer functions of neurons are step functions, which is similar to the biological neurons, but this is rarely used in practice in artificial intelligent netwotks because after going through each neurons, the signal properties become more unaccurate.

- Linear network: similar to perceptron network, except that the transfer function is linear, which gives unbounded output values, exclusively applied for linear problems. This network uses the LMS training rule, which is more powerful than Peceptron training rule. Linear network can adapt to the variation of the environment. It can be adjusted step by step according to the input vector and the desired vector, with purpose of finding the appropriate weights and threshold values such that the sum of squares of errors is minimized. This kind of network is often used in adaptive filters, digital signal processing and controlling systems. This is the simplest kind of neural networks that can be applied in practice.

- Backpropagation network: Backpropagation network is implemented based on the Widrow-Hoff traning rule generalized for multilayer networks with various nonlinear transfer function. This network consists of a thresholding device, a sigmoid layer (layer with sigmoid transfer function) and an output linear layer, so it can simulate any function with finite discrete number of samples. Once it is trained correctly, this network can adapt to a variety of input signals, even if these signals have never been trained. Usually, the input signal will have a corresponding output that have the similar form of the output of trained signal. Due to this generalization property, we can train the network by using representative input/output pairs and it still have good adaptation with other untrained signals.

- Radial Basis Network: Radial basis networks require more conditions than the standard feedforward backpropagation networks, however, as a compensation, the design process takes less time than the standard feedforward networks. These networks will have good performance as long as there are enough trained vectors. This property itself restrict the ability of applying Radial Basis networks in equalizers. Furthermore, the number of neurons in radial basis network is proportional to the size of input space and the complexity of problem, hence, radial basis networks often have larger size than backpropagation networks. Another drawback is that radial basis networks are very slow because of the huge number of computations, which requires a large memory space. From these reasons, the authors decide not to simulate radial basis networks in this project. The only reasonable application of radial basis networks is classification problem.

- Feedback networks: Feedback networks contain some links from the neurons at current stage to the neurons of previous stage. These networks can be unstable and oscillate complicatedly. Feedback networks are currently interested issue of many researchers but they are not efficient in practical problems.

- Seft-Organnizing Networks: These networks have ability of learning, finding the correlation between input signals to output corresponding responses. The neurons of these networks can recognize the groups of similar inputs, and self-organize to determine the appearance frequency of input vectors. Therefore, seft-organizing networks are used to classify the input vectors, and are suitable for pattern recognization and classification.

Here in this project, we use linear netwok and backpropagation network to design an equalizer that can learn adaptively the property of data sequences, given a predefined training sequence. The network can vary its weights continuously to detect correct data sequence. The simulaton program is designed with various training methods; each network, each training method can have customized configure. According to simulation results, the backpropagation network is the best choice for digital signal processing due to its ability of generalization.

3.6.3 Flow charts of the simulation algorithm

The simulation program is run step by step, following the flow chart in fig. 3.13.

illustration not visible in this excerpt

Fig. 3.13. Simulation procedure

Flow chart of main program algorithm is shown in fig. 3.14.

illustration not visible in this excerpt

Fig. 3.14. Main program flow chart

3.6.4 Using the simulation program

On the user interface of GSM simulation program, click on the button "Transmitting, receiving and noise processing", the following windows will appear on the screen, allow users to input the simulation parameters:

illustration not visible in this excerpt

Fig. 3.15. User interface for transceiving simulation

This window prompts for the following parameters:

- Carrier Frequency
- Bit Rate
- Length of Data
- Type of data (Random or deterministic)
- Carrier Amplitude
- Modulation method (BPSK/QPSK)
- Type of noise (White Noise/Fading/Cochannel)
- Noise parameters (SNR/Mean and Variance, number of noise sources)

Click button Back to return to main window or click button "CONTINUE" to go to another window. This window allows user to input ([6]) :

- Types of Neural Network:
- Bayesian regularization
- Levenberg-Marquardt
- One step secant
- Quasi Newton
- Scale conjugate gradient
- Powell-Beal
- Polak-Ribieure
- Fletcher-Reeves
- Resilient backpropagation
- Adaptive learning rate
- Gradient descent momentum
- Gradient descent
- Adaptive linear

- Learning function: Gradient descent hay Gradient descent momentum
- Number of inputs
- Number of layers
- Error performance functions:


- Number of neurons in each layer

- Transfer function of each layer

illustration not visible in this excerpt

Fig. 3.16. Neural network parameters

To change the configure of network, first choose the number of layers and number of delay taps, then on the space below appear some popupmenus, allowing users to adjust the parameters for each network layer. This program supports up to 10 layers.

After input all required parameters, click button "RUN". At first, the training plot will be displayed.


Excerpt out of 136 pages


Simulation models of mobile communication networks
Ton Duc Thang University
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
10680 KB
This book is written as a documentation for a telecommunication research project at Ton Duc Thang University. The project was highly appreciated by the evaluating committee and this documentation is now being used as a supplementary lecture notes for the undergraduate course "Communication Systems" at Ton Duc Thang University.
Quote paper
MEng Phuong Tran (Author), 2007, Simulation models of mobile communication networks, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: Simulation models of mobile communication networks

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free