Final Draft 0.docx

Mekelle University Ethiopian Institute of Technology Mekelle (EiT-M) School of Electrical and Computer Engineering Dimensioning and Planning of Multi RAT Radio Network

Submitted By Group Member

1. Ataklti Gebremichael 2. Freweyni Hagezom 3. Hailay Gebremedhin

Id. No 161608/06 262040/06 162213/06

A thesis submitted in partial fulfillment of the requirements for BSc in Electrical and Computer Engineering (stream of electronics and communication engineering)

Advisor Name Mr. Merkebu Tekaw (MSc)

May, 2018

Abstract The evolution of mobile services are taking place with considerable faster rate starting from second generation GSM services, third generation UMTS services to fourth generation LTE technology. The concept of multi RAT could incorporate these radio access technologies in one to support multiple functional purpose. But with this pace the operators always have a concern of planning the network. Network planning is a never ending task, planning network with limited number of user is not the issue but the issue is to plan a network that also allows future growth and expansion. The planning ensures the customers to use the network services wherever they are. This is an ongoing process. This paper deal on the process the multiple radio access network specifically 2G, 3G and 4G radio network planning. The study considers Adi haki and kedamay weyane sub cities in Mekelle, Ethiopia, in view of new deployment scenario. After analysis of the data taken from the sub city and using the nominal parameters in the detailed planning, the study resulted in target area coverage prediction and capacity evaluation in terms of a given subscriber future growth. Then after evaluating the performance in terms of prediction by signal level, overlapping zone, throughput and interference results shows that a better network coverage with an optimum network capacity can be achieved.

1

Acknowledgment We would like to thank the almighty God for letting us finish this thesis and endless blessing throughout our life. Moreover, this could not be successfully conducted without an essential and valuable assistant many other peoples. Therefore, we are also grateful to express our sincere thanks and gratitude to our advisor, Merkebu Tekaw (MSc) for his indispensable suggestions, tremendous support and encouragement. And we are also grateful to the examiners (lecturers of school of electrical and computer engineering) for invaluable critics towards the betterment of the thesis during progress presentation. The last but not the least we are also grateful for everyone who engaged their hands for successfully finished of this thesis.

2

Table of Contents Abstract ................................................................................................................................................. 1 Acknowledgment .................................................................................................................................. 2 Table of Contents .................................................................................................................................. 3 List of Figures ....................................................................................................................................... 5 List of Tables......................................................................................................................................... 6 List of Acronyms ................................................................................................................................... 7 CHAPTER ONE ................................................................................................................................... 9 INTRODUCTION ............................................................................................................................. 9 1.1

Background ........................................................................................................................... 9

1.2

Literature Survey ................................................................................................................. 10

1.3

Statement of Problem .......................................................................................................... 11

1.4

Objective ............................................................................................................................. 11

1.4.1

General Objective ...................................................................................................... 11

1.4.2

Specific Objectives ..................................................................................................... 12

1.5

Methodology ....................................................................................................................... 12

1.6

Significance of the Paper for the Society ............................................................................ 13

1.7

Organization of Thesis ........................................................................................................ 14

1.8

Scope of the Thesis.............................................................................................................. 14

CHAPTER TWO................................................................................................................................. 15 RADIO ACCESS TECHNOLOGIES OVERVIEW....................................................................... 15 2.1

Introduction ......................................................................................................................... 15

2.2

Global System for Mobile (GSM) ....................................................................................... 15

2.2.1

GSM Architecture ..................................................................................................... 15

2.2.2

Multiple Accessing in GSM ...................................................................................... 15

2.2.3

GSM Channels ........................................................................................................... 16

2.3

Universal Mobile Telecommunications System (UMTS) ................................................... 16

2.3.1

UMTS Network Architecture ................................................................................... 16

2.3.2

UMTS Operation Modes and Multiple Access ....................................................... 17

2.4

Long Term Evolution (LTE) ............................................................................................... 18

2.4.1

LTE Network Architecture ...................................................................................... 18

2.4.2

LTE Physical Layer................................................................................................... 19

2.4.3

LTE FDD Frame Structure ...................................................................................... 21

2.4.4

LTE MIMO Basics .................................................................................................... 22

CHAPTER THREE ............................................................................................................................. 24 3

MULTI RAT RADIO NETWORK PLANNING ........................................................................... 24 3.1

Introduction ......................................................................................................................... 24

3.2

Multi RAT Network Planning ................................................ Error! Bookmark not defined.

3.3

Site Survey .......................................................................................................................... 25

3.4

GSM Radio Network Planning............................................................................................ 26

3.4.1

GSM Coverage Planning .......................................................................................... 26

3.4.2

GSM Capacity Planning ........................................................................................... 34

WCDMA Radio Network Planning .................................................................................... 38

3.5

3.5.1

WCDMA Coverage Planning ................................................................................... 38

3.5.2

WCDAM Capacity Planning .................................................................................... 44

LTE Network Planning ....................................................................................................... 46

3.6 3.6.1

LTE Coverage Planning ........................................................................................... 46

3.6.2

LTE Capacity Planning ............................................................................................ 52

CHAPTER FOUR ............................................................................................................................... 56 Simulation Results and Discussion ................................................................................................. 56 4.1

Designing Multi RAT Network .............................................. Error! Bookmark not defined.

4.1

Performance Evaluation of Planned GSM Network ........................................................... 58

4.2

Performance Evaluation of Planned UMTS Network ......................................................... 61

4.3

Performance Evaluation of Planned LTE Network ............................................................. 64

4.4

Comparison of Multi RAT by Effective Service Analysis .................................................. 67

CHAPTER FIVE ................................................................................................................................. 70 Conclusions and Recommendations for Future Work......................................................................... 70 5.1

Conclusion............................................................................................................................... 70

5.2

Recommendation for Future Work.......................................................................................... 71

Reference............................................................................................................................................. 72

4

List of Figures Figure 2.1: GSM Network Architecture .............................................................................................. 15 Figure 2.2: UMTS Network Architecture ........................................................................................... 17 Figure 2.3: LTE Evolved Packet System (EPS) architecture .............................................................. 19 Figure 2.4: Frequency-time representation of an OFDM Signal ......................................................... 20 Figure 2.5: LTE FDD Frame and Slot Structure ................................................................................. 22 Figure 2.6: Physical Resource Block and Resource Element.............................................................. 23 Figure 2.7: MIMO Transmission......................................................................................................... 23 Figure 3.1: Network planning process steps ....................................................................................... 24 Figure 3.2: Link Budget Parameters.................................................................................................... 27 Figure 3.3: Lower tail of normal distribution curve ............................................................................ 30 Figure 4.1 Digital map of Mekelle City .............................................................................................. 56 Figure 4.2 Computational zone of Kedamay Weyane and Adi Haqi in ATOLL ................................ 57 Figure 4.3 GSM Coverage Prediction by Signal level ........................................................................ 58 Figure 4.4 GSM Coverage area with Signal level histogram ................. Error! Bookmark not defined. Figure 4.5 a) GSM Overlapping zone of transmitter b) Histogram..................................................... 59 Figure 4.6 a) GSM Coverage by C/ (I+N) Level and b) Histogram.................................................... 60 Figure 4.7 GSM Effective Service Area Analysis............................................................................... 61 Figure 4.8 a) UMTS Coverage prediction by Signal level b) Histogram .......................................... 62 Figure 4.9 a) UMTS Overlapping zone of transmitter b) Histogram .................................................. 63 Figure 4.10 a. UMTS Total Noise Level Analysis b. Histogram ........................................................ 63 Figure 4.11 UMTS: Effective Service Area Analysis ........................................................................ 64 Figure 4.12 Coverage prediction by signal level in DL using histogram ............................................ 65 Figure 4.13 Coverage prediction by overlapping zone ....................................................................... 65 Figure 4.14 Coverage prediction using throughput ............................................................................. 67 Figure 4.15 LTE effective service analysis ......................................................................................... 67 Figure 4.16 Comparison of Effective Service Area Analysis by UMTS vs LTE ............................... 68 Figure 4.17 Comparison of Effective Service Area Analysis by GSM vs UMTS .............................. 68 Figure 4.18 Comparison of Effective Service Area Analysis by LTE vs GSM .................................. 69

5

List of Tables Table 3.1: Regions of Planning ........................................................................................................... 26 Table 3.2: Cable loss per 100m ........................................................................................................... 29 Table 3.3: GSM link budget calculation ............................................................................................. 33 Table 3.4: Subscribers for GSM .......................................................................................................... 35 Table 3.5: Assumption taken for GSM capacity dimensioning........................................................... 35 Table 3.6: GSM Channel Distribution Strategy .................................................................................. 37 Table 3.7: GSM Best site selection of coverage and capacity ............................................................ 37 Table 3.8: Required Eb/No Values...................................................................................................... 39 Table 3.9: Receiver parameters ........................................................................................................... 41 Table 3.10: Calculated 3G Link Budget .............................................................................................. 43 Table 3.11: Throughput per user at busy hour calculation .................................................................. 45 Table 3.12: UMTS Cell Load Dimension Result ................................................................................ 45 Table 3.13: UMTS Best site selection of coverage and capacity ........................................................ 45 Table 3.14: Penetration losses and Standard deviation of slow fading typical dense urban value. ..... 48 Table 3.15: Uplink link budget parameters ......................................................................................... 50 Table 3.16: Downlink Link budget parameters ................................................................................... 50 Table 3.17: Clutter parameters ............................................................................................................ 51 Table 3.18: List of PRBs . ................................................................................................................ 52 Table 3.19: LTE Users Category ......................................................................................................... 53 Table 3.20: Total average throughput per subscriber at busy hour ..................................................... 53 Table 3.21: The total amount of sites .................................................................................................. 55

6

List of Acronyms 2G 3G 3GPP 4G AMC BPL BS BSS BW C/I CCU CINR CN CP CPICH CS DL EIRP eNodeB EPC EPS ETSI E-UTRAN FDD FDMA FFT FR FSPL GB GSM HR HSPA HSS HSUPA LB LCD LTE MAPL MCS MIMO MME MS

Second Generation Third Generation 3rd Generation Partnership Project Fourth Generation Adaptive Modulation and Coding Building Penetration Loss Base-Stations Base Station Subsystem Bandwidth Channel to Interference Ratio Central Control Unit Carrier to Interference plus Noise Ratio Core Network Cyclic Prefix Common Pilot Channel Circuit Switched Down Link Effective Isotropic Radiated Power Enhanced node B Evolved Packet Core Evolved Packet System European Telecommunications Standards Institute Evolved Universal Terrestrial Radio Access Frequency Division Duplex Frequency Division Multiple Access Fast Fourier Transform Full Rate Free Space Propagation Loss Giga Byte Global System for Mobile Half Rate HSPA High Speed Packet Access (HSDPA +HSUPA) Home Subscriber Server High Speed Downlink Packet Access Link Budget Low Delay Constrained Data Long Term Evolution Maximum Allowed Path Loss Modulation Coding Scheme Multiple Input Multiple Output Mobility Managment Entity Mobile Station

7

MSC NAS NGMN NSS OFDM OSS PRB PS QoS RAN RAT RB RE RF RLB RNP RNSs RRU RSRP SAE SC-FDMA SIM SIMO SINR SNR SRC TCH TDD TDMA TMA TSL UDD UE UL UMTS UTRAN WCDMA

Mobile Switching Centers Non-Access Stratum Next Generation Mobile Network Network and Switching Subsystem Orthogonal Frequency Division Multiplex Operation and Support Subsystem Physical Resource Block Packet Switched Quality of Service Radio Access Network Radio Access Technology Resource Block Resource Element Radio Frequency Radio Link Budget Radio Network Planning Radio Network Subsystems Remote Radio Units Referance Signal Received Power System Architecture Evolution Single Carrier Frequency Division Multiple Access Subscriber Identity Module Signle Input Multiple Output Signal to Interfernce and Noise Ratio Signal to Noise Ratio Single Radio Controller Traffic Channel Time Division Duplex Time Division Multiple Access Tower Mounted Amplifier Time Slot Unconstrained Delay Data User Equipment Uplink Universal Mobile Telecommunication System UMTS Terrestrial Radio Access Network Wideband Code Division Multiple Access

8

CHAPTER ONE INTRODUCTION 1.1 Background A cellular network or mobile network is a communication network where the medium used to transmit data or voice from the transmitter to the receiver is natural medium like air and water. The network is distributed over land areas called cells, each served by at least one fixedlocation transceiver, known as a cell site or base station. Multi radio access technologies are technologies which give radio access to the customers, which means they provide cellular network to the users. Generally multi radio access technology incorporates the first generation or GSM, UMTS, and the fourth one which is Long Term Evolution. The networks support more simultaneous users than would be possible without deploying cellular solution using efficient utilization of the spectrum over the network coverage area. However the overall service comes out after a proper dimensioning and planning. The idea of network planning is a complicated process made up of several phases putting final target to define the network design, which let built a cellular network. The designing procedure can be an extension of the existing network or a new network to be launched. The difficulty during the planning is to combine all the requirements in an optimal way and designing a costeffective network since the cellular system consists various components such us User Equipment (UE), Base-Stations (BS), mobile switching centers (MSC) and so on. [1]. The network planning process consists of several phases, the first stage is preplanning or dimensioning covers the assignments and preparation before the actual network planning is started, dealing what kind of services will the network provide, what kind of requirements the different services impose on the network, the basic network configuration parameters and so on.[1] Takes an input from the dimensioning as initial network configuration. Nominal planning would follow. It starts for the site survey, finding the real site locations. Once it supplemented i.e. it has all the data related to the geographical properties and the estimated traffic volumes at different points of the area will be incorporated into a digital map, which consists of different pixels, each of which records all the information about the selected site locations.

9

Based on the propagation model, allowable maximum path loss is calculated the link budget parameters, to define the cell range and coverage threshold. From the point of the digital map and the link budget a computer simulations will evaluate the different possibilities to build up the radio network. The goal is to achieve as much coverage as possible with the optimal capacity. The coverage and the capacity planning are essentially important in the whole radio network planning. The coverage planning determines the service range, and the capacity planning determines the number of to-be-used base stations and their respective capacities. In the third phase, constant adjustment will be made to ensure an optimal operation of the network. Finally the multi RAT network radio plan is ready to be deployed in the area to be covered and served. An investigation of the performance of multi RAT network with prediction of effective service area and in general the three technology will proceed to be evaluated in terms of several parameters such as coverage signal level, overlapping zone, throughput, and interference following. 1.2 Literature Survey A dozen of literatures for the basis of building a strong foundation has been reviewed. Tibebu Mekonnen[1]: deals on Dimensioning and Planning of Multi RAT Radio Network for Future Deployment in Bahir Dar City, it has been tried to consider the limitations of the current telecom services according to their capacity and coverage. And also deals with the procedure of how to carry out the radio network planning for 2G, 3G and 4G systems. The general steps and methods for wireless radio network planning are first addressed. Then the issues of radio network planning for multi radio access technologies including GSM, UMTS & LTE are discussed with special focus on the coverage, capacity and frequency planning. Bethelhem Seifu[2]: LTE Radio Network Planning Modeling for the Case of Addis Ababa. In the paper states that the coverage estimation is done with consideration of the real environment information at its nominal stage to obtain better estimations. The propagation modeling is done using COST 231 W/I model with inclusion of additional parameters obtained from the real environment/terrain model which improves the coverage estimation. Yiming Sun, 2004 [5]: Radio Network Planning for 2G and 3G. Focused on the procedures of how to carry out the radio network planning for 2G and 3G systems, specially the link budget calculation. The general steps and methods for wireless radio network planning are 10

first addressed. Then the issue of radio network planning is discussed with special focus on the 2G and 3G networks, as well as a comparison between 2G and 3G radio network planning processes is summarized at the end. A. Benjamin Paul & Sk.M. Subani, 2012 [7]: Code Planning of 3G UMTS Mobile Networks using ATOLL Planning Tool”, International Journal of Engineering Research & Technology (IJERT). This paper involves on simulation exercise on planning of 3G UMTS network with the help of Atoll planning software tool. It involves planning of coverage, quality& capacity of UMTS Network which uses WCDMA in radio interface between 3G base station and the User equipment. It also involves planning of scrambling codes for 3G WCDMA Network. Jaana Laiho [12]: a doctoral thesis is published by Helsinki University of technology radio laboratory publications in July 2002. It focuses on Wide Band Code Division Multiple Access (WCDMA) (Frequency Division Duplex (FDD) mode) radio network planning and optimization process. For the planning part, the link budget model, radio network planning process using Radio Network Planning (RNP) tool is thoroughly discussed. 1.3 Statement of Problem The issues encountered as defect stand form view angle of two points. It is obvious that the advancement of technology lets an evolution. Since the existing system was install with a lot of investment in case something new happen in middle. A study how do these technologies behave, there compatible in one situation should been considered. The concept of multi RAT is similarly lies in integrating the various radio access technology in common. Investigating the characteristics and interaction between those technologies is crucial. Making business is the other angle since operators are curious in minimize their investment for infrastructure while the equipment vendor’s wants to sell more in turn. Therefore these two extreme interests need to be tradeoff through an effective planning methods which has an attribute of determining the minimum network resources required to meet coverage, capacity and quality demands of the operator. 1.4 Objective 1.4.1 General Objective The general objective of this work is dimensioning and planning of multi RAT access network and investigating its performance.

11

1.4.2 Specific Objectives Specifically this thesis is focused to the following specific radio network dimensioning and planning:  Dimensioning the 2G, 3G and 4G cellular network for case of Kedamay weyane and Adi haki sub cities based on link budget calculation  Modelling the 2G and 3G networks and analyze in terms of signal strength, interference and overlapping zones.  Modelling the 4G network and analyze in terms of signal strength, throughput, interference, and overlapping zones.  Modelling the multi RAT network and analyze in terms effective service area. 1.5 Methodology The employed methods to achieve the objectives of this thesis goes procedurally on these basic as follow steps: 

Review related literatures: starting from finding out papers, journals and articles related areas of research with this paper help to fine tune the statement of problem. The problem that needs to be addressed through the current work needs to be well understood. Reviewing of similar literature helps to widen the viewpoints for the statement of problem.



Study the multi radio access network technologies: making an understand is the initial thing so that the challenges associated with the radio network planning and stuff like that would be explored. The architectures of each multi RAT, there multiple access, the channel and the frame structure of the technologies are reviewed.



Sample Site exploring: doing survey to collect data from the live cellular network, identify and collect the required information from the ground.



System design: includes studying problems of network and putting down system flow, by analyzing the modeling techniques and design a system that can co-plan GSM/UMTS/LTE as multi RAT environment. And identify all necessary inputs for simulation including the digital map of the area.



Practice on the planning tools: the coverage prediction and capacity simulation are made using a commercial radio network planning tools known to be Attol is required. And to use the planning tool effectively working knowledge of the tools should be developed.

12



Simulation and analysis: Based on coverage prediction and capacity simulation result the efficiency of the planning output is evaluated.



Analysis and Interpretation of the results: putting a conclusion of Multi RAT deployment scenarios and overall performance will be analyzed based on the likely outcome.

Generally the points or the procedure included in the methodology followed are illustrated as in the Figure 1.1.

Figure 1.1: Flow Chart of the Methodology 1.6 Significance of the Paper for the Society The significance of this thesis straightforward it could quench the thirsty of great quality services. Service provider needs to provide high quality service with in very low cost and it needs to get high profit from its services, so as the company would gain the profits it has been required. Through an efficient planning introduced here an effective way of resource utilization would be headed anything without allowance of the operator vendors will not install any kit for their own reason.

13

1.7 Organization of Thesis This paper has contains five chapters from the beginning of the first chapter which deals about background of the study, literature survey, objective and methodology. The second chapter also consist of the general over view of these three radio access technologies. In third chapter the multi RAT network planning is discussed from the capacity and coverage point of view. At the end the simulation result discussion and the conclusion and recommendation for future work are put in the fourth and fifth chapters respectively. 1.8 Scope of the Thesis The scope of this thesis lies on the coverage and capacity planning of Multi radio access network (RAT). Accounting the actual morphology and topography details of Adi haki and Kedamay weyane an investigate measure of the performance has been taken. the optimizing procedure, planning and dimensioning of some other features of multi RAT like scrambling code, frequency planning as well as core network dimensioning stuffs are out of our range.

14

CHAPTER TWO RADIO ACCESS TECHNOLOGIES OVERVIEW 2.1 Introduction In this section the general introduction and the overall architecture of radio access technologies are discussed. The intended technologies for network planning are GSM, UMTS and LTE. 2.2 Global System for Mobile (GSM) GSM was launched in the early 1990s, as one of the first truly digital systems for mobile telephony. It was specified by ETSI and originally intended to be used only in Europe GSM, later on has evolved to be more or less the first truly global standard for mobile communication. Even though it is relatively old, it is still being rolled out all over the world. 2.2.1 GSM Architecture GSM system comprises three subsystems named Mobile station (MS), Base-Station Subsystem (BSS), Network and Switching Subsystem (NSS) and Operation and Support Subsystem (OSS).

Figure 2.1: GSM Network Architecture 2.2.2 Multiple Accessing in GSM The cellular systems use various mechanisms to allow multiple users accessing the same radio spectrum at the same time. FDMA and TDMA are the most common ones. The FDMA system divides the available spectrum into several frequency channels. Individual users are allocated two channels for uplink and downlink communication so that no other user could allocated 15

the same channels at the same time. In TDMA system the entire available bandwidth is shared by single user at a time only for short periods. The frequency channel is divided into timeslots so that it’s allocated periodically. GSM is based on TDMA technology, each frequency channel is divided into several time slots and each user is allocated one or more slots. 2.2.3 GSM Channels In the air interface there are two types of channels named physical and logical channels. Generally the physical channel is all the time slots of BTS and also it’s divided in two types, Half Rate (HR) and Full-Rate (FR). The Logical channel refers to specific type of information that is carried by the physical channel. It’s also divided into two types these are traffic channels and control channels. Basic Channel Structure The radio spectrum in GSM 900 is separated into 124 radio channels, each of these radio channels then separated into eight time‐divided channels called time slots (TSL). 2.3 Universal Mobile Telecommunications System (UMTS) It is one of the third generation technologies that uses Wideband Code Division Multiple Access (WCDMA) as the underlying standard. It support two basic modes FDD and TDD, variable transmission rates, inter cell asynchronous operation, adaptive power control, etc. are some of the key features. 2.3.1 UMTS Network Architecture UMTS network architecture consists of Core Network (CN), UMTS Terrestrial Radio Access Network (UTRAN) and User Equipment (UE). UE contains the mobile phone and the SIM (Subscriber Identity Module) card called Universal SIM (USIM). It has a specific data of subscribers that enables the authenticated entry of the subscriber into the network. UTRAN consists of one or more RNSs (radio network subsystems), which in turn consist of base stations Node B’s and RNCs (radio network controllers). The RNS performs all of the radio resources and air interface management functionalities. Radio Network Subsystem (RNS) is the equivalent of the previous Base Station Subsystem or BSS in GSM. It provides and manages the air interface for the overall network.

16

Radio Access Network (RAN) consist of the Base Stations (BS) or Node B and Radio Network Controllers (RNCs). The major functions of the BS are closed loop power control, physical channel coding, modulation/demodulation, error handling, etc., while RNCs are radio resource control/management, power control, channel allocation, ciphering, etc. Core Network (CN) provide switching, routing and transit for user traffic. It also contains the databases and network management functions. Uu Interface is WCDMA radio interface through which the UE accesses the fixed part of the system. It is probably the most important open interface in UMTS. IU interface connects UTRAN to CN. It is an open interface too that divides the system into radio-specific UTRAN and CN which handles switching, routing and service control.

Figure 2.2: UMTS Network Architecture 2.3.2 UMTS Operation Modes and Multiple Access UMTS might work in two different modes i.e. TDD and FDD in other words the channels in the UL and DL will be managed in two different ways: In FDD mode two pairs of frequency bands are used at the same time, one for UL and the other for DL. Its uses WCDMA, the carried services being characterized by their symmetric traffic, like voice. In the TDD mode, both the UL and DL use the same frequency, through a scheme of Time Division - Code Division Multiple Access (TD-CDMA) in unpaired bands, which will be advantageous to handle services with asymmetric traffic, like Internet one. The wide bandwidth of WCDMA offers an inherent performance gain over the previous cellular systems, since it reduces the fading of the radio signal. It uses coherent demodulation in UL, a feature that was not implemented in cellular CDMA systems. [14]. 17

2.4 Long Term Evolution (LTE) LTE is a standard wireless communication of high-speed data for mobile and data terminals. It is based on the GSM/EDGE and UMTS/HSPA network technologies, increasing the capacity and speed using a different radio interface together with core network improvements. All LTE devices have to support Multiple Input Multiple Output (MIMO) transmissions. The interfaces between network nodes in LTE are Internet Protocol (IP) based. 2.4.1 LTE Network Architecture The simplified network architecture with open interfaces of LTE is introduced to be all-IP based. The architecture is designed to be more simplified and flat as compared to the previous 3GPP releases. Since LTE is the evolution of UMTS, its equivalent components are named Evolved Universal Terrestrial Radio Access (E-UTRA). This is the air interface includes the user equipment (UE) and Evolved Universal Terrestrial Radio Access Network (E-UTRAN) and it is used to describe RAN. System Architecture Evolution (SAE) has been introduced in the new architecture instead of a radio controller. The combination of the EPC, E-UTRA, and E-UTRAN is called Evolved Packet System (EPS). As shown in Figure 2.3 the LTE network architecture and the network components are: Evolved Packet Core (EPC) In LTE both Circuit-Switched (CS) and Packet-Switched (PS) mobile core sub-domains are unified as a single IP domain and becomes Evolved Packet Core (EPC). All-IP mobile core network specified by 3GPP Release 8 for LTE. User Equipment (UE) is refers to the LTE mobile station. The UE categories stand for an abstract grouping of common UE radio access capabilities and are defined in 3GPP 36.306. The maximum possible bit rate of UL ranges from 5Mbps (Cat. 1) to 75Mbps (Cat. 5). Evolved-UTRAN (E-UTRAN) consist of the enhanced NodeB (eNodeB) which handles the radio communications between the mobile and the evolved packet core. Each eNodeB is a base station that controls the mobiles in one or more cells. Evolved Packet Core (EPC) contains the home subscriber server (HSS), which is a central database that contains information about all the network operator’s subscribers.

18

Figure 2.3: LTE Evolved Packet System (EPS) architecture Evolved Packet System (EPS) Architecture EPC + eUTRAN builds the Evolved Packet System (EPS). LTE/SAE is specified from Release 8. The name of the actual Radio Access Network (RAN) is eUTRAN and for Core Network (CN) is Enhanced Packet Core (EPC). The eUTRAN supports use of different MIMO (Multiple Input Multiple Output) multiple antenna configurations. This increases the data rates and spectrum efficiency. One of the objectives E-UTRAN is to simplify and reduce the number of interfaces. Mobility Management Entity (MME) is the main control element in the EPC used to process signaling between the CN and the UE. The protocols running between the UE and the CN are called as the Non-Access Stratum (NAS) protocols. The S-GW (Serving Gateway): is responsible for IP packet transferring. It acts as a router, and forwards data between the base station and the packet data network (PDN) gateway. 2.4.2 LTE Physical Layer The design of LTE physical layer is heavily influenced by requirements of high peak transmission rate (100 Mbps DL or 50 Mbps UL), spectral efficiency, and multiple channel bandwidths (1.25-20 MHz), so that an Orthogonal Frequency Division Multiplex (OFDM) was selected as the basis for the physical layer to fulfill the requirements,. 19

OFDMA and SC-FDMA LTE has a multiple access scheme of Orthogonal Frequency Division Multiple Access (OFDMA) in the downlink and Single-Carrier Frequency Division Multiple Access (SCFDMA) in the uplink. The OFDM signal can be generated by using the Fast Fourier Transform (FFT). The available spectrum is divided into multiple, mutually orthogonal subcarriers. The OFDM technique applied for a signal with 5 MHz bandwidth shown in Figure 2.4 [6].

Figure 2.4: Frequency-time representation of an OFDM Signal In the frequency domain, the 5MHz bandwidth is divided into a high number of closely spaced orthogonal subcarriers. The subcarriers in LTE have a constant spacing of 15 kHz. In EUTRA, the downlink modulation schemes can be QPSK, 16QAM and 64QAM. In E-UTRA, the guard interval is a cyclic prefix (CP) which is inserted prior to each OFDM symbol. A group of subcarriers is called a sub-channel. Adaptive Modulation and Coding (AMC) In cellular systems, the quality of the received signal by UE depends on the channel quality from serving cell, level of interference from other cells, and noise level. To optimize system capacity and coverage for a given transmission power, the transmitter should try to match the information data rate for each user to the variations in the received signal. This is commonly referred to as link adaptation and is typically based on Adaptive Modulation and Coding (AMC). The AMC consists of the modulation Scheme and code rate. Modulation Scheme: Low-order modulation (i.e. few data bits per modulated symbol, e.g. QPSK) is more robust and can tolerate higher levels of interference but provides a lower transmission bit rate. High-order modulation (i.e. more bits per modulated symbol, e.g. 64QAM) offers a higher bit rate but is more prone to errors due to its higher sensitivity to interference, noise and channel 20

estimation errors, it is useful only when the Signal to Interference and Noise Ratio (SINR) is sufficiently high. Code rate: For a given modulation, the code rate can be chosen depending on the radio link conditions: a lower code rate can be used in poor channel conditions and a higher code rate in the case of high SINR [13]. LTE supports the following modulation techniques in the downlink and uplink: 

64 Quadrature Amplitude Modulation (64 QAM) which uses 64 different quadrature and amplitude combinations to carry 6 bits per symbol



16 Quadrature Amplitude Modulation (16 QAM) which uses 16 different quadrature and amplitude combinations to carry 4 bits per symbol



Quadrature Phase Shift Keying (QPSK) which used 4 different quadrature’s to send 2 bits per symbol [15, 16].

In LTE each subcarrier is modulated with a conventional modulation scheme depending on the channel condition. LTE uses QPSK, 16QAM, or 64QAM. The FFT sizes of 128, 256, 512, 1024 and 2048, corresponding to LTE channel bandwidth of 1.25, 2.5, 5, 10 and 20MHz are used. Guard intervals are inserted between each of the symbols to prevent inter-symbol interference at the receiver caused by multipath delay spread in the radio channel [16]. Spectrum Flexibility depending on regulatory aspects in different geographical areas, radio spectrum for mobile communication is available in different frequency bands in different bandwidths. LTE can be deployed with bandwidths ranging from approximately 1.25MHz up to approximately 20MHz. Furthermore, LTE can operate in both paired and unpaired spectrum by providing a single radio access technology that supports frequency-division duplex (FDD) as well as time division duplex (TDD) operation. 2.4.3 LTE FDD Frame Structure The LTE FDD frame structure is illustrated in Figure 2.5 for normal cyclic prefix (CP). Each LTE FDD radio frame is 𝑇𝑓 = 307200 × Ts = 10ms long and consists of 20 slots of length 𝑇𝑠𝑙𝑜𝑡 = 15360 × 𝑇𝑠 = 0.5ms, numbered from 0 to 19. For LTE FDD, 10 sub-frames are available for downlink transmission and 10 for uplink transmissions in each 10ms interval. UL and DL transmissions are separated in the frequency domain [16, 17].

21

Figure 2.5: LTE FDD Frame and Slot Structure Resource Blocks (RB) A physical resource block (PRB) is used to describe the physical resource in the time/ frequency grid. Figure 2.6 illustrates the LTE time/frequency grid definitions. The PRB consists of 12 consecutive subcarriers and lasts for one slot, 0.5ms. Each subcarrier is spaced by 15 kHz. The NRB UL and NRB DL parameter are used to define the number of RB (resource blocks) for uplink and downlink respectively. Each resource block consists of NSCRB subcarriers for standard operation is set to 12 or a total of 180 kHz lasting in a 0.5ms slot. The resource element (RE) is the smallest defined unit, which consists of one OFDM subcarrier during one OFDM symbol interval. Each RB consists of 12 × 7 = 84 REs in the case of normal CP and 72 REs for extended CP. The maximum RB is 100 and corresponds to the transmission bandwidth while 20MHz is the channel bandwidth. The number of subcarriers depends on the system BW (i.e. 1.4→72, 3→180, 5→300, 10→600, 15→900, 20→1200) [16]. 2.4.4 LTE MIMO Basics Multiple Input Multiple Output (MIMO) refers to the use of multiple antennas at the transmitter and receiver side. There are two functionality modes of MIMO in which different gains can be achieved. The Spatial Multiplexing mode: allow transmitting different streams of data simultaneously on the same resource blocks by exploiting the spatial dimension of the radio channel so that the data rate or capacity is increased. Spatial Diversity: used to exploit diversity and increase the robustness of data transmission. Each transmitter antenna transmits essentially the same stream of data, so the receiver gets replicas of the same signal [13]. Only the spatial multiplexing mode is concerned in this thesis while calculating the LTE capacity and data rate. As shown in Figure 2.7 taking 4 x 4 antenna configuration as an instance, where each receiver antenna may receive the data streams from all transmit antennas. 22

Figure 2.6: Physical Resource Block and Resource Element The transmission relationship can be described with a Transmission Channel Matrix H. The coefficients ℎ𝑖𝑗 stands for transmit antenna 𝑗 to receive antenna 𝑖 thus describing all possible paths between transmitter and receiver sides.

Figure 2.7: MIMO Transmission Suppose the receiver vector is 𝑦, the transmitter vector is 𝑥, the noise vector is 𝑛 and 𝐻 is the transmission channel matrix. Then MIMO transmission described by the Equation 2.1. 𝑦 = 𝐻𝑥 + 𝑛

2.1

In M x N antenna configuration, the number of data streams which can be transmitted in parallel over the MIMO channel is given by the minimum value of M and N. It’s limited by the rank of the transmission matrix H. For example, a 4 x 4 MIMO system could be used to transmit four or fewer data streams. 23

CHAPTER THREE MULTI RAT RADIO NETWORK PLANNING 3.1 Introduction It is refers to the process of designing a multiple radio access network structure and determining network elements subject to various design requirements. In this thesis the network planning is associated with dimensioning and planning.

Figure 3.1: Network planning process steps The network planning procedure is a complicated process consisting of several phases. The final target is to define the network design, which is then built as a cellular network. Generally the process is divided into five main steps starting from preplanning, planning, detailed planning, acceptance and optimization. The input for the preplanning phase is the network planning criteria which is then used as an input for the main activity of preplanning or dimensioning. The following are basic inputs for dimensioning: 

Coverage requirements, the signal level for outdoor and indoor with the coverage probabilities.



Quality requirements, call blocking.



Subscriber information, traffic per user, busy hour value and Services.

The result of dimensioning has two aspects, it tells the minimum number of base stations due to coverage or capacity reasons. The planning phase takes input from the dimensioning, initial network configuration. This is the basis for nominal planning, which means radio network coverage and capacity planning with a planning tool. Detailed planning covers frequency, neighbor and parameter planning. After detailed planning the network is ready for verification and acceptance, which finishes the prelaunch activities. After the launch the activities continue with optimization [2]. The basic requirements are to meet the coverage and quality targets. Coverage targets the geographic area where the network covers with agreed location of probability. The quality 24

targets are related to factors such as success of the call, drop call ratio, which should not exceed the agreed value and the success ratio for the call setup and for handovers [1]. A good plan should address the following issues provision of required capacity, optimum usage of the available frequency spectrum, minimum number of sites, provision for easy and smooth expansion of the network in future & provision of adequate coverage of the given area, for a minimum specified level of interference [5, 18]. The detailed part of radio network plan can be sub-divided into three sub-plans: Link budget calculation, coverage, capacity planning and spectrum efficiency and Parameter planning. Link Budget Calculations Link budget (LB) calculations give the loss in the signal strength on the path between the mobile station antenna and base station antenna. And radio link budget (RLB) analysis should be done for both uplink and downlink communications [2]. Coverage Planning The objective of coverage planning phase in coverage limited network areas is to find a minimum amount of cell sites with optimum locations for producing the required coverage for the target area. The basic input information for coverage planning includes coverage regions, coverage threshold values on per regions (outdoor, in-car, indoor), Antenna, preferred antenna line system specifications, preferred base station specification & activities such as propagation modeling, field strength predictions and measurements are usually referred to as coverage planning. Capacity Planning The steps for calculating the network capacity are: 

Find the maximum no of carriers per cell that can be reached for the different regions based on the frequency reuse patterns and the available spectrum.



Calculate the capacity of the given cell using blocking probability and the number of carriers.



Finally the sum of all cell capacities gives the network capacity.

3.2 Site Survey The purpose of site survey is to identify the different environmental factors that directly or indirectly affect the radio network planning process. In this thesis, Kedamay Weyane and Adi Haqi sub cities of Mekelle are considered as the area of planning and its environmental factors are listed in Table 3.1. 25

Table 3.1: Regions of Planning SN.

Name

Popn. Size

Area (km2)

Density(per km2)

No. kebeles

Remark

1 2 3

Kedamay Weyane Quiha Hadinet

32,035 45,627 49,566

7.09 16.33 10.67

5,005.5 2,794 2,725

4 4 5

Dense Urban Sub Urban Sub Urban ( aynalem & debri)

4

Adi Haqi

51,590

4.9

13,228

4

Dense Urban

5

Hawelti

61,507

17.32

3,551.2

5

Urban

6

Semen

53,057

37.35

1,420.5

5

Urban

7

Ayder

40,878

12.59

3,246.86

5

Sub Urban

The selected area covers around 11.99 km2 with a population of 83,625 from total of 454,207. 3.3 GSM Radio Network Planning The first considered multi RAT network is GSM, which is divided into a lot of cells, and usually a base station is planted in the center of each cell. For the sake of easy analysis, the cells are represented as neighboring hexagons, while in reality they can be of any kind of forms and overlap with each other. The size of each cell, when fixed, will usually stay stable. The one important feature of GSM network planning is both the coverage and capacity planning are independent. The coverage planning depends on the received signal strength, while the capacity planning depends mainly on the frequency allocation [1, 5]. 3.3.1 GSM Coverage Planning The radio network dimensioning parameters have an impact on each other so it is good to decide the parameters for optimal result. In this thesis a macro cells with three sector sites are considered that spans a 1 to 35km and is characterized by an outdoor antenna [3, 5]. Link Budget Calculations The radio link budget aims to calculate the cell coverage area. One of the required parameters is radio wave propagation to estimate the propagation loss between the transmitter and the receiver. The other required parameters are the transmission power, antenna gain, cable losses, receiver sensitivity and margins, as shown in Figure 3.2.

26

Figure 3.2: Link Budget Parameters When defining the cell coverage area, the aim is to balance the uplink and downlink powers. The links are calculated separately and are different from the transmission powers. The GSM link budget parameters are classified into four types these are system parameter, transmitter parameters, receiver parameters, and margin reservation System Parameter Carrier frequency: radio waves of different frequencies have different propagation models and different transmission losses. System bandwidth: In a GSM system, the receiver bandwidth is 200 kHz (that is, 53dBm) Data rate: The full rate and half rate of GSM voice service is given as 9.6Kbps and 4.Kbps respectively. Transmitter Parameter MS and BTS powers are important along with the sensitivities. The MS TX (transmission) power is defined by the MS class in ETSI specifications. For MS class 4 (GSM 900) the maximum TX power is 2W. BTS TX power depends on the BTS type and vendor. The TX power is adjustable, which enables the link budget to be balanced. Antenna gains is dependent on the antenna type and whether the antenna is omnidirectional or directional. The antenna gain is around 16 – 20dBi when there is a widely used antenna with 60–65◦ horizontal half power beam width and 5–10 vertical half power beam width. In the link budget calculations the MS antenna the gain is 0dBi. Body Loss is a loss generated due to signal blocking and absorption when a terminal antenna is close to the human body. It’s always taken 3dB and 0dB for the case of MS and BTS respectively. Receiver parameter 27

Noise spectral density is calculated as: 𝑠𝑝𝑒𝑐𝑡𝑟𝑎𝑙 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 = 𝑘𝑇 = −174 𝑑𝐵𝑚⁄𝐻𝑧 𝑖𝑛 𝑟𝑜𝑜𝑚 𝑡𝑒𝑚𝑝𝑟𝑎𝑡𝑢𝑟𝑒(290𝐾)

3.1

Noise power also called thermal noise power is produced by the thermal movement of electrons. It is given by: 𝑁𝑖 = 𝑘𝑇𝐵

3.2

C/I required by TCH: is the SNR requirement on the air interface. In narrowband system, C/I is the requirement of receiver baseband demodulation performance. The target value varies depending on the propagation environment, mobility speed, and coding rate. According to the GSM protocol, C/I should be greater than or equal to 9dB [6]. Noise figure: is a method to measure the noise added when a signal is pass through BTS or/and MS receiver. In the case of ideal receiver nose figure is F = 1 or 0dB. In actual situations, a receiver has noise and the output noise power is greater than signal power. Thus, the SNR is worse and F > 1. As defined in GSM protocol, the noise figure of a base station receiver is 8 dB and 10dB for MS receiver. 𝐹=

𝑆𝑖 ⁄𝑁𝑖 𝑆𝑖 ⁄(𝑘𝑇𝑜 𝐵) = 𝑆𝑜 ⁄𝑁𝑜 𝐸𝑏 ⁄𝑁𝑜

3.3

𝑤ℎ𝑒𝑟𝑒 𝑁𝑖 = 𝑘𝑇𝑜 𝐵

Receiver sensitivity is a measure of how well the receiver performs and defined as the power of the weakest signal the receiver can detect. The receiver sensitivity is given by: 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟 𝑠𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 = 𝑛𝑜𝑖𝑠𝑒 𝑠𝑝𝑒𝑐𝑡𝑟𝑎𝑙 𝑑𝑒𝑛𝑠𝑖𝑡𝑦(𝑑𝐵𝑚⁄𝐻𝑧) + 𝑏𝑎𝑛𝑑𝑤𝑖𝑑𝑡ℎ(𝑑𝐵𝐻𝑧) + 𝑛𝑜𝑖𝑠𝑒 𝑓𝑖𝑔𝑢𝑟𝑒(𝑑𝐵) + 𝐶⁄𝐼 (𝑑𝐵) 𝐸𝑏 𝑆𝑖 = 𝐹𝐾𝑇𝑜 𝐵 3.4 𝑁𝑜 BTS sensitivity has a general recommendation value of -106dBm as it is specified on the ETSI GSM recommendation 05.05. MS sensitivity is also specified in the ETSI recommendation 05.05, where the receiver sensitivity value for MS class 4 the recommended value is -102dBm. The MS sensitivity can also be calculated using the information of receiver noise F and minimum Eb/N0. The value for the noise is 10dB and the minimum Eb/N0 is 8dB, as defined in the ETSI recommendation 03.30. The receiver sensitivity Si is solved from Equation 3.3 at room temperature T = 290 K, and BW = 200 kHz (53dBm). 𝐹=

𝑆𝑖 ⁄𝑁𝑖 𝑆𝑖 ⁄(𝑘𝑇𝐵) = , 𝑆𝑜 ⁄𝑁𝑜 𝐸𝑏 ⁄𝑁𝑜

𝑤ℎ𝑒𝑟𝑒 𝑁𝑖 = 𝑘𝑇𝐵

28

𝑆𝑖 =

𝐸𝑏 𝐹𝐾𝑇𝐵 𝑁𝑜

𝐽 = 9𝑑𝐵 + 10𝑑𝐵 + (290𝐾 × 1.38 × 10−23 ) + 53𝑑𝐵𝑚 𝐾 = 9𝑑𝐵 + 10𝑑𝐵 − 174𝑑𝐵𝑚 + 53𝑑𝐵𝑚 = 102𝑑𝐵𝑚 Interference degradation margin: describes the loss due to frequency reuse. Therefore the frequency reuse rate corresponds to the degradation margin value. The suggested value for the interference degradation margin in dense urban areas have a value of 4 – 5dB according to the ETSI recommendation 03.30. Margin Reservation Diversity gain can be used for correcting unbalance between the uplink and downlink. The typical way to arrange diversity is to have it in the BTS reception. One basic method is to separate receiver antennas vertically or horizontally which is known to be space diversity. The diversity decreased fading effect and gain achieved can be around 5dB. Cable and connector losses are case specific and need to be measured or calculated separately. The cable losses can be seen from Table 3.2. An individual connector gives a loss of around 0.1dB, but depending on the cable installations there can be several. Table 3.2: Cable loss per 100m Cable type ½ inch 7/8 inch 1 ¼ inch

Loss/100m 1800MHz 10 dB 6 dB 4.5 dB

900MHz 7 dB 4 dB 3 dB

Slow fading margin / Shadow fading margin: also named slow attenuation. It follows a lognormal distribution in the calculation of radio coverage. To reach the specified coverage probability, during network planning, certain power margin must be reserved for BS or MS receivers to reduce the attenuation effect. Shadow fading standard deviation is related to electromagnetic wave propagation environment. In dense urban areas, the shadow fading standard deviation is about 10dB. Coverage probability: refers to the probability that the quality of communication between terminals in radio coverage edge (or inside coverage) and the base station meet the requirement. Coverage probability can also be classified into area coverage probability and edge coverage probability. 29

Edge coverage probability: is used to evaluate the reliability of communication links in shadow fading environment. It is an index for determining the coverage quality. To determine the location probability (i.e. describes the probability of the receiver being able to capture the signal) a distribution for the received signal has to be defined. The slow fading variations in the average received signal level are normally distributed, which is illustrate in Figure 3.3.

Figure 3.3: Lower tail of normal distribution curve In radio propagation, for a given distance, the path loss changes quickly and can be regarded as a random variable in lognormal distribution. To improve cell coverage, fade margin should be considered in link budget. The distribution function for slow fading is: Ƥ(r) =

1 √2𝜋𝜎

𝑒

−

(𝑟−𝑟𝑚 )2 2𝜎2

3.5

Where r is the random variable and rm the mean value of it, and σ is the standard deviation, which is measured in dB. The slow fading is described by the normal random variable r [2, 6]. The location probability can be expressed by an equation, which is upper tail probability of Equation. The location probability for upper tail probability can be expressed as follows: Ƥ 𝑥𝑜 =

𝑥𝑜

1 √2𝜋𝜎

∫ 𝑒

−

(𝑟−𝑟𝑚 )2 2𝜎2

𝑑𝑟

3.6

−∞

The probability Ƥ𝑥𝑜 gives the location probability at a certain point when the random variable r exceeds some threshold xo: Ƥ 𝑥𝑜 =

1 √2𝜋𝜎

∞

∫ 𝑒 𝑥𝑜

−

(𝑟−𝑟𝑚 )2 2𝜎2

1 𝑥𝑜 − 𝑟𝑚 𝑑𝑟 = [1 − 𝑒𝑟𝑓 ( )] 2 𝜎√2

3.7

The location probability can be expressed as well as the lower tail in Equation 3.7, and therefore the probability can be calculated below a certain margin. The planning target for the 30

location probability is normally 90-95% over the whole cell area [6]. The location probability, slow fading margin (xo - rm), maximum path loss and cell range are all connected. The cell range is dependent on the maximum allowed path loss and therefore improvement in the location probability causes a decrease in the cell range. This leads to calculation of the coverage threshold, which is the minimum allowed downlink signal strength at the cell edge with a certain location probability. For the coverage threshold calculations are needed for the MS isotropic power, propagation model with calculation parameters, standard deviation, area type correction factor and building penetration loss. Using the standard deviation and location probability the value of the slow fading margin is first calculated. Area type correction factors come from the propagation model tuning measurements. Building penetration loss is needed in the case where indoor coverage thresholds are calculated. ETSI recommendation 03.30 suggests values for the average building penetration loss (BPL); in urban areas it is approximately 18 dB for 900 MHz and in rural areas around 10 dB, due to the smaller size of buildings. For an indoor coverage threshold calculation the deviation of building penetration loss indoor is also needed, which is used as the standard deviation indoor. The approximate value for the BPL deviation is 10dB. For indoor environment, the standard deviation of propagation loss random variable is around 10dB. Then, the slow fading margin of 90% edge coverage probability is as follows: 𝑥𝑜 − 𝑟𝑚 = 1.29𝜎 = 1.29 × 10 = 12.9𝑑𝐵

3.8

In GSM network planning 12.9 dB margin is reserved to ensure a 90% edge coverage probability. Table 3.1: Common edge coverage probability and shadow fading margin Edge coverage Probability (%)

70%

75%

80%

85%

90%

95%

98%

Shadow fading margin (dB)

0.53σ

0.68σ

0.85σ

1.04σ

1.29σ

1.65σ

2.06σ

Area coverage probability: is the percentage of area of the location where receiving signal strength is greater than receiving threshold to the total area in a round region with radius R. The 90% edge coverage probability corresponds to 95% area coverage probability and shadow fading margin is 12.9dB. Fast fading margin (Rayleigh fading margin) is a type of multipath wave interference generated because the propagation is reflected by scattering objects (mainly buildings) or

31

natural obstacles (mainly forest) around the MS (within 50-100 wavelength). Fast fading always produce standing wave field. Penetration loss refers to building loss that is associated with building style and structure, such as concrete structure, brick structure, window size, style and so on. Maximum allowable path loss The Equation 3.9 is a typical one to determine maximum allowable path loss for a radio communications system. 𝑚𝑎𝑥𝑖𝑚𝑢𝑚 𝑎𝑙𝑙𝑜𝑤𝑎𝑏𝑙𝑒 𝑝𝑎𝑡ℎ 𝑙𝑜𝑠𝑠(𝑀𝐴𝑃𝐿) = 𝑇𝑋 𝐸𝐼𝑅𝑃 + 𝑅𝑋 𝐿𝑜𝑠𝑠𝑒𝑠 − 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟 𝑆𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 − 𝑇𝑜𝑡𝑎𝑙 𝑚𝑎𝑟𝑔𝑖𝑛𝑠

3.9

𝑤ℎ𝑒𝑟𝑒 𝑇𝑥 𝐸𝐼𝑅𝑃 = 𝑇𝑋 𝑝𝑜𝑤𝑒𝑟 + 𝑇𝑋 𝑎𝑛𝑡𝑒𝑛𝑛𝑎 𝑔𝑎𝑖𝑛 − 𝑏𝑜𝑑𝑦 𝑙𝑜𝑠𝑠 𝑎𝑛𝑑 𝑎𝑛𝑑 𝑇𝑜𝑡𝑎𝑙 𝑚𝑎𝑟𝑔𝑖𝑛 = 𝑠𝑙𝑜𝑤 𝑓𝑎𝑑𝑖𝑛𝑔 𝑚𝑎𝑟𝑔𝑖𝑛 + 𝑓𝑎𝑠𝑡 𝑓𝑎𝑑𝑖𝑛𝑔 𝑚𝑎𝑟𝑔𝑖𝑛 + 𝑖𝑛𝑡𝑒𝑟𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑚𝑎𝑟𝑔𝑖𝑛 + 𝑏𝑢𝑖𝑙𝑑𝑖𝑛𝑔 𝑜𝑟 𝑣𝑒ℎ𝑖𝑐𝑙𝑒 𝑝𝑒𝑛𝑒𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑙𝑜𝑠𝑠 As a result of the link budget calculation of Table 3.3, there is a difference in the power budget of the uplink and downlink calculations, so that the downlink path loss exceeds the uplink power loss. This is implies that the area covered by the base station antenna radiations is more than the area covered by the mobile station antenna, thereby giving more coverage in the downlink direction Propagation Model Selection The propagation model for GSM 900MHz Okumura-Hata is selected considering the worst case scenario for better radio network planning. Cell Radius Calculation for GSM 900 MHz The Okumura-Hata model for path loss prediction is given by 𝑃𝐿 = 𝐴 + 𝐵𝑙𝑜𝑔10(𝑓) − 13.82𝑙𝑜𝑔10(𝐻𝑏) − 𝑎(𝐻𝑚) + [44.9 − 6.55𝑙𝑜𝑔10(𝐻𝑏)]𝑙𝑜𝑔10(𝑑) + 𝐿𝑜𝑡ℎ𝑒𝑟 𝑤ℎ𝑒𝑟𝑒

𝐴 = 69.55, 𝐵 = 26.16,

𝑙𝑜𝑔10(𝑓) = log10(900) = 2.954,

𝐵𝑙𝑜𝑔10(𝑓) = 26.16 x 2.954 = 77.276

32

Table 3.3: GSM link budget calculation Scenario Parameter

Cell radius

Margin reservation

Receiving end

Transmitting end

TX power Antenna gain Cable losses Body loss Combiner loss EIRP Antenna gain Cable losses + connector Body loss Noise spectral density, Ni Noise figure Bandwidth C/I Receiver sensitivity Area coverage probability Edge coverage probability Slow fading standard deviation Slow fading margin Fast fading margin Interference degradation margin Indoor penetration loss Sum of margins Max path loss Frequency band Propagation model

Unit W dBm dBi dB dB dB dBm dBi dB dB dBm/Hz dB dBHz dB dBm % % dB dB dB dB

Dense urban UL DL MS BS 2 20 33.01 43.01 0 18 0 3 3 0 0 3 30.01 55.01 18 0 3 0 0 3 -174 -174 10 7 53.01 53.01 9 8 -102 -106 95 90 10 12.9 5 5 4 4

Formula

A B C D E F = A+B-C-D-E G H I J K L M N = J+K+L+M

O P Q

dB dB dB MHz

5 26.9 120.11

0 21.9 136.11 900 Okumura – HATA

R S = O+P+Q+R T = F+G-H-I-N-S

Cell radius

Km

0.662

𝑢 = 10[(𝑈−126.42)/35.21]

Cell radius output

Km

1.885 0.662

13.82𝑙𝑜𝑔10(𝐻𝑏) = 13.82𝑙𝑜𝑔10(30) = 20.41 𝑎(𝐻𝑚) = [1.1𝑙𝑜𝑔10(𝑓) − 0.7]𝐻𝑚 − [1.56𝑙𝑜𝑔10(𝑓) − 0.8] = [1.1 ∗ 2.954 − 0.7] ∗ 1.5 − [1.56 ∗ 2.954 − 0.8] = 0.016 [44.9 − 6.55𝑙𝑜𝑔10(𝐻𝑏)]𝑙𝑜𝑔10(𝑑) = [44.9 − 6.55𝑙𝑜𝑔10(30)]𝑙𝑜𝑔10(𝑑) = [44.9 − 6.55 ∗ 1.48]𝑙𝑜𝑔10(𝑑) = 35.21𝑙𝑜𝑔10(𝑑) 33

𝐿𝑜𝑡ℎ𝑒𝑟 = +2𝑑𝐵 𝑃𝐿 = 𝐴 + 𝐵𝑙𝑜𝑔10(𝑓) − 13.82𝑙𝑜𝑔10(𝐻𝑏) − 𝑎(𝐻𝑚) + [44.9 − 6.55𝑙𝑜𝑔10(𝐻𝑏)]𝑙𝑜𝑔10(𝑑) + 𝐿𝑜𝑡ℎ𝑒𝑟 = 69.55 + 77.276 − 20.41 − 0.016 + 35.21𝑙𝑜𝑔10(𝑑) = 128.42 + 35.2𝑙𝑜𝑔10(𝑑) 𝑑900𝑀𝐻𝑧(𝑘𝑚) = 10[(𝑃𝐿−128.42)/35.21]

3.10

Required number of BTS’s A tri-sector cells in a single base station are considered to provide precise coverage for the selected regions. The coverage area of the tri-sector base station is determined using the following formula for R is radius of single cell [1, 2]. 9√3 2 𝑅 3.11 8 Now the coverage per base station can be predict using Equation 3.11. Thus the computed cell 𝑆𝑖𝑡𝑒 𝐶𝑜𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑟𝑒𝑎 =

radius is 0.662 km. Consequently the coverage area of a single site would be 0.8546 km 2. Therefore a totally of 14 GSM 900MHz base stations are computed from Equation 3.12 for total coverage area of 11.99km2. 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐵𝑇𝑆 =

𝑡𝑜𝑡𝑎𝑙 𝑐𝑜𝑣𝑒𝑟𝑎𝑔𝑒 𝑎𝑟𝑒𝑎 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑠𝑖𝑛𝑔𝑙𝑒 𝐵𝑇𝑆

3.12

3.3.2 GSM Capacity Planning Capacity dimensioning is a very important process in the network rollout as it defines the number of base stations required and their respective capacities. There are three essential parameters required for estimated traffic, average antenna height, and frequency usage. Traffic Estimates is based on theoretical estimates or assumptions, and on studies of existing networks. Traffic in the network is dependent on the user communication rate and user movement in the network. The user communication rate means how much traffic is generated by the subscriber and for how long. The estimated traffic in the network is given in terms of ‘Erlangs’. One erlang (1 Erl) is defined as the amount of traffic generated by the user when he or she uses one traffic channel for one hour (this one hour is usually the busy hour of the network). Another term that is frequently used in network planning is ‘blocking’. Blocking describes the situation when a user is trying to make a call and is not able to reach a dialed subscriber owing to lack of resources. Generally, it is assumed an average of two call per hour and the call lasts an average of 90 seconds per call and individual voice traffic intensity can be 𝐴𝑢 = 𝐻 × 𝜆

3.13 34

Where 𝜆 is the average arrival rate (call request/time), and 𝐻 is average holding time Erlang B offers no queuing for call requests in other words for every requests service if any is available the user is given an immediate access to a channel otherwise the requesting user is blocked without access and is free to try again later. The Erlang B formula determines the probability that a call is blocked and is a measure of the GOS for a trunked system which provides no queuing for blocked calls [4]. The dense urban regions covers about 11.99 km2 with a population of 83,625 from total of 454,207 population of Mekelle city. A sixty percent of the population are assumed to be GSM subscriber so that it becomes total of 50,175 mobile users. Table 3.4: Subscribers for GSM Sub cities

Total population

Adihaki and kedamay weyane

192,047

Dense urban population number 83,625

Number of Subscribers 50,175

The following assumptions has been made in the planning: Table 3.5: Assumption taken for GSM capacity dimensioning Parameter GOS Average calls per hour Average call duration Traffic of the cell Pages per cell SMS ratio

Value 2% 2 calls per hour 90 seconds 35 erlang 2 0.1

Now the individual voice traffic intensity can be calculated using Equation 3.13 as follow: 𝐴𝑢 =

2 𝑐𝑎𝑙𝑙𝑠 𝑠𝑒𝑐𝑜𝑛𝑑𝑠 × 45 3600 𝑠𝑒𝑐𝑜𝑛𝑑𝑠 𝑐𝑎𝑙𝑙

𝐴𝑢 = 0.025𝑒𝑟𝑙𝑎𝑛𝑔 𝑜𝑟 25𝑚𝑖𝑙𝑙𝑖 𝑒𝑟𝑙𝑎𝑛𝑔 𝑝𝑒𝑟 𝑢𝑠𝑒𝑟 Since RF Carriers Needed depend upon the number of subscribers to be served, average traffic per subscriber, grade of service so that the total number of subscribers are 50,175. The total system traffic is given as Equation 3.14: 𝐴 = 𝑈𝐴𝑢

3.14

Where: 𝐴𝑢 traffic intensity per user and 𝑈 number of subscribers 𝐴 = 50,175 × 25𝑚𝐸 = 1254.375 Step 1: Calculate Pages per second: The call rate (Calls/hour) becomes 35

=

1254.375 × 3600 = 100,350 45

Average number of mobile terminated calls = 20% = 20,070 Hence number of pages per hour = 40,140 Number of pages per second = 40,140/3600 = 11.15 Step 2: Calculate NPCH (Number of Paging Channels): 𝑁𝑃𝐶𝐻 =

𝑃𝑎𝑔𝑒 𝑝𝑒𝑟 𝑠𝑒𝑐𝑜𝑛𝑑 2 × 4.25 11.15 = = 1.312 2 × 4.25

3.15

Step 3: Calculate NAGCH (Number of AGCH Channels): For this we need λ values λcall =

a∗3600 90

= 3200 calls/hr in the cell



λSMS = λcall × SMS = 3200 × 0.1 = 320



λLU = λcall × 𝐿 = 3200 × 2 = 6400



λ AGCH = (λcall + λSMS + λLU ) 3600 =



NAGCH = 2×4.25 = 0.3242

1

3200+320+6400 3600

=2.7556 per second

2.7556

Step 4: Calculate NPAGCH (Number of PAGCH Channels): 𝑁𝑃𝐴𝐺𝐶𝐻 =

𝑁𝑃𝐶𝐻 + 𝑁𝐴𝐺𝐶𝐻 𝑈𝐶𝐶𝐶𝐻

3.16

Where, UCCCH is the % utilization of control channels Typically, UCCCH= 0.33, NPAGCH = (1.312 + 0.3242)/0.33 = 4.9581 Blocks NAGCH i.e. we need a minimum of 5 CCCH blocks. Step 5: Calculate SDCCH Requirement: The need of SDCCH channel is crucial during call set-up, location updates & SMS. Hence Erlang offered for SDCCH Channel is given by: λ𝑐𝑎𝑙𝑙 × 𝑇𝑐 + λ𝐿𝑈 × (𝑇𝐿𝑈 + 𝑇𝑔 ) +

λ𝑆𝑀𝑆 (𝑇𝑆𝑀𝑆 + 𝑇𝑔 ) 3600

3.17

The values of time variables 𝑇𝑔 , 𝑇𝑐 , 𝑇𝐿𝑈 , 𝑇𝑆𝑀𝑆 are given with values of 4, 5, 4, and 6 seconds respectively. Therefore the SDCCH Channel Erlang becomes: (3200 × 5 + 6400 × 8 + 320 × 10)/3600 = 19.5556𝐸 Step 6: Calculate SDCCH Channels: A GoS = 2% & E = 19.5556 the number of SDCCH Channels are 27 from Erlang B table. A single timeslot accommodates eight SDCCH Channels, so that three independent Timeslots & one combined timeslot are needed to accommodate 27 SDCCH Channels. 36

Step 7: Number of RF Carriers need Cell Traffic = 50E at 2% GOS 60TCH required Time slots required for traffic channels = 60 Time slots required for control channels = 4 Total Time slots required 60 + 4 = 64 So RF Carriers needed = 64/8 = 8 Table 3.6: GSM Channel Distribution Strategy Channel Distribution Strategy Carrier channel

SDCCH

Control channel

Surplus channel

TCH user

traffic/ cell

PDCH channel

TCH ERL

TCH user

1

8

SDCCH/4

1

7

117

2.0592

1

2.28

91

1 2 3

8 16 24

SDCCH/8 SDCCH/8 SDCCH/8

2 2 2

6 14 22

91 327 595

1 1 2

1.66 7.4 13.18

66 296 527

4

32

2*SDCCH/8

3

29

841

1.6016 5.7552 10.472 14.801 6

2

19.27

770

5

40

2*SDCCH/8

3

37

1130

19.888

3

25.53

1021

6

48

2*SDCCH/8

3

45

1424

4

31.92

1276

7

56

3*SDCCH/8

4

52

1685

25.062 4 29.656

4

38.39

1535

8

64

3*SDCCH/8

4

60

1985

34.936

4

45.87

1834

A total of 64 channels are required four of them are belong to control channel and the rest would be a surplus channel. As a result the number of subscribers per site is 1985*3 = 5955. 𝑅𝑒𝑞𝑢𝑖𝑟𝑒𝑑 𝑛𝑜. 𝑜𝑓 𝑠𝑖𝑡𝑒𝑠 =

𝑇𝑜𝑡𝑎𝑙 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 𝑇𝑜𝑡𝑎𝑙 𝑛𝑜. 𝑜𝑓 𝑎𝑐𝑡𝑖𝑣𝑒 𝑢𝑠𝑒𝑟 𝑝𝑒𝑟 𝑠𝑖𝑡𝑒

=

3.18

50,175 = 8.4257 ≈ 9 5,955

Therefore the minimum site required to meet 50E would be 9 sites since the number of site form coverage dimensioning has been 14 sites. The bottleneck is the dimensioned with highest number so that a total of fourteen number of sites are needed to fulfil our requirements. Table 3.7: GSM Best site selection of coverage and capacity Dimensioning Coverage

Number of sites calculated 14 sites

Capacity

9 sites

Best selection 14 sites

37

3.4 WCDMA Radio Network Planning The fundamental process for coverage planning in the WCDMA system is quite similar to that of the GSM system. However, propagation models need to be adjusted to take into consideration the WCDMA technology. 3.4.1 WCDMA Coverage Planning The link budget is performed as part of coverage dimensioning process to acquire the maximum allowable path loss based on the given planning scenes. And the acquired maximum path loss will be further used to calculate the maximum cell radius based on the radio propagation model. In all these processes the final target is to estimate the number of required bases station to cover the target coverage area [7]. Link budget parameters Based on the signal propagation path from the transmitting end to receiving end, there are two paths uplink and a downlink path. UMTS system could be uplink limited or downlink limited based on the system loading. In a lightly loaded system, the User equipment transmits power sets a coverage limitation, therefore, it is uplink limited. In a heavily loaded system, the base station transmits power limits the number of user equipment it can serve therefore it is downlink-limited. Generally, the uplink is limiting link in terms of radio bearer coverage. But practically the link budget analysis is also performed on the downlink path to verify the assumption [8]. The link budget calculation involves various link parameters associated with gains, losses and design margins of the particular signal transmission path. The major link parameters are [17]. Chip and information rates: system chip rate is 3.84 Mchips/s whereas the information rate depends upon the service. For speech the information rate is 12.2 kbps while for data services it can be 64kbps, 128kbps, 384kbps or HSUPA. The processing gain is defined by the equation: 𝑝𝑟𝑜𝑐𝑒𝑠𝑠𝑖𝑛𝑔 𝑔𝑎𝑖𝑛 = 10 𝑥 log

𝑐ℎ𝑖𝑝 𝑟𝑎𝑡𝑒 𝐵𝑖𝑡 𝑟𝑎𝑡𝑒

3.19

UE velocity: is an important parameter as it impacts the uplink and downlink Eb/No requirements. The Eb/No figures are based upon propagation models assuming 120 km/h and 3 km/h for vehicular and pedestrian environments, respectively. These two UE velocities are used in link budget were deemed to provide the most challenging cases for planning. At 3 km/h the UE suffers from fast fading and therefore power control is required to track this

38

fading environment. At 120 km/h, power control is unable to track the fast fading, and therefore the Eb/No value is now affected by the interleaving performance [11, 16, 19]. Transmitter Parameters UEs Maximum Transmits Power (dBm) The maximum transmits power of the mobile station is dictated by the UE power class. Usually voice-centric UE is class 3 or 4 and data-centric UE is class 3. Therefore the transmit power parameter for the uplink path is specified as 21dBm for voice terminal while for data terminal it is specified as 24dBm [11]. BTS Transmit Power (dBm) The common pilot channel (CPICH) is used by the base station to provide a reference to all mobile stations and aid the channel estimation at the terminals. For this reason, the CPICH channel is used to define the maximum possible coverage of a particular cell. In this thesis the typical CPICH power configuration of a cell taken as 33dBm [8, 16]. Required Eb/No The Eb/No value assumption reflect the performance of the BTS receiver for Uplink link budget calculation, and the value may vary from one vendor to the other. As shown in Table3.8 the Eb/No taken from Huawei product documentation accordingly [10, 14]. Body loss (dBm) Body loss occurs at the UE side. Its value depends on usage habit of the user. The default setting of body loss is 3dB for speech service and 0dB for data services because the UE is far away from the human body [8]. Table 3.8: Required Eb/No Values CS 12.2k CS 64K PS 64K PS 128K PS 384K

Downlink 7.5 5.2 4.8 4.5 4.3

Uplink 4.2 2.7 1.6 1.1 0.6

Antenna gains (dBi) Normally antenna gain for the user equipment is taken as 0dBi whereas for the BTS antenna the value is chosen based on the type of antenna selected for each propagation scenario. For

39

this particular case or research, the Base station antenna selected for all propagation scenarios has an 18dBi gain [15]. EIRP represents the effective isotropic radiated power from the transmitter antenna. In the case of the uplink it is computed from the Equation 3.19: 𝐸𝐼𝑅𝑃𝑈𝐿 = 𝑁𝑜𝑑𝑒𝐵 𝑡𝑟𝑎𝑛𝑠𝑚𝑖𝑡𝑡𝑒𝑟 𝑝𝑜𝑤𝑒𝑟 + 𝑈𝐸 𝑎𝑛𝑡𝑒𝑛𝑛𝑎 𝑔𝑎𝑖𝑛 − 𝑈𝐸 𝑏𝑜𝑑𝑦 𝑙𝑜𝑠𝑠

3.20

Receiver Parameters Noise Figures Noise figure is an index used to evaluate whether noise performance of the amplifier is good or not. It is expressed by Noise Figures (N) and defined as the ratio of input Signal-to-Noise ratio (SNR) and output Signal-to-Noise ratio (SNR) of an amplifier. The noise figure for Huawei Base station DBS3900 is taken as 2.1dB for the case where there is no TMA (Tower mounted Amplifier) [17]. Thermal Noise Spectral Density is -174dBm/Hz in room temperature (300K).The thermal noise density is computed from the Equation 3.1: 𝑁 = 𝑘𝑇 = 1.38 𝑥 10−23 𝑥 290 𝑥 103 = 4 𝑥 10−23 𝑁𝑑𝐵 = 10 log10 (4 𝑥 10−23) = −174𝑑𝐵𝑚/𝐻𝑧 Thermal Noise Power is computed from the Equation 3.2: 𝑁𝑖 = 𝑘𝑇𝐵 Where, k is Boltzmann constant, which equals 1.38 x 10-23J/K, T is absolute temperature 290K, and B is system bandwidth. 𝑁 𝑐ℎ𝑖𝑝𝑟𝑎𝑡𝑒 = 1.38 𝑥 10−23 𝑥 290 𝑥 103 𝑥 106 𝑖

𝑠

= 1.54 𝑥 10−11 𝑁𝑖(𝑑𝐵) = 10 log10 (𝑘𝑇𝐵) = 𝑁𝑑𝐵 + 𝑁𝑑𝐵 = −108.16 𝑑𝐵𝑚 Receiver Noise Power / Total Effective Noise Power: The total effective noise at the receiver, receiver noise power, is computed from the sum of the thermal noise power and the receiver noise figure. 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟 𝑛𝑜𝑖𝑠𝑒 𝑝𝑜𝑤𝑒𝑟 = 𝑡ℎ𝑒𝑟𝑚𝑎𝑙 𝑛𝑜𝑖𝑠𝑒 𝑝𝑜𝑤𝑒𝑟 + 𝑛𝑜𝑖𝑠𝑒 𝑓𝑖𝑔𝑢𝑟𝑒

3.21

The total effective noise indicated in Table 3.9 is computed from the sum of the thermal noise density and the receiver noise figure.

40

Table 3.9: Receiver parameters Parameter (a) Rx antenna gain (dBi) (b) Cable and connector losses (dB) (c) Noise figure (dB) (d) Thermal noise density (dBm/Hz) (e) Total effective noise (dBm/Hz) (f) receiver thermal sensitivity (dBm)

UE 0 0 0 -174

Node B 18 0.5 7 -174

(c) + (d) (e) + 10 log (information rate) + Eb/Io

It is used to define the noise floor when computing the receiver sensitivity. Similarly, the receiver thermal sensitivity is computed based on Equation 3.22. 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟 𝑡ℎ𝑒𝑟𝑚𝑎𝑙 𝑠𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 = 𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 𝑛𝑜𝑖𝑠𝑒 𝑝𝑜𝑤𝑒𝑟 + 𝑟𝑒𝑞𝑢𝑖𝑟𝑒𝑑

𝐸𝑏 ⁄𝑁 − 𝑝𝑟𝑜𝑐𝑒𝑠𝑠𝑖𝑛𝑔 𝑔𝑎𝑖𝑛 3.22 𝑜

Loading Factor and interference margin The link budget includes an interference margin that is based upon the load factor and is given by Equation 3.23: 𝐼𝑛𝑡𝑒𝑟𝑓𝑒𝑟𝑎𝑛𝑐𝑒 𝑚𝑎𝑟𝑔𝑖𝑛 = 10𝑥 log(1/(1 − 𝑙𝑜𝑎𝑑 𝑓𝑎𝑐𝑡𝑜𝑟))

3.23

The load factor is the ratio of actual load to the pole capacity. In this thesis, the typical uplink load factors of 50% are assumed for all propagation scenario. Other Margins and Gains Slow fading margin Slow fading or shadowing is the variation in the local mean of the received signal that is caused by terrain irregularity and is typically lognormally distributed. In this thesis, the standard deviation of shadowing is taken as 10 dB in a dense urban propagation environment. Fast fading margin refers to the attenuation of the signals due to multipath reflections and diffractions. The short-term average of fast fading can typically be represented by a Rayleigh distribution. In slow moving environments, the UE’s closed-loop fast power control can effectively compensate fast fading. It requires an appropriate headroom in the UE transmission power. For the link budget calculation, 2.0 dB is considered for all propagation scenarios. Soft Handoff Gain Depending upon the degree of slow fading correlation between BTSs, soft handoff results in a reduction of the required slow fading margin. In addition, due to macro diversity combining, soft handoff provides gain against fast fading by reducing the required Eb/No. Typical values for soft handoff gain ranges 2-4dB. The values of 4dB is considered for Dense urban [1, 2, 3]. 41

Receiver antenna diversity gain It is assumed that the receiver antenna diversity gain is taken into account in the required Eb/No values [13, 15]. Penetration loss The building penetration losses are assumed to be dependent upon the building construction. In all the propagation environment a mean penetration loss is assumed to provide good indoor coverage for the outdoor macro cells. The 15dB is assumed for dense urban [17, 15]. Cable and Connector Losses Cable and connector losses in the link budget attribute to the losses in RF jumper cable that connects Remote Radio Unit (RRUs) of the BTS to the antenna system. RF jumpers transmit signals between a BTS and an antenna system. The RF jumper comes with a standard length of 2m, 3m, 4m, 6m, and 10m. The 3m standard length is selected. Accordingly a cable loss of 0.5 dB is considered [16, 17]. Link Budget Calculations The fundamental principle remains the same as described in GSM link budget. However, in WCDMA radio networks the link budget calculations need to be done individually for voice and various data rates (e.g. 64 kbps, 128 kbps, 384 kbps & HSPA). This section presents the link budget templates for each radio bearer i.e. 12.2kbps speech, 64kbps data, 128kbps data, 384kbps data and HSPA. The data service link budgets are presented for Low Delay Constrained Data (LCD) services. It is assumed that the Unconstrained Delay Data (UDD) services have identical link budgets i.e. there is no difference between the packet switched and circuit switched radio bearer link budgets Maximum Allowed Path Loss (MAPL) The Maximum Allowed Path Loss (MAPL) for each service and in every environment can be computed for uplink and downlink. Uplink and downlink load factors are assumed based upon the traffic density expectation. The transmit EIRP and receiver sensitivities are calculated as the net gains and losses of the radio link. The uplink is generally the limiting link in terms of radio bearer coverage. Nevertheless the downlink is checked to verify this assumption. Required number of NodeB sites As it is shown on the Table 3.10, PS384K has the smallest coverage. Whereas AMR 12.2 (voice service) and HSUPA service have the largest coverage. In this regard the limiting service is PS384K but it is more conservative to ensure continues voice coverage in this 42

design. Hence, the site density is calculated based on the cell radius for voice service. The cell radius becomes 0.62 km, Thus the coverage area of a single NodeB using Equation 3.11 would be 1.95(0.62km*0.62km) = 0.74958 km2. More over from the perspective of dimensioning, a total of 16 NodeB’s are computed from Equation 3.12 to provide enough 3G services.

Receiver characteristics

Transmitter characteristics

Table 3.10: Calculated 3G Link Budget Parameters Services Bit rate Chip rate Target load TX power Antenna gain Body loss Feeder loss EIRP Thermal noise spectral density Thermal noise power Receiver noise figure Effective noise power Processing gain Required Eb/No Receiver sensitivity Interference margin Antenna gain Feeder loss Body loss Power control headroom Standard deviation Soft handover gain Shadow fading margin Penetration loss Max allowable outdoor path loss Outdoor coverage radius Frequency Band Max allowable indoor path loss Indoor coverage radius

Units CS12.2 Kbps 12.20 Mchip/s 3.84 % 50 dBm 21 dBi 0 dB 3 dB 0 dBm 18 dBm/Hz -174

CS64K 64.00 3.84 50 21 0 0 0 21 -174

PS64K 64.00 3.84 50 21 0 0 0 21 -174

PS128K 128.00 3.84 50 21 0 0 0 21 -174

PS384K 384.00 3.84 50 21 0 0 0 21 -174

HSUPA 200.00 3.84 50 24 2 0 0 26 -174

dBm dB dBm dB dB dBm dB dBi dB dB dB

-108.16 2.10 -106.06 24.98 4.20 -126.84 3.01 18 0.5 0 2

-108.16 2.10 -106.06 17.78 2.70 -121.14 3.01 18 0.5 0 2

-108.16 2.10 -106.06 17.78 1.60 -122.24 3.01 18 0.5 0 2

-108.16 2.10 -106.06 14.77 1.10 -119.73 3.01 18 0.5 0 2

-108.16 2.10 -106.06 10 0.60 -115.46 3.01 18 0.5 0 2

-108.16 2.10 -106.06 -13 -119.06 -119.06 3.01 18 0.5 0 2

dB dB dB dB dB

10 4 16.4 15 144.93

10 4 16.4 15 142.23

10 4 16.4 15 143.33

10 4 16.4 15 140.82

10 4 16.4 15 136.55

10 4 16.4 15 145.15

Km

1.67

1.4

1.5

1.28

0.97

1.694

MHz dB

1950 129.93

1950 127.23

1950 128.33

1950 125.82

1950 121.55

1950 130.15

Km

0.62

0.53

0.56

0.48

0.36

0.64

43

3.4.2 WCDAM Capacity Planning The purpose of capacity dimensioning is to estimate the approximate base station number needed from the capacity perspective. Similar to the link budget, the capacity estimation should be performed from the uplink and downlink based on the traffic model and service traffic demand. Individual voice traffic intensity for UMTS is similar with GSM, and based on Equation 3.13. 𝐴𝑣 =

2 𝑐𝑎𝑙𝑙𝑠 𝑠𝑒𝑐𝑜𝑛𝑑𝑠 × 45 3600 𝑠𝑒𝑐𝑜𝑛𝑑𝑠 𝑐𝑎𝑙𝑙

𝐴𝑣 = 0.025𝑒𝑟𝑙𝑎𝑛𝑔 𝑜𝑟 25𝑚𝑖𝑙𝑙𝑖 𝑒𝑟𝑙𝑎𝑛𝑔 UMTS Traffic Model Now the traffic model consist of 25mErl per subscribers is considered for voice traffic, 10GB per month per user for dongle (load heavy user), 1GB per month per for Smart phone user and also the following assumptions are used as our dimensioning requirement: 

% of daily traffic at busy hour is 10% and down link ratio 70%



Active users is assumed to be 70%



From the total data users the average load heavy user are assumed to be 30% and 70% is set for smart phone user.



This traffic per user includes normal traffic, signaling traffic and additional soft handover traffic.

The data rate per user at a busy hour is calculated based 𝑑𝑎𝑡𝑎 𝑣𝑜𝑙𝑢𝑚𝑒 𝑝𝑒𝑟 𝑚𝑜𝑛𝑡ℎ 𝑝𝑒𝑟 𝑢𝑠𝑒𝑟@𝐵𝐻(𝑘𝑏𝑝𝑠) =

𝑑𝑎𝑡𝑎 𝑣𝑜𝑙𝑢𝑚𝑒 𝑝𝑒𝑟 𝑚𝑜𝑛𝑡ℎ 𝑝𝑒𝑟 𝑢𝑠𝑒𝑟[𝐺𝐵𝑦𝑡𝑒] × 𝐵𝑢𝑠𝑦 ℎ𝑜𝑢𝑟 𝑟𝑎𝑡𝑖𝑜[%] × 1024 × 1024 30 𝑑𝑎𝑦𝑠 × 3600 𝑠𝑒𝑐

3.24

The uplink to downlink data ratio is 70% and selected based on the network statistics. At last, the data throughput per user calculated based on Equation 3.25 and 3.26 for uplink and downlink: 𝐷𝐿 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑝𝑒𝑟 𝑢𝑠𝑒𝑟@𝐵𝐻[𝑘𝑏𝑝𝑠] = 𝑎𝑐𝑡𝑖𝑣𝑒 𝑢𝑠𝑒𝑟 𝑑𝑎𝑡𝑎 𝑣𝑜𝑙𝑢𝑚𝑒 𝑝𝑒𝑟 𝑢𝑠𝑒𝑟@𝐵𝐻[𝑘𝑏𝑝𝑠] × 70%

3.25

𝑈𝐿 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑝𝑒𝑟 𝑢𝑠𝑒𝑟@𝐵𝐻[𝑘𝑏𝑝𝑠] = 𝑎𝑐𝑡𝑖𝑣𝑒 𝑢𝑠𝑒𝑟 𝑑𝑎𝑡𝑎 𝑣𝑜𝑙𝑢𝑚𝑒 𝑝𝑒𝑟 𝑢𝑠𝑒𝑟@𝐵𝐻[𝑘𝑏𝑝𝑠] × 30%

3.26

HSDPA and HSUPA user per carrier calculation To calculate the total subscriber, an active user ratio of 70% is considered. According to orange telecom laboratory test, the HSDPA average throughput per cell is 3.6Mbps, and 44

HSUPA average throughput per cell is 1.9Mbps. Therefore, the supported subscribers per cell results as follow: Table 3.11: Throughput per user at busy hour calculation Type

Data Dongle 10

Data per month (GB) Proportion (%) Data Volume per month per user Data Volume per month per user (GB) Busy hour ratio (%) Data Volume per month per user @BH (kbps) Active user ratio (%)

Data SP 1

30% 3

70% 0.7 3.7 10% 29.43 70%

Active user Data volume per user @BH(kbps)

29.43

DL/UL Ratio DL Throughput per user @BH (kbps) UL Throughput per user @BH (kbps)

70% 20.6 8.83

Table 3.12: UMTS Cell Load Dimension Result Bear

Cell Load

Traffic per subs

Active Subs/Cell

Total Subs/Cell

HSDPA+Voice

3.6 Mbps

5.97 kbps

603

862

HSUPA+Voice

1.9 Mbps

3.85 kbps

494

706

Therefore, an active subscriber per cell would support 1097 subscribers for DL and UL FDD mode communication, the total population had been forecasted for the next three years with rate of 0.25. Finally the required number of cells for our case becomes: 𝑅𝑒𝑞𝑢𝑖𝑟𝑒𝑑 𝑛𝑜. 𝑜𝑓 𝑐𝑒𝑙𝑙𝑠 =

𝑇𝑜𝑡𝑎𝑙 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑇𝑜𝑡𝑎𝑙 𝑛𝑜. 𝑜𝑓 𝑎𝑐𝑡𝑖𝑣𝑒 𝑢𝑠𝑒𝑟 𝑝𝑒𝑟 𝑐𝑒𝑙𝑙 =

3.27

31360 = 29 1097

As a result of capacity planning a total of ten sites or tri sector are required. Therefore the required sites number for a specific area should be chosen to be the maximum number of sites obtained from coverage and capacity planning calculations to satisfy the traffic requirements of both coverage and capacity so according to the results of Table 3.13 the sixteen sites are selected. Table 3.13: UMTS Best site selection of coverage and capacity Planning Coverage

Number of sites calculated 16 sites

Capacity

10 sites

Best selection 16 sites

45

3.5 LTE Network Planning The third multi RAT network consists evaluation of downlink and uplink radio link budgets. The maximum path loss is calculated based on the required SINR level at the receiver, taking into account the extent of the interference caused by traffic. The minimum of the maximum path losses in UL and DL directions is converted into cell radius, by using a propagation model appropriate to the deployment area. Capacity planning gives an estimate of the resources needed for supporting a specified offered traffic with a certain level of QoS (e.g. throughput or blocking probability). Theoretical capacity of the network is limited by the number of base stations-eNodeB installed in the network. Cell capacity in LTE is impacted by several factors, which includes interference level, packet scheduler implementation and supported modulation and coding schemes [7, 8]. 3.5.1 LTE Coverage Planning The link budget calculations estimate the maximum allowed signal attenuation, called path loss, between the mobile and the base station antenna. The maximum path loss allows the maximum cell range to be estimated with a suitable propagation model, such as Cost231–Hata model. The cell range gives the number of base station sites required to cover the target geographical area. The link budget calculation can also be used to compare the relative coverage of the different systems. Dimensioning procedure Link budget and coverage planning is calculated for both cases "UL & DL". The procedure steps are [4]: 1. Calculate the Max Allowed Path Loss (MAPL) for DL and UL. 2. Calculate DL & UL cell radiuses by using the propagation model & MAPL. 3. Determine the appropriate cell radius 4. Calculate the site coverage area and the required sites number. LTE Link budget Parameters The parameters can be grouped as propagation related, equipment related parameters, and LTE specific parameters. Propagation related parameters include the penetration loss, body loss, feeder loss, and background noise. The equipment related parameters are the specification given by the manufacturer such as transmitter power, receiver sensitivity, and antenna gain. The LTE specific parameters include interference margin, fast fading margin, edge coverage rate and MIMO type used. In this section the main parameters are discussed.

46

eNodeB Maximum Total Transmitter Power refers to the transmit power value per transmitting (TX) path. Typical value is either 43dBm (20W) or 46dBm (40W) [7]. eNodeB Antenna Gain depends on the clutter type and coverage requirement. The low gain antenna (15–17 dBi) can be used in dense urban and urban clutters while a high gain antenna (18–20 dBi) can be used in rural areas and highways to extend the RF coverage. UE Maximum Total Transmitter Power refers to the UE transmit power which depends on the power class of the UE. Currently only one power class is defined in 3GPP TS 36.101, class 3 with maximum transmitter power of 23dBm. UE Antenna Gain based on the specifications of 3GPP, UE(s) are assumed to have an integral antenna only with a gain of 0 dBi for each antenna port(s) [21]. Body Loss is a loss generated due to signal blocking and absorption, when UE antenna is close to the body of the user. For UE such as USB dongle, Wi-Fi device, and an LTE fixed router the position of the antenna is far from the user’s body, thus the body loss is ignored. Whereas for mobile terminals (smart phones for instant) they are close to the user’s body and the body loss must be considered in the link budget, has a typical value of 3dB. Feeder Loss considers the losses due to RF feeder, RF jumpers and connectors in the path between the antenna and the eNodeB. In a distributed eNodeB, the baseband unit and RF unit (Remote Radio Units (RRU)) are separated, only the loss of the jumper between the RRU and antenna is considered with typical value of 0.5dB loss. On the other hand, if the eNodeB is not a distributed type, feeder losses may be 3dB or more according to the feeder and its connector’s characteristics [20]. Equivalent Isotropic Radiated Power (EIRP) indicates the power that would be radiated by the theoretical isotropic antenna to achieve the peak power density observed in the direction of maximum antenna gain. In LTE the EIRP in the DL is calculated based on the total number of RB due to the OFDMA while in the UL the allocated RB are only used due to the SCFDMA (3 RBs). The EIRP per subcarrier in the DL and UL are calculated as follows [7]: 𝑆𝐶 𝐸𝐼𝑅𝑃𝐷𝐿 = 𝑃𝑒𝑁𝐵(𝑠𝑐) + 𝐴𝐺𝑒𝑁𝐵 − 𝐹𝐿 + 𝑀𝐺

3.28

𝑆𝐶 𝐸𝐼𝑅𝑃𝑈𝐿 = 𝑃𝑈𝐸(𝑠𝑐) + 𝐴𝐺𝑈𝐸 − 𝐵𝐿

3.29

Where 𝑃𝑒𝑁𝐵(𝑠𝑐) and 𝑃𝑈𝐸(𝑠𝑐) are the power per subcarrier in the DL and UL respectively, 𝐴𝐺𝑒𝑁𝐵 is the eNodeB antenna gain, 𝐹𝐿 is feeder loss, 𝑀𝐺 is the MIMO gain, 𝐴𝐺𝑈𝐸 is the UE antenna gain, and the 𝐵𝐿 is the body loss. 47

Cell Edge User Throughput is the minimum net single UE target throughput requirement to be achieved service at the cell edge. Accordingly it can limit the minimum Modulation and Coding Scheme (MCS) to be used. This parameter usually provided by the network operator based on the required services at cell edge. A typical value for UL can be 512 kbps to 1 Mbps where as in the DL it can be 1 Mbps up to 4 Mbps. Signal to Interference Noise Ratio (SINR) is the threshold of the receiver that can demodulate the signal for the UL and it is related to the MCS for the DL. SINR values are obtained from the system level simulation result and it depends on the receiver design. Thus SINR is a vendor specific parameter. Noise Figure is a key factor to measure the receiver performance. It is defined as the ratio of the input SINR at the input end to the output SINR at the output end of the receiver. The noise figure depends on the bandwidth and the eNodeB capability. A typical value for the noise figure is between 6 to 8 dB [15]. Receiver Sensitivity determines the signal level or threshold at which the RF signal can be detected with a certain quality. The receiver sensitivity per subcarrier can be calculated as follows [15]: 𝑅𝑥𝑠(𝑠𝑐) = 𝑆𝐼𝑁𝑅 + 𝑁𝐹 + 𝑁𝑃 + 10 log(𝑆𝐶)

3.30

Where SINR is the threshold of the receiver that can demodulate the signal, NP is the density of the thermal white noise power which is -174dBm/Hz, SC is the subcarrier and it is 15 KHz in LTE, and NF is the noise figure in dB. Penetration Loss indicates the fading of radio signals due to building obstruction from an indoor terminal to the eNodeB and vice versa. It depends on the nature of the buildings and the clutter type of the targeted coverage area. Table 3.14: Penetration losses and Standard deviation of slow fading typical dense urban. Clutter type

Penetration loss range (dB)

Dense urban

19 – 25

Standard deviation of slow fading (dB) 10

Shadow Fading Margin indicates the fading due to obstruction like building. The effect of shadow fading minimized by ensuring a certain edge coverage probability, certain allowance is required. This allowance is called the “slow fading margin” or the “shadow fading margin”. The Standard deviation of slow fading is used to determine the slow fading difference and it

48

shows the distribution of the radio signal strength at different test points at similar distance from the transmitter [22]. Frequency Band: Band 3 (1800 MHz) is selected to use since the band is the most promising LTE band as it can be used for nationwide coverage with dense urban convergence. Reference Signal Received Power (RSRP) is used to measure the coverage of the LTE cell on the DL. The UE will send RRC measurements reports that include RSRP values in a binned format. The reporting range of RSRP is defined from -140 to -40dBm with 1dB resolution. The main purpose of RSRP is to determine the best cell on the DL radio interface and select this cell as the serving cell for either initial random access or intra-LTE handover. In this thesis, the RSRP is assumed to be greater than/or equal to -110dBm in order to have a better signal strength through the selected area. Cell-edge Coverage Probability is assumed to have a cell-edge coverage probability of 80%, and hence a fading margin of 7dB is used during link budget calculation. Area Coverage Probability tells how much the target area will be covered by the planned network. The assumed 80% coverage probability implies the target network would cover 80% of the selected area for the considered value of RSRP. Based on the above described the main link budget parameters and the common parameters, a MAPL would be calculated as the difference between the EIRP and the overall loss. Propagation model selection COST231-Hata model is selected in the calculation of the cell radius. 𝑃𝐿 = 46.3 + 33.9𝑙𝑜𝑔10(𝑓) − 13.82𝑙𝑜𝑔10(ℎ𝑏 ) − 𝑎ℎ𝑚 + [44.9 − 6.55𝑙𝑜𝑔10(ℎ𝑏 )]𝑙𝑜𝑔10(𝑑) + 𝑐𝑚

3.31

Where frequency 𝑓 = 1800𝐻𝑧, height of BTS ℎ𝑏 = 30𝑚, height of MS ℎ𝑟 = 1.5𝑚 d radius of the site, and 𝑐𝑚 is correction factor in this case it is 2. 

𝑎ℎ𝑚 = [1.11𝑙𝑜𝑔10(𝑓)]ℎ𝑟 − [1.5𝑙𝑜𝑔10(𝑓) − 0.8] = 1.11 log(1800) 1.5 − 1.5 log(1800) = −0.8287



[44.9 − 6.55 log(ℎ𝑏)] log(𝑑) + 𝑐𝑚 = [44.9 − 6.55 log(30)] log(𝑑) + 2 35.225 log(𝑑) + 2 = 46.3 + 33.9𝑙𝑜𝑔10(𝑓) − 13.82𝑙𝑜𝑔10(ℎ𝑏 ) = 46.3 + 33.9 log(1800) − 13.82 log(30) = 136.240 49

𝑃𝐿 = 136.240 − 0.287 + 35.225 log(𝑑) + 2 = 137.953 + 35.225 log(𝑑) 𝑑 = 10

(𝑃𝐿−137.953) 35.225

3.32 Table 3.15: Uplink link budget parameters Unit

General parameter Morphology Duplex mode Data channel type System bandwidth User Environment MIMO scheme Cell edge user throughput(kbps) Transmitter side Tx power (dBm) Tx antenna gain (dBi) Cable loss (dB) EIRP (dBm ) Receiver side – UE eNode B noise figure

Formula

Typical value DU FDD PUSCH 15 Indoor 1x1 384

MHz

Kbps

dBm dBi dB dBm

A B C E =A+B-C

23 0 0 23

dB

G

2.2

Feeder loss

dB

H

0.5

Thermal noise

dBm

I

−174

Noise power per subcarrier Required SINR (dB)

dBm dB

J= I+𝟏𝟎 𝐥𝐨𝐠(𝑩𝑾 ∗ 𝟏𝟎𝟎𝟎) K

-132.17 2.17

L = 𝟏𝟎𝒍𝒐𝒈(PRB per TTI*12)

17.8

Number of received subcarrier Receiver sensitivity eNode B Rx Gain

dBm dBi

M= K+G+L+J N

-110 18

Cable loss MAPL= E - M + N - O = 148

dB

O

3

Therefore Table 3.15 and Table 3.16 shows that the maximum allowable path loss has been computed for uplink and downlink case with result of 148 and 161.73 respectively. As a result of this the MAPL in the DL path is greater this points out that the area covered by the eNodeB antenna radiation is more than the area covered by the UE antenna. Considering Table 3.17 the radius of the single site and the required number of sites are calculated using Equation 3.32 and Equation 3.33.

50

Table 3.16: Downlink Link budget parameters Unit General parameter Morphology Duplex mode Data channel type System bandwidth User Environment MIMO scheme Cell edge user throughput(kbps) Transmitter side Tx power (dBm) Tx antenna gain (dBi) Cable loss (dB) TMA insertion loss(dB) Total Tx power increase(dB) EIRP (dBm ) Receiver side UE noise figure Thermal noise Thermal noise density Required SINR

Typical value DU FDD PDSCH 15 Indoor 4x4 1024

MHz

Kbps

dBm dBi dB dB dB dBm

A B C D E F=A+B-C-D

46 17.5 3 0.5 3 63

dB dBm

G H=10𝐥𝐨𝐠(𝒌 ∗ 𝑻𝒐 ∗ 𝟏𝟎𝟎𝟎)

7 -174

dBm dB

I= H+𝟏𝟎 𝐥𝐨𝐠(𝑩𝑾 ∗ 𝟏𝟎𝟎𝟎) J

-132.17 -3.06

K = 𝟏𝟎𝒍𝒐𝒈(PRB per TTI*12)

29.5

L= J+G+K+I M N

-98.73 0 0

Number of received subcarrier Receiver sensitivity Rx antenna gain Body loss MAPL= F-L+M-N = 161.73

Formula

dBm dBi dBm

Table 3.17: Clutter parameters Parameter Interference margin (dB) Standard deviation Edge coverage probability Shadow Fading margin (dB) Gain against shadowing

Value 3 10 77% 7 2.8

Average Penetration loss (dB)

20

𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝐴𝑙𝑙𝑜𝑤𝑎𝑏𝑙𝑒 𝑃𝑎𝑡ℎ 𝐿𝑜𝑠𝑠(𝑀𝑃𝐴𝐿) = 𝐿𝑀𝐴𝑋 − 𝑠𝑢𝑚 𝑜𝑓 𝑐𝑙𝑢𝑡𝑡𝑒𝑟 𝑙𝑜𝑠𝑠𝑒𝑠 + 𝑔𝑎𝑖𝑛 𝑎𝑔𝑎𝑖𝑛𝑠𝑡 𝑠ℎ𝑎𝑑𝑜𝑤𝑖𝑛𝑔 = 148 − 30 + 2.8 = 120.8 51

Therefore the MAPL of the uplink is equal with 120.8dB. From this it is easy to identify the number of sites. (120.8−137.953) 35.225

𝑑 = 10

= 0.3258

The coverage area of the eNodeBs (i.e. = 1.95𝑅 2 = 1.95(.3258)2 = 0.207𝑘𝑚2 ) 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑒𝑁𝐵𝑠 = =

𝑡𝑜𝑡𝑎𝑙 𝑐𝑜𝑣𝑒𝑟𝑎𝑔𝑒 𝑎𝑟𝑒𝑎 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑠𝑖𝑛𝑔𝑙𝑒 𝑒𝑁𝐵𝑠

3.33

11.99 𝑘𝑚2 = 57.9 0.207𝑘𝑚2

Finally an approximately value a total of 58 sites are required. 3.5.2 LTE Capacity Planning Capacity planning calculates the average cell throughput and average number of subscribers that can be supported for a given cell range. A simple approach is done on the results obtained from both SINR-based and the traffic demand analysis to get the average cell throughput. Calculation of LTE Data Rate Factors that affect peak data rate are system bandwidth, multiplexing technique, modulation and coding schemes and UE category. Available bandwidth: 10% of the given bandwidth is used as a guard band, therefore the effective bandwidth would be 90%. Carrier spacing in LTE is 15 kHz and there are 12 subcarriers in a resource block. From this, bandwidth of one resource block will be 15 kHz * 12 = 180 kHz and the number of resource blocks can be calculated as: No of resource blocks =

𝐸𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 𝑏𝑎𝑛𝑑𝑤𝑖𝑑𝑡ℎ 180𝑘𝐻𝑧

3.34

For LTE system the numbers of resource blocks are listed in Table 3.12. The PRB is calculated by dividing the effective bandwidth to the bandwidth of single RB, like the effective 9𝑀𝐻𝑧

bandwidth of 10MHz is 9MHz so 𝑃𝑅𝐵 = 180𝐾𝐻𝑧 = 50. Table 3.18: List of PRBs. System bandwidth (MHz)

1.4

3

5

10

Sub-carrier bandwidth (kHz)

15

Physical resource block (PRB) bandwidth (kHz)

180

Number of available PRBs

6

12

25

50

15

75

20

100

The daily traffic can be estimated as a percentage of the busy hour traffic. In this thesis it is assume that the busy hour traffic is 10% of the daily traffic. Three types of service packages 52

are provided, golden, silver and bronze service package, each service has its own quality. The month service package, the DL and UL peak data rates, and the package percentage, all of these characteristics are shown in Table 3.20. The traffic ratio of the UL and DL in terms of the total traffic is chosen to be 20% for UL and 80% for DL. The number of subscribers must be specified in order to continue the analysis, the population number is considered to be 83,625 and 15% (i.e. 12544) of them are assumed as LTE subscribers. Table 3.19: LTE Users Category Package Type Gold Silver Bronze

Month service package (GB) 20 15 10

Package Percentage 10% 40% 50%

Total average throughput per subscriber must be calculated in order to calculate the average throughput per site. 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑝𝑒𝑟 𝑠𝑢𝑏𝑈𝐿+𝐷𝐿 (𝑘𝑏𝑝𝑠) 8𝑏𝑖𝑡 ∗ 𝐵𝐻 𝑟𝑎𝑡𝑖𝑜 𝑏𝑦𝑡𝑒 𝑁𝑜 𝑜𝑓 𝑑𝑎𝑦𝑠 ∗ 𝑡𝑖𝑚𝑒 𝑖𝑛 𝑠𝑒𝑐𝑜𝑛𝑑

𝑀𝑜𝑛𝑡ℎ𝑙𝑦 𝑠𝑒𝑟𝑣𝑖𝑐𝑒 𝑝𝑎𝑐𝑘𝑎𝑔𝑒 ∗ =

3.35

Table 3.20: Total average throughput per subscriber at busy hour Package type

Month service package (GB)

Package Traffic ratio percentage UL DL

Average throughput per subscriber (UL+DL)

Total average throughput (kbps)

Total average throughput per subscriber for (kbps) DL 77.03 77.03 77.03

Golden Silver

20 15

10% 40%

20% 20%

80% 80%

148.15 111.11

14.815 44.444

UL 19.25 19.25

Bronze

10

50%

20%

80%

74.074

37.037

19.25

Total average throughput per subscriber at busy hour (kbps): ∑ (𝐴𝑣𝑒 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡𝑝𝑒𝑟 𝑠𝑢𝑏 @𝐵𝐻𝑈𝐿+𝐷𝐿 ∗ packet percentage)

3.36

= 14.815 + 44.444 + 37.037 = 96.296 Data rate calculation of FDD-DL The peak capacity throughput per sector calculated considering a 15 MHz LTE system with 4×4 MIMO and 1×2 SISO configuration in downlink and uplink sides respectively. 64QAM 4

and 6 code rate: firstly the number of resource elements (RE) in a sub frame is being calculated. 53



1RB = 12 sub-carriers



1 sub-frame = 1ms



1time slot = 7 OFDM symbols (when normal CP length is used)



1 OFDM symbol = 6 data bits if 64 QAM is used as modulation scheme in downlink side and 4 data bits if 16 QAM is used in the uplink side. No of bits in sub − frame = No of RBs ∗ No of sub − carriers in a RB ∗ No of slots in a sub frame ∗ No of modulation symbols in a slot ∗ No of data bits in 1 OFDM symbol

𝑓𝑜𝑟 𝐷𝐿 75 ∗ 12 ∗ 2 ∗ 7 ∗ 6 = 75,600𝑏𝑖𝑡𝑠 𝑓𝑜𝑟 𝑈𝐿 75 ∗ 12 ∗ 2 ∗ 7 ∗ 4 = 50,400𝑏𝑖𝑡𝑠 Data rate =

𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑖𝑡𝑠 𝑖𝑛 𝑠𝑢𝑏𝑓𝑟𝑎𝑚𝑒 𝐷𝑢𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 1 𝑠𝑢𝑏𝑓𝑟𝑎𝑚𝑒

𝐷𝑎𝑡𝑎 𝑟𝑎𝑡𝑒(𝐷𝐿) = 𝐷𝑎𝑡𝑎 𝑟𝑎𝑡𝑒(𝐷𝐿) =

75600 1𝑚𝑠 50400 1𝑚𝑠

∗ No of antennas ∗ coding rate

∗ 4 ∗ 4/6 = 201.6𝑀𝑏𝑝𝑠 3

∗ 1 ∗ 4 = 37.8𝑀𝑏𝑝𝑠

Peak Throughput per sector calculation for DL communication: The data rate in the down link is 151.2Mbps, but there are a lot of overheads related to control signaling to be subtracted from. 

Pilot overhead (4 TX antennas) = 14.29%



Common channel overhead (adequate to serve 1 UE/sub frame) = 10%



CP overhead = 6.66%



Guard band overhead = 10%

The total DL overhead for the 15 MHz channel is 14.29% + 10% + 6.66%+ 10% = 40.95%. The remaining percentage would be (100% - 40.95%=59.5). Thus 𝑇ℎ𝑒 𝑝𝑒𝑎𝑘 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡𝐷𝐿 = 0.595 ∗ 151.2 ∗ 106 = 119.95Mbps Peak Throughput per sector for UL communication calculation: Similarly, the overheads related to the control signaling would be subtracted using a single transmission antenna.  Pilot overhead = 14.3%  Random access overhead = 0.625%  CP overhead = 6.66%  Guard band overhead = 10% The total UL overhead for the 15 MHz channel is 14.3% + 0.625% + 6.66%+ 10% = 31.585%. So 68.4 % of the peak data rate in the uplink becomes: 54

Peak data rate UL = 0.684*37.8*106 = 25.85 Mbps Peak throughput per site calculation for UL and DL: 𝑇𝑜𝑡𝑎𝑙 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑝𝑒𝑟 𝑠𝑖𝑡𝐷𝐿 = 3 ∗ 𝐷𝐿 𝑑𝑎𝑡𝑎 𝑟𝑎𝑡𝑒 𝑝𝑒𝑟 𝑠𝑒𝑐𝑡𝑜𝑟 = 3 ∗ 119.95𝑀𝑏𝑝𝑠 = 359.85𝑀𝑏𝑝𝑠 𝑇𝑜𝑡𝑎𝑙 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑝𝑒𝑟 𝑠𝑖𝑡𝑈𝐿 = 3 ∗ 𝑈𝐿 𝑑𝑎𝑡𝑎 𝑟𝑎𝑡𝑒 𝑝𝑒𝑟 𝑠𝑒𝑐𝑡𝑜𝑟 = 3 ∗ 25.85𝑀𝑏𝑝𝑠 = 77.56 𝑀𝑏𝑝𝑠 Having the total throughput per sit in both UL and DL will help us to calculate the total number of subscriber per site: Max. Number of subscriber per site 𝑇𝑜𝑡𝑎𝑙 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑝𝑒𝑟 𝑠𝑖𝑡 = 3.37 𝑇𝑜𝑡𝑎𝑙 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑡ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 𝑝𝑒𝑟 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 359.85𝑀𝑏𝑝𝑠 𝑀𝑎𝑥. 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 𝑝𝑒𝑟 𝑠𝑖𝑡𝑒𝐷𝐿 = = 4672 𝑠𝑢𝑏 77.03𝑘𝑏𝑝𝑠 77.56𝑀𝑏𝑝𝑠 𝑀𝑎𝑥. 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 𝑝𝑒𝑟 𝑠𝑖𝑡𝑒𝑈𝐿 = = 4029 𝑠𝑢𝑏 19.25𝑘𝑏𝑝𝑠 Since the total number of subscribers are considered to be 12,544 subscribers. Assuming the population growth rate 4% and penetration rate of 15% a total of 13046 subscribers are expected for LTE service usage. So the number of sites in the coverage area becomes: 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓𝑠𝑖𝑡𝑒𝑠 𝐷𝐿 = =

𝑇𝑜𝑡𝑎𝑙 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 𝑛𝑢𝑚𝑏𝑒𝑟 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑟𝑒𝑞𝑢𝑖𝑟𝑒𝑑 𝑎𝑟𝑒𝑎 𝑀𝑎𝑥. 𝑛𝑜. 𝑜𝑓 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 𝑝𝑒𝑟 𝑠𝑖𝑡𝑒

13046 =3 4672

𝑇𝑜𝑡𝑎𝑙 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 𝑛𝑢𝑚𝑏𝑒𝑟 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑟𝑒𝑞𝑢𝑖𝑟𝑒𝑑 𝑎𝑟𝑒𝑎 𝑀𝑎𝑥. 𝑛𝑜. 𝑜𝑓 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑏𝑒𝑟 𝑝𝑒𝑟 𝑠𝑖𝑡𝑒 13046 = =4 4029 Table 3.21: The total amount of sites 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓𝑠𝑖𝑡𝑒𝑠 𝑢𝑙 =

Type of analysis

Total number of sites

Best selection

Coverage analysis Capacity analysis

58 sites 4 sites

58 sites

The Table 3.21 shows that the sites in coverage analysis are greater than the sites in the capacity analysis so that 58 sites are going to be selected.

55

CHAPTER FOUR SIMULATION RESULTS AND DISCUSSION 4.1 Introduction An atoll simulation software is going to be used to find out the analytical results from coverage and capacity dimensioning. In this thesis the simulation environment is used to investigate the RAN nominal planning of GSM/UMTS/LTE networks. The cell planning tools require a digital map which is a geographical map of the planned area. It is the most basic and initially required tool for RNP since it contains information about, the land usage, the height of obstacles and vector data. A digital map with topographic data, morphographic data and building location and height data, in the form of raster maps is the minimum requirement.

Figure 4.1: Digital Map of Mekelle City Computational Zone It is a special zone where the intended calculations are being carries out. It also defines the area where the coverage prediction results will be displayed. Therefore it allows to restrict the coverage prediction result to the part of the network aimed to work on. so in the case of this thesis the computational zone which incorporate Kedamay Weyane and Adi Haqi is shown in Figure 4.2.

56

Figure 4.2: Computational Zone of Kedamay Weyane and Adi Haqi in ATOLL 4.1 Modeling and Analysis of the Multi RAT Network The multiple radio access networks consist of GSM, UMTS and LTE is illustrated on the Atoll simulation software window.

Figure 4.3: Model of the Multi RAT Radio Network The computational zone of the digital map as shown in the Figure 4.3 depict that of the three sector yellow colored base stations, the red colored NodeBs and the ----- colored eNodeB are generally belongs to Multi RAT technologies of GSM, UMTS and LTE respectively.

57

Analysis and Interpretation of the Simulation Results In this section manipulating the performance, coverage predictions have been performed using parameters like Signal level, Throughput, Overlapping Zones and Carrier to Interference plus Noise Ratio (CINR). 4.1.1 Performance Evaluation of Planned GSM Network a. GSM Coverage Prediction by Signal level The prediction by signal level show in Figure 4.4 (a) tells the prediction of the best signal strength at each pixel within the computation zone. The signal prediction result has an acceptable coverage from the simulation result.

a)

b)

Figure 4.4: a) GSM Coverage Prediction by Signal level b) Histogram The histogram result shown in Figure 4.4 (b) depicts more areas within the computation area are covered by strong signal level and its signal values are ranges from -65dBm to -70dBm. Therefore it has come to be greater than the receiver sensitivity. Other areas are also covered by acceptable signal level comparing design receiver sensitivity signal level -105dBm. 58

b. Coverage Prediction on Overlapping Zones

The Figure 4.5 composed of pixels that are covered by the signal of at least two transmitters. The coverage is calculated for the cell with the highest power for a transmitter with more than one cell.

a)

b) Figure 4.5: a) GSM Overlapping zone of transmitter b) Histogram The Figure 4.6 show that the 76.009% of the area are served by single cell without need of hand off stuffing. The rest would be 19.41%, 4.33%and 0.251% served with more than two cells. As result of this 100% of the area becomes to be served by at least a single transmitter. c. Coverage prediction by C/ (I+N) Level This parameter is used to analyze the signal quality. The prediction are based on the predicted signal level from the best server and the predicted signal levels from other cells (interference) at each pixel.

59

The simulator calculates the serving transmitter for each pixel based on the downlink reference signal level. The serving transmitter is determined according to the received reference signal level from the cell with the highest power. If more than one cell cover the pixel, the one with the lowest order is selected as the serving cell. Then, depending on the prediction definition, it calculates the interference from other cells, and finally calculates the 𝐶 ⁄(𝐼 + 𝑁). If the 𝐶 ⁄(𝐼 + 𝑁) is higher than 𝐶 ⁄(𝐼 + 𝑁) threshold the pixel is colored. The Carrier to Interference plus Noise Ratio (CINR) is the ratio of the signal carrier best servings for the intervention seemed at all other sites/sectors, plus all the noise. If a signal goes below the level of noise, it cannot be decoded and no useful information can be recovered from it.

a)

b) Figure 4.6: a) GSM Coverage by C/ (I+N) Level and b) Histogram The histogram result shows the comparison of coverage percentage area versus 𝐂⁄(𝐈 + 𝐍) level value. The 98.07% of the areas within the computation zone are covered by 15dB to 21.2dB. Which satisfies the standard value. 60

d. Prediction by Effective Service Area Analysis The effective service area prediction calculates where a service actually is available for the probe mobile. The Figure 4.8 shown below refers to the colored area are where an effective service is being provided.

Figure 4.7: GSM Effective Service Area Analysis 4.1.2 Performance Evaluation of Planned UMTS Network a. Coverage Prediction by Signal level The prediction of the best signal strength at each edge within the designed dense urban computational zone is shown in the Figure 4.9 (a).

a) 61

b) Figure 4.8: a) UMTS Coverage prediction by Signal level b) Histogram The histogram statistical result shows in Figure 4.9 (b) 65.405% of the areas within the computation zone are covered by -100dBm to -90dBm strong signal level which is better as compared to the calculated receiver sensitivity signal value -120dBm. The remaining percentage area also covers by 1.36%, 2.907% and 30.328% of signal level intervals [-110, 100], [-90,-80] and [-80,-70] respectively. Therefore a total of the computational zone is covered with greater signal levels in which the receiver can just sense it. b. Coverage Prediction on Overlapping Zones The Figures 4.10 (a) depicts that the coverage prediction of the overlapping zones of UMTS and The histogram illustrate in Figure 4.10 (b) the 63.17% of the computational zone are served by a single server where as 27.02%, 8.45%, 1.36% are being served with two, three and four servers respectively. So that the total area or 100% of the area are served by at least a single server.

a) 62

b) Figure 4.9: a) UMTS Overlapping zone of transmitter b) Histogram c. Coverage Prediction by Total Noise Level Analysis

a) UMTS Total Noise Level Analysis

b) Histogram Figure 4.10: a) UMTS Total Noise Level Analysis b) Histogram 63

The Figures 4.11 (b) depict that the 98.33 percentage of the area has a maximum noise level of -80 to -70dBm rages which is 10-11 to 10-10 watt. The rest 1.67% has a noise level of -90 to 80dBm. So that the noise level is much smaller as compared to the transmitted power. d. Prediction by Effective Service Area Analysis The effective service area is the intersection zone between the pilot reception area, and the uplink and downlink service areas. In other words, the effective service area prediction calculates where a service actually is available for the probe mobile. The Figure 4.12 shown below refers to the colored area where an effective service is being provided.

Figure 4.11: UMTS Effective Service Area Analysis 4.1.3 Performance Evaluation of Planned LTE Network a. LTE Coverage Prediction based on the signal level Coverage prediction by signal level that is shown in Figure 4.13 tells us the prediction of the best signal strength at each edge within the designed dense urban computation zone. The simulation signal level values are greater than our design receiver sensitivity signal value. This shows all the customers with in the selected coverage are can sense the signal from the sites, which means any customer can be served by being in every places of the considered area because. The designed receiver sensitivity was around -110dBm.

64

Figure 4.12: Coverage prediction by signal level in DL using histogram The above figure shows the 24.775% of customers with in the dense urban get signal streangth btween -80, -75, 40.552 % of the users get -75,-70 signal streangth, 8.893% of them get -85, -80 dBm, 23.436% of the customers get very strong signal than the others which is geater than -70dBm and less than 3% of the users get signal level between -95, -85dBm. Therefore from this all it is observed that about 97.656% of the users get very excellent signal level above 85dBm. b. Coverage prediction by overlapping zone Overlapping zones as shown on Figure 4.14 are composed of pixels, they are composed of pixels that are covered by the signal of at least two transmitters. For a transmitter with more than one cell, the coverage is calculated for the cell with the highest power.

Figure 4.13: Coverage prediction by overlapping zone c. Analyzing the Signal Quality (Coverage by C/ (I+N) Level The carrier to interference plus noise ratio (CINR) is the ratio of the signal carrier best servings for the intervention seemed at all other sites/sectors, plus all the noise. If a signal goes below the level of noise, it cannot be decoded and no useful information can be recovered from it. A 65

good signal is important for high data rate communications. Coverage prediction by C/(I+N) level calculates the co-channel interference as well as the adjacent channel interference, which is reduced by the adjacent channel suppression factor defined in the Frequency Bands table [20].

Figure 4: coverage analysis with c/(I + N) and its Histogram result The histogram result of Figure 4 shows the relationship between coverage percentage area and 𝐶/(𝐼+𝑁) value. As it is shown in the figure 75.18% of the coverage percentage has CINR in between 17 and 18 dB and 24.29% of the area owns CINR in between 18 and 19 dB. Therefore the 99.47% of the area within the computation zone are covered by 17 to 19 dB CINR Level value which satisfies the standard value. d. Coverage prediction by Throughput (DL) Downlink and uplink throughput coverage predictions calculate and display the channel throughputs and cell capacities based on C/ (I+N) and bearer calculations for each pixel. These coverage predictions can also display aggregate cell throughputs. The simulation result determines the total number of symbols in the downlink and the uplink sub frames from the input parameter tables. Then, Atoll determines the bearer at each pixel and multiplies the bearer efficiency by the number of symbols in the frame to determine the peak MAC channel throughputs. The cell capacity is equal to channel throughput when the maximum traffic load is set to 100%, and is equal to a throughput limited by the maximum allowed traffic loads otherwise. Cell capacities are, therefore, channel throughputs scaled down to respect the maximum traffic load limits. The per-user throughput in DL and UL is calculated by dividing the DL and UL cell capacity by the number of DL and UL users of the serving cell respectively, but in uplink, the per-user throughput is smaller than the DL per-user throughput. Finally the simulation result shows the DL individual user throughputs within the dense urban area is acceptable as shown in Figure 4.

66

Figure 4.14: Coverage prediction using throughput e. Prediction by Effective Service Area Analysis

Figure 4.15: LTE effective service analysis 4.2 Comparison of Multi RAT by Effective Service Analysis a. Comparison between UMTS and LTE The figure below show that the computational zone has been covered almost 100%. The red color take the most percentage of the area which are the pixels covered by the two radio access and the rest pixels are covered by each predictions.

67

Figure 4.16: Comparison of Effective Service Area Analysis by UMTS vs LTE b. Comparison between GSM and UMTS The pixels of the computational zone has been covered almost 100% as show figure 4.17. The red color take the most percentage of the area which are the pixels covered by the two radio access and the rest pixels are covered by each predictions.

Figure 4.17: Comparison of Effective Service Area Analysis by GSM vs UMTS c. Comparison between LTE and GSM Also the pixels of the computational zone has been covered almost 100% by these two radio access technologies. The red color as show in the figure 4.18 indicates areas which are the pixels covered by LTE and GSM, the rest pixels are covered by each radio access predictions.

68

Figure 4.18: Comparison of Effective Service Area Analysis by LTE vs GSM

69

CHAPTER FIVE CONCLUSIONS AND RECOMMENDATIONS

5.1

Conclusion

In a cellular network industry radio network planning shall compromise the important factors such as network coverage, capacity, and QoS. Furthermore, the radio network dimensioning parameters have an impact on each other and therefore it is important to decide the emphasis in order to get an optimal dimensioning result within the agreed parameter ranges. Based on system, transmitter and receiver parameters link budget has been determined. Since it the initial point for the coverage dimensioning. The essential parameters for capacity planning start with estimated traffic, average antenna height, and frequency usage of the multi radio access technologies. This thesis was done to develop a multiple radio network for Kedamayweyane and Adihaki sub cities of Mekelle city. As a result of the capacity and the coverage dimensioning 14 GSM BTS’s, 16 WCDMA NodeB’s and 58 LTE eNode’s are selected so that it is enough to cover 11.99km2 of the dense urban clutters. The output of this planning process by using radio network prediction and simulation can verify and adjust the coverage and capacity planning results. The dimensioning output such as cell radius of sites and required number of sites are some of the main important factors for our final simulation results. The simulation output includes coverage prediction area with best signal level, overlapping zone, C/I and throughput per each radio access technology that verify the planned multi RAT radio network can handle the cellular traffic of sub cities.

70

5.2

Recommendation for Future Work

As there is always a need to see a certain research topics using different approaches and methods, the following issues are recommended as future works: 

In this study the cell range of the target network has been taken from selecting the worst case scenario due to lack of the drive test data. Hence a further study can be made by incorporating the propagation model comparison after tuning each model parameters using the target deployment area drive test data.



This study focused on the network planning of the radio access technologies. One can further study the various features of these technologies as well as the core network part.



And also one can do in the optimizing process for efficient and effective dimensioning and planning outcome.

71

Reference [1]. Tibebu Mekonnen, "Dimensioning and Planning of Multi RAT Radio Network for Future Deployment in Bahir Dar City," International Journal of Scientific & Engineering Research, vol.7, Issue 11, pp.1557-1676, 2016. [2]. Ajay R. Mishra, “Fundamentals of Cellular Network Planning and Optimization 2G/ 2.5G/ 3G... Evolution to 4G”, John Wiley & Sons Ltd, 2004. [3]. Ajay R. Mishra, “Advanced Cellular Network Planning and optimization 2G/2.5G/3G Evolution to 4G”, John Wiley & Sons Ltd, England, 2007. [4]. Bethelhem Seifu, “LTE Radio Network Planning, Modeling Approaches for the Case of Addis Ababa”, Addis Ababa Ethiopia, 2012. [5]. Yiming Sun, “Radio Network Planning for 2G and 3G”, Seminar on Topics in Communications Engineering, Munich University of Technology, 2004 [6]. Abdullah Abdulwahab Mohammed Ghaleb, Adnan Sae ed Ahmed Kaid, Hassan waze ’ a Hassan Ali, Haith em Abdu Saif Esmaeel, Mahmoud Abdullah Mohammed Sadeq,Waseem mohammed Abdulsalam, Younis Taher Kaid Ahmed, “LTE Network Planning and Optimization”, Taiz University, Project, 2014 [7]. A. Benjamin Paul & Sk.M. Subani,”Code Planning of 3G UMTS Mobile Networks using ATOLL Planning Tool”, International Journal of Engineering Research & Technology (IJERT) , 2012 [8]. M.Elbagir and K.H.Bilal, "LTE Radio Planning Using Atoll Radio Planning and Optimization Software," International Journal of Science and Research, vol.3, Issue-10, 2014. [9]. J. Cerny and J. Masopust, "Long Term Evolution Networks Planning," 2016 International Conference on Applied Electronics (AE), 2016. [10].

Zewdu Gurmu, "Long Term Evolution (LTE) Radio Network Dimensioning: Case of

Addis Ababa City, " M.S.thesis, Addis Ababa University, 2015. [11].

Reshma Begum Shaik, T.Krishna Chaitanya, “Simulation of GSM Mobile Networks

Planning Using ATOLL Planning Tool”, International Journal of Engineering and Innovative Technology (IJEIT), 2012 [12].

J. Laiho and Achim acker, “Radio Network Planning and Optimization for UMTS”,

John Wiley & Sons Ltd, 2nd Edition, 2006. [13].

Liang Zhang, “Network Capacity, Coverage Estimation and Frequency Planning of

3GPP Long Term Evolution”, Linköping University Master Thesis, 2010

72

[14].

Models and Algorithms for UMTS, WLANs, and Ad Hoc Networks”, Linkoping

University, 2007 [15].

Hossam S. Hassanin, Abd-Elhamid M. Taha, Najah Abu Al, “LTE, LTE- Advanced,

and WiMAX: towards IMT-advanced networks”, John Wiley & Sons, Ltd 1st ed, 2012 [16].

Ayman Elnasbar, Mohamed A. El-Saidny & Mahmoud R. Sherif, “Design,

Deployment & Performance of 4G LTE Networks: A Practical Approach, John Wiley & Sons, Ltd, 2014 [17].

Abdul Basit, Syed, “Dimensioning of LTE Network Description of Models and

Tool, Coverage and Capacity Estimation of 3GPP Long Term Evolution radio interface” Helsinki University of Technology, Master Thesis, 2009 [18].

Richa Budhiraja, Jitendra Singh Jadon, “Mobile Radio Network Planning: A Survey”

International Journal of Emerging Technologies in Computational and Applied Sciences [19].

Kassir Hussain, “UTRAN Macro Link Budget” Hutchison 3G Engineering

Documentation, Radio Design Group, 2001 [20].

Harri Holma and Antti Toskala, “LTE for UMTS – OFDMA and SC-FDMA Based

Radio Access”, John Wiley & Sons Ltd., 2009 [21].

Huawei, “Long Term Evolution (LTE) Radio Access Network Planning Guide”,

Huawei Technology, Co., Ltd, 2008 [22].

Tewodros Mamo, “Nominal and Detailed LTE Radio Network Planning: Case of

Kirkos Sub”, may 2016, Addis Ababa [23].

Aalto University Department of Communication and Networking, "LTE Radio Link

Budget", J. Hämäläinen, 2015, Unpublished

73