hoonesiseste raadiovÕrkude planeerimineavots/bwa_1a/deniss_syropjatov_mag.pdf · hoonesiseste...

TALLINNA TEHNIKAÜLIKOOL Raadio- ja sidetehnika instituut Kood: IRT84LT HOONESISESTE RAADIOVÕRKUDE PLANEERIMINE IN-BUILDING RADIO NETWORK PLANNING Deniss Sõropjatov Töö on tehtud telekommunikatsiooni õppetooli juures Juhendaja: Avo Ots Kaitsmine toimub Infotehnoloogia teaduskonna kaitsmiskomisjonis Autor taotleb tehnikateaduste magistri nimetust Esitatud: 27.06.2005. Kaitsmine: 2005. Tallinn 2005

2

REFERAAT Käesoleva magistritöö eesmärgiks on uurida võimalusi, mille puhul sisekatvuse lahenduses ühendatakse hajutatud antennisüsteemi mitu operaatorit, mis kasutavad erinevaid sagedusalasid (GSM 900, GSM 1800, UMTS). Töös on välja töötatud disaini metodoloogia ja disaini kriteeriumid, mis vastaksid nii GSM kui ka UMTS nõudmistele. Disaini metodoloogia puudutab nii võrgu mahu dimensioneerimist kui ka katvuse planeerimist. Töö tulemusena on leitud vastavus GSM kärje ja UMTS kärje suuruste (võrgu mahu mõttes) vahel. Selline vastavus võimaldab võrgu mahu dimensioneerimist, mis on otstarbekas mõlemale süsteemile (GSM ja UMTS). Töös on määratletud sisekatvuse dimensioneerimise kriteeriumina kasutatav piirav mobiilside süsteem (mis on väikseima lubatud kaoga raadioliideses ning vastavalt ka väikseima kärje katvusalaga), ning selle süsteemile esitatavad nõudmised . Praktiline rakendus on toodud magistritöö lõpus. Magistritöö seletuskiri on esitatud 93 leheküljel, sisaldab endas 51 joonist ja 9 tabelit ning on kirjutatud inglise keeles. Võtmesõnad: Sisekatvuse lahendus (IBS), hajutatud antennisüsteem (DAS), levimudelid sisetingimusteks, mahu dimensioneerimine, katvuse dimensioneerimine.

3

ABSTRACT The aim of current thesis is to investigate the possibility of simultaneous multi-band (GSM 900, GSM 1800 and UMTS) and multi-operator access to an in-building coverage solution, which is based on distributed antenna system; to develop design methodology for this solution and to find design criteria, which fulfill requirements for both GSM and UMTS; and to present practical application of this methodology. Design methodology concerns both capacity and coverage aspects. As a result of this paper, firstly, certain mapping between cell size (capacity wise) in GSM and WCDMA is found in order to allow relatively easy capacity dimensioning optimal for both systems; and, secondly, the limiting system (the one with the lowest allowable path loss and the lowest cell range accordingly) is determined and the requirements for this system are defined to be the design criteria for cell coverage dimensioning. Practical applications of the design methodology and design criteria are presented in the end of this thesis. Current thesis is presented on 93 pages, containing also 51 illustrations and 9 tables and is written in English language. Keywords: In-Building coverage Solution (IBS), Distributed Antenna System (DAS), Indoor propagation models, capacity dimensioning, coverage dimensioning.

4

PREFACE Since the mid 1990s, the cellular communications industry has witnessed explosive growth. Wireless communications networks have become much more pervasive than anyone could have imagined when the cellular concept was first developed in the 1960s and 1970s. The worldwide cellular and personal communication subscriber base surpassed 600 million users in late 2001, and the number of individual subscribers is projected to reach 2 billion (about 30% of the world’s population) by the end of 2006! Indeed, most countries throughout the world continue to experience cellular subscription increases of 40% or more per year. The widespread adoption of wireless communications was accelerated in the mid 1990s, when governments throughout the world provided increased competition and new radio spectrum licenses for personal communications services (PCS) in the 1800-2000 MHz frequency bands. [27] The rapid worldwide growth in cellular telephone subscribers has demonstrated conclusively that wireless communications is a robust, viable voice and data transport mechanism. The widespread success of cellular has led to the development of newer wireless systems and standards for many other types of telecommunication traffic besides mobile voice telephone calls. For example, third generation cellular networks are being designed to facilitate high-speed data communications traffic in addition to voice calls. New standards and technologies are being implemented to allow wireless networks to replace fiber optic or copper lines between fixed points several kilometers apart (fixed wireless access). Similarly, wireless networks have been increasingly used as a replacement for wires within homes, buildings and office settings through the deployment of wireless local area networks (WLANs) and In-Building wireless communications Solutions (IBS). The evolving Bluetooth modem standard promises to replace troublesome appliance communication cords with invisible connections within a person’s personal workspace. Used primarily within buildings, WLANs and Bluetooth generally do not require a license for spectrum use. These license-free networks provide an interesting dichotomy in the wireless market, since ad-hoc high data rate networks are being deployed by individuals within buildings without a license, whereas wireless carriers who own the spectrum licenses for mobile cellular telephone service have focused on providing outdoor voice coverage and have been slow to provide reliable in-building coverage and high data rate services to their cellular subscribers. While it might still be too early to tell, it appears that the in-building wireless access market is becoming a huge battleground between licensed and unlicensed services, and this is prompting the architects of today’s popular cellular standards to design for high data rate packet-based networking capabilities in the next generation of cellular technology.

5

TABLE OF CONTENTS

LIST OF FIGURES......................................................................................................................................... 7

LIST OF TABLES........................................................................................................................................... 8

USED ABBREVIATIONS .............................................................................................................................. 9

INTRODUCTION ......................................................................................................................................... 10

1 IN-BUILDING COVERAGE SOLUTION OVERVIEW .................................................................. 12 1.1 PASSIVE DAS.................................................................................................................................... 12 1.2 ACTIVE DAS..................................................................................................................................... 13 1.3 MULTI-OPERATOR AND MULTI-BAND PASSIVE DAS ......................................................................... 14

2 IN-BUILDING SYSTEM CAPACITY DIMENSIONING ................................................................ 16 2.1 TRAFFIC MODELING FOR COMMUNICATION SYSTEMS ....................................................................... 17

2.1.1 Traffic engineering for cellular systems .................................................................................. 21 2.2 CELL CAPACITY DIMENSIONING FOR UMTS ..................................................................................... 23

2.2.1 Traffic estimation ..................................................................................................................... 23 2.2.2 Grade of service....................................................................................................................... 23 2.2.3 Definitions................................................................................................................................ 23 2.2.4 Capacity calculations............................................................................................................... 26

2.3 CELL CAPACITY DIMENSIONING FOR GSM........................................................................................ 28 2.3.1 Traffic estimation ..................................................................................................................... 28 2.3.2 Grade of service....................................................................................................................... 28 2.3.3 Capacity calculations............................................................................................................... 29

2.4 CONCLUSION OF CAPACITY DIMENSIONING....................................................................................... 29 3 IN-BUILDING SYSTEM CELL COVERAGE DIMENSIONING................................................... 30

3.1 RADIO WAVE PROPAGATION PREDICRION AND MODELING ................................................................ 30 3.1.1 The three main propagation mechanisms ................................................................................ 31

3.1.1.1 Reflection.............................................................................................................................................31 3.1.1.2 Diffraction............................................................................................................................................32 3.1.1.3 Scattering .............................................................................................................................................32

3.1.2 Large-scale indoor propagation models .................................................................................. 32 3.1.2.1 Free space propagation model..............................................................................................................34 3.1.2.2 Partition losses (same floor).................................................................................................................36 3.1.2.3 Partition losses between floors .............................................................................................................37 3.1.2.4 Log-distance path loss model ...............................................................................................................38 3.1.2.5 Log-normal shadowing ........................................................................................................................38 3.1.2.6 Determination of percentage of coverage area .....................................................................................40 3.1.2.7 Attenuation factor model......................................................................................................................42 3.1.2.8 Keenan-Motley model..........................................................................................................................43 3.1.2.9 Modified Keenan-Motley model ..........................................................................................................43

3.1.3 Small-scale indoor propagation models .................................................................................. 45 3.1.3.1 Rayleigh fading model .........................................................................................................................47 3.1.3.2 Ricean fading model ............................................................................................................................48

3.2 CELL COVERAGE DIMENSIONING FOR GSM ...................................................................................... 49 3.2.1 Definitions................................................................................................................................ 49 3.2.2 Margins.................................................................................................................................... 50 3.2.3 GSM 900 cell coverage dimensioning...................................................................................... 53 3.2.4 GSM 1800 cell coverage dimensioning.................................................................................... 54

3.3 CELL COVERAGE DIMENSIONING FOR UMTS.................................................................................... 55 3.3.1 Definitions................................................................................................................................ 55 3.3.2 Margins.................................................................................................................................... 57 3.3.3 Uplink UMTS cell coverage dimensioning............................................................................... 57 3.3.4 Downlink UMTS cell coverage dimensioning.......................................................................... 60 3.3.5 Final UMTS cell coverage ....................................................................................................... 61

3.4 CONCLUSION OF CELL COVERAGE DIMENSIONING ............................................................................ 61

6

4 IBS DESIGN EXAMPLE...................................................................................................................... 62 4.1 DESIGN PROCESS .............................................................................................................................. 63 4.2 NOMINAL ANTENNA PLAN ................................................................................................................ 64 4.3 SURVEY ............................................................................................................................................ 68 4.4 COVERAGE MEASUREMENTS............................................................................................................. 69

4.4.1 Transmitter equipment ............................................................................................................. 69 4.4.2 Receiver equipment.................................................................................................................. 69 4.4.3 Measurement procedure .......................................................................................................... 70

4.4.3.1 Floor 0..................................................................................................................................................70 4.4.3.2 Office floors .........................................................................................................................................76

4.5 CABLE DISTRIBUTION DESIGN ........................................................................................................... 82 5 CONCLUSIONS AND FUTURE WORK ........................................................................................... 87

6 REFERENCES....................................................................................................................................... 88

A. PASSIVE DAS COMPONENTS ...................................................................................................... 90 A.1 COMBINING BOX .............................................................................................................................. 90 A.2 ANTENNAS........................................................................................................................................ 91 A.3 FEEDER CABLES ................................................................................................................................ 92 A.4 TAPPERS ........................................................................................................................................... 92 A.5 SPLITTERS......................................................................................................................................... 93

7

LIST OF FIGURES FIGURE 1.1 PASSIVE DISTRIBUTED ANTENNA SYSTEM [7]................................................................................. 12 FIGURE 1.2 ACTIVE DISTRIBUTED ANTENNA SYSTEM [7].................................................................................. 13 FIGURE 1.3 MULTI-OPERATOR AND MULTI-BAND PASSIVE DAS ARCHITECTURE. ........................................... 14 FIGURE 2.1 MULTI-SERVER QUEUING MODELS [27]......................................................................................... 18 FIGURE 2.2 PROBABILITY OF LOSS FOR INFINITE SOURCES, LCC SYSTEMS. [27].............................................. 22 FIGURE 3.1 SMALL-SCALE AND LARGE-SCALE FADING [21]. ............................................................................ 31 FIGURE 3.2 SCATTER PLOT OF MEASURED DATA AND CORRESPONDING MMSE PATH LOSS MODEL FOR MANY

CITIES IN GERMANY [24]. ......................................................................................................................... 39 FIGURE 3.3 THE AREA COVERAGE AS A FUNCTION OF �/N FOR DIFFERENT VALUES OF THE PERIMETER

COVERAGE. ............................................................................................................................................... 42 FIGURE 3.4 DIAGRAMS SHOWING APPROXIMATE INDOOR PROPAGATION LOSS AT 900 MHZ FOR A WALL

ATTENUATION OF 0.2 & 0.5 DB/M [7]....................................................................................................... 44 FIGURE 3.5 LOSS BETWEEN FLOORS [7]. ........................................................................................................... 45 FIGURE 3.6 A TYPICAL RAYLEIGH FADING ENVELOPE AT 900 MHZ [10]. ........................................................ 47 FIGURE 3.7 RAYLEIGH PROBABILITY DENSITY FUNCTION (PDF) [21]................................................................ 48 FIGURE 3.8 PROBABILITY DENSITY FUNCTION OF RICEAN DISTRIBUTIONS [21]................................................ 49 FIGURE 3.9 THE CUMULATIVE NORMAL DISTRIBUTION FUNCTION [7]. ............................................................. 51 FIGURE 4.1 VERTICAL STRUCTURE OF THE BUILDING. ...................................................................................... 62 FIGURE 4.2 IBS DESIGN WORKFLOW. ............................................................................................................... 64 FIGURE 4.3 CELL SPLITTING. ............................................................................................................................ 65 FIGURE 4.4 NOMINAL ANTENNA PLAN FOR FLOOR 2......................................................................................... 66 FIGURE 4.5 NOMINAL ANTENNA PLAN FOR FLOOR 0......................................................................................... 67 FIGURE 4.6 RBS POSITION AND CABLE RAISERS............................................................................................... 68 FIGURE 4.7 MEASUREMENT CONFIGURATION................................................................................................... 69 FIGURE 4.8 COVERAGE AREA OF ANTENNA 0.1 (BI-DIRECTIONAL) IN THE GARAGE ON FLOOR 0. ..................... 71 FIGURE 4.9 COVERAGE AREA OF ANTENNA 0.2 (DIRECTIONAL) IN THE GARAGE ON FLOOR 0........................... 71 FIGURE 4.10 COVERAGE AREA OF ANTENNA 0.3 (DIRECTIONAL) IN THE GARAGE ON FLOOR 0......................... 72 FIGURE 4.11 COVERAGE AREA OF ANTENNA 0.4 (OMNI) IN THE GARAGE ON FLOOR 0...................................... 72 FIGURE 4.12 COVERAGE AREA OF ANTENNA 0.5 (OMNI) IN THE TECHNICAL PART OF FLOOR 0......................... 73 FIGURE 4.13 COVERAGE AREA OF ANTENNA 0.6 (OMNI) IN THE TECHNICAL PART OF FLOOR 0......................... 74 FIGURE 4.14 COVERAGE AREA OF ANTENNA 0.7 (OMNI) IN THE TECHNICAL PART OF FLOOR 0......................... 74 FIGURE 4.15 COVERAGE AREA OF ANTENNA 0.8 (OMNI) IN THE TECHNICAL PART OF FLOOR 0......................... 75 FIGURE 4.16 FINAL ANTENNA PLAN FOR FLOOR 0............................................................................................. 75 FIGURE 4.17 COVERAGE AREA OF ANTENNA 2.1 (OMNI) ON FLOOR 2............................................................... 76 FIGURE 4.18 COVERAGE AREA OF ANTENNA 4.1 (OMNI) ON FLOOR 4............................................................... 77 FIGURE 4.19 COVERAGE AREA OF ANTENNA 4.2 (OMNI) ON FLOOR 4............................................................... 77 FIGURE 4.20 COVERAGE AREA OF ANTENNA 4.3 (OMNI) ON FLOOR 4............................................................... 78 FIGURE 4.21 COVERAGE AREA OF ANTENNA 4.4 (OMNI) ON FLOOR 4............................................................... 78 FIGURE 4.22 COVERAGE AREA OF ANTENNA 4.5 (OMNI) ON FLOOR 4............................................................... 79 FIGURE 4.23 COVERAGE AREA OF ANTENNA 3.1 (OMNI) ON FLOOR 4............................................................... 79 FIGURE 4.24 COVERAGE AREA OF ANTENNA 3.2 (OMNI) ON FLOOR 4............................................................... 80 FIGURE 4.25 FINAL ANTENNA PLAN FOR FLOOR 1............................................................................................. 81 FIGURE 4.26 FINAL ANTENNA PLAN FOR FLOOR 2............................................................................................. 81 FIGURE 4.27 ANTENNA EIRP CALCULATIONS. ................................................................................................. 83 FIGURE 4.28 ANTENNA AND COMPONENTS LOCATIONS DIAGRAM FOR FLOOR 2. .............................................. 84 FIGURE 4.29 FINAL CELL SPLITTING DIAGRAM. ................................................................................................ 85 FIGURE 4.30 SAMPLE OF TRUNKING DIAGRAM. ................................................................................................ 86 FIGURE A.1 ERICSSON MULTI-OPERATOR COMBINING BOX. ........................................................................... 90 FIGURE A.2 OMNI-DIRECTIONAL ANTENNA...................................................................................................... 91 FIGURE A.3 DIRECTIONAL ANTENNA................................................................................................................ 91 FIGURE A.4 BI-DIRECTIONAL ANTENNA. .......................................................................................................... 92 FIGURE A.5 COAXIAL FEEDER CABLES. ............................................................................................................ 92 FIGURE A.6 PRINCIPAL TAPPER DIAGRAM. ....................................................................................................... 93 FIGURE A.7 TAPPER. ........................................................................................................................................ 93 FIGURE A.8 SPLITTER....................................................................................................................................... 93

8

LIST OF TABLES TABLE 2.1 PART OF ERLANG’S B-TABLE, YIELDING THE TRAFFIC (IN ERLANGS) AS A FUNCTION OF THE GOS

(COLUMNS) AND NUMBER OF TRAFFIC CHANNELS (ROWS)........................................................................ 21 TABLE 2.2 DTX GAIN [7]. ................................................................................................................................ 24 TABLE 3.1 AVERAGE SIGNAL LOSS MEASUREMENTS REPORTED BY VARIOUS RESEARCHERS FOR RADIO PATHS

OBSTRUCTED BY COMMON BUILDING MATERIAL [21]. .............................................................................. 37 TABLE 3.2 AVERAGE FLOOR ATTENUATION FACTOR IN DB IN TWO OFFICE BUILDINGS [21]............................. 38 TABLE 3.3 REQUIRED SIGNAL STRENGTHS IN AREAS OF DIFFERENT INTERFERENCE. ........................................ 52 TABLE 3.4 SOME TYPICAL EXAMPLES OF RADIO ENVIRONMENTS REPRESENTING LOW, MEDIUM AND HIGH

LEVEL OF INTERFERENCE. ......................................................................................................................... 53 TABLE 4.1 NUMBER OF USERS IN THE BUILDING (PRESENTED PER FLOOR)........................................................ 62 TABLE A.1 COMBINING BOX TECHNICAL SPECIFICATIONS................................................................................ 90 TABLE A.2 ANTENNAE TECHNICAL SPECIFICATIONS. ....................................................................................... 91

9

USED ABBREVIATIONS AMR - Adaptive Multi Rate BCCH - Broadcast Control Channel BER - Bit Error Rate BS - Base Station DAS - Distributed Antenna System CDMA - Code Division Multiple Access CPL - Car Penetration Loss CS - Circuit Switched DCCH - Dedicated Control Channel DL - DownLink DTCH - Dedicated Traffic Channel DTX - Discontinuous Transmission EIRP - Effective Isotropic Radiated Power ERP - Effective Radiated Power ETSI - European Telecommunication Standards Institute FAF - Floor Attenuation Factor FER - Frame Error Rate GoS - Grade of Service GPRS - General Packet Radio Service GSM - Global System for Mobile communications IBS - In-Building System LCC - Lost Calls Cleared LCD - Lost Calls Delayed LCH - Lost Calls Held LI - Local Interface LOS - Line-Of-Sight MCPA - Multi Carrier Power Amplifier MS - Mobile Station NLOS - Non-Line-Of-Sight PCS - Personal Communications Services PDF - Probability Density Function PS - Packet Switched RAB - Radio Access Bearer RB - Radio Bearer RBS - Radio Base Station RF - Radio Frequency RMS - Root Mean Square RU - Remote Unit SDCCH - Standalone Dedicated Control Channel SRB - Signaling Radio Bearer TDMA - Time Division Multiple Access TEMS - Test Mobile Station TRU - Transceiver Unit TU - Typical Urban UE - User Equipment UL - UpLink UMTS - Universal Mobile Telecommunications System WCDMA - Wideband CDMA, Code Division Multiple Access WLAN - Wireless Local Area Network

10

INTRODUCTION With the rapidly growing number of cellular telephone subscribers it becomes more and more important to provide good quality of service for the users inside the buildings, because people (especially urban inhabitants) are actually spending most of the time indoors. Current cellular systems generally perform quite poor in the in-building environment and cannot meet high requirements set on the quality of service. It gives a rise in demand for the IBS, which should provide dedicated coverage and capacity for the users indoors. Current paper concerns one of the possible In-Building Solutions. This solution constitutes passive Distributed Antenna System (DAS), which allows simultaneous multi-band (GSM 900, GSM 1800 and UMTS) and multi-operator access. Technical overview of the passive DAS and problem definition is provided in Chapter 1. In DAS under consideration second generation (GSM) and third generation (UMTS) systems should coexist. Mentioned systems use different access technologies and thus cell capacity of these systems is calculated differently and affected by different factors. Capacity aspects of both systems should be considered. Cells need to be dimensioned so that they provide enough capacity for GSM and UMTS. Chapter 2, firstly, adduces traffic modeling for communication systems. Secondly, certain mapping between cell size (capacity wise) in GSM and WCDMA is found in order to allow relatively easy capacity dimensioning optimal for both systems. In-building solution capacity dimensioning methodology and its implementation for both GSM and UMTS are also addressed in this chapter. As it was mentioned above, IBS under consideration allows simultaneous multi-band and multi-operator access. Several mobile telephony systems can be combined into the same antenna system: GSM 900, GSM 1800 and UMTS. All these systems use different frequency bands for operation and moreover – GSM and UMTS use completely different technologies for transmitting data over the air. This means that mentioned systems have different requirements for received signal strength and attenuation in the radio propagation environment is also different. When dimensioning coverage of the systems these aspects have to be considered and design criteria suitable for all mentioned systems needs to be found. Chapter 3 describes basic radio wave propagation aspects and propagation models. Multi-band indoor cell coverage dimensioning methodology and criteria are also presented in this chapter. Chapter 4 adduces example of IBS design – from nominal (based on cell coverage dimensioning results) to final antenna plan. Coverage test results, which are used for nominal antenna plan verification, are also presented here. Chapter 5 concludes the thesis along with providing some future research directions In current thesis we investigate different aspects related to design of multi-band and multi-operator IBS. As a result of this paper, firstly, certain mapping between cell size (capacity wise) in GSM and WCDMA is found in order to allow relatively easy capacity dimensioning optimal for both systems; and, secondly, the limiting system (the one with the lowest allowable path loss and the lowest cell range accordingly) is determined and the requirements for this system are defined to be the design criteria for cell coverage dimensioning.

11

I would like to take this opportunity to express my gratitude to the members of Ericsson team, who worked together with me in the area of IBS design and provided great support in writing this thesis.

12

1 IN-BUILDING COVERAGE SOLUTION OVERVIEW During the last years there has been a growing demand for mobile coverage within buildings. In-building solutions are designed and built to meet the requirements of the customer’s needs. The main purpose of the in-building solution is to provide a certain building complex with dedicated coverage and capacity for the users of mobile systems. There are different approaches of providing mobile coverage inside buildings. One of them is using a Distributed Antenna System. The in-building distributed antenna system solutions can be divided into two main categories: passive or active solutions. It should be mentioned in advance that current thesis would concentrate on Passive Distributed Antenna system. Active DAS will be briefly mentioned as an alternative. 1.1 PASSIVE DAS Passive DAS uses coaxial feeder cables and RF power splitters, couplers and tappers in order to distribute the RF power from a centralized radio base station to a network of distributed antennas within a building. This provides a very reliable and robust system. The passive DAS requires only a minimum of maintenance since there are no active components. With this kind of distribution the radiated power from each antenna can be kept very low. Passive DAS components are discussed in more detail in Appendix A. Figure 1.1 shows architecture of a Passive DAS. Figure 1.1 Passive distributed antenna system [7].

RBS

Antennas

FeederCables

Splitters/Tappers

Jumpers

13

1.2 ACTIVE DAS In an active distributed antenna system, active components, such as Local Interfaces (LI) and Remote Units (RU) are introduced in the antenna distribution network. Single mode twin fiber interconnects these components. The task of LI is to transform RF signal coming from radio base station into optical signal and distribute it over optical fibers to several Remote Units. In opposite direction (from RU to RBS) LI performs reverse conversion. RU, in turn, converts optical signal, coming from LI, back to RF signal and distributes it over coaxial cable to antennas. RU performs reverse conversion in the direction from antennas to LI. In active systems, a combination of active and passive components is normally used. The distribution to each remote unit, or central point, is made through an active fiber system. Passive co-axial feeders and antennas then handle the distribution from each remote unit on a particular floor. An Active DAS can be used for distribution of RF signals over long distances (some kilometers) and/or to a large number of antennas. It is also a viable solution for installations where building constraints make it troublesome or impossible to install passive systems. A fiber optical distribution system can easily be used for signal distribution between a number of buildings in an area or a complex. In comparison with the Passive solution, Active DAS is usually more expensive and requires more maintenance, since it includes active components. Figure 1.2 depicts Active DAS architecture. Figure 1.2 Active distributed antenna system [7].

Antennas

FeederCables

Jumpers

LI

RU

RU

RU

RBS

OpticalFiber Cable

RemoteUnits

Local Interface

14

1.3 MULTI-OPERATOR AND MULTI-BAND PASSIVE DAS In order for the solution to allow simultaneous multi-band and multi-operator access, certain combining equipment needs to be used. Combining Box (see Appendix A) can be used as the first interface between the operators’ equipment (radio base station) and the DAS. This device allows several networks or operators utilize the same in-building network. The basic design principle of the combining box is that it has twelve inputs and four outgoing antenna ports. There are four inputs for each of the frequency bands: for example 900, 1800 and 2000 MHz, which can correspond to GSM 900, GSM 1800 and UMTS. The twelve signals are combined within the box and sent out to every antenna port with one-fourth power. This means that every antenna port has one fourth of all twelve signals. Further information and technical specifications on Combining Box can be found in Appendix A. Further, RF signal is distributed to the antennas distributed in the building by means of coaxial cables connected to the antenna ports of the combining box. Passive power splitters and tappers are used to split RF power between branches of the antenna system. Figure 1.3 depicts architecture of the multi-operator and multi-band Passive DAS. Figure 1.3 Multi-operator and multi-band passive DAS architecture. Combining Box enables GSM 900/1800 and UMTS to coexist in the same passive distributed antenna system. These two mobile telephony systems use different medium access technologies: GSM uses Time Division Multiple Access (TDMA) and UMTS uses

Power splitters and antennas

Coaxial Cables

F1

F2

F3

F4

F5

GSM 900

GSM 1800

WCDMA

GSM 1800

WCDMA

CombiningEquipmentCombiningEquipment

15

Wideband Code Division Multiple Access. Moreover, frequency bands used by the systems are also different. These facts put certain requirements on the physical DAS itself and provide problems to be considered during design of the IBS:

1. All components in the DAS, such as: antennas, feeders, splitters etc, must support the GSM 900, GSM 1800 and UMTS frequency bands.

2. Cell capacity in UMTS and in GSM is defined differently and depends on different factors. GSM cell capacity depends on the amount of time-slots available in that cell; and is basically only limited by certain hardware in the RBS. UMTS, on the other hand, is an interference limited system and cell capacity is rather limited by the radio access technology itself, that by RBS hardware. In UMTS cell capacity and cell coverage are in tight relation with each other. This situation has to be considered during IBS cell capacity dimensioning, because GSM and UMTS will use the same physical DAS.

3. In addition to different radio access technologies, GSM 900/1800 and UMTS use different frequency bands. These both factors have impact on cell coverage. Mentioned systems have different requirements for received signal strength and attenuation in the radio propagation environment is also different. In UMTS, as it was already mentioned, cell capacity and cell coverage depend on each other.

First point in the list is solved by using DAS components described in Appendix A. Cell capacity dimensioning methodology, optimal for both GSM and UMTS is adduced in Chapter 2. Third problem is concerned in Chapter 3. Certain cell coverage dimensioning criteria, which fulfill requirements for all systems, are found.

16

2 IN-BUILDING SYSTEM CAPACITY DIMENSIONING When dimensioning the capacity of an indoor cell, one must take its application into account. Two different categories of indoor cells can be identified [7]:

1. Public indoor cells. Indoor cells which cover public buildings such as shopping centres and airport terminals.

When dimensioning the capacity in public indoor cells, the same methods and requirements that are used in the surrounding macrocell network should be applicable. However, it is important to keep in mind that in some buildings the number of subscribers might fluctuate very much during the day and between different days. Some examples of this are sports grounds, convention centres and shopping malls.

2. Business indoor cells. Indoor cells covering areas such as offices. In some

applications the coverage, capacity and quality demands in business indoor cells may be considerably higher than in public indoor cells.

Business indoor cells may be installed in order to improve the performance of the existing radio network, or as a substitute for the fixed telephone network. The demands on network accessibility as well as coverage are likely to be higher in business indoor cells, especially if the indoor cell is to work as a substitute for the fixed telephony.

Current paper will concentrate on business indoor cells. As DAS under consideration supports both second generation (GSM) and third generation (UMTS) systems, capacity aspects of both systems should be considered. Cells need to be dimensioned so that they provide enough capacity for both GSM and UMTS. GSM is nowadays mainly used for circuit switched communications. That makes capacity calculations for GSM quite straight forward. But on the other hand UMTS is a system that supports different circuit and packet switched services. That makes capacity dimensioning for UMTS much more complicated and requires quite precise user traffic profile. As the experience shows, this user profile is not always available or is based mainly on assumptions due to the fact, that 3G networks are not yet widely spread (by the time this thesis was written). That is why user profiles tend to be very uncertain and quickly changing with time. The following clauses will address traffic and cell capacity calculations. Certain mapping between cell size (capacity wise) in GSM and WCDMA has to be found in order to allow relatively easy design of a solution optimal for both systems. In current paper the following strategy is used. Cells have to be dimensioned in a way to be able to handle traffic for both scenarios:

�� 100% of the users will use GSM only. �� 100% of the users will use UMTS only.

17

It is assumed that the system is primarily intended for speech communications. Data services are considered to be delivered on best effort basis and involve only packet switched services. It means that speech will be the only circuit switched service. In UMTS cell capacity is rather limited by the radio access technology itself, than by BS hardware configuration. In GSM it is the other way around – cell capacity is virtually limited by the BS configuration. Due to this, calculations are firstly performed in order to find the amount of speech users that UMTS cell can serve. Then it will be calculated how much capacity is left for the data services. Secondly, calculations are performed to define configuration of GSM cell, which can serve the same amount of speech users as UMTS cell. Before proceeding with capacity dimensioning, traffic modeling will be addressed in the following section. 2.1 TRAFFIC MODELING FOR COMMUNICATION SYSTEMS Traffic engineering concepts were developed in the design of telephone switches and circuit-switching telephone networks, but the concepts equally apply to cellular networks. Consider a cell able to handle N simultaneous users (capacity of N channels) that has L potential subscribers (L mobile units). If L < N, the system is referred to as nonblocking; all calls can be handled all the time. If L > N, the system is blocking; a subscriber may attempt a call and find the capacity fully in use and therefore be blocked. For a blocking system, the fundamental performance questions are as follows:

1. What is the degree of blocking; that is, what is the probability that a call request will be blocked? Alternatively, what capacity (N) is needed to achieve a certain upper bound on the probability of blocking?

2. If blocked calls are queued for service, what is the average delay? Alternatively, what capacity is needed to achieve a certain average delay?

Two parameters determine the amount of load presented to a system [27]:

�� :� the mean rate of calls (connection requests) attempted per unit of time, �� :h the mean holding time per successful call.

The basic measure of traffic is the traffic intensity, expressed in a dimensionless unit, the Erlang [27]:

2.1 hA �� (2.1)

The parameter A, as a measure of busy-hour traffic, serves as input to a traffic model. The model is then used to answer questions such as those posed in the beginning of this section. There are two key factors that determine the nature of the model:

�� The manner in which blocked calls are handled; �� The number of traffic sources.

18

(a) Infinite source, lost calls cleared.

(b) Infinite source, lost calls delayed.

(c) Finite source, lost calls cleared.

(d) Finite source, lost calls delayed

Figure 2.1 Multi-server queuing models [27].

19

Blocked calls may be handled in one or two ways. First, blocked calls can be put in a queue awaiting a free channel; this is referred to as lost calls delayed (LCD), although in fact the call is not lost, merely delayed. Second, a blocked call can be rejected and dropped. This in turn leads to two assumptions about the action of the user. If the user hangs up and waits some random time interval before another call attempt, this is known as lost calls cleared (LCC). If the user repeatedly attempts calling, it is known as lost calls held (LCH). For each of these blocking options, formulas have been developed that characterize the performance of the system. For cellular systems, the LCC model is generally used and is generally the most accurate. The second key element of a traffic model is whether the number of users is assumed to be finite or infinite. Figure 2.1 illustrates this difference. For an infinite source model, there is assumed to be a fixed mean arrival rate. For the finite source case, the arrival rate will depend on the number of sources already engaged. In particular, if the total pool of users is L, each of which generates calls at an average rate of L/� , then, when the cell is totally idle, the arrival rate is � . However, if there are K users occupied at time t, then the instantaneous arrival rate at that time is LKL /)( �� . Infinite source models are analytically easier to deal with. The infinite source assumption is reasonable when the number of sources is at least 5 to 10 times the capacity of the system. Let us consider the use of various traffic models for system sizing. For each of the four models of Figure 2.1, formulas have been derived for the quantities of interest. The most important relationships are summarized in the following formulas. Infinite sources, lost calls cleared [27]:

2.2

��

�N

x

x

N

xA

NA

P

0 !

! . (2.2)

Infinite sources, lost calls delayed [27]:

2.3

ANN

NA

xA

ANN

NA

PNN

x

x

N

��

��

��

� !!

!)0(1

0

. (2.3)

htANePtP /)()0()( ��

htANetP /)(

2 )( ��

ANh

PD�

� )0(1

ANh

D�

�2 .

20

Finite sources, lost calls cleared [27]:

2.4

��

�

� �

� �

�

� �

� �

�N

x

x

M

Mx

LN

L

P

N

0

1

1

(2.4)

where

)!1(!)!1(1

xLxL

x

L

��

��

� �

� �

)1( PALA

M��

� .

Finite sources, lost calls delayed [27]:

2.5

��

�

��

�

�

��

��

�

� �

�

��

L

NxNx

xN

x

x

L

NxNx

x

NxLM

NL

Mx

LNxL

MNL

P

)!(!!)!(!

!

)0(1

0

(2.5)

where

ALA

PALA

M��

��

�1)1(1

.

The following denominations are used in the above formulas: A - offered traffic, Erlangs, N - number of servers, P - probability of loss (blocking), P(>0) - probability of delay greater than 0, P(>t) - probability of delay greater than t, P2(>t) - probability of delay greater than t on calls delayed, D1 - mean delay, all calls, D2 - mean delay, delayed calls. The formulas are based on the following assumptions [27]:

�� Poisson arrivals; �� Exponential holding time (not needed for infinite sources, LCC); �� Equal traffic intensity per source; �� Calls served in order of arrival (for delay calculations).

21

Even with these assumptions, it can be seen that the formulas involve lengthy summations. In the early development phase of communications systems, much of the work of traffic theorist laid in simplifying assumptions to the point that the equations could be calculated at all. The results were and are published in tables. Unfortunately, the tendency would be to try to use the available tables in situations whose assumptions did not fit any of the tables. The problem is now alleviated with the use of the computer. Nevertheless, the tabular results are still useful for quick and rough sizing. Several parameters in the formulas above warrant comment. For LCC systems, P is the probability that a call request will be cleared or lost. This is the ratio of calls unable to obtain service to the total call requests; in telephone traffic, it is also called grade of service (GoS). For LCD systems, an arriving call will be delayed rather than cleared. In this case, P(>0) is the probability that a call request will find the system fully utilized and hence be delayed. P(>t) is the probability that any call request is delayed by an amount greater than t, whereas P2(>t) is the probability that a call that is delayed will be delayed by an amount greater than t. 2.1.1 Traffic engineering for cellular systems Considering basic access algorithms for cellular systems and user behavior, cellular systems are blocking systems and LCC traffic model generally has the best compliance and is the most accurate. For cellular systems, GoS is always greater than 0, which allows certain blocking probability in a system. That means, that the number of users (sources) in a cellular system is usually much higher than the capacity of the system. Traffic model with infinite sources is applicable in this case. Concluding mentioned above, LCC traffic model with infinite sources has the best matching for cellular systems. The equation (2.2) is also referred to as Erlang B. This equation is easily programmed, and tables of values are readily available. Given the offered load and number of servers, the grade of service can be calculated or determined from a table. More often, the inverse problem is of interest: determining the amount of traffic that can be handled by a given capacity to produce a given grade of service. Another problem is to determine the capacity required to handle a given amount of traffic at a given grade of service. Table 2.1 provides an extract of an erlang B table. Figure 2.2 plots the probability of loss as a function of offered load with the number of servers as a parameter. Table 2.1 Part of Erlang’s B-table, yielding the traffic (in Erlangs) as a function of the GoS (columns) and number of traffic channels (rows).

N .007 .008 .009 .01 .02 .03 .05 .1 .2 n 1 .00705 .00806 .00908 .01010 .02041 .03093 .05263 .11111 .25000 1 2 .12600 .13532 .14416 .15259 .22347 .28155 .38132 .59543 1.0000 2 3 .39664 .41757 .43711 .45549 .60221 .71513 .89940 1.2708 1.9299 3 4 .77729 .81029 .84085 .86942 1.0923 1.2589 1.5246 2.0454 2.9452 4 5 1.2362 1.2810 1.3223 1.3608 1.6571 1.8752 2.2185 2.8811 4.0104 5 6 1.7531 1.8093 1.8610 1.9090 2.2759 2.5431 2.9603 3.7584 5.1086 6 7 2.3149 2.3820 2.4437 2.5009 2.9354 3.2497 3.7378 4.6662 6.2302 7 8 2.9125 2.9902 3.0615 3.1276 3.6271 3.9865 4.5430 5.5971 7.3692 8

22

Figure 2.2 Probability of loss for infinite sources, LCC systems. [27]

23

2.2 CELL CAPACITY DIMENSIONING FOR UMTS 2.2.1 Traffic estimation Unlike the situation of public indoor cells it should be fairly easy to estimate the traffic in a business indoor cell. This can be done by studying telephone bills or by making an inquiry. Note that if the business indoor cell is to substitute the fixed telephone network, the traffic per subscriber in the cell is likely to become very high (typically 100 mE/subscriber). The number of subscribers per area/floor is important to retrieve from the customer for a good calculation of needed traffic capacity. 2.2.2 Grade of service The demand on network accessibility is likely to be higher in business indoor cells than in public indoor cells. A typical value of GoS might be 1%. If the cell is to substitute the fixed telephone network, the performance must be similar to this and the GoS must be very low, more than 0.5% will probably not be tolerated. 2.2.3 Definitions Several UMTS specific definitions need to be clarified for capacity calculations Activity factor and DTX gain A radio link between the UE and the RBS consists of a Dedicated Traffic Channel (DTCH) and a channel used for radio resource control signaling sent on the Dedicated Control Channel (DCCH). When one of the channels (DTCH or DCCH) is not transmitting, less interference is generated on the air interface and more capacity can be expected. The capacity gain depends on the activity factor of the DCCH and DTCH channel or channels. In this thesis the activity factor is assumed to be 10% for DCCH, 50% for the speech/AMR DTCH and 100% for all other DTCH channels. Due to the low activity factor for DCCH, the DTX gain can be large, especially in cases when the data rate on DTCH is low. The resulting DTX gains are given in Table 2.2.

24

Table 2.2 DTX gain [7].

DTX gain RB configuration

UL DL

Speech/AMR 12.2 kbps RB1 + 3.4 kbps SRB2 69% 102%

AMR 7.95 kbps RB1+ 3.4 kbps SRB2 69% 102%



28.8 kbps CS RB3 + 3.4 kbps SRB 7% 5%

64 kbps CS RAB3 + 3.4 kbps SRBs on DPDCH/DPCH2 7% 5%

8 kbps PS RB3 + 3.4 kbps SRB2 5% 5%


32 kbps PS RB3+ 3.4 kbps SRB2 5% 5%


128 kbps PS RB3 + 3.4 kbps SRB2 – 2%

384 kbps PS RB3 + 3.4 kbps SRB2 – 1%

57.6 kbps CS RAB3 + 3.4 kbps SRBs for DCCH2 6% 6%

Streaming 16 kbps PS RB3 + 8 kbps PS RB3 + 3.4 kbps SRB2 6% –

Streaming 64 kbps PS RB3 + 8 kbps PS RB3 + 3.4 kbps SRB2 – 6% 1 50% activity factor 2 10% activity factor 3 100% activity factor during packet transmission

The DTX gain in Table 2.2 has been included in the Mpole figures presented further elsewhere in this work. Pole capacity The pole capacity, Mpole, is the theoretical limit for the number of simultaneous RABs that a cell can support. In the uplink at this limit the interference level in the system is infinite and thus the coverage reduced to zero. In the downlink it corresponds to an infinite BS output power and no common channels. Mpole values that are used in current thesis are based on simulations, which were performed by Ericsson. Complete results of the simulations can found in [7]. Loading In WCDMA dimensioning, it is necessary to define the concept Loading:

2.6 poleMM

Loading � (2.6)

where M is the number of simultaneous users in the cell. For a multi-service system where the services utilize different types of RABs, the equation can be written as:

25

2.7 ...

MM

MM

MM

Loading,pole,pole,pole

��3

3

2

2

1

1

(2.7)

where Mn - the number of simultaneous users for the n:th RAB, Mpole,n - the uplink pole capacity for the n:th RAB. Average served traffic and average loading due to speech users The number of channels required to handle certain offered traffic with certain GoS is calculated, as it was mentioned before, using Erlang-B formulas. This number of required channels is basically valid only for the “busy hour” situation. The real utilization of the channels will be lower than 100%, leaving some free capacity. Term average served traffic is defined to characterize this:

2.8 )1( GoSAA offeredaverage �� (2.8).

Then average loading can be defined as:

2.9 pole

averageaverage M

ALoading �

(2.9).

Uplink load limit A WCDMA system cannot be loaded up to 100%. To secure a well performing network the load used in the dimensioning process should not exceed the following numbers [7]: UL system load limit: 60% DL system load limit: 75% In the uplink this limit is related to the fact that the noise rise increases rapidly at the loads above 60% and there is a risk of the system becoming unstable. In the DL the load limit is more related to the power available in the power amplifier. A maximum system limit that depends on the pole capacity is not relevant as a dimensioning criterion. At the cell borer, where the background noise is the limitation, the power amplifier sets the limit in how much the cell can be loaded. The admission control currently has an admission threshold of 15 W (for a 20 W BS) where new users are not allowed into the cell. This is required in order to ensure that current users have enough headroom in order to move within the cell but also to allow for mobility of users between cells. Therefore a power “headroom” is introduced. This headroom is currently defined to 5 W, meaning that not more than a loading of up to 15 W average MCPA (Multi Carrier Power Amplifier) power or 75% of max power, is allowed in the homogeneous network.

26

2.2.4 Capacity calculations UMTS belongs to Code Division Multiple Access systems, where all transmitters are simultaneously transmitting on the same frequency and are identified by special scrambling codes. It these systems capacity is different for up- and downlink. Final cell capacity is defined by the limiting factor: whether uplink or downlink. It is assumed that:

�� Speech traffic generated by office user is equal to 100 mE (see Section 2.1); �� GoS = 0.5% (see Section 2.1); �� Load limits for are: 60% Uplink, and 75% Downlink.

Uplink Cell Capacity Considering uplink load limit and downlink Mpole value for speech in indoor environment [7] equation (2.6) becomes:

576.0 speechM

� .

Then the maximum number of simultaneous speech users (channels) in the cell is equal to:

342.346.057 ��speechM . According to Erlang-B formula (GoS=0.5%) UMTS cell speech capacity (maximum offered traffic) is equal to 22.336 Erl. The maximum number of speech users a UMTS cell can support is equal to:

22336.2231.0

336.22��

ErlErl

N speech .

Offered traffic generated by speech users per cell is equal to:

ErlErl 3.221.0223 �� According to equations (2.8) and (2.9) average uplink loading generated 223 speech users is equal to:

%93.38%10057

)005.01(3.22

,,

, ��

��uplinkspeechpole

speechaverageaverage M

ALoad

Obtained average Load value is lower than the uplink loading limit by ~21%. Roughly speaking, this means that a UMTS cell can serve 223 speech subscribers, and there is still some free uplink capacity for best effort packet switched services. The following gives an indication of how much uplink capacity is left for the data services, considering the loading limit. The following figures are calculated considering, that all left capacity is used for one data service only.

27

PS 64/64

mEN

MLoadUserperTraffic

PSusers

uplinkPSpolePS 4.9

223102107.0

__64/64_

,64/64,64/64 �

��

�� .

Which gives the following Data Volume per Hour:

MbitKbitKbpsRateRAB

UserperTrafficKbitHourperVolumeData PSPS

2.276.21653600640094.0(sec)3600)(_

__)(___ 64/6464/64

��

��

PS 64/128

mEN

MLoadUserperTraffic

PSusers

uplinkPSpolePS 4.9

223102107.0

__128/64_

,128/64,128/64 �

��

�� .


MbitKbitKbpsRateRAB


2.276.21653600640094.0(sec)3600)(_

__)(___ 128/64128/64

��

��

PS 64/384

mEN

MLoadUserperTraffic

PSusers

uplinkPSpolePS 4.9

223102107.0

__384/64_

,384/64,384/64 �

��

�� .


MbitKbitKbpsRateRAB


2.276.21653600640094.0(sec)3600)(_

__)(___ 384/64384/64

��

��

Downlink Cell Capacity Considering downlink load limit (75%) and downlink Mpole value, which is equal to 110 [7] for speech in indoor environment, one can say, that uplink is the limiting link for speech service capacity wise. Then offered traffic generated by speech users per cell in the downlink is the same as in the uplink and equals to:

ErlErl 3.221.0223 �� According to equations (2.8) and (2.9) average downlink loading generated 223 speech users is equal to:

%17.20%100110

)005.01(3.22

,,

, ��

��downlinkspeechpole

speechaverageaverage M

ALoad

Obtained average Load value is lower than the downlink loading limit by ~55%. Roughly speaking, this means that a UMTS cell can serve 223 speech subscribers, and there is still a lot of free downlink capacity for best effort packet switched services.

28

The following gives an indication of how much downlink capacity is left for the data services, considering the loading limit. The following figures are calculated considering, that all left capacity is used for one data service only. PS 64/64

mEN

MLoadUserperTraffic

PSusers

downlinkPSpolePS 3.44

223185486.0

__64/64_

,64/64,64/64 �

��

�� .


MbitKbitKbpsRateRAB


2.1072.102063600640443.0(sec)3600)(_

__)(___ 64/6464/64

��

��

PS 64/128

mEN

MLoadUserperTraffic

PSusers


22395486.0

__128/64_

,128/64,128/64 �

��

�� .


MbitKbitKbpsRateRAB


2.1068.1018336001280221.0(sec)3600)(_

__)(___ 128/64128/64

��

��

PS 64/384

mEN

MLoadUserperTraffic

PSusers


22325486.0

__384/64_

,384/64,384/64 �

��

�� .


MbitKbitKbpsRateRAB


8.676.677336003840049.0(sec)3600)(_

__)(___ 384/64384/64

��

��

2.3 CELL CAPACITY DIMENSIONING FOR GSM 2.3.1 Traffic estimation The same speech traffic per subscriber as for UMTS will be considered (see Section 2.2.1). 2.3.2 Grade of service The same GoS (see Section 2.1) for speech service as for UMTS will be considered (see Section 2.2.2).

29

2.3.3 Capacity calculations GSM belongs to TDMA systems. Each GSM radio carrier is divided into 8 time slots. Every time slot represents a Physical Channel. Generally, when dimensioning capacity of the GSM cell, the number of required Transceiver Units (TRU or TRX) is calculated. TRU is a part of GSM Radio Base Station, responsible for radio transmitting, radio receiving, power amplification and signal processing. One TRU includes all functionality needed for handling one radio carrier (i.e. the 8 time slots in one TDMA frame). The following initial data is used:

�� 223 subscribers per busy hour; �� Cell control channel configuration: BCCH + 2 SDCCH/8; �� One timeslot is reserved for GPRS; �� Speech traffic generated by office user is equal to 100 mE; �� GoS = 0.5 %.

With the assumed configuration we get 22.3 Erlangs of offered traffic. According to Erlang B table the required number of traffic channels is 34. The required number of TRUs is equal to:

.575.48

4348

___,

��

�

��

�ChannelsGPRSChannelsControlChannelsTraffic

N GSMTRUs

2.4 CONCLUSION OF CAPACITY DIMENSIONING The calculations above show that capacity-wise a UMTS cell can provide resources to support the same speech traffic as a GSM cell configured with 5 TRUs. Considering loading limitations there is a capacity margin for data services. In a design a number of necessary DAS cells should always be estimated. If the system appears to be capacity limited it is not practical to consider GSM cells dimensioned with more than 5 TRUs, as in this case WCDMA system connected to the same DAS will not provide necessary capacity resources.

30

3 IN-BUILDING SYSTEM CELL COVERAGE DIMENSIONING As it was mentioned above, IBS under consideration allows simultaneous multi-band and multi-operator access. Several mobile telephony systems can be combined into the same antenna system: GSM 900, GSM 1800 and UMTS. All these systems use different frequency bands for operation and moreover – GSM and UMTS use completely different technologies for transmitting data over the air. This means that mentioned systems have different requirements for received signal strength and attenuation in the radio propagation environment is also different. When dimensioning coverage of the systems these aspects have to be considered. The methodology used is as follows. Maximum path loss is calculated using the Received Signal Strength requirement for every system. Then the limiting system (the one with the lowest allowable path loss and the lowest cell range accordingly) is determined and the results for this system are defined to be the design criteria. Before proceeding with coverage dimensioning, radio wave propagation in indoor environment will be discussed in the following section 3.1. 3.1 RADIO WAVE PROPAGATION PREDICRION AND MODELING The mechanisms behind electromagnetic wave propagation are diverse, but can generally be attributed to reflection, diffraction and scattering. Most cellular radio systems operate in urban areas where there is no direct line-of-sight path between the transmitter and the receiver, and where the presence of high-rise buildings causes severe diffraction loss. Due to multiple reflections from various objects, the electromagnetic waves travel along different paths of varying lengths. The interaction between these waves causes multipath fading at a specific location, and the strengths of the waves decrease as the distance between the transmitter and the receiver increases. Propagation models have traditionally [21] focused on predicting the average received signal strength at a given distance from the transmitter, as well as the variability of the signal strength in close spatial proximity to a particular location. Propagation models that predict the mean signal strength for an arbitrary transmitter-receiver separation distance are useful in estimating the radio coverage area of a transmitter and are called large-scale propagation models, since they characterize signal strength over relatively large transmitter-receiver separation distances. On the other hand, propagation models that characterize the rapid fluctuations of the received signal strength over very short travel distances (a few wavelengths) or short time durations (on the order of seconds) are called small-scale or fading models. As mobile moves over very small distances, the instantaneous received signal strength may fluctuate rapidly giving rise to small-scale fading. The reason for this is that the received signal is a sum of many contributions coming from different directions. Since the phases are random, the sum of the contributions varies widely; for example, obeys a Rayleigh fading distribution. In small-scale fading, the received signal power may vary by as much as three or four orders of magnitude (30 or 40 dB) when the receiver is moved by only a fraction of a wavelength. As the mobile moves away from the transmitter over much larger distances, the local average received signal will gradually decrease, and it is this local average signal level that is predicted by large-scale propagation models. Typically, the

31

local average received power is computed by averaging signal measurements over a measurement track of 5� to 40� [21]. Figure 3.1 illustrates small-scale fading and the more gradual large-scale variations for an indoor radio communication system. Notice in the figure that the signal fades rapidly (small-scale fading) as the receiver moves, but the local average signal changes much more gradually with distance. Section 3.1.2 covers large-scale propagation and presents a number of common methods used to predict received power in indoor environment. Section 3.1.3 treats small-scale fading in the mobile radio environment. Figure 3.1 Small-scale and large-scale fading [21]. 3.1.1 The three main propagation mechanisms Reflection, diffraction and scattering are the three main propagation mechanisms, which impact propagation in a mobile communication system. These mechanisms are briefly explained in this section. Received power (or its reciprocal, pathloss) is generally the most important parameter predicted by large-scale propagation models based on the physics of reflection, scattering and diffraction. Small-scale fading and multipath propagation (discussed in section 3.1.3) may also be described by the physics of these three basic propagation mechanisms. 3.1.1.1 Reflection Reflection occurs when a propagating electromagnetic wave impinges upon an object, which has very large dimensions when compared to the wavelength of the propagating wave. Reflections occur from the surface of the earth and from buildings and walls. When a radio wave propagating in one medium impinges upon another medium having different electrical properties, the wave is partially reflected and partially transmitted. If the plane wave is incident on a perfect dielectric, part of the energy transmitted into the second medium and part of the energy is reflected back into the first medium, and there is no loss of energy in absorption. If the second medium is a perfect conductor, then all incident

32

energy is reflected back into the first medium without loss of energy. The electric field intensity of the reflected and transmitted waves may be related to the incident wave in the medium of origin through the Fresnel reflection coefficient � �� [21]. The reflection coefficient is a function of the material properties, and generally depends on the wave polarization, angle of incidence, and the frequency of the propagating wave. 3.1.1.2 Diffraction Diffraction occurs when the radio path between the transmitter and receiver is obstructed by a surface that has sharp irregularities (edges). The secondary waves resulting from the obstructing surface are present throughout the space and even behind the obstacle, giving rise to a bending of waves around the obstacle, even when a line-of-site path does not exist between transmitter and receiver. At high frequencies, diffraction, like reflection, depends on the geometry of the object, as well as the amplitude, phase, and polarization of the incident wave at the point of diffraction [21]. Diffraction allows radio signals to propagate around the curved surface of the earth, beyond the horizon, and to propagate behind obstructions. Although the received field strength decreases rapidly as a receiver moves deeper into the obstructed (shadowed) region, the diffraction field still exists and often has sufficient strength to produce a useful signal. The phenomenon of diffraction can be explained by Huygen’s principle, which states that all points on a wavefront can be considered as point sources for the production of secondary wavelets, and that these wavelets combine to produce a new wavefront in the direction of propagation. Diffraction is caused by the propagation of secondary wavelets into a shadowed region. The field strength of a diffracted wave in the shadowed region is the vector sum of the electric field components of all the secondary wavelets in the space around the obstacle [21]. 3.1.1.3 Scattering Scattering occurs when the medium through which the wave travels consists of objects with dimensions that are small compared to the wavelength, and where the number of obstacles per unit volume is large. Scattered waves are produces by rough surfaces, small objects, or by other irregularities in the channel [21]. The actual received signal in a mobile radio environment is often stronger than what is predicted by reflection and diffraction models alone. This is because when a radio wave impinges on a rough surface, the reflected energy is spread out (diffused) in all directions due to scattering. Objects such as lampposts and trees tend to scatter energy in all directions, thereby providing additional radio energy at a receiver [21]. 3.1.2 Large-scale indoor propagation models The indoor radio channel differs from the traditional mobile radio channel in two aspects – the distances covered are much smaller, and the variability of the environment is much greater for a much smaller range of transmitter-receiver separation distances. It has been

33

observed that propagation within buildings is strongly influences by specific features such as the layout of the building, the construction materials, and the building type. This section outlines models for path loss within buildings. Indoor radio propagation is dominated by the same mechanisms, which were described earlier: reflection, diffraction, and scattering (see Section 3.1.1). However, conditions are much more variable. For example, signal levels vary greatly depending on whether interior doors are open or closed inside a building. Where antennas are mounted also impacts large-scale propagation. Antennas mounted at desk level in a partitioned office receive vastly different signals than those mounted on the ceiling. Also, the smaller propagation distances make it more difficult to ensure far-field radiation for all receiver locations and types of antennas. The field of indoor radio propagation is relatively new, with the first wave of research occurring in the early 1980s. Cox [4] at AT&T Bell Laboratories and Alexander [1] at British Telecom were the first to carefully study indoor path loss in and around a large number of homes and office buildings. Excellent literature surveys are available on the topic of indoor propagation [19], [12]. In general, indoor channels may be classified either as line-of-sight or obstructed, with varying degrees of clutter [20]. Some of the key models are presented in the following sections. Attending to their nature, the propagation models can be classified [9] as:

�� Empirical models; �� Semi empirical or semi deterministic models; �� Deterministic models.

The empirical models are described by equations derived from statistical analysis of a large number of measurements. These methods are simple and do not require detailed information about the environment. They are also easy and fast to apply because the estimation is usually obtained from closed expressions. On the other hand, they cannot provide a very accurate estimation of the path loss. The deterministic models are based on the application of well-known electromagnetic techniques to a site-specific description of the environments. The environmental description is obtained from the building databases. From these data, one can obtain a description of the scenario in terms of canonical or primitive entities, which can be managed efficiently by the electromagnetic theory used. Different degrees of accuracy can be found in the environmental description. Most of the deterministic models are based on ray-tracing electromagnetic methods. Finally, semi empirical or semi deterministic models are based on the equations derived from the application of deterministic methods to generic indoor models. Sometimes, the equations have been corrected experimentally in order to improve their agreement with the measurements. The resulting equations are functions of the characteristics of the areas surrounding the antennas and certain specific characteristics of the scenario. These methods require more detailed information about the environment than the empirical methods but not as much as the deterministic models. With the required data known, they

34

are easy and fast to apply because, as with the empirical models, the results are obtained from closed expressions. This thesis focuses on empirical and semi empirical propagation models. 3.1.2.1 Free space propagation model The free space propagation model is used to predict received signal strength when the transmitter and receiver have a clear, unobstructed line-of-sight path between them. As with most large-scale radio wave propagation models, the free space model predicts that received power decays as a function of the transmitter-receiver separation distance raised to some power (i.e. a power law function). The free space power received by a receiver antenna, which is separated from a radiating transmitter antenna by a distance d, is given by the Friis free space equation [18]:

3.1 � � 22

2

4 d

GGPP rtt

r�

�� (3.1)

where tP - transmitted power,

rP - received power, � - wavelength, d - distance between transmitting and receiving antenna,

tG - transmitting antenna gain,

rG - receiving antenna gain. The gain of the antenna is is related to its effective aperture, Ae, by [21]:

3.2 2

4�� eA

G � . (3.2)

The effective aperture Ae is related to the physical size of the antenna, and � is related to the carrier frequency by [21]:

3.3 c

cfc

��

�2

�� (3.3)

where f - carrier frequency in Hertz,

c� - carrier frequency in radians per second, c - speed of light given in meters/s. The Friis free space equation (3.1) shows that the received power falls off as the square of the transmitter-receiver separation distance. This implies that the received power decays with distance at a rate of 20 dB/decade.

35

An isotropic radiator is an ideal antenna, which radiates power with unit gain uniformly in all directions, and is often used to reference antenna gains in wireless systems. The effective isotropic radiated power (EIRP) is defined as:

3.4 ttGPEIRP � (3.4)

and represents the maximum radiated power available from a transmitter in the direction of maximum antenna gain, as compared to an isotropic radiator. In practice, effective radiated power (ERP) is also used to denote the maximum radiated power as compared to a half-wave dipole antenna (instead of an isotropic antenna). Since a dipole antenna has a gain of 1.64 (2.15 dB above an isotropic), the ERP will be 2.15 dB smaller that the EIRP for the same transmission system [21]. In practice, antenna gains are given in units of dBi (dB gain with respect to an isotropic antenna) or dBd (dB gain with respect to a half-wave dipole). Path loss or propagation loss pathL is the parameter commonly used to characterize the local average signal in mobile channels. It is defined as the relationship between the transmitted power of the transmitter antenna and the received power by the receiver antenna. Path loss may or may not include the effect of the antenna gains. In most cases it is expressed in dB, so [18]:

3.5 )()(log10 dBPdBPPP

L rtr

tpath �� . (3.5)

Therefore combining equations (3.1) and (3.5), the path loss for free-space propagation when antenna gains are included becomes:

3.6 � �

rtrt

path GGfdKGG

dL log10log10log20log20

4log10 2

22

��

� �

��

��

(3.6)

where f - frequency, K - constant that depends on the units of d and f (for example, when d is in

kilometers and f is in megahertz, K = 32.44. When the antenna gains are not included in the path loss definition, (3.6) reduces to:

3.7 � �

fdKd

Lpath log20log204

log10 2

22

��

� �

��

��

. (3.7)

The Friis free space model is only a valid predictor for Pr for values of d, which are in the far-field of the transmitting antenna. The far-field, or Fraunhofer region, of a transmitting antenna is defined as the region beyond the far-field distance df, which is related to the largest linear dimension of the transmitter antenna aperture and the carrier wavelength. The Fraunhofer distance is given by [21]:

36

3.8 �

22Dd f � (3.8)

where D - largest physical linear dimension of the antenna. Additionally, to be in the far-field region, df must satisfy [21]:

Dd f and

�fd . Furthermore, it is clear that equation (3.1) does not hold for d = 0. For this reason, large-scale propagation models use a close-in distance, d0, as a known received power reference point. The received power, Pr, at any distance d > d0, may be related to Pr at d0. The value Pr(d0) may be predicted from equation (3.1), or may be measured in the radio environment by taking the average receiver power at many points located at a close-in radial distance d0 from the transmitter. The reference distance must be chosen such that it lies in the far-field region, that is, ,0 fdd � and d0 is chosen to be smaller than any practical distance used in the mobile communication system. Thus, using equation (3.1), the received power in free space at a distance greater that d0 is given by [21]:

3.9 frr ddddd

dPdP ��

� �

�� 0

20

0 )()( . (3.9)

In mobile radio systems, one can find that Pr may change by many orders of magnitude over a typical coverage area. Because of the large dynamic range of received power levels, often dBm or dBW units are used to express received power levels. Equation (3.9) may be expressed in units of dBm or dBW by simply taking the logarithm of both sides and multiplying by 10. For example, if Pr is in units of dBm, the received power is given by:

3.10 � � � �f

rr ddd

dddP

dP ��

� �

��

�

��

�� 0

00 log20W001.0

log10dBm (3.10)

where Pr(d0) is in units of watts. The reference distance d0 for practical systems using low-gain antennas in the 1-2 GHz region is typically chosen to be 1 m in indoor environments, so that the numerator in equations (3.9) and (3.10) is a multiple of 10. This makes path loss computations easy in dB units. 3.1.2.2 Partition losses (same floor) Buildings have a wide variety of partitions and obstacles, which form the internal and external structure. Houses typically use a wood frame partition with plaster foard to form

37

the internal walls and have wood or non-reinforced concrete between floors. Office buildings, on the other hand, often have large open areas (open plan), which are constructed by using moveable office partitions so that the space may be reconfigured easily, and use metal reinforced concrete between floors. Partitions that are formed as part of the building structure are called hard partitions, and partitions that may be moved and which do not span to the ceiling are called soft partitions. Partitions vary widely in their physical and electrical characteristics, making it difficult to apply general models to specific indoor installations. Nevertheless, researchers have formed extensive databases of losses for a great number of partitions, some examples are shown in Table 3.1. Table 3.1 Average signal loss measurements reported by various researchers for radio paths obstructed by common building material [21].

Material type Loss (db) Frequency All metal 26 815 MHz

Aluminum siding 20.4 815 MHz Foil insulation 3.9 815 MHz

Concrete block wall 13 1300 MHz Fade observed when transmitter turned a right angle

corner in corridor 10-15 1300 MHz

Light machinery < 1 m2 1-4 1300 MHz General machinery 1-2 m2 5-10 1300 MHz Heavy machinery > 2 m2 10-12 1300 MHz

Metal stairs 5 1300 MHz Light textile inventory 3-5 1300 MHz Heavy textile inventory 8-11 1300 MHz

Metallic inventory 4-7 1300 MHz Metallic inventory racks 4-9 1300 MHz

Empty cardboard inventory boxes 3-6 1300 MHz Concrete block wall 13-20 1300 MHz

Ceiling duct 1-8 1300 MHz 2.5 m storage rack with small metal parts (loosely

packed) 4-6 1300 MHz

4 m metal box storage 10-12 1300 MHz 5 m storage rack with paper products (loosely packed) 2-4 1300 MHz

5 m storage rack with large paper products (tightly packed)

6 1300 MHz

5 m storage rack with large metal parts (tightly packed)

20 1300 MHz

Semi-automated assembly line 5-7 1300 MHz 0.6 m square reinforced concrete pillar 12-14 1300 MHz

Stainless steel piping 15 1300 MHz Concrete wall 8-15 1300 MHz

3.1.2.3 Partition losses between floors The losses between floors of a building are determined by the external dimensions and materials of the buildings, as well as type of construction used to create the floors and the external surroundings [26], [25]. Even the number of windows in a building and the presence of tinting (which attenuates radio energy) can impact the loss between floors.

38

Table 3.2 illustrates values for floor attenuation factors (FAF) in two buildings. It can be seen that the attenuation between one floor of the building is greater than the incremental attenuation caused by each additional floor. After about five or six floor separations, very little additional path loss is experienced. Table 3.2 Average floor attenuation factor in dB in two office buildings [21].

3.1.2.4 Log-distance path loss model This is an empirical model that estimates the path loss (in decibels) as follows [18]:

3.11

�Xdd

ndLL pathpath ��

� �

��

00 log10)(

(3.11)

where d - transmitter-receiver separation,

)( 0dLpath - reference path loss,

0d - distance at which path loss is referred to (in meters), n - path loss exponent, which depends on the environment,

�X - normal random variable with a standard deviation of � (in decibels). The reference path loss can be measured or calculated using free-space path loss expression. The term �X accounts for the environmental clutter. The parameters of the model [ )( 0dLpath , n and � ] depend on the characteristics of the scenario. Several types of indoor environments in LOS and NLOS situations have been analyzed and, for each case, values at different frequencies have been obtained from measurements [2]. At several frequencies between 900 and 4000 MHz, the values of n varied between 1.6 and 3.3 and the values of � between 3.0 and 14.1 dB. Due to its simplicity, this model has been widely used in indoor environments. Also, it has been used in outdoor microcell environments [8]. 3.1.2.5 Log-normal shadowing Measurements have shown that at any value of d, the path loss Lpath at a particular location is random and distributed log-normally (normal in dB) about the mean distance-dependent value [5], [3]. In model in equation (3.11) �X accounts for this fact. �X is a zero-mean Gaussian distributed random variable (in dB) with standard deviation � (also in dB).

Building FAF (dB) Office building 1: Through one floor 12.9 Through two floors 18.7

Through three floors 24.4 Through four floors 27.0 Office building 2: Through one floor 16.2 Through two floors 27.5

Through three floors 31.6

39

The log-normal distribution describes the random shadowing effects, which occur over a large number of measurement locations which have the same transmitter-receiver separation, but have different levels of clutter on the propagation path. This phenomenon is referred to as log-normal shadowing. Simply put, log-normal shadowing implies that measured signal levels at a specific transmitter-receiver separation have a Gaussian (normal) distribution about the distance-dependent mean of equation (3.11), where the measured signal levels have values in dB units. The standard deviation of the Gaussian distribution that describes the shadowing also has units in dB. Thus, the random effects of shadowing are accounted for using the Gaussian distribution. The close-in reference distance d0, the path loss exponent n, and the standard deviation � , statistically describe the path loss model for an arbitrary location having a specific transmitter-receiver separation. In practice, the values of n and � are computed from measured data, using linear regression such that the difference between the measured and estimated path losses is minimized in a mean square error sense over a wide range of measurement locations and transmitter-receiver separations. The value of )( 0dLpath in equation (3.11) is based on either close-in measurements or on a free space assumption from the transmitter to d0. An example of how the path loss exponent is determined from measured data follows. Figure 3.2 illustrates actual measured data in several cellular radio systems and demonstrates the random variations about the mean path loss (in dB) due to shadowing at specific transmitter-receiver separations. Figure 3.2 Scatter plot of measured data and corresponding MMSE path loss model for many cities in Germany [24].

40

Since Lpath is a random variable with a normal distribution in dB about the distance-dependent mean, so is Pr. The probability density of Pr (with normal distribution in dB) can be described as [7]:

3.12 ��

��

� �� 2

2

2)(

exp2

1)(

��rr

r

PPPp . (3.12)

3.1.2.6 Determination of percentage of coverage area It is clear that due to random effects of shadowing, some locations within a coverage area will be below a particular desired received signal threshold. It is often useful to compute how the boundary coverage relates to the percent of area covered within the boundary. For a circular coverage area having radius R from transmitter antenna, let there be some desired received signal threshold� . We are interested in computing )(�U , the percentage of useful service area (i.e. the percentage of area with a received signal that is equal or greater than � ), given a known likelihood of coverage at the cell boundary. Letting d = r represent the radial distance from the transmitter, it can be shown that if ])(Pr[ �rPr is the probability that the random received signal at d = r exceeds the threshold � within an incremental area dA, then )(�U can be found by [15]:

3.13

! !" "" ��

��

��

�2

0 022 �)(Pr

1)(Pr

1)(

R

rr drdrrPR

dArPR

U . (3.13)

!�)(Pr rPr can be calculated by integrating equation (3.12):

3.14

.2

))]/log(10)(([

21

21

2

)(21

21

])(Pr[

00

�

�

�

� ��

��

� �

� ��

�

�

�

��

drdLPerf

rPerfrP

patht

rr

(3.14)

In order to determine the path loss as referenced to the cell boundary (r = R), it is clear that:

)(log10log10)( 00

dLRr

ndR

nrL pathpath ��

� �

��

� �

��

and equation (3.14) may be expressed as:

41

3.15 .2

))]/log(10)/log(10)(([

21

21

])(Pr[ 00

�

�

�

� ��

�

��

RrdRdLPerfrP patht

r (3.15)

If we let 2/))/log(10)((( 00 �� dRdLPa patht �� and 2/)log10( �enb � , then:

3.16

.ln1

21

)(0

2 " �

� �

� ��R

drRr

baerfrR

U � (3.16)

By substituting )/log( Rrbat �� in equation (3.16), it can be shown that:

3.17 �

� �

��

��

��

� �

� ��

�

� �

� ��

bab

erfb

abaerfU

11

21exp)(1

21

)( 2� . (3.17)

)(�U denotes the probability of having a received signal strength value above a certain

threshold within an area, i.e. )(�U means the area coverage. )(�P will then denote the probability of having a received signal strength value above a certain threshold on the cell border, i.e. )(�P means the perimeter coverage. )(�P is obtained by substituting r= R into equation (3.15):

3.18 .2

))]/log(10)(([

21

21

)( 00

�

�

�

� ��

�

��

dRdLPerfP patht (3.18)

By choosing the signal level such that ��)(RPr (i.e. a = 0), equation (3.17) can be shown to be:

�

� �

��

��

��

� �

��

� �

��b

erfb

U1

11

exp121

)( 2�

Equation (3.17) may be evaluated for a large number of values of � and n, as shown in Figure 3.3 [22] For example, if n = 4 and 8�� dB, and if the boundary is to have 75% boundary coverage (75% of the time the signal is to exceed the threshold at the boundary), then the area coverage is equal to 90%. If n = 2 and 8�� dB, a 75% boundary coverage provides 86% area coverage. If n = 3 and 9�� dB, then 50% boundary coverage provides 71% area coverage.

42

Figure 3.3 The area coverage as a function of �/n for different values of the perimeter coverage. 3.1.2.7 Attenuation factor model This is an empirical indoor model [25] that predicts the propagation path loss (in decibels) on the same floor or through different floors. The path loss is given by:

3.19 FAF

dd

ndLL pathpath ��

� �

��

00 log10)(

(3.19)

where n - path loss exponent for observers located on the same floor, FAF - represents the floor attenuation (in decibels), which accounts for the

propagation through different floors. For propagation through multiple floors, when the path loss exponent 1n is known, an alternative expression is:

3.20 .log10)(

010

�

� �

��

dd

ndLL pathpath (3.20)

Experimental values of n, n1 and FAF for various types of buildings can be found in [25]. As an example, the values of the floor attenuation factor (in decibels) varied between 12.9 and 16.2 for transmission through one floor, between 18.7 and 27.5 through two floors, and between 24.4 and 31.6 through three floors. A typical value for n is 2.8. The values of n1 varied between 4.19 and 5.22. This model is also known as the one-slope model because it assumes that the path loss depends, linearly, on the logarithm of the separation between transmitter and receiver.

43

A modification of the model is proposed in [6] where the path loss is given by:

3.21

FAFddd

dLL pathpath ��

� �

�� #

00 log10)(

(3.21)

where # - attenuation factor determined experimentally for different indoor

environments.

This model is known as the linear-slope model because the logarithmic path loss depends linearly on the separation between transmitter and receiver. Experimental values from [6] at 850, 1700 and 4000 MHz show a variation in the attenuation factor between 0.62 and 0.47 dB/m for a four-story building and between 0.48 and 0.23 dB/m for a two-story building. 3.1.2.8 Keenan-Motley model A more sophisticated method that considers the attenuation through the individual walls and floors is given by [17]:

3.22 ��

��J

jwjwj

I

ififipathpath LkLkdnLL

110, log10 (3.22)

where 0,pathL - attenuation at the reference distance (1 meter),

n - path loss exponent, d - distance transmitter-receiver,

fiL - attenuation through floors of the type i,

fik - number of floors of the type i between the transmitter and the receiver,

wjL - attenuation through walls of type j,

wjk - number of walls of the type j between the transmitter and the receiver.

In this model, 0,pathL and n tend to be the values of the free-space conditions

( 2,00, �� nLpath ). Typical values for the attenuation through floors can be found in Table

3.2 and [25]. The values of the attenuation through walls depend strongly on the type of partitions used (see Table 3.1 for more details). 3.1.2.9 Modified Keenan-Motley model Another way of using Keenan-Motley model is presented in equation (3.23) [7]:

3.23 )(log20log2044.32 bwwffpath ddDLkLkfdL �� (3.23)

where f - frequency (MHz), d - distance transmitter-receiver (km),

44

fL - attenuation through floor,

fk - number of floors between the transmitter and the receiver,

wL - attenuation through walls,

wk - number of walls between the transmitter and the receiver, D - linear attenuation factor (dB/m) (for distances above breakpoint, 0.2 dB/n

is typically added), db - indoor breakpoint (m) (typically 65 m). First part of equation (3.23) represents the free space formula. Free space loss plus attenuation in walls can be further simplified, as shown in Figure 3.4. Figure 3.4 Diagrams showing approximate indoor propagation loss at 900 MHz for a wall attenuation of 0.2 & 0.5 dB/m [7].

45

At 1800 MHz the loss is 6 dB higher (see free-space formula, equation (3.7)), thus 50 has to be changed to 56 in the formula. At 2100 MHz the loss is 7.4 dB higher, thus 50 has to be changed to 57.4 in the formula. A simplified way of estimating the loss between floors is shown in the diagram in Figure 3.5. Figure 3.5 Loss between floors [7]. 3.1.3 Small-scale indoor propagation models Small-scale fading, or fast fading, is used to describe the rapid fluctuations of the amplitudes, phases, or multipath delays of a radio signal over a short period of time or travel distance, so that large-scale path loss effects may be ignored. Fading is caused by interference between two or more versions of the transmitted signal, which arrive at the receiver at slightly different times. These waves, called multipath waves, combine at the receiver antenna to give a resultant signal, which can vary widely in amplitude and phase, depending on the distribution of the intensity and relative propagation time of the waves and the bandwidth of the transmitted signal. Multipath in the radio channel creates small-scale fading effects. The three most important effects are [21]:

�� Rapid changes in signal strength over a small travel distance or time interval; �� Random frequency modulation due to varying Doppler shifts on different multipath

signals; �� Time dispersion (echoes) caused by multipath propagation delays.

Usually fading occurs because there is no single line-of-sight path to the base station antenna. Even when a line-of-sight exists, multipath still occurs due to reflections from the surrounding structures. The incoming radio waves arrive from different directions with different propagation delays. The signal received by the mobile at any point in space may consist of a large number of plane waves having randomly distributed amplitudes, phases,

46

and angles of arrival. These multipath components combine vectorially at the receiver antenna, and can cause the signal received by the mobile to distort or fade. Even when a mobile receiver is stationary, the received signal may fade due to movement of surrounding objects in the radio channel. If objects in the radio channel are static, and motion is considered to be only due to that of the mobile, then fading is purely a spatial phenomenon. The spatial variations of the resulting signal are seen as temporal variations by the receiver as it moves through the multipath field. Due to constructive and destructive effects of multipath waves summing at various points in space, a moving receiver can pass through several fades in a small period of time. In a more serious case, a receiver may stop at a particular location at which the received signal is in deep fade. Maintaining good communications can then become very difficult, although passing people walking in the vicinity of the mobile can often disturb the field pattern, thereby diminishing the likelihood of the received signal remaining in a deep null for a long period of time. Figure 3.1 shows typical rapid variations in the received signal level due to small-scale fading as a receiver is moved over a distance of a few meters. Due to the relative motion between the mobile and the base station, each multipath wave experiences an apparent shift in frequency. The shift in received signal frequency due to motion is called the Doppler shift, and is directly proportional to the velocity and direction of motion of the mobile with respect to the direction of arrival of the received multipath wave. Hashemi claims in [13] that since “rapid motions and high velocities typical of the mobile user are absent in the indoor environment […] Doppler shift is therefore negligible.” [14] In order to gain more understanding of the multi-path fading phenomena, it is necessary to study the mathematical principles behind it. In this section, two mathematic models are discussed. These two mathematic models are Rayleigh and Rician Fading Models. The multi-path fading phenomena can result in a signal traversing different paths and combined to form a signal with different amplitude than the intended signal as equation (3.24) suggests. The received signal amplitude V(x) is often irregular, but it can be modeled as a random variable to help us predict the system’s performance. To facilitate the understanding of this random variable, it is often normalized in the following manner [16]:

3.24 !)()(xVE

xVr � (3.24)

where E[V(x)] - the average of V(x) received. Although, V(x) varies differently in slow fading, in a much smaller scale fast fading is much more significant and E[V(x)] can be considered as a constant. The following two mathematical models are suitable for different scenarios of multi-path fading.

47

3.1.3.1 Rayleigh fading model In mobile radio channels, the Rayleigh distribution is commonly used to describe the statistical time varying nature of the received envelope of a fading signal, or the envelope of an individual multipath component. Figure 3.6 shows a Rayleigh distributed signal envelope as a function of time. The Rayleigh distribution has a probability density function given by [21]:

3.25 � �

� �$%

$&

'

(

)**�

� �

��

�

00

02

exp)( 2

2

2

r

rrr

rp �� (3.25)

where � - the RMS value of the received voltage signal before envelope detection

2� - time-average power of the received signal before envelope detection. Figure 3.6 A typical Rayleigh fading envelope at 900 MHz [10].

48

Figure 3.7 illustrates the Rayleigh pdf. Figure 3.7 Rayleigh probability density function (pdf) [21]. 3.1.3.2 Ricean fading model When there is a dominant stationary (nonfading) signal component present, such as a line-of-sight propagation path, the small-scale fading envelope distribution is Ricean. In such a situation, random multipath components arriving at different angles are superimposed on a stationary dominant signal. At the output of an envelope detector, this has the effect of adding a dc component to the random multipath. Just as for the case of detection of a sine wave in thermal noise [23], the effect of a dominant signal arriving with many weaker multipath signals gives rise to Ricean distribution. As the dominant signal becomes weaker, the composite signal resembles a noise signal, which has an envelope that is Rayleigh. Thus, the Ricean distribution degenerates to a Rayleigh distribution when dominant component fades away. The Ricean distribution is given by [21]:

3.26 � �

� �$%

$&

'

(

��

� �

��

� �

� ��

�

00

0,02

exp)( 202

22

2

r

rAA

IArr

rpr

�� (3.26)

where A - denotes the peak amplitude of the dominant signal

)(0 �I - modified Bessel function of the first kind and zero-order. The Ricean distribution is often described in terms of a parameter K, which is defined as the ratio between the deterministic signal power and the variance of the multipath. It is given by � �22 2/ �AK � or, in terms of dB [21]:

3.27 2

2

2log10

�A

K � dB (3.27)

49

The parameter K is known as Ricean factor and completely specifies the Ricean distribution [21]. As �)++ KA ,0 dB, and as the dominant path decreases in amplitude, the Ricean distribution degrades to a Rayleigh distribution. Figure 3.8 shows the Ricean pdf ( �)�K dB (Rayleigh) and 6�K dB. For K >> 1, the Ricean pdf is approximately Gaussian about the mean probability density function (pdf)). Figure 3.8 Probability density function of Ricean distributions [21]. 3.2 CELL COVERAGE DIMENSIONING FOR GSM When planning a system it is not sufficient to use mobile station or base station sensitivity level as a planning criterion. Various margins have to be added in order to obtain the desired coverage. In this section these margins are discussed and the cell coverage dimensioning criteria for GSM are presented. 3.2.1 Definitions Required signal strength To the sensitivity level of an MS, margins have to be added to compensate for Rayleigh fading, interference and body loss. The obtained signal strength is what is required to perform a phone call in a real-life situation and will be referred to as SSreq and can be expressed as [7]:

3.28 BLIFRFMSSS mmsensreq �� argarg (3.28)

where Mssens - MS sensitivity (-104 dBm [7]), RFmarg - Rayleigh fading margin, Section 3.2.2, IFmarg - interference margin, Section 3.2.2, BL - body loss, Section 3.2.2. Design level Extra margin have to be added to SSreq to handle the log-normal fading (see Section 3.1.2.5). This margin depends on the environment and on the desired area coverage. The obtained signal strength is what should be used when planning the system and it will be

50

referred to as the design level, SSdesign. This signal strength is the value that should be obtained on the cell border. The design level can be calculated from:

3.29 argmreqdes LNFSSSS �� (3.29)

where LNFmarg - log-normal fading margin, Section 3.2.2. 3.2.2 Margins Rayleigh (or Ricean) fading margin Rayleigh (or Ricean) fading is due to multipath interference and occurs especially in environments where there is high probability of blocked sight between transmitter and receiver. The distance between two adjacent fading dips is approximately 2/� . Small scale fading was discussed in more detail in section 3.1.3. The required sensitivity performance of GSM in terms of FER, BER is specified for each type of channel and at different fading models (called channel models). The channel models reflect different types of propagation environment and different MS speeds. The sensitivity is measured under simulated Rayleigh fading conditions for all the different channel models and the sensitivity is defined as the level where the required quality performance is achieved. In a noise limited environment the sensitivity is the one mentioned in Section 3.2.1. This would mean that Rayleigh fading is already taken into consideration in the sensitivity definition. However, the GSM specification [11] allows worse quality for slow MSs (3 km/h) than for fast moving MSs. The sensitivity performance at fading conditions corresponding to an MS speed of 50 km/h in an urban environment (called Typical Urban, TU50), is in accordance with good speech quality, while the sensitivity performance for slow MSs at TU3 does not correspond to acceptable speech quality. In order to obtain good speech quality even for slow mobiles, an extra margin, RFmarg, is considered when planning. From experience, 3 dB margin seems adequate. In a frequency hopping system the Rayleigh fading dips are leveled out and there should be no need for a Rayleigh fading margin. But since a Broadcast Control Channel (BCCH) never hops, the Rayleigh fading margin is recommended in cell coverage estimations, regardless of using frequency hopping or not. Recommended Rayleigh fading margin (RFmarg) = 3 dB [7] Log-normal fading margin The signal strength value computed by wave propagation equations can be considered as a mean value of the signal strength in a small area with a size determined by the resolution and accuracy of the model. Assumed that the fast fading is removed, the local mean value of the signal strength fluctuates in a way not considered in the prediction algorithm. This deviation of the local mean in dB compared to the predicted mean has nearly a normal distribution. Therefore this variation is called log-normal fading (see Section 3.1.2.5).

51

The received signal strength is a random process and it is only possible to estimate the probability that the received signal strength exceeds a certain threshold. In the result from a prediction, 50% of the locations (for example at the cell borders) can be considered to have a signal strength that exceeds the predicted value. In order to plan for more than 50% probability of signal strength above the threshold � , a log-normal fading margin, LNFmarg, is added to the threshold during the design process. Jakes’ formulas A common way to calculate LNFmarg is to use Jakes’ formulas, shown in Section 3.1.2.5. LNFmarg can be expressed as [7]:

3.30 �� ))]/log(10)(([ 00arg dRdLPLNF pathtm

(3.30)

According to equation (3.18), the relation between the perimeter coverage, )(�P , and LNFmarg is written as:

3.31 .22

121

)( argarg

�

� �

�,�

�

� �

��

�� mm LNFLNF

erfP (3.31)

Here )(t, is the cumulative normal distribution function with its variable t. If )(�P is known, LNFmarg can thus be obtained by using [7]:

3.32 �� tLNFm arg (3.32)

The variable t can be found in a normal distribution function table. Figure 3.9 The cumulative normal distribution function [7]. If the perimeter coverage for example is 75 %, t will be 0.67, see Figure 3.9. For a standard deviation value of 8 dB, LNFmarg then will be 5.4 dB, according to equation (3.32). If a guaranteed availability of 75 % at the cell border and 90 % within the cell area is required,

52

an extra margin of 5.4 dB has to be added to the signal threshold, with the assumptions mentioned above. Simulation of log-normal fading margin in a multi-cell environment A disadvantage with Jakes’ formulas is that it does not take the effect of many servers into account. The presence of many servers at the cell borders will reduce the required log-normal fading margin. This is because the fading patterns of different servers are fairly independent. If the signal from one server fades down below the sensitivity level a neighbor cell can fill out the gap and rescue the connection. In order to find the log-normal fading margins in a multi-cell environment, simulations have been performed, see [7] for more details regarding simulation results. Interference margin The plain receiver sensitivity depends on the required carrier to noise ratio (C/N). When frequencies are reused, the received carrier power must be large enough to combat both noise and interference, that means C/(N+I) must exceed the receiver threshold. In order to get an accurate coverage prediction in a busy system, an interference margin (IFmarg) is defined. In areas with a high level of interference the required interference margin has to be higher, and in areas with low interference it has to be lower. The question is only how high and how low. In order to avoid the complex matter of measuring the level of interference and from these data calculate SSreq, a simplified design procedure is proposed. The idea is that by studying the surrounding radio environment it should be possible to classify the level of interference in the cell as low, medium or high and from this classification say what the required signal strength should be. In Table 3.3, proposed values of SSreq are given for each of the mentioned interference levels. Table 3.3 Required signal strengths in areas of different interference.

Level of interference Required level (SSreq) Low -85 dBm

Medium -75 dBm High -65 dBm

Now the procedure of determining the required signal strengths seems easy enough if one could just say what is the level of interference in the area in question. It is very difficult to give any general rules for how this should be determined. By studying how tight the frequency re-use is, how long the site-to-site distance is and how many sites that are within line-of-sight etc. Some sort of estimation should however be possible. Some “typical radio environments” representing the different levels are given in Table 3.4.

53

Table 3.4 Some typical examples of radio environments representing low, medium and high level of interference.

Low interference Medium interference High interference Area with a re-use factor

larger than 12 and the indoor cell is situated in a building

of the same height as the surroundings.

Area where the outdoor coverage is poor. Dedicated indoor frequencies with low

re-use.

Something in between.

Area with a frequency reuse of 9 or tighter and the indoor

cell is situated in a high building compared to the

surroundings.

This method of determining the required signal strength contains some uncertainties. However, compared to the case where the interference is measured a lot of work is saved. Interference measurements are troublesome and it is not obvious how to interpret the results. These results depend heavily on the present cell plan and may change substantially at re-planning. However in cases where the interference cannot be estimated, measurements can be used as a last resort. According to the test results [7] the C/I requirements for good speech quality (for business applications) are 18 dB (average for a spot in the cell) in a non-hopping system. For a frequency hopping system the C/I requirement can be decreased to16 dB. Body loss margin The human body has several effects on the MS performance compared to a freestanding mobile phone [7]:

1. The head absorbs energy. 2. The antenna efficiency of some MSs can be reduced. 3. Other effect may be a change of the lobe direction. This effect can be neglected in

the link budget since no mobile antenna gain is used. The body loss recommended by ETSI [11] is 3 dB for 900 and 1800 MHz. A need for a somewhat higher margin for 900 MHz has been indicated. 5 dB is the Ericsson recommendation [7]. 3.2.3 GSM 900 cell coverage dimensioning In this section GSM 900 link budget will be calculated in order to estimate the maximum allowable pathloss. The following initial data is considered:

�� Received Signal Strength requirement is assumed to be –75 dBm for an office building, which corresponds to the area with medium level of interference (see Section 3.2.2).

�� Coverage probability requirement is assumed to be 95% of the area.

54

�� EiRP at the antennas is assumed to be 10 dBm. With the values inserted and considering initial data equation (3.28) becomes:

dBm

dBdBdBdBmBLIFRFMSSS mmsensreq

75

5213104argarg

��

��

One can see that interference margin for GSM 900 in the area with medium level of interference is equal to 21 dB. In order to assure coverage probability of 95%, corresponding log-normal fading margin should be added. Simulated log-normal fading margin plots were used with 9)( �iLNF� [7].

With the values inserted equation (3.29) becomes:

dBmdBdBmLNFSSSS mreqdes 2.698.575arg ��

The maximum allowable path loss for the GSM 900 can be calculated as:

dBdBdBmSSEIRPL despath 2.792.6910(max) ��

3.2.4 GSM 1800 cell coverage dimensioning In this section GSM 1800 link budget will be calculated in order to estimate the maximum allowable pathloss. The following initial data is considered:

�� Received Signal Strength requirement is assumed to be –75 dBm for an office building, which corresponds to the area with medium level of interference (see Section 3.2.2).

�� Coverage probability requirement is assumed to be 95% of the area. �� EiRP at the antennas is assumed to be 10 dBm.

With the values inserted and considering initial data equation (3.28) becomes:

dBm

dBdBdBdBmBLIFRFMSSS mmsensreq

75

3233104argarg

��

��

One can see that interference margin for GSM 1800 in the area with medium level of interference is equal to 23 dB. In order to assure coverage probability of 95%, corresponding log-normal fading margin should be added. Simulated log-normal fading margin plots were used with 9)( �iLNF� [7]. With the values inserted equation (3.29) becomes:

dBmdBdBmLNFSSSS mreqdes 2.698.575arg ��

55

The maximum allowable path loss for the GSM 1800 can be calculated as:

dBdBdBmSSEIRPL despath 2.792.6910(max) �� 3.3 CELL COVERAGE DIMENSIONING FOR UMTS When planning a UMTS system, it is (similarly to GSM system) not sufficient to use mobile station or base station sensitivity level as a planning criterion. Various margins have to be added in order to obtain the desired coverage. Most of the margins are similar to those used in GSM cell coverage dimensioning. But some specific margins have to be defined for UMTS. In this section these margins, additional definitions and the cell coverage dimensioning criteria for UMTS are presented. It should be noted, that similarly to capacity, UMTS cell coverage has to be calculated separately for up- and downlink. Final cell coverage is defined by the limiting factor: whether uplink or downlink. 3.3.1 Definitions Noise rise The more loaded the system, the more interference will be generated. This will have the effect that the receiver noise floor is higher in a loaded system as compared to an unloaded system. The increase is often referred to as uplink noise rise or cell breathing and is denoted IUL. It can be calculated from the relative uplink system load as follows [7]:

3.33

�

� �

��

ULUL Loading

I-1

1log10

[dB]

(3.33)

4.5

where LoadingUL - uplink system loading [%]. Near-far effect The near-far effect can be described as a problem that occurs when a mobile is connected to a base station (RBS A) far away, hence transmitting with high power, and at the same time close to another base station (RBS B). Due to the low path loss, RBS B will receive high signal strength on the uplink, which will be perceived as interference [7]. For an indoor cell the near far problem can be somewhat different. In an indoor cell it is possible for the UE to be positioned very close to the RBS antennas due to the way an indoor antenna system is built where normally the antennas are attached to the ceiling. This can result in very low path loss for a UE positioned just under an antenna. Ideally the received power in the RBS receiver will be the same for all UEs (provided they are on the same RAB), independent of their pathloss to the RBS. If the mobile cannot down-regulate the power sufficiently, the excessive power from the UE will be perceived

56

as interference on the uplink. The lowest standardized output power for the UE is –50 dBm. Not to introduce unnecessary noise, the UE must be able to down-regulate to a power level so that the signal strength is no more than the RBS sensitivity at the RBS reference point. The excessive power will be perceived as interference. This in turn forces all the other UEs in the cell to combat this interference by raising their output power by the same amount, which will result in a raised total interference level in the cell. Due to the raised interference level the cell border moves towards the antenna, since UEs that are far away do not receive enough power to maintain required Eb/N0 values. However the capacity in the cell is unchanged. Some consequences of the near far problems are thus: �� The cell will shrink; the uplink link budget is degraded. �� More interference will be emitted towards other cells since the general

noise level in the cell is raised. Note that even though it is possible to calculate a certain noise rise for a UE close to an antenna, the probability that the UE actually is positioned just under the antenna has to be taken into account. The near far problem might seem severe when calculations are done, but in reality the probability could be very small that the UE actually is positioned in the “worst case spot”. It could also be that the UE just passes under the antennas for a short interval of time. Thus the perceived near far problems could be much more limited than what is indicated by “worst case” calculations. The magnitude of the problem will vary from case to case and will depend on factors such as: �� Feeder lengths

Short feeder gives less attenuation to the RBS, hence more interference will be introduced than when long feeders are used (with higher attenuation).

�� Antenna placements If the antennas are placed in positions where the UEs are positioned only a limited amount of time, the problems will be much less than when placing antennas at spots, which are more frequented by UEs.

�� Amount of antennas If an indoor antenna system is made in such a way that the coverage from the antennas will overlap; the result will be less coverage reduction when the interference level increases. On the other hand, with many antennas the probability that a UE is positioned close to an antenna will increase. Thus the number of antennas has to be chosen carefully.

�� Service The RBS sensitivity for low data rates is higher compared to that for high data rates. Therefore the impact of near far problem will depend on the type of service.

The recommended way to deal with near far problems is to introduce a margin in the uplink link budget [7], see section 3.3.2. This will ensure that the antennas will be placed close enough to minimize any coverage holes due to near far problems. Another way to handle near-far problems is to use attenuators. Note that attenuators should be avoided if possible. There are several reasons for this: firstly, poor inter-modulation performance of these components; secondly, it will degrade the downlink coverage as well.

57

Instead of using attenuators in a coaxial distributed antenna network, it is recommended to add more antennas. That will result in the required attenuation and will also improve the general performance of the indoor cell. 3.3.2 Margins Near-far margin As it was mentioned in section 3.3.1 the recommended way to deal with near-far problem is to introduce a margin in the uplink link budget in order to deal with the raised interference. This margin will be referred to as near-far margin and denoted as NFmarg. Power control margin When the UE approaches the cell border the path loss increases. The power control loop responds by increasing the UE power until the loop saturates and it is transmitting with full power. Hence power control is no more operating. This will result in reduced sensitivity for the RBS. To cater for this effect a margin is included in uplink link budget. This margin is called power control margin, PCmarg, and is dependent on channel model as well as on UE speed. Recommended power control margin for indoor environment is 4.5 dB [7]. Log-normal fading margin General description of the log-normal fading can be found in section 3.1.2.5 and log-normal fading definition in section 3.2.2. In contrast to GSM, in UMTS system the UE has the possibility to be connected to several RBSs at the same time. This allows reducing the log-normal margin as compared to GSM, where only hard handover is possible. The reduction is referred to as soft handover gain. The soft handover gain is approximately 0.3 – 2 dB on the uplink [7]. The soft handover gain is depending on coverage probability and site configuration. Simulations have been performed to find log-normal fading margins for UMTS, see [7] for more details regarding simulation results. Log-normal fading margin for 95% area coverage in indoor environment ( 9�� ) is equal to 5 dB [7]. Body loss The body loss margin recommended by ETSI is 3 dB for 1900 MHz [7]. Generally, body loss is not applied for data services since the users will most likely not have the terminal by the ear. 3.3.3 Uplink UMTS cell coverage dimensioning The maximum path loss for an indoor cell between an UE and the Rx reference point in the RBS can be calculated as [7]:

3.34 argargarg(max) mDASantmmULsensUEpath NFLGBLPCLNFIRBSPL �� (3.34)

58

where (max)pathL - maximum path loss due to propagation in the air. The cell range can

be calculated based upon this figure, UEP - the UE output power (21 dBm for speech, 24 dBm for data),

sensRBS - the RBS sensitivity. Depends on the RAB,

ULI - the noise rise in uplink, Section 3.3.1,

argmLNF - the indoor log-normal fading margin, Section 3.3.2,

argmPC - the power control margin, Section 3.3.2,

BL - the body loss, Section 3.3.2, antG - the sum of RBS antenna gain and UE antenna gain,

DASL - total losses in the distributed antenna system,

argmNF - the near-far margin, Section 3.3.2. First of all, near-far margin need to be calculated. The following initial data is considered:

�� Total DAS losses normally should not exceed 30dB in order to achieve acceptable EIRP (based on experience from IBS design)

�� Omni antennas are used in DAS (antenna gain, antG is 2 dBi) �� Speech 12.2 kbps is transmitted from the UE, leading to an RBS sensitivity of �� –120.2 dBm [7]. This situation provides the worst case from the near-far point of

view; as for other services the RBS sensitivity is lower �� The UE is immediately under an antenna (the min path loss, minp,L is 40 dB)

The minimum signal strength at the RBS reference point can be calculated as:

3.35 DASantpUE,RBS, LGLPSS �� min,minmin (3.35)

With values inserted equation (3.35) becomes:

dBmdBdBidBdBmSS RBS, 1183024050min �� . This will give us the following near-far margin for speech service:

dBdBmdBmRBSSSNF sensRBSm 2.22.120118min,arg ��

RBS sensitivity for other services is worse than –117 dBm [7]. Therefore there is no need to add argmNF in case of other services, that speech.

After having defined near-far margin, we can calculate uplink link budget for different services.

59

Uplink link budget for speech service The following initial data is considered:

�� UE output power, UEP is 21 dBm for speech 12.2 kbps;

�� RBS sensitivity, sensRBS is –120.2 dBm for speech 12.2 kbps [7]; �� The highest possible load in uplink is equal to 60% of poleM , see Section 2.2.3;

�� Total losses in the distributed antenna system, 30�DASL dB;

�� Omni antennas are used in DAS (antenna gain, antG is 2 dBi). Uplink noise rise in this case is equal to (see equation (3.33)):

.46.0-1

1log10 dBIUL �

�

� �

��

With the values inserted, equation (3.34) becomes:

.5.942.2302

35.4542.12021(max)

dBdBdBdBi

dBdBdBdBdBmdBmLpath

��

��

Uplink link budget for data services The following initial data is considered:

�� UE output power, UEP is 24 dBm for all PS services;

�� RBS sensitivity, sensRBS is –115.5 dBm for all PS services [7]; �� The highest possible load in uplink is equal to 60% of poleM , see Section 2.2.3;


�� Omni antennas are used in DAS (antenna gain, antG is 2 dBi). Uplink noise rise in this case is equal to (see equation (3.33)):

.46.0-1

1log10 dBIUL �

�

� �

��

With the values inserted, equation (3.34) becomes:

.980302

05.4545.11524(max)

dBdBdBdBi

dBdBdBdBdBmdBmLpath

��

��

Conclusion of Uplink budget calculations According to uplink budget calculations, speech has lowest allowable pathloss. Thus, speech is the limiting service in uplink.

60

3.3.4 Downlink UMTS cell coverage dimensioning

The downlink equations are more complex than the uplink ones. For the downlink it is not as easy to separate the coverage and capacity in the way that was done for the uplink. The main difference as compared to the uplink is that the UEs in the downlink share one common power source and that the interference is different for each user. Thus the cell range is not dependent only on how many UEs there are in the cell but also on the distribution of the UEs. In order to facilitate the dimensioning process, curves have been generated based on the equations. The curves display the cell load (M/Mpole) versus the cell range. Simulated curves can be found in [7]. Typical parameter values have been used and 20% of the power has been allocated to control channels. A homogenous user distribution has been assumed. To account for non-homogenous distributions a 5 W headroom should be used. In the downlink capacity charts, the Tx reference point of the RBS is moved out to the indoor antenna, which makes the design of the indoor cell solution independent. Knowing system load in downlink, the EIRP and the type of interior, the maximum allowed pathloss in downlink between the antenna and the UE is determined by the chart. Note that the indoor log normal fading margin and body loss have intentionally been left out of the DL capacity charts and are included in the DL linkbudget in order to make it possible to change these parameters if needed. For this reason when using the charts, output power at the antenna should be presented by adapted EIRP equation [7]:

3.36 argdim mantDASadapted LNFBLGLPEIRP �� (3.36)

where dimP is the power level to be used for dimensioning the system. Using RBS with maximum output power of 20 W and adding a headroom margin of 5 W, dimP is 15 W (41.8 dBm).

The following initial data is considered:

�� RBS output power, dimP is 41.8 dBm; �� The highest possible load in downlink is equal to 75% of poleM , see Section 2.2.3;


�� Omni antennas are used in DAS (antenna gain, antG is 2 dBi); �� Body loss is considered to be 3 dB (worst case).

Adapted EIRP in this case is equal to (see equation (3.36)):

.8.5532308.41 dBmEIRPadapted ��

61

Based on the curves from the DL capacity chart for interior environment type “Narrow” [7] the maximum allowable downlink pathloss between the antenna with adapted EiRP = 5.8 dBm and the UE, at a load of 75% is 95 dB ( dBLpath 95(max) � ). 3.3.5 Final UMTS cell coverage As a it can be seen from sections 3.3.3 and 3.3.4 UMTS indoor cell with the stated initial data is uplink limited. As balance between uplink and downlink limitations the maximum allowable pathloss between antenna and the UE is to be taken equal to 94.5 dB.

3.4 CONCLUSION OF CELL COVERAGE DIMENSIONING In previous sections maximum allowable pathloss has been calculated for all systems: GSM 900, GSM 1800 and UMTS. In order to be able to compare these results it is necessary to take system carrier frequency into account, as it will affect the cell range. According to the free space formula (see Section 3.1.2.1) UMTS with its carrier frequency at 2100 MHz, will have the most attenuation in the air interface. It is followed by GSM 1800 (1800 MHz) and then GSM 900 (900 MHz). Difference in air interface due to frequency is as follows (compared to UMTS):

�� GSM 1800 has 1.3 dB lower loss, compared to UMTS; �� GSM 900 has 7.3 dB lower loss, compared to UMTS.

To compare maximum allowable pathloss for different systems we need to equalize results that were presented in sections 3.2.3, 3.2.4 and 3.3.5. Equalized path loss is as follows:

3.37 dBLL UMTSpathUMTSequalizedpath 5.94(max),,(max), �� (3.37)

3.38 dBdBLL GSMpathGSMequalizedpath 5.863.72.793.7900(max),900,(max), �� (3.38)

3.39 .5.803.12.793.11800(max),1800,(max), dBdBLL GSMpathGSMequalizedpath �� (3.39)

According to equations (3.37), (3.38), (3.39) GSM 1800 has the lowest equalized allowable path loss and thus the smallest cell range. GSM 1800 is the limiting system in this case and DAS cell coverage should be designed according to GSM 1800 coverage requirements. It should be noted that losses in the DAS would be lower for GSM 900 and GSM 1800 than for WCDMA when using the same DAS, because the losses in the cables and combining equipment also depend on the frequency. This means that performance of the GSM systems will be somewhat better than presented above. But considering big difference in maximum allowable path loss, different attenuation in the DAS can be neglected.

62

4 IBS DESIGN EXAMPLE Now, when IBS design methodology and criteria are defined, let us look at practical implementation. An office building in Stockholm is taken as an example. It is an 8-floor building. Information regarding the number of users is presented in Table 4.1. Table 4.1 Number of users in the building (presented per floor).

Building floor Number of

people Floor 0 10 Floor 1 40 Floor 2 60 Floor 3 60 Floor 4 60 Floor 5 60 Floor 6 60 Floor 7 40 Total 390

Figure 4.1 depicts the vertical structure of the building. Figure 4.1 Vertical structure of the building.

Floo 0. Garage below the ground.

Floor 1. Main entrance, restaurantand offices.

Floor 2. Offices.

Floor 3. Offices.

Floor 4. Offices.

Floor 5. Offices.

Floor 6. Offices.

Floor 7. Offices.

63

The following initial information is used for the design:

�� Multi-band operation (GSM 900, GSM 1800 and UMTS); �� Area with medium interference (SSreq = -75 dBm for GSM 1800, see Section 3.2.2)

is considered for the office floors above the ground and area with low interference (SSreq = -85 dBm for GSM 1800, see Section 3.2.2) is considered for the underground garage;

�� Area coverage requirement is 95%; �� Target EIRP is 10 dBm; �� Traffic per user is 100 mE; �� GoS = 0.5% (for speech); �� Cell control channel configuration is BCCH + 2 SDCCH/8 (for GSM); �� One timeslot is reserved for GPRS (for GSM).

All passive DAS equipment mentioned in the following sections is described in Appendix A. 4.1 DESIGN PROCESS In this section an overview of IBS design activities is shown.

64

The purpose of this process is to create an antenna plan, which works “on paper”. An antenna plan includes antenna locations and types, design of the feeder network as well as cell splitting. Some antenna locations may be unsuitable from an implementation point of view and may have to be adjusted. If major changes are needed a creation of a new nominal antenna plan is required. Since predictions inside buildings contain uncertainties such as wall losses it is necessary to control the actual coverage from the nominal antenna. Once final antenna positions are defined, cable distribution is to be designed in order to achieve balanced EIRP at the antennas. When the nominal antenna plan has been accepted, the system shall be implemented. This action is not further described in this thesis.

Figure 4.2 IBS design workflow. 4.2 NOMINAL ANTENNA PLAN The purpose of this section is to show how a nominal antenna plan is made. This is done in two steps:

Creationof a

NominalAntenna plan

Survey

Coveragemeasure-

ments

CableDistribution

design

Implementation

Not O

K

Not O

K

OK

OK

65

1. Cell splitting. Number of cells required to fulfill traffic requirements is based on the number of users in the building and traffic generated by each user. Cell splitting is based on the cell capacity dimensioning calculations, presented in Chapter 2.

2. Path loss estimations. This can be done with building floor plans by using

prediction models mentioned in Section 3.1.2. Taking into account initial data and cell capacity calculations for UMTS and GSM, performed in Chapter 2 (especially Sections 2.2 and 2.3), one can see that the maximum optimal number of users per cell should not exceed 223 for this case. According to Table 4.1 total number of people in the building is 390. That means that capacity-wise the number of needed IBS cells should be at least:

275.1223390

___ ��cellsneededofNumber .

Taking into account distribution of users per floor, the following cell splitting is proposed: Cell 1: Floor 0, Floor1, Floor 2 and Floor 3 (total 170 users); Cell 2: Floor 4, Floor 5, Floor 6 and Floor 7 (total 220 users). Each cell is connected to a separate combining box (see Appendix A.1). Proposed cell splitting is shown in Figure 4.3. Figure 4.3 Cell splitting.

Floo 0. Garage below the ground.

Floor 1. Main entrance, restaurantand offices.

Floor 2. Offices.

Floor 3. Offices.

Floor 4. Offices.

Floor 5. Offices.

Floor 6. Offices.

Floor 7. Offices.

Cell 1. 170 users.Combining Box nr. 1.

Cell 2. 220 users.Combining Box nr. 2.

66

Design level for office floors is equal to (see Section 3.2):

SSdes = SSreq + 5.8 = -69.2 dBm. Maximum allowable path loss becomes (see Section 3.2):

dBdBdBmSSEIRPL despath 2.792.6910(max) �� . For the sake of simplicity, modified Keenan-Motley model (presented in Section 3.1.2.9, Figure 3.4) will be used for nominal antenna plan. Based on experience, wall loss in office buildings is approximately 1 dB/m (based on experience). Path loss for GSM 1800 is then equal to:

Antenna coverage range is equal to:

m8.163.1

4.572.793.1

4.57(max) ��

��

� pathLD .

Figure 4.4 shows nominal antenna plan for one of the office floors of the building (all office floors have similar layout). Antenna in elevator hall is intended for elevator coverage only. Figure 4.4 Nominal antenna plan for floor 2 Floor 0 in building under consideration can be split into two parts: garage part and technical part, which consists of a number of laboratories and some technical rooms.

.)3.01(4.57 DL path ��

67

Technical part of floor 0 is assumed to have similar radio characteristics and interference situation as office space. Garage part is more isolated from the outdoor network, thus SSreq = -85 dBm. In the garage, car penetration loss (CPL = 6 dB [7]) should be taken into account.

dBm2.7368.585arg �� CPLLNFSSSS mreqdes . Maximum allowable path loss in this case is:

dBdBdBmSSEIRPL despath 2.832.7310(max) �� . Garage area is a relatively open space environment, but due to presence of certain obstructions like pillars, separations, walls and low ceilings, Friis free space model (see section 3.1.2.1) will not give reliable results. Instead, modified Keenan-Motley model (presented in Section 3.1.2.9, Figure 3.4) will be used. Based on experience, linear losses in underground parking is approximately 0.2 dB/m. Path loss for GSM 1800 is then equal to:

Antenna coverage area is equal to:

m6.515.0

4.572.835.0

4.57(max) ��

��

� pathLD .

Figure 4.5 shows nominal antenna plan for floor 0. Figure 4.5 Nominal antenna plan for floor 0

.)3.02.0(4.57 DLpath ��

68

4.3 SURVEY The nominal configuration may not be suitable for the building. Reasons for this can be that there are obstacles deteriorating the radiation pattern or that there are problems mounting the antennas at these locations. Note that the landlord often has an opinion about the antenna locations. These locations should be accepted before final installation. This will save unnecessary work. It is also necessary to consider the effect of non-optimal antenna locations. During survey installation aspects have to be considered. It is usually controlled that the antennas are not placed in positions that make the cable running difficult, for example with a concrete wall between the antenna location and the cable ladder. Nominal antenna positions are reviewed considering that the omni-directional antennas are ceiling mounted and the directional antennas are wall mounted in general. During survey it is also necessary to identify position where RBS will be placed and find locations of cable raisers. This information will be further used during cable distribution design (see section 4.5). For the building under consideration, Figure 4.6 shows selected RBS position and places of cable raisers. Figure 4.6 RBS position and cable raisers.

69

4.4 COVERAGE MEASUREMENTS Coverage measurements are performed to verify the antenna configuration so that it provides sufficient coverage. The antenna locations that are determined according to section 4.2 are then tested one or several at a time. A test transmitter is connected to the antenna and the signal strength is measured with a mobile receiver. In Figure 4.7 the measurement configuration is shown. According to conclusions made in Section 3.4, GSM 1800 is the limiting system in terms of coverage. Due to this fact coverage test measurements are performed in GSM 1800 band. Figure 4.7 Measurement configuration. 4.4.1 Transmitter equipment For the generation of test signals it is suitable to use one or several GSM TEMS Transmitters. The GSM TEMS Transmitter is a small unit that transmits in the GSM downlink band. The output power is fixed to 17dBm. Output power can be adjusted to the required value by applying attenuators at the output of the transmitter. A complete editable BCCH is transmitted while the other 7 time slots contain an unmodulated carrier. 4.4.2 Receiver equipment The receiver used is TEMS Light equipment. This is a TEMS mobile connected to a small pen-operated PC. The TEMS light program is software with the possibility to log fixed points by marking with the pen on a scanned map. The information in the log files is displayed on the scanned map as color marks, associated to a window with more information concerning each mark. For this kind of tests, information regarding received signal strength is the main aspect that should be looked upon.

70

4.4.3 Measurement procedure

1. Measurement frequency or frequencies need to be selected. Usually frequencies allocated to the cell, or any other “quiet” frequencies are used.

2. The selected frequencies are measured with the TEMS mobile (test transmitters turned off) to ensure that the choice of frequencies is not bad. This can be done in just one or two representative spots.

3. The transmitted power is adjusted by the attenuators. Usual EIRP of an indoor antenna is 10-15 dBm.

4. The output power may be additionally controlled by a power meter. 5. The test transmitters are connected to the temporarily mounted antennas. 6. Engineer with TEMS Light equipment moves away from the antenna and searches

for coverage borders. Borders and coverage holes are noted on a floor plan. This limit can be set in the TEMS Light to enable an easy measurement procedure.

7. The described measurement procedure is repeated with the remaining antenna locations.

8. If coverage holes exist in the present cell plan, more antennas should be added or nominal antenna positions reconfigured.

It should be noted, that log-normal fading is included in the measurement results, so it is SSreq that should be measured. Nominal antenna plan, presented in section 4.2 will be verified, using measurement procedure, presented above. 4.4.3.1 Floor 0 Coverage area of bi-directional antenna in the garage was tested first. Results of the measurements are presented in Figure 4.8. As it can be seen from measurement results, antenna in this position provides good coverage in the main garage hall, but coverage in the entrance tunnel and secondary parking sections is insufficient. It can be seen that secondary parking sections are separated by very heavy walls (this area was intended to be a bomb shelter), which provide approximately 25 dB attenuation. Good coverage in the entrance tunnel is important, because this is a pre-requisite for seamless handover from outdoor network to indoor. Due to mentioned coverage holes, alternative antenna positions were tested as shown in Figure 4.9, Figure 4.10 and Figure 4.11.

71

Figure 4.8 Coverage area of antenna 0.1 (bi-directional) in the garage on floor 0. Figure 4.9 Coverage area of antenna 0.2 (directional) in the garage on floor 0.

72

Figure 4.10 Coverage area of antenna 0.3 (directional) in the garage on floor 0. Figure 4.11 Coverage area of antenna 0.4 (omni) in the garage on floor 0.

73

Coverage tests showed that the best composite coverage is provided by antenna positions 0.2 and 0.4. Secondly, antenna positions in the technical part of floor 0 were tested. Figure 4.12 shows coverage area of an omni antenna in position according to nominal plan. Figure 4.12 Coverage area of antenna 0.5 (omni) in the technical part of floor 0. Test measurements showed, that antenna in this position does not provide necessary signal strength in the leftmost area. So, antenna positions alternative to nominal plan were tested as shown in Figure 4.13, Figure 4.14 and Figure 4.15.

74

Figure 4.13 Coverage area of antenna 0.6 (omni) in the technical part of floor 0. Figure 4.14 Coverage area of antenna 0.7 (omni) in the technical part of floor 0.

75

Figure 4.15 Coverage area of antenna 0.8 (omni) in the technical part of floor 0. According to the test measurements antennas in positions 0.6, 0.7 and 0.8 are more suitable. Final antenna positions for floor 0 are presented in Figure 4.16. Figure 4.16 Final antenna plan for floor 0.

76

4.4.3.2 Office floors Firstly, nominal antenna position in the middle square was tested. Floor two was selected for the measurements. Coverage test results are presented in Figure 4.17. Test showed that antenna in position 2.1 provides acceptable coverage in the middle square. The rest of office floors have similar layout in the middle square – rooms along the sides of the square and open central part (thin walls denominate just soft partitions of approximately 1.5 meters high). Figure 4.17 Coverage area of antenna 2.1 (omni) on floor 2. Left and right squares have more walls in the central part. There is also concrete staircase in each of these squares. Test measurements were performed on floor four. Figure 4.18 depicts coverage area of an antenna at the nominal position in the left square. As it can be seen antenna in position 4.1 does not provide sufficient coverage in the left square. Coverage holes appear in the rooms in the corners of the square, especially behind the staircase, which gives high attenuation. Shadow behind the staircase is better illustrated in Figure 4.19.

77

Figure 4.18 Coverage area of antenna 4.1 (omni) on floor 4. Figure 4.19 Coverage area of antenna 4.2 (omni) on floor 4.

78

Alternative antenna positions were tested as shown in Figure 4.20 and Figure 4.21. Coverage tests showed that antennas in positions 4.3 and 4.4 achieve good composite coverage. Figure 4.20 Coverage area of antenna 4.3 (omni) on floor 4. Figure 4.21 Coverage area of antenna 4.4 (omni) on floor 4.

79

An additional coverage test was performed for nominal antenna position in the middle of the building. This antenna is intended for elevator coverage. Figure 4.22 shows coverage provided by this antenna in the office square. Figure 4.22 Coverage area of antenna 4.5 (omni) on floor 4. In order to verify propagation through the floor, antennas were placed on floor three and coverage area of these antennas was measured on floor four. Measurement results are presented in Figure 4.23 and Figure 4.24. Figure 4.23 Coverage area of antenna 3.1 (omni) on floor 4.

80

Figure 4.24 Coverage area of antenna 3.2 (omni) on floor 4. As one can see from the coverage plots, antennas in the corners of the square (like positions 4.3, 4.4 and 3.1) provide good coverage through the floor. Coverage hole appears only in the center of the square in this case. So, in simple terms, antennas in positions 4.3 and 4.4 provide good coverage in the corners of the square on floor 5. Only one antenna in the center is then required on floor 5. In order to minimize the number of antennas in the building, it is possible to alter antenna configuration every other floor: one floor having antennas in the center of the squares and the other having two antennas in the corners of the left and right squares. Final antenna positions for office floors are presented in Figure 4.25 and Figure 4.26.

81

Figure 4.25 Final antenna plan for floor 1. Figure 4.26 Final antenna plan for floor 2.

82

Antenna plan of floor 1 is repeated on floors 3, 5 and 7. Antenna plan of floor 2 is repeated on floors 4 and 6 respectively. 4.5 CABLE DISTRIBUTION DESIGN When final antenna positions have been defined, cable distribution is designed. Normally, one passive coaxial cable branch can feed 5-10 indoor antennas, depending on building shape and the distance from the RBS location to the antennas. During cable distribution design, capacity calculation results and decided cell splitting (see Section 4.2) should also be considered. Passive DAS can be whether power or capacity limited. The limiting factor decides the final cell splitting. Note, that when using Ericsson combining box (see section A.1), there can be maximum 4 main cable branches per cell. From a radio propagation perspective it is preferable to have vertical cell splitting. During this activity, cable routing from RBS to each of the antennas is defined on the floor plan. Splitting places are also defined. Types (sizes) of the coaxial cable and types of splitters (refer to Appendix A for equipment description) are selected in a way to achieve as equal EIRP at the antennas as possible. As it was previously mentioned in this thesis, recommended EIRP at the indoor antennas is 10 - 15 dBm. Antenna EIRP calculations can be presented in an Excel table (see Figure 4.27).

83

Figure 4.27 Antenna EIRP calculations.

84

After achieving equal EIRP (within precision of 5- dB) for all antennas, antenna and component locations diagram as well as trunking diagram can be finalized. Figure 4.28 depicts sample antennae, splitters and cable locations on a building floor plan. Figure 4.28 Antenna and components locations diagram for floor 2.

85

Figure 4.29 depicts final cell splitting diagram for the building. Figure 4.29 Final cell splitting diagram.

86

Figure 4.30 illustrates sample cable distribution (trunking) diagram. Figure 4.30 Sample of trunking diagram.

87

5 CONCLUSIONS AND FUTURE WORK In conclusion we can summarize this thesis with the following points:

�� We investigated the possibility of simultaneous multi-band (GSM 900, GSM 1800 and UMTS) and multi-operator access to an in-building coverage solution, which is based on distributed antenna system. Combining equipment, which enables these possibilities is described in Appendix A.

�� Multi-band IBS cell capacity dimensioning methodology was developed.

Calculations showed, that UMTS is the limiting system in terms of cell capacity. Certain mapping between GSM cell capacity and UMTS cell capacity was found: it is not practical to consider GSM cells dimensioned with more than 5 TRUs, as in this case WCDMA system connected to the same DAS will not provide necessary capacity resources.

�� Multi-band IBS cell coverage dimensioning methodology was developed and

design criteria found. According to calculations GSM 1800 has the lowest equalized allowable path loss and thus the smallest cell range. GSM 1800 is the limiting system in terms of cell coverage and DAS cell coverage should be designed according to GSM 1800 coverage requirements.

�� We have also provided a practical example of an IBS design, based on the

presented methodology and criteria.

�� When studying technical specifications of the passive DAS components, one can see that most of them support frequencies up to 2500 MHz. This means that there is a possibility to also combine WLAN into the same IBS. On the market there already certain equipment (for example WLAN Injector from Ericsson) available, which enables combining WLAN, together with GSM and UMTS, into the same DAS. WLAN integration could be the topic of future research.

88

6 REFERENCES

1. Alexander, S. E. Radio propagation within buildings at 900 MHz. - Electronics letters, Vol. 18, No. 21, 1982, pp. 913-914.

2. Andersen, J. B., T. S. Rappaport, and S. Yoshida. Propagation measurements and models for wireless communication channels. – IEEE Communication Magazine, Nov. 1994, pp. 42-49.

3. Bernhardt, R. C. Macroscopic Diversity in Frequency Reuse Systems. IEEE Journal on Selected Areas in Communications, Vol-SAC 5, June 1987, pp. 862-878.

4. Cox, D. C., Murray, R. R., and Norris, A. W. Measurements of 800 MHz radio transmission into buildings with metallic walls. - Bell systems technical journal, Vol. 62, No. 9, November 1983, pp. 2695-2717.

5. Cox, D. C., Murray R., and Norris A. 800 MHz attenuation measured in and around suburban houses. - AT&T Bell laboratory technical journal, Vol. 673, No. 6, July-August 1984.

6. Devasirvatham, D. M. J., C. Banerjee, M. J. Krain, D. A. Rappaport. Multi-frequency radiowave propagation measurements in the portable radio environment. – IEEE International Conference on Communications, 1990, pp. 1334-1340.

7. Ericsson technical documentation. 8. Feuerstein, M. J., K. L. Blackard, T. S. Rappaport, S. Y. Seidel, and H. H. Xia. Path

loss, delay spread, and outage models as functions of antenna height for microcellular system design. – IEEE transactions on vehicular technology, Vol. 43, No. 3, Aug. 1994, pp. 487-498.

9. Fleury, B. H., and P. E. Leuthold. Radio Propagation in Mobile Communications: An Overview of European Research. – IEEE Communications Mag., Vol. 34, No. 2, Feb. 1996, pp. 70-81.

10. Fung, V., Rappaport, T. S., and Thoma, B. Bit error Simulation for 4/� DQPSK Mobile Radio Communication Using Two-ray and Measurement-based Impulse Response Models. - IEEE Journal on Selected Areas in Communication, Vol. 11, No. 3, April 1993, pp. 393-405.

11. GSM 03.30 (Phase 2+). Radio Network Planning Aspects. - ETSI, version 8.3.0, 1999.

12. Hashemi, H. The indoor radio propagation channel. - Proceedings of the IEEE, Vol. 81, No. 7, July 1993, pp. 943-968.

13. Hashemi H. Impulse response modeling of indoor radio propagation channels. - IEEE Journal on Selected Areas in Communications, Vol. 11, No 7, Sep. 1993, pp. 967-978.

14. Marko I. Silventoinen. Indoor GSM Base Station Systems. Doctor’s thesis, Helsinki University of Technology, 1998, 158 p.

15. Jakes, W. C. Jr. Microwave mobile communications. Wiley-Interscience, 1974. 16. Javorski, A. Wideband CDMA physical channel models. Masters thesis, Tallinn,

2002, 72 p. 17. Keenan, J. M., A. J. Motley. Radio coverage in buildings. - British Telecom

Technology, Vol. 8, No. 1, January, 1990, pp. 19-24. 18. Manuel F. Catedra, Jesus Perez-Arriaga. Cell planning for wireless

communications. Artech House, Inc., 1999, 199 p. 19. Molkdar, D. Review of radio propagation into and within buildings. - IEE

Proceedings, Vol. 138, No. 1, February 1991, pp. 61-73. 20. Rappaport, T. S. Characterization of UHF multipath radio channels in factory

buildings. IEEE transactions on antennas and propagation, Vol. 37, No. 8, August 1989, pp. 1058-1069.

89

21. Rappaport, Theodore S. Wireless communications: principles and practice, 2. ed. Prentice-Hall, 2002, 707 p.

22. Reudink, D. O. Properties of Mobile Radio Propagation Above 400 MHz. - IEEE Transactions on Vehicular Technology, Vol. 23, No. 2, November 1974, pp. 1-20.

23. Rice, S. O. Statistical properties of a Sine Wave Plus Random Noise. Bell Systems Technical Journal, Vol. 27, January 1948, pp. 109-157.

24. Seidel, S. Y., Rappaport, T. S., Jain, S., Lord, M., and Singh, R. Path Loss, Scattering and Multipath Delay Statistics in Four European Cities for Digital Cellular and Microcellular Radiotelephone. - IEEE Transactions on Vehicular Technology, Vol. 40, No. 4, November 1991, pp. 721-730.

25. Seidel, S. Y., Rappaport, T. S. 914 MHz path loss prediction models for indoor wireless communications in multifloored buildings. - IEEE Transactions on Antennas and Propagation, Vol. 40, No. 2, Feb. 1992, pp. 207-217.

26. Seidel, S. Y., et al. The impact of surrounding buildings on propagation for wireless in-building personal communications system design. – 1992 IEEE vehicular technology conference, Denver, May 1992, pp. 814-818.

27. Stallings, William. Wireless communications and networking. Prentice-Hall, 2002, 584 p.

90

A. PASSIVE DAS COMPONENTS A.1 COMBINING BOX The Ericsson Multi-operator Combining Box combines 4 input ports on each of 3 frequency bands, into 4 antenna ports, giving a number of operators access to the same indoor Distributed Antenna System (DAS). Figure A.1 illustrates Ericsson Combining Box. Figure A.1 Ericsson Multi-operator Combining Box. Low frequency input: 824–960 MHz. This makes it possible to use it for TDMA/CDMA 850 MHz and GSM 900. (Additional filtering may be required depending on actual frequencies used.) Mid frequency input: GSM 1800 band 1710–1880 MHz. High frequency input: UMTS band 1920–2170 MHz. Technical specifications of the combining box are presented in the following table. Table A.1 Combining box technical specifications.

Electrical Specifications Frequency range input ports: 824–960 MHz 4 ports

1710–1880 MHz 4 ports 1920-2170 MHz 4 ports

Frequency range antenna ports (4 ports): 824–960 / 1710–1880 / 1920-2170 MHz

Isolation between 2 input ports of the same band: � 20dB Isolation between input ports for different bands: >70 dB

Return loss: > 18 dB Insertion loss (total of the “one to four port”

coupling and insertion loss): 824-960 7.5 ± 1 dB 1710-1880 8.0 ± 1 dB 1920-2170 8.0 ± 1 dB

Max composite input power (per input port): 100 W (50 dBm) Intermodulation (IM3): * – 105 dBm ( * -142 dBc)

Impedance in/out: 50 Ohm

91

A.2 ANTENNAS Antennas provide the coverage within the building by being positioned at strategic locations to maximize the performance. Antennas tend to come in various shapes and sizes, which allow for flexibility of choice during design. There are three main types of antennas used in the DAS: omni-directional antennas, directional antennas and bi-directional antennas. Technical specifications of the antennas are presented in Table A.2. Table A.2 Antennae technical specifications.

Description Gain Frequency Antenna Dir. 90

deg. 7 dBi 824-960 +

1710-2500MHz Antenna Omni 2 dBi 876-960 +

1710-2500MHz Antenna Bi-directional

5 dBi 824-960 + 1710-2170MHz

The following figures depict mentioned antenna types.

Figure A.2 Omni-directional antenna.

Figure A.3 Directional antenna.

92

Figure A.4 Bi-directional antenna. A.3 FEEDER CABLES The feeder cables are used to provide the required coverage within the building. The cable is run within the vertical riser/s of a building and along the horizontal sections (between ceilings and floors). The various type of feeder cable used are listed below: Co-axial cable: Co-axial cable is used as a medium over which the RF signal can travel. The varying sizes of cable used for IBS are 3/8 inch cable, 1/2 inch cable, 7/8 inch cable and 1&1/4 inch cable – the cable size is used dependant upon the length of the feeder runs. Radiating Cable: Radiating cable is similar in size to co-axial cable except that the cable has cut slots at regular intervals along the cable. This enables the RF to propagate evenly over the feeder distance. Thus, radiating cable does not require the need for antennas. Radiating cable is never run within the vertical risers – it is only run along the horizontal sections. The following figure depicts examples of coaxial feeder cables. Figure A.5 Coaxial feeder cables. A.4 TAPPERS A tapper is a RF device used to split the RF power in an unequal manner. Tappers are used along various sections along the feeder runs to:

�� Allow for the installation of an antenna; �� To extract a percentage of the RF power on every floor.

93

Figure A.6 Principal tapper diagram. Figure A.7 Tapper. A.5 SPLITTERS A splitter is a RF device used to split the RF power equally. Splitters may be used on every floor to split the signal equally along two, three or four different directions. Figure A.8 Splitter.

Coupled Loss

Insertion Loss Input

hoonesiseste raadiovÕrkude planeerimineavots/bwa_1a/deniss_syropjatov_mag.pdf · hoonesiseste...

Documents