Chapter3 Radio Propagation Theory

Wireless Network Planning Table of Contents

Table of Contents

Chapter 3 Radio Propagation Theory...........................................................................................13.1 Keystone of Radio Propagation...........................................................................................13.2 Radio Propagation Environment.........................................................................................3

3.2.1 Frequency Division Introduction...............................................................................3

3.2.2 Fast Fading and Slow Fading...................................................................................3

3.2.3 Propagation Loss......................................................................................................63.3 Radio Propagation Model..................................................................................................103.4 Correction for propagation model......................................................................................16

3.4.1 CW Basics..............................................................................................................16

3.4.2 CW Test Method.....................................................................................................16

3.4.3 Correction for Propagation Model and Instance......................................................183.5 Doppler Effect and its Impact on Handover.......................................................................193.6 Fresnel Zone.....................................................................................................................223.7 ASSET Software Introduction............................................................................................24

i

Wireless Network Planning Chapter 3 Radio Propagation Theroy

Chapter 3 Radio Propagation Theory

1.1 Basic knowledge of Radio Propagation

When planning and constructing a mobile communication network, we must understand the features of the electric waves to define the frequency band, frequency allocation, radio coverage, communication probability calculation, electromagnetic interference between systems and final parameters of radio devices. It is the keystone for system design, efficient use of frequency spectrum and EMC (Electronic Magnetic Compatibility).

It is well known that the radio wave can be transmitted from the transmitting antenna to the receiving antenna in multiple modes: forward wave or free space wave, earth wave or surface wave, troposphere reflecting wave, and ionosphere wave.

As shown in Figure 3-1, as far as electronic propagation is concerned, the simple method between transmitters and receivers is free space propagation. Free space refers to isotropy (identical in axes characters) and uniformity (even texture) in such zone. Other names for free space are forward wave or stadia wave. As shown in Figure 3-1(a), forward wave transmits along straight lines, so that it can be used for communication between satellite and exterior space. In addition, this definition is also used for stadia propagation in land (between two microwave towers), as shown in Figure 3-1(b).

The second method is the earth wave or surface wave. Earth wave is the combination of three waves: the forward wave, backward wave and surface wave. The Surface wave transmits along the earth surface. Some energy from the transmitting antenna can directly reach to the receiver; some energy reaches to the receiver after reflecting on the earth surface; some reaches to the receiver through surface wave. Surface wave transmits on the earth surface. Since the earth surface is not ideal for propagation, some of the energy is absorbed by the ground. When energy is absorbed by the ground, it can cause ground current. Such three surface waves are shown in Figure 3-1(c).

The third method is that troposphere reflecting wave. It is generated from the troposphere layer, which is a heterogeneous medium, changing with time because of air conditions. Its reflection factors decrease with an increase of height. Such reflection factors with gradual change cause electric wave bending, as shown in Figure 3-1(d). Troposphere method is applied to radio communication with the wavelength less than 10 meters (frequency larger than 30MHz). The fourth method is propagation through ionospheric reflection. When the electric wave is less than 1 meter in length (frequency larger than 300 MHz), the troposphere layer is the reflected body. The radio wave reflected from troposphere layer might have one or more leaps, as shown in Figure 3-1(e). Such propagation is used for long-distance communication. Besides reflection, troposphere layer can generate electric wave scattering because of uneven refractive rate. In addition, meteors in troposphere layer can also scatter electric waves. Like the troposphere layer, ionosphere layer also has the feature of continuous fluctuation, and such fluctuation is rapid fluctuation at random. Cellular system radio propagation is adopted the second method of electric wave propagation. It will be discussed in the following parts.

1


(b)

(c) (d)

(e)

(a)

Ground wave

Scatterer

Ionization layer

(a) Forward wave transmits along straight line

(b) Stadia communication application

(c) Earth wave propagation

(d) Troposphere lay scatters radio wave irregularly. (e) Radio wave transmits through ionosphere reflection

Figure 3-1 Different Propagation Modes

There are two reasons for propagation study when designing cellular system: first, it provides necessary tool for calculating signal level covering different cells. In most cases, coverage area is , therefore earth wave propagation can be adopted in such condition. Secondly, it can calculate monkey-chatter interference and cochannel interference.

There are three methods for predicatingsignal level radio coverage: the first one is pure theory, which is applied to separate objects, such as mountain and other solid objects. However, it ignores the irregularity of the Earth. The second one is based upon measurement in various environments, including irregular landform and man-made obstacles, especially the higher frequency and lower mobile antenna commonly existing in mobile communication. The third method is the improved model upon the above two methods, which considers the influence of mountains and other obstacles upon the measurement and the refraction law.

In the cellular system, there are at least two propagation models: the first one is FCC suggested model; the second one, established by Okumura, considers the actual experience data.

2


1.2 Radio Propagation Environment

1.2.1 Frequency Division Introduction

The Radio frequency from 3Hz to 3000GHz are separated into 12 bands, as shown in Figure the following table. Frequency in different frequency spectrum has different propagation characteristics. As to mobile communication, we only pay attention to UHF spectrum.

Frequency Classification Designation3 to 30Hz

30 to 300Hz Extremely Low Frequency ELF

300 to 3000Hz Voice Frequency VF

3 to 30KHz Very-low Frequency VLF

30 to 300KHz Low Frequency LF

300 to 3000KHz Medium Frequency MF

3 to 30MHz High Frequency HF

30 to 300MHz Very High Frequency VHF

300 to 3000MHz Ultra High Frequency UHF

3 to 30GHz Super High Frequency SHF

30 to 300GHz Extremely High Frequency EHF

300 to 3000GHz

1.2.2 Fast Fading and Slow Fading

According to the last section, in a typical cellular mobile communication environment, direct path between the receiver and the transmitter is obstructed by buildings or other objects. Thus, communication between the cellular base station and mobile station completes not through direct path but many other paths. In UHF frequency, the main propagation mode for electromagnetic wave from the transmitter to the receiver is scattering, i.e. reflection from the surface of building or refraction from artificial and natural objects, as shown in Figure 3-2.

3


① building reflected wave ② diffracted wave ③ forward wave ④ ground reflected wave

Figure 3-2 Multipath Propagation Model

All the signal components compose a multi standing wave, the signal level of which increases or decreases with corresponding changes of the components. The synthesis signal level fades 20 to 30dB in a few car bodies away, the difference between the maximum and the minimum is about 1/4 wavelength. A great number of propagation paths result in so called multipath phenomenon, whose synthesis amplitude and phase will undergo great fluctuation with the movement of mobile stations. Usually, such phenomenon is called multipath fading or fast fading, as shown in Figure 3-3. Essentially, multipath fading is a fast change. Besides, such propagation character causes time dispersion phenomenon. The distribution of deep fading point in space is approximately half wavelength away (900MHz is 17cm, 1800 or 1900Mhz is 8cm). If the mobile antenna is at the deep fading point at that time (when mobile user in a car stay at the deep fading point because of redlight, we call it Redlight Problem), voice quality is very poor. Therefore, related technologies like hopping should be applied to solve this problem.

Studies show that if the mobile cell receives the amplitude, phase and angle of respective component at random, then the azimuth angle of the synthesis signal and the probability density function of amplitude are as follows:

p 12 0 ≤ ≤ 2 (3-1)

pr r 2 e r2

2 2 r ≥ 0 (3-2)

Among them, “r” is the standard deviation. (3-1) and (3-2) represent the azimuth

angle is even distribution between 0 to 2, while the probability density function of electric field abides by Rayleigh Distribution. Therefore, multi path is also called

4


Rayleigh fading. As to this fast fading, the base station adopts the methods of time diversity, frequency diversity and space diversity (polarity diversity). Time diversity mainly adopts the methods of symbol interleave, error code checking and correcting. Different code has different anti-fading characteristics. As to the air channel coding of GSM mobile communication, please see related GSM protocol. The basic of frequency diversity theory is the correlation bandwidth, i.e. after more than an interval between two frequencies, their space fading characteristics are considered irrelevant. A large number of test data shows that such irrelevancy can be obtained if the interval between the two frequencies is larger than 200 KHz; frequency diversity mainly adopts spread spectrum. In GSM mobile communication, hopping is simply applied to obtain hopping gain, while in CDMA mobile communication, each channel works in wide band (narrow band CDMA is 1. 25 MHz), which actually, is a spread frequency communication. Space diversity mainly adopts the master diversity antenna receiving method. Signals the base station receiving from the master and diversity channels are respectively combined after equalization through the Maximum Likelihood Sequence Equalizer (MLSE). Such master diversity receiving effect is guaranteed by the irrelevancy received by the master diversity. Irrelevance refers to the signals received respectively by the master antenna and diversity antenna having no fading at the same time. It requires that the spacing between the master and diversity antenna is 10 times more than the radio signal wavelength (the antenna spacing is more than 4 meters in GSM900), or adopting polarity diversity to guarantee the signals received by the master and diversity antenna having different fading characteristics. Mobile station (mobile phone) has no such space diversity function with only one antenna. The equalizing ability to different ranges (time window) of the base station receiver is also a form of space diversity. In CDMA communication, when soft switching is performed, the mobile station and multi base stations communicate at the same time to select the best signal for handover, such is also a form of space diversity.

A great number of studies shows that the average signal levels received by the mobile station, except for fast Rayleigh fading in instantaneous value, appear slow changes as changing position, such change is called slow fading, as shown in Figure 3-3. It is caused by the shadow effect, and also called shadow fading. Buildings, forest and topographical relief in the way of radio propagation will cause shadow in electromagnetic field. The medium value of receiving signal level will change when electromagnetic shadow is produced by different obstacles the mobile station encounters. The change is depended upon the obstacle condition and working frequency; changing rate has relations with obstacles and driving speed.

By studying this fading law, it shows that its medium value variation abides by Logarithmic Normal Distribution.

Additionally, radio refraction coefficient changes as the climate conditions change with times, as well as slow changes in vertical gradient of atmosphere dielectric constant, which results in slow changes in signal level medium value in the same place as time changing.

Statistics show that such medium value variation also abides by Logarithmic Normal Distribution. The distribution standard deviation is rt. Variation of signal medium value in a larger range of distribution with time and place all abides by Logarithmic Normal Distribution, so that their synthesis distribution still abides by Logarithmic Normal Distribution. When communicating in land, usually, signal medium value variation as time changing is less than that as place changing, so that such slow fading can be ignored, r=rL. However, in fixed-point communication, slow fading shall be considered.

5


Figure 3-3 Fast fading and slow fading

In general, there are two influences in cellular environment: the first one is fast fading; and the second one is slow changes in receiving signal level resulted from directly visible path, i.e. long-term signal level change. That is to say, the channel works in fast fading in accordance with Rayleigh distribution, and superimposes amplitude with signal to meet with slow fading in Logarithmic Normal Distribution.

1.2.3 Propagation Loss

In propagation studies, signal level received by specialized receiver is a major feature. Owing to the interference of propagation path and landform, propagation signal level decreases. Such signal level decrease is called propagation loss.

In radio propagation studies, first study the characteristics of the two antennas in free space (homogeneous medium with isotropy, no absorption, zero electric conductivity). Take the ideal omnidirectional antenna as an example. The propagation loss of free space is:

Lp 32.4 20 lglg(fMHz ) 20 lglg(d km ) (3-3)

Among which, f is frequency, d is distance (kilometers). In the above equation, propagation loss is in inverse proportion to d. When d doubles, free space path loss increases by 6 dB. Meanwhile, when wavelength λ decreases (increase frequency f), path loss increases. We can compensate these losses by increasing radiation and receiving antenna gain. If the working frequency is already known, (3-3) can be also written as:

Lp L0 10 lglg(d km ) (3-4)

Of the equation, 2, is called path loss slope. In the actual cellular system,

according to measurement result, value ranges from 3 to 5.

Having the equation of path loss in free space, the actual propagation can be considered between the two antennas on plain but imperfect surface. Suppose the whole propagation path surface is absolutely plain (without refraction). The antenna

height of the mobile station is hc and hm respectively (A represents hc, and B

represents hm), as shown in Figure 3-4.

6


A

A

A

A'

B

B

B

(a)

(b)

(c)

(a) multireflection (b) simple reflection (c) mapping method of finding path difference between stadia and ground reflection

Figure 3-4 Propagation on Plain Surface

As compared with the path loss in free space, propagation path loss on plain grounds is:

Lp 10 lglgd 20 lglghc 20 lglghm (3-5)

Of which, 4. This equation shows that if antenna height doubles, 6 dB can be compensated for loss; while the receiving power of the mobile station changes with the fourth power of distance, i.e. if distance doubles, the power received reduced by 12 dB.

Various landforms and ground objects differ greatly, so the impact on radio propagation loss in mobile communication also varies. It is impossible to have absolutely plain landform in actual application. Such complex landforms can be divided into two types: “quasi smooth landform” and “irregular landform”.

“Quasi smooth landform” refers to the landform with gentle rolling topography, rolling height less than or equal to 20 meters as well as slight difference in average surface height. Okumura defines the rolling height as the difference between 10% and 90% of rolling topography 10 kilometers ahead of the mobile station. CCIR defines it as the difference between over 90% and over 10% of rolling topography 10 to 50 kilometers ahead of the receiver. Other landforms are generally called “irregular landform”, which can be divided into the following types based upon their conditions: hills, separated mountains, slopping landform and water-and-land mixed landform and so on.

When analyzing propagation loss in urban areas and their nearby areas, we can also classify “irregular landform” by congestion in regions as open area, dense urban area, medium urban area and suburb area.

7


In general, we also analyze diffraction loss when analyzing propagation loss in mountainous area or dense urban areas with close skyscrapers. Diffraction loss is used to measure the height of obstacles and antenna. The obstacle height must be compared with propagation wavelength. As to the same obstacle, the diffraction loss to long wavelength is less than that to short one. When predicating path loss, we can view these obstacles as pointed obstacles, i. e. “knife-shaped”. Loss can be calculated by the method commonly used in physical optics. Two kinds of obstacles shown in Figure 3-5. Under the first condition, no obstacles appear in stadium path at H. Under the second condition, obstacles appear in radio path. In the first condition, we assume that the height of obstacle is negative number, while positive number in the second condition. Diffraction loss can be calculated through the diffraction constant v, which is known from the following equation.

v H 2/ 1/d1 1/d2 (3-6)

The approximate value of diffraction loss can be calculated from the following equations:

F 0 v 1

20 lglg0.5 0.62v 0 v 1

20 lglg0.5e0.45v 1 v 0

20 lglg0.4 0.12 0.1v 0.382 2.4 v 1

20 lglg 0.225/v v 2.4 (3-7)

8


(a) Negative height (b) positive height

Figure 3-5 Radio propagation past the cutting edge

9


1.3 Radio Propagation Model

Propagation model is very important. It is the basic of mobile communication in cell planning. Its value is to guarantee accuracy and to save labor, expense and time. Before planning a cellular system in an area, it is an essential task to select the cellular station address with signal coverage so as to avoid interference. If predictive method is not adopted, then the only one is cut-and-try method, which is carried out through actual measurement. Measure the coverage area of cellular station address to select the best one from all the suggested solutions. It is money-wasting and labor-wasting by adopting this method. We can easily select the best layout solution for cellular station address with accurate predictive method through computer calculation, by comparing and evaluating the performance of all the solutions output from the computer. Therefore, we can say that the accuracy of propagation model not only influences on whether the cell planning is proper, but also on whether the operators can invest rationally to satisfy users’ needs. With a vast territory, radio propagation environment is various in provinces and cities. For instance, propagation environment and propagation models have great differences between cities in plain area and the ones in hills area. Therefore, to ignore different factors of landforms, physiognomy, buildings and vegetation and consider experience will only result in network problems of coverage and quality or in resource wasting because of too close base stations. A good mobile radio propagation model is flexible to adjust according to different landforms, such as plains, hills and mountains, or different man-made environment, such as open areas, suburb and urban areas, etc. These environmental factors, involved in many variables in propagation model, play an important role. Therefore, it is not easy to form a good mobile radio propagation model. In order to improve models, statistical method is used to measure a large number of data and correct models. Correction for propagation model will be introduced in section 3. 4.

Also, a good model should be easy to use. Models should be clear enough not to give users any subjective judgment and explanation, for different predictive value can be deduced from that in the same area. A good model shall have good recognition and acceptability. Using different models might have different structures. Good recognition is very important.

Most of models predict path loss in radio propagation path. Therefore, propagation environment plays an important role in radio propagation model. Main factors involved in propagation environment in a specific area are:

Natural (mountains, hills, plains and water area);

Quantity, height, distribution and material characteristics of man-made buildings;

Characteristics of vegetation in the area;

Climate conditions;

Conditions of natural and man-made electromagnetic noise.

In addition, radio propagation model is affected by system working frequency and mobile station movement. In the same area, different working frequency results in different receiving signal fading; stationary mobile station differs high-speed moving mobile station in propagation environment. Generally, it is divided into two types: outdoor propagation model and indoor propagation model. Commonly used models are shown in Table 3-1.

10


Table 3-1 Common Propagation Models

Model name Scope of ApplicationOkumura-Hata Applied to 150-1000 MHz macro cellular predicationCost231-Hata Applied to 150-2000 MHz macro cellular predication

Cost231 Walfish-Ikegami Applied to 900 and 1800 MHz micro cellular predication

Keenan-Motley Applied to 900 and 1800 MHz indoor predication

Used in ASSET planning Applied to 900 and 1800 MHz macro cellular predication

Below is the brief introduction of Okumura-Hata model and Cost231-Hata model as well as the propagation model used in ASSET network planning software. Hata model is composed of the average data measured in Japan. Path loss value in general areas can be approximately represented with the following equation:

Lp 69.55 26.16 lglg f 13.82 lglghb

(44.9 6.55 lglghb ) lglg d AOkumurahm (3-8-1:Okumura-Hata)

Lp 46.3 33.9 lglg f 13.82 lglg hb

(44.9 6.55 lglghb ) lglg d ACost231hm Cm (3-8-2:Cost231-Hata)

Lp---Path loss from the base station to the mobile station, unit: dB

f ---Carrier wave frequency, unit: MHz;

hb---Antenna height of the base station, unit: m;

hm---Mobile station antenna height (1-10 m), having average value 1. 5 m, unit: m;

d ---Distance between mobile stations, unit: km;

Cm--The value is 0dB in medium-size cities or in suburb with medium woods density, while 3 dB in big cities.

AOkumurahm－－ (1.1lglg f 0.7)hm (1.56lglgf 0.8)MS height correction, value in medium sized citiesAOkumurahm－－ (1.1lglg f 0.7)hm (1.56lglgf 0.8)MS height correction, value in medium sized cities;

The value in big city is 3.2logloglog11.75hm2 4.97 (with frequency more than 400MHz);

ACost231hm＝（1.1 lglg f 0.7）hm (1.56 lglg f 0.8);

In suburb area, propagation model can be revised as

Lps Lp 2[lglg(f /28)]2 5.4Urban areaLps Lp 2[lglg(f /28)]2 5.4Urban area

(3-9)

In open areas, the propagation is revised as

Lpo Lp 4.78(lglg f)2 18.33 lglg f 40.94Urban areaLpo Lp 4.78(lglg f)2 18.33 lglg f 40.94Urban area

(3-10)

In the actual radio propagation environment, various relief shall be considered, which is considered in ASSET planning software to improve propagation model. Consider

11


various ground objects and relief having influence on radio propagation in actual environment so as to guarantee the accuracy of prediction result.

The model expression is as follows:

Lp K1 K2 lglgd K3(hm ) K4 lglghm K5 lglg(Heff )

K6 lglg(Heff ) lglgd K7 diffn Kclutter

In the above expression (the following expressions are applied to macro cell):

K1---the constant related to frequency;

The center of medium-size city:

K1=69. 55+(26. 16+1. 56lg(Fc))-0. 8 {Fc=150-1000MHz}

K1=46. 3+(33. 9+1. 56)lg(Fc)-0. 8 {Fc=1500-2000MHz}

Center of big city

K1=69. 55+26. 16lg(Fc) {Fc=150-1000MHz}

K1=46. 3+Cm+(33. 9+1. 56)lg(Fc)-0. 8 {Fc=1500-2000MHz}

Suburb area:

K1=69. 55+(26. 16+1. 56lg(Fc))-0. 8 -2(log(Fc/28))2 - 5. 4{Fc=150-1000MHz}

K1=46. 3+(33. 9+1. 56)lg(Fc)-0. 8 -2(log(Fc/28))2 - 5. 4{Fc=1500-2000MHz}

Open area:

K1=69. 55+(26. 16+1. 56lg(Fc))-0. 8-4. 78(log(Fc))2+18. 33log(Fc)-40. 94 {Fc=150-1000MHz}

K1=46. 3+(33. 9+1. 56)lg(Fc)-0. 8-4. 78[log(Fc)]2+18. 33log(Fc)-40. 94 {Fc=1500-2000MHz}

K2---Distance fading constant;

K3、K4---The revision coefficient of mobile station antenna height;

K5、K6---The revision coefficient of base station height;

K7---The revision coefficient of diffraction;

Kclutter---The revision coefficient of ground object in the prediction is: the field density of the prediction point is revised based upon the clutter type of that point, and has nothing to do with the clutter type in the propagation path. And all the losses in the propagation path lie in the medium value loss;

d---Distance between the base station and the mobile station, unit: km;

hm、heff---The available height of mobile station and base station antenna, unit: m.

As to the radio propagation in different areas and cities, K value will have different value owing to different landform and relief as well as different city environment. K value and some clutter fading values used in radio propagation analysis in medium-size cities are shown in Table 3-2.

12


Table3-2 K parameter value

K parameter name Parameter value

K1150/900MHz Urban, 160/1800MHz Urban

146/900MHz Large city, 163/1800MHz Large cityK2 44. 90

K3-2. 54/900MHz Urban,-2. 88/1800MHz Urban

0/900MHz Large city,-2. 88/1800MHz Large cityK4 0. 00K5 -13. 82K6 -6. 55K7 -0. 8

Clutter fading valueInland Water -3. 00

Wetland -3. 00Open Areas -2. 00Rangeland -1. 00

Forest 13. 00Industrial & Commercial Areas 5. 00

Village -2. 90Parallel_Low_Buildings -2. 50

Suburban -2. 50Urban 0

Dense urban 5High Building 16

Medium value of propagation loss can be calculated according to these K values. However, thanks to the complicated environment, some revision is required. Building loss is to be considered when the cellular mobile communication is used indoors. Building loss refers to the functions of wall structure (steel, glass and bricks, etc), building height, building direction, percentage coverage of the window area. Owing to complicated variables, building loss can be only calculated based upon the surrounding environment. Below are some conclusions we draw:

The average penetration loss in urban buildings is more than those in suburb areas and remote areas.

Loss in the area with window zone is generally less than that without window zone.

Loss in the open area within buildings is less than that in the wall area with corridors.

Fading in street wall with aluminum support frame is more than that without aluminum support frame.

Loss in the building with isolation only added to the ceiling is less than that in the building with isolation both added to the ceiling and inside walls.

There are two frequencies in GSM mobile communication system, i.e. e. 900MHz and 1800MHz. Different frequency results in different propagation characteristics. The longer the wavelength is, the less the diffraction loss is. While the relation between wavelength and penetration loss is worth further study, or is uncertain. In addition, indoor radio components are the superimposition of penetration components and diffraction components, and the diffraction accounts for the majority. Therefore, generally speaking, 1800MHz level difference between indoors and outdoors is larger than 900MHz. However, the problems of complicated propagation environment and the direction of incident wave make it impossible to quantize indoor-and-outdoor level difference. The best method is to test indoor-and-

13


outdoor level difference in a specific environment, so as to optimize the plan.

The average floor penetration loss refers to the function of the floor height. According to record data, the slope of loss line is -1. 9dB/story. The average penetration loss in the first floor is about 18dB in urban area, and 13dB in suburb area. The measurement of specific floors shows that loss characteristic inside buildings can be treated as a waveguide with fading. For example, when radio propagates along the corridor direction, which is vertical to the outdoor window, the loss can reach to 0. 4dB/m.

Tunnel propagation loss shall be considered when calculating radio propagation in tunnels. At this moment, simply regard the tunnel as a wave-guide with loss. The experiment result shows that propagation loss in a specific distance reduces as the frequency increases. When the working frequency band is below 2GHz, the relation between the loss curve and working frequency show exponential fading. As to GSM frequency, it can be approximately considered that loss and distance appear the inverse exponential change of fourth power, i.e. e. if the distance between the two antennas doubles, then the loss increases by 12dB.

Besides, the influence of leaves on propagation in UHF frequency shall be considered. Studies show that, in general, the signal loss in summer is about 10 dB more than that in winter, vertically polarized signal loss is more than the horizontally-polarized one, for leaves flourish in summer.

Radio battle-sight distance might be very far in wide coverage, such as desert or sea. The earth curvature shall be considered under such conditions. Assume that the earth

radium is (unit: m, the equator radium is 6378000m), hm、heff is the height of mobile

station antenna and the base station antenna respectively, the unit is m,hb is the height of base station antenna, the unit is also m, then the battle-sight range of radio

wave is d (unit: m).

d 2 heff + 2 hm

Per contra, if the expectation coverage range is known (when path loss is not the major factor), the base station height can be calculated.

14


1.4 Correction for propagation model

1.4.1 CW Basics

Correction for propagation model is required to obtain the radio propagation model in accordance with local actual environment and to improve the accuracy of coverage predication so as to lay a good foundation for network planning. CW test, say, continuous wave test, is a necessary step for model correction by correcting data obtained from CW test and digital map. The information of latitude and longitude of these test data and incoming level form the data source of model correction.

Random process theory is used to analyze mobile communication propagation, which can be expressed as follows:

r(x) m(x)r0(x) (3-11)

In which, x is distance, r(x) is incoming signal; r0(x) is Rayleigh fading; m(x) is local value, i. e. the mixture of long-term fading and space propagation loss, which can be expressed as follows:

m(x) 12L

xL

x L r(y)dy

(3-12)

In which, 2L is the average sample interval length, also called intrinsic length.

CW test is aimed to obtain the local average value of various locations in an area on whole way, that is, the difference between r (x) and m(x) is as small as possible. Therefore, The influence of Raleigh fading must be removed so as to obtain the local average value. When a group of signal data r (x) is averaged, if the intrinsic length 2L is too short, then the influence of Raleigh fading still exists; if 2L is too long, then the normal fading will be averaged. Therefore, in CW test, to determine 2L has great influence on the degree of approximation between the tested data and the actual local average, as well as on the accuracy of the propagation model prediction corrected through CW test. Li Jianye, a famous communication expert, has proved that, in GSM system, the intrinsic length is 40 wave lengths; the difference between the tested data and the actual local value is less than 1 dB by sampling 50 sampling points (the test equipment and the error of digital map are ignored).

1.4.2 CW Test Method

I. Select station for CW test

Before testing, test stations and quantity need to be determined. According to experiences, at least 5 test stations in big cities with dense population; as to medium-size and small-size cities, 1 test station is enough, which is mainly depended on the antenna height of test base station and the effective radiant power (EIRP). The principle of station selection is to cover ground objects as many as possible (these ground objects come from digital map)

In actual test, proper test stations can be selected according to the following standards:

(1) The antenna height is above 20 meters;

15


(2) The antenna height is above 5 meters over the nearest obstacle.

5m

Figure 3-6 Diagrammatic representation of station selection standard

The obstacle here refers to the highest building at the top of which the antenna is located. The building as a station shall be higher than the average height of the surrounding buildings.

II. CW test preparation

CW test first needs a test base station to transmit RF signal with or without FM modulation, then make a drive test by using CW test equipment. The base station includes transmitting antenna, feeder cable, power amplifier and HF signal source. The test system includes test receiver, GPS receiver, distance measuring instrument, test software as well as portable PC. The sampling rate of the test receiver is as fast as possible.

After the equipment of base station is installed in the selected test station, using the power meter to measure the forward power and reflection power of the antenna. Calculate the EIPR of the test base station. The calculation formula is below:

EIRP 10 lglg[P_forward(mW) P_reflect(mW)]

Tx_Antenna_Gain Rx_Antenna_Gain

Rx_Feeder_Loss (3-13)

In which, P_forward is forward transmitting power, P_reflected is reflection power, Tx_Antenna_Gain is the transmitting antenna gain of the test station (dBi), Rx_Antenna_Gain is the antenna gain of the test receiver (dBi), Rx_Feeder_Loss is the feeder cable loss of the test receiver.

After normal installation and debugging of the base station equipment, record the EIRP of the base station. Use GPS to measure the latitude and longitude of the station; use triangulation method to measure the height of the building, and use angle instrument to test the slope angle of the antenna. The antenna height is the height of the building plus antenna mast height and half the antenna height. Sweep frequency by using portable test equipment to ensure the normal work of the test base station equipment, without any interference signal in surroundings.

III. CW test

There are three sampling ways of the professional CW test equipment: sampling by time, pulse and distance. General test equipment samples by time only. Test by distance sampling can meet the Theorem of Lee’s requirement of sampling 36~50 sampling points with 40 wave lengths. The measure accuracy is very high. Speed is

16


not strict in distance sampling, but there exists an upper speed limit. The upper speed limit (Vmax) has relation with the maximum sampling speed of CW equipment:

Vmaxmaxmax 0.8 /Tsample (3-14)

During the test, test paths with various ground objects are selected as random drive test. When the mobile station is within the distance of 3km away from the test base station, the receiving signals are affected greatly by the building structure around the base station, and the antenna height. The intensity difference between the signal level parallel to the signal propagation direction and that vertical one is around 10dB. Therefore, when testing on the street within 3km in radium of the base station, it is better to sample the same amount of samples in longitudinal and lateral streets to remove their effects. Test paths should not be selected on highways and on the wide and flat streets, but on the narrow streets. Sample as much data as possible in each test base station. Generally, it is better to test in each station over 4 hours. Stop recording when the car stops for redlight.

The landform and ground objects are fixed within a period of time, so that in a deterministic base station, the local average value is determined in a deterministic location. The local average value is the data tested through CW test expectation, which is also the closest value to propagation model predication value.

1.4.3 Correction for Propagation Model and Instance

Digital map is needed for model correction. The digital map used in mobile communication contains the geographical information such as relief height and ground usage, which effect radio propagation in mobile communication. It is the important fundamental data for planning software in model correction, coverage prediction, interference analysis and frequency planning.

Propagation models developed for computer aided analysis are different, but based on Okumura basic models, and provide modified parameters. Below is the specific method of model correction based upon the above-mentioned ASSET planning software. It needs to be pointed out that if the model parameters of the city similar to the existing landform and ground objects, they can be directly applied to planning prediction. It is unnecessary to redo CW test and model correction, thus saving labor.

Parameters from K1 to K7 in ASSET model are determined by specific propagation environment, K(clutter) is the correction factor depended on different ground objects. Different ground objects determine different K(clutter), these K parameters are gradually fitted from CW test data. When CW test data obtained, K parameters can be acquired in two ways: K parameter testing method and the minimum variance method.

Among a great many of K parameters in the standard model, the degree of influence of each K parameter is different. By analyzing the models, we know that K1 and K(clutter) are constant, which has nothing to do with the propagation distance and antenna height; K3 and K4 are the height modifying factor of the mobile station. The mobile station has slight changes in height (about 1. 5m), so that K3 and K4 can be eventually classified as micro-adjustment in the final stage, while the adjustment of K2, K5 and K6 are determined by specific test data and test path.

17


1.5 Doppler Effect and its Impact on Handover

In GSM system, the relation of frequency change caused by Doppler effect is given through the following formula:

(1) the base station is the frequency source f, the frequency fˊreceived by the mobile phone is

fˊ=f(1±V/c) (3-15)

In the formula, v is the travel speed of MS, c is the radio signal propagation speed (3E8 m/sec)

Select “+” when MS moves towards the base station and select “-” when it is away from the base station.

(2) MS is the frequency source f, and the frequency fˊreceived by the base station is

fˊ=f/(1±U/c) (3-16)

In the formula, u is the travel speed of MS, c is the radio signal propagation speed (3E8 m/sec)

Select “-” when MS moves towards the base station and select “+” when it is away from the base station.

Below are several special conditions discussed:

(1) MS moves towards BTS at the speed of v, as shown in Figure 3-7.

Figure 3-7 MS moves towards BTS

The signal frequency of BTS is f1. Through FCH channel of BCCH channel, BTS can control MS to synchronize the frequency with BTS. MS receives the signal frequency f2 because of the Doppler Effect, and transmits f2 to the base station. f3 is the frequency received by BTS because of the Doppler Effect. Below is the formulas based on the above-mentioned:

f2=f1(1+v/c)

f3=f2/(1-v/c)

f3=f1(1+v/c)/(1-v/c)=f1(c+v)/(c-v)

The relative frequency change is (f3-f1)/f1=2v/(c-v) (3-17)

18


(1) MS moves away from BTS at the speed of v, as shown in Figure 3-8.

Figure 3-8 MS moves away from BTS

The signal frequency of BTS is f1. Through FCH channel of BCH channel, BTS can control MS to synchronize the frequency with BTS. MS receives the signal frequency f2 because of the Doppler Effect, and transmits f2 to the base station. Frequency f3 received by BTS because of the Doppler Effect, below are the formulas based upon the above-mentioned formula:

f2=f1(1-v/c)

f3=f2/(1+v/c)

f3=f1(1-v/c)/(1+v/c)=f1(c-v)/(c+v)

The relative frequency change is (f3-f1)/f1=-2v/(c+v) (3-18)

The travel speed of MS is slow as compared with signal propagation speed; therefore relative frequency change is almost the same in these two conditions except for the opposite direction. Frequency increases in the first condition, while decreases in the second one.

The relation between the relative frequency and MS speed can be illustrated in Figure 3-9.

Figure 3-9 Graph of relation between the relative frequency and MS speed

The graph shows that when MS speed is 100km/h, the relative frequency change is 0. 19ppm. As to 900M frequency, the deviation is 171Hz, while 342Hz as to 1800M.

(3) MS moves between the two base stations at the speed of v, as shown in Figure 3-10. When handover is performed, the deviation is the superimposition of the above

19


two conditions. MS obtains the monitoring information of the BCCH channel of the neighboring cells through BA table, controls MS to adjust its frequency and a certain number of kHz to monitor the neighboring cell level. Thus, it might appear Doppler frequency changes, which make MS unable to receive the signals of the neighboring cells correctly. Take the Figure 3-10 as an example, MS monitors BTS1 level, the signal f2ˊ received by MS might appear between the two MS adjustment frequencies. So that MS cannot correctly monitor BTS1 signal level. On the other hand, RXlev information reported from SACCH shall be transmitted at least once every 30s. Such long time information report will also result in abnormally monitoring the neighboring cells level, which causes unsuccessful handover. The frequency change caused by the Doppler Effect will effect the signal frequency f1(c+v)/(c-v) received by the base station, which will receive data by f1 sampling clock. Receiving data error might be another reason for effecting handover.

Figure 3-10 MS moves between the two base stations

20


1.6 Fresnel Zone

There are direct wave and reflected wave in the propagation path from the transmitter to the receiver, and the electric direction of the reflected wave is just opposite to the original with the phase difference of 180 degree; if the antenna height is relatively low and the distance is relatively far, the difference between the direct way path and the reflected wave path is small, then the reflected wave will cause destruction. In

addition, the path difference between the direct wave and the reflected wave is 2h thr

d ,

the phase difference is 4 h thr

d , h t ,hr refer to the height of the transmitter and

the receiver above the ground respectively, d is the horizontal distance from the transmitter to the receiver, as shown in Figure 3-11.

Figure 3-11 Graphs of Direct incidence and reflection

Ignore part of the signals from the transmitting point to the receiver through ground wave propagation (signals in ultra-high frequency and very-high frequency band can be ignored), then the square of the ratio of the total receiving field density an the free space density (unit: V/m) is:

Erec

Efs2 4 sinsinsin(

2 ) 4 sinsinsin22 hthr

d (3-19)

The formula shows that n is a natural number, when is（2n－1） , it can

generate 6dB signal power gain; while when is 2n , the two signals can be offset. The change from this point is caused or caused together by the change of antenna height and propagation distance.

The simulation result shows that when d is less than 4hthr

, 2 is more than

2 , then

the gain obtained swings as the mobile station moves towards the base station; when

d is more than 4hthr

, 2 is less than

2 , the gain won’t swing as the mobile station is

away from the base station.

In the actual propagation environment, the first Fresnel zone definition contains some ellipsoids of reflection points, on these reflection points, the path difference between

the reflected wave and direct way is half a wavelength, say, 2 less than

2 , as shown

in Figure 3-12. The first Fresnel zone is the main propagation zone, when obstacles

21


don’t block the first Fresnel zone, the diffraction loss is least. As to a point in the path

with d in length, its radium of the first Fresnel zone (the distance to the transmitter is

d t , and dr is the distance to the receiver) is:

h0m dtdr

d 548dtkmdrkm

dkmfMHz (3-20)

Figure 3-12 The radium of the first Fresnel zone

Take an example to illustrate that: in typical cities, a point in the path with the coverage range is 2km; suppose that the distance from this point to the transmitting antenna is 100m, as to the frequency of 900MHz, this point’s first Fresnel zone is h0 5m.

On the definition basis of the first Fresnel zone, define the nth Fresnel zone as the reflection-point set, in which its propagation is half wavelength more than the n-1 th; the phrase difference between the two reflection paths is 180 degree. The radium of the nth Fresnel zone is:

hnm ndtdr

d 548ndtkmdrkm

dkmfMHz (3-21)

If the direct path jumps over the wavy terrains and ground buildings, then the reflected wave will have positive effect on direct wave; otherwise it might become the obstructive multi-path interference. The obstructive effect grows as the frequency increases. Therefore, the height of antenna shall be built as high as possible above the ground. This conclusion will be applied to the below-mentioned antenna project designing. As a matter of fact, according to experience, if 55% of the first Fresnel zone, used for stadia microwave link designing remain unobstructed, then the conditions of other Fresnel zones won’t affect the diffraction loss.

22


1.7 ASSET Software Introduction

Below is the brief introduction about the above-mentioned ASSET planning software.

ASSET software is the network planning software designed by Aircom company. By using ASSET software, we can configure the system hardware parameters, network capacity, frequency allocation and complete the network design (such as coverage prediction, traffic analysis, neighboring cell allocation, frequency plan, interference analysis and microwave propagation and so on) and simulate the network operation effect to guide the project construction.

Before using the ASSET software, we need to prepare and know the following information:

Digitalized map with proper ratio of accuracy. The accuracy of the digitalized map includes 20m, 50m, 100, 5m. 20m accuracy is applied to urban area and suburb area, 50m or 100m accuracy can be applied to rural area, while 5m accuracy is generally used for micro cellular planning;

Network and base station information mainly includes the configuration of MSC and BSC, latitude and longitude of the base station, antenna type, feeder system parameters;

Design consideration includes the purpose of this planning, network hierarchy, frequency range and frequency reuse mode, cell frequency hierarchy, configuration of network functional parameters.

ASSET network planning is carried out on the basis of digital map. Digital map is a map for record and storage in digital form; digital map is convenient to store, transmit and update, which can be transformed into paper map by processing in computer, or displayed on the computer screen by visual processing. Owing to different storage structure, digital map can be divided into vector digital map and grid digital map (such as scanning map). In order to cover prediction, we usually use vector digital map. Map data usually is composed of three data types: digital elevation model (DEM), digital object model (DOM) and linear vector model (LDM). Construction vector data is also applied to micro cellular prediction. Digital ground elevation model and ground object disaggregated model has related to prediction. Digital ground elevation model is used to describe the basic relief of this area and directly participate in the calculation of radio propagation model; ground object disaggregated data is used to describe planar ground coverage, such as forest, lakes, open area, industrial area, downtown, high building area and so on, and used to calculate radio propagation path loss; LDM is used to describe the relation between the plane distribution and the space of linear ground objects, including highway, streets, rivers and so on. DEM data and DOM data adopts the grid data format, each grid represents a sampling point; while LDM adopts vector data.

Before officially beginning planning, the following work needs to be done:

(1) Define the parameters related to propagation model and feeder system, and input antenna database

(2) Define layer, determine the frequency reuse mode

(3) Add network element with multi methods (MSC, BSC, BTS and cell layer)

(4) Improve various parameters in the database

23


Thus, we can make use of ASSET software to complete the entire planning process.

24

Chapter3 Radio Propagation Theory

Documents