-
4. Base Station Antenna Siting and Measurement• This chapter
starts by presenting the design procedure used for siting base
station antennas in
cellular systems. The site location is based on statistical data
of the propagation characteristics, which are summarized as a set
of compact equations derived from these data. This is called the
Okumura-Hata estimation method.
1. This chapter describes this estimation method in detail,
together with another method that is appropriate for the design of
small cells.
2. Following site location, the base station antenna is then
designed to satisfy the specifications. The design procedure for a
cellular system is the second topic in the chapter.
3. At the development stage, the designed base station antenna
characteristics are evaluated by making use of prototype models.
The measurement method, particularly for cellular base station
antennas, is the third topic described in the chapter.
4. A unique measurement for base station antennas is the high
power test. This chapter describes the special features associated
with the high power treatment of the antenna and its measurement
method in detail.
5. The last topic in the chapter is the evaluation of the
performance of base station antennas. Tests needed before product
shipping include electrical, mechanical and reliability performance
checks. These tests are described in detail together with practical
examples. 1
-
4.1 Base Station Antenna Siting• The location of the base
station is the most important issue for cellular system design.
• The coverage area of a cellular base station ranges from a
radius of 0.5 km to 20 km, and is determined by the number of
expected users inside the coverage area and the geographical
features in the locality.
• The Okumura-Hata method is widely used to estimate the
propagation loss from a base station with a cell radius between 1
km and 20 km.
• The Okumura-Hata estimation method described in this section
has been used recently in micro-cellular systems with a cell radius
below 0.5 km, which are employed to increase the user capacity of a
single base station.
• The propagation loss estimation method is also presented for a
cell of a similar small size. The selection of base station
location is based on these propagation loss estimations.
2
-
4.1.1 Design of Macro Cell Siting Using the Okumura-Hata
Curve
• In mobile communication systems, the propagation
characteristics are predicted by an estimation method based on
large quantities of data obtained from propagation loss
experiments, as described in section 1.2.2.
• Geographical features inside a range of several tens of
kilometers are roughly divided into three different types described
below, and the propagation loss is estimated for each type. This
technique is well known as Okumura’s curve.
• The variation of the propagation loss evaluated from
measurement data at 150 MHz, 450 MHz, 900 MHz, and 1500 MHz defines
the original Okumura curve. This section explains how this method
can be used to estimate the propagation loss, as described in
section 1.1 [1].
• Geographical features may be divided roughly into irregular
and semi-flat. Irregular features are classified into hills,
isolated mountains, inclined geographical features and so on.
• Land mobile communication services are however mainly used in
locations with semi-flat geographical features. Within semi-flat
geographical features, an undulating plain is defined as having
heights less than 20 m. Local geographical features and buildings
are described by a compensation value based on a place where the
undulations are small.
3
-
Fig. 4.1 Okumura’s curve for city area in 900 MHz band• Such
semi-flat geographical features are classified as being in the
city, in the suburbs, and in
open areas. An interpretation for a city area, where the
buildings have more than two stories, is that the proportion of
buildings is equivalent to 23 % of the area [2].
• Fig. 4.1 shows an example of Okumura’s curve in the 900 MHz
band for a city area. Distances below 10km from the base station
are shown to have a log scale path loss, while greater distances
have a linear scale loss; this is a unique feature of Okumura’s
curve.
Effective height of base station antenna(m)
free space
210
200
190
180
170
160
150
140
130
120
110
100
901 2 5 10 20 30 40 50 60 70 80 90 100
305070100150200
450300
6008001000
log scale linear scale
Distance(km)
Path loss between dipole antennas[dB]
4
-
• It is convenient to calculate data using a closed form
equation, although propagation loss can be read directly from the
Okumura curve. The approximate equation of the Okumura curve, as
derived by Hata, is shown below [3].
• The propagation loss LP in a city area is expressed by an
equation at a frequency f MHz (150 ∼ 1500 MHz), for a mobile
station antenna of height h m (1∼10 m), with a distance between
base station and mobile station of r km (1∼20 km) as follow:
(4.1)• where the antenna effective height a(hm) is defined as
the compensated antenna height of the
mobile terminal. (4.2)
• For large cities:f ≤ 200 MHz (4.3a)
f ≥ 400 MHz (4.3b)
dBrhhahfL bmbp 10101010 log)log55.69.44()(log82.13log16.2655.69
−+−−+=
)8.0log56.1()7.0log11.1()( 1010 −−−×= fhfha mm
1.1)54.1(log29.8)( 210 −= mm hha
97.4)75.11(log2.3 210 −= mh
5
-
Fig. 4.2 Definition of antenna effective height
BS antenna
0 3km 15km
Average heighthm
Earth surface
• As shown in Fig. 4.2, the antenna effective height is
approximated by the height of the average surface of the earth at a
point where the mobile terminal is used at between 3 km and 15 km
from the base station antenna
• Next, the propagation loss is considered in areas which may be
classified as suburban or pastoral and these include areas such as
those along a road where trees and houses disturb the propagation
of electromagnetic waves. Under these conditions:
dB (4.4)• In open areas where high trees, buildings, and other
obstructions do not exist within a range of
300∼400m of the mobile station in the direction of arrival, the
propagation loss is given as:dB (4.5)
4.5)}28/({log2}{ 210 −−= fcityLL pp
94.40log33.18}{log78.4}{ 102
10 −+−= ffcityLL pps
6
-
Fig. 4.3 Propagation loss estimated by Okumura-Hata method
100
110
120
130
140
150
160
170
180
190
200
Path loss between dipole antennas [dB]
200 300 400 500 600 700 800 900 1000
Frequency [MHz]
1500150
distance=20km
distance=10km
distance=5km
distance=2km
Effective antenna height of base station 50m
Antenna height of mobile 1.5m
• The propagation loss estimated by the Okumura-Hata method is
shown in Fig. 4.3 as a function of frequency. A coverage area with
a radius of more than 1 km of a base station can be obtained using
this propagation loss estimation method.
7
-
• It is possible to calculate the output power and antenna gain
of the base station from the propagation loss estimated by the
Okumura-Hatte method. If Pt, Gt and Ms represent the transition
power, antenna saturation gain and system margin, respectively, the
received electric power Per at the cell edge can be formulated as
follows:
(4.6)• The system margin is between 10 and 20 dB and is caused
by the unexpected leveling down
of the received electric field strength due to fading.
• The noise level at the location of the base station determines
the cell edge of a cellular system.The noise level in the frequency
range of 800–2000 MHz is between 0 and 10 dBµV, whichmay be
converted into the unit of dBm by the conversion equation, or as
follow:
(4.7)• where a feed line having an impedance of 50Ω is
assumed.
• The propagation loss at 900 MHz may be computed for base
station antenna with heightshb=50m and hm=1.5m.
• If the cell radius of the base station in big cities is
assumed to be 3 km, the propagation loss Lpbecomes 139 dB from
(4.1) through (4.3). If the minimum electric field strength
received atthe cell edge is 10dBµV=-103dBm and the system margin
Ms=20dB, then the followingequation applies:
(4.8)
sptter MLGPP −−+=
01.113)()( −= VdBVdBP emr µ
5620139103 =++−=++=+ spertt MLEGP
8
-
• An antenna gain of 13 dB is necessary for this estimation if
the output power of the base station is 20 W, which is equivalent
to 43 dBm.
• Based on this procedure and other factors such as the coverage
area of the base station and the gain of the base station antenna,
the transmitter output can then be determined.
• After the above statistical approach, Ikegami introduced a
propagation model consisting of free space propagation loss and
reflected/detracted loss by buildings [4].
• Walfisch derived the propagation loss assuming an urban
environment as semi-infinite multi screen [5]. In this model, the
diffracted wave at the screen top is a dominant factor at received
signal at mobile terminals.
• These two models are integrated as “COST-231 Walfisch-Ikegami”
model, which is widely used in recent design of cellular system
[6][7].
9
-
Antenna System Design: ExamplesPropagation loss at 1500 MHzthe
height of the base station antenna hb = 30 mthe height of the
mobile station hm = 1.5 m. the cell radius of the base station in
cities is 1.5 km
the propagation loss Lp at 1500 MHz is ? dB by Okumura-Hata
curvethe minimum electric field strength received at the cell edge
is 10 dBµV = -103 dBmthe system margin Ms = 20 dB
?=++=+ spertt MLEGP dB
An antenna gain of ? dB is necessary for this estimation if the
output power of the base station is 20 W, which is equivalent to 43
dBm.
10
-
4.1.2 Break Point of a Micro cell• Okumura’s curve is derived
for early land mobile communication systems for which the
coverage area of the base stations is assumed to be from several
kilometers to 10 km in radius.
• Therefore, it is difficult to apply to a case where the
coverage area is below 1 km from a base station, such as a micro
cellular system. This small cell sized is used to increase the
capacity and high-speed data transmission capability.
• The propagation loss of micro cells is based on line of sight
propagation (defined in sec. 1.2.1), and a break point is defined
at the position where the propagation loss changes rapidly
[8][9].
• The break point can be explained by the two-wave model, by
considering the direct path and the ground reflected path. The
earth is approximated as a flat surface, as shown in Fig. 4.4,
where the propagation distance is denoted by d1 for the direct path
between transmission and reception point, and d2 for the ground
reflected path.
d1
d2
d
hb
hm
Fig. 4.4 Two-wave model with flat earth
11
-
• When the heights of the transmitting point hb and the
receiving point hm are very small as compared to the distance d
between the transmitting and receiving points, path difference
between these two paths is given by the following:
(4.9)
• The propagation loss increases rapidly when the path
difference given by (4.9) is more than a half-wavelength, because a
reflection wave from the ground cancels a direct wave, due to the
180o phase difference.
• The point where the path difference satisfies the above
condition is called a break point, and the corresponding distance,
d2-d1, is defined as:
(4.10)
• However, because fading causes large local fluctuations in the
measured electric field, the position of the break point cannot be
clearly determined from measurement of the propagation loss.
222212 )()( mbmb hhdhhddd −+−++=−
−+−
++=
22
211
211
dhh
dd
hhd mbmb
dhh mb2≅
λmbhhb
4=
12
-
• Assume that a break point exists in an area between 100m and
500m from a base station, and an example measurement result is
shown in Fig. 4.5. From measurements at frequencies 457.3 MHz, 2.2
GHz and 4.7 GHz along a checker-shaped road in the city, the break
point position bm is given as [10]:
(4.11)• Quoted propagation loss estimations are based on
measured data at 457.3 MHz, 2.2 GHz, and
4.7 GHz in Yokohama, where main roads are checker-shaped. The
scaling factor of 1.9 is derived by the measurements.
• This reference has been selected because the frequency range
is broad and the closed form formula well approximates the
propagation loss before and after the break point.
• The propagation loss in the line of sight region is inversely
proportional to the distance r (free space loss), and attenuates
according to 1/r4 after the break point. If the same expression as
the Okumura curve is taken as the propagation loss, it may be
formulated by denoting frequency as f MHz and K as a constant
before the break point (db. The value of K is 27.6 in this
example.
λmbhhb
9.1=
KfdLbb +−−= log20log20
KchhbLbdL mbbbba +−+−=+−= log20)9.1log(20log40log40
13
-
Fig. 4.5 Propagation loss and break point
(a) 457.3 MHZ, (b) 2.2 GHz, and (c) 4.7 GHz.14
-
• The value of propagation loss estimated by the two-way model
is smaller than that obtained from the measured results because of
the effective raising of the earth surface by vehicles on the road
and the blocking of the propagation path by signboards and
pedestrian overpasses.
• These two factors are shown in Fig.4.6 and are introduced to
the propagation loss as:
(4.14a)
(4.14b)
• From the measurement carried out in Sapporo city, it is
reported that the above estimation agrees well when values of
s=0.001, and ho =1 m are used for daytime and ho 0.5 m for
nighttime [11]. This formula can estimate the propagation loss
without dividing the path by the break point.
rm
rjkrjksr
d reR
reerL
rmoo −−− += )(
4)(
1
1
απλ
{ }22 )()( omobrm hhhhrr −+−+=
Fig. 4.6 Two-wave model usingeffective earth surface
15
-
4.1.3 Propagation Loss Outside the line of Sight
• In a propagation model for a micro cellular system, there may
be a shape turn in a road at a corner and this results in a
non-line of sight scenario.
• Beyond the corner, the waves encounter multiple reflections
and diffractions the corner, as shown in Fig. 4.7. The effects of
the diffracted waves are usually small and can be neglected.
• The propagation loss may be divided into the loss at the
corner and that beyond the corner. In general, the loss at the
corner varies from 10 dB to 30 dB depending on the propagation
conditions, the distances between the transmitting and receiving
points, and the corner.
• The loss may be expressed in terms of the distance d2 from the
corner to the receiving point, as:
dB (4.15)• Reference [10] indicates that α has little frequency
dispersion and its recurrence curve may be
approximated by the measurement data as:dB (4.16)
• where d1 is the distance from the transmitting point to the
corner.
nnl KdL += 2logα
0.51log5.14 1 −= dα
16
-
• A further method for estimating the propagation loss uses the
total distance r=d1+d2 without considering the differences for the
regions inside or outside the line of sight.
• In this case, the propagation loss is the same as free space
loss for distance up to 30 m away from the transmitting point, and
is proportional to 1/r5 from there on.
• The corresponding expression does not depend on the location
of the break point.
dB (4.17)
Fig. 4.7 Propagation path at intersection
6.2730log30log20log50 ++−−= fdLnl
17
-
4.1.4 Propagation Loss Estimation by the Ray Tracing Method
• A large number of propagation experiments are required to
derive the approximate equation for the estimation of propagation
loss.
• Recently, the propagation loss, particularly within micro
cells, has been estimated by numerical simulations using the
ray-tracing method. The ray tracing method approximates
electromagnetic waves by optical rays, and is applicable for
simulation of reflection, diffraction, and scattering by buildings
and other obstacles.
• It is difficult to model all the real influences precisely.
For example, it is impossible to include all the influences of
street trees, building’s windows and window frames and moving
people inside buildings in the calculation model. However, accurate
modeling is indispensable when considering the precise changes in
electric field level at the receiving point.
• On the other hand, a relatively simple calculation model can
give the median level of electric field strength at the receiving
point and the average propagation loss [12].
• For example, when most of the buildings in a large city are
higher than the height of a base station antenna, the propagation
loss along a street can be approximated by a U-shaped waveguide,
for which all the surfaces are concrete walls, as shown in Fig.
4.8.
18
-
• The electric field strength at the receiving point in the
U-shaped guide is calculated as the sum of the direct waves from
the transmitting point and the reflected waves from the walls.
• A calculated result obtained using such an approximate model
is shown in Fig. 4.9. Once the height of the base station and the
radiation pattern of the base station antenna are known a priori,
propagation characteristics can be calculated. It is therefore an
effective technique for estimating the propagation phenomena in
micro cellular systems.
W
s
hb
0.1 1 10 100 1000-90
-60
-30
r (m)
dB
Fig. 4.8 U-shaped guide for ray tracing simulation Fig. 4.9
Calculated electric field distribution from ray tracing method: f =
1.9GHz, W = 10, s = 0.1, hb = 2.5, hm = 1.6 (m).
V-pol.
H-pol.
19
-
Side view
Strip lineSubstrate
Ground plane
Foam
Mounting plateTop view
Tunnel
Flat Antenna Base Station
E/O O/E
Optical Fiber
Unit
Crank-shaped ground plane Notched ground plane
Bi-Directional Notch Antenna
20/58
-
Tunnel Booster Antenna
21/58
-
Two modes of Bi-directional Notch
p : probe length
f1: Dominant mode is patch mode resonance.f2: Secondary mode is
probe mode resonance (p≒λ/4).
f1
f2
22/58
-
Current Distribution of Two modes
patch mode probe modecurrent flow
Patch mode currents radiate large cross pol. in E plane.
Patch mode currents should be suppressed for probe mode.
E plane
H plane
23/58
-
λ/2
Half wavelength dipole antennaOut of phaseexcitation
A Pair of Dipole Antenna for Bi-Directional Radiation
Pattern
Current direction
24/58
-
Radiation Pattern of Two Modes
a
d
b
y
pc
z
sp
x a=140, b=40, c=50, d=20,p=40, sp=60, g=30 [mm].
-20
-10
00ﾟ
90ﾟ
180ﾟ
270ﾟ
patch mode probe mode
-20
-10
00ﾟ
90ﾟ
180ﾟ
270ﾟ
patch mode probe mode
H (zx) planeE (yz) plane
25/58
-
-20
-10
00゜
90゜
180゜
270゜
notch without metal planenotch with metal plane
Notch Antenna
Metal Plane
z
x
y
θ
φ
Effect of Mounting Ground Plane
The radiation beam is tilted up by the finite sized ground
plane.
26/58
-
0 [degree]
Antenna
30 [degree]
60 [degree]•Antenna pattern ：sinαθ
θ [degree] 0 30 60Coverage Area [m] (800MHz) ±170 ±93 ±65
Coverage Area [m] (1.5GHz) ±140 ±96 ±60
Coverage Area of Tilted Beam Antenna
27/58
-
-120
-100
-80
-60
-40
-20
0
0.1 1 10 100 1000Distance [m]
Peak Value
Line of -20[dB]
Coverage
Definition of Coverage AreaPr
opag
atio
n lo
ss (d
B)
Monoploe antenna
28/58
-
p1=60(0.3λ), p2=10(0.05λ), p3=80(0.3λ),
m1=120, m2=260, h=40 [mm]
h
p1p2 p3
m1
m2
Notch Antenna
Metal Plane
Parasitic elements
z
x
y
θ
φ
Notch Antenna with Parasitic Elements
29/58
-
-20
-10
00゜
90゜
180゜
270゜
no element one element two elements three elements
Parasitic Elements 0 1 2 3Amplitude 1 0.21 0.21 0.24Phase
[degrees] 0 16.06 -1.05 -15.28Distance 0 0.3λ 0.35λ 0.4λ
Εφ (yz plane) radiation pattern
z
x
y
θ
φ
[dB]
Number of Parasitic Element
Phase and amplitude of parasitic elements
30/58
-
Undersea Tunnel Cellular Phone Service
31/58
-
4.6ｋｍ 4.8ｋｍ 4.4ｋｍ
15.1ｋｍ
Tokyo-Wan Aqua Line
32/58
-
1.5GHz (T)
1.5GHz (R)
800MHz (T)
800MHz (R)
Strip directors for gain enhancement (3 dB up)500×479×125 mm
High Gain Tunnel Booster Antenna
33/58
-
High Gain Antenna for W-CDMA
34/58
-
Low Cost Tunnel Booster SystemAdvantage:Very low transmission
loss by optical fiberDrawback:High cost of E/O, O/E unit
OpticalFiber
Booster by outside antenna
Bidirectional Antenna
Tunnel
E/O,O/E unit
Base Station
TunnelBaseStation
Receiver35/58
-
8.5 [m]
2.5 [m]
Receiver
f= 818[MHz]
V-pol. for mobile terminal
V/H-pol. for base station
Tunnel length 1005 m
Mt. Mt.
Road
Open space
52[m]19[m]
★
Base station
Propagation Measurement in Tunnel
36/58
-
52[m]h
-150
-140
-130
-120
-110
-100
-90
0 100 200 300 400 500 600 700 800 900
Prop
agat
ion
Loss
[dB
]
Distance [m]
h= 3 [m]h= 6 [m]h=14 [m]
Propagation Loss Inside Tunnel
37/58
-
2-D Simulation ModelConcrete wall @ 5GHz
1.00.5 jr −=ε0.00.1 jr −=µ
)(1.0 mD =
)tanh( DjkZ rrr
rin εµε
µ=
iin
iinTE Z
Zθθ
coscos
+−
=Γ
i
jkRmn
TEi ReE
i−+Γ=
A point of symmetry
38
-
Convergence Characteristics
3-reflection5-reflection
10-reflection
CW incidence
39
-
Image Method
∏−⋅=l
lirti RjkddiGiGE ,)exp(4)()(
πλ
22
1i
iP E E
=
= = ∑
ici
iciTE
θεθ
θεθ2
2
sincos
sincos
−+
−−=Γ
icic
icicTM
θεθε
θεθε2
2
sincos
sincos
−+
−−=Γ
R : Reflection coefficient, d : Distance, θ : Incident angle, G
: Directivity gain of antenna 40
-
Definition of Reflection
Incident Angle [deg]
Ref
lect
coe
ffici
ent [
dB]
0 10 20 30 40 50 60 70 80 90-60
-50
-40
-30
-20
-10
0
TE TM
Reflection coefficient
2 2
2 2
cos sin
cos sini i
TE
i i
n
n
θ θ
θ θ
− −Γ =
+ −
2 2 2
2 2 2
cos sin
cos sini i
TM
i i
n n
n n
θ θ
θ θ
− −Γ =
+ −2 60rn jε σλ= −
iθ
EH E
H
Reflection plane41
-
Transverse Electric (TE) Incidence
Incidence/Transmission
εr1=2, εr2=1θi=40°
"http://www-antenna.ee.titech.ac.jp/~hira/hobby/edu/em/polarization/rhcpw/index-j.html"
Reflection/Transmission
All components 42
-
Transverse Magnetic (TM) Incidence
εr1=1, εr2=2θi=80°
"http://www-antenna.ee.titech.ac.jp/~hira/hobby/edu/em/polarization/rhcpw/index-j.html"
Incidence/Transmission Reflection/Transmission
All components43
-
5.0ｘ7.0ｘ3.0 (m)Tx(2.0，1.0，1.0)Rx z=1.5 xy
planef=5.2GHz5-reflection
3-D Simulation Results
44
-
Time Delay Profile
10×10×3 (m)Tx(2,2,1)Rx(8,8,1.5)10-reflection
45
-
4.2 Design of Cellular Base Station Antennas
• This section presents the base station antenna structures
including the feeding network, and a synthesis method for pattern
shaping to give the desired characteristics. As uplink diversity
scheme is mandatory for base station systems, the design procedures
for space and polarization diversity are also described in this
section.
4.2.1 Antenna Array Structure and Feeding Circuit
• The horizontal radiation pattern is often omni-directional,
but a sector shaped pattern is sometimes used to increase the
system capacity.
• Such sector system covers the whole of the horizontal plane by
dividing it into several sectors. The sector pattern is given by
the antenna elements supplemented by reflector or parasitic
elements in front of the antenna, as shown in Fig. 4.10.
• A narrow sector pattern with a 60° half-power beam width
requires a two-element array in the horizontal direction, excited
in phase.
46
-
Reflector
Di-pole
Parasitic element
Fig. 4.10 Sector beam antenna using reflector and parasitic
element
(a) tournament array (b) colinear arrayFig. 4.11 Feeding network
for base station antenna array
47
-
• The antenna radiation pattern in the vertical plane consists
of a shaped beam to control the coverage area of the base station
and constitutes an array antenna in the vertical direction.
• A simple pattern used for cellular systems is a pencil beam to
increase the directive gain in the horizontal direction. This
pattern is given by an array in which the phases and amplitudes of
all the elements are the same.
• A cooperate-feed network, shown in Fig. 4.11 (a), provides the
in-phase excitation. This feed network is useful for high gain
antennas of more than 15 dBi, although an antenna gain of about 10
dBi may be obtained with a co-linear array shown in Figure 4.11
(b).
• Such co-linear array consists of several half-wavelength
coaxial cables connecting the inner conductor of the lower element
to the upper outer conductor in order to change the phase
alternatively at the edge.
• Although this antenna structure is very simple, the input
impedance is not matched at the frequency to give the required
radiation pattern. Impedance matching is then obtained by an
external matching network.
• Typical antenna elements for cellular systems are printed
dipole, microstrip and printed slot, as shown in Fig. 4.12. The
antenna element is selected according to the design of base station
and its cost. The feeding network is usually made with microstrip
line.
48
-
Fig. 4.12 Antenna elements for base station antenna
49
-
4.2.2 Beam Tilting and Propagation Characteristics
• To increase the subscriber capacity in current cellular
systems, a smaller coverage area is required for each base
station.
• In the frequency division multiple access (FDMA) system, the
same frequency can be used in near cells. A restriction in coverage
area is also seen in the code division multiple access (CDMA)
system, because the increase in the number of users in the same
cell raises the base noise level.
• Therefore, a similar cell form is required for both FDMA and
CDMA systems. To restrict the cell size, the beam tilting technique
shown in Fig. 4.13 is effective.
• In the FDMA system, the null position of the antenna pattern
is directed towards another cell in order to minimize the
interference between cells using the same frequency band.
• Using tilted beams, the propagation loss becomes larger than
that without tilting, as shown in Fig. 4.14, which also shows the
effectiveness of using tilted beam antennas
50
-
• For a system in which the antenna height of the base station
is h (m) and the cell radius is a(m), the tilt angle θt may be
determined as:
(4.18)• For example, the tilt angle is θt = 3.8° for a system in
which h=100 m and a=1.5 km. For
practical designs, which include the interference margin between
adjacent cells, the real tilt angle is 1 to 2°, which is larger
than the value given by (4.18).
• A design example of the vertical plane pattern for a base
station antenna is shown in Fig. 4.15. Denoting θ=0° as the
horizon, the lower region 0
-
0
10
20
30
40
50
60
70
0 5 10 15 20 25
Received electric field strength [dB V/m]
Distance from the base station [km]
m
tilt angle=0
tilt angle=3
o
o
Fig. 4.13 Beam tilting effect of base station antenna Fig. 4.14
Propagation loss by antenna main beam tilting
52
-
Beam Tilting Effect of Base Station Antenna
( )aht /tan 1−=θ
θt = 3.8° , h = 100m, a = 1.5km
θt = 1.15° , h = 30m, a = 1.5km
53
-
θt
0 +90-90
Side lobelevel
cosec θregion
2Taylordistributionregion
0
-10
-20
-30
-40
-50
dB
-90 -60 -30 0 30 60 90degree
Fig. 4.15 Example of vertical plane pattern for base station
antenna
Fig. 4.16 Desired radiation pattern in vertical plane: Number of
element=16; beam tilt angle=4° and sidelobe level=–13dB
θtcosec2θ
54
-
dx
y
θsinkdθ
Co-Phase Excitation for Tilt Beam
θsinkd−
θsin2kd
θsin2kd−
( )okd θθδ sinsin −=Phase difference for each antennaoθ
Direction of maximum radiation
Phase difference is required to be 30º for θ0=10 º, 16º for θ0=5
º @d=λ/2.55
-
Sub array
1
2
4
3
5
6
7
Radiationelement
Feednetwork
Phase shifter
Power divider
Feeding Network and Phase Shifter
56
-
Phase Shifter Arrangement
θ=10º as the maximum tilt angle
δm=30º as the maximum phase difference
om kd θδ sin=
4-element sub array needs 4× δm=120ºas the phase difference
between sub array
θ=1º, δm=3º, 4× δm=12º
Wide variable range required57
-
Variable Phase Shifter by Sandwich Dielectric Plates
Meander line
Moving direction
Moveable dielectric plate
Dielectric substrate
58
-
Moveable dielectric plate
Meander tri-plate line
#1
#1‘ #2‘
#2
Shield box
Shield boxMoving direction
Top view Side view
Variable phase shifter by sandwich dielectric plates
59
-
Theory of Phase Shifter
W
s
p
筐体 誘電体 (εd、厚さ d )
t
εb
Metal plate Dielectric plate (thickness d)
εbεe = b = p + s
1－2d / b+ (εb /εd )×( 2d / b)
Effective dielectric constant
bεe = εb εd≫εb
b – 2dεd >90
Thickness ratio betweendielectric materials
Guide wavelength oftri-plane line is;
ee ε
λλ 0=
60
-
Photo of Variable Phase Shifter
61/44
-
• To obtain the amplitude Ii and phase φi for the desired
antenna pattern, the error value should be minimized. Discreet
points of M, in the following vector, express the desired pattern
in the vertical plane:
(4.20)
• Using (4.19), g (θ) may be given in matrix form from and
as:
(4.21)
• If the error function to minimize the value of δ is ε, Gd and
G are defined as:
(4.22)
• The condition for the excitation coefficient of an to minimize
ε is then given by the leastsquares method as:
(4.23)
=
)(
)( 11
MdM
d
d
g
gG
θ
θ
=
=
NMN
N
MM a
a
A
A
A
A
g
gG
11
1
111
*
1 11∑ ∑∑
= ==
−
−=
M
m
N
nnmndm
N
nnmndm aAgaAgε
01
*
1* =
−=
∂∂ ∑ ∑
= =
N
nmn
N
nnmndm
n
AaAgaε
62
-
Antenna Pattern SynthesisSymmetrical array using Fourier series
expansion
∑∑∑==
−
−=
∆+=++=N
nno
N
n
jnn
N
n
jnno naaeaeaag
111)cos(2)( δδθ
oa 1a 2a 3a1a2a3aElement amplitude
( ) ( )( ) ( )
( ) ( ) 3322312
33
221
321
84622
cos3cos421cos22cos2
3cos22cos2cos2)(
xaxaxaaaaaaaa
aaaag
o
o
o
++−+−=
−+−++=
+++=
δδδδ
δδδθ7-element array
∆= cosx63/44
-
Binomial Radiation Pattern ∆= cosx( )
Nxg
+=
21θ
3=N ( ) ( )323
33181
21 xxxxg +++=
+=θ
323120 881,4
83,62
83,2
81 aaaaaa ==−=−=
641,
323,
41,
165
3210 ==== aaaa
7-element array
64/44
-
Binomial Radiation Pattern
0
0.25
0.5
0.75
1
-90 -60 -30 0 30 60 90
2λ
=d
No side lobe
65/44
-
4.2.3 Sector Antennas
• The sector zone allocation requires a fan-shaped pattern in
the horizontal plane. However, an ideal fan-shape cannot be
obtained from a real antenna and the span of a sector is specified
by the half power beam width. This real sector pattern may be
approximated by denoting the half power beam width as φb [14],
where:
-φe ≤ φ ≤ φe dB (4.23a)
-π ≤ φ≤ φe, φe ≤ φ ≤π dB (4.23b)
(4.23c)
• Current cellular systems use three-sector zones, where the
half power beam width is 90° or 120°. The beam width depends on the
system design.
• Although a narrow beam increases the user number, it sometime
causes blind regions at the cell edge. Therefore, the base station
location is important. Six sectors are also used because such an
arrangement doubles the capacity of the three-sector system
[15].
( ))2/cos(log10)( 10 φθ mg =
FB−=
( ) ,)2/cos(log103
10 b
mφ
−=
=
−− me
21 10cos2φ
66
-
Photo of Base Station Antenna
67
-
Horizontal Radiation Pattern
0
90
180
270
3020100
(dB)0
90
180
270
3020100
(dB)
: Ideal Radiation Pattern
68/44
-
Dual band antennaShort stub
Reflector
Feed network circuit
Dual Band Base Station Antenna
69
-
Reflector
Dual band antenna (900/1500 MHz)Parasitic element
(2GHz)
Shorted stub
Triple Band Antenna Element
70
-
Horizontal Radiation Pattern of Triple Band Antenna
900MHz 1500MHz 2000MHz
127° 113° 113°
71
-
Antenna Arrangement
72
-
4.2.4 Diversity Antennas• Up-link diversity reception is
mandatory for base station antennas in order to combat multi-
path fading.
• The diversity reception described in sec. 1.3.1 is also used
to compensate for power imbalance between uplink and downlink,
because the uplink signal is weaker than that of the downlink.
• For a system with space diversity, the same antenna is
installed with a spacing of more than 10 wavelengths (λ).
• The correlation coefficient for this space diversity, as shown
in Fig. 4.17, is calculated by assuming that the uplink signals
originate from inside the angle ψm with uniform distribution. The
correlation coefficient of ρe is evaluated from the definition of
(1.30) as
(4.25)
• The calculated correlation is shown in Fig. 4-18. For a design
value of ρe1.5° is obtained from propagation measurements in the
cities [16].
,sin2
=
XX
eρ αλπ sin2
Ψ=
dX m
73
-
ψm
α
0
0.2
0.4
0.6
0.8
1.0
0 5 10 15 20
Interval of the antennas d/
f =1.0H
o
f =1.5H
o
f =2.0H
o
l
Correlation coefficientr e
ψm=1.0°
ψm=1.5°
ψm=2.0°
d/λ
Fig. 4.17 Coordinate system for space diversity antenna Fig.
4.18 Correlation coefficient of space diversity antenna
74
-
• Mobile telephone services started as vehicular mounted mobile
terminals. Developments in electronics technology have resulted in
very small and lightweight portable handsets.
• As a result, everyone now uses a handset placed near the ear
at a tilted angle. As derived from statistical data [17], the tilt
angle is about 60°, and this decreases the vertical radiation
component by 6 dB, and the horizontal component by 5 dB more than
the vertical component.
• This operating condition results in an increase in the
horizontal electric field component in the uplink, and polarization
diversity is therefore desired for base station designs.
• Vaughan measured a polarization diversity scheme in the 400
MHz band using a tilted dipole as a mobile terminal [18], and
showed the correlation coefficient between horizontal and vertical
components to be almost zero in both the city and the suburbs.
• Propagation experiments by Turkmani at 1.8 GHz resulted in a
diversity gain of 1 to 2 dB for vertically-oriented dipole antennas
at the mobile terminal and 3 to 5.2 dB for 45° tilted antennas
[19].
• The measured correlation coefficient quoted by Lotse at 1.8
GHz was less than 0.2 [20], and the diversity gain depended the
average received signal level.
• The polarization diversity gain measurement for the case with
a human operator using a handset was 7 dB greater than the space
diversity in the line of sight region, and 1 dB greater non line of
sight region [21]. 75
-
• The polarization diversity system using vertical and
horizontal components has the merit of minimizing the antenna
installation space, and has been adopted for Japanese cellular
systems. An example of this diversity antenna is shown in Fig.
4.19.
• Another polarization diversity system uses slant polarization
of ±45°. The correlation coefficient is about 0.5 and smaller than
the diversity in the above, but the received signal levels for both
diversity ports are equal, which gives the same diversity gain as
with vertical and horizontal diversity systems. This slant
diversity system is widely used in cellular systems.
Fig. 4.19 Vertical and horizontal diversity base station
antenna; (a) horizontal polarization patch antenna, (b) vertical
polarization patch antenna, and (c) phase shifter.
76
-
4.3 Base Station Antenna Measurements• Base station antenna
measurements are classified into low power tests and high power
tests.
• In the low power tests, the antenna radiation pattern and the
input impedance characteristics are measured. A similar procedure
as that used for handset terminal antenna can also be used to
measure the input impedance of the base station antenna, and this
is presented in Chapter 2.
• For base station antenna pattern measurement, it is necessary
for perform measurements in an open site or near field measurement
is needed when the antenna has a high gain.
• The beam width and the position of the null point should be
verified for the designed coverage area of the cell, then, this
section presents the measurement method for high gain antennas that
cannot be measured in an anechoic chamber.
• In any case, the passive inter modulation (PIM) causes serious
problems especially when the base station antenna is operated at
high power levels.
• This section describes the definition of PIM and its
measurement method. Finally, the section also presents the
inspection process of the base station antenna as a product.
77
-
4.3.1 Antenna Pattern Measurements in Open Sites
• The distance R between the transmit and receive antennas
should satisfy the condition R ≥2D2/λ, defined by (2.11), to
measure the far field radiation, as presented in Chapter 2, where
the maximum dimension of the antenna is denoted as D.
• For a cell radius under 3km, the vertical radiation pattern of
the base station antenna can be controlled by electronically
adjusting the phase of the array elements. Since the dimension in
the vertical direction becomes more than 10λ for the pipe-shaped
form, and the measurement distance R requires more than 200λ as a
far field condition, in the 900 MHz band, a distance of more than
67 m is necessary, which is difficult to obtain in an indoor
facility.
• Measurement in an open site is therefore necessary, and the
transmit antenna must be installed on a steel tower or a building,
as shown in the Fig. 4.20, in order to satisfy the above far field
condition.
• For radiation pattern measurement, it is necessary to prevent
interference from the radio link that is actually in-use and to
suppress reflections from surrounding buildings.
78
-
• In order to avoid receiving waves reflected from the ground,
the antenna aperture size dshould satisfy the following equation,
where hr is the height of the receive antenna (the antenna used for
the measurements) and R is the distance between the transmit and
the receive antennas:
(4.26)• The antenna aperture size d is defined as the largest
dimension of the test antenna.
• If the aperture size of the transmit antenna is much greater
than the wavelength, and the antenna aperture is assumed to have a
uniform current distribution, the half power beam width 2θh of the
transmitting antenna may be approximated by the equation shown
below.
• This equation was derived so that the edge of the half power
beam points to the tower base of the receiving antenna:
(4.27)
• When the required distance is not available for the
measurements, a diffraction fence may be installed on the central
ground between the transmit and the receive antennas to avoid the
influence of the waves reflected from the ground.
• The diffraction fence should be mounted so as to not interrupt
the main beam between the transmit and receive antennas, and it
should be a saw-shape fence in order to give knife edge
diffraction.
rhRd
2886.0
21 λ
×≥
dhλθ 886.02 ≅
79
-
d
hr
Test antenna
R
d
hr
Test antenna
R
ht
Direct path
Reflection path
Fig. 4.21 Far-field measurement using ground reflection
Fig. 4.20 Far-field measurement using a tall tower
80
-
• A measurement method which uses the reflection from the ground
actively is shown in Fig. 4.21. The antenna under test is installed
on the tower, and the height of receive antenna near the ground is
adjusted so that the maximum electric field strength is
received.
• In this method, it is assumed that the reflection coefficient
at the ground has a value close to –1 when both the transmit and
the receive antennas are located just above the ground.
• Measurement can be made in similar conditions by adjusting the
height for an arbitrary frequency. Based on the condition that the
phase difference in the image antenna becomes (λ/2), the height hr
of the antenna under test may be determined as:
(4.26)
• For the slant range measurement, the antenna under test is
installed on the top of the high steel tower, as shown in Fig.
4.22, and the measurement antenna on the ground is pointed towards
it with a large elevation angle.
• The reflection waves from the ground are radiated from the
direction of the side lobes, and this is an effective method to
determine the direct path component in the measurement.
• However, the waves reflected from the surrounding buildings
need to be avoided.
tr h
Rh4λ
=
81
-
R
Test antenna
Fig. 4.22 Far-field measurement using slat range
82
-
4.3.2 Near-Field Antenna Pattern Measurement• The problems
associated with open site measurements include noise from the
surroundings of
the site, the influence of weather, and the requirement for a
license to transmit electromagnetic waves at high power levels. A
technique for measuring high gain antennas in a stable condition is
the near field measurement described in this section.
4.3.2.1 Vector Kirchhoff Integral and Wave Spectrum
Expansion
• If a virtual boundary S encloses a radiating test antenna, the
equivalent electric and magnetic currents J and M are given by the
tangential magnetic and electric fields on the surface S according
to the expressions:
(4.29)• The vector Kirchhoff integral given below gives the
antenna far field, using the measured
tangential magnetic and electric field, as:(4.30)
• where Zo is the characteristic impedance in free space. The
procedure requires calibrated, ideal probes for the measurement of
both the tangential electric and magnetic fields over the surface
S. However, the computation time required to perform the numerical
integral in (4.30) is very large [22].
tHnJ ×= nEM ×= t
( )∫ ⋅−
×+×=S
jko
jkr
dSeZr
jke '4
)( r'rJrMrrΕπ
83
-
• A different approach which reduces the computation time is
given by wave vector expansion and the Fast Fourier Transform
(FFT). The electric field radiated by the antenna is expressed as a
summation of plane waves expanded in the wave spectrum domain,
as:
(4.31)
• where r and k denote the observation point vector and the wave
vector, respectively, and are expressed by the unit vectors, ex,
ey, ez in the x, y, z coordinate system as:
(4.32)• The wave number kz given below, which does not diverge
as z→∞, is:
(4.33)• where k is the wave number in free space.
• If the measurement region in the xy plane is denoted by A, and
the tangential (x, y) component of the wave vector source and the
vector in the region A are kt and rt, as shown in Fig. 4.23, the
following equations may be derived:
(4.34)
(4.35)• where the subscript T denotes the tangential (x, y)
components. Equation (4.35) indicates that
the function of fT(kx, ky) may be evaluated by measuring the
tangential electric field components in region A.
yxyx dkdkjkkrE ∫ ∫+∞
∞−
⋅−= )exp(),(21)( rkfπ
,zyx eeer zyx ++= zyx eeek zyx kkk ++=
222 kkkjk yxz −+−=
∫ ⋅−=A
yxyx dkdkjkkyx )exp(),(21),( TTTA rkfE π
∫ ⋅=A
yx dxdxyjyxkk )exp(),(21),( TTAT rkEf π
84
-
• After evaluating fT(kx, ky), the antenna far field may be
calculated using (4-31), assuming the observation vector r to be
sufficiently large. After variable transformation, (4-31) can be
rewritten as:
(4.36)
(4.37)
(4.38)
• where the outgoing wave, kz ≥ 0, is only considered in the
variable transformation. When kr islarge enough for the antenna far
field region, (4.36) may be approximated using the saddle-point
method [23], which gives:
(4.39)
∫ ∫−
=2/
0
2
0
)exp(),(cossin21)(
π π
βαααπ
ddr
jkrkk yxfrE
θφθφθ cos,sinsin,cossin rzryrx ===
,,
αβαβα cos,sinsin,cossin kkkkkk zyx ===
rjkrkkjk yx
)exp()sinsin,cossin(cos)( −= φθφθθ Tf frE
Fig. 4.23 Coordinate system for region A, and wave vector 85
-
4.3.2.2 Near-Field Measurement• In an example of large aperture
size, such as a parabolic reflector antenna, the
electromagnetic
fields are defined differently, by distance in the vicinity, in
the near field region and in the far field region, respectively, as
shown in Fig. 4-24.
• Accordingly, the equivalent theory for the electromagnetic
field in each region gives the radiation field. For field
measurements in the vicinity of the antenna, the mutual coupling
between the measurement probe and the antenna under test becomes
very high and disturbs the electromagnetic field radiated from the
antenna.
• Therefore, measurements in the vicinity of the antenna are
very difficult. Although it is possible to exclude this mutual
coupling by probe correction techniques, measurements in this
region are not generally performed because of the instability of
the measured field.
• Next to the region in the vicinity of the antenna is the near
field. A suitable probe for near field measurement consists of an
open-ended rectangular waveguide, which does not have an ideal
omni-directional pattern.
• The non-ideal performance can be rectified easily by probe
correction [22], and the far field can then be evaluated from the
electromagnetic field in the near field region.
86
-
• Although near field measurements require a space enclosing the
antenna, a closed edge plane is not always necessary when the
measured field strength is small enough to neglect.
• The probe’s scanning surface for near field measurements can
be planar, cylindrical, or spherical, as shown in Fig. 4.25.
• Planar surface scanning covers only half the space during the
measurements, and is appropriate for high gain antenna measurements
such as parabolic antennas.
• Cellular base station antennas and marine radar antennas have
fan shaped patterns with a narrow beam in one plane and a wide beam
in the other, and cylindrical scanning is appropriate for near
field measurements on these antennas.
• Spherical scanning has the advantage that low gain antennas
can be measured by scanning enclosed surfaces.
• The scanning area (A×B) and the sampling interval (∆x, ∆y) are
discussed in the planar scanning method and are shown in Fig. 4.26.
The size of the scanning area is determined by the requirement to
obtain more than 35 dB dynamic range in the scanning area, without
the need to interpolate the measured data.
• In this case, the dynamic range is defined by the difference
between the maximum electric field strength and the edge field
strength. The maximum data point spacing should be such that ∆x =
λo/2 and ∆y = λo/2 for planar scanning, using the sampling theorem
[22]. 87
-
R
Antenna
Vicinityregion
0 3λ
Near fieldregion
2a /λ2
Far fieldregion
Fig. 4.24 Electromagnetic fields in vicinity, near-field and
far-field regions
3λ 2a2/λ
Fig. 4.25 Probe scanning surface for near field measurement Fig.
4.26 Scanning area size and sampling interval for planar scanning
88
-
4.3.2.3 Far Field Evaluation in Near Field Measurements
• If the x and y components of the measured electric field are
Ex(x,y) and Ey(x,y), respectively, the far field radiation patterns
are given by eqs. (4.34) and (4.35) and the coordinate
transformation from (x, y, z) to (r, θ, φ) as:
(4.40)
(4.41)
,i=x,y (4.42)
• The total number of observed discretized electric fields are
2M and 2N in the x and ydirections, respectively, and Si may then
be expressed as:
xm=m∆x, yn=n∆y (4.43)
• In the above calculations, the Fast Fourier Transform (FFT)
can be used to decrease thecomputation time. An outline of FFT is
given in the Appendix.
)sincoscoscos()exp(cos φθφθθθ yx SSrjkrjkE +−=
)cossin()exp(cos φφθφ yx SSrjkrjkE −−−=
( ){ }∫ ∫− −
+=2/
2/
2/
2/
sinsincossinexp),(21 B
B
A
AoooooooAii dydxyxjkyxES φθφθπ
( ){ }∑∑− −
∆∆+=M
M
N
NnommAii yxyxjkyxES φθφθπ
sinsincossinexp),(21
89
-
4.3.3 High-Power and Passive Inter Modulation Tests
4.3.3.1 Definition of Passive Inter Modulation
• Passive inter modulation (PIM) is a technique in which a given
frequency product is generated by combining several different
frequencies which are inputs to a circuit with non-linear
input-output characteristics.
• This non-linear phenomenon is observed at the junction between
a connector and a transmission line, where there is a contact
between different metals.
• The base station antenna is used for both transmission and
reception and its transmit power level is several watts per
channel.
• The uplink received electric field strength is between 10 and
60 dBµV, which represents a difference from the transmit power of
around 100 dB.
• The received field strength is equivalent to –103 to –53 dBm
in the 50 Ω transmission line system. This power imbalance becomes
more than 140 dB with multi-channel input to base station antennas,
and causes the PIM in the passive device.
90
-
Passive Inter Modulation (PIM)
power imbalance > 140 dB
Non-linear circuitf1 f1
f2 f2f3
at the junction between a connector and a transmission line
(different metals)
10 and 60 dBµV(–103 to –53 dBm )
several watts per channel
BS
Junction at connector
Small gap
91
-
PIM at Base Station
BS
ft1fr1
Duplex system
Diversity reception
f
ft1
uplink
ft2fr1
downlink
PIMfr1’Real uplink signal
92
-
• Most current cellular systems are duplex, with the transmit
and receive frequency bands allocated at a certain interval.
• A further receive antenna is necessary for the base station,
in addition to the antennas for transmission and reception, because
of the uplink diversity system. A duplex cellular system uses the
frequency bands for transmission and reception simultaneously.
• When the PIM occurs at the antenna for both transmission and
reception, it appears on the receive frequency band and the receive
channel cannot then be used, due to the interference at the base
station.
• Although the antenna is a passive device, very weak electric
discharge phenomena occur at the junctions of the antenna parts
during high power operation. Such discharges often appear at the
surface of the junctions that are made of different metals.
• This phenomenon causes the antenna to have a non-linear
input-output response that may be regarded as a two-port circuit
with transmitter input and free space output.
• The output signal Y is given as the weighted summation of the
input signals X by the expression:
(4.44)∑∞
=
=++++=0
33
2210
k
kk XaXaXaXaaY
93
-
• If the two input signals have different angular frequencies,
ω1=2πf1 and ω2=2πf2:(4.45)
• where A1 and A2 are the respective amplitudes of the two
signals and the phase angle f of X2 is normalized with respect to
that of X1.
• Substituting (4.45) into (4.44), the components of the
non-linear output signals may be obtained. The angular frequency
components appearing at the output port may be expressed as follow,
for terms up to Xi3, for example:
1. ω1 and ω2 from the term for X1
2. 2ω1, 2ω2, and from the term for X2
3. ω1, ω2, 3ω1, 3ω2 and from the term for X3
• Taking the M-signal input for X, each power series from the
output of the circuit with the input-output response of (4.44) may
be expressed as (4.46) by denoting each input signal as
. Then:(4.46)
• where ni is an integer given by the expression , in which the
sum of all the combinations of ni is included. Then, the angular
frequency ωk of the PIM signal corresponding to this non-linear
response is given as:
(4.47)
)cos(,cos 222111 φωω +== tAXtAX
tAX iii ωcos=
∑∑ =
=
= Mn
nMM
nnkM
m
k XXXnMn
kXX
,1
22
11
1 !!1!
Mnnnk +++= 21
kkk nnn ωωωω ±±= 221194
-
• The parameter k is the order of PIM. The PIM signals for odd
orders higher than the 3rd causeserious problems in current
cellular systems. The interference condition for the PIM may
berewritten by denoting p as a natural number as:
(4.48)• The order of PIM is 2p±1, and these frequencies cause
the most serious problems to down
link frequency. Might be worthwhile to mention that the odd
order PIM frequency is close to the input frequencies
• Fig. 4.27 shows the allocation of uplink and downlink
frequency bands for cellular systems with upper and lower limit
edge frequencies of f1 and f2, respectively.
• The frequency interval should be allocated so that it is not
affected by the third and fifth order PIM, as shown in Fig. 4.27.
However, the PIM should take into account of ideal frequency
allocations.
• PIM signals are expressed as an absolute value in dBm and a
relative value in dBc, normalized by the input signal. For example,
for a PIM level of –100 dBm the relative PIM value is -143 dBc,
normalized by two input signals of 43 dBm.
( ) ( )211122 , ωωωωωωωω −+=−+= pp PIMPIM
95
-
Fig. 4.28 Block diagram of setup for antenna PIM measurement
SG1 AMP
SG2 AMP
Diplxer
DUT
BPF
DL
SA
SG = signal generator, AMP = amplifier, DUT = device under test,
DL = dummy load ,BPF = band-path filter, SA = spectrum analyzer
ft1
ft2
ft1 ft2
fr1’
DL for connector & cable
Open air for antenna
96/44
-
f
Up linkDown link
3rd PIM
f1 f22f1 -f2 2f2 -f1 3f2 -2f1 4f2 -3f13f1 -2f24f1 -3f2
5rd PIM
7rd PIM
Fig. 4.27 Frequency allocation for higher order
The PIM signals for odd orders higher than the 3rd cause serious
problems in current cellular systems. The interference condition
for the PIM may be rewritten by denoting p as a natural number
as:
( )122 ωωωω −+= pPIM ( ) 121 ωωωω +−= pPIMThe order of PIM is 12
±p
PIM in dBmrelative value in dBc,
PIM level of –100 dBmrelative PIM value is -143 dBc, (two input
signals of 43 dBm)
97/44
-
4.3.3.2 PIM Measurement Setup• Fig. 4.28 shows the block diagram
of the PIM measurement setup. The measurement system
consists of three units connected to the three ports of the
duplexer, which operates as a transmission port, reception port and
common port for both transmission and reception.
• After amplification to the required power level, the two
continuous signals P1 and P2, at different frequencies, are input
to the transmit port of the duplexer through the power combiner.
The device under test (DUT) is connected to the common port for
both transmission and reception.
• When the DUT is used as a connector and a transmission cable,
the power is dissipated in a dummy load. When the DUT is an
antenna, the antenna under test is set up inside an anechoic
chamber.
• With this measurement setup, the PIM signal detected at the
reception port is measured by a spectrum analyzer. If the PIM
signal level is less than the noise level of the spectrum analyzer,
a low noise amplifier (LNA) is inserted to adjust the receiving
level.
• By inserting a band-pass filter prior to the LNA, the direct
coupling from transmitting port to receiving port is eliminated.
The band-pass filter suppresses the undesired PIM occurring in the
LNA and the spectrum analyzer.
98
-
• When the DUT is a connector and a transmission cable, a
terminating load is used, as shown in Fig. 4.29. When a resistor is
used as a dummy load this also tends to have PIM within it.
• Therefore, a long coaxial cable, similar to a semi-rigid
cable, is often used instead of a resistor. The coaxial cable of
length between 50 and 100 m is usually long enough to be used as a
dummy load.
• This long cable then dissipates most of the power. After
confirming that there is no PIM excitation at the interconnection
between the duplexer and DUT, the PIM may be measured using the
above setup.
• During PIM measurements on antennas, metallic objects should
not be located near the antenna under test in order to avoid the
occurrence of PIM by induced current excitation.
• The PIM for antennas should preferably be measured inside an
anechoic chamber, although measurements can also be made in open
sites. In open site measurements, the main beam of the antenna is
directed towards the sky.
• PIM phenomena may occur at several points, but the PIM signals
may be difficult to observe, due to phase cancellation by each
other. To avoid this phase cancellation, the PIM should be measured
at several different frequencies.
99
-
SG1 AMP
SG2 AMP
Diplxer
DUT
BPF
SA
Fig. 4.29 Block diagram of setup for connector and cable PIM
measurement: SG, signal generator; AMP, amplifier; DUT, device
under test; BPF, band-path filter; SA, spectrum analyzer; DL, dummy
load.
100
-
4.3.3.3 PIM Tests for Products• PIM measurement procedures for
commercial base station antennas are described in the
following.
• Using the measurement setup described in section 4.3.3.2, a
plastic hammer is used to knock each part of the antenna under
test.
• Three kinds of measured PIM phenomena may be observed, these
being high level PIM occurring continuously, high level PIM which
occurs only during knocking, variable level PIM which occurs as a
step function due to knocking.
• It should be noted that the measured PIM disappears after
knocking. Using this knocking test, the problem parts where the PIM
occurs can be found. It should be noted with caution, however, that
the PIM phenomena sometimes reappear after a period of time.
101
プレゼンタープレゼンテーションのノート木槌 wooden hammer, wooden mallet
-
4.4 Product Inspection• This section presents production
inspection tests which should be undertaken before base
station antennas are shipped. Similar tests are presented in
sec. 2-6 for mobile handset terminals.
• It is necessary for the products to satisfy the
specifications. In addition to the electrical performance, the
input characteristics and the radiation pattern, mechanical and
reliability tests (for example, vibration and moisture-proof tests)
are also required, since the products are going to be used a real
environment.
• The antennas discussed in this section are for outdoor base
stations as well as for small indoor base stations. Outdoor
antennas are usually used for cellular coverage, and indoor
antennas for in-building mobile communication systems or for micro
cellular systems.
• The tests presented here are not carried out on all base
station antennas and in most cases, the operator and the
manufacturer decide the product inspection procedures to be
adopted, depending on the environment in which the antennas are
intended to be used.
• This section presents example specification data, but their
samples are mainly based on Japanese standards. The environment in
Japan is very harsh for base station antennas, because there are
earthquakes, typhoons, and heavy snow falls in metropolitan areas.
Inspection tests should therefore be chosen appropriately to match
the environment in each country.
102
-
4.4.1 Radiation pattern4.4.1.1 Directivity Gain• The absolute
gain (dBi) or the relative gain (dBd) of an antenna in the
direction of maximum
radiation is measured as the antenna gain. The measurement
procedure for antenna gain is described in Section 2.2.3.1.
4.4.1.2 Radiation Pattern• The specifications for the radiation
pattern for base station antennas, defined on the horizontal
and vertical planes or on the E plane and H plane, are: cross
polarization ratio (XPR), as defined in sec. 1.2.5, and half power
beam width.
• The XPR is required to be more than 20 or 25 dB on the
specified radiation plane. The beam width is defined by the
deviation due to the frequency dependence and varying quality of
the product. For example, the edge level of half power beam width
for 60° is 3±1 dB, or 3±0.5 dB at φ=±30°. The difference from the
definition written in textbooks is due to the inclusion of the
above deviation [24].
• In addition, the polarization in the measurement is based on
the slant polarization diversity antenna being measured by means of
a ±45° tilted antenna to coincide with the polarization for both
transmission and reception. For a space diversity antenna system
using only the vertical polarization, the antenna pattern
characteristics are tested only for vertical polarization. 103
-
4.4.1.3 Beam Tilt Angle• To adjust the coverage area of the base
station, the antenna main beam is tilted an angle θt
below the horizon. This angle is then defined as the tilt
angle.
4.4.1.4 Front-to-Back Ratio• In sector zone cellular systems,
the front to back (FB) ratio is defined as the ratio between
the
forward main beam level and the backward level. The definition
of FB ratio is the forward radiation level at φ=0° divided by the
backward level at φ=180°.
• Another definition is obtained by taking the backward
radiation as the maximum radiation level in the region of 120° ≤ φ
≤ 240° or 90° ≤ φ≤ 270°, where φ=0° corresponds to the forward
direction.
• In the ±45° tilted polarization antenna, the backward
polarization is orthogonal to the front one. It should be carefully
checked in the FB ratio measurement.
104
-
4.4.2 Antenna Input-Port Characteristics4.4.2.1 Voltage Standing
Wave Ratio• The voltage standing wave ratio (VSWR), defined in
section 2.1.1, at the input port of base
station antennas is required to be less than 2.0 in the
specified frequency band.
4.4.2.2 Power-Resistance Capacity• For the power-resistance
test, the input characteristics of the antenna should not vary for
a
specified input power. An example of input power might be 30 W
per input port.
• The power-resistance test is carried out at the center
frequency of the specified frequency band. The procedures used in
the power test should not cause any heating problems and the VSWR
at the input port should not vary at high power input.
4.4.2.3 Insulation Resistance• Typically, an insulation
resistance of more than 100 MΩ is required at the antenna input
port,
with the resistance measured by connecting the probes of an
insulation resistance tester between the center conductor and the
outer conductor at the input connector [25]. The maximum voltage of
the tester in this example is 500 V.
105
-
4.4.2.4 High Voltage-Resistance Performance• After applying 1000
V AC for 60 seconds between the center conductor and the outer
conductor at the antenna input port, it should be verified that
the antenna characteristics, such as VSWR, have not changed. The
test procedure is the same as for the power-resistance test.
• The tests for both the insulation resistance and power
resistance are not applied to the antenna with a direct
short-circuit, as it would be the case for a log-periodic
antenna.
4.4.2.5 Conductivity Performance• As a conductivity performance
test, the AC resistance between the inner and outer conductors
at the antenna input port should be less than 100 mΩ.
4.4.2.6 Other Performance Tests• In the case of multi-port base
station antennas, such as polarization diversity antennas and
multi-band antennas, the mutual coupling between the antenna
ports is defined as the isolation characteristics, as described in
section 2.1.2.
• The isolation should be more than 30 dB, or 25 dB in the range
of specified frequency band at the two input/output ports of a
polarization diversity antenna.
106
-
4.4.3 Feed Circuit• Recent base station antennas adjust their
coverage areas by changing their tilt angle in the
vertical plane, and both mechanical and electrical methods can
be used for beam tilting.
• In the electrical method, the base station antenna consists of
several sub-array units, and the phase of each is electronically
changed for beam tilting. Therefore, the antenna system has two
components, the antenna itself and a feed circuit for phase
adjustment.
• Three types of phase adjusting system may be used for the feed
circuit: remote control, manual control and fixed advance. Digital
remote control is mostly based on specifications by Antenna
Interface Specification Group (AISG) [26].
• The performances of all the above control schemes are similar
and one or other will be selected, depending on the cost of the
base station(s). The phase control unit is located separately from
the antenna unit, as shown in Fig. 4.30.
• This feed circuit is tested independently of the antenna unit.
The measurements made are the same as for antenna measurements, and
include VSWR, power-resistance capability, insulation resistance,
high voltage-resistance performance, conductivity performance and
mutual coupling between input ports.
• These tests are not conducted on antennas operated at a fixed
beam tilt angle, and those in which phase differences are provided
to each antenna element. 107
-
Phasecontrol unit
Sub array unit
Radome
Feeding cable
Antena unit
Fig. 4.30 Phase control unit and base station
108
-
4.4.4 Appearance and Weight • The dimensions and total weight of
each antenna part should be within specification. The
nameplate, logo and painting color should also be confirmed.
4.4.4.1 Antenna Construction• The antenna construction, as shown
in the specifications, must be verified.
4.4.4-2 Assembly Instructions• The assembly instructions are
required to be sufficiently comprehensive that mistakes will
not
be made during installation.
4.4.4.3 Range of Mechanical Adjustment• If the beam tilt
adjustment is performed mechanically, the range of adjustment
available must
be verified. In addition, the adjustment range of the movable
polarization plane should also be checked, if applicable.
109
-
4.4.4.4 Wind pressure resistance• Depending on the wind pressure
on the antenna, it should be confirmed that the radome and
the other antenna parts have sufficient mechanical strength to
withstand a wind velocity v up to 60m/s. This check is usually
carried out by numerical calculation.
• If the wind pressure is q (kgf/m2) and the wind pressure
coefficient is C, the wind pressure loading P (kgf/m2) is:
(4.49)• where
(4.50)• and h is the antenna height. The wind pressure
coefficient C for antenna pillars with square
and circular cross sections has values of 2 and 0.8,
respectively.
• Therefore, a circular column is preferable from the viewpoint
of wind pressure. In a typical example, the wind pressure loading
may be larger than 350 kgf/m2.
• In addition, the allowed rolling vibration range is within ±2
degrees for a wind velocity of 60 m/s. Furthermore, no noise should
be generated by the wind.
4.3.4.5 Other Tests• If a phase control unit is added to the
antenna system, its tolerance should also be checked.
CqP =
)16(60),16(120 44 mhhqmhhq ≤=>=
110
-
4.4.5 Reliability Examination• The reliability of antennas is
examined according to the standard regulations of the
International Electrotechnical Commission (IEC) and typical
tests involved are described in this section.
4.4.5.1 Thermal Test (Heat Cycle Test)• The antenna
characteristics are examined in a heat cycle test under conditions
in which the
humidity is more than 90 % and the temperature range is from
-30° C to +60° C.
• The actual heat cycle is as follows: After the temperature has
been raised from a room temperature of +20° C, the antenna is held
at the highest temperature of +60° C for several hours.
• The temperature is then dropped to the lowest level of -30° C
and is held at this level for several more hours. It is then raised
again to the highest temperature.
• This procedure represents one cycle of the thermal test.
During a complete test program, the input VSWR is checked after
tens of test cycles.
111
-
• For small indoor base station antennas that can be tested
inside a compact box, the variations of temperature with time
during a heat cycle test are as described below [27].
• The temperature is raised from the room level of +20° C to
+50° C in one and a half hours, and this high temperature level is
held for three hours.
• The temperature is then dropped from +50° C to -20° C in three
hours, and the temperature held at -20° C for three hours. Finally,
the temperature is raised again to +20° C in one and a half hours.
This variation corresponds to one heat cycle.
• When automatic temperature control is not available, the
following is used as one heat cycle: The temperature is held at
+50° C for four hours after raising it from room level, then a
temperature of -20° C is maintained for four hours.
• Finally, the antenna under test is left at room temperature
for a further two hours. The antenna characteristics, such as the
input VSWR, are evaluated several times during the heat cycle.
112
-
4.4.5.2 Heat(Cold) Resistance Performance• Vehicular mounted
antennas are tested for a few hours in a heat-resistance test at
90° C and a
cold–resistance test at -40° C. As a heat shock test, the
heat/cold-resistance test covers a period of three hours and is
repeated several times. However, this test is not often
performed.
4.4.5.3 Waterproofing and Moisture-Proof Performance• The
waterproofing test is carried out on the product before shipping.
The moisture-proof test
is applicable to both the parts and the materials of the
antennas.
• An example of a waterproofing test is as follows: In the same
installation as intended for real use, water is sprinkled all over
the antenna using a nozzle mounted above the antenna under test at
a height of from 300 to 500 mm.
• As an example of the sprinkling rate, the quantity of water
may be 10 ± 0.5l per minute, with a pressure of between 50 and 150
kPa for more than one hour. On completion of this waterproofing
test, it is confirmed that the insulation resistance at the antenna
input port has not changed.
• A sealed type antenna used for small base stations may be
immersed into a water tank at a depth of more than 400 mm for 24
hours, and the penetration of water is then confirmed by breaking
up the antenna.
113
-
• The moisture-proof performance of small base station antennas
can be tested using a box at constant temperature and humidity.
• As an example of a typical test cycle, the antenna is left
inside the box at a temperature of 40° C and a humidity of 80% for
96 hours, it is then placed in an area at room temperature and
humidity for an hour. After this cycle, the antenna input VSWR
characteristics should not have changed.
• It should be noted that these waterproof and moisture-proof
tests are applied only to a sample of the products [28]. The tests
carried out before shipping include VSWR measurement and appearance
checking.
4.4.5.4 Vibration Test• For antennas used in areas subject to
vibration, such as on highways, a vibration test is
required.
• Before the vibration test, the mechanical resonance frequency
is measured by vibrating the antenna at variable frequency in both
the vertical and horizontal directions, with an accelerometer
attached to each part of the antenna.
• An example measurement setup is shown in Fig. 4-31.
114
-
• The antenna is then vibrated at the measured resonant
frequency in the x, y, and z directions. In a typical test example,
a force of 5G is applied for an hour in the z direction and for
half an hour in the x and y directions. After the vibration test,
it must be confirmed that the VSWR has not changed.
(a) Probe attachment for vibration test (b) Vibration test
setup
Fig 4-31 Vibration test
115
-
References1. “Propagation data and prediction methods for the
terrestrial land mobile service using frequency range 30 MHz to 3
GHz,” CCIR SG-5, Report
567-3 (MOD F).2. Kozono S. and Watanabe K., “Influence of
environmental buildings on UHF band mobile radio propagation,” IEEE
Trans. Commun., COM-25,
no. 10, pp. 1133-1143, Oct. 1977.3. Hata M., “Empirical formula
for propagation loss in land mobile radio service,” IEEE Trans.
Veh. Technol., VT-29, no. 3, pp.317-325, 1980.4. Ikegami, F., et.
al., “Propagation factors controlling mean field strength on urban
streets,” IEEE Trans. Antennas Propagt., vol. 22, no. 8, pp.
822-829, Aug., 1984.5. Walfisch, J., et. al., “A theoretical
model of UHF propagation in urban environment,” IEEE Trans.
Antennas Propagt., vol. 36, pp. 1788-1796,
Dec. 1988.6. European Cooperation in the Field of Scientific and
Technical Research EURO-COST 231, “Urban transmission loss models
for mobile radio in
the 900 and 1800MHz band,” Revision 2, The Hague, Netherlands,
1991.7. ITU-R, Recommendation P. 1411-4, “Propagation data and
prediction methods for the planning of short-range outdoor radio
communication
systems and radio local area networks in the frequency range
300MHz to 100GHz,” 2007.8. Green E. and Hata, M. “Micro cellular
propagation measurements in an urban environment,” 1991 IEEE
International Sym. On Indoor and
Mobile Radio Communications, King’s College London, pp. 23-25,
Sept. 1991.9. Blackar, K. L. et. al., “Path loss and delay spread
models as functions of antenna height for micro cellualr system
design,” Proc. 1992 IEEE Veh.
Technol., Conf., Denver, CO, pp. 333-337, May 1992.10. Oda, Y.,
Tanaka T., and Sato K. "Microwave Band LOS Path Loss
Characteristics in Microcellular Mobile Communications,"
International
Symposium on Antennas and Propagation, ISAP'96, Chiba, Japan,
pp. 1097-1100, Sep. 199611. Oda Y. and Tsunekawa K., "Advanced LOS
Path Loss Model in Microwave Mobile Communications," ICAP'97,
pp.2.170-2.173, April 199712. Fujimori K. and Arai H. "Indoor
Propagation Characteristic Including Radiation Pattern and
Polarization of Base Station" Conf. Dig., 1997
IEEE AP-S, Int, Symp, Montreal, Canada, pp.2006-2009, Jul.,
1997.13. Taylor T. T., Design of line-source antennas for narrow
beamwidth and low sidelobe,” IRE Trans. AP. Vol. AP-3, pp. 16-28,
Jan 1955.14. Ebine Y. "Dual frequency base station antennas for PDC
systems in Japan." IEEE AP-S, Int, Symp, And USNC/URSI National
Radio Science
Meeting, Orland, Florida, pp.564-567, July, 1999.15. Fujii T.
et. al., “Handling capacity expansion of land mobile communication
system,” NTT Review, vol. 2, no. 3, pp. 55-61, Mar. 1990.16. Yamada
Y., Ebine Y. and Tsunekawa K. , ”Base and mobile station antennas
in land mobile radio systems,” IEICE Japan, Trans., Vol, E74,
no.
10, pp. 3202-3209, June 1991.17. Taga T. and Tsunekawa K., "A
built-in antenna for 800 MHz band portable radio units," in Proc.
of ISAP’85, Kyoto, 121-1, pp.425-428, 1985.18. Vaughan R.,
"Polarization diversity in mobile communications", IEEE Trans. VT,
Vol. 39, No.3, pp.177-186, Aug. 199019. Turkmani A.M.D., et. al.
"An experimental evolution of the performance of two branch space
and polarization diversity schemes at 1800 MHz",
IEEE Trans. VT. Vol.44, No 2. pp.318-326, May 199520. Lotse F.,
Berg J-E., Forssen U., and Idahl P., "Base station polarization
diversity reception in macrocellular system at 1800 MHz", IEEE
VTS
46th Vehicular Technology Conference Proceeding, Vol.3, pp.
1643-1646, April 28-May 1, 1996.21. Arai H. and Nakano M., "Up-link
polarization diversity and antenna gain measurement of hand held
terminal at 900 MHz", MDMC’94, Niigata,
Japan, Nov. 21-22, 1994. 116
-
References (continued)22. Yaghjian, A.D., “An overview of
near-field antenna measurement,” IEEE Trans. AP, vol. AP-34, no. 1,
pp.30-45, 1986.23. Kong, J, “Electromagnetic wave theory,” Second
Edition, John Wiley & Sons, Inc. pp. 307-312, 1990.24. Kuga,
N., Tokyo Institute of Polytechnics, May 2000, personal
communication .25. Nakano, M., KDDI R&D LAB, June 2012,
personal communication .26. http://www.aisg.org.uk/27. Sugisawa,
H., NHK Spring Co., Ltd., April 2000, personal communication .28.
Matsuoka, T., Nihon Dengyo Kousaku Co., Ltd., April 2000, personal
communication .
117
http://www.aisg.org.uk/
4. Base Station Antenna Siting and Measurement4.1 Base Station
Antenna Siting4.1.1 Design of Macro Cell Siting Using the
Okumura-Hata CurveFig. 4.1 Okumura’s curve for city area in 900 MHz
bandスライド番号 5Fig. 4.2 Definition of antenna effective heightFig. 4.3
Propagation loss estimated �by Okumura-Hata methodスライド番号 8スライド番号
9Antenna System Design: Examples4.1.2 Break Point of a Micro
cellスライド番号 12スライド番号 13Fig. 4.5 Propagation loss and break
pointスライド番号 154.1.3 Propagation Loss Outside the line of
Sightスライド番号 174.1.4 Propagation Loss Estimation �by the Ray Tracing
Methodスライド番号 19Bi-Directional Notch AntennaTunnel Booster
AntennaTwo modes of Bi-directional NotchCurrent Distribution of Two
modesA Pair of Dipole Antenna for Bi-Directional Radiation
PatternRadiation Pattern of Two ModesEffect of Mounting Ground
PlaneCoverage Area of Tilted Beam AntennaDefinition of Coverage
AreaNotch Antenna with Parasitic ElementsNumber of Parasitic
ElementUndersea Tunnel Cellular Phone ServiceTokyo-Wan Aqua
LineHigh Gain Tunnel Booster AntennaHigh Gain Antenna for W-CDMALow
Cost Tunnel Booster SystemPropagation Measurement in
TunnelPropagation Loss Inside Tunnel2-D Simulation ModelConvergence
CharacteristicsImage MethodDefinition of Reflection Transverse
Electric (TE) IncidenceTransverse Magnetic (TM) Incidence3-D
Simulation Results Time Delay Profile4.2 Design of Cellular Base
Station Antennasスライド番号 47スライド番号 48Fig. 4.12 Antenna elements for
base station antenna4.2.2 Beam Tilting and Propagation
Characteristicsスライド番号 51スライド番号 52Beam Tilting Effect of Base
Station Antenna スライド番号 54Co-Phase Excitation for Tilt Beam Feeding
Network and Phase ShifterPhase Shifter ArrangementVariable Phase
Shifter by Sandwich Dielectric PlatesVariable phase shifter by
sandwich dielectric platesTheory of Phase ShifterPhoto of Variable
Phase Shifterスライド番号 62Antenna Pattern SynthesisBinomial Radiation
Pattern Binomial Radiation Pattern4.2.3 Sector AntennasPhoto of
Base Station AntennaHorizontal Radiation PatternDual Band Base
Station AntennaTriple Band Antenna ElementHorizontal Radiation
Pattern of Triple Band AntennaAntenna Arrangement4.2.4 Diversity
Antennasスライド番号 74スライド番号 75スライド番号 764.3 Base Station Antenna
Measurements4.3.1 Antenna Pattern Measurements in Open Sitesスライド番号
79スライド番号 80スライド番号 81スライド番号 824.3.2 Near-Field Antenna Pattern
Measurementスライド番号 84スライド番号 854.3.2.2 Near-Field Measurementスライド番号
87スライド番号 884.3.2.3 Far Field Evaluation in Near Field
Measurements4.3.3 High-Power and Passive Inter Modulation
TestsPassive Inter Modulation (PIM)PIM at Base Stationスライド番号
93スライド番号 94スライド番号 95Fig. 4.28 Block diagram of setup for antenna
PIM measurementFig. 4.27 Frequency allocation for higher
order4.3.3.2 PIM Measurement Setupスライド番号 99スライド番号 1004.3.3.3 PIM
Tests for Products4.4 Product Inspection4.4.1 Radiation
patternスライド番号 1044.4.2 Antenna Input-Port Characteristicsスライド番号
1064.4.3 Feed Circuitスライド番号 1084.4.4Appearance and Weight スライド番号
1104.4.5 Reliability Examinationスライド番号 112スライド番号 113スライド番号
114スライド番号 115スライド番号 116スライド番号 117
