A A NTENNA NTENNA S S YSTEM YSTEM D D ESIGN ESIGN SIGMA IGMA W WIRELESS IRELESS T TECHNOLOGIES ECHNOLOGIES L LTD TD MCKEE EE A AVENUE VENUE FINGLAS INGLAS DUBLIN UBLIN 11 11 I RELAND RELAND PHONE HONE: : INT INT 353 1 8142050 353 1 8142050 FAX AX: : INT INT 353 1 8142051 353 1 8142051 EMAIL MAIL : : [email protected]WWW WWW. SIGMA SIGMA. IE IE
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
AANTENNANTENNA S SYSTEMYSTEM D DESIGNESIGN
SSIGMAIGMA W WIRELESSIRELESS T TECHNOLOGIESECHNOLOGIES L LTDTD
2.4.1 Diversity Gain Explained................................................................................................................142.4.2 Experimental Results.......................................................................................................................152.4.3 Different Diversity Schemes Described...........................................................................................152.4.4 How does Space Diversity work?....................................................................................................162.4.5 How Does Polarisation Diversity Work?........................................................................................17
3 ANTENNA CHARACTERISTICS FOR OPTIMUM PERFORMANCE.............................................18
3.1 HIGH TRAFFIC DENSITY...............................................................................................................................193.2 MEDIUM / LOW TRAFFIC DENSITY.................................................................................................................20
4 GOOD TETRA ANTENNA SYSTEM DESIGN PRACTICE.................................................................20
4.1 RECEIVER ISOLATION FROM TRANSMITTERS.....................................................................................................204.2 OMNIDIRECTIONAL DIVERSITY APPLICATIONS..................................................................................................21
4.2.1 Introduction.....................................................................................................................................214.2.2 Sample Power Balance Calculation................................................................................................214.2.3 ‘Top of Mast’ Omni plus Two Offsets ............................................................................................224.2.4 ‘Top of Mast’ Omni plus Three Panels ..........................................................................................234.2.5 Side Mount ‘Omnidirectional’ Diversity Array...............................................................................264.2.6 Two Sector Hybrid Sector System...................................................................................................27
5 MOUNTING CRITERIA FOR OPTIMUM PERFORMANCE.............................................................28
5.1 ELECTRICAL................................................................................................................................................285.1.1 Background.....................................................................................................................................285.1.2 Dolphin Measurement.....................................................................................................................295.1.3 Practical measurement....................................................................................................................295.1.4 Summary of results..........................................................................................................................305.1.5 Analysis of results............................................................................................................................315.1.6 Test limitations................................................................................................................................325.1.7 Conclusions.....................................................................................................................................32
6 ANTENNA / SYSTEM INTEGRATION...................................................................................................34
6.1 LOW DENSITY SYSTEM.................................................................................................................................346.2 MEDIUM DENSITY.......................................................................................................................................356.3 HIGH DENSITY............................................................................................................................................36
Sigma Wireless Technologies 2 November 2000.
Charts
CHART 1 EFFECT OF GAIN.........................................................................................................................6
CHART 2 OMNI ANTENNA ON ONE METRE MAST AT VARIOUS SPACINGS...............................7
CHART 3 OFFSET ANTENNA ON ONE METRE MAST AT VARIOUS SPACINGS..........................8
CHART 4 BUILD UP OF SECTORED SITE COVERAGE........................................................................9
CHART 5 ILLUSTRATION OF INTERFERENCE IN RE-USING FREQUENCIES...........................10
CHART 6 - AN ELECTRICALLY DOWN-TILTED PANEL ANTENNA MECHANICALLY UP-
Chart 8 Illustration Of difference between Electrical and Mechanical Down-Tilt
2.4 DIVERSITY
Chart 9 illustrates the variability of the strength of a received signal coming from a mobile
transmitter over time, in to both polarisations of a dual slant polarised antenna. Signals usually
arrive at the receiver via multiple paths (see below). This receiver diversity can be used to enhance
systems performance. This is particularly useful when the system requires talkback from low
powered handheld devices. This technique ensures that the network receives the same signal at
least twice (dual receiver mode) which is then manipulated either by an additive or a selective
process to ensure a better net received signal to noise ratio.
Signal Strength In Dual Polar Antenna with Distance Travelled
-120
-100
-80
-60
-40
-20
0
Time Travelling
Sig
na
l S
tre
ng
th
Left Polar
Right Polar
Chart 9 Variability of the Signal Strength coming from a mobile transmitter over time
Sigma Wireless Technologies 13 November 2000.
Quoting from reference ii may help to understand the complexities of propagation in the mobile
radio environment: - "Radio wave propagation in the mobile radio environment is described by
dispersive multi-path caused by reflection, diffraction and scattering. Different paths may exist
between a BS and a MS due to large distant reflectors and/or scatterers and due to scattering in the
vicinity of the mobile, giving rise to a number of partial waves arriving with different amplitudes
and delays. Since the mobile will be moving, a Doppler shift is associated with each partial wave,
depending on the mobile's velocity and the angle of incidence. The delayed and Doppler shifted
partial waves interfere at the receiver causing frequency and time selective fading on the
transmitted signal."
The available antenna diversity options are: - Space -Vertical.
Space - Horizontal
Polarisation (Usually dual polarisation)
The principle is the same for each, in that the receiving base station has a choice of two signals on
the incoming path. The process on average yields a ‘gain’ on the receive path.
A separate paper which looks at diversity gain versus antenna spacing is available during the
second quarter of 2000.
2.4.1 Diversity Gain Explained
Diversity gain only operates on the up-link (Mobile Station to Base Station). It is required because
portables usually have one watt transmit power towards the base, but bases can be up to 40-Watts
back to the mobile. The measurement test involves a mobile and a base station with special test
software in it.
A typical test route is driven; the mean bit-error rate is measured at the base, using a vertically
polarised antenna of equivalent gain to the antenna under test. The route is then driven again, using
either two vertically polarised antennas spaced apart, or the two halves of a cross polarised antenna
(as two separate tests) each being fed into separate receivers. The mobile transmit power is
reduced in steps until the same bit error rate is achieved at the base as was measured in the
reference drive. The amount by which the power is reduced is the equivalent Diversity Gain of the
base antenna configuration chosen for the test.
Sigma Wireless Technologies 14 November 2000.
2.4.2 Experimental Results
During Sigma’s initial TETRA antenna development work, tests were performed to determine if
diversity gain existed in the 400MHz band. From these experimental results, we know that the
diversity gain of a cross-polarised antenna in a suburban environment is about 4dB. The diversity
gain of two vertically polarised antennas horizontally spaced at 5.5 metres is about 4.5dB. We
also know that in a high-density urban environment the gains are increased by a further 1dB. In
open countryside, there is some small gain improvement (over a single antenna) for both
configurations. For three antenna diversity, it is possible to assume that there is at least a 1.5dB
improvement over two antenna diversity. Thus, the diversity gain of a particular antenna
configuration will also depend on the type of environment in which it is being used.
Most cellular operators have, over the years, done experiments to assess the gain obtained with
different diversity schemes, and some have published the results. The findings are generally
similar, but never identical. One set of results is given here iii.
Area Type Estimated Diversity Gain with 45
Slanted Antenna
Estimated Diversity Gain with Space
Diversity
Urban, Indoor 3.7 dB 5.0 dB
Urban, Outdoor 4.7 dB 3.3 dB
Suburban, Indoor 4.0 dB 3.7 dB
Suburban, Outdoor 5.7 dB 4.7 dB
Rural 2.7 dB 5.3 dB
Table 1 - Diversity Gain as a Function of Operating Environment
2.4.3 Different Diversity Schemes Described
• Horizontal space diversity requires that approximately 5.5 meters should horizontally
separate two antennas. Reducing this space reduces the gain and the final gain obtained
depends on the antenna height above surrounding terrain as well as the spacing between the
antennas. This is the optimum situation electrically, but in reality access to the required space
is limited. The greater the antenna separation, the less likely that fades will occur in both
antennas simultaneously. If optimum diversity techniques are used in the base station, expect a
minimum of 3dB diversity gain for two antennas and 4.7dB gain from three antennas.
• Vertical space diversity can be easier to implement, but again the requirement is for
approximately 10 metres vertical separation between two antennas to give the best
Sigma Wireless Technologies 15 November 2000.
improvement over a single antenna (similar to that given by horizontal spacing). Most of the
diversity advantage is lost at 4 metres. One reason for this failure is that the coverage area of
the two antenna systems is very different at this spacing. This will cause many problems trying
to balance the signal quality received at the base with that received at the mobile / portable.
Another disadvantage of this type of diversity is that the two received signal are not the same
strength at the antenna, causing a reduction in diversity gain.
• Dual-polar diversity is achieved using a single antenna structure with two sets of dipoles
positioned at +/- 45 degrees to each other. The dipoles positioned in this way produce of
typically of 3 to 4.5dB better than a single vertically polarised antenna of similar dimensions.
The gain of these antennas is usually specified as Co-Polar gain i.e. the gain measured at +/-45
Degrees. The 2 to 4dB gain improvement is relative to this gain. If, however, you measure the
antenna gain vertically polarised, it will be 3dB less than that measured at +/-45o. The vertical
space occupied by a dual polarised antenna of a given Co-Polar gain is the same as for a
vertically polarised antenna of the same gain.
RedFeed
BlueFeed
Chart 10 Illustration of Dual Polarisation Diversity
2.4.4 How does Space Diversity work?
To achieve diversity at least two receivers are required. These will receive signals from diverse
sources - two antennas. These antennas will provide a separate signal to each receiver, this signal
comes from the same original source - the portable / mobile (called a mobile in the following
discussion) but via different paths. These antennas will need to be positioned on a mast in a
suitable position to allow them to appear as two separate diverse sources of the same signal. The
Sigma Wireless Technologies 16 November 2000.
greater the distance between the antennas horizontally, the less likely that a signal fade (received
from a moving mobile) from one antenna will occur at the same time as a signal fade from the other
antenna. Thus, the diversity gain (reducing the effect of these fades) increases as the separation
increases and relies on the concept that the strength of the two signals should be nearly equal on
average. On average, if the two signal strengths are not equal, then the full diversity gain cannot be
achieved.
At 900 MHz, antennas are generally regarded as being at optimum separation at 2.75 metres. At
400 MHz, this optimum is generally regarded as occurring at five and a half Metres, which is often
difficult to achieve in practical situations. The required separation is in fact a function of effective
antenna height. A separate paper which looks at diversity gain versus antenna spacing is available
during the second quarter of 2000.
The correlation coefficient between the amplitude envelope of the received signals depends on the
antenna spacing. To give an adequately low coefficient (0.7), the antennas should be at the same
height and spaced at least 5.5 metres apart. In other words, the long-term correlation between the
amplitude of the received signals should be high, but the instantaneous value of the correlation
should be very low. (The lowest short term correlation coefficient achievable with two antennas is
approximately 0.5, which is adequate to achieve expected diversity gain. The lower the short-term
correlation coefficient, the better the diversity gain). If these criteria are met by the antenna
system, (and the receivers receive equal amplitude signals on average in the long term), the gain
achieved by two receivers over one can be up to 5dB and by three receivers is up to 7dB, dpending
on the surrounding propagation environment.
2.4.5 How Does Polarisation Diversity Work?
As a RF signal travels from a moving mobile towards the base antenna, it will follow multiple
paths. The obvious one is directly from the mobile antenna to the base antenna. However this path
is often obstructed, a more indirect path may give a better signal. There will be many paths and
each will be due to reflections. These reflections will change the polarisation of the signal. The
amount by which the signal's polarisation is changed depends on the angle of incidence at the
reflection point. In most instances the signal will be partly reflected and partly refracted at the
surface of the 'reflecting' material. There will be multiple signals propagating from the mobile to
the base and, as it moves, the points at which the signals are reflected will be constantly changing.
Thus, the polarisation and strength of the incident signals at a single base antenna will be constantly
changing in a similar manner.
Sigma Wireless Technologies 17 November 2000.
If two receiving base antennas occupy the same space, but receive signals at different polarisations
(+/- 45o), the RF signals coming from these antennas will be diverse or different (see Chart 9 on
page 13) as the signals incident on the two base antennas from the mobile are of different
polarisations. One antenna characteristic of importance in this regard is Cross-Polar
Discrimination, which quantifies the ability of the antenna to discriminate between the polarisations
of the signals impinging on them. If this is at 15dB or better the correlation coefficient is at 0.58
and it falls rapidly below this value. Thus an antenna with the ability to discriminate between
opposite polarisations at better than 15dB across the field-of-view of the antenna, will have an
adequately low correlation coefficient to achieve the 3 to 5dB diversity gain improvement.
3 Antenna Characteristics for Optimum Performance
The role of an antenna system is driven by the need to balance the conflicting requirements of
electrical performance (gain, pattern, tilt), physical dimensions (restricted space available on masts)
and product costs.
• Having wide coverage from sites reduces network costs, but also reduces capacity. Traffic
density will steer you either to Omni, Offset Omni or sectored panels.
• Reduced Physical dimensions make it easy to get lower cost mast space, but reduces electrical
performance. Gain, Front to Back ratio and bandwidth are affected by the dimensions
• Sector planning will influence the choice of tilt. The cell size and frequency re-use plan will
dictate the level and type of tilt required (mechanical or electrical). In addition, the ‘pattern’
control ensures predictable coverage in a range of environments.
Chart 11 below shows the process for selecting the optimum antenna system depending on the
application:
Sigma Wireless Technologies 18 November 2000.
Sectored
• Gain• Receiver Diversity• Front to Back Ratio• Bandwidth
Omni
• Gain• Diversity arrays• Bandwidth
Low powerdevices
Mast space/costs
Yagi
Traffic Density
Coverage
Chart 11 Process for Selecting Optimum Antenna System
3.1 High Traffic Density
Dense traffic systems typically will be required in urban environments where simultaneous
communication is the important issue. The use of multi-frequency sector arrays ensures more
capacity in the area covered by the antenna array.
Taking the sectored approach will require an array of antennas at each site and as mast space is at a
premium, care needs to be taken to ensure that the panel antennas have the smallest dimensions
possible while delivering good electrical performance. Key parameters at risk as you try to reduce
the overall dimensions include, gain (length), front to back ratio (width), bandwidth (depth) and
horizontal beamwidth.
On the inbound side, receiver diversity is used to balance the system. Cell sizes can be maximised
using dual polarised antennas or arrays of space diversity antennas. Space diversity may be used
with Omni as well as panel antenna arrays and offers the maximum electrical performance possible.
However, from a practical point of view the current practice is to use dual polar panel antennas in
high traffic density sites and use arrays where space is more readily available.
Sigma Wireless Technologies 19 November 2000.
3.2 Medium / Low Traffic Density
Omnidirectional antennas can be set up to work in an omni or an offset mode. The choice depends
on the type of mast on which it is placed. If a ‘top of mast’ position is available, the result is
optimum. The main issue is the gain of the antenna and where the inbound signals are low then
either the outbound power needs to be restricted (small cells) or receiver arrays need to be used in
conjunction with a separate Tx antenna.
4 Good Tetra Antenna System Design Practice
Good design practice ensures that the antenna system is compatible with the infrastructure and
offers additional benefits through delivery of: -
1. Lower Costs
2. Higher redundancy
3. Optimum channel usage (capacity)
These benefits can be derived using a combination of good mounting practices and infrastructure
enhancement. The following outlines some possible antenna configurations, along with possible
explanations for choice of configuration. These are only examples and the final choices taken will
be determined by the system designer. The diversity gain shown in the examples is 3.5 for
illustration purposes only. It is up to the system designer to avail of the currently available
information on diversity and the radio environment to decide on the value to apply in a particular
circumstance. The examples are given primarily to stimulate thought and not to be final solutions.
4.1 Receiver Isolation from Transmitters
In section 6.5.1 of Reference i, the level of blocking for a base station is -25dBm. With a
transmitter level of +47 dBm, the isolation between two antennas with transmitter into one antenna
and receiver into the other will need to exceed 72dB. This does not take into account any additional
filtering, or the fact that some manufacturer's equipment will exceed minimum TETRA
requirements. In fact many system designers use a band-pass duplexer on the transmit / receive
antenna and use a separate band-pass filter for each receiver. (Some even use half of the duplexer
for this purpose, as this reduces the number of different filter types on and individual site). In all
antenna configurations given below this requirement for isolation must be taken into account as
must any additional insertion losses in the receiver and transmitter paths.
Sigma Wireless Technologies 20 November 2000.
4.2 Omnidirectional Diversity Applications
4.2.1 Introduction
GSM networks give priority to site capacity. Thus, the use of GSM sectored sites using panel
antennas is widespread and has led to the use of polarisation diversity, in preference to space
diversity.
In contrast to GSM, TETRA networks will place a lower emphasis on capacity and will seek to
maximise cell size (and minimise the use of frequencies) using omnidirectional antenna arrays.
Most of the work in the area of diversity was originally done for GSM frequencies, where the
horizontal diversity spacing is only 3m.
The examples given below are presented to show that there are many different ways to achieve
omni coverage from towers. They will give the designer some idea of how to go about coming up
with the design that is optimum for his own network.
A separate paper which looks at diversity gain versus antenna spacing is available during the
second quarter of 2000.
4.2.2 Sample Power Balance Calculation
Table 2 below shows an example non diversity power balance calculation for a system with a
single antenna of 5dB gain connected to a base station with a duplexer of 1dB insertion loss and a
combiner and filtering with a loss of 4dB. There is no diversity gain. The figures are only
representative and should only be used as a guide. It assumes a 3-Watt portable. The base station
transmit power is adjusted to balance the outbound path with the inbound path. This adjustment of
the base station power applies in all examples in this section.
In the table below BS is the Base Station and MS is the Mobile Station.
Sigma Wireless Technologies 21 November 2000.
BS -> MS MS->BSTx Power 42 dBm 35 dBmCombiner/Filter Losses -5 dB -1 dBMobile Antenna Gain 1 dBi 1 dBiBase Antenna Gain 5 dB 5 dB BaseCable Losses -2 dB -2 dBDiversity Gain 0 dB 0 dBRx Sensitivity -112 dBm -115 dBmResultant System Gain 153 dB 153 dB
Tx
Combiner / Filter
Duplexer
Rx
Main Antenna
Table 2- System Gain in Reference Example
4.2.3 ‘Top of Mast’ Omni plus Two Offsets
This method uses a single omni at the top of the mast and two offset four stack arrays positioned so
that their tops are positioned below the omni and are mounted about three metres off each side of
the mast. Table 2 shows the resultant increase in system gain. Setting the Transmitter Power in the
Base Station (BS) four and a half dB higher than in the reference example (section 4.1)
counterbalances this extra gain and the reduced loss.
This configuration will suit triangular masts of up to around three metres. Above this size, the
pattern of the offset antennas will become increasingly distorted. The exact amount of distortion
depends on how far away from the tower the antennas are mounted and the exact nature of the
tower's construction. (See MAST POSITION on Page 7).
BS -> MS MS->BSTx Power 45.5 dBm 35 dBmCombiner/Filter Losses -4 dB 0 dBMobile Antenna Gain 1 dBi 1 dBiBase Antenna Gain 5 dB 5 dBBase Cable Losses -2 dB -2 dBDiversity Gain 0 dB 3.5 dBRx Sensitivity -112 dBm -115 dBmResultant System Gain 157.5 dB 157.5 dB
Tx Rx1 Rx2
Main Antenna Other Two Antennas
Table 3 - System Gain for Two Offsets plus an omni.
4.2.3.1 Diversity Gain in this configuration
The pattern of the omni antenna at the top of the mast is shown in black in the drawing below. This
is 3.5dB less gain than peak gain of the offset antennas. At the 90 and the 270-degree areas, of the
Sigma Wireless Technologies 22 November 2000.
pattern shown below, the gain is 3.5 dB higher than that of the Transmit antenna. As we sweep
around towards the 0o and the 180o positions, we see two antennas with nearly equal gains.
The offset antennas are spaced at 5.5 metres from each other and would give a diversity gain of
3.5dB, in the direction where their gains are equal (at the 0o and the 180o positions). As we rotate
away from this the gains of the two antennas become markedly different, so the diversity gain will
be reduced in these directions but the gain of individual antennas becomes closer to 8.5dB. Thus,
overall we will get an apparent improvement of about 3.5dB over a single 5dB antenna mounted on
the top of the mast.
In such a configuration, consideration should be given to providing sufficient isolation between the
transmit antenna at the top of the tower and the receive antennas on the side of the tower. (See 4.1
Receiver Isolation from Transmitters on page 20)
0 - 3 - 6 - 1 0
- 1 5- 2 0
d B
0
9 0
1 8 0
2 7 0
Chart 12 'Top of Mast' Omni Plus Two Offsets.
4.2.4 ‘Top of Mast’ Omni plus Three Panels
This method uses a single omni at the top of the mast and three panel antennas set at 1200 to each
other around the mast. The transmitter feeds the omni at the top and the three receivers operate
from the three panel antennas around the mast. Setting the Transmitter Power in the Base Station
(BS) 3.1dB higher than in the reference example above counterbalances this extra gain. This
Sigma Wireless Technologies 23 November 2000.
arrangement best suits triangular masts over two to three metres per side. It can be used on any
size tower, as the system works better as the distance between the panels increases above 5.5
metres.
BS -> MS MS->BSTx Power 45 dBm 35 dBm
Combiner/Filter Losses -4 dB 0 dBAntenna Gain 5 dB 5 dBDiversity Gain 0 dB 3 dBRx Sensitivity -112 dBm -115 dBm
Resultant System Gain 158 dB 158 dB
Tx
Combiner / Filter
Rx1 Rx2 Rx3
Transmit Antenna
Three Panel Antennas
Table 4- System Gain for Omni plus Three Panels.
4.2.4.1 Diversity Gain in this configuration
The pattern of the omni antenna at the top of the mast is shown in black in the drawing below. At
the zero point, the first panel antenna (red pattern) has a gain that is 3.1dB higher than the omni.
As we sweep clockwise around towards the 60-degree point, the gain of this antenna drops by
about 7 dB. If the panels are spaced at centres of 5.5 metres or more, the resultant diversity gain is
about 3 to 4dB (counting both the red and the blue pattern). This leaves the net resultant gain in
this direction about 3 to 4dB down on the peak of the red pattern (-7dB + 3 and -7dB + 4). In a
similar manner, the net gain will also be reduced at 1800 and 3000 for the other coloured patterns.
Thus, overall we will get an apparent improvement of about 3.1dB over a single 5dB antenna
mounted on the top of the mast, with a reduction of approximately 3dB in net gain at the overlap of
the patterns.
In such a configuration, consideration should be given to providing sufficient isolation between the
transmit antenna at the top of the tower and the receive antennas on the side of the tower. (See 4.1
Receiver Isolation from Transmitters on page 20).
Sigma Wireless Technologies 24 November 2000.
0 -3 -6 -10
-15-20
dB
0
90
180
270
Chart 13 'Top of Mast' Omni plus Three Panels
If we assume that the panels are mounted at the end of an extension pole on each side of a
triangular mast, Table 5 below gives an indication
of the length of the extension poles required to give
a total of 6 metres between the centres of the
panels. Note that this is end to end of the extension
poles and takes into account the fact that the
electrical centres of these antennas are about
100mm forward of the backplane and the brackets
mount the antenna about 200mm further away from
the vertical mounting pole. Thus, the extension
poles net length could be reduced by up to 400mm,
without affecting the spacing too adversely.
Sigma Wireless Technologies 25 November 2000.
Chart 14 Extension Pole for Omni Plus 3
Panels
Tower Side Size (Metres)
Pole Extension length (Metres)
1 32 2.43 1.94 1.35 0.656 0
Table 5 - Extension Pole Length Vs Tower Size
4.2.5 Side Mount ‘Omnidirectional’ Diversity Array
This antenna array comprises a pair of four stack antennas, one of which is used for Tx while both
are used for receiving to achieve space diversity gain. The antenna system functions identically to
and has the advantages of the ‘Offset omni’ (See page 7).
The antenna configuration has the radiation pattern described below. Network planners would
direct the pattern peak in the appropriate direction.
First Antenna
Rx Multicoupler A
Duplex Filter
Tx Combiner
1A 2A nA1 2 n
N by 4 Voice ChannelsN by RF ChannelsTwo Diverse antennas shown
Second Antenna
Rx Multicoupler B
1B 2B nB
Chart 15 Side Mount ‘Omnidirectional’ Diversity
Array
4.2.5.1 Diversity Gain in this configuration
The radiation pattern is shown in Magenta as both the
Tx and Rx patterns overlap completely. The dipole
arrays of both antennas need to be pointed in the same
Sigma Wireless Technologies 26 November 2000.
0 - 3 - 6 - 1 0
- 1 5- 2 0
d B
0
9 0
1 8 0
2 7 0
Chart 16 - Side Mount ‘Omnidirectional’
Diversity Array Gain
direction (note that they are shown in the centre of the radiation pattern and it is worth noting their
directions). The diversity gain of the antennas will yield a 3dB improvement in all directions
around the mast compared with a single offset antenna and the magenta represents the combination
of the red and blue radiation patterns.
In such a configuration, consideration should be given to providing sufficient isolation between the
transmit antenna and the receive antennas on the other side of the tower, especially receiver 'B'.
(See 4.1 Receiver Isolation from Transmitters on page 20).
4.2.6 Two Sector Hybrid Sector System
In some circumstances, it may be appropriate to install a hybrid sector array which operates as two
discretely separate sectors. The first part of the sector uses a cross-polarised panel antenna in one
direction and the second part of the sector uses two separate stacked dipole arrays using space
diversity. The pair of stacked dipoles is used in the classic way, one antenna Tx/Rx with a
duplexer and the other antenna providing the second Rx path. The stacked dipole in its offset
configuration is Omnidirectional but skewed in one direction. This should be pointed in the
opposite direction to the panel antenna. The resulting coverage is egg shaped and slightly offset in
the direction of the panel. It has been used where a high traffic density is required and where this
shape of coverage is not a disadvantage (Such as at Motorway Junctions, with the heavier traffic in
the direction of the Panel).
Second Antenna
1C 2C 4C
1A 2A 4A1 2 43
Half of First AntennaSecond Half not Illustrated
for Clarity
Rx Multicoupler ATx Combiner A
Example shows two sector cell – One Offset andone panel system
Rx Multicoupler B
1B 2B 4B
Tx Combiner B
1B 2B 4B
Duplex Filter Duplex Filter
3B
Third Antenna
Rx Multicoupler C
First Sector - Panel Second Sector – Two Offset
Chart 17 Two-Sector Hybrid Sector System
Sigma Wireless Technologies 27 November 2000.
4.2.6.1 Diversity Gain in this configuration
The radiation pattern for the stacked array is shown in Magenta as both the Tx and Rx patterns
overlap completely, as in the ‘two offset’ mode.
The dipole arrays of these two antennas need to
be pointed in the same direction (note that they
are shown in the centre of the radiation pattern
and it is worth noting their directions). The
diversity gain of the antennas will yield a 3dB
improvement in all directions, for this antenna
pair over a single antenna of the same type. The
magenta represents the combination of the red
and blue radiation patterns. On the opposite
side of the mast a single cross-polarised panel
antenna is mounted on a different set of
channels. This also has a diversity gain of 3dB
relative to the nominal gain of the antenna. This
antennas radiation pattern is shown in green.
5 Mounting Criteria for Optimum Performance
5.1 Electrical
5.1.1 Background
The following testing was carried out by Dolphin Telecommunications. Where several services
share a radio site there will inevitably be some potential for interference between the various
signals present. This section examines the possibility of signals radiated from a TETRA 80º panel
being received by a 900 MHz base station through coupling between the two panels. The TETRA
band is not harmonically related to either of the TACS or GSM frequency bands and so
interference caused by the direct impact of harmonics of the TETRA carrier has been discounted.
It is therefore assumed that the primary mechanism for concern would be blocking of the 900 MHz
receiver by the TETRA transmitter.
GSM interference level recommendation 05.05 requires that a GSM base station is able to operate
normally in the presence of an interfering out of band signal at a power level of 0 dBm at the GSM
Sigma Wireless Technologies 28 November 2000.
0 - 3 - 6 - 1 0
- 1 5- 2 0
d B
0
9 0
1 8 0
2 7 0
Chart 18 - Two Sector Hybrid Sector System
Gain
receiver antenna terminals. It is therefore desirable to keep any TETRA signal presented to the
receiver terminals of the GSM receiver below this level. The TETRA EIRP is limited by licensing
requirements to less than +47 dBm. A minimum acceptable loss in the coupling between the two
panels is therefore 47 dB.
5.1.2 Dolphin Measurement
While the antenna characteristic data can be obtained for each antenna, the published data refers to
the performance within the intended band of operation. Since coupling between the 400 MHz and
900 MHz systems includes reception of the potential interference outside the intended band of
operation of the antenna, it is difficult to predict the resulting coupling theoretically.
5.1.3 Practical measurement
In the absence of a theoretical calculation the coupling between two example systems was
measured to indicate the levels involved. Two GSM panel antennas, both cross-polar units with
different H-plane beamwidths, were examined in close proximity to a TETRA 80º panel antenna to
simulate the situation on a shared mast. The measurements were made using a tracking
generator/analyser combination to establish the isolation between the two antennas. The
measurement sweep encompassed all anticipated TETRA frequencies in using the range 380 - 430
MHz. To identify the worst-case coupling between the two panels, each measurement was
repeated using the second polarisation on the antenna so that the cases corresponding to a co-polar
and cross-polar coupling were both taken into account.
T E T R Ap a n e l
8 0d e g r e e s
H B W
G S Mp a n e l
8 5d e g .
H B W
4 2 t o 6 2 d B l o s s
4 7 t o 6 8 d B l o s s
a t T E T R Af r e q u e n c i e s
a t G S Mf r e q u e n c i e s
1 . 0 - 1 . 5 5 m e t r e s
Chart 19 - Measurement configuration for horizontal separation
Sigma Wireless Technologies 29 November 2000.
For comparison purposes, a measurement of the coupling between the two panels in the GSM
frequency range 870 – 960 MHz, with the test generator connected to the GSM panel, was also
taken.
Tests were carried out with the panels on a common azimuth in each case. Two horizontal
separations (1m and 1.55m) were tested to examine the effect of increasing the separation between
the two panels. To investigate the possibility of mast sharing where the TETRA panel is installed
below a GSM installation, a separate measurement was made of the coupling when the two panels
are positioned end-to-end. The measurement was made with a single vertical separation of 1m
TETRApanel
80degrees
HBW
GSMpanel
85deg.
HBW
47 to 68 dB
loss
1.0
m
etre
s
at T
ET
RA
fre
qu
en
cie
s
Chart 20- Measurement configuration for vertical separation
5.1.4 Summary of results
TETRA to GSM, 1m horizontal separation
Radiating panel Polarisation TX band RX panel Polarisation Minimum
isolation
Maximum
isolation
TETRA 80º +45º TETRA GSM 60º +45º 42 dB 50 dB
TETRA 80º +45º TETRA GSM 60º -45º 48 dB 51 dB
TETRA 80º +45º TETRA GSM 85º -45º 45 dB 51 dB
TETRA 80º -45º TETRA GSM 85º -45º 43 dB 49 dB
Sigma Wireless Technologies 30 November 2000.
TETRA to GSM, 1.55m horizontal separation
Radiating panel Polarisation TX band RX panel Polarisation Minimum
isolation
Maximum
isolation
TETRA 80º -45º TETRA GSM 85º -45º 48 dB 62 dB
TETRA to GSM, 1m vertical separation, TETRA below
Radiating panel Polarisation TX band RX panel Polarisation Minimum
isolation
Maximum
isolation
TETRA 80º -45º TETRA GSM 85º -45º 51 dB 67 dB
TETRA 80º -45º TETRA GSM 85º +45º 54 dB 76 dB
GSM to TETRA, 1m horizontal separation
Radiating panel Polarisation TX band RX panel Polarisation Minimum
isolation
Maximum
isolation
GSM 60º -45º GSM TETRA 80º -45º 51 dB 61 dB
GSM 60º -45º GSM TETRA 80º +45º 47 dB 68 dB
5.1.5 Analysis of results
The coupling between the two panels is affected by the polarisation relationship, despite the fact
that the receiving panel is not tuned for the receiving frequency and that the ‘side-on’ orientation
does not obviously suggest any correlation. The results show that the isolation is reduced when the
two elements in use are on the same polarisation.
The isolation between the two panels is noticeably increasing with separation – in the test carried
out the minimum isolation improved by 5 dB as the distance was increased from 1m to 1.55m
without changing any other parameters. The results suggest that the required TETRA to GSM
isolation is achieved with an inter-panel horizontal separation of as little as 1.55 metres.
For vertical separation, the test results suggested that the required isolation was achieved with a
separation of 1m in all polarisation combinations.
Sigma Wireless Technologies 31 November 2000.
5.1.6 Test limitations
1. The test antennas shared a common azimuth in all configurations. While the data sheet polar
pattern will probably not be applicable in such close proximity to the antenna, it would be
reasonable to assume that the isolation would be reduced if the TETRA panel were turned
towards the GSM panel.
2. The tests were carried out with the two panels mounted horizontally on a non-conductive
support. While the forward beam patterns should have been into free space, it is possible that
building reflections may be responsible for some of the minor frequency-dependant
perturbations observed. Further work to establish the limitations of the test set-up would be
appropriate, however it would seem likely that the isolation figure indicated would be increased
if reflective paths between the two panels are reduced or eliminated.
3. The tests were carried out with two GSM panel antenna types. For more general applicability,
it would be appropriate to repeat the exercise with other products.
4. The figures indicated assumed that there is no additional bandpass frequency filtering in front
of the GSM receiver, providing additional attenuation at the TETRA frequencies.
5.1.7 Conclusions
From the data collected so far, it would be appropriate to use an inter-panel spacing guideline
requiring horizontal separation of at least 2m between TETRA and GSM, if the sector orientations
are on a similar azimuth. It is suggested that the vertical separation should be at least 1m.
Sigma Wireless Technologies 32 November 2000.
5.2 Physical Mounting Criteria
The antenna system most appropriate for a given application is governed by:
• Mast type and position.
• Traffic Density
• Radio system configuration
5.2.1 Sectored Panel array
1. Pole mount
This approach involves the three antennas mounted at the same height, each arranged at 120
degrees to each other. Either the antennas may be vertically polarised or dual polarised where
diversity gain is required.
2. Mast mount
The panel may be mounted on each leg of a triangular tower, again at the same level as before.
Sigma Wireless Technologies 33 November 2000.
3. Building mount
The panel antennas may be mounted at the edge of a building giving sufficient clearance from the
rooftop and held in position by a steel structure. The positioning of the panels should be as with a
mast, in terms of level and orientation. Free space in front of the antenna should be provided for at
least 20M extending down at an angle of 30 degrees. Avoid roof edge obstructions.
6 Antenna / System Integration
6.1 Low Density System.
This type of system uses a simple approach to radio coverage and shows how a single antenna may
be used to allow two-way communication with four RF channels. The duplex filter allows
simultaneous Tx/Rx operation and is only restricted by the RF power requirements dictated by the
number of channels in use. Below is a representation of how a non-diversity system functions,
using one antenna. This antenna could be Omnidirectional, sectored panel or directional. The use
of one antenna in this arrangement reduces the cost of antennas, plus the cost of mast space.
Sigma Wireless Technologies 34 November 2000.
Antenna
Rx Multicoupler
Duplex Filter
Tx Combiner
1 2 n1 2 n
N by 4 Voice ChannelsN by RF Channels
6.2 Medium Density
The schematic below shows a system using antenna diversity. The Transmitters and the first set of
receivers are connected to one antenna and the second set of receivers is connected to the second
antenna. Therefore, there are diverse sources for the signals being fed into the receivers. Note
there is a practical limit to the number of transmitters that can be fed into the duplexer and the
antenna. This is determined by the PEAK power of the transmitters (the peak power of a
transmitter in TETRA is 6 dB above nominal power and therefore, there is a limitation set by the
voltage as well as the power capabilities of these elements). See also 4.1 Receiver Isolation from
Transmitters for more information on protection of Rx multi-coupler.
The use of an antenna system, in this way enables the use of large cells with the benefit of diversity
gain on the receive path. This could be a single antenna with dual slant polarisation or two
antennas.
First Antenna
Rx Multicoupler A
Duplex Filter
Tx Combiner
1A 2A nA1 2 n
N by 4 Voice ChannelsN by RF ChannelsTwo Diverse antennas shown
Second Antenna
Rx Multicoupler B
1B 2B nB
Sigma Wireless Technologies 35 November 2000.
6.3 High Density
The illustration below shows how a large site might be configured. Half of the transmitters would
be fed into one antenna system and the other half would be fed into the second antenna. All the
receivers would be fed from each antenna to give diversity.
First Antenna
Rx Multicoupler ATx Combiner
1A 2A 8A1 2 4
Example Shows 8 Tx and 16 RxRedundancy in Duplexers and in Tx Combiners
Second Antenna
Rx Multicoupler B
1B 2B 8B
Tx Combiner
5 6 8
Duplex Filter Duplex Filter
3 7
The key benefit of this configuration is to spread the power load between the two antennas and in
that way to give greater redundancy in case of antenna failure, it offers diversity gain as before.
Receiver protection for the second one is also increased by such a configuration.
Sigma Wireless Technologies 36 November 2000.
i ETS 300-392-2 First Edition, March 1996 Section 10.3.3
ii ETS 300-392-2 First Edition, March 1996 Section 6.6.3.1
iii Jaana Laiho Steffens, Jukka Lempeiainen et al., “Experimental Evaluation of Polarisation Diversity Gain at Base
Station End in GSM900 Network”, IEE Transactions, Vehicular Technology 0-7803-4320-4/98 Pages 16-20.