CALIBRATION OF ACTIVE ANTENNA ARRAYS FOR MOBILE
TELECOMMUNICATIONS
Field of the Invention
The present invention relates to antenna arrays employed in mobile
telecommunications systems, and in particular to the phase and/or amplitude
calibration of RF signals in active antenna arrays.
Background Art
In wireless mobile communications, active, or phased array, antenna
systems are emerging in the market, which are used for beam steering and
beam forming applications. Active antenna systems allow increase of network
capacity, without increasing the number of cell sites, and are therefore of high
economical interest. Such systems comprise a number of individual antenna
elements, wherein each individual antenna element transmits RF energy, but
adjusted in phase relative to the other elements, so as to create a beam
pointing in a desired direction. It is essential for the functionality of the system
to be able to measure, control and adjust the phase coherency of the signal
being radiated from the various individual antenna elements of the antenna
array.
In Figure 1 a known active antenna system is depicted, formed from
several individual transceiver elements 4. A digital baseband unit 6 is coupled
to each transceiver element, and each transceiver element comprises a
transmit path 8 and a receive path 10. Each path is coupled to an antenna
element 12. The transmit path 8 processes a signal from baseband unit 6 and
includes a digital to analog converter DAC, a power amplifier PA, and a
Diplexer/Filter 15. The receive path 10 processes signals received from
antenna element 12, and comprises Diplexer/Filter 5, a low noise amplifier
LNA, and an analog to digital converter ADC.
Each transceiver element generates an RF signal which is shifted in
phase either electronically or by RF-phase shifters relative to the other
transceiver elements. Each antenna element thereby forms a distinctive phase
and amplitude profile 14, so that a distinctive beam pattern 16 is formed. It is
therefore necessary to align or calibrate all signal phases and amplitudes from
the individual transceiver elements at the point where they are transmitted by
the antenna elements. To align all transceivers, a common reference is
required. The transmitted signal is then compared in phase and amplitude with
the reference.
To provide a phase and amplitude reference, two different methods have
been used:
. The signal of one element of the array is used as reference and all
other signals are adjusted so that the required coherency to the reference
element is achieved. This method usually requires (depending on the size of the
array and accuracy) very complex algorithms to mutually adjust the elements,
because the adjustment relies on mutual coupling of the elements, which is
weak for elements at larger distances. Or a factory-calibration is used, which is
complicated to recalibrate if, e.g. during the operation of the array, any phase or
amplitude changes in the RF-signal-generation and transmission occurs. This
method also requires a dedicated receiver unit, which is able to receive the
transmitted signals from the other antenna elements. If receive calibration is
also required, a dedicated transmitter is needed for a test signal. The additional
receiver and transmitter increase cost and the associated algorithms require
extra computational resources.
2. A star-distribution network, wherein a reference is generated in a
central unit, which is then distributed to all transceivers, and each transceiver is
aligned with the reference. This method is the preferred ones for smaller arrays
(number of elements10) due to the simpler algorithms required. Critical for the
central reference generation calibration method is that the accuracy of the
reference distribution is high. Each error in terms of phase or amplitude in the
reference will be carried forward to the transmitted/received signal itself. To
accurately distribute the phase reference, a centrally generated reference signal
is split into a set number of signal paths. Each such path is connected to the
respective reference signal input of each transceiver unit of the array by
respective transmission lines, the transmission lines being of nominally equal
length. This method suffers from three draw backs:
a) Each transmission line has to be of at least half the length of the array
size. That means even if an element is located very close to the reference
signal generator, it requires a long cable. This increases cost unnecessarily and
the volume and weight of the network.
b) The number of transceiver elements is limited to the preset number of
signal paths. The network has to be designed for a specific number of elements,
which leads to inflexibility.
c) The mechanical accuracy of the transmission line lengths has to be
great, that is the tolerances must be small, in view of the requirements for
phase and amplitude accuracy of the array itself. For example, for a mobile
communication base station antenna with eight to ten elements operating at a
frequency of approx 2GHz, the required phase accuracy is in the order of ±3°
among elements. This corresponds to an approximate accuracy of the total line
length of ±0.9mm of a Teflon-filled 50 Ohm-coaxial cable with a total length of
approx 700mm (the array itself is approx 1400mm long). To ensure this kind of
accuracy in a mass production environment is expensive, especially if e.g.
thermal expansion during the operation of the antenna and varying bending
radii of the different lines within the antenna structure are also taken into
account.
Summary of the Invention
The present invention provides an active antenna array for a mobile
telecommunications network, comprising a plurality of radio elements, each
including a transmit and/or a receive path coupled to an antenna element, and
each including comparison means for comparing phase and/or amplitude of
transmitted or received signals with reference values in order to adjust the
characteristics of the antenna beam, and including a feed arrangement for
supplying reference signals of amplitude and/or phase, the feed arrangement
including a waveguide of a predetermined length, which is coupled to a
reference signal source, and which is terminated at one end in order to set up a
standing wave system along its length, and a plurality of coupling points at
predetermined points along the length of the waveguide, which are each
coupled to a said comparison means of a respective said radio element.
In accordance with the invention, at least in a preferred embodiment, it is
possible to overcome or at least reduce the above noted problems, and to
provide an accurate distribution mechanism for phase and amplitude reference
signals for calibration of active antenna arrays for mobile communications. The
distribution mechanism in addition in a preferred embodiment is mechanically
robust and cost-effective.
In the present invention, at least in a preferred embodiment, a reference
source signal of phase and/or amplitude is coupled to a finite length of a
transmission line, which is terminated such as to set up a standing wave within
the transmission line length. As is well-known, in a length of transmission line
or other waveguide terminated at one end with its characteristic impedance,
radiated travelling waves will progress along the line and be absorbed in the
terminating impedance. For all other terminations however, some radiation will
not be absorbed, but be reflected from the end, and will set up a standing wave
system, where the resultant wave amplitude changes periodically along the
length of the waveguide (there will in addition be time variation of the voltage
value at each point along the line as a result of wave oscillation /phase rotation).
The amount reflected depends on the terminating impedance, and in the limiting
cases of short circuit and open circuit, there will be a complete reflection. In
other cases, there will be partial reflection and partial absorption.
The standing wave signal may be sampled at predetermined tapping or
coupling points along the length of the line, which all have the same amplitude
and phase relationships, or at least a known relationship of phase and
amplitude. As preferred, such coupling points occur at or adjacent voltage
maxima/minima within the standing wave, where the change of voltage with
respect to line length is very small. Hence, the requirement for mechanical
accuracy in positioning of the coupling point is much reduced as compared with
the star-distribution network arrangement described above.
These coupling points may each be connected by a respective flexible
short length of line of accurately known length to respective comparators in
respective transceiver elements (more generally radio elements). Short lengths
of flexible cable, all of the same length, may be formed very accurately as
compared with the known star-distribution network above.
ln a preferred embodiment, said waveguide may be formed as a plurality
of sections of waveguide of predetermined length, interconnected by releasable
couplings; this permits scaling to any desired size of antenna.
An application of the invention is for frequencies of the order of GHz,
usually up to 5GHz, that is microwave frequencies in the mobile phone
allocated bands, where coaxial cable is generally used as a transmission line.
However the invention is applicable to other frequencies, greater and smaller,
and coaxial cable may be replaced by other waveguide and transmission line
constructions such as hollow metallic waveguides, tracks on a printed circuit, or
any other construction.
Brief Description of the Drawings
A preferred embodiment of the invention will now be described, by way of
example only, with reference to the accompanying drawings, wherein:
Figure 1 is a schematic diagram of a known active antenna array
comprising a number of transceiver elements;
Figure 2 is a schematic diagram of a means of distributing a reference
signal to respective transceivers of an active antenna array, incorporating the
known star-distribution network;
Figure 3 is a schematic diagram of progression of a travelling
electromagnetic wave along a transmission line length, having its free end
terminated with a matching impedance;
Figure 4 is a schematic diagram of a standing electromagnetic wave
along a transmission line, which has its free end terminated with a short circuit;
Figures 5a, 5b, and 5c are diagrammatic views of a length of
transmission line with coupling points formed by capacitive coupling ports, for
use in a preferred embodiment of the invention;
Figure 6 is a schematic view of a feed arrangement of a reference signal
to transceiver elements of an active antenna, in accordance with a preferred
embodiment of the invention;
Figure 7 is a schematic block diagram of a means for phase and
amplitude adjustment within a transceiver element of the active array of Figure
Figure 8 is a schematic diagram of a modification of the preferred
embodiment, forming a distribution arrangement for 2-D arrays.
Description of the Preferred Embodiment
In the following description, where reference is made to the transmit
path, it will be appreciated the invention can be used in the same way to provide
a reference for the receive path. The invention is applicable both to transmit
and receive cases.
Referring to Figure 2 , this shows a means of distributing a reference
signal of phase and amplitude to the individual transceivers of a n active
antenna array. A centrally generated reference signal 20 (VCO PLL) is split in
a n N-way-power divider 22 (1 :N-splitter) and connected to the reference input of
each transceiver unit 24 by respective transmission lines 26 of equal length I.
Length I is nominally equal to half the length of the array . This forms the
known star-distribution network, and any change of the line length results in a
change of the phase length, giving rise to the disadvantages noted above. This
is due to the travelling nature of the wave propagation on the line: the phase
change is proportional to the length which the wave travels along the line:
= (360/ ), where is the wavelength of the radiation in the transmission
line. If one looks at a travelling wave at a certain snap-shot in time, the phase
changes with the position along the transmission line, a s indicated in Figure 3.
In Figure 3, voltage values are shown existing along the line at time intervals
- - As is well known the measured voltage value is dependent on the
amplitude A and phase of the electromagnetic wave, and in the travelling
wave of Figure 3, the measured voltage will vary, with time, at each point on the
line between +A and -A. In Figure 3, the line length is terminated with the
matching impedance of the transmission line, so that all the energy of the
travelling wave is absorbed. If however a line length is terminated with a n
impedance other than a matching impedance, then a standing wave system
may be set up.
A standing wave arrangement is shown in Figure 4 . Such a standing
wave can be generated along a line 40 by feeding it with a signal 42 from one
end and shorting the signal at the other end 44. This short enforces a voltagenull
at the end of the line. The same energy that travels along the line is fully
reflected at the short and travels backwards towards the source. If the line is
lossless (or reasonable low loss), this leads to a standing wave on the line.
Thus, the voltage value at any point along the line now depends on time, but the
phase of the wave does not vary along the line, rather the amplitude A of the
electromagnetic wave varies cyclically along the length of the line, between
maxima and minima, (positive and negative peaks), the maxima being spaced
apart one wavelength of the wave, as shown. The first minimum occurs at a
distance of /4 from the shorted end. At any given point along the line e.g. x 1
and x2 the amplitude is different. The maximum voltage occurs at the same
point in time as the minimum.
If the voltage on the line is now sampled by couplers 46 with a low
coupling coefficient in order not to interfere with the standing wave, then the
maximum at each coupler output occurs at the same time (even they may differ
in amplitude). If it ensured that each coupler is spaced in a distance of 1,
where is the wavelength of the radiation in the transmission line, then it is also
ensured, that the amplitude at each coupler output is equal. If different
amplitudes are desired, not necessarily equal, other distances than can be
chosen.
In accordance with the invention, this arrangement of couplers attached
to a line having a standing wave, may be used to transmit an amplitude and
phase reference signal to the individual antenna elements of an active array
system. Each coupler is attached to a respective transceiver by a short length of
cable, of accurately known length. A primary advantage of this arrangement is
that it avoids the strict requirements of mechanical accuracy of the star
distribution arrangement of Figure 2. To minimize the amplitude difference
between coupling or tapping points, it is desirable to space the couplings in a
distance of d= (+/4) from the shorted end; this places each coupling in a
voltage-peak of the standing wave. Since the voltage distribution along the line
follows a sinusoidal function, and the derivative of the sinusoidal function near
the maximum/minimum value is zero, the sensitivity of the amplitude of the
coupled signal to the physical position of the coupling point is minimal.
This arrangement overcomes shortcomings of the star-distribution
arrangement, since the reduced dependence of the phase reference on the
physical location of the coupling point along the line reduces the manufacturing
cost and increases the accuracy of the system according to the invention as
compared to a star-network. The signal may be transported from the coupling
port to the reference comparator in the respective transceiver by a much shorter
cable (e.g. in the order of several cm instead of several ten cms of the star
network) and therefore be manufactured much more precisely. Due to the
shorter cable lengths, the costs of the cables/line between the reference-line
and the comparator are also reduced. The dependence of the amplitude of the
coupled signal is minimized by placing the coupling ports at distances
d=(NA+A/4). For example, at 2GHz and a Teflon filled line, a misplacement of
the coupling point from the voltage maximum of +/-5mm corresponds to a shift
of 16.8°. With cos(16.8°)=0.95 this reduces the coupled amplitude by
20*log(0.95)=0.38dB, which is about half of the permitted tolerance in amplitude
accuracy for mobile communication antennas. Therefore the required
mechanical accuracy has been reduced from a sub-mm-level tolerance to a
level of several mm tolerance. It is much easier to achieve a sub-mm- or mmaccuracy
on a short connection line between the standing wave line and the
transceiver than on a line which is orders of magnitude longer, as in a starnetwork.
In Figures 5a, 5b, and 5c a preferred form of coaxial line is shown, which
is incorporated a distribution arrangement for amplitude and phase reference
signals according to the invention. In Figure 5a, a transmission line, which is a
coaxial line 50 with a shorted free end 52, is coupled to a reference source 54.
The line has a series of spaced capacitive coupled coaxial coupling or tapping
ports 56. A perspective view of a coupling port is shown in Figure 5b. In
Figure 5c, a part-sectional view of a physical implementation of the transmission
line is shown, comprising a length of air-filled coaxial line 60, which has a length
equal to one wavelength of the transmission signal (a 2Ghz signal has a
wavelength of the order of 15 cm in free space). One end has a male coupling
connector 62, and the other end a female coupling 64, for coupling to identical
sections of coaxial line, in order to provide a composite line of desired length.
The length 60 has a capacitive coupling port 66, having an electrode pin 68
which is adjustable in its spacing from a central conductor 70. The coupling
coefficient can be tuned to a desired value by the length of the coupling pin
protruding into the standing wave line.
In the illustrated case of the standing wave line filled with air, the
distance between the ports 56 is A0=c0/f with AO being the wavelength in free
space. In antenna arrays the distance of antenna elements is usually between
0.5 AO and 1A0, so that no gratings lobes occur in the array-pattern. In mobile
communication antenna arrays this distance is usually in the order of -0.9 AO. It
is beneficial, that the distance between the coupling-ports for the reference
signal matches the element distance, so the length of the wave guide that
connects the coupling ports with the comparator-input is minimized. This is
possible with the invention, by adapting the effective dielectric permittivity Eeff
used in the standing wave line such, that the physical length lc between the
couplings equals approximately the element distance d between the antenna
elements: 0.9 A0=d A0/(square root(Eeff)). This is possible by using e.g. foammaterial
in the coaxial line as a dielectric and adjusting the dielectric permittivity
by the density of the foam.
Figure 6 shows a preferred embodiment of a distribution arrangement for
reference signals of amplitude and phase to an active antenna system. The
embodiment incorporates the coaxial line of Figures 5, and similar parts to
those of earlier Figures are denoted by the same reference numeral. In this
embodiment the coupling or coupling ports 56 are separated by an effective
distance of 0.9 A, and each coupling port 56 is connected by a short (of the
order of a few cms, and short in relation to the length of line 50) flexible coaxial
cable 72 to a respective transceiver (radio) element 4, which includes a
comparator 100 and which is coupled to an antenna element 12. The lengths of
the cables 72 are precisely manufactured to be equal.
The arrangement for processing the phase and amplitude reference
signal within a transceiver (radio) element is shown in Figure 7. A Digital
baseband unit 80 provides signals, which include digital adjustment data, to a
DAC 8 1, which provides a transmission signal for up-conversion in an
arrangement comprising low-pass filters 82, VCO 84, mixer 86, and passband
filter 88. The up-converted signal is amplified by power amplifier 90, filtered at
92, and fed to antenna element 94 via an SMA connector 96. To achieve phase
calibration and adjustment, a directional coupler 98 senses the phase and
amplitude A, of the output signal. This is compared in a comparator 100 with
phase and amplitude references A ref, r at 102, to provide an adjustment value
104 to base band unit 80. Alternatively, if analog adjustment is required, a
vector modulation unit 106 is provided in the transmission path. Thus, the
comparator output 104 is fed back either to a digital phase shifter and
adjustable gain block 80 or an analog phase shifter and gain block 106, to
adjust the phase and amplitude of the transmitted signal until its phase and
amplitude matches the reference value.
The arrangement of capacitive coupling points of Figure 5, that is simple
envelope detectors for the standing wave detection, may leave a 180° phase
ambiguity. This ambiguity may be resolved by employing two similar standing
wave lines, working with same frequency signals, but fed with, e.g., 90°phase
difference (i.e., T/4 time difference). Then, detection can comprise using two
detectors against ground, or using one detector between the two lines.
An advantage of the distribution means of preferred embodiments of the
present invention is that it is scalable: the line can be designed as a single
mechanical entity, or as a modular system, which is composed of several
similar elements, which can be connected to each other. If more coupling points
are required, the line length is increased by simply adding more segments.
In a modification, a distribution system for 2-dimensional arrays is
provided. This is shown in Figure 8, where a first line 110, as shown in Figures
5, is coupled at each coupling point 112 to further coaxial lines 1 4 , each line
114 being disposed at right angles to line 110, and each line 114 being as
shown in Figures 5 and having further coupling points 116. Coupling points 116
are connected to respective transceiver elements of a two dimensional active
array.
In a further modification, by choosing a symmetrical implementation of
the coupling points about the mid-point of the standing wave line, the accuracy
can be improved further. Any error occurring in phase or amplitude is now
symmetrical about the center of the array. If any phase or amplitude error
occurs now along the reference coupling points (e.g. due to aging effects of the
line), the symmetry of the generated beam is nevertheless ensured and no
unwanted beam tilt effect occurs. Further, a temperature gradient along an
active antenna array does not affect phase accuracy of the signals distributed to
the respective antenna radiator modules. In practical operation, the uppermost
antenna can easily experience an ambient temperature 20-30 degrees higher
than the one of the lowest element. This can cause a few electrical degrees
phase shift difference in a coaxial cable.
Thus the mechanism of the invention, at least in its preferred
embodiment, overcomes the noted shortcomings of the prior art and may
provide the following advantages:
Scalability (in 1D and 2D). The invention may therefore be ideal for the
design of antenna arrays of varying sizes, depending on the required gain,
output power and beam width of the system.
The required mechanical accuracy may be reduced theoretically
completely if it is used for phase reference distribution. In cases where it is
used also as an amplitude reference, the required mechanical accuracy is
decreased from a sub-mm-level to a level of several mm.
The cost, weight and volume of the preferred form of reference
distribution of the invention is reduced as compared to the prior art.
The description and drawings merely illustrate the principles of the
invention. It will thus be appreciated that those skilled in the art will be able to
devise various arrangements that, although not explicitly described or shown
herein, embody the principles of the invention and are included within its spirit
and scope. Furthermore, all examples recited herein are principally intended
expressly to be only for pedagogical purposes to aid the reader in
understanding the principles of the invention and the concepts contributed by
the inventor(s) to furthering the art, and are to be construed as being without
limitation to such specifically recited examples and conditions. Moreover, all
statements herein reciting principles, aspects, and embodiments of the
invention, as well as specific examples thereof, are intended to encompass
equivalents thereof.
CLAIMS:
1. An active antenna array for a mobile telecommunications network,
comprising a plurality of radio elements (4), each including a transmit
and/or a receive path (8, 10) coupled to a respective antenna element
(12), and including comparison means (100) for comparing phase and/or
amplitude of transmitted or received signals with reference values in
order to adjust the characteristics of the antenna beam, and including a
feed arrangement for supplying reference signals of amplitude and/or
phase, the feed arrangement including a waveguide (50) of a
predetermined length, which is coupled to a reference signal source (54),
and which is terminated at one end (52) in order to set up a standing
wave system along its length, and a plurality of coupling points (56) at
predetermined points along the length of the waveguide, which are each
coupled to a said comparison means of a respective said radio element.
2. An array as claimed in claim 1, wherein said waveguide comprises a
length of coaxial cable.
3. An array as claimed in claim 2, wherein said coupling points each
comprise a capacitive coupling port (66).
4. An array as claimed in claim 3, wherein each capacitive coupling port is
adjustable (68) in order to adjust the coupling coefficient with the central
conductor (70) of the coaxial cable.
5. An array as claimed in claim 2, 3 or 4, wherein the coaxial cable has a
dielectric filling which may be adjusted in characteristics to alter the
wavelength of radiation in the line.
6. An array as claimed in any preceding claim, wherein the coupling points
are spaced apart by a distance of equal to or less than 1, where is the
wavelength in free space of the reference signal.
7. An array as claimed in claim 6, wherein the coupling points are spaced
apart by a distance of about 0.9, where is the wavelength in free
space.
8. An array as claimed in any preceding claim, wherein the waveguide
comprises a plurality of waveguide sections (60) of predetermined length,
interconnected by releasable couplings (62, 64).
9. An array as claimed in any preceding claim, wherein each coupling point
is located at or adjacent a voltage maximum or minimum in the standing
wave system.
10.An array as claimed in any preceding claim, wherein the coupling points
are spaced from the terminated end by a distance d= (+/4), where 
is the wavelength in the waveguide.
11.An array as claimed in any preceding claim, wherein the terminated end
comprises a short circuit.
12.An array as claimed in any preceding claim, wherein each coupling point
is connected to a said comparison means by a length of waveguide
which is short in relation to the length of the first mentioned waveguide.
13.An array as claimed in any preceding claim, wherein the array is two
dimensional and including a further plurality of waveguides ( 1 14), each
as claimed in claim 1, wherein each waveguide of said further plurality
has an end which is not terminated coupled to a respective coupling
point ( 1 12) of said first-mentioned waveguide, said first mentioned
waveguide extending in a different direction to that of said further plurality
of waveguides.
14.An array as claimed in any preceding claim, wherein the feed
arrangement includes a second waveguide of a predetermined length
which is terminated at one end in order to set up a standing wave system
along its length, and a plurality of coupling points at predetermined points
along the length of the waveguide, which are each coupled to a said
comparison means of a respective said radio element, wherein the
waves in the first and second waveguides have predetermined time
phase difference.
15. An array as claimed in any preceding claim, wherein the coupling points
of the waveguide are symmetrically arranged about the mid-point of the
length of the waveguide.