Inverse beamformer for inverting the action of existing beamformer in communication system

Abstract

Beamforming system and method which includes an inverse beamformer with first ports and second ports, a beamformer with beam ports and antenna ports, and antennas. The inverse beamformer may receive transmit signals provided to first ports, generates intermediate signals based on the transmit signals, and output the intermediate signals from the second ports. The beamformer's beam ports may couple to the second ports and the antenna ports, and may receive the intermediate signals at the beam ports, generate antenna signals based on the intermediate signals, and output the antenna signals at the antenna ports. The antennas may each receive an antenna signal from an antenna port, and transmit the antenna signal into space. The inverse beamformer may generate the intermediate signals in a manner that is mathematically inverse to the beamformer's generation of the antenna signals so that the antenna signals approximate the transmit signals up to respective phase shifts.

Description

FIELD OF THE INVENTION

The present invention relates generally to antennas for wireless data communications networks, and more specifically to beamforming antenna systems.

BACKGROUND OF THE INVENTION

Modern wireless communications systems place great demands on the antennas used to transmit and receive signals, especially at cellular wireless base stations. Antennas are required to produce a carefully tailored radiation pattern with a defined beamwidth in azimuth, so that, for example, the wireless cellular coverage area has a controlled overlap with the coverage area of other antennas.

In addition to a defined azimuth beam, such antennas are also required to produce a precisely defined beam pattern in elevation; in fact the elevation beam is generally required to be narrower than the width of the azimuth beam.

It is conventional to construct such antennas as an array of antenna elements so as to form the required beam patterns. Such arrays require a feed network to split signals for transmission into components with the correct phase relationship to drive the antenna elements; when receiving, the feed network doubles as a combiner.

An array consisting of a single vertical column of antenna elements is commonly used as a building block at cellular radio base stations. Such a column antenna can be designed to produce the required narrow elevation beam, and will typically be designed to give azimuth coverage of a sector in a cellular wireless network. In a simple configuration, three such column antennas are deployed at a base station to give coverage to three sectors; this is a form of a spatial division multiple access (SDMA) system, in which the capacity of a cellular wireless system is enhanced by enabling a given frequency band to be used substantially independently by wireless links which are spatially separated.

FIG. 1 illustrates a conventional tri-cellular deployment of base stations. A number of cell sites 1a . . . 1g are deployed to give wireless coverage over a given area. It can be seen that there are three radiation beams roughly equally spaced in azimuth angle at each cell site (for example, in the case of cell site 1a, there are three radiation beams 3a, 3b, 3c). Further capacity increases can be achieved by sub-dividing the azimuth plane more finely in angle, for example to form a hex-sectored plane, as shown in FIG. 2 (in the case of cell site 1a there are six hex-sector radiation beams 5a . . . 5f).

The required azimuth beam patterns for a system such as that illustrated by FIG. 2 can be implemented by the use of multiple column antennas in combination with a beamformer. The beamformer couples together the column antennas in the appropriate amplitude and phase relationship to give the required beam patterns. Such a beamformer will typically be a passive device that may be used for both transmission and reception of signals. Typically, in such a system, the column antennas are referred to as azimuth antenna elements or simply antenna elements.

FIG. 3 illustrates a system in which four azimuth antenna elements 7a . . . 7d are combined to give two beam outputs at ports 11a, 11d for a case in which two beams are required per sector. Each beam may be connected to a respective radio transceiver 27a, 27b, which will typically be connected to a telecommunications network 29 such as the public switched telecommunications network (PSTN). It is known, as shown by FIG. 4, to implement the beamformer 8 by the use of a Butler Matrix 14. The beamformer 8 has 4 antenna element ports 9a . . . 9d, typically connected to an array of antenna elements (as illustrated in FIG. 3 and indicated by reference numerals 7a . . . 7d). It is known to combine pairs of beam ports of the Butler Matrix 20a, 20c and 20b, 20d using 3 dB hybrid couplers 13a, 13b and phase shifters 16a, 16b to produce two beams at beam ports 11a, 11d. However, a beamformer as illustrated by FIGS. 3 and 4 suffers from complexity, involving the use of 6 hybrid couplers and four phase shifters. It should be noted that a beamformer as illustrated by FIGS. 3 and 4 forms beams by combining the contributions from each of the antenna element ports 9a . . . 9d in such a way that signals from an antenna element port cannot be constructively combined at a beam port while signals from another antenna element port are destructively combined at that beam port. Accordingly, it is not possible to access at a beam port the signals received by an antenna element but not those from another antenna element.

It may be convenient to locate the beamformer 8 close to the antenna elements 7a . . . 7d, which will typically be located on an antenna tower. It may also be advantageous in terms of cost, size and performance to integrate the beamformer with the antenna elements, contained within the same enclosure.

However, the integration of a beamformer with its associated antenna elements may present disadvantages in terms of potential upgrade strategies if such strategies require access to the individual antenna elements. In order to access the individual antenna elements, an operator would require to climb the tower and modify or replace the beamformer; in the case of an integrated system this may not be possible, necessitating the replacement of the integrated unit 6. The replacement and re-alignment of antenna elements may be costly; accordingly the lack of an economical upgrade path may limit the deployment of an otherwise attractive integrated beamformer and antenna system.

It is an object of the present invention to provide methods and apparatus which addresses these disadvantages.

SUMMARY OF THE INVENTION

In accordance with aspects of the present invention, there is provided methods and systems according to the appended claims.

More specifically, in one aspect there is provided a method of receiving signals from a first antenna element, said first antenna element providing input to a beamformer, the beamformer being arranged to receive input from at least one other antenna element and being arranged to generate at least two beams as output therefrom, the method comprising the steps of:

combining said at least two beam outputs at a connecting port such that said signals from said first antenna element are constructively combined at the connecting port; and

combining said at least two beam outputs at the connecting port such that signals from antenna elements other than the first antenna element providing input to the beamformer are destructively combined at the connecting port; and

configuring the connecting port so as to provide access to individual said signals received by said antenna elements.

The connecting port provides access to signals at an individual antenna element without the need to remove the beamformer, which may for example be beneficial in situations where access to the beamformer is difficult or costly or where the beamformer is physically integrated with the antenna elements.

Constructively combining signals is a process of combining signals substantially in phase so that the magnitude of the resultant signal is maximised. Destructively combining signals is a process of combining signals in such a way that they cancel, so that the magnitude of the resultant signal is minimised.

Preferably, the beams formed by the beamformer are orthogonal beams. The benefit of forming orthogonal beams is that the signal loss between the antenna element and the connecting point is minimised.

In one arrangement the beamformer can be arranged to form three output beams from a combination of input from three antenna elements, at least one of the antenna elements being said first antenna element, the method further comprising combining said three output beams at the connecting port. In such an arrangement the beamformer can be configured according to the following steps, so as to achieve the aforementioned constructive and destructive combining of signals at the connecting port:

combining signals received by the first antenna element with signals received by a third of said three antenna elements, in which combining comprises in-phase combining, to provide a third output beam;

combining signals received by the first antenna element with signals received by the third antenna element, in which said combining comprises anti-phase combining, so as to provide a first intermediate signal;

combining signals received by a second of the three antenna elements with the first intermediate signal such that the intermediate signal is at minus ninety degrees phase to the signals received by the second antenna element, so as to provide the first output beam; and

combining signals received by the second antenna element with the first intermediate signal such that the intermediate signal is at ninety degrees phase to the signals received by the second antenna element, to provide the second output beam.

In a yet further arrangement, the beamformer can be arranged to form four output beams from a combination of input from four antenna elements, at least one of the antenna elements being the first antenna element, the method further comprising combining the four output beams at the connecting port.

Conveniently, a further beamformer can be used to combine the beams output from the beamformer. The further beamformer may be connected as an inverse beamformer, that is to say the input beam ports of the further beamformer are connected to the output beam ports of the beamformer, in which case an individual input port of the further beamformer provides the connecting port. The benefit of using a further beamformer is that a similar technology may be employed to that used to implement the beamformer, thereby potentially giving cost savings in design and construction.

Preferably, the further beamformer provides a further connecting port, thereby ensuring that a radio system that requires access to more than one individual antenna element may be provided with such access via a suitable further connecting port. Examples of such radio systems are multiple in multiple out (MIMO) systems, diversity combination and adaptive beamforming systems.

Conveniently the connecting port, or in the case of arrangements that include the further beamformer, the further connecting port will be connected to a receiver.

Advantageously, the further beamformer may be implemented by an array of weighting elements, which weight, in amplitude and phase, the beam outputs from the beamformer and combine the weighted components at a connecting port. The advantage of implementing the further beamformer by an array of weighting elements is that the implementation may be economical, for example the weighting elements and combination may be implemented in digital signal processing.

The beamformer arrangement can also be arranged to transmit signals from antenna elements; in one arrangement, this involves transmitting signals from a first antenna element, said first antenna element receiving input from a beamformer, the bcamformer being arranged to input signals to at least one other antenna element and comprising a set of beam ports for receiving signals to be transmitted via said antenna elements, the set of beam ports being connected to a connecting port external to said beamformer, the method comprising:

coupling signals to said beam ports of the beamformer such that signals received at the connecting port are constructively combined at said first antenna element; and

coupling signals to said beam ports such that signals received at the connecting port are destructively combined at antenna elements other than said first antenna element,

thereby enabling a transmitter connected to said connecting port to transmit from said first antenna element and not to transmit from the other antenna elements.

Further features and advantages of the invention will become apparent from the following description of preferred embodiments of the invention, given by way of example only, which is made with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram showing a conventional tri-cellular cellular wireless deployment;

FIG. 2 is a schematic diagram showing a conventional hex-sectored cellular wireless deployment;

FIG. 3 is a schematic diagram showing a conventional beamforming antenna system;

FIG. 4 is a schematic diagram showing a conventional beamformer utilising a Butler Matrix and combined beam ports;

FIG. 5 is a schematic diagram showing a three column beamformer providing two output beams according to an embodiment of the invention;

FIG. 6 is a schematic diagram showing an embodiment of the invention in a MIMO system;

FIG. 7 is a schematic diagram showing an embodiment of the invention in a four-branch MIMO system;

FIG. 8 is a schematic diagram showing an embodiment of the invention in an adaptive SDMA system;

FIG. 9 is a schematic diagram showing operation of a four port coupler with an input to port A;

FIG. 10 is a schematic diagram showing operation of a four port coupler with an input to port D;

FIG. 11 is a schematic diagram showing operation of a four port coupler with an input to port B;

FIG. 12 is a schematic diagram showing operation of a four port coupler with an input to port C;

FIG. 13 is a schematic diagram showing a beamformer according to an embodiment of the invention;

FIG. 14 is a schematic diagram showing a beamformer according to an embodiment of the present invention and an inverse beamformer according to an embodiment of the present invention;

FIG. 15 is a schematic diagram showing a beamformer according to an embodiment of the present invention and an inverse beamformer according to an embodiment of the present invention implemented as an array of weighting networks;

FIG. 16 is a schematic diagram showing an embodiment of the present invention comprising a four element antenna array and a MIMO transceiver;

FIG. 17 is a schematic diagram showing an embodiment of the present invention comprising a four element antenna array and an adaptive SDMA beamformer;

FIG. 18 is a schematic diagram showing an embodiment of the present invention comprising a four element antenna array and an inverse beamformer; and

FIG. 19 is a schematic diagram showing an embodiment of the present invention comprising a four element antenna array and an inverse beamformer implemented as an array of weighting network.

DETAILED DESCRIPTION OF THE INVENTION

In general, the present invention is directed to methods and apparatus that enhance the capacity of wireless communications between a base station and remote stations. The invention will be described in the context of a cellular wireless system, but it is to be understood that this example is chosen for illustration only and that other applications of the invention are possible.

FIG. 5 illustrates a first embodiment of the invention, which relates to a beamformer that produces two beams from a three element antenna array. Three antenna elements 7a, 7b and 7c, which may be typically vertical columns of antenna elements, are connected to beamformer 8, which may typically be integrated with the antenna elements and may be within the same enclosure. In a one deployment, the antenna elements 7a, 7b, and 7c and beamformer 8 will be sited on a tower at a cellular radio cell site. Three such integrated units would be deployed at a cell site, so that each may give coverage to an approximately 120 degree sector. Other arrangements are of course possible: indeed there is no inherent requirement for the coverage of a respective unit to be 120 degrees. As shown in the Figure, three outputs are provided from the beamformer 8, but only two of the beam ports 11a, 11c are connected to a respective radio transceiver 27a, 27b. The radio transceivers may be integrated into a cellular radio base station, and are typically used to enable the capacity of the cellular system to be increased by the simultaneous use of the same radio resource in terms of frequency and time within the area of coverage provided by the two beams generated by the beamformer 8. As per the prior art arrangements described above, the radio transceivers can be connected to a telecommunications network such as the public switched telecommunications network (PSTN) 29.

Typically, the internal structure of the beamformer will produce the same number of beams as there are antenna element ports; if fewer beams are required for use, then the unwanted beam ports (11b) may be simply terminated with an appropriately matched impedance. Conventionally unwanted beam ports are terminated internally to the beamformer, as illustrated in FIG. 4 by the terminations 15a and 15b which are connected to the unwanted ports 11b and 11c, since there is no need to have external access to them and the provision of external connectors would add unnecessary cost. However, in the embodiment of the invention illustrated in FIG. 5, the third beam port 1 lb is preferably provided with a connector external to the enclosure 6, so that the port may be simply terminated with an appropriate impedance for initial deployment, while enabling future upgrade options as will be discussed with reference to later figures.

It should be noted that the system shown in FIG. 5 shows only one polarisation channel; typically the system would be duplicated in parallel using antenna elements of an orthogonal polarisation. There is thus the capability to provide a two-branch MIMO system, using the two orthogonal polarisation channels. That is to say there are two transceivers at the base station on orthogonal polarisations and preferably two transceivers at a remote unit; the two transceivers would be used to enable orthogonal data paths to provide extra payload capacity. However, with the system as shown in FIG. 5, it is not possible to provide more than two MIMO transceivers at a base station.

FIG. 6 illustrates the use of the beamformer 8 of FIG. 5 in a MIMO radio system that provides two branches per polarisation per sector. Whilst for 2-branch MIMO transceivers the polarisations relating to one column of elements would typically be used, FIG. 6 shows use of two end columns, each of which has a common polarisation associated therewith; such an arrangement is depicted for the purposes of exemplifying an embodiment of the invention. The arrangement additionally includes an inverse beamformer 12 connected to the beam ports 11a, 11b and 11c of the beamformer 8. In this arrangement the beam port 11b of beamformer 8 that was terminated in the system of FIG. 5 has its termination removed and is connected to the inverse beamformer. The inverse beamformer 12 receives the beams formed by the beamformer 8 and transforms them to signals that are output via individual element ports 25a, 25c. Individual element ports may also be referred to as connecting ports, or connection points.

On receive, the system operates as follows: signals are received by antenna elements 7a . . . 7c and transformed to beams that are output via beam ports 11a, 11b, 11c, at the output of the bcamformer 8, each beam corresponding to a radiation pattern received by the antenna array 7a . . . 7c. The beam outputs are then transformed by the inverse beamformer 12 to represent the signals originally received by the antenna elements 7a . . . 7c. In the case of the system of FIG. 6, two outputs 25a, 25c of the inverse beamformer 12 are made available externally, carrying signals corresponding to the signals received by two of the antenna elements. In the case illustrated, the two outputs correspond to the end elements 7a and 7c of the array. The output of the inverse bcamformer that is not made available externally may be terminated internally. It should be noted that the relative phases of the signals at the output of the inverse beamformer 12 may not be the same as the relative phases of the signals received by the antenna elements 7a, 7b, 7c.

For many applications (such as MIMO) the relative phase of the signals on these outputs is not important; however, if the relative phases on the individual element ports are required to correspond with the relative phases of the signals at the antenna array, then this can be achieved by the addition of phase shifting elements in line with the individual element ports 25a, 25c. The system has the characteristic that signals received by an individual antenna element 7a are connected to a respective individual element output of the inverse beamformer 25a, whereas signal received by the other elements 7b, 7c in the array are not connected to that individual element output 25a. The combination of the beamformer 8 and the inverse beamformer 12 thus in effect gives access to signals received by the antenna elements 7a . . . 7c without the need to remove the beamformer 8 from the system. This is particularly advantageous in arrangements where the beamformer 8 and the antenna elements 7a . . . 7c are integrated into a unit such that removal of the beamformer unit may not be feasible.

On transmit, the system operates in the reverse manner of the operation on receive. Signals connected to individual element ports 25a, 25c are transformed to give inputs to the beam ports 11a . . . 11c of the beamformer 8 such that the outputs of the beamformer to antenna elements 7a . . . 7c correspond to the signals input to the respective individual element ports 25a, 25c. That is to say that a signal input to an individual element port 25a of the inverse beamformer will be transmitted from the respective antenna element 7a and not from the other antenna elements 7b, 7c. As in the case of receive, the relative phase of the signals transmitted from the antenna elements may not be the same as the relative phases between the signals connected to the respective individual elements ports of the inverse beamformer. If the relative phases on the individual element ports are required to correspond with the relative phases of the signals at the antenna array, then this can be achieved by the addition of phase shifting elements in line with the individual element ports 25a, 25c as was mentioned for the case of receive.

FIG. 7 shows how the system of FIG. 6 may be applied to a system receiving orthogonal polarisations A and B to provide a MIMO system with four branches, that is to say potentially four orthogonal signal paths. Signals received by the antenna elements 7a . . . 7c on polarisation A are connected to the antenna ports of a beamformer 8a and signals received by the antenna elements 7a . . . 7c on polarisation B are connected to the antenna ports of a second beamformer 8b. Examples of orthogonal states of polarisations include pairs of linear states such as nominally vertical and horizontal, or nominally +/−45 degrees, and pairs of circular states such as left and right hand polarisation. Similar to the system of FIG. 6, the beamformer transforms an array of antenna element signals at each polarisation to an array of beams at the respective polarisation, and the array of beams at the respective polarisations are transformed by the inverse beamformers to an array of individual element signals at the respective polarisations. In the system illustrated in FIG. 7, it can be seen that four ports 25a, 25c, 25d, 25f are available to a MIMO transceiver 17, corresponding to polarisation A and polarisation B on each of two antenna elements 7a . . . 7c. As a variant, all or a sub-set of the individual element ports on one or both polarisations could be used to provide a port for a MIMO transceiver 17.

FIG. 8 shows an embodiment of the invention in which a beamformer 8 and an inverse beamformer 12 are used in conjunction with an adaptive SDMA beamformer 18 connected to transceivers 28a . . . .28n. As in the example of FIG. 6, the inverse beamformer provides individual element ports corresponding to the signals on the antenna elements 7a . . . 7c. The adaptive SDMA beamformer 18 functions in a conventional manner so as to weight each individual element signal in amplitude and phase, where the weighting may be typically applied in Cartesian format as Inphase and Quadrature components, in order to form appropriate beams to provide coverage enabling base station transceivers 28a . . . 28n to communicate with respective transceivers. The transceivers may be user equipment, sharing between beams the same radio resource in terms of frequency and time. Typically, the function of the inverse beamformer 12 and the adaptive SDMA beamformer 18 may be implemented using digital signal processing.

FIGS. 9 to 12 illustrate the operation of a four port hybrid coupler, two of which are shown in FIG. 4 as parts 13a, 13b and which may also be simply referred to as a four port coupler, and which is also commonly referred to as a 3 dB coupler or a 90 degree hybrid coupler. Typically, the coupler may be constructed from two parallel coupled transmission lines, but other constructions of the coupler are well known in the art. As is known in the art, a four port coupler can be used in beamformers and inverse beamformers to combine signals, and in particular enable signals input thereto to be combined in a configurable manner.

The operation of a four port hybrid coupler comprising two pairs of ports according to an embodiment of the invention will now be described; referring to FIG. 9, ports A and B form a first pair and ports C and D form a second pair. The principle of operation is that signals applied to a port A of the first pair are split equally in power between two other ports, B, C, one from each pair, and none of the signals from this port A are transmitted to a fourth port D (this being the remaining port of the first pair). Thus signals received by a given port A of a pair of ports are output, equally in power, via the other port B of the pair and via a port C of another pair of ports.

The signals on the two ports B, C between which the power is split differ in phase by 90 degrees; the port C which is unpaired with the input port A carries signals with a −90 degree phase difference to those received by the paired port B. In the convention adopted in FIG. 9, the two ports B, C from which power is output are shown on the opposite side of the coupler 13 from the input A, but this may not correspond with the physical arrangement in a physical device.

FIGS. 10, 11 and 12 illustrate the operation of the four port coupler of FIG. 9 when signals are input to each of the other three ports B, C, D. The principle of operation as outlined above applies, so that in FIG. 10 signals are applied to port D, and are split between ports C and B with signals at port B being at −90 degrees to those on port C. No signals are transmitted to port A. FIG. 11 illustrates the case in which signals are input to port B, and are split between ports A and D with signals at port D being at −90 degrees to those on port A. No signals are transmitted to port C. In FIG. 12 signals are input to port C, and are split between ports A and D with signals at port A being at −90 degrees to those on port D, while no signals are transmitted to port B. It should be understood that the operation as described assumes that coupler ports are matched to an appropriate terminating impedance.

If the four port coupler 13 is used as a combiner, then signals present on each output of the coupler may be obtained by vector addition; if signals that are 180 degrees out of phase with each other but that are otherwise identical arrive at a given output port, then the signals will cancel so that the port appears isolated. Accordingly the signal power will be transmitted to another port at which the signals do not cancel.

FIG. 13 shows a three element beamformer 8 according to an embodiment of the invention. The beamformer 8 uses two four port couplers 13a, 13b and a −90 degree phase shifter 16a. The −90 degree phase shifter 16a may, for example, be implemented by a physical delay such as a length of printed track. It can be seen that in comparison with the conventional four element beamformer design of FIG. 4, the three element beamformer 8 of FIG. 13 involves fewer components and is therefore potentially smaller and cheaper to implement. In addition, since the addition of any single component introduces a loss of signal due to non-ideal implementation, a system with fewer components is advantageous since it can help reduce signal loss. It is important to minimise signal loss in a beamformer design, since, on receive, any loss of signal before the first amplifier stage impacts the signal to noise ratio of the system and on transmission any loss of signal is wasteful of expensive amplifier resource.

The operation of the beamformer 8 of FIG. 13 will now be described in operation for transmission of signals (it will be appreciated that the beamformer 8 is bi-directional and so can also be used for receive and indeed for both receive and transmission). The beamfoiuier 8 has three beam ports 11a, 11b and 11c. In an application such as that illustrated by FIG. 5 requiring the formation of only two beams, the third beam output as indicated by reference numeral 11c may be terminated by an appropriate impedance 15.

The flow of signals from the first beam port as indicated by reference numeral 11a will now be described.

Signals entering at an amplitude of 1 into port A of the second four port coupler 13b are split into a component designated as a reference phase of 0 degrees at port B and a component at −90 degree phase at port C. The signal is split equally in power, that is to say the amplitude of each is half of the square root of two. Signals from port B of the second four port coupler 13b are connected to the second antenna element port 9b, and have an amplitude of half of the square root of two and at a phase of 0 degrees. Signals from port C of the second four port coupler 13b are connected to port A of the first four port coupler 13a and have an amplitude of half of the square root of two and a phase of −90 degrees. The signals leave port B of the first four port coupler 13a and have an amplitude of one half and a phase of −90 degrees; these signals are connected to the first antenna element port 9a. Signals leave port C of the first four port coupler 13a and have an amplitude of one half and a phase of −180 degrees; these signals arc phase shifted by a further −90 degrees by the phase shifter 16a and are then connected to the third antenna element port 9c, at an amplitude of one half and a relative phase of −270 degrees, that is to say equivalent to +90 degrees relative to signals entering port A of the first four-port coupler 13a.

As a result, the antenna array 7a . . . 7c to which the antenna ports 9a . . . 9c are connected is excited as follows: the phase on signals on the first 7a, second 7b and third 7c antenna elements respectively is −90, 0, +90 degrees and the amplitude is 0.5, 0.707, 0.5 respectively. If the antenna elements 7a . . . 7c are half a wavelength apart in the azimuth plane at the frequency of operation of the antennas, then the excitation of the antenna elements results in a beam at −30 degrees from boresight (that is, closer in angle to the line from the centre of the array to the first element than to the line from the centre of the array to the third element), where boresight is an angle perpendicular in azimuth to the array 7a . . . 7c. On receive, signals will be received from a similar beam.

The flow of signals from the second beam port as indicated by reference numeral 11b will now be described.

Signals entering at an amplitude of I into port D of the second four port coupler 13b are split into a component designated as a reference phase of 0 degrees at port C and a component at −90 degree phase at port B. The signal is split equally in power, that is to say the amplitude of each is half of the square root of two. Signals from port B of the second four port coupler 13b are connected to the second antenna element port 9b, at an amplitude of half of the square root of two and at a phase of −90 degrees. Signals from port C of the second four port coupler 13b are connected to port A of the first four port coupler 13a with an amplitude of half of the square root of two and a phase of 0 degrees. The signals leave port B of the first four port coupler 13a with an amplitude of one half and a phase of 0 degrees; this signal is connected to the first antenna element port 9a. Signals leave port C of the first four port coupler 13a with an amplitude of one half and a phase of −90 degrees; this signal is phase shifted by a further −90 degrees by the phase shifter 16a and then connected to the third antenna element port 9c, at an amplitude of one half and a relative phase of −180 degrees.

As a result, the antenna array 7a . . . 7c is excited as follows: the phase on signals on the first 7a, second 7b and third 7c antenna elements respectively is 0, −90, −180 degrees and the amplitude is 0.5, 0.707, 0.5 respectively. If the antenna elements are half a wavelength apart in the azimuth plane at the frequency of operation of the antennas, then the excitation of the antenna elements results in a beam at 30 degrees from boresight. On receive, signals will be received from a similar beam.

Hence it can be seen that the signals output via the first and second beam ports 11a, 11b form a pair of beams, one at −30 degrees and the other at +30 degrees to boresight. It will be appreciated that this pair of beams is well suited to give coverage within a 120 degree cellular base station sector such as in the system illustrated by FIG. 2; for example the two beams could provide beams 5a, 5b in FIG. 2.

The amplitude taper in the excitation across the array 7a . . . 7c, in which the centre element 7b has a higher amplitude than that of the end elements 7a, 7c, is beneficial in reducing sidelobe levels of the beams compared to the beam that would be formed if the elements were excited with equal amplitudes. Lower sidelobe levels in turn are beneficial in reducing interference between beams and therefore improving the capacity of a cellular wireless system.

It can be shown that the beams produced by the beamformer of FIG. 13 are orthogonal. Orthogonality of antenna beams is a well known concept in the art, and has the effect that a correlation between the beams across azimuth angles is zero. A result of the orthogonality of the beams is that the beamformer is ideally loss-less.

It should be noted that the amplitudes set out in the examples above are in arbitrary units and do not take account of implementation losses. Also, the phases quoted do not account for transmission delays through components and between components (except where specifically mentioned). A practical implementation would typically be laid out so that transmission paths are equalised in terms of delay from each beam port to each antenna port, as far as is possible. Techniques for the equalisation of transmission delays are well known in the art; for example, lengths of transmission line may be designed with the required delay characteristics.

The beamformer design need not necessarily be passive, as shown; instead, amplifiers may be inserted between stages if signal gain is required.

The third beam port 11c of the beamformer 8 illustrated by FIG. 13 produces a third beam, which can be shown to be orthogonal to the first and second beams. This beam has three lobes and so does not constitute a conventional beam that would be used for a cellular wireless system, since it is conventional to use beams with single lobes. In applications that require only two single lobe beams, such as that illustrated in FIG. 2, this third beam may be unused and the beam port 11 c may simply be terminated with an appropriate impedance 15.

The flow of signals from the third beam port as indicated by reference numeral 11c will now be described.

Signals entering at an amplitude of 1 into port D of the first four port coupler 13a leave port B of the first four port coupler 13a with an amplitude of half of the square root of two and a phase of −90 degrees; this signal is connected to the first antenna element port 9a. Signals leave port C of the first four port coupler 13a with an amplitude of half of the square root of two and a phase of 0 degrees; this signal is phase shifted by a further −90 degrees by the phase shifter 16a and then connected to the third antenna element port 9c, at an amplitude of half of the square root of two and a relative phase of −90 degrees.

As a result, the antenna array 7a . . . 7c is excited as follows: the phase on signals on the first 7a, second 7b and third 7c antenna elements respectively is −90, no signal, −90 degrees and the amplitude is 0.707, 0, 0.707 respectively. If the antenna elements are half a wavelength apart in the azimuth plane at the frequency of operation of the antennas, then the excitation of the antenna elements results in a beam at boresight, with additional lobes either side of boresight. On receive, signals will be received from a similar beam.

FIG. 14 shows a beamformer 8 used in conjunction with an inverse beamformer 12. As mentioned above, when the beamformer 8 is combined with the inverse beamformer 12, signals received on antenna element ports 9a . . . 9c of the beamformer 8 are transmitted to respective individual element ports 25a . . . 25c of the inverse beamformer, while signals received on individual element ports 25a . . . 25c of the inverse beamformer are transmitted to respective antenna element ports 9a . . . 9c of the beamformer 8.

The operation of the inverse beamformer 12, having two four port couplers 13a, 13b, in conjunction with the beamformer 8 configured as described above with reference to FIG. 13 will now be described.

Signals entering the first individual element port 25a of the inverse beamformer 12 are connected to port A of the fourth four port coupler 13d. Signals leave port B of the fourth coupler 13d at a phase of 0 degrees and signals leave port C of the fourth coupler 13d at a phase of −90 degrees.

Signals from port B of the fourth coupler 13d are connected to port D of the third four port coupler 13c. Signals leave port B of the third coupler 13c at a phase of −90 degrees and port C at a phase of 0 degrees, which then undergo a further −180 degree phase shift in the phase shifter 16b so that the signals are presented to port D of the second four port coupler 13b at −180 degrees phase, which is equivalent to 180 degrees phase, and leave port B of the second four port coupler 13b at 90 degrees phase and port C of the second four port coupler 13b at 180 degrees phase.

Signals entering the second four port coupler 13b from port A leave port B at −90 degrees and leave port C of the second four port coupler 13b at −180 degrees, that is equivalent to 180 degrees. Accordingly, the signals at port B of the second four port coupler 13b arriving from ports A and D of the second four port coupler 13b are equal amplitude and in anti-phase and cancel; no signals are thus transmitted to the second antenna element 9b. However, the signals at port C of the second port coupler 13b arriving from ports A and D of the second coupler 13b are in phase and reinforce at a phase of 180 degrees.

The final signal combination occurs in the first four port coupler 13a; it will be appreciated that the signals arriving at port A and port D of this first four port coupler 13a are at the same amplitude as each other. Signals arriving at port A of the first four port coupler 13a, at 180 degrees, leave port B 180 degrees and leave port C at 90 degrees. Signals arriving at port D of the first four port coupler 13a, at −90 degrees, leave port B at 180 degrees and leave port C at −90 degrees. Accordingly, the signals at port B of the first four port coupler 13a arriving from ports A and D are in phase and reinforce; the signals are thus transmitted to the first antenna element 9a. The signals at port C of the first four port coupler 13a arriving from ports A and D are equal amplitude and in anti-phase and cancel, so that no signals are transmitted to the third antenna port 9c.

Hence it can be seen that the first, second, third and fourth couplers 13a . . . 13d are arranged such that signals transmitted into the first individual element port 25a of the inverse beamformer 12 are transmitted to the first antenna element port 9a of the beamformer 8 and not to the other antenna element ports 9b, 9c.

Similarly, signals from each of the individual element ports 25a . . . 25c arrive only at the respective antenna element ports 9a . . . 9c, while signals received at each of the antenna element ports 9a . . . 9c arrive only at the respective individual element ports 25a . . . 25c.

FIG. 15 shows a beamformer 8 and an inverse beamformer 12, in which the inverse beamformer 12 is implemented by an array of weighting elements 20a . . . 20i. Each weighting element functions conventionally to weight signals in amplitude and phase, where the weighting may be typically applied in Cartesian format as Inphase and Quadrature components. The implementation of FIG. 15 can achieve the same functional result as that shown in FIG. 14 by the application of appropriate weighting element values to weighting elements 20a . . . 20i. Each individual element port 25a . . . 25c is connected to an array of weighting elements, specifically a subset of the elements 20a . . . 20i, each of which is connected to a beam port 24a . . . 24c. For example, the first individual element port 25a is connected to a first subset of weighting elements W₁₁ 20a, W₁₂ 20b and W₁₃. 20c. Weighting element W₁₁ 20a, is connected to the first beam port 24a, weighting element W₁₂ 20b is connected to the second beam port 24b and weighting element W₁₃ 20c is connected to the third beam port 24c.

The values of the weighting elements are arranged such that on receive, signals originating at the antenna element port 9a corresponding with the individual element port 25a are combined constructively at the respective individual element port 25a, whereas signals originating at the other antenna element ports 9b, 9c are combined destructively at the respective individual element port 25b, 25c.

On transmit, the same weighting values may be used as on receive to achieve the desired connection between an antenna element port 9a and a respective individual element port 25a.

The weighting network may be implemented via a physical device, which may be bi-directional, allowing the use of the inverse beamformer for transmit, receive, or for both. Alternatively, the weighting network may be implemented via digital signal processing components.

An example of the weighting values that could be used with the beamformer 8 design as shown in FIG. 15 is as follows. It should be noted that the amplitude values are in arbitrary units proportional to signal voltage and that the phase values in degrees are relative to other values appropriate to the individual element 25a . . . 25c. That is to say, W₁₁ 20a, W₁₂ 20b and W₁₃ 20c should have the phase values listed, relative to each other. However, the relative phase between different subsets of weighting elements is not significant. For example, the phase relationship between W₁₁ 20a and W₂₁ 20d is not significant.

	Amplitude	Phase
	W11	0.5	90
	W12	0.5	0
	W13	0.707	90
	W21	0.707	0
	W22	0.707	90
	W23	Not driven
	W31	0.5	180
	W32	0.5	90
	W33	0.707	0

FIG. 16 illustrates an embodiment of the invention using a four element beamformer 8 in a MIMO radio system that provides two branches per polarisation per sector. The beamformer 8 may be embodied as shown in FIG. 4. An inverse beamformer 12 is provided that is connected to the beam ports 11a, 11b, 11c and 11d of the beamformer 8. The beam ports 11b and 11d of beamformer 8 that were terminated in the system of FIG. 3 and FIG. 4 have had their terminations removed and are used to connect to the inverse beamformer. The inverse beamformer 12 receives the beams formed by the beamformer 8 and transforms them to signals that are output via individual element ports 25a, 25d, respectively corresponding to the antenna elements 7a and 7d. The two individual element ports are connected to a MIMO transceiver 17 which is itself connected to a telecommunications network 29. FIG. 16 shows the use of a single polarisation for clarity; however the system would typically employ dual polarisations as illustrated in FIG. 7 for the three element case, so that the MIMO receiver 17 would be connected to two element ports for two polarisation states giving access to 4 independent channels.

FIG. 17 illustrates an embodiment of the invention using a four element beamformer with an adaptive SDMA beamformer, which operates in a similar manner to the three element beamformer of FIG. 8.

FIG. 18 shows a further embodiment of the invention, in which the inverse beamformer 12 is embodied as a second beamformer 8b; this second beamformer 8b can be embodied as the beamformer 8a that is connected to the antenna array 7a . . . 7d, with the addition of a −180 degree phase shifter 16a in series with the third beam port 11g and another −180 degree phase shifter 16b in series with the fourth beam port 11h of the second beamformer 8b. Thus, individual element ports 25a . . . 25d of the inverse beamformer 12 correspond to the respective antenna element ports 9e . . . 9h of the second beamformer 8b, and in the embodiment shown, the first and second beam ports 24a and 24b correspond to the first and second beam ports 11e and 11f respectively of the second beamformer 8b. The third and fourth beam ports 24c and 24d of the inverse beamformer 12 are connected to the phase shifters 16a, 16b that are connected to third and fourth beam ports 11g and 11h of the second beamformer 8b respectively.

It should be noted that a phase shift of 180 degrees has the same meaning in this context as a phase shift of −180 degrees. It should also be noted that the phase shifters could alternatively be placed in series with the first and second beam ports 11e, 11f of the second beamformer 8b.

FIG. 19 shows a beamformer 8 and an inverse beamformer 12, in which the inverse beamformer is implemented by an array of weighting elements 20a . . . 20p for a four antenna element arrangement; this is implemented using similar principles to the three element design described above with reference to FIG. 15.

The above embodiments are to be understood as illustrative examples of the invention. It is to be understood that any feature described in relation to any one embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the embodiments, or any combination of any other of the embodiments. Furthermore, equivalents and modifications not described above may also be employed without departing from the scope of the invention, which is defined in the accompanying claims.

Claims

1. A system comprising: an inverse beamformer including a first plurality of ports and a second plurality of ports, wherein the inverse beamformer is configured to: receive transmit signals provided respectively to the ports of the first plurality,generate intermediate signals based on the transmit signals; andoutput the intermediate signals respectively from the ports of the second plurality;a beamformer including a plurality of beam ports and a plurality of antenna ports, wherein each of the beam ports is coupled to a respective one of the ports of the second plurality, wherein the beamformer is configured to: receive the intermediate signals respectively at the beam ports;generate a plurality of antenna signals based on the intermediate signals; andoutput the antenna signals respectively at the antenna ports;a plurality of antennas, wherein each of the antennas is configured to receive a respective one of the antenna signals from a respective one of the antenna ports and to transmit the antenna signal into space, wherein the inverse beamformer is configured to generate the intermediate signals in a manner that is mathematically inverse to the beamformer's generation of the antenna signals so that the antenna signals approximate the respective transmit signals up to respective phase shifts.

2. The system of claim 1, further comprising a transmitter coupled to the first plurality of ports of the inverse beamformer, wherein the transmitter is configured to generate the transmit signals based on data received from a network, wherein the transmit signals together form part of a multiple-input multiple-output (MIMO) transmission.

3. The system of claim 1, wherein the beamformer and the antennas are configured as parts of an integrated unit.

4. The system of claim 3, wherein the antennas are vertical antenna arrays, wherein the integrated unit serves a sector of a cell in a cellular communication system.

5. The system of claim 1, wherein the beamformer and the antennas are situated on a tower, wherein the inverse beamformer is not on the tower.

6. The system of claim 1, wherein the inverse beamformer is configured to generate each of the intermediate signals as a corresponding weighted sum of the transmit signals.

7. The system of claim 1, wherein the inverse beamformer comprises a network of four-port hybrid couplers and phase shifters.

8. The system of claim 1, further comprising: a spatial division multiple access (SDMA) beamformer coupled to the first ports of the inverse beamformer.

9. A method comprising: generating a plurality of intermediate signals based on a plurality of transmit signals, wherein said generating the plurality of intermediate signals is performed by an inverse beamformer;generate a plurality of antenna signals based on the plurality of intermediate signals, wherein said generating the plurality of antenna signals is performed by a beamformer;transmitting the antennas signals into space using a respective plurality of antennas, wherein said generation of the intermediate signals is performed in a manner that is mathematically inverse to the said generation of the antenna signals so that the antenna signals approximate the respective transmit signals up to respective phase shifts.

10. The method of claim 9, wherein the inverse beamformer receives the transmit signals from one or more transmitters, wherein the transmit signals together form part of a multiple-input multiple-output (MIMO) transmission.

11. The method of claim 9, wherein the inverse beamformer receives the transmit signals from a spatial division multiple access (SDMA) beamformer.

12. The method of claim 9, wherein the beamformer and the antennas are configured as parts of an integrated unit.

13. The method of claim 12, wherein the antennas are vertical antenna arrays, wherein the integrated unit serves a sector of a cell in a cellular communication system.

14. The method of claim 9, wherein the beamformer and the antennas are situated on a tower, wherein the inverse beamformer is not on the tower.

15. A method for modifying a communication system including one or more transmitters, a beamformer and a plurality of antennas, wherein the beamformer is coupled to the plurality of antennas, the method comprising: exposing beam ports of the beamformer by decoupling transmit ports of the one or more transmitters from two or more of the beam ports and by removing one or more terminations from an additional one or more of the beam ports, wherein the one or more terminations are external to the beamformer;coupling the transmit ports of the one or more transmitters to respective element ports of an inverse beamformer;coupling first ports of the inverse beamformer respectively to the beam ports of the beamformer, wherein the inverse beamformer is configured to: receive transmit signals respectively through the element ports;generate intermediate signals based on the transmit signals;output the intermediate signals respectively through the first ports;wherein the beamformer is configured to: receive the intermediate signal respectively through the beam ports;generate antenna signals based on the intermediate signals; andoutput the antennas signals onto the respective antennas;wherein the inverse beamformer is configured to generate the intermediate signals in a manner that is mathematically inverse to said beamformer's generation of the antenna signals, so that the antenna signals approximate the respective transmit signals up to respective phase shifts.

16. The method of claim 15, wherein said exposing the beam ports, said coupling the transmit ports and said coupling the first ports are performed without disturbing alignment of the antennas.

17. The method of claim 15, wherein the beamformer and the antennas are situated on a tower, wherein said exposing the beam ports, said coupling the transmit ports and said coupling the first ports are performed without a person climbing the tower.

18. The method of claim 15, wherein the beamformer and the antennas are included as parts of an integrated unit.

19. The method of claim 15, wherein said modifying the communication system enables the one or more transmitters to perform MIMO transmission through the antennas.

20. The method of claim 15, further comprising: coupling a spatial division multiple access (SDMA) beamformer between the transmit ports of the one or more transmitters and the element ports of the inverse beamformer.