CN116686231A

CN116686231A - Establishing wireless communication in a beam forming system by selecting from a predetermined plurality of antenna weight vectors

Info

Publication number: CN116686231A
Application number: CN202180088602.5A
Authority: CN
Inventors: 奈杰尔·乔纳森·理查德·金; 蒂鲁马兰·穆蒂亚; 马修·富勒
Original assignee: Cambium Networks Ltd
Current assignee: Cambium Networks Ltd
Priority date: 2020-12-11
Filing date: 2021-12-10
Publication date: 2023-09-01

Abstract

Wireless communication is established between a first station (8) and a second station (9) in a wireless communication system, the first station (8) having a beam forming network (23) configured to form a continuous beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors. The orientations of the beams are arranged in a grid comprising a plurality of rows. The beams of each row are spaced apart in angular position such that at least one beam in a respective row is positioned intermediate the positions of two beams of an adjacent row. Successive beams are formed to transmit a first message using the selected first subset of antenna weight vectors. If a first message in a first beam (38) is received at the second station (9), a second subset of the antenna weight vectors selected to form beams adjacent to the first beam (38) is used to form further successive beams (39 a,39 b).

Description

Establishing wireless communication in a beam forming system by selecting from a predetermined plurality of antenna weight vectors

Technical Field

The present invention relates to establishing wireless communication between a first station and a second station in a wireless communication system, the first station having an antenna comprising an array of antenna elements and a beam forming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors.

Background

As the demand for increased bandwidth continues and the cost of radio frequency electronics decreases, the market for wireless systems operating at higher and higher frequencies increases. Particularly for fixed wireless access systems, radio stations are required to have high antenna gain to provide sufficient system gain to establish long range communications, which may be one kilometer or more, at higher frequencies, 20GHz or more, and up to 60GHz or even more. In order to provide a high gain antenna beam, an array of antenna elements may be generally provided in which the amplitude and/or phase of each antenna element is controlled by a beamformer to produce a beam, which may be selected from preconfigured beams. However, establishing initial wireless communications in such systems can be difficult, in part, due to the narrow beam, particularly under adverse atmospheric conditions (such as rain), which can introduce additional transmission loss.

Disclosure of Invention

According to a first aspect of the present invention there is provided a method of establishing wireless communication between a first station and a second station in a wireless communication system, the first station having an antenna comprising an array of antenna elements and a beamforming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors, the method comprising:

Providing a predetermined plurality of antenna weight vectors at the first station, the predetermined plurality of antenna weight vectors configured to form a plurality of beams, the plurality of beams oriented in a grid comprising a plurality of rows, the beams of each row being spaced apart in angular positions in a row on the first axis such that at least one beam in a respective row is positioned midway between the positions of two beams in adjacent rows on the first axis;

selecting a first subset of the predetermined plurality of antenna weight vectors for use at the first station;

forming, at the first station, a continuous beam in a first time sequence using a first subset of the predetermined plurality of antenna weight vectors to transmit a first message;

a further successive beam is formed at the first station using a second subset of the predetermined plurality of antenna weight vectors selected to form beams adjacent to the first beam in accordance with receiving the first message at the second station using the first beam of the first station.

Providing a predetermined plurality of antenna weight vectors provides computational efficiency by allowing non-real-time computation of the weight vectors. Providing a limited number of beams allows an efficient search procedure to find the best beam for communication between the first station and the second station. The orientations of the plurality of beams are arranged in a grid comprising a plurality of rows, the beams of each row being spaced apart in angular position in a row on the first axis such that at least one beam in a respective row is positioned midway between the positions on the first axis of two beams in an adjacent row on the first axis, which provides a beam pattern having a greater gain in the area between the beams than provided in a conventional rectangular beam grid. The beams are arranged in a plurality of approximately equilateral triangles. Such an arrangement allows communications to be established that are more tolerant of losses due to adverse atmospheric conditions, such as rain. Selecting a first subset of the predetermined plurality of antenna weight vectors for use by the first station allows for initial searching of the beam in a time efficient manner. In particular, if both the first station and the second station are performing a search, the search time may be approximately proportional to the square of the number of beams in a subset of one station, which is the time it takes on average when the situation in which two stations simultaneously form beams to the other station occurs. The first message is transmitted at the first station using a first subset of the predetermined plurality of antenna weight vectors, the successive beams formed in the first time sequence allowing identification of which subset of beams can be received at the second station. The second station may send an acknowledgement message back to the first station if the first message has been received. In a time division duplex system in which beams are formed to alternately transmit and receive beams in a time sequence, an acknowledgement message may be transmitted using the same beam used to transmit a first message. Once a first beam has been selected from the subset of beams, a second subset of the predetermined plurality of antenna weight vectors selected to form beams adjacent to the first beam is used at the first station to form a further continuous beam. This allows a fine search of the beams to identify if one of the beams in the second subset has better alignment than the first beam with the second station. This can be determined by a measurement of the signal-to-noise ratio or the received signal strength.

In an example, the angular positions of the beams of each row of the grid in the row are spaced apart by a first angular interval on the first axis, the angular positions of the beams of each row being offset on the first axis by half of the first angular interval relative to the angular positions of the beams in an adjacent row.

Spacing the beams on a row by a constant angular spacing along the row provides a slightly reduced drop in gain in the region between the beams toward the end of the row compared to in the middle of the row, as the beamwidth of the beams toward the end of the row may be greater, but the angular spacing between the beams remains constant. This helps to counteract this effect, which is that the gain of the beam towards the centre of the row end may be lower than the gain of the beam towards the centre of the row.

In an example, each row is spaced apart from an adjacent row by a first angular interval of 30 degrees cosine on a second axis perpendicular to the first axis, whereby each of the plurality of beams is arranged as an equilateral triangle with two adjacent beams. This provides a beam pattern of greater gain in the area between the beams than a conventional rectangular beam grid.

In an example, the first subset is selected to form selected beams on alternating rows of the grid, the selected beams of each alternating row being spaced apart by a first angular interval that is twice the angular position of each alternating row on the first axis, and the angular position of the selected beams of each alternating row being offset by the first angular interval on the first axis, whereby each selected beam is arranged as an equilateral triangle with two adjacent selected beams.

The selection of the beam subset allows for a time efficient procedure for establishing the initial communication. The arrangement of the equilateral triangles allows the beam pattern to have a larger gain in the area between the beams of the beam subset than provided in a conventional rectangular beam grid.

In an example, if the first beam or the second beam is not at an edge of the grid, the second subset is selected to form a ring of at least six beams surrounding the respective first beam and second beam.

This provides an efficient procedure for selecting the best beam to use after the initial communication is established.

In an example, the first axis is an azimuth angle. The array of antenna elements may have 8 columns of elements and 8 rows of elements, the spacing between the antenna elements in each row of elements and each column of elements being substantially one half wavelength at the operating frequency of the wireless communication system. Each predetermined antenna weight vector may provide a respective phase shift for each antenna element and the beam grid may comprise 120 beams arranged in 8 rows. The predetermined plurality of antenna weight vectors are configured to form a plurality of beams in a grid of beams extending substantially +/-40 degrees in azimuth and +/-20 degrees in elevation at the first station.

In an example, the array of antenna elements of the first station is arranged to feed the sub-reflectors of the offset gli-high antenna arrangement such that a plurality of beams are formed from the main reflector of the offset gli-high antenna arrangement. The predetermined plurality of antenna weight vectors may be configured to form a plurality of beams from a main reflector plate of the gli-high antenna arrangement by forming a plurality of feed beams from an array of antenna elements, each feed beam corresponding to a respective one of the plurality of beams from the main reflector plate.

Providing a biased gurley antenna arrangement provides a convenient way of increasing the beam gain provided by an array of antenna elements. The secondary reflector may be offset on a vertical axis relative to the center of the primary reflector dish to allow the beam to be formed over a wide range of azimuth angles without being blocked by the secondary reflector or array. Providing a predetermined plurality of weight vectors is computationally efficient by allowing the computation of the weight vectors to be performed prior to forming the beam. Arranging the orientations of the plurality of beams in a grid comprising a plurality of rows allows forming successive beams at different azimuth angles and at the same elevation angle, which allows a convenient way of forming trial beams, e.g. establishing initial communication between a subscriber module and an access point of a wireless communication system. This also allows a convenient way to reselect the beam to track the movement of the subscriber module due to wind loading, for example if the subscriber module is mounted on a pole above the subscriber premises. Providing a predetermined plurality of weight vectors such that the relationship between the azimuth and elevation directions of each feed beam and the azimuth and elevation directions of the corresponding beam from the main reflector dish is a nonlinear function of azimuth and elevation, allows the plurality of beams formed from the subscriber module to be arranged in a series of straight rows in the grid by arranging the feed beams from the array of antenna elements into a twisted grid. This allows the processor to apply a simple algorithm to steer the beam by selecting the beam in a straight line.

In an example, the predetermined plurality of antenna weight vectors are configured to form a plurality of feed beams such that the orientations of the plurality of beams are arranged in a distortion grid comprising a plurality of curved rows, each curved row providing a monotonic change in azimuth along the curved row and a non-monotonic change in elevation along the curved row. Each curved row may have an elevation offset between the center of the curved row and either end of the curved row. The elevation offset of a curved row may be equal to +/-50% of the elevation angular spacing between the curved row and an adjacent curved row at the center of the curved row. This provides a plurality of beams as a grid comprising a plurality of rows with a well-approximated constant elevation angle. In general, each curved row may have a greater elevation angle at the center of the curved row than at either end of the curved row. In an example, each curved row has an approximately parabolic dependence of elevation versus azimuth angle within +/-50% of a true parabola.

In an example, the first station is a subscriber module and the second station is an access point of a fixed wireless access communication system.

In an example, the wireless communication system is a time division duplex wireless system.

In an example, the wireless communication system has an operating frequency of at least 50GHz.

In an example, the second station has an antenna comprising an array of antenna elements and a beamforming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors, wherein the method comprises:

providing, at a second station, a second predetermined plurality of antenna weight vectors configured to form a second plurality of beams, the orientations of the second plurality of beams being arranged in a second grid comprising a plurality of rows, the beams of each row being spaced apart by a second angular interval at an angular position in the row, the angular position of the beams of each row being offset by half of the second angular interval relative to the angular positions of the beams in an adjacent row;

selecting a subset of the second predetermined plurality of antenna weight vectors for use at the second station;

forming, at the second station, successive beams in a second time sequence using a subset of the second predetermined plurality of antenna weight vectors: and

the method further includes forming a further continuous beam at the first station using a second subset of the predetermined plurality of antenna weight vectors selected to form a beam adjacent to the first beam, and forming a further continuous beam at the second station using a second subset of the second predetermined plurality of antenna weight vectors selected to form a beam adjacent to the second beam, based on receiving the first message at the second station using the first beam of the first station and the second beam of the second station.

In an example, the or each beamforming network is configured to form each beam for transmission and reception.

According to a second aspect of the present invention there is provided a first wireless station configured to establish wireless communication between a first station and a second station in a wireless communication system, the first station having an antenna comprising an array of antenna elements and a beamforming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors, the first wireless station configured to:

A further continuous beam is formed at the first station using a second subset of the predetermined plurality of antenna weight vectors selected to form beams adjacent to the first beam in accordance with the first message being received at the second station using the first beam of the first station.

In an example, the first wireless station comprises an offset gli-y antenna arrangement, wherein the array of antenna elements of the first station is arranged to feed sub-reflectors of the offset gli-y antenna arrangement such that a plurality of beams are formed from main reflector plates of the offset gli-y antenna arrangement.

Further features and advantages of the invention will become apparent from the following description of examples of the invention with reference to the accompanying drawings.

Drawings

For a more easy understanding of the invention, examples of the invention will now be described with reference to the accompanying drawings, in which:

fig. 1 is a schematic diagram illustrating a wireless communication system having an access point and subscriber modules, each subscriber module having an array of antenna elements;

fig. 2 illustrates a first wireless station configured to form a plurality of beams;

fig. 3 shows a first wireless station and a second wireless station, each configured to form a plurality of beams;

fig. 4a shows a wireless station with an array of antenna elements and a beamformer;

Fig. 4b shows an array of antenna elements of fig. 4 b;

fig. 5 shows a rectangular grid of preconfigured antenna beams;

fig. 6 shows an antenna diagram of a rectangular grid of antenna beams of fig. 5;

fig. 7 shows a detail of fig. 6;

fig. 8 shows a grid of pre-configured beams in a triangular arrangement;

fig. 9 shows an antenna gain diagram of the triangular mesh of fig. 8;

fig. 10 shows a detail of fig. 9;

fig. 11 shows an antenna gain diagram of a triangular mesh of the beam subset of fig. 8 formed from a subset of a predetermined plurality of antenna weight vectors;

fig. 12 shows a grid of preconfigured beams with a pitch triangle arrangement proportional to the beam width on a row;

fig. 13 shows an antenna gain diagram of the grid of fig. 12;

fig. 14 is a schematic diagram showing the principle of operation of a biased gurley antenna arrangement with a planar array of antenna elements as feed;

FIG. 15 shows a sub-reflector of a multiple feed beam formed to feed a biased Grignard antenna arrangement;

fig. 16 shows a schematic cross-sectional view of a biased gurley antenna arrangement;

FIGS. 17a and 17b are schematic diagrams showing the shape of the primary and secondary reflector dish in vertical and horizontal cross-sections, respectively;

Fig. 18 shows an oblique perspective view of a first wireless station with an offset gurley antenna arrangement;

fig. 19 is a plan view of a first wireless station looking from the direction of a radio frequency main beam in which the offset gray antenna arrangement is configured to form;

FIG. 20 shows a grid of a plurality of preconfigured beams formed by a main reflector of a biased Grignard antenna arrangement;

FIG. 21 is an antenna diagram of the grid of FIG. 20;

fig. 22 shows a plurality of feed beams from an array of antenna elements as feeds to a secondary reflector of a gurley antenna system; and

fig. 23 is a flow chart of a method according to an example.

Detailed Description

Examples of the invention are described in the context of a terrestrial fixed radio access wireless communication system operating in the 57-66GHz band according to IEEE 802.11 ay. In the described example, the wireless communication system is a time division duplex wireless system. However, it should be understood that embodiments of the present invention may relate to other applications and to other frequency bands.

Fig. 1 shows in a schematic plan view a wireless communication system with an access point 1 and subscriber modules 2a, 2b, 2c, 3a, 3 b. The access point 1 as shown in the figure covers two sectors, with two planar arrays of antenna elements 4, 5 arranged to cover a first sector, and further two antenna arrays 6, 7 arranged to cover a second sector. Each of the arrays is arranged to form a beam within an azimuth angle of about +/-40 degrees and an elevation angle of +/-20 degrees of the boresight direction of the array, that is, perpendicular to the plane of the array. Two arrays covering one sector are arranged to have visual axis directions that differ by about 80 degrees so that beams can be formed in successive angular sectors of about 160 degrees using the two arrays. Each element of the array of antenna elements is connected to a beamformer, which may be in the form of a commercially available beamformed radio frequency integrated circuit, arranged to apply a selected weight vector comprising the respective transmission phase of each element of the array. For example, the array of antenna elements may be an 8 x 8 array of patch antenna elements spaced about a half wavelength apart. The beamformer may generally be arranged to form a beam selected from a plurality of preconfigured beams, in the example 120 preconfigured beams. The preconfigured beams may be distributed over angular sectors of about +/-40 degrees in azimuth and about +/-20 degrees in elevation.

In the fixed wireless access wireless communication system shown in fig. 1, the access point 1 is typically located on a tower, the subscriber modules may be a mix of high gain 3a, 3b and lower gain 2a, 2b, 2c subscriber modules, typically fixed to a pole mounted on a subscriber premises, which may be, for example, a commercial or private residential premises. The lower gain subscriber modules 2a, 2b, 2c have an antenna arrangement comprising an array of antenna elements similar to those used at the access point and may be mounted relatively close to the access point, typically within a few hundred meters. The high gain subscriber modules 3a, 3b have an antenna arrangement comprising an array of antenna elements similar to or identical to those used at the access point, but which array is used as a feed for a biased gurley antenna arrangement, which provides improved antenna gain and a narrower antenna beam,

in one example, for the lower gain subscriber modules 2a, 2b, 2c, the array of antenna elements may be an array of identical 8 x 8 patch antenna elements used at the access point, and the beamformer may be further arranged to form beams selected from 120 preconfigured beams distributed over angular sectors of about +/-40 degrees azimuth and +/-20 degrees elevation. To establish communication at the first installation, the lower gain subscriber modules 2a, 2b, 2c are essentially aimed in the direction of the access point and the best beam to use can be selected by scanning the possible beams at the subscriber modules and the possible beams at the access point, which can be an exhaustive search of each combination of beams so that the best beam at the subscriber module and the best beam at the access point can be selected.

The higher gain subscriber modules 3a, 3b may be installed at a greater distance from the access point, for example, at a distance of 1km or more. The higher gain antenna arrangement can overcome the effects of greater signal loss due to greater propagation distance, as well as signal loss due to oxygen absorption and rain in the approximately 60GHz band.

The high gain subscriber modules 3a, 3b typically use the same array of antenna elements and the same beamforming arrangement as used at the access point 1 and the lower gain subscriber modules 2a, 2b, 2c as a feed for the biased gurley antenna system. The beams generated by the array of antenna elements are reflected by the sub-reflectors of the offset grignard antenna system onto the main reflector dish to generate a beam from the main reflector dish that is narrower than the beams generated by the array. For example, the beams produced by the array are approximately +/-8 degrees between 3dB points and the beams transmitted or received by the primary reflector dish are approximately 0.7 degrees between 3dB points. This reduced beamwidth increases the gain, which may provide about 22dB gain increase compared to the gain of the antenna array alone. For this arrangement, the total gain of the antenna arrangement of the high gain subscriber module may be about 44dBi (decibels compared to isotropic). The high gain antenna arrangement results in a reduction of the angular sector over which the beam can be formed. In the above example, the preconfigured beams may be distributed over angular sectors of about +/-2 degrees in azimuth and +/-1 degree in elevation from the primary reflector dish. As with the lower gain subscriber modules, the same technique that uses beam scanning at the access point and subscriber module is used to find the best beam. Because of the narrower beam and the smaller angular sector in which the beam can be steered, the subscriber module is typically first installed using an optical sight attached to the high gain subscriber module in an orientation in which the angular sector in which the beam can be steered includes the access point direction.

Fig. 2 shows a first wireless station 8, which may be a subscriber module or an access point, configured to form a plurality of beams 10 using an array of antenna elements 13, and a second wireless station 9, which is typically the other of the access point and the subscriber module, configured to form a fixed beam 11. In this case, only the first wireless station 8 performs scanning of the first beam.

Fig. 3 shows a first wireless station 8 and a second wireless station 9, each configured to form a plurality of beams 10, 12. The first wireless station may be a subscriber module and the second wireless station may be an access point. In another configuration, more than one access point may be used to form a mesh communication system. In this case, both the first wireless station 8 and the second wireless station 9 may be access points, and each wireless station will select the best beam for use with the search procedure already described.

Fig. 4a shows a wireless station, which may be a first wireless station and/or a second wireless station. The radio station has an array 22 of antenna elements and a beamformer 23. Fig. 4b shows that the array of antenna elements 22 comprises a two-dimensional planar array of elements. The antenna element may be a conventional patch antenna element formed from a conductive metal film carried on a non-conductive substrate such as ceramic tile or conventional printed circuit board material. Two or more planar arrays may be arranged with different boresight directions in azimuth to allow beams to be formed over a larger angular sector. The beamformer weights the signals transmitted and/or received by each antenna element. Typically, the weights are transmission phase values. The transmission phase value is typically quantized, e.g. allowing switching in steps of 90 degrees. The application of the transmission phase value to the signal may be achieved by using a switchable transmission delay. The combiner/divider tree connects each element of the beamformer to the radio modulator/demodulator 25 via a frequency converter stage to convert the received signal and/or signals for transmission to or from a digital format. The processor and controller 26 controls the beam selection and acquisition stages. The preconfigured beams are stored in a memory 24 for application to the beamformer 23 under control of a processor 26. The processor may be implemented using conventional digital technology and may be implemented in software, firmware, or cloud-based processing. The use of phase weights applied to signals received and/or transmitted by antenna elements to form beams is well known in the art. For example, a beam having a conventional sine function beam shape may be formed by applying weights to antenna elements having uniform gain and appropriate phase slope across the array in each dimension to steer the beam in a desired direction. The antenna array and beamformer may be commercially available products such as the samsung SWL-QD46 module.

Fig. 5 shows that a predetermined plurality of antenna weight vectors may be configured to form a plurality of beams 35a, 35b, etc., the orientations of which are arranged in a rectangular grid. For clarity, the reference numerals are shown as examples of only a few beams, but each of the illustrated beams is one of a plurality of beams. This method allows searching for beams for acquisition using conventional searching in a two-dimensional plane arranged in rows and columns as shown in the figure. By this method, it is possible to easily search by adding an index in each orthogonal dimension (typically in azimuth and elevation).

Fig. 6 shows an antenna gain diagram of the beam configuration of fig. 5, showing the maximum gain that can be achieved by selecting the best beam at each point in the grid. Fig. 7 shows a portion of fig. 6 in more detail. It can be seen that in valleys 36 between the peaks of beam 35, the gain is about 2.4dB lower than the gain of the beam peaks. The profile is stepped by 0.2 dB.

In contrast, fig. 8 shows that a predetermined plurality of antenna weight vectors may be configured to form a plurality of beams 27a, 27b, etc., the orientations of which are arranged in an arrangement of equilateral triangles. For clarity, the reference numerals are shown as examples of only a few beams, but each of the beams shown therein is one of a plurality of beams, that is, each of the beams numbered 1-120. The numbers in circles 28a, 28b, etc. are the beams in the first subset of the plurality of beams for initial searching, while the numbers in squares 39a, 39b, etc. are the beams in the second subset of the plurality of beams for fine searching. As can be seen in fig. 8, the beams of each row are spaced apart at angular positions in the row on the first axis 32 such that at least one beam in the respective row is positioned midway between the positions on the first axis of the two beams of the adjacent row. For example, beam number 22 in row 34b is centered between azimuth axis 32 between beam numbers 9 and 10 in row 34 a.

Fig. 9 shows an antenna gain diagram for the beam configuration of fig. 8, showing the maximum gain that can be achieved by selecting the best beam at each point in the configuration. Fig. 10 shows a portion of fig. 9 in more detail. The profile is stepped by 0.2 dB. It can be seen that in valleys 37 between the peaks of beam 27, the gain is about 1.8dB lower than the gain of the beam peaks. This arrangement improves the minimum available gain by about 0.6dB over the rectangular arrangement of figure 5. This improvement provides additional link margin to allow communication and acquisition under adverse atmospheric conditions.

In order to establish wireless communication between a first station and a second station in a wireless communication system, the following method may be used. The first station has an antenna comprising an array of antenna elements and a beamforming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors. The predetermined plurality of antenna weight vectors may be referred to as a codebook.

The predetermined plurality of antenna weight vectors are configured to form a plurality of beams. As illustrated in fig. 8, the position of each of the plurality of beams is shown superimposed. The orientations of the beams are arranged in a grid comprising a plurality of rows 34a, 34b, etc., the beams of each row being angularly spaced apart in this row by a first angular spacing 29 on the first axis 32, in this example an azimuth angle. The angular position of each row of beams is offset 30 by half the first angular interval 29 on the first axis 32 relative to the angular position of the adjacent row of beams. For example, it can be seen that the angular position of each beam 27a, etc. in row 34a is offset from the beam in the adjacent row 34b by half the spacing between beams 29. The beam numbers 1-120 shown in fig. 8 are arbitrary.

As shown in fig. 8, each row is spaced from an adjacent row by a first angular interval 29 times the cosine 30 degrees on a second axis perpendicular to the first axis 32 such that each of the plurality of beams is arranged in an equilateral triangle with two adjacent beams. For example, row 34a is separated from row 34b by a first angular interval 29 multiplied by the cosine 30 degrees. In practice, the position of the beam will be affected by errors due to the accuracy of the beam forming weights, taking into account the quantization effects.

For an initial acquisition search, a first subset of a predetermined plurality of antenna weight vectors is selected, as shown in circular beams 28a, 28b, etc. In this case, 30 beams, i.e. 1 out of 4 beams, are selected for the subset.

Fig. 11 shows an antenna gain diagram of a triangular grid of beam subsets 28c, 28d, etc., formed by a subset of a predetermined plurality of antenna weight vectors. A subset, referred to as a first subset, is selected to form selected beams on alternating rows of the grid, the selected beams of each alternating row being spaced apart by twice the first angular interval on the first axis and the angular position of the selected beams of each alternating row being offset by the first angular interval on the first axis. This would arrange each selected beam in the subset as an equilateral triangle with two adjacent selected beams. This arrangement provides a minimum gain of about-5.5 dB between the beam peaks of the subset. This provides an improvement in the link margin of the initial acquisition due to the tighter packing of the triangular arrangement compared to the minimum gain between the peaks of the rectangular beam grid.

In the alternative, another proportion of beams other than 1/4, for example 1/9, may be selected for the subset. In each case, the triangular arrangement shows advantages.

During acquisition, successive beams are formed in a first time sequence at a first station using a first subset of a predetermined plurality of antenna weight vectors to transmit a first message.

In accordance with receiving the first message at the second station using the first beam of the first station, an accurate search is performed using further successive beams formed at the first station using a second subset of the predetermined plurality of antenna weight vectors selected to form beams adjacent to the first beam. As marked with squares in fig. 8, if the first beam or the second beam is not at the edge of the grid, the second subset is selected to form a ring of at least six beams surrounding the respective first beam and second beam. The second subset of beams is shown in fig. 8 by the beams marked by squares 39a, 39b, etc. surrounding the first beam 38 selected during the initial acquisition. This provides an efficient procedure for selecting the best beam to use after the initial communication is established.

As already mentioned, the array of antenna elements of the first station may be arranged to feed a sub-reflector of a biased gli-high antenna arrangement to increase the antenna gain. This is beneficial for subscriber modules, access points, or wireless stations arranged in a mesh arrangement where greater gain is required.

Fig. 12 shows another arrangement in which the grid of preconfigured beams has a pitch in the triangle arrangement that is proportional to the beam width on the rows. It can be seen that the beams centered in rows 27f, 27g have closer angular spacing than the beams near the ends of rows 27h, 27 i. The spacing is constant in terms of the ratio of the 3dB beamwidth of the beams. This may provide a way to reduce the number of beams that need to be searched during acquisition. In this way, the gain reduction between the regions between beams can be kept approximately constant over the beam grid.

Fig. 13 shows in schematic form an antenna gain diagram of the grid of fig. 12.

Fig. 14 is a schematic diagram showing the principle of operation of a biased glicah antenna arrangement having a main reflector dish 43 and a sub-reflector 42. The array of antenna elements 41 is used to feed the sub-reflector 42 with radio frequency radiation forming a first beam having a first beamwidth. The amplitude and/or phase of the signals fed to/received from the respective elements of the array are arranged to have appropriate values to form a beam having the desired direction and beam width. The amplitude and/or phase of the signals fed to/received from the respective elements is typically controlled by a beamformer implemented by a radio frequency integrated circuit. The effect of the combination of the primary reflector dish 43 and the secondary reflector 42 is to increase the gain of the first beam, producing a second beam of reduced beamwidth. For example, the first beam may have a beam width of about 8 degrees, measured as the angular distance between points of the radiation beam having a gain 3dB lower than the gain of the beam center, and the second beam may have a beam width of about 0.5 degrees.

Fig. 15 shows a plurality of feed beams 45 formed by an array of antenna elements 41 to feed a sub-reflector 42 of a biased gurley antenna arrangement to produce a plurality of beams 46 from a main reflector dish 43. It can be seen that a given deviation of the feed beam from the direction perpendicular to the array will result in a smaller deviation of the beam from the main reflector dish 43. Thus, the angular sector of beam 46 from the main reflector dish may be formed narrower than the angular sector of beam 45 from the main reflector dish. Each feed beam corresponds to a corresponding one of the plurality of beams from the main reflector dish.

Fig. 16 shows an example of an implementation of the offset grignard antenna arrangement in an example of a high gain subscriber module, showing a planar array of sub-reflectors 42 and antenna elements 41, the antenna elements 41 being arranged for transmitting radio frequency signals to the sub-reflectors 42 and/or for receiving a feed of radio frequency signals from the sub-reflectors 42. The conductive support blocks are configured to support a planar array of antenna elements 41. The support block is formed as a first end of a feed support member 47, which is directly connected at an end opposite to the first end to a support body 48 configured to support the main reflector dish 43.

Fig. 17a shows a typical profile in a vertical cross-section through a bias gurley antenna arrangement in a plane similar to the cross-section plane of fig. 16. The reflector surfaces of the primary reflector dish 43 and the secondary reflector 42 are shown. A practical implementation may comprise a decreasing section of the theoretical curve shown in fig. 17a and 17 b. The offset gli arrangement is arranged such that the array of sub-reflectors and antenna elements does not obscure the sector of the beam that is intended to be formed from the main reflector. Typically, the secondary reflector is vertically offset such that the azimuth sector is unobstructed. In general, in a fixed wireless access system, the range of elevation angles required for beam forming is smaller than the range of azimuth angles. A planar array of antenna elements 41 is also shown. Fig. 17b shows a typical profile in a horizontal cross section through an offset gli-high antenna arrangement, again showing the reflector surfaces of the main reflector dish 43 and the sub-reflector 42, and the planar array of antenna elements 41. The main reflector dish 43 has a parabolic shape in both vertical and horizontal cross-sections. The sub-reflector dish 42 also has a parabolic shape in both vertical and horizontal cross-sections.

Fig. 18 shows an oblique perspective view of a wireless station with an offset glicah antenna arrangement in an example, showing an aperture 50 for visually aligning the wireless station with a second wireless station, a primary reflector dish 43 and a non-conductive housing 49 encapsulating the secondary reflector and its support.

Fig. 19 is a view of an offset gurley antenna arrangement seen from a radio frequency main beam direction in an example, the offset gurley antenna arrangement configured to form the main beam. It can be seen that the main reflector dish 43 is substantially rectangular in plan view, as seen in a direction parallel to the main beam of radio frequency in which the offset gli-high antenna arrangement is configured to form. The primary reflector dish 43 may be formed of stamped metal. This arrangement has been found to provide a compact design with high radio frequency gain. The sub-reflector covered by the housing 49 may also have a substantially rectangular shape in plan view, viewed in a direction parallel to the main beam of radio frequency in which the offset gli-high antenna arrangement is configured to form.

In the case of the gurley antenna arrangement, a predetermined plurality of antenna weight vectors are configured to form a plurality of beams from the main reflector dish 43, as shown in fig. 20. Fig. 20 shows that the position of each of the plurality of beams is overlapped in an arrangement similar to that of fig. 8. In the case of fig. 18, the reduced set of beams is used for communication, shown within an ellipse 35. At least some of the rows are truncated at each end. The resulting truncated row is longest towards the center of the high elevation value range.

Similar to the case of fig. 8 without the offset gli-high antenna system, the orientation of the beams is arranged in a grid comprising a plurality of rows 34, the beams of each row being spaced apart in angular position in this row by a first angular interval 29 on a first axis 32, in this example an azimuth angle. The angular position of each row of beams is offset 30 by half the first angular interval 29 on the first axis 32 relative to the angular position of the adjacent row of beams. For example, it can be seen that each beam in row 34 is offset by half the spacing between beams 29 relative to the angular position of the beams in the adjacent row.

The beam numbers selected from the range of 1-120 shown in fig. 20 relate to an alternative example of the order in which each beam is addressed during the search, wherein every fourth digital beam is selected to form the first subset of beams. The second subset of beams may be selected by selecting the first 10 and the last 10 beams in numerical order of the numbered system. This selects a second subset of beams comprising at least six beams surrounding the selected beam. This provides a simple algorithm for selecting the beam.

As shown in fig. 20, on a second axis perpendicular to the first axis 32, each row is spaced from an adjacent row by a first angular interval 29 multiplied by a cosine of 30 degrees such that each of the plurality of beams is arranged in an equilateral triangle with two adjacent beams. In practice, the position of the beam will be affected by errors due to the accuracy of the beam forming weights, taking into account the quantization effects.

For an initial acquisition search, a first subset of the predetermined plurality of antenna weight vectors is selected, as shown in the encircled beam 28. In this case 26 beams are selected for the subset.

Similar to fig. 9, fig. 21 shows an antenna gain diagram of the beam arrangement of fig. 17, showing the maximum gain that can be achieved by selecting the best beam at each point in the arrangement. The profile is stepped by 1 dB.

Similar to the beam formed directly by the array of antenna elements in the case shown in fig. 8, during acquisition, successive beams are formed in a first time sequence at a first station using a first subset of a predetermined plurality of antenna weight vectors to transmit a first message.

The first beam is used as a basis for a finer refinement procedure based on the first message being received at the second station using the first beam of the first station and an acknowledgement from the second station that the first message has been received. In the case where the first beam is used for both transmission and reception in a time division duplex arrangement, an acknowledgement from the second station may be carried in the first beam. An accurate search is performed using a further successive beam formed at the first station using a second subset of the predetermined plurality of antenna weight vectors selected to form beams adjacent to the first beam. As shown in fig. 17, if the first beam or the second beam is not at the edge of the grid, the second subset is selected to form a ring of at least six beams surrounding the respective first beam and second beam. The second subset of beams is shown in fig. 17 by the beams marked by squares 39a, 39b, etc. surrounding the first beam 38 selected during the initial acquisition. This provides an efficient procedure for selecting the best beam to use after the initial communication is established.

Fig. 22 shows an arrangement of feed beams generated by an array of antenna elements to feed a secondary reflector and to generate a beam arrangement from the primary reflector dish shown in fig. 20 and 21. Each feed beam number in fig. 22 produces a correspondingly numbered beam from the main reflector dish as shown in fig. 20.

It can be seen that the relationship between the azimuth and elevation directions of each feed beam and the azimuth and elevation directions of the corresponding beam from the main reflector dish is a nonlinear function of azimuth and elevation. As shown in fig. 22, the predetermined plurality of antenna weight vectors are configured to form a plurality of feed beams such that the orientations of the plurality of beams are arranged in a twisted grid comprising a plurality of curved rows 36a, 36b, etc., each curved row providing a monotonic change in azimuth 32 along the curved row and a non-monotonic change in elevation 33 along the curved row. As shown in the figures, each curved row 36a, 36b, etc. has an elevation offset between the center of the curved row and either end of the curved row. In this example, the elevation offset of a curved row is equal to +/-50% of the elevation angle spacing between the curved row and an adjacent curved row. In this example, each curved row has a greater elevation angle at the center of the curved row than at the ends of the curved row, and in this example, each curved row has an approximately parabolic dependence of elevation angle on azimuth angle within +/-50% of a true parabola.

Examples of equations that may be used to correlate the azimuth and elevation directions of each feed beam (x 1, y 1) with the azimuth and elevation directions of the corresponding beam from the main reflector dish (x, y) are as follows:

x1＝x ² ((0.12-0.0052y)y+0.17)+0.21x ³ +x((0.16y+0.55)y-13.)+y((-0.27y-0.55)y-0.27)-0.38

y1＝x ² ((0.061-0.047y)y-1.7)+0.077x ³ +x((0.030y+0.00061)y-0.46)+y((0.45y-0.48)y-14.)-5.8

the above equation has been found to be a useful approximation.

Both the first wireless station and the second wireless station may form multiple beams during the process of establishing communication. In this case, the second station has an antenna comprising an array of antenna elements and a beam forming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors. The method includes providing, at a second station, a second predetermined plurality of antenna weight vectors configured to form a second plurality of beams, the orientations of the second plurality of beams being arranged in a second grid comprising a plurality of rows, the beams of each row being spaced apart at a second angular interval over an angular position in the row, the angular position of the beams of each row being offset from the angular position of the beams in an adjacent row by half of the second angular interval. A subset of the second predetermined plurality of antenna weight vectors is selected for use at the second station. Successive beams are formed in a second time sequence at the second station using a subset of the second predetermined plurality of antenna weight vectors. The method further includes forming a further continuous beam at the first station using a second subset of the predetermined plurality of antenna weight vectors selected to form a beam adjacent to the first beam, and forming a further continuous beam at the second station using a second subset of the second predetermined plurality of antenna weight vectors selected to form a beam adjacent to the second beam, based on receiving the first message at the second station using the first beam of the first station and the second beam of the second station.

Fig. 23 is a flow chart of a method in an example according to steps S23.1, S23.2, S23.3 and S23.4.

In an example, the wireless communication system may have an operating frequency of at least 50 GHz. In other examples, the wireless communication system may have an operating frequency greater than 28GHz, such as 28GHz.

It is to be understood that any feature described in relation to any one example 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 examples, or any combination of any other of the examples. 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 method of establishing wireless communication between a first station and a second station in a wireless communication system, the first station having an antenna comprising an array of antenna elements and a beamforming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors, the method comprising:

providing a predetermined plurality of antenna weight vectors at the first station, the predetermined plurality of antenna weight vectors configured to form a plurality of beams, the plurality of beams being oriented in a grid comprising a plurality of rows, the beams of each row being spaced apart at angular positions in the row on a first axis such that at least one beam in a respective row is positioned midway between the positions of two beams on adjacent rows on the first axis;

forming, at the first station, a continuous beam in a first time sequence using the first subset of the predetermined plurality of antenna weight vectors to transmit a first message;

a further continuous beam is formed at the first station using a second subset of the predetermined plurality of antenna weight vectors selected to form beams adjacent to the first beam, based on receiving a first message at the second station using the first beam of the first station.

2. The method of claim 1, wherein the beams of each row of the grid are spaced apart in angular positions in the row by a first angular interval on a first axis, the angular positions of the beams of each row being offset in the first axis by half the first angular interval relative to the angular positions of beams of an adjacent row.

3. The method of claim 2, wherein each row is spaced from an adjacent row by 30 degrees of cosine on the first angular interval times a second axis, the second axis being perpendicular to the first axis, whereby each of the plurality of beams is arranged in an equilateral triangle with two adjacent beams.

4. A method according to claim 2 or 3, wherein the first subset is selected to form selected beams on alternating rows of the grid, the angular positions of the selected beams of each alternating row being spaced apart by twice the first angular spacing on the first axis, and the angular positions of the selected beams of each alternating row being offset by the first angular spacing on the first axis, whereby each selected beam is arranged as an equilateral triangle with two adjacent selected beams.

5. The method of any preceding claim, wherein if the first or second beam is not at an edge of the grid, the second subset is selected to form a ring of at least six beams surrounding the respective first and second beams.

6. A method according to any preceding claim, wherein the first axis is azimuth.

7. A method as claimed in any preceding claim, wherein the array of antenna elements has 8 columns of elements and 8 rows of elements, the spacing between the antenna elements in each row of elements and each column of elements being substantially half a wavelength at the operating frequency of the wireless communication system.

8. A method as claimed in any preceding claim, wherein each predetermined antenna weight vector provides a respective phase shift for each antenna element.

9. A method according to any preceding claim, wherein the grid of beams comprises 120 beams arranged in 8 rows.

10. The method of any preceding claim, wherein the predetermined plurality of antenna weight vectors are configured to form the plurality of beams in a grid of beams at the first station, the beams extending substantially +/-40 degrees in azimuth and +/-20 degrees in elevation.

11. The method of any of claims 1 to 9, wherein the array of antenna elements of the first station is arranged to feed a sub-reflector of a biased gurley antenna arrangement such that the plurality of beams are formed from a main reflector dish of the biased gurley antenna arrangement.

12. The method of claim 11, wherein the predetermined plurality of antenna weight vectors are configured to form a plurality of feed beams from a main reflector dish of a gurley antenna arrangement by forming the plurality of beams from an array of antenna elements, each feed beam corresponding to a corresponding one of the plurality of beams from the main reflector dish.

13. The method of claim 12, wherein the relationship between the azimuth and elevation directions of each feed beam and the azimuth and elevation directions of the corresponding beam from the main reflector dish is a nonlinear function of azimuth and elevation.

14. The method of claim 12 or 13, wherein the predetermined plurality of antenna weight vectors are configured to form the plurality of feed beams such that the orientations of the plurality of beams are arranged in a distorted grid comprising a plurality of curved rows, each curved row providing a monotonic variation in azimuth along the curved row and a non-monotonic variation in elevation along the curved row.

15. The method of claim 14, wherein each curved row has an elevation offset between a center of the curved row and either end of the curved row.

16. The method of claim 15, wherein the elevation offset of a curved row is equal to +/-50% of an elevation angular spacing between the curved row and an adjacent curved row.

17. The method of claim 15 or 16, wherein each curved row has a greater elevation angle at the center of the curved row than at either end of the curved row.

18. The method of any one of claims 15 to 17, wherein each curved row has an approximately parabolic correlation of elevation angle and azimuth angle within +/-50% of a true parabola.

19. A method as claimed in any preceding claim, wherein the first station is a subscriber module and the second station is an access point of a fixed wireless access communication system.

20. A method according to any preceding claim, wherein the wireless communication system is a time division duplex wireless system.

21. A method as claimed in any preceding claim, wherein the wireless communication system has an operating frequency of at least 50GHz.

22. The method of any preceding claim, wherein the second station has an antenna comprising an array of antenna elements and a beamforming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors, wherein the method comprises:

providing, at the second station, a second predetermined plurality of antenna weight vectors configured to form a second plurality of beams, the orientations of the second plurality of beams being arranged in a second grid comprising a plurality of rows, the beams of each row being spaced apart by a second angular interval at angular positions in the row, the angular positions of the beams of each row being offset from the angular positions of beams in an adjacent row by half of the second angular interval;

in accordance with receiving a first message at a second station using the first beam of the first station and the second beam of the second station, forming a further continuous beam at the first station using the second subset of the predetermined plurality of antenna weight vectors selected to form beams adjacent to the first beam, and forming a further continuous beam at the second station using the second subset of the second predetermined plurality of antenna weight vectors selected to form beams adjacent to the second beam.

23. A method as claimed in any preceding claim, wherein the or each beamforming network is configured to form each beam for transmission and reception.

24. A first wireless station configured to establish wireless communication between a first station and a second station in a wireless communication system, the first station having an antenna comprising an array of antenna elements and a beamforming network configured to form a beam using an antenna weight vector selected from a predetermined plurality of antenna weight vectors, the first wireless station configured to:

Providing the predetermined plurality of antenna weight vectors at the first station, the predetermined plurality of antenna weight vectors configured to form a plurality of beams, the plurality of beams being oriented in a grid comprising a plurality of rows, the beams of each row being spaced apart at angular positions in the row on a first axis such that at least one beam in a respective row is positioned midway between the positions of two beams on adjacent rows on the first axis;

25. The first wireless station of claim 23, comprising an offset gli-high antenna arrangement, wherein the array of antenna elements of the first station is arranged to feed sub-reflectors of the offset gli-high antenna arrangement, thereby forming the plurality of beams from a main reflector dish of the offset gli-high antenna arrangement.