WO2023230081A1

WO2023230081A1 - Method and apparatus for maintaining beam weights for ongoing beamforming in a satellite communications system

Info

Publication number: WO2023230081A1
Application number: PCT/US2023/023260
Authority: WO
Inventors: David J. Hancharik; Parker A. Robinson
Original assignee: Viasat, Inc.
Priority date: 2022-05-24
Filing date: 2023-05-23
Publication date: 2023-11-30

Abstract

A disclosed beamforming technique reduces the frequency of full recalculation for a set of beam weights while still allowing the beam weights to track dynamic conditions between full recalculations via computationally simpler evaluations performed in a beam strength domain. Transformation into the beam strength domain allows for the computation of beam adaptations between full recalculations based on gradient projections of beam contours. An example implementation involves beamforming in a satellite communications system, with the disclosed technique used to reduce the computational load that beamforming imposes on the system.

Description

METHOD AND APPARATUS FOR MAINTAINING BEAM WEIGHTS FOR ONGOING BEAMFORMING IN A SATELLITE COMMUNICATIONS SYSTEM

TECHNICAL FIELD

[0001] The present invention relates to satellite communications systems and, in particular, relates to beamforming in a satellite communications system.

BACKGROUND

[0002] Beamforming in the context of radio signals refers to directional transmission or reception. Transmit beamforming, for example, involves splitting a signal into multiple copies and applying respective weights to the copies, such that the simultaneous transmission of the weighted copies from respective antenna elements in an antenna array results in a far-field pattern of constructive and destructive interference that produces a beam having a desired shape and direction. Reception beamforming operates similarly, with an incoming signal received on multiple antenna elements producing per-element signals which are weighted in such a way as to create a directional reception sensitivity when the per-element signals are combined.

[0003] Among its many advantages, beamforming allows for more efficient use of available spectrum, based on reusing frequency, polarization, or other resources across multiple simultaneous beams according to a defined reuse pattern that minimizes interference between adj cent beams. Consider an example scenario of an overall satellite service area divided into a plurality of user beam coverage areas, with the involved satellite communications system using a plurality of simultaneous user beams to serve user terminals in the respective user beam coverage areas and employing patterns of frequency and polarization reuse across the plurality of user beams. U.S. Patent No. 10,079,636 B2, issued 18 September 2018, discloses various approaches to beamforming in a satellite communications system.

[0004] Along with its various advantages, beamforming imposes a number of challenges. One challenge is performing the many computations needed to determine the beam weights used to form the beams. Beam weight computations typically rely on a channel matrix containing coefficients representing current estimates of the involved propagation channels. These channel matrices may be quite large, meaning that the channel matrix inversion operations performed for beam-weight determination involve many calculations. Conventional approaches to maintaining updated beam weights involve recalculation on a periodic basis at relatively short intervals, e.g., on the order of milliseconds. Beamforming in the case of large channel matrices thus requires a large number of floating point operations per second — FLOPS — which in turn implies high costs and power consumption that may be impractical in processing-limited platforms.

SUMMARY

[0005] A disclosed beamforming technique reduces the frequency of full recalculation for a set of beam weights while still allowing the beam weights to track dynamic conditions between full recalculation via computationally simpler evaluations performed in a beam strength domain. Transformation into the beam strength domain allows for the computation of beam adaptations between full recalculations based on gradient projections of beam contours. An example implementation involves beamforming in a satellite communications system, with the disclosed technique used to reduce the computational load imposed on the system by beamforming.

[0006] One embodiment comprises a method of maintaining beam weights for ongoing beamforming via a satellite of a satellite communications system. The method includes: (a) performing a full recalculation of the beam weights at first intervals, each full recalculation being based on a channel state matrix reflecting then current propagation channel measurements; (b) transforming the beam weights as determined at each full recalculation into reference beam strengths having azimuthal and elevational components; and (c) performing adaptations of the beam weights at subintervals of the first intervals, based on changes in the reference beam strengths observed across successive first intervals.

[0007] Another embodiment comprises an apparatus configured for maintaining beam weights for ongoing beamforming via a satellite of a satellite communications system. The apparatus includes interface circuitry and processing circuitry. The interface circuitry is configured to receive channel state information reflecting then current propagation channel measurements. The processing circuitry is configured to: (a) perform a full recalculation of the beam weights at first intervals, each full recalculation being based on a channel state matrix determined from the channel state information; (b) transform the beam weights as determined at each full recalculation into reference beam strengths having azimuthal and elevational components; and (c) perform adaptations of the beam weights at subintervals of the first intervals, based on changes in the reference beam strengths observed across successive first intervals.

[0008] Of course, the present invention is not limited to the above features and advantages. Indeed, those skilled in the art will recognize additional features and advantages upon reading the following detailed description, and upon viewing the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 is a block diagram of a transmit (forward) beamforming system, such as may be used in a satellite communications system.

[0010] Figure 2 is a block diagram of a receive (return) beamforming system, such as may be used in a satellite communications system.

[0011] Figure 3 is a block diagram of a satellite communications system configured for forward beamforming, according to one embodiment..

[0012] Figure 4 is a diagram of a timing structure used for full and adaptive beam weight calculations — adaptive beamforming — in a satellite communications system according to an example embodiment.

[0013] Figure 5 is a diagram of an example beam weight matrix in an aperture weight domain.

[0014] Figure 6 is a block diagram of domain transformations and associated operations for adaptive beamforming according to an example embodiment.

[0015] Figure 7 is a diagram of example beam contours.

[0016] Figure 8 is a logic flow diagram of a method of maintaining beam weights in a satellite communications system according to an example embodiment.

[0017] Figure 9 is a logic flow diagram illustrating an example processing loop associated with the method depicted in Figure 8.

[0018] Figure 10 is a block diagram of a satellite communications system according to another example embodiment.

[0019] Figure 11 is a block diagram of a satellite communications system according to another example embodiment.

[0020] Figure 12 is a block diagram of an apparatus for maintaining beam weights in a satellite communication system according to an example embodiment.

DETAILED DESCRIPTION

[0021] Figure 1 depicts a transmit (TX) beamforming system 10 for forward beamforming in a satellite communications system, according to one embodiment. “Forward” refers to transmissions by the satellite communications system for user terminals served by the satellite communications system. Later drawings depict example satellite communications systems and user terminals, but these elements are not shown in Figure 1 to avoid clutter. [0022] The forward user beams 12 are “spot beams,” meaning that they are relatively narrow, focused beams of radiofrequency (RF) energy that transmit the involved signals on corresponding geographic areas, which are referred to as forward user beam footprints or forward user beam coverage areas 14. Each forward user beam 12 has a beam center 16, representing the geographic location of maximum beam strength and, more generally, each forward user beam 12 has a defined beamwidth. For example, the 3 dB beamwidth is the angle between the two points on either side of the maximum signal strength where the signal strength drops by half. Each forward user beam 12 may be visualized in terms of beam contours, which are defined as regions of equal beam strength within the corresponding forward user beam coverage area 14.

[0023] Only three forward user beams 12 appear in the diagram, labeled as 12-1, 12-2, and 12-3. Here and elsewhere, suffixing appears only where needed for clarity. Thus, the reference number “12” refers to any given forward user beam, or any given forward user beams, whereas “12-x” refers to a particular forward user beam, with “x” being “1 ,” “2,” and so on. While only three forward user beams 12 are shown, the involved satellite communications system may generate many hundreds of forward user beams 12 simultaneously and the illustrated arrangement of circuitry may be “sized” for forming large numbers of beams.

[0024] Simultaneous formation of multiple forward user beams 12 relies on a reuse pattern, for example, to limit interference between adjacent beams. For example, adjacent beams may be at different radiofrequencies and/or may use different signal polarizations, with the same frequencies and polarizations used across a potentially large set of forward user beams 12 according to a defined reuse pattern. Thus, each forward user beam 12 is a RF beam having a controlled beam shape and direction and is transmitted using a selected frequency and/or polarization. Beam direction may be expressed in terms of azimuthal and elevational angles and each forward user beam 12 carries forward traffic for user terminals assigned to it. Typically, the particular user terminal(s) assigned to a given forward user beam 12 are those user terminals operating in the forward user beam coverage area 14 of the given forward user beam 12.

[0025] Forming the forward user beams 12 relies on the satellite communications system receiving or otherwise determining channel state information (CSI) corresponding to user terminals at geographic locations at or near the beam center targets of the nominal user beam coverage areas. The satellite communications system uses the CSI to determine transmit beam weights that yield the desired shape and orientation for the forward user beams 12. Relative movement between the satellite and respective user terminals or other changing conditions requires repeatedly updating the forward beam weights used to form the forward user beams 12. [0026] The process of ongoing re-computation may be referred to as “maintaining” the beam weights. The transmit beamforming system 10 embodies an advantageous technique that reduces the computational burden of maintaining beam weights, with the reduced-complexity determinations still offering near-optimal results. The technique complements a variety of beamforming scenarios but particular advantages accrue in the context of “constrained” platforms, such as where the payload of a satellite includes the transmit beamforming system 10, for performing onboard forward beamforming. Use of the technique reduces the otherwise potentially immense computation burden imposed on the processing circuitry of the satellite, thus allowing for reduced cost, size, and power.

[0027] Reviewing example details for beamforming aids understanding of the advantageous technique for maintaining beam weights. Illustrated beamforming involves a number of forward beam signals 18. Each forward user beam 12 can be understood as the beamformed transmission of a corresponding one of the forward beam signals 18. Thus, each forward beam signal 18 carries forward user traffic for user terminals assigned to the corresponding forward user beam 12 and in the example context of three forward user beams 12-1, 12-2, and 12-3, there are three forward beam signals 18-1, 18-2, and 18-3.

[0028] The forward beamforming circuitry 20 includes splitter circuitry 22 that splits each forward beam signal 18 into a corresponding set of duplicate signals 24. Each such set includes N duplicate signals, where N is an integer number equaling the number of antenna elements used for transmit beamforming. Thus, Figure 1 illustrates the splitting of the forward beam signal 18-1 into a corresponding set of duplicate signals 24-1, the splitting of the forward beam signal 18-2 into a corresponding set of duplicate signals 24-2, and the splitting of the forward beam signal 18-3 into a corresponding set of duplicate signals 24-3. For clarity, there is a set of N duplicate signals 24 formed for each forward user beam 18, with each duplicate signal 24 in each such set targeted for transmission from a respective one among N antenna elements of the transmit antenna array involved in the beamforming.

[0029] Weighting circuitry 26 applies a first set of N forward beam weights, a beam weight matrix denoted as FBW1, to the duplicate signals 24-1, to form weighted signals 28-1. The beam weights FBW1 are calculated so that simultaneous transmission of the weighted signals 28-1 from respective antenna elements of the transmit antenna array used for transmit beamforming results in far field signal superpositions that form the first forward user beam 12-1. Using a mathematical framing, the beam weights FBW1 are complex values — amplitude and phase — contained in a matrix, where each element of the matrix is an amplitude/phase value to be applied to a particular one among the duplicate signals 24-1, for transmission from a corresponding one among the N antenna elements of the involved transmit antenna array. [0030] The same logic holds for each forward user beam 12 being formed by the satellite communications system. Thus, the forward beam weights FBW2 are calculated so that simultaneous transmission of the weighted signals 28-2 from respective antenna elements of the transmit antenna array results in far field signal superpositions that form the second forward user beam 12-2, and the forward beam weights FBW3 are calculated so that simultaneous transmission of the weighted signals 28-3 from respective antenna elements of the transmit antenna array results in far field signal superpositions that form the third forward user beam 12- 3.

[0031] An adaptive beam weight calculator 30 calculates the forward beam weights using an advantageous technique that reduces the overall computational load associated with maintaining the forward beam weights on a dynamic basis, to account for changing operational conditions. Briefly, the adaptive beam weight calculator 30, which is a computational apparatus such as a Central Processing Unit (CPU) and supporting circuitry, maintains the beam weights (forward and/or reverse) for ongoing beamforming via a satellite of the satellite communications system 10 by: (a) performing a full recalculation of the beam weights at first intervals, each full recalculation being based on a channel state matrix reflecting then current propagation channel measurements; (b) transforming the beam weights as determined at each full recalculation into reference beam strengths having azimuthal and elevational components; and (c) performing adaptations of the beam weights at subintervals of the first intervals, based on changes in the reference beam strengths observed across successive first intervals.

[0032] Determination of the channel state matrix, for example, relies on the adaptive beam weight calculator 30 receiving channel state information (CSI) 32 that indicates the relevant propagation channels, or otherwise provides a basis for estimating those channels. For example, for each forward user beam coverage area 14, there is at least one designated user terminal (DUT) at or near the geographical location representing the target location for the beam center 16. These DUTs receive reference signals transmitted by the satellite communications system and return corresponding channel estimates to the satellite communications system, with the channel estimates or information derived from them being the CSI 32 provided to the adaptive beam weight calculator 30. [0033] Combining circuitry 34 receives the respective set of weighted signals 28 for each forward user beam 12 to be formed simultaneously by the satellite communications system. In the three-beam example of Figure 1 , then, the combining circuitry 34 receives the first set of weighted signals 28-1 corresponding to the first forward beam signal 18-1, the second set of weighted signals 28-2 corresponding to the second forward beam signal 18-2, and the third set of weighted signals 28-3 corresponding to the third forward beam signal 18-3. These weighted signals may be referred to as “forward beam element signals 28,” to denote that their simultaneous transmission from respective array elements of the involved transmit antenna array result in beam formation.

[0034] The combining circuitry 34 linearly combines the forward beam element signals 28 that are all targeted to the same antenna array element, to produce a corresponding combined forward beam element signal 36. As such, the combining circuitry 34 outputs one set of combined forward beam element signals corresponding to the multiple sets of forward beam element signals 28 input to it. In notation form, let besl 1 denote the forward beam element signal in the first set of forward beam element signals 28-1 that is targeted for transmission from a first antenna element of the transmit antenna array. Continuing that notation, bes21 denotes the forward beam element signal from the second set of forward beam element signals 28-2 that is targeted to the first antenna element, and bes31 denotes the forward beam element signal from the third set of forward beam element signals 28-3 that is targeted to the first antenna element. The combining circuitry 34 linearly combines besl 1, bes21, and bes31 to form one combined forward beam element signal 36 targeted for transmission from the first antenna element. The same is done with respect to each antenna element.

[0035] Transmit (TX) signal path circuitry 38 couples the combined forward beam element signals 36 to a transmit antenna array 42 as forward antenna element signals 40. The forward antenna element signals 40 may differ from the combined forward beam element signals 36 in terms of filtering, frequency, and power level. For example, in one or more embodiments, the TX signal path circuitry 38 provides a respective signal path for each combined forward beam element signal 36 that includes low-noise amplification, filtering, frequency shifting, e.g., to a user downlink radiofrequency (RF) band, and power amplification, for formation of the corresponding forward antenna element signal 40.

[0036] Simultaneous transmission of the forward antenna element signals 40 from the respective antenna elements 44 of the TX antenna array 42 results in simultaneous formation of the forward user beams 12 represented in the forward antenna element signals 40. In the three- beam example, the radiated forward antenna element signals 46 result in far field signal superpositions — patterns of constructive and destructive wavefront interference — that forms the three forward user beams 12-1, 12-2, and 12-3.

[0037] Figure 2 illustrates example beamforming by a return beamforming system 50 of a satellite communications system. In the “return” direction, the satellite communications system receives transmissions from user terminals. The return beamforming system 50 is implemented onboard a satellite for example and employs the above-mentioned advantageous technique for reducing the computational burden of maintaining return user beams to account for changing conditions.

[0038] Although the satellite communications system may form many return user beams, Figure 2 illustrates an example case of three return user beams 52, shown as return user beam 52-1, return user beam 52-2, and return user beam 52-3. Each return user beam 52 has a corresponding beam footprint or return user beam coverage area 54 having a corresponding beam center 56. The return user beam coverage areas 54 may or may not be the same as the forward user beam coverage areas 14 in terms of the illuminated geographic areas.

[0039] One point to keep in mind is that, unlike the forward user beams 12, which are realized in free space via signal superpositions, the return user beams 52 are realized in the processing domain and do not exist in free space. In this regard, each return user beam 52 can be understood as a directional reception sensitivity, by which the return beamforming system 50 increases the signal-to-noise-and-interference ratio (SINR) for user uplink signals originating from user terminals operating in a corresponding one of the return user beam coverage areas 54. Thus, the return beam signal 58-1 contains user uplink signals from user terminals in a first return user beam coverage area 54, the return beam signal 58-2 contains user uplink signals from user terminals in a second return user beam coverage area 54, and the return beam signal 58-3 contains user uplink signals from user terminals in a third return user beam coverage area 54.

[0040] In operation a receive (RX) antenna array 60 receives a simultaneous plurality of user uplink signals 62 originating from respective ones among user terminals operating in the satellite service area. The user uplink signals 62 are received as respective signal superpositions on individual antenna elements 64 of the receive antenna array 60. Because of the geometric spacing of the antenna elements 64, each antenna element 64 receives a slightly different combination of user uplink signals 62, with each antenna element 64 outputting a respective one among a set of return antenna element signals 66. RX signal path circuitry 68 converts the set of return antenna element signals 66 into a set of received signals 70 by performing any one or more of filtering, frequency translation, and low-noise amplification.

[0041] Splitting circuitry 72 creates duplicates of the set of received signals 70, with one set for each return user beam 52 to be formed in the processing domain. In the simplified three-beam example, the splitting circuitry 72 duplicates the set of received signals 70 three times, outputting a first set of duplicated received signals 74-1 that will be used to form the first return beam signal 58-1, a second set of duplicated received signals 74-2 that will be used to form the second return beam signal 58-2, and a third set of duplicated received signals 74-3 that will be used to form the third return beam signal 58-3.

[0042] Forming the first return beam signal 58-1 involves weighting circuitry 76 applying a corresponding set of return beam weights RB W 1 to the first set of duplicated received signals 74-1, to obtain a corresponding set of return beam element signals 78-1. Each signal in the set of return beam element signals 78-1 is the weighted version of a corresponding one among the first set of duplicated received signals 74-1. The set of return beam weights RBW1 has a weight value for each signal in the set of duplicated received signals 74-1, with those values maintained by the adaptive beam weight calculator 30 using dynamically updated CSI 80. The CSI 80 comprises or is determined from a return channel matrix containing complex values representing the return propagation channels from each of one or more DUTs operating in each return user beam coverage area 54.

[0043] Combining circuitry 82 combines the first set of return beam element signals 78-1 to form the first return beam signal 58-1, and similar processing applies with combining the second set of return beam element signals 78-2 to form the second return beam signal 58-2 and combining the third set of return beam element signals 78-3 to form the third return beam signal 58-3. Again, there may be many more return user beams 52 and corresponding return beam signals 58 in actual implementation of the return beamforming system 50, and the return beamforming circuitry 84 will be “sized” accordingly. Also, as previously noted, the return user beams 52 exist in the processing domain, based on the described processing of the received signals 70, with each return beam signal 58 effectively extracting a subset of the user uplink signals received by the RX antenna array 60, with each subset being those user uplink signals originating from user terminals in a particular one among the return user beam coverage areas 54 that logically subdivide the larger satellite service area.

[0044] Each of one or more satellites in a satellite communications system may incorporate the transmit beamforming system 10 and/or the return beamforming system 50, to perform forward and/or return beamforming. Rather than using “and/or” explicitly, any reference hereafter to “A or B” shall be understood to mean only A, only B, or both A and B, unless otherwise stated, or unless the mutually exclusive use of “or” is clear from the context.

[0045] One or more elements used in forward beamforming may be reused in return beamforming and vice versa. For example, the transmit antenna array 42 may be the same as the receive antenna array 60, although there may be different sets of antenna element feeds for the forward and return directions. Different signal frequencies or polarizations may be used in forward beamforming as compared to return beamforming.

[0046] Figure 3 illustrates an example satellite communications system 100 in the context of forward beamforming. The satellite communications system 100 includes a ground segment 102 and a space segment 104.

[0047] Elements of the ground segment 102 include one or more satellite access nodes (SANs) 106 and one or more ground nodes 108, which provide processing of forward and return traffic for user terminals served by the satellite communications system 100. For example, the satellite communications system 100 receives incoming user traffic 110 from one or more external networks 112, such as the Internet. The ground node(s) 108 identify the individual user terminals targeted by the incoming user traffic 110, determine the corresponding forward beam associations of those user terminals, and include in each forward beam signal 18 the user traffic targeted to the particular user terminals served by the corresponding forward user beam 12.

[0048] Thus, the ground node(s) 108 send one or more forward beam signals 18 to each SAN 106 and each SAN transmits a feeder uplink signal 114 that conveys the one or more forward beam signals 18 provided to the SAN 106. All such signals may be continuous in nature, meaning that the signals are generated and transmitted on an ongoing basis, for delivering dynamic streams of user traffic.

[0049] A satellite 120 receives the feeder uplink signal(s) via a feeder link antenna system 122, which may include one or more reflectors and corresponding sets of antenna feeds for receiving each feeder uplink signal 114. The satellite 120 includes a bus 124 that provides power distribution, spacecraft control, etc., and further includes a payload 126. The payload 126 includes, for example, the transmit beamforming system 10 introduced in Figure 1. Thus, the depicted user link antenna system 128 may comprise or otherwise include the transmit antenna array 42 shown in Figure 1, for simultaneous transmission of forward antenna element signals 40 formed from the forward beam signals 18 received by the satellite 120 via the one or more feeder uplink signals 114 incoming to the feeder link antenna system 122. [0050] In this way, the satellite 120 serves various user terminals 130 via corresponding forward user beams 132, with three forward user beams 132-1, 132-2, and 132-3, and three corresponding forward user beam coverage areas 134-1, 134-2, and 134-3, shown by way of example. Although labeled using the reference number “132,” the depicted forward user beams 132 may be understood as being consistent with the forward user beams 12 shown in Figure 1, in terms of how they are generated or otherwise formed. Thus, the adaptive beam weight calculator 30 is onboard the satellite 120 in one or more embodiments, maintaining the forward beam weights used to form the forward user beams 132 with reduced computational burden, according to the advantageous technique disclosed herein.

[0051] The satellite 120 in one or more embodiments is a bent pipe satellite providing nonprocessed signal paths that do not include signal demodulation and remodulation, but may include frequency translation, filtering, amplification, etc. With respect to the advantageous technique disclosed herein, the adaptive beam weight calculator 30 may be onboard the satellite 120 or may be implemented in the ground segment 102. In the latter case, the ground segment 102 transmits the beam weights to the satellite 120 for application in the satellite 120.

[0052] Figure 4 illustrates an aspect of the advantageous technique according to an example embodiment, in which the adaptive beam weight calculator 30 performs “full recalculations” of the beam weights at first intervals and, between each full recalculation, performs one or more “adaptations” based on defining second intervals within each first interval. The second intervals thus subdivide the first intervals and, although the adaptations do not produce optimal values for the beam weights, they produce near-optimal values and they therefore allow a longer first interval. The approach lowers the overall computational burden by reducing how often the adaptive beam weight calculator 30 must perform the optimal full recalculation of beam weights. [0053] In more detail, the full recalculations require inversion of the channel matrix representing the complex propagation channels between the satellite communications system 100 and the respective user terminals 130 and containing information about the path gains, phase shifts, and time delays experienced by the involved signals as they propagate through the environment. Each full recalculation involves, for example, a Minimum Mean Square Error (MMSE) optimization of the beam weights. Determining the beam weight values so as to optimally account for the complex propagation channels requires inversion of the channel matrix and the inversion operations entail a large number of floating point operations (FLOPs).

[0054] In typical applications, maintaining beam weights requires a relatively fast update interval, such as every 10 to 100 milliseconds. Performing a full recalculation for each update may therefore be impractical, at least in constrained computing environments such as may exist onboard satellites. The disclosed technique can be understood as performing these full recalculations at a slower rate and performing simpler adaptations between the full recalculations, with the simplifications based on transformation from the “aperture weight domain” in which beam weights are conventionally expressed.

[0055] Figure 5 illustrates an example set of beam weights (“BW”) in the aperture weight domain. The BW set is a matrix of complex values (“V”) corresponding to respective antenna elements in the geometric array of antenna elements within the involved antenna array. Such elements may be laid out in an X, Y grid, for example, with each antenna element thus having a unique pair of X and Y coordinates within the overall array. Figure 5 suggests this by showing the values as VXI.YI, V\2.Y2, and so on, with Figure 5 assuming R columns of antenna elements and 5 rows of antenna elements.

[0056] Figure 6 illustrates an example implementation of the first-interval/second-interval approach discussed with respect to Figure 4 and shows the beam-weight domain transformations used in one or more embodiments of the advantageous technique for maintaining a set of beam weights used to form a corresponding beam, with applicability to forward beams and return beams and with the understanding that the technique may be used in parallel with respect to any number of beams.

[0057] The set of beam weights is maintained by performing full recalculations on a recurring basis, according to a first periodicity, and by performing adaptations between the full recalculations. Each full recalculation updates the set of beam weights optimally. Each such optimal set is “saved” for use in performing the simpler adaptations. In this context, “B” refers to the beam weights expressed in the aperture weight domain, “W” refers to the beam weights as transformed into the beam strength domain, and Q refers to the gradients, with reference to the below gradient operation for transforming from the beam strength domain into the beam gradient domain:

where 9 and are unit vectors in the 9 and directions. Further, the inverse transformation from the beam gradient domain back to the beam strength domain is as follows:

[0058] Each optimally-calculated set of beam weights (full recalculation) is transformed into the beam strength domain, which expresses the beam weights in terms of beam angle — azimuthal (9) and elevational (0). This transformation involves applying a two-dimensional (2D) Fast Fourier Transform (FFT) to each optimally calculated set of beam weights, to obtain the corresponding expression of those weights in the beam strength domain. Then, these saved sets of beam weights in the beam strength domain are evaluated across successive first intervals to determine the changes. These changes are expressed in a beam gradient domain as A0 A0 At ' At’ and can be understood as capturing how the beam gradients of the beam being formed change over successive full recalculations.

[0059] Beam gradients, in the context of satellite spot beams, refer to the rate of change in the signal strength, or power density, as one moves away from the center of the beam. The beam gradient is a measure of how quickly the signal strength decreases when moving from the position corresponding to the peak strength of the beam towards the edges of the beam. Figure 7 illustrates example beam gradients — denoted by the gradient lines 150 — for a given forward user beam 132 having a corresponding beam center 152. If the beam center 152 is targeted to the location of a moving user terminal, for example, then the beam angles will change from one full recalculation to the next, to account for movement of the user terminal, and the rate and magnitude of these changes will be reflected in the beam gradient domain. As such, the adaptations between each full recalculation may be determined on an interpolative basis, which may be understood as predicting the incremental changes need for the beam weights in between the full recalculations.

[0060] Figure 8 illustrates a method 800 of operation by an apparatus, such as the adaptive beam weight calculator 30 introduced in Figure 1 , for maintaining beam weights for ongoing beamforming via a satellite of a satellite communications system. Here, the term “beam weights” shall be understood as referring to one or more sets of beam weights, or a linear combination of such sets, with each set of beam weights being the complex weighting values used to form a respective beam.

[0061] The method 800 includes: performing (Block 802) a full recalculation of the beam weights at first intervals, each full recalculation being based on a channel state matrix reflecting then current propagation channel measurements; transforming (Block 804) the beam weights as determined at each full recalculation into reference beam strengths having azimuthal and elevational components; and performing (Block 806) adaptations of the beam weights at subintervals of the first intervals, based on changes in the reference beam strengths observed across successive first intervals.

[0062] Figure 9 illustrates a method 900 that can be understood as representing example details performed in the context of the method 800. At each new first interval (YES from Block 902), the apparatus performs a full recalculation of the beam weights (Block 904) and computes and saves the corresponding reference beam strengths (Block 906). Saving the reference beam strengths means saving the just calculated values for a set of beam weights in the beam strength domain. Processing continues with subinterval processing following the full recalculation, with the apparatus at each new subinterval (YES from Block 908) performing an adaptation of the beam weights via gradient projection (Block 910). That is, the apparatus performs a linear or non-linear interpolation to update the values of the set of beam weights, with the interpolation being based on the observed rate of change of the beam gradients — i.e., based on how the beam contours are changing from across respective full recalculations. The two most recent recalculations may be considered or a running window of recalculations may be considered, e.g., with weightings to reflect recency. See the “gradient projection” operation shown in Figure 6, by which the beam weight adaptations are determined.

[0063] In at least one embodiment, the beam weights comprise a combination of respective sets of beam weights corresponding to respective beams among a plurality of beams being simultaneously formed. Correspondingly, transforming the beam weights as determined at each full recalculation comprises transforming each respective set of beam weights into a respective reference beam strength, such that performing adaptations of the beam weights at the subintervals comprises, for each subinterval, adapting each respective set of beam weights based on the changes observed in the corresponding respective beam strength, and combining the adapted reference sets of beam weights.

[0064] Each full recalculation is a joint calculation of the respective sets of beams weights, in one or more embodiments, for optimization of a signal performance metric, such as SINR, Bit Error Rate (BER), or Null-Depth. Each full recalculation is a Minimum Mean Square Error (MMSE) estimation, for example, where the full recalculations require inversion of the channel state matrix. Conversely, by operating in the beam gradient domain and relying interpolative estimations, the adaptions performed between the full recalculations do not require inversion of the channel state matrix.

[0065] The ongoing beamforming comprises, for example, beamforming from an antenna array of a satellite, wherein processing circuitry located either in the satellite or in a ground segment of the satellite communications network performs the full recalculations of the beam weights and the adaptations of the beam weights. In at least one embodiment, the ongoing beamforming comprises end-to-end beamforming from a plurality of terrestrial transmitters of a satellite communications system to a plurality of terrestrial receivers of the satellite communications system. Here, the satellite operates as a multi-path signal relay between the plurality of terrestrial transmitters and the terrestrial receivers, and processing circuitry in a ground segment of the satellite communications system performs the full recalculations of the beam weights and the adaptations of the beam weights.

[0066] In one or more embodiments of the method 800, the ongoing beamforming is forward link beamforming, for formation of a plurality of forward user spot beams, each forward user spot beam conveying user traffic for one or more respective user terminals within a corresponding forward user beam coverage area. Here, maintaining the beam weights for the ongoing beamforming accounts for any one or combination of: movement of the satellite, movement of the respective user terminals, and movement of interfering transmitters within an overall satellite service area encompassed by the plurality of forward user spot beams.

[0067] Alternatively, the ongoing beamforming is return link beamforming, for formation of a plurality of return user spot beams, each return user spot beam conveying user traffic from one or more respective user terminals within a corresponding return user beam coverage area. Here, maintaining the beam weights for the ongoing beamforming accounts for any one or combination of: movement of the satellite, movement of the respective user terminals, and movement of interfering transmitters within an overall satellite service area encompassed by the plurality of return user spot beams. Of course, in at least one embodiment, the involved satellite communications system performs the advantageous technique in both forward link and return link beamforming.

[0068] In terms of performing the full recalculation of beam weights on first intervals and performing the simpler adaptations on second intervals between respective first intervals, the adaptations are performed either as linear or non-linear adaptations. The linear/non- linear decision may depend on determining whether the changes in the reference beam strengths observed relative to a given first interval or given subset of first intervals are linear or non-linear. One or more embodiments include tracking the observed changes in reference beam strengths as a beam gradient that is determined or updated based at least in part on the observed changes between successive pairs of first intervals. The beam gradient may be a linear beam gradient. See, e.g., Figure 6 for beam gradient details. [0069] In at least one embodiment, performing the adaptations with respect to a given first interval comprises determining the beam gradient based on the reference beam strengths corresponding to the most recent full recalculation and at least the next most recent full recalculation, and, for each subinterval within the given first interval: computing a beam strength as a function of the beam gradient; transforming the computed beam strength into new beam weights; and setting the beam weights as the new beam weights. Setting the beam weights as the new beam weights includes applying the new beam weights in an ongoing beamforming process active in the satellite communications system. In other words, saying that the beam weights are adapted or fully recalculated can be understood as saying that the values used for ongoing beamforming are adapted or fully recalculated.

[0070] Figure 10 illustrates the satellite communications system 100 according to an embodiment in which the advantageous technique for maintaining beams weights with reduced computational burden involves the aforementioned end-to-end beamforming. Whereas Figure 1 illustrated the determination of beam weights corresponding to respective antenna elements of a transmit antenna array, such as may be carried onboard a satellite, with end-to-end beamforming, each one among a plurality of geographically distributed SANs 106 acts as a “transmitting element.” Hence, for beamformed transmission of a given forward beam signal in the end-to-end beamforming context, that signal is transmitted from each participating SAN 106 according to a SAN-specific beam weight.

[0071] In more detail, in end-to-end beamforming, the satellite 120 operates as a multipath bent-pipe relay between a plurality of SANs 106 participating in the end-to-end beamforming and the population of user terminals 130 served by the satellite communications system 100. For simplicity, only one forward user beam 132 is shown and only one user terminal 130 is shown within the corresponding forward user beam coverage area 134. However, it should be understood that end-to-end beamforming may be performed with many simultaneous forward user beams 132, with each beam serving a respective plurality of user terminals 130. Further, the advantageous technique applies directly in the context of end-to-end beamforming in the return direction, in which end-to-end beamforming is used to “form” return user beams in the processing domain.

[0072] The satellite 120 as depicted in Figure 10 is configured for end-to-end beamforming in the forward direction by including a plurality of forward transponders 160. Each forward transponder 160 is a non-processed (bent-pipe) signal path through the satellite 120. Each such signal path may include low-noise amplification, filtering, frequency translation, and power amplification. The input end of each forward transponder 160 couples with a respective receive antenna element 162 among a plurality of such elements.

[0073] Each receive antenna element 162 receives a unique signal superposition 164 comprising the superposition of beam- weighted forward uplink signals 166 transmitted by the respective SANs 106. The signal superposition 164 impinging on each receive antenna element 162 is unique because the uplink propagation paths between each receive antenna element 162 and each SAN 106 are different. Each forward transponder 160 transmits a forward downlink signal 168 corresponding to the unique signal superposition 164 received by the forward transponder via a respective transmit antenna element 170, i.e., each forward transponder 160 relays the signal superposition 164 received by the forward transponder 160, with power amplification and possibly with filtering and frequency translation, for transmission in a selected downlink frequency band.

[0074] The far-field superpositions of the radiated forward downlink signals 168 result in one of more forward user beams 132. Each such forward user beam 1 .32 is represented in the satellite communications system 100 by a corresponding set of beam weights, as described above. In the end-to-end beamforming context, each beam weight in the set of beam weights for a given forward user beam 132 represents the weighting of the corresponding forward beam signal as transmitted by a respective one among the SANs 106. That is, rather than each beam weight in the set of beam weights corresponding to transmission from a respective antenna element in an antenna array, each beam weight corresponds to transmission by a respective one of the SANs 106.

[0075] To appreciate this, Figure 10 shows the ground node(s) 108 receiving incoming user traffic 172 from the one or more external networks 112, with the ground node(s) 108 identifying the user terminals 130 targeted by the incoming traffic 172, identifying which forward user beams 132 the targeted user terminals 130 are assigned to, and correspondingly directing the user traffic to ongoing generation of forward beam signals 174. Here, each forward beam signal 174 is transmitted as a corresponding forward user beam 132 and thus contains the forward user traffic to be delivered to the user terminals 130 served by the corresponding forward user beam 132.

[0076] The adaptive beam weight calculator 30 in this embodiment maintains the set of beam weights corresponding to each forward user beam 132 and provides these sets of beam weights to a beamformer 176 and does so using the advantageous technique disclosed herein. The beamformer 176, which comprises dedicated or programmatically configured processing circuitry, performs the previously described splitting/duplicating of each forward beam element signal, to obtain corresponding sets of beam element signals 178, which are weighted for transmission by the respective SANs 106.

[0077] Taking a single forward beam signal 174 as an example and assume that there are M SANs, where M is an integer greater than two (and possibly much greater). The forward beam signal 174 is split into M duplicate signals, each duplicate signal destined for transmission by a respective one of the M SANs. Thus, to form the corresponding set of M beam element signals 178, each one of the M duplicate signals is weighted by a SAN-specific weight among a set of M forward beamforming weights. These resulting weighted signals are simultaneously transmitted by the M SANs, resulting in particular signal superpositions 164 at the satellite 120, which in turn result in particular downlink signal superpositions providing for formation of the corresponding forward user beam 132.

[0078] The channel matrix in the end-to-end beamforming context encompasses the end-to- end forward propagation channels, which include multipath uplink propagation from the M SANs 106 to the receive antenna elements 162, propagation through each forward transponder 160, and multipath downlink propagation from respective ones of the transmit antenna elements 170 to respective ones of the user terminals 130 served by the satellite communications system 100. In addition to such arrangements being additionally or alternatively applied in the return direction, the advantageous technique applies to a wide variety of beamforming arrangements, including ground-based beamforming (GBBF), of which end-to-end beamforming is a particular species, and onboard beamforming such as previously discussed in the context of Figure 3.

[0079] As noted, onboard beamforming has at least two flavors, including a first approach where the satellite 120 includes the adaptive beam weight calculator 30, meaning that the satellite 120 calculates the beam weights, and a second approach where the ground segment 102 includes the adaptive beam weight calculator 30, meaning that the ground segment 102 calculates the beam weights and forwards them to the satellite 120. These variations also exist in the return direction, with one example of return beamforming shown in Figure 11.

[0080] Merely as an example, Figure 11 illustrates three return user beams 180, labeled as return user beams 180-1, 180-2, and 180-3 having corresponding return user beam coverage areas 182-1, 182-2, and 182-3. User terminals 130 among the overall population of user terminals 130 served by the satellite communications system 100 transmit individual user uplink signals 184 and a receive antenna array 190 onboard the satellite 120 includes a plurality of individual antenna elements (not shown) that receive respective superpositions of the user uplink signals 184. The satellite 120 includes the return beamforming circuitry 84 shown in Figure 2. However, the adaptive beam weight calculator 30 may be onboard the satellite 120 or in the ground segment 102. In the latter case, the ground segment 102 provides the return beam weights to the satellite 120.

[0081] In either case, a feeder link transmit antenna 192 of the satellite 120 transmits one or more return beam signals 194 formed via return beamforming, to each of one or more SANs 106. In turn, the SANs 106 transfer the return beam signals 194 or the contents thereof as signals 196 provided to the ground node(s) 108, which processes them for recovery of outgoing user traffic 198.

[0082] Figure 12 illustrates an example apparatus 200 according to one embodiment, where the apparatus is configured for operation as an adaptive beam weight calculator 30 according to the advantageous technique disclosed herein. In at least some embodiments, the apparatus 200 performs at least some of the beamforming operations, e.g., it implements at least a portion of the transmit beamforming system 10 shown in Figure 1 or the return beamforming system 50 shown in Figure 2. In one example, the apparatus 200 is a computer node in the ground segment 102 of a satellite communications system 100. In another example, the apparatus 200 is onboard a satellite 120 and implemented as part of the payload 126.

[0083] The apparatus 200 includes processing circuitry 202, which comprises one or more microprocessors 204 and may include one or more hardware accelerators 206. Example hardware accelerators 206 include any one or more of Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), or Field Programmable Gate Arrays (FPGAs) that contain specialized hardware providing high-speed signal processing operations for matrix operations or other aspects of beam weight computation.

[0084] The apparatus 200 further includes storage 208. The storage 208 comprises one or more types of computer-readable media, such as any one or more types of volatile memory and one or more types of non-volatile memory. Examples include DRAM for program execution and working data storage at run time, and FLASH memory for non-volatile retention of configuration data, program code, etc. The storage 208 in one or more embodiments stores one or more computer programs 210 and data 212, which may include configuration data and working data. [0085] In at least one embodiment, the microprocessor(s) 204 are specially adapted to maintain beam weights for a satellite communications system 100 according to the advantageous technique disclosed herein, based on their execution of computer program instructions stored in the storage 208. Such operations may include instructions for the hardware accelerator(s) 206. [0086] Additionally, the apparatus 200 includes interface circuitry 214, such as an Ethernet interface or other communication interface. In at least one embodiment, the apparatus 200 uses the interface circuitry 214 to output beam weights as updated according to full recalculations and adaptations. The apparatus 200 may also use the interface circuitry 214 to receive CSI for ongoing channel estimation and corresponding beam weight determinations.

[0087] Broadly, then, in an example embodiment, the apparatus 200 is configured for maintaining beam weights for ongoing beamforming via a satellite 120 of a satellite communications system 100. The apparatus 200 includes interface circuitry 214 that is configured to receive CSI reflecting then current propagation channel measurements, and further includes processing circuitry 202. The processing circuitry 202 is configured to: (a) perform a full recalculation of the beam weights at first intervals, each full recalculation being based on a channel state matrix determined from the channel state information; (b) transform the beam weights as determined at each full recalculation into reference beam strengths having azimuthal and elevational components; and (c) perform adaptations of the beam weights at subintervals of the first intervals, based on changes in the reference beam strengths observed across successive first intervals. Of course, the processing circuitry 202 of the example apparatus 200 may be configured to perform any one or combination of the operations described herein for maintaining beam weights, such as described in the context of the methods 800 and 900.

[0088] Notably, modifications and other embodiments of the disclosed invention(s) will come to mind to one skilled in the art having the benefit of the teachings presented in the foregoing descriptions and the associated drawings. Therefore, it is to be understood that the invention(s) is/are not to be limited to the specific embodiments disclosed and that modifications and other embodiments are intended to be included within the scope of this disclosure. Although specific terms may be employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.

Claims

CLAIMS What is claimed is:

1. A method of maintaining beam weights for ongoing beamforming via a satellite of a satellite communications system, the method comprising: performing a full recalculation of the beam weights at first intervals, each full recalculation being based on a channel state matrix reflecting then current propagation channel measurements; transforming the beam weights as determined at each full recalculation into reference beam strengths having azimuthal and elevational components; and performing adaptations of the beam weights at subintervals of the first intervals, based on changes in the reference beam strengths observed across successive first intervals.

2. The method according to claim 1, wherein the beam weights comprise a combination of respective sets of beam weights corresponding to respective beams among a plurality of beams being simultaneously formed, and wherein transforming the beam weights as determined at each full recalculation comprises transforming each respective set of beam weights into a respective reference beam strength, such that performing adaptations of the beam weights at the subintervals comprises, for each subinterval, adapting each respective set of beam weights based on the changes observed in the corresponding respective beam strength, and combining the adapted reference sets of beam weights.

3. The method according to claim 2, wherein each full recalculation is a joint calculation of the respective sets of beams weights, for optimization of a signal performance metric.

4. The method according to claim 2 or 3, wherein each full recalculation is a Minimum Mean Square Error (MMSE) estimation.

5. The method according to any one of claims 1-4, wherein the full recalculations require inversion of the channel state matrix, and wherein the adaptions do not require inversion of the channel state matrix.

6. The method according to any one of claims 1-5, wherein the ongoing beamforming comprises beamforming from an antenna array of the satellite, wherein processing circuitry located either in the satellite or in a ground segment of the satellite communications network performs the full recalculations of the beam weights and the adaptations of the beam weights.

7. The method according to any one of claims 1-5, wherein the ongoing beamforming comprises end-to-end beamforming from a plurality of terrestrial transmitters of the satellite communications system to a plurality of terrestrial receivers of the satellite communications system, with the satellite operating as a multi-path signal relay between the plurality of terrestrial transmitters and the terrestrial receivers, and with processing circuitry in a ground segment of the satellite communications system performing the full recalculations of the beam weights and the adaptations of the beam weights.

8. The method according to any one of claims 1-7, wherein the ongoing beamforming is forward link beamforming, for formation of a plurality of forward user spot beams, each forward user spot beam conveying user traffic for one or more respective user terminals within a corresponding forward user beam coverage area, and wherein maintaining the beam weights for the ongoing beamforming accounts for any one or combination of: movement of the satellite, movement of the respective user terminals, and movement of interfering transmitters within an overall satellite service area encompassed by the plurality of forward user spot beams.

9. The method according to any one of claims 1-7, wherein the ongoing beamforming is return link beamforming, for formation of a plurality of return user spot beams, each return user spot beam conveying user traffic from one or more respective user terminals within a corresponding return user beam coverage area, and wherein maintaining the beam weights for the ongoing beamforming accounts for any one or combination of: movement of the satellite, movement of the respective user terminals, and movement of interfering transmitters within an overall satellite service area encompassed by the plurality of return user spot beams.

10. The method according to claim 1, further comprising, with respect to any given first interval or any given subset of the first intervals, performing the adaptations either as linear or non-linear adaptations, in dependence on determining whether the changes in the reference beam strengths observed relative to the given first interval or given subset of first intervals are linear or non-linear.

11. The method according to any one of claims 1-10, further comprising tracking the observed changes as a beam gradient that is determined or updated based at least in part on the observed changes between successive pairs of first intervals.

12. The method according to claim 11, wherein the beam gradient is a linear gradient.

13. The method according to claim 11 or 12, wherein performing the adaptations with respect to a given first interval comprises determining the beam gradient based on the reference beam strengths corresponding to the most recent full recalculation and at least the next most recent full recalculation, and, for each subinterval within the given first interval: computing a beam strength as a function of the beam gradient; transforming the computed beam strength into new beam weights; and setting the beam weights as the new beam weights.

14. The method according to claim 13, wherein setting the beam weights as the new beam weights includes applying the new beam weights in an ongoing beamforming process active in the satellite communications system.

15. An apparatus configured for maintaining beam weights for ongoing beamforming via a satellite of a satellite communications system, the apparatus comprising: interface circuitry configured to receive channel state information reflecting then current propagation channel measurements; and processing circuitry configured to: perform a full recalculation of the beam weights at first intervals, each full recalculation being based on a channel state matrix determined from the channel state information; transform the beam weights as determined at each full recalculation into reference beam strengths having azimuthal and elevational components; and perform adaptations of the beam weights at subintervals of the first intervals, based on changes in the reference beam strengths observed across successive first intervals.

16. The apparatus according to claim 15, wherein the beam weights comprise a combination of respective sets of beam weights corresponding to respective beams among a plurality of beams being simultaneously formed, and wherein, for transforming the beam weights as determined at each full recalculation, the processing circuitry is configured to transform each respective set of beam weights into a respective reference beam strength, such that performing adaptations of the beam weights at the subintervals comprises, for each subinterval, adapting each respective set of beam weights based on the changes observed in the corresponding respective beam strength, and combining the adapted reference sets of beam weights.

17. The apparatus according to claim 16, wherein each full recalculation is a joint calculation of the respective sets of beams weights, for optimization of a signal performance metric.

18. The apparatus according to claim 16 or 17, wherein each full recalculation is a Minimum Mean Square Error (MMSE) estimation.

19. The apparatus according to any one of claims 15-18, wherein the full recalculations require inversion of the channel state matrix, and wherein the adaptions do not require inversion of the channel state matrix.

20. The apparatus according to any one of claims 15-19, wherein the ongoing beamforming comprises beamforming from an antenna array of the satellite, and wherein the apparatus is located either in the satellite or in a ground segment of the satellite communications network.

21. The apparatus according to any one of claims 15-19, wherein the ongoing beamforming comprises end-to-end beamforming from a plurality of terrestrial transmitters of the satellite communications system to a plurality of terrestrial receivers of the satellite communications system, with the satellite operating as a multi-path signal relay between the plurality of terrestrial transmitters and the terrestrial receivers, and with the apparatus included in a ground segment of the satellite communications system.

22. The apparatus according to any one of claims 15-21, wherein the ongoing beamforming is forward link beamforming, for formation of a plurality of forward user spot beams, each forward user spot beam conveying user traffic for one or more respective user terminals within a corresponding forward user beam coverage area, and wherein maintaining the beam weights for the ongoing beamforming accounts for any one or combination of: movement of the satellite, movement of the respective user terminals, and movement of interfering transmitters within an overall satellite service area encompassed by the plurality of forward user spot beams.

23. The apparatus according to any one of claims 15-21, wherein the ongoing beamforming is return link beamforming, for formation of a plurality of return user spot beams, each return user spot beam conveying user traffic from one or more respective user terminals within a corresponding return user beam coverage area, and wherein maintaining the beam weights for the ongoing beamforming accounts for any one or combination of: movement of the satellite, movement of the respective user terminals, and movement of interfering transmitters within an overall satellite service area encompassed by the plurality of return user spot beams.

24. The apparatus according to claim 15, wherein, with respect to any given first interval or any given subset of the first intervals, the processing circuitry is configured to perform the adaptations either as linear or non-linear adaptations, in dependence on determining whether the changes in the reference beam strengths observed relative to the given first interval or given subset of first intervals are linear or non-linear.

25. The apparatus according to any one of claims 15-24, wherein the processing circuitry is configured to track the observed changes as a beam gradient that is determined or updated based at least in part on the observed changes between successive pairs of first intervals.

26. The apparatus according to claim 25, wherein the beam gradient is a linear gradient.

27. The apparatus according to claim 25 or 26, wherein, for performing the adaptations with respect to a given first interval, the processing circuitry is configured to determine the beam gradient based on the reference beam strengths corresponding to the most recent full recalculation and at least the next most recent full recalculation, and, for each subinterval within the given first interval: compute a beam strength as a function of the beam gradient; transform the computed beam strength into new beam weights; and set the beam weights as the new beam weights.

28. The apparatus according to claim 27, wherein, to set the beam weights as the new beam weights, the processing circuitry is configured to output the new beam weights via the interface circuitry, for application in an active beamforming process being performed by the satellite communications system.