Method and apparatus for communication using large beam MIMO phased arrays - Patents.com

Abstract

The present invention provides a method and device for long-range (e.g., satellite to ground) wireless communication. A two-dimensional antenna array is operated in pairs to generate interleaved beams that are elongated in one direction and aligned in another. These beams can be provided using a fixed analog beamforming array. An active beamformer can be used to form beams directed within each beam group. Different ports of the analog beamforming array can be operated to generate beams in different groups, or different bi-sector beams within the same group. A simplified parametric-based channel state information precoding operation that uses direction of arrival for MIMO processing is also provided.

Description

[Technical field]

[0001] The present invention relates generally to long-range wireless communications, for example between ground equipment and satellites, such as low-Earth orbit (LEO) satellites, and more particularly to MIMO phased antenna arrays, and related methods and apparatus, mounted on such satellites, for example. [Background technology]

[0002] [Cross reference] This application claims the benefit of and priority to U.S. Non-Provisional Patent Application No. 17 / 195,089, filed March 8, 2021, entitled “Method and Apparatus for Communications Using Massive Beam MIMO Phased Arrays,” the contents of which are incorporated herein by reference in their entirety.

[0003] [background] Various low-orbit satellite constellations are planned or about to be launched to support a variety of applications, such as improved global connectivity, high-speed and secure communications, and global Internet access. Of particular interest are high-throughput satellite (HTS) systems, which are capable of providing significantly higher data rates compared to traditional satellite services. These systems traditionally use multiple steerable spot beams aimed at targets to improve data capacity with high-gain transmit / receive antennas, which allow for better frequency reuse. However, traditional approaches to generate these high-gain steerable beams require large, complex, and expensive phased antenna arrays. Thus, large phased array approaches have both cost and physical limitations.

[0004] In recent years, high data rate mobile communications (between ground equipment) have been developed that employ large phased arrays and use precoding processes. These developments involve techniques such as massive multiple-input multiple-output (MIMO) communications. Using this method, ground stations can use a large number of antenna elements (electromagnetic radiators) to generate multiple steerable, independent, high gain beams. Massive MIMO as currently developed is well suited for high capacity ground-based cellular systems where transmitters and receivers are relatively close together.

[0005] One of the key components of current massive MIMO systems is the use of pilot signals, which require complex signal processing for each given target device. The pilot signals are communicated from the transmitter to the receiver, which provides feedback to the transmitter to support channel estimation, which is critical for a proper MIMO configuration. For terrestrial communications where the target devices are confined within cells of relatively small size (e.g., a few kilometers), a relatively successful approach using pilot signals has been developed. However, for space-to-ground communications, or networks spanning significant distances, this same approach may not provide sufficient signal strength to achieve the required processing performance. In addition, the use of such pilot signals is problematic due to the propagation delays involved over larger distances relative to the coherence time of the channel, and other factors specific to satellite networks, such as the high relative speed of the satellites. Furthermore, antenna arrays dedicated to use over long distances leave room for improvement.

[0006] There is therefore a need to provide a communication method and apparatus that overcomes or mitigates one or more limitations in the prior art.

[0007] This background information is provided to identify information believed by the applicant to be of possible relevance to the present invention. No admission is necessarily intended, nor should it be construed, that any of the preceding information is prior art to the present invention. Summary of the Invention

[0008] It is an object of embodiments of the present invention to provide a method, apparatus, and system for communication using a large beam MIMO phased array mounted on a satellite, such as a LEO satellite. Although the present invention is described primarily with respect to a satellite implementation, it should be noted that embodiments of the present invention can be incorporated into other communication devices, such as terrestrial, airborne, orbital, or space-based devices. Various aspects related to the present invention are described below. It should also be noted that embodiments of the present invention can be provided using a synergistic combination of multiple aspects related to the present invention as described herein. For example, embodiments of the present invention can include multiple antenna arrays, multiple beamformer arrays, and an active beamformer, where multiple beamformer arrays generate narrowly interleaved beam groups, and the active beamformer is used to steer the beams using a given beam group. In some embodiments, simplified CSI operations can also be used in combination with the above.

[0009] The present invention provides a multiple antenna (e.g., MIMO, or massive MIMO, etc.) method and apparatus that incorporates both analog and digital precoding (e.g., in conjunction with beamforming). This makes the method and apparatus more suitable for long-range communications, such as, for example, LEO and high-throughput satellite (HTS) communications applications. A simplified channel state information (CSI) processing approach can be used, facilitating simpler and faster processing compared to conventional massive MIMO phased array approaches.

[0010] According to some embodiments, a communication device is provided that includes one, two or more sets of antenna arrays. Each set of antenna arrays may be a pair of antenna arrays. The antenna arrays may be positioned in close proximity with an inter-element spacing of less than or equal to half the operating wavelength. Each one of each set of antenna arrays is configured and steered to generate a plurality of beams. Each beam may have a cross-section that, when combined, assumes an elongated (e.g., elliptical, etc.) shape. The elongated shape may be significantly narrower in width than in length. The shape may be elongated in a primary direction of movement of the communication device (e.g., the direction of a satellite's orbit, etc.). Different beams in a given antenna array are spaced apart from each other, and beams from different antenna arrays in the same set (e.g., pair, etc.) are interchanged with each other. Each set of antenna arrays is thereby configured to generate a plurality of beams that may be substantially contiguous, with the property that adjacent beams are generated by different antenna arrays in the same set of antenna arrays. This property may apply to all pairs of adjacent beams.

[0011] According to a further embodiment or any of the above embodiments, there is provided a bi-sector beamspace massive MIMO phased array communication device, for example comprising N sets of N antenna elements each, for a total of N 2The present invention includes a two-dimensional phased antenna array having elements. A plurality of panels each having such a phased antenna array may be provided. For ease of explanation, the array may be described as having N rows and N columns. For example, N may be equal to 8 or 16. Further, a plurality of beamformer networks are provided and coupled to the array elements. The beamformer networks collectively form a portion of the beamformer array. Each beamformer network may be implemented using analog radio frequency (RF) circuitry and may operate in a fixed configuration. Each beamformer network may be configured to generate N orthogonal beams, different beams of which are pointed in different (e.g., azimuth) directions. A subset of M, less than N, of such beams may be selected and used for communication. Two or more sets of antenna arrays (as described above) may be operated together to generate a plurality of substantially orthogonal elongated beams. Additional active beamformers may be used to operate multiple sets of antenna arrays together. Pairs or multiple antenna arrays can be operated together to generate smaller tracking beams by combining beams from a given group or from multiple common direction clusters. The tracking beams can be steered in a given (e.g., elevation) direction. In various embodiments, beam steering and beam tracking along the long axis of an elongated beam can be implemented using linear massive MIMO processing.

[0012] According to a further embodiment or any of the above embodiments, the beamformer array includes a plurality (e.g., N / 2 rows, etc.) of hybrid combiners as well as N beamformer networks, where N is the number of rows of antenna elements in the array. Each beamformer network can be a Butler matrix or similar beamformer. In such an embodiment, each beamformer network can have N beam ports and N antenna ports operatively coupled to respective rows of antenna elements in the phased array. The N beamformer networks can be implemented in parallel, each coupled to a different row of antenna elements. A pair of beamformer networks can be coupled together using a row of hybrid combiners and operated together to generate N forward and N backward beams in a bi-sector fashion.

[0013] According to a further embodiment or any of the above embodiments, the beamformer array includes N / 2 specialized beamformer networks, where N is the number of rows of antenna elements in the array. Each beamformer network is coupled to two (e.g., adjacent) rows of antenna elements. Each beamformer network includes a number of beam ports that is less than the total number of antenna elements (and corresponding antenna ports) in the two rows. For example, the number of beam ports may be approximately one-quarter of the total number of antenna ports. The beamformer networks in a given array may all be similar or substantially identical. However, two different configurations of the beamformer network (e.g., using different phase shifter settings) may be provided and used in two different complementary antenna arrays to generate sets of different beams that are interleaved together. Hybrid combiners configured and operated to form forward and backward pointing (bi-sector) beams may also be used in such embodiments.

[0014] According to a further embodiment or any of the above embodiments, there is provided a communication device configured to generate (at least) two orthogonal and independent steerable bi-sector beams. The beams are bi-sector in the sense that one of the beams is directed to the front of a given area and the other beam is directed to the rear of the same given area. The two beams may overlap in the sense that the front and rear may overlap. In some embodiments, the bi-sector beams are generated using a hybrid coupler as described above. The hybrid coupler may be a 90° hybrid coupler. Any amplitude and phase taper may also be used in the beamformer. The two bi-sector beams may have forward and rear pretilt angles of 30° each at the operating frequency. The pretilt angle is the angle relative to an axis perpendicular to the plane of the array. This presents an effective inter-element spacing of half a wavelength. In some embodiments, each of the two bi-sector beams is steerable from 0° to 45° with a radiation efficiency of at least 85%. Using a combination of two beams oriented at an angle close to the axis perpendicular to the plane of the array, coverage within half-power width (HPBW) can be achieved.

[0015] According to a further embodiment or any of the above embodiments, there is provided a communication device including two independent N×2N (e.g., N=8) orthogonal beamforming networks having low sidelobes and low beam combining factor (BCF).

[0016] According to further embodiments or any of the above embodiments, a simplified parametric-based channel state information (CSI) precoder, and associated methods, are provided that use direction of arrival of signals from target devices for massive MIMO processing.

[0017] According to an embodiment of the present invention, a wireless communication device is provided. The device includes first and second separate two-dimensional arrays each having a plurality of antenna elements, and first and second beamformer arrays (e.g., fixed analog arrays, etc.). The first beamformer array is operatively coupled to the first antenna array and configured to generate a first plurality of beams by operating the first antenna array as a phased array. The first plurality of beams is separated into a plurality of spaced-apart beam groups, e.g., spaced apart in an azimuth direction. Similarly, the second beamformer array is operatively coupled to the second antenna array and configured to generate a second plurality of beams by operating the second antenna array as a second phased array. The second plurality of beams is separated into a plurality of additional spaced-apart beam groups interleaved with the plurality of spaced-apart beam groups from the first array. Each beam group typically includes a plurality of overlapping beams, which can operate together for further beamforming. In various embodiments, each beamformer array includes a plurality of beamformer networks arranged in parallel. In such an embodiment, the beamformer array is configured to provide a set of beam ports, each of which is operatively coupled to a pair of adjacent rows of antenna elements via one of the beamformer networks, or via a coupled pair of beamformer networks.

[0018] According to further embodiments or any of the above embodiments, each of the beamformer networks is coupled to a different respective row of antenna elements, and each beamformer array includes one or more rows of hybrid couplers, each row of hybrid couplers being coupled to a respective pair of the plurality of beamformer networks, and the beam ports being ports of the hybrid couplers. In some further embodiments, different hybrid couplers steer beams in different ones of the beam groups, and each hybrid coupler includes a pair of ports, each one of the pair of ports steer different bi-sector beams in the same one of the beam groups.

[0019] According to a further embodiment or any of the aforementioned embodiments, each of the beamformer networks comprises a subset of the set of beam ports coupled to different respective pairs of adjacent rows of the antenna elements, the number of beam ports in the subset being less than the number of antenna elements in the adjacent row pairs. In a further embodiment, each of the beamformer networks comprises a plurality of hybrid couplers arranged in a series-parallel configuration, a plurality of power splitters respectively coupled to pairs of antenna elements, including one antenna element in each of the adjacent row pairs, and a plurality of phase shifters, at least some of the phase shifters being arranged between the power splitters and the antenna elements.

[0020] According to a further embodiment or any of the aforementioned embodiments, particularly in combination with where each beamformer network is coupled to a different respective pair of adjacent rows of the aforementioned antenna elements, the plurality of antenna elements and the plurality of additional elements are each arranged with an inter-element spacing that is equal to or less than half the operating wavelength to provide a closely spaced array.

[0021] According to further embodiments or any of the above embodiments, each beam group is composed of individual beams collectively having a respective combined cross section extending in a first (e.g., elevation) direction, the different beam groups being aligned parallel to each other in a second (e.g., azimuth) direction perpendicular to the first direction. In some further embodiments, the apparatus is carried on a satellite moving relative to the Earth, the first direction being oriented such that the respective combined cross sections are in a portion of the Earth's surface extending parallel to the satellite's direction of travel. In some embodiments, each beam group is a bi-sector beam group including one, two or more forward beams and one, two or more rearward beams. The forward beams collectively provide a first portion of said combined cross section, the first portion being located towards a first end of the combined cross section. The rearward beams collectively provide a second portion of the combined cross section, the second portion being located towards a second end of the combined cross section. The first end and the second end are opposite each other along a first direction, and the first and second portions may be (partially) overlapping or non-overlapping. The forward beam may be orthogonal to the rearward beam. The forward beam may be oriented at a different angle relative to the rearward beam.

[0022] According to further embodiments or any of the preceding embodiments, an active beamformer is provided, for example as part of the apparatus as described above. The active beamformer may be operably coupled to the first beamformer array. The active beamformer is configured to generate one or more tracking beams by at least partially combining a plurality of beams belonging to one of a plurality of spaced beam groups using digital precoding. In some further embodiments, each beam group is composed of individual beams collectively having a respective combined cross-section that is elongated in a first direction, and the active beamformer is configured to direct the tracking beam in the first direction. In some further embodiments, the apparatus includes, in addition to the first and second arrays, one or more additional antenna arrays and one or more additional beamformer arrays each operably coupled to a respective one of the additional antenna arrays. Each of the additional beamformer arrays is configured to generate an additional copy of the first plurality of beams by operating each one of the additional antenna arrays as an additional phased array. Each of these additional copies overlaps with the first plurality of beams. In such an embodiment, the active beamformer is further operably coupled to each of the additional beamformer arrays, and the active beamformer is configured to generate the one or more tracking beams by further combining the plurality of beams with one or more additional beams belonging to the additional replicas of the first plurality of beams.

[0023] According to a further embodiment or any of the above embodiments, each beam group is composed of individual beams collectively having a respective combined cross-section extending in a first direction. The different beam groups are aligned parallel to one another in a second direction perpendicular to the first direction. Further, the rows of antenna elements extend parallel to one another in the second direction and are spaced apart from one another in the first direction.

[0024] According to further embodiments or any of the above embodiments, the channel state estimator is provided, for example (but not necessarily) as part of the device as described above. The channel state estimator is configured to estimate a predominantly line-of-sight wireless communication channel state between the host device and the remote wireless communication device. The estimation is based on the direction of arrival of the signal transmitted by the remote wireless communication device. The estimation is further based on the geometry of the antenna arrays (e.g., the first antenna array and the second antenna array, etc.) using statistical channel state information and a parametric channel model. In some embodiments, when the channel state estimator forms part of the device as described above, the beams are mutually orthogonal and estimating the wireless communication channel state includes performing a plurality of separate linear precoding operations.

[0025] According to an embodiment of the present invention, there is provided a system including two or more devices, including one or more devices as described above, one or more target devices, or a combination thereof.

[0026] According to an embodiment of the present invention, there is provided a method for wireless communication involving operating the apparatus or system as described above or elsewhere herein.

[0027] The embodiments are described above in conjunction with the aspects of the invention in which they can be implemented. Those skilled in the art will understand that the embodiments can be implemented in conjunction with the aspect in which they are described, but can also be implemented with other embodiments of that aspect. Where the embodiments are mutually exclusive or otherwise incompatible with each other, the embodiments will be apparent to those skilled in the art. Some embodiments are described in connection with one aspect, but may also be applicable to other aspects, as will be apparent to those skilled in the art. [Brief description of the drawings]

[0028] Further features and advantages of the present invention will become apparent from the following detailed description when considered in conjunction with the accompanying drawings. [Figure 1a] FIG. 1 illustrates a steerable bi-sector (N×N) beamspace massive MIMO array in accordance with one embodiment. [Figure 1b] FIG. 13 illustrates a steerable bi-sector (N×N) beamspace massive MIMO array in accordance with another embodiment. [Diagram 2] 1 is a graph illustrating a typical beam pattern for orthogonal beams generated by a Butler matrix, according to one embodiment. [Diagram 3] 1 is a graph showing a beam pattern of a composite beam resulting from the superposition of N×N Butler orthogonal beams in accordance with one embodiment. [Figure 4] 1 is a graph showing a beam pattern of a first orthogonal composite beam spatially phased array, referred to herein as Array A, in accordance with one embodiment. [Diagram 5] 13 is a graph showing the beam pattern of a second orthogonal composite beam spatially phased array, referred to herein as Array B, in accordance with one embodiment. [Figure 6a] FIG. 1 illustrates an 8×16 orthogonal analog beamformer for an embodiment of phased array A. [Figure 6b] 6b is a table showing phase shift values ​​for the various phase shifters in FIG. 6a; [Figure 7a] FIG. 14 illustrates an 8×16 orthogonal analog beamformer for an embodiment of phased array B. [Figure 7b] 7b is a table showing phase shift values ​​for the various phase shifters in FIG. 7a. [Figure 8a] 13 is a graph showing 12 of 16 orthogonal beams with low beam combining coefficients for a phased array (A+B) combining phased array A and phased array B for coverage up to ±40 degrees along the azimuth axis Az in accordance with one embodiment. [Figure 8b]8b is a table showing effective excitation amplitudes and phase tapers for an 8×16 beamforming network according to the embodiment in FIG. 8a, with beams labeled L1 through L6. [Figure 8c] 8b is a table showing effective excitation amplitudes and phase tapers for an 8×16 beamforming network according to the embodiment in FIG. 8a, with beams labeled R1 through R6. [Figure 9] FIG. 1 illustrates a general coordinate system for a 2D massive MIMO phased array in accordance with one embodiment. [Figure 10] FIG. 1 is a conceptual diagram of a low-orbit VHTS massive MIMO phased array, according to one embodiment. [Figure 11] FIG. 1 illustrates a dual-panel orthogonal beam spatial massive MIMO phased array in accordance with one embodiment. [Figure 12] 1 is a graph showing the forward and rear radiation patterns resulting from a 90° hybrid that effectively splits the radiation pattern of a single element into two sector orthogonal beams. [Figure 13] 13 is a graph showing the array output pattern of a bi-sector orthogonal steerable beam over a range of ±45° in the elevation plane, excluding angles near 0° elevation, according to one embodiment. [Figure 14] 13 is a graph showing the combined pattern for two bi-sector steerable beams near 0° elevation angle according to an embodiment where the low earth orbit satellite is operating at 30 GHz at an altitude of 350 km. [Figure 15a] 1 is a table showing m-MIMO performance data for pilot signals using two 320 mm square array panels according to one embodiment. [Figure 15b] 1 is a table showing actual m-MIMO performance data for pilot signals using two array panels of 320 mm square, where the m-MIMO processing has a wider bandwidth and a larger number of array panels, such as 2×5.

[0029] It should be noted that throughout the accompanying drawings, like features are identified by like reference numerals. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0030] As used herein, the term "about" should be interpreted as including variations from the nominal value, for example, a variation of + / - 10% from the nominal value. It should be understood that such variations are always included in any given value provided herein, whether or not specifically stated.

[0031] The embodiments of the present invention provide methods and apparatus related to beam-space massive MIMO phased antenna arrays. Phased arrays can have high gain and can be steerable. Massive MIMO technology has not been implemented for satellite communications before. This is due to, for example, long transmission distances (and large delays) and high transmission losses. Such factors make it difficult to obtain CSI feedback for use in channel estimation, which is typically used in MIMO systems. The embodiments of the present invention involve the use of large phased arrays using massive MIMO processing for long-distance communications using a combination of analog and digital precoding. This can result in stronger pilot signals for CSI processing and greatly simplified CSI processing. The simplification can involve reducing two-dimensional digital precoding operations to a combination of separate, simpler linear processing operations. This is particularly well suited for applications in low-orbit high-throughput satellite (HTS) communications.

[0032] FIG. 1a illustrates a single antenna array 102 provided in accordance with an embodiment of the present invention. This can be referred to as an M steerable bi-sector (N×N) beam-space massive MIMO array. The illustrated antenna array is applicable, for example, to single-polarized applications, using, for example, single-polarized antennas (radiators). However, it should be understood that the present invention can be easily extended to dual-polarized applications by using dual-polarized radiators. The antenna array is a two-dimensional array having a number of elements 110, for example, arranged in N rows (such as row 105, for example) of N radiators each. For convenience of explanation and in various embodiments of the present invention, each row (oriented vertically in FIG. 1a) extends in the azimuth direction, and the rows are parallel to each other and spaced apart from each other in the elevation direction. However, it should be understood that other directions and orientations can also be used. The antenna array may be a planar array, or an array that can be arranged on a curved surface. In operation, a beam space of M orthogonal beams (or beam groups) (out of a possible N) can be formed, with each of the M beams pointed differently in the azimuth direction. In the currently illustrated embodiment, these beams or beam groups are generated in part using a plurality of beamformer networks 120 (referred to as N×N beamformers), each having N antenna ports and N beam ports. Although N beam ports are available, implying N beams, in various embodiments only M beams, which is less than N, are used, with the number M being selected based on, for example, scan angle requirements in the azimuth direction.

[0033] According to various embodiments, the inter-element spacing of the antenna array 102, as well as the inter-element spacing of the other antenna arrays in the device, is equal to or less than half the operating wavelength of the antenna array. As such, the antenna array can be a closely spaced array. The array can also be configured as a non-resonant array. Such an embodiment is particularly useful, for example, when implementing the beamformer network in Figures 6a and 6b.

[0034] Inter-element spacings of less than half a wavelength typically introduce high mutual coupling between antenna elements. Tightly coupled phased arrays are designed and operated with this high mutual coupling in mind. An example is the Tightly Coupled Dipole Array (TCDA), which has been recently introduced in the literature. See, for example, JT Logan, RW Kint, MY Lee, MN Vouvakis, "A New Class of Planar Ultra-Wideband Modular Antenna Arrays with Improved Bandwidth," IEEE Transactions on Antennas and Propagation, February 2018, Vol. 66, No. 2, pp. 692-701, and JJH Wang, "Traveling Wave Antenna Arrays (TWAA) with Multi-Octave Scanning Gain Bandwidth," 2016 IEEE International Symposium on Phased Array Systems and Technology (PAST), 2016.

[0035] A typical phased array is designed using independently resonating radiators that are either resonant in a specific frequency range or are narrowband resonators. In a tightly coupled phased array, the entire array behaves in a substantially uniform manner, although each radiator does not necessarily behave as a resonator in the range of operating frequencies.

[0036] More specifically, each beamformer network 120 is coupled to a different corresponding set (e.g., a row, or a pair of closely spaced rows, etc.) of N antenna elements 110 in the same antenna array 102. The beamformer network may be a Butler matrix, or a structure similar to a Butler matrix. In an alternative embodiment, a beamformer network as described with respect to Figures 6a and 7a is used. In such a case, a hybrid combiner 155 may also be included and used to form a bi-sector beam, for example, in the elevation direction. Such an embodiment is described in more detail elsewhere herein.

[0037] Beamformer networks are generally constructed using analog components and are fixed in operation, e.g., phase shifts are substantially uncontrollable or at least controllable only in a calibration capacity. Beamformer networks may employ, for example, power splitters / combiners, phase shifters, and hybrid combiners. Because the components of the beamformer network are fixed, i.e., uncontrollable, beam steering is generally accomplished using additional active beamformers, which are described elsewhere herein.

[0038] The beam ports of the beamformer network are coupled to hybrid (e.g., 3 dB) combiners 155. In a row Operatively coupled to the hybrids are 90° hybrid couplers, also referred to as quadrature hybrids, or simply hybrids. For clarity, in FIG. 1a, each row of hybrids is shown extending in the elevation direction, but this is merely for ease of illustration. The hybrid couplers are four-port directional couplers that provide equal power splitting / combining between the input and output ports. In some embodiments, the 90° hybrid couplers operate with a resulting 90° phase shift between the output ports. As shown, each hybrid has a first pair of ports that face towards the antenna and beamformer network, and a second pair of ports that face away from the beamformer network. The second pair of ports can be coupled to a separate active beamformer. Each row of hybrids is connected to two different pairs of beamformer networks by connecting the first pair of ports on the hybrids to the beam ports of the beamformer network. In various embodiments, two different beamformer networks connected to the same row of hybrids are adjacent to each other, for example, in the sense that they are coupled to adjacent rows of antenna elements in the array.

[0039] Thus, multiple beamformer networks are arranged in parallel and coupled to different respective sets of antenna elements. The line is , are connected to respective (adjacent) pairs of beamformer networks. Different beam ports of the beamformer network 120, and therefore different ports of the hybrid combiner 155, steer beams in different beam groups. Different beam ports of the same hybrid combiner steer different bi-sector beams in the same one of the beam groups due to such different ports directing RF energy in different directions to and from the beamformer network pair.

[0040] Multiple beamformer networks 120 coupled to the same antenna array, along with hybrid combiners 155 connected to the multiple beamformer networks, form part of an overall beamformer array. Each antenna array may be operatively coupled to its own corresponding beamformer array.

[0041] A beamformer array operatively coupled to the antenna array 102 is configured to generate a first plurality of beams by operating the antenna array 102 as a phased array. The beams may be transmit beams, receive beams, or a combination thereof. As will be described in more detail below, the first plurality of beams is separated into a plurality of spaced apart beam groups. That is, adjacent groups may be separated by a gap of a given width. Each beam group includes its own respective plurality of overlapping beams.

[0042] A second antenna array and a second beamformer array are provided and configured similarly to, but not identical to, the antenna array and beamformer array as described above. In particular, the second beamformer array is operatively coupled to the second antenna array. The second antenna array includes a plurality of additional antenna elements and is typically separate from the first antenna array. For example, the first and second antenna arrays can be spaced apart. The first and second arrays can be substantially identical in terms of element design and arrangement. The second beamformer array is configured to generate a second plurality of beams by operating the second antenna array as a second phased array. The second plurality of beams are separated into a plurality of additional spaced apart beam groups. In particular, the plurality of additional spaced apart beam groups are interleaved with the plurality of spaced apart beam groups in the first antenna array and the first beamformer. In various embodiments, two beam groups are interleaved. However, theoretically, three or more beam groups can be interleaved if the beamformer structure is configured to accommodate this configuration. The use of multiple separate arrays promotes beam orthogonality.

[0043] Additional antenna arrays and corresponding additional beamformer arrays may also be provided, each operatively coupled to a respective one of the additional antenna arrays. Each of the additional beamformer arrays is configured to generate additional copies of the first or second plurality of beams. This is accomplished by operating the additional arrays as further phased arrays in the same manner as described above with respect to the first and second beamformer arrays and the corresponding arrays. The additional arrays may be substantially identical to the arrays already described above. Each of the additional copies overlaps (e.g., coincides with or is directed in the same direction as) the first or second plurality of beams. As will be described elsewhere herein, multiple beamformer arrays, e.g., a first beamformer array with copies of overlapping / coincident beams and one, two, or more additional beamformer arrays, may be operated together using an active beamformer to generate steerable tracking beams. These steerable tracking beams may be more tightly focused and controllable to direct RF energy to communication with a specific target device.

[0044] In various embodiments, each beam group is composed of individual beams collectively having respective combined cross sections that are elongated in a first direction, which may be an elevation direction. The different beam groups are aligned parallel to one another in a second direction, which may be perpendicular to the first direction and may be an azimuth direction. In some embodiments, for example, when the device is carried on a satellite moving relative to the Earth, the first direction is oriented such that the respective combined cross sections are part of the Earth's surface, and the combined cross sections are elongated parallel to the satellite's orbital direction.

[0045] The beam groups in the second beamformer array may be wide enough to substantially fill the gaps between the beam groups in the first beamformer array, and thus, in various embodiments, the beams in the first and second beamformer arrays (i.e., the beam groups and the additional beam groups) may together form an overall set of substantially contiguous beams that are aligned parallel to one another.

[0046] FIG. 11 shows an example of two interleaved sets of beams that are elongated in a first (e.g., elevation) direction and aligned parallel to one another in a second (e.g., azimuth) direction. The elongated cross-section can be elliptical, as shown, or another shape. Each ellipse 1120 and 1130 corresponds to a beam. The unshaded ellipse 1120 corresponds to a plurality of spaced beams in one of the arrays, e.g., array A1, while the shaded ellipse 1120 corresponds to a plurality of spaced beams in another complementary array, e.g., array B1. Arrays A2 through Am can also generate beams corresponding to the unshaded ellipse 1120, and arrays A2 through Am can also generate beams corresponding to the shaded ellipse 1130.

[0047] In various embodiments, each beam group includes at least two different bisector beams. A first one of the bisector beams is directed toward a first (e.g., forward) portion of the combined cross section (e.g., toward the bottom of the ellipse 1120 or 1130 in FIG. 11 ), and a second one of the bisector beams is directed toward a second (e.g., rear) portion of the combined cross section (e.g., toward the top of the ellipse 1120 or 1130 in FIG. 11 ). Thus, the bisector beams are directed in different directions, but typically overlap. Each group can include multiple copies of each of the bisector beams.

[0048] Thus, each beam group may be a bi-sector beam group having one, two or more forward beams and one, two or more rear beams. The forward beams may be substantially identical to each other and the rear beams may be substantially identical to each other. The forward beams collectively provide a first portion of the combined cross section of the group. This first portion is directed toward a first end of the combined cross section. Similarly, the rear beams collectively provide a second portion of said combined cross section. This second portion is directed toward a second end of the combined cross section. The first and second ends are opposite each other along a first direction, and the first and second portions may overlap. Alternatively, the first and second portions may not overlap, although this may affect beamforming options if it is desired to form a beam directed toward a central portion of the combined cross section (e.g., using an active beamformer). Since the beams all originate from substantially the same location (as seen by a distant target), directing the beams to different locations in different ways is performed by varying the angle of the beams. For example, the front beam is directed at a different angle relative to the rear beam.

[0049] In various embodiments, the different beams are orthogonal. For example, the forward beam may be orthogonal to the rear beam. Additionally or alternatively, the different beams may be orthogonal to each other. Orthogonality is used to describe spatial orthogonality in beamforming, for example, as follows: For advanced communications using multi-beam transmission, the excitation of the phased array must satisfy the following conditions for optimal performance:

[0050]

number

[0051]

number

[0052] Where:

[0053]

number

[0054] Equation (1) above defines the port orthogonality between the two beams, where L i and R i represents the excitation of two complex beams. If the two ports are perfectly orthogonal, this sum will be zero. Typically, it is desirable to keep this effect below -20 dB. Equation (2) above represents the beam coupling factor (BCF) between the two beams. Furthermore, the radiation pattern E R and E L It constrains the complex excitation so that coupling through is minimized. Typically, the BCF should be less than -15 dB.

[0055] Figure 1b illustrates an embodiment similar to Figure 1a. Apparatus 170 has six array panels operatively coupled thereto, corresponding to M=3 panel pairs. One such panel 175 is shown in more detail. Panel 175 generates a plurality of spaced apart beam groups, e.g., group 177. Another panel 185 generates a plurality of additional spaced apart beam groups, e.g., group 187 (dashed line), which is interleaved with group 177.

[0056] Digital beamforming is performed on rows of beam ports, e.g., row 157, to perform active beamforming in the elevation direction. There are three groups (pairs of rows) of antenna elements 190, each of which generates a beam group including forward beam 192 and rear beam 194. The groups overlap each other. Active beamforming in the elevation direction is performed at least in part using a combination of these three beam groups. Beamforming can also involve additional overlapping beam groups from other panels. Different rows of beam ports, e.g., rows 157 and 159, can be used to generate beams that are pointed in different directions in the azimuth direction. Different ports of the same hybrid, e.g., ports 162 and 164, can be used to generate different beam groups of the same beam group, forward beam 192 and rear beam 194.

[0057] Typically, an N×N beamformer such as a Butler matrix produces N orthogonal beam patterns with relatively high side lobes, which may be −13 dB. FIG. 2 shows, as an example, a typical beam pattern of orthogonal beams, such as may be produced by a 16×16 Butler matrix. The N orthogonal beam patterns may have high side lobes. The beams are labeled 1R through 6R for beams angled gradually outward to the right and 1L through 6L for beams angled gradually outward to the left.

[0058] The resulting beams of such an N×N beamformer are typically mutually orthogonal, but the beam spatial pattern can have a high beam combining factor (BCF). This can increase the interference between adjacent beams and reduce the overall capacity of the system. In FIG. 2, the beams are orthogonal with an intersection of −4 dB. However, the side lobe level (SLL) is only −13 dB. As a result, the beam combining factor between all beams is greater than −10 dB and the correlation coefficient is greater than 0.3. This problem can be solved by superimposing two or three of the N×N orthogonal beams to form a new set of orthogonal beams. In this case, the side lobe level can be reduced to less than −15 dB, which is more acceptable from the perspective of the BCF of the pattern.

[0059] FIG. 3 shows the beam pattern of a composite beam formed by superimposing N×N orthogonal beams generated by a Butler matrix beamforming network. Only a limited number of beams are shown, and these beams are included in the angular range of interest. The beams in FIG. 3 can be realized by superimposing pairs of beams in FIG. 2. For example, the two leftmost beams 6R and 5R in FIG. 2 can be superimposed to form the leftmost beam 6R-5R in FIG. 3. The composite beams are labeled in the format 1R-1L to 6L-5L, etc., to indicate what combination of beams in FIG. 2 each composite beam is formed from. The composite beams can be formed by combining multiple beam outputs with appropriate amplitudes and phases using an RF combining circuit. In FIG. 3, the composite beam SLL can be less than -15 dB, and the correlation coefficient can be less than 0.3, with less than 2 dB crossover.

[0060] However, such a composite beam is only orthogonal to beams that are two or more beams apart in beam space. Thus, in an embodiment of the present invention, two separate and / or independent phased arrays, which may be referred to as Array A and Array B, may be used to cover the entire beam space in the azimuth direction, with each array generating a beam pattern of alternating beams (as in FIG. 11). That is, Array A may generate a first set of spaced beams, and Array B may generate a second set of spaced beams that are interleaved with the first set of spaced beams. Array A may correspond to a first antenna array and a first beamformer array, and Array B may correspond to a second antenna array and a second beamformer array.

[0061] 4 is a graph showing the beam pattern of a first orthogonal synthetic beam spatially phased array, referred to herein as Array A, according to one embodiment. The synthetic beams are the same as the corresponding beams in FIG. 3, but only the beams actually formed by Array A are shown.

[0062] 5 is a graph showing the beam pattern of a second orthogonal composite beam spatially phased array, referred to herein as Array B, according to one embodiment. The composite beams are the same as the corresponding beams in FIG. 3, but only the beams actually formed by Array B are shown.

[0063] Together, array A and array B can form a complete set of 16×16 orthogonal beam spatial patterns with low BCF in the azimuth direction. Unlike conventional massive MIMO arrays, the output at each transmit / receive (Tx / Rx) port of the array according to one embodiment can have significantly higher signal strength due to the high-gain beam patterns. As a result, the array can cover a significantly larger range because each pilot signal from the beam port can have a significantly better signal-to-noise ratio (SNR) at the individual beam port.

[0064] FIG. 6a illustrates an 8×16 quadrature analog beamformer network for an embodiment of phased array A. Along the bottom are eight beam ports, labeled 1L through 8R. Along the top are two subsets of 15 antennas each, labeled A2 through B16 (A1 and B1 are not required in this implementation). Antenna elements A2 through A16 may correspond to elements in a first row of an antenna array, and antenna elements B2 through B16 may correspond to elements in a second row of the same antenna array, typically adjacent to the first row. Boxes (e.g., 610, etc.) represent hybrid couplers, and circles (e.g., 615, etc.) represent phase shifters. The “Y” junctions (e.g., 620, etc.) may be power splitters. The various phase shifts are given in multiples of φ, which may be approximately −11.25°. The phase shifts α2 to α16 and β2 to β16 are given in FIG. 6b.

[0065] Figure 7a illustrates an 8x16 quadrature analog beamformer network for one embodiment of phased array B. The structure and phase shifts are similar to Figure 6a. The phase shifts α2 to α16 and β2 to β16 are given in Figure 7b.

[0066] The beamformer networks in Figures 6a and 7a (with associated phase shifts as in Figures 6b and 7b) are implemented with a 7 dB amplitude taper, although this can be changed in other embodiments. The beamformer networks in Figures 6a and 7a (with associated phase shifts as in Figures 6b and 7b) are simplified compared to a conventional Butler beamformer network (or a pair of Butler beamformer networks coupled together using a row of hybrids). For example, each of the beamformer networks in Figures 6a and 7a requires 12 hybrid combiners each (24 in total), whereas a 16-beam Butler beamformer network would require a total of 4 x 8 = 32 interleaved hybrid combiners. Thus, a reduction of at least 8 hybrid combiners is realized. The reduction in complexity and number of components also translates into a reduction in associated losses. Furthermore, using the two beamformer networks in Figures 6a and 7a to generate a pattern of interleaved beams (e.g., in azimuth) results in beams that are orthogonal and have low correlation due to low sidelobes, a situation that is not easily achieved using conventional Butler-type beamformer networks.

[0067] The beamformer networks in Figures 6a and 7a can be used as beamformer networks in a given beamformer array. Each beamformer network is typically coupled to two different adjacent rows of antennas, where the antennas have an inter-element spacing of less than half the operating wavelength. The beamformer networks can include or be used with hybrids to generate bi-sector beams (e.g., forward and rearward beams in elevation) similar to those in Figure 1a. The beamformer networks in Figures 6a and 7a can be used with antenna arrays having an inter-element spacing of less than half the operating wavelength.

[0068] For example, two adjacent beamformer networks in Figures 6a and 7a can be connected to a row of hybrid combiners similar to the configuration in Figure 1a. In this case, the spacing of the array elements may be required to be less than a quarter of the operating wavelength. As a result, the effective spacing between adjacent beamformer networks remains less than half a wavelength. Thus, in some embodiments, only N / 4 hybrids are required for bi-sector array beamforming. Note that in some embodiments, only N / 2 azimuth beamformers may be required.

[0069] Thus, in some embodiments, as shown, for example, in FIG. 1a, each pair of rows of antenna elements is connected to a different beamformer network in two respective beamformer networks, and the two beamformer networks are interconnected using a row of hybrid combiners. In other embodiments, as shown, for example, in FIGS. 7a and 7b, each pair of rows of antenna elements is connected to a single interconnected beamformer network that is used to operate both rows of antenna elements. In both cases, the two rows of antenna elements are coordinately controlled for beamforming. The two different sets of antenna elements coupled to a pair of interconnected beamformer networks or to a single interconnected beamformer network may be two different rows of antenna elements. These rows may be rows of the same corresponding antenna array. These columns may run parallel to each other in one direction (e.g., azimuth) and spaced apart in another direction (e.g., elevation).

[0070] All of the embodiments as illustrated in Figures 1a, 6a, and 7a can be described as follows: Each beamformer array includes multiple beamformer networks arranged in parallel. In Figure 1a, each beamformer network is coupled to a single row of antenna elements, and pairs of beamformer networks are connected together using a row of hybrid couplers. In Figures 6a and 7a, each beamformer network is coupled to a pair of rows of antenna elements. The beamformer array in each case is configured to provide a set of beam ports. Each of the beam ports is operatively coupled to a pair of adjacent rows of antenna elements via one of the beamformer networks or via a coupled pair of beamformer networks. In Figure 1a, the beam ports are ports of the hybrid couplers that face away from the antenna array. In Figures 6a and 7a, the beam ports are directly integrated into the beamformer networks.

[0071] More specifically, in the case of Fig. 1a, each of the beamformer networks is coupled to a different respective row of antenna elements, and each beamformer array includes one or more rows of hybrid couplers. Each row of hybrid couplers is connected to a respective pair of beamformer networks, and as described above, the nominal beam ports in the array are the ports of the hybrid couplers. In a further embodiment, different hybrid couplers steer beams in different ones of the beam groups, and each hybrid coupler has a pair of ports. As described elsewhere herein, each one of the pair of ports steers different bi-sector beams in the same one of the beam groups.

[0072] More specifically, in the case of Figures 6a and 7a, each of the beamformer networks is coupled to a different respective pair of adjacent rows of antenna elements. Each beamformer network includes a subset of the set of beam ports, where the number of beam ports in the subset is less than the number of antenna elements in said pair of adjacent rows. For example, in Figures 6a and 7a, there are a maximum of 8 beam ports (e.g., 1L, 2L, ..., 8R), but there are 30 (or 32) antenna elements. In fact, the number of beam ports is less (e.g., almost half) than the number of antenna elements in a single row.

[0073] In even more detail, for Figures 6a and 7a, each of the beamformer networks can be characterized as follows: Each of the beamformer networks includes a plurality of hybrid couplers arranged in a series-parallel configuration. In the illustrated example, there are three rows of four hybrids each with phase shifters between the rows of hybrids. Each of the beamformer networks includes a plurality of power splitters each coupled to a pair of antenna elements, with one antenna element from each pair of rows. That is, each power splitter is coupled to a pair of antenna elements, one antenna element from each row. The top row of hybrid couplers is connected to the antenna elements via the power splitters. The beamformer networks also include a plurality of phase shifters. At least some of the phase shifters are disposed between the power splitters and the antenna elements. The phase shifters have different phase settings. There can be a plurality of cascaded sets of power splitters, with phase shifters coupled between the hybrid couplers and the power splitters and between different sets of cascaded power splitters.

[0074] In an embodiment, for example, corresponding to the implementation of the beamforming network in Fig. 6a-7b, 12 of 16 orthogonal and low BCF beams can be generated from the combined operation of Array A and Array B. This can define an azimuth range of up to ±40 degrees. The beams can have low beam crossing at more than 10 dB to generate low side lobes (less than -15 dB). This can result in a significantly lower BCF for each set of phased arrays. With such a beamforming network (BFN), only 24 hybrids are needed in total for the two arrays, instead of 32 hybrids each for a conventional 16x16 BFN. Furthermore, due to the arrays having an effective amplitude taper, only 15 radiating elements are needed instead of the typical 16 radiating elements.

[0075] 8a is a graph showing 12 of 16 orthogonal and low BCF beams in a combined phased array (A+B) of phased array A and phased array B for up to ±40 degree coverage along the azimuth (Az) axis, according to one embodiment. Solid lines represent beams of phased array A, e.g., beam R2 is accessible at port 2R in FIG. 6a, etc. Dashed lines represent beams of phased array B, e.g., beam R1 is accessible at port 1R in FIG. 7a, etc.

[0076] In an embodiment, each one of Array A and Array B may require two rows of the same array with different phase distributions in the final stage of the beamforming network (BFN). This is shown in both Figures 6a and 7a. Due to the effective amplitude taper and phase taper (e.g., as in Figures 8b and 8c), such an array architecture can generate orthogonal beams with low side lobes.

[0077] Figure 8b is a table showing effective excitation amplitude and phase tapers for an 8x16 beamforming network, where the beams are labeled L1 through L6 and correspond to the beams in Figure 8a, according to one embodiment. The amplitudes and phases for various antenna combinations (shown as in Figures 6a and 7a) are shown for the various beams.

[0078] Figure 8c is a table showing effective excitation amplitude and phase tapers for an 8x16 beamforming network, according to one embodiment, where the beams are labeled R1 through R6 and correspond to the other beams in Figure 8a. The amplitudes and phases for the various antenna combinations (shown as in Figures 6a and 7a) are shown for the various beams.

[0079] Having described various aspects relating to analog beamforming and antenna arrays, techniques for active (eg, digital) beamforming and channel state information estimation according to embodiments of the present invention will now be described.

[0080] In an embodiment, an analog (e.g., 8x16) BFN can effectively divide the azimuth coverage space into multiple orthogonal beam spaces with high gain beams. The beams can have fixed orientations or they can be selected to be orthogonal to the flight direction of the satellite on which the device is mounted. Target devices can then be assigned based on beam position in azimuth. However, using digital beamforming techniques, the target devices for each beam can be tracked along the flight path.

[0081] Channel estimation in a massive MIMO system can be achieved by estimating a channel transfer function using available channel state information (CSI). However, due to long propagation delays (e.g., between target devices and satellites), typically several milliseconds, and due to high moving speeds of satellites and target devices, it is not feasible to obtain instantaneous CSI at the transmitter side of a satellite system. In an embodiment, channel estimation can be accomplished based on a parametric channel model by using the direction of arrival (DOA) of the signal. Once the DOA of the incident signal is determined, which can be assumed in many cases, but mostly with line-of-sight transmission, the relative excitation phase of the radiators in the 2D phased array can be approximated. In an embodiment, this approach can be based on determining the response vector of the phased array. In an embodiment, a simple linear precoding process can be derived for 2D massive MIMO based on a common array configuration.

[0082] FIG. 9 illustrates a general coordinate system for a 2D massive MIMO phased array, according to an embodiment. An XYZ coordinate system is shown, where the X axis can extend, for example, in the azimuth direction of the antenna array, and the Y axis can extend, for example, in the elevation direction of the antenna array. The two-dimensional antenna array is arranged in the XY plane, the pointing angle of the reference antenna is illustrated in the (Uo, Vo) direction, and the direction 910 of the target device i is shown to have components of (Ui, Vi) when projected onto the XY plane. The angle θ i and Φ i points in the direction 910 of target device i using spherical coordinates.

[0083] For line-of-sight (LOS) transmission, the signal received by the satellite's 2D phased array antenna from a number P of target devices on the ground can be expressed as:

[0084]

number

[0085] Where: β i (t) is the channel gain for target device i, F(θ i ,Φ i , t) is the alignment factor in time for the satellite phased array, R(t-τ i ) is the delay τ i is a transmission signal from target device i having N(t) is noise, for example, additive white Gaussian noise (AWGN).

[0086] The frequency response corresponding to Y(t) is:

[0087]

number

[0088] Where: β i is the complex channel gain, F(θ i ,Φ i ) are the complex 2D array coefficients, R(φ i ) is the delay φ i a transmitted signal having W is the AWGN noise.

[0089] Based on FIG. 9, the general form for the array coefficients of a uniform 2D phased array can be expressed as follows:

[0090]

number

[0091] Where: x and y are the axes of the 2D array, k0 is the propagation constant, m is the number of array elements, d x and d y are the inter-element spacings in the x and y directions, u and v are the scan axes; (u o ,v o ) is the directivity angle of the reference antenna, (u i ,v i ) is the position of target device i given by:

[0092]

number

[0093]

number

[0094] By performing analog beamforming along the u scan axis, a set of orthogonal beams can be obtained with the following general (1 / x)sin(x) beam pattern (beam space in the u direction):

[0095]

number

[0096] The above equation can be expressed as a set of N orthogonal beams in the u direction, where the value for each beam depends only on the position of the target device i and the beam number m. Therefore, the signal received by the 2D phased array can be expressed as:

[0097]

number

[0098] The N beams in the u-scan direction are orthogonal and have low beam coupling, so they can be treated as spatially separated sectors. Therefore, the total received signal can be treated separately in each beam space m. That is,

[0099]

number

[0100]

number

[0101] The satellite moves at a constant high speed v sat In an embodiment where the terminal is moving at a relatively slow speed and the Doppler shift is a phase shift factor having a Doppler radian angle defined as

[0102]

number

[0103] That is,

[0104]

number

[0105] Each received beam signal is then:

[0106]

number

[0107] The beam steering vector can be defined as follows:

[0108]

number

[0109]

number

[0110]

number

[0111] Here, {A·S(t)} n is an n-dimensional linear precoder, which can be easily implemented using one of a variety of processing methods, including digital Fourier transform (DFT), multiple signal classification (MUSIC), and rotationally invariant signal parameter estimation technique (ESPRIT). Thus, for each target device i, complex massive MIMO processing can be reduced to n-dimensional linear precoding.

[0112] According to the above, a channel state estimator may be provided in the device disclosed herein. The channel state estimator is configured to estimate the state of a (mainly line-of-sight) wireless communication channel between the device and a remote wireless communication device. The estimation may be based on the direction of arrival of a signal (e.g., a pilot signal, etc.) transmitted by the remote wireless communication device, and further based on the geometry of the first antenna array and the second antenna array. The estimation may be obtained using statistical channel state information and a parametric channel model. The estimation may be performed as described above with respect to FIG. 9 and the disclosed equations.

[0113] The channel state estimation can be described as using statistical channel state information, e.g., using a parametric model. In this approach, two-dimensional pilot signal processing (as would be required for a conventional two-dimensional antenna array) is replaced by multiple linear processing operations. The processing operations are based on the direction of arrival of the (e.g., pilot) signal, in an operation that is configured based on the geometry of the array (as represented by the array coefficients). The direction of arrival points in one direction, e.g., in the elevation direction. This leads to increased speed and reduced simplicity. Because the pilot signal is received through a beam, rather than at the level of a single antenna element, the signal strength of the pilot signal is significantly increased compared to conventional methods. Because there are multiple separate beams that are substantially decoupled, the processing is simplified to a set of one-dimensional linear processing operations. In particular, the pilot signal processing is reduced to N parallel linear precoding operations using statistical CSI.

[0114] In some embodiments, estimating the wireless communication channel conditions comprises performing a plurality of separate linear precoding operations, as described above.

[0115] Also according to the above, an active (e.g., digital) beamformer can be provided coupled to one, two or more beamforming arrays. The active beamformer can be configured to generate one or more tracking beams by at least partially combining, using digital precoding, a plurality of beams belonging to one of a plurality of spaced beams, which are generated by the antenna array or antenna arrays by combining the beamforming array or beamforming arrays. The active beamformer can combine beams of a given beam group, or a set of overlapping beam groups, using different phases to generate a steerable beam.

[0116] Since the beam groups can have narrow cross sections and extend in a first (e.g., elevation) direction, beam steering can be substantially one-dimensional, i.e., in the first direction. This can simplify steering. Thus, in various embodiments, each beam group (of the beamformer array and antenna array) is composed of individual beams that collectively have a respective combined cross section that extends in the first direction. The tracking beam of the active beamformer can then be steered in the first direction.

[0117] FIG. 10 depicts a conceptual diagram of a low-orbit VHTS massive MIMO phased array mounted on a satellite 1000 shown in three different locations, according to one embodiment. The satellite can generate multiple orthogonal elliptical high-gain beams (or beams) 1010a, 1010b, 1010c, 1010d, 1010e, 1010f, 1010g, and 1010h in the azimuth direction 1030. This is shown in more detail in FIG. 11. Beams 1010a, 1010c, 1010e, and 1010g can be generated by one array or set of arrays, and beams 1010b, 1010d, 1010f, and 1010h can be generated by another array or set of arrays. A target device on the ground can then be tracked by the satellite using massive MIMO processing between each path. As an example, the cross-section of the beam (or group of beams) at the intersection point on the Earth may be, for example, 900 km long (corresponding to the elevation direction 1020) and 24 km wide (corresponding to the azimuth direction). The beam or group of beams is thus such that its cross-section is significantly longer than it is wide (e.g., in some embodiments at least 10 times longer, and in some further embodiments 30 to 50 times longer).

[0118] In one embodiment, the phased array can be arranged to have dual-panel orthogonal beam-space massive MIMO. Two sets of complementary array panels, Array A and Array B, can be used, which can generate alternating beams that are mutually orthogonal. Multiple sets of such two complementary array panels can be provided. To achieve high data rates through massive MIMO processing, a total of M pairs of panels can be used. Each array panel includes an antenna array and, optionally, a corresponding beamforming array. Alternatively, the beamforming array can be separate from the panel, provided sufficient coupling between the beamforming array and the antenna array can be performed.

[0119] FIG. 11 illustrates a dual-panel orthogonal beam spatial massive MIMO phased array according to an embodiment. A brief description of FIG. 11 has already been provided. However, it is noted here that FIG. 11 illustrates an apparatus having a chassis 1100, a first set of antennas and beamforming arrays A1, A2, A3, A4, ..., and Am, and a second set of antennas and beamforming arrays B1, B2, B3, ..., and Bm. In general, m can be any number greater than or equal to 1 or 2. Each of the antennas and beamforming arrays of the first set is configured to generate a copy of a spaced beam group 1120, while each of the antennas and beamforming arrays of the second set is configured to generate a copy of a spaced beam group 1130. In particular, the beam groups are elongated in a first direction and typically each include multiple beams. Furthermore, the multiple copies of the groups provide multiple overlapping beams. All overlapping beams corresponding to a given group can be steered together using an active beamformer to create a composite beam that is steerable in a first direction. Beam steering in a second direction can be achieved by using a different beam group.

[0120] According to an exemplary embodiment corresponding to FIG. 11, multiple (e.g., 16) elliptical beams per satellite can be provided. Beam steering or tracking can be performed along the major axis of each ellipse, massive MIMO (M-MIMO) steering can be performed, or both. Such a configuration can potentially improve handover complexity. In various embodiments, 8 to 16 parallel elliptical beams or beams can be provided per satellite through beamforming. 50 ports per beam (800 ports per satellite) can be provided for HPBW=2 degrees. M array panels per satellite can be used for M-MIMO or tracking processing to achieve effective beams with HPBW<1 degree (along the orbit). This can incorporate ultra-wideband active phased arrays with bandwidths >5 GHz, dual polarization, and wide scan angles >±45. Phased arrays with good active impedance and inter-element spacing <λ / 2 can define scan angles up to 45°.

[0121] A low-orbit satellite can move at a relatively high speed along a fixed path, typically hundreds to thousands of kilometers above the ground. For a user (target device) on the ground, a low-orbit satellite appears in view only for a short period (minutes) and can move at a constant speed along a predictable straight line direction. As such, the satellite beam may be required to track the target device on the ground as it moves. This may require the capability of wide-angle beam steering in the elevation direction. Since the satellite's viewing angle is expected to be about ±40°, the array beam should be capable of steering at angles greater than ±45°. To achieve this goal using massive MIMO precoding in the elevation direction, the output of each individual azimuth beam may be required to generate a pilot signal with an angular range of ±45° and sufficient strength to achieve the minimum SNR for massive MIMO processing. The output of each orthogonal beam at a given azimuth angle may then be combined to form a beam of a typically steerable phased array. However, due to the wide beam angle in elevation (±45°), this may not produce sufficient signal strength at each azimuth beamformer.

[0122] To improve the signal strength of the pilot signal and form two high-gain orthogonal steerable elevation beams, an embodiment of the present invention utilizes a 90° hybrid coupler (also called a hybrid) at the output of each orthogonal azimuth beam. An embodiment including a 90° hybrid is shown in FIG. 1a. Each 90° hybrid can effectively split the radiation pattern of a single element into two sectors, a forward radiating sector and a rearward radiating sector.

[0123] FIG. 12 is a graph showing the forward and rearward radiation patterns resulting from a 90° hybrid that effectively splits the radiation pattern of a single element into two sector orthogonal beams. The forward radiation beam pattern 1210 and the rearward radiation beam 1220 are mutually orthogonal, and they can be steered with a pretilt angle of almost or exactly 30° at the operating frequency where the array spacing is exactly half a wavelength. These two beam patterns can have suitable (e.g., just right) power half-widths (HPBWs) to allow beam steering to use array processing from 0° to 45° with a minimum radiation efficiency of 85%. The overall gain of each steerable beam can be increased by a factor of two (3 dB) since the coverage angle of each beam can be reduced by half.

[0124] Figure 13 is a graph showing the array output pattern of a bisectored orthogonal steerable beam over a range of ±45° in the elevation plane, excluding angles near 0° elevation. The beam is bisectored in that the rear beam (labeled NDIR) and the forward beam (labeled PDIR) are angularly separated.

[0125] Thus, when bi-sector forward and rear beams are used, each beam can be used by itself (or with the other forward or rear beam, respectively) to perform beam steering towards a forward or rear portion of a given area, e.g., towards one end of the ellipse in Figure 11. The forward and rear beams can be combined to perform beam steering towards a central portion of a given area, e.g., towards the center of the ellipse in Figure 11.

[0126] FIG. 14 is a graph showing a combination pattern for two bi-sector steerable beams near 0° elevation angle according to an embodiment in which a low-earth orbit satellite is operating at 30 GHz at an altitude of 350 km. At low elevation angles (less than 1 HPBW), the two beams can overlap. As a result, the radiation efficiency of each beam can be reduced due to high coupling. Although the two beams can have high beam coupling near 0° elevation angle, they can be excited separately without coupling loss because the two beams remain orthogonal. Combining the two beams can provide a substantially complete recovery in terms of efficiency loss in each individual beam.

[0127] In one embodiment, m-MIMO performance can be achieved using pilot signals using two array panels of 320 mm square.

[0128] FIG. 15a is a table showing actual m-MIMO performance data for pilot signals using two 320 mm square array panels.

[0129] Using m-MIMO processing with significantly wider bandwidth and a larger number of array panels can significantly improve performance.

[0130] FIG. 15b is a table showing actual m-MIMO performance data for pilot signals using two 320 mm square array panels, where the m-MIMO process has a wider bandwidth and a larger number of array panels, five pairs of panels.

[0131] In view of the above, a long-range wireless communication device is provided that can be adapted for use in a satellite system. For example, the device can be used to provide a VLEO satellite-based massive MIMO array for a direct link between a handset and a satellite, for example using Ku / Ka / V / Q band transmission. For a very low altitude satellite constellation at an altitude of 350 km, an embodiment of the present invention can be used to provide a direct link between a handset device and a satellite, for example using beam-space massive MIMO with BER=1e-05.

[0132] In various embodiments, pilot signals from target devices using lower order modulation codes such as BPSK / QPSK may be used while data communications may utilize higher order modulation codes. A parametric channel estimation method is developed herein based on DOA for fast linear precoding.

[0133] Embodiments of the present invention provide for a multi-beam M-MIMO arrangement that may reduce the complexity of antennas and beamforming arrays and may provide flexibility in improving overall performance.

[0134] Embodiments of the present invention can potentially be used in communications systems providing over 30 Gbps total throughput per beam per satellite using massive MIMO phased arrays with sizes on the order of 1.6 x 0.64 m, available spectral bandwidth of 5 GHz, and transmit power of 200 mW per panel.

[0135] In some cases, unlike conventional spot beam m-MIMO phased arrays, embodiments of the present invention use parametric channel estimation for fast signal processing. A convenient method for linear precoding for parametric channel estimation based on user DOA is disclosed herein. This channel estimation method is applicable to both TDD and FDD transmissions.

[0136] In some cases, compared to conventional massive MIMO systems, embodiments of the present invention provide significantly improved pilot signal strength, improving quality of service and enabling longer service distances. The pilots are monitored on a per-beam basis, rather than on a per-antenna element basis.

[0137] In some embodiments related to low-earth-orbit satellite applications, if the beam steering direction (e.g., elevation direction) of the array is set to be the same as the satellite's orbit direction, then beam tracking for a given user reduces to a simple linear one-dimensional precoding process once the appropriate azimuth beam is selected. Compared to a full two-dimensional spot beam phased array, this allows for significantly simpler and more energy-efficient multi-user operation.

[0138] Embodiments of the present invention are easily scalable. The capacity of the system can be increased by adding more array panels without a significant increase in complexity and processing time. Adding panels increases the degrees of freedom available for beamforming using active beamformers and M-MIMO processing in at least one (e.g., elevation) direction. This promotes narrower beamwidths, higher gain, or both, with a proportional reduction in beamwidth with the number of panels used. In one embodiment, the HPBW can be approximately 0.2° in the constellation extension direction (e.g., elevation direction).

[0139] It should be noted that although the present invention has been described as beams being generated, due to the principles of antenna reciprocity, beams may be implemented for use in transmission, reception, or both.

[0140] Although the invention has been described with reference to particular features and embodiments thereof, it is apparent that various modifications and combinations can be made without departing from the invention. Accordingly, the specification and drawings should be considered merely as illustrative of the invention as defined by the appended claims, and it is intended to cover any modifications, variations, combinations, or equivalents that are included within the scope of the invention.

Claims

1. 1. A wireless communication device, comprising: a first two-dimensional antenna array including a plurality of antenna elements; a second two-dimensional antenna array including a plurality of additional antenna elements and separated from the first two-dimensional antenna array; a first beamformer array operatively coupled to the first two-dimensional antenna array and configured to generate a first plurality of beams by operating the first two-dimensional antenna array as a phased array, the first plurality of beams being separated into a plurality of spaced-apart beam groups; a second beamformer array operatively coupled to the second two-dimensional antenna array and configured to generate a second plurality of beams by operating the second two-dimensional antenna array as a second phased array, the second plurality of beams being separated into a plurality of additional spaced-apart beams interleaved with the plurality of spaced-apart beams; Equipped with each beam group includes a plurality of overlapping beams; each beamformer array including a plurality of beamformer networks arranged in parallel, the beamformer array configured to provide a set of beam ports, each of the beam ports operatively coupled to a pair of adjacent rows of the antenna elements via one of the beamformer networks or via a coupled pair of the beamformer networks; the plurality of antenna elements and the plurality of additional antenna elements are respectively arranged with an inter-element spacing that is equal to or less than half an operating wavelength; Device.

2. 2. The apparatus of claim 1 , wherein each of the beamformer networks is coupled to a different respective row of antenna elements, each beamformer array further comprising one or more rows of hybrid combiners, each row of hybrid combiners connected to a respective pair of the plurality of beamformer networks, and the beam ports are ports of the hybrid combiners.

3. 3. The apparatus of claim 2, wherein different hybrid couplers steer beams in different ones of the beams, each hybrid coupler including a pair of ports, each one of the pair of ports steer different two-sector beams in the same one of the beams.

4. 4. The apparatus of claim 1, wherein each of the beamformer networks is coupled to a different respective pair of adjacent rows of the antenna elements and includes a subset of the set of beam ports, the number of beam ports in the subset being less than the number of antenna elements in the pair of adjacent rows.

5. 5. The apparatus of claim 4, wherein each of the beamformer networks includes a plurality of hybrid couplers arranged in a series-parallel configuration, and a plurality of power splitters respectively coupled to pairs of antenna elements include one antenna element in each of the adjacent rows, and a plurality of phase shifters, at least some of the phase shifters being disposed between the power splitter and the antenna elements.

6. 6. The apparatus of claim 1, wherein each beam group is comprised of individual beams collectively having respective combined cross sections extending in a first direction, and wherein different beam groups are aligned parallel to one another in a second direction perpendicular to the first direction.

7. 7. The apparatus of claim 6, wherein the apparatus is carried on a satellite moving relative to the Earth, and the first direction is oriented such that each combined cross section is a portion of the Earth's surface that is elongated parallel to the direction of motion of the satellite.

8. 8. The apparatus of claim 6 or 7, wherein each beam group is a bi-sector beam group including one, two or more forward beams that collectively provide a first portion of the bonding cross-section, the first portion facing a first end of the bonding cross-section, and each beam group further includes one, two or more rearward beams that collectively provide a second portion of the bonding cross-section, the second portion facing a second end of the bonding cross-section, the first end and the second end facing each other along the first direction, and the first and second portions being overlapping or non-overlapping.

9. The apparatus of claim 8 , wherein the forward beam is orthogonal to the rearward beam, and the forward beam is directed at a different angle relative to the rearward beam.

10. 10. The apparatus of claim 1, further comprising an active beamformer operably coupled to the first beamformer array, the active beamformer configured to generate one or more tracking beams by at least partially combining a plurality of beams belonging to one of the plurality of spaced apart beam groups using digital precoding.

11. 11. The apparatus of claim 10, wherein each beam group is comprised of individual beams collectively having a respective combined cross-section that is elongated in a first direction, and the active beamformer is configured to steer the tracking beams in the first direction.

12. one or more additional antenna arrays; and one or more additional beamformer arrays operatively coupled to respective ones of the additional antenna arrays, each of the additional beamformer arrays configured to generate additional replicas of the first plurality of beams by operating respective ones of the additional antenna arrays as further phased arrays, each of the additional replicas being associated with the one or more additional beamformer arrays overlapping the first plurality of beams. Further equipped with 12. The apparatus of claim 10 or 11, wherein the active beamformer is operatively coupled to each of the additional beamformer arrays, the active beamformer configured to generate the one or more tracking beams by further combining the plurality of beams with one or more additional beams belonging to the additional replicas of the first plurality of beams.

13. 13. The apparatus of claim 1, wherein each beam group is composed of individual beams collectively having a respective combined cross-section elongated in a first direction, different beam groups are aligned parallel to each other in a second direction perpendicular to the first direction, and the rows of antenna elements extend parallel to each other in the second direction and are spaced apart from each other in the first direction.

14. 14. The apparatus of claim 1, further comprising a channel state estimator configured to estimate a predominantly line-of-sight wireless communication channel state between the apparatus and the remote wireless communication device based on a direction of arrival of a signal transmitted by the remote wireless communication device, and further based on the geometric shapes of the first two-dimensional antenna array and the second two-dimensional antenna array using statistical channel state information and a parametric channel model.

15. 15. The apparatus of claim 14, wherein the beams are mutually orthogonal, and estimating the primarily line-of-sight wireless communication channel conditions comprises performing a plurality of distinct linear precoding operations.

16. 16. The apparatus of claim 1, wherein the first beamformer array and the second beamformer array are fixed analog beamformer arrays.

17. 1. A method for wireless communication, comprising: operating a first two-dimensional antenna array including a plurality of antenna elements arranged with an inter-element spacing that is equal to or less than half of an operating wavelength; operating a second two-dimensional antenna array, the second two-dimensional antenna array including a plurality of additional antenna elements separated from the first two-dimensional antenna array by an inter-element spacing that is equal to or less than half of an operating wavelength; operating a first beamformer array operatively coupled to the first two-dimensional antenna array and configured to generate a first plurality of beams by operating the first two-dimensional antenna array as a phased array, wherein the first plurality of beams are separated into a plurality of spaced-apart beams; operating a second beamformer array operatively coupled to the second two-dimensional antenna array and configured to generate a second plurality of beams by operating the second two-dimensional antenna array as a second phased array, the second plurality of beams being separated into a plurality of additional spaced-apart beams interleaved with a plurality of spaced-apart beams; Equipped with each beam group includes a plurality of overlapping beams; each beamformer array including a plurality of beamformer networks arranged in parallel, the beamformer array configured to provide a set of beam ports, each of the beam ports operatively coupled to a pair of adjacent rows of the antenna elements via one of the beamformer networks or via a coupled pair of the beamformer networks; method.

Images