Hybrid adaptive antenna array

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
26Forward
Citations 
0
Petitions 
1
Assignment
First Claim
1. A hybrid antenna array comprising:
 (a) a plurality of digital branches, each digital branch including;
(i) an analogue beamforming subarray, each subarray having a plurality of antenna elements, a phase shifter adapted to apply a phase shift to the signal from each antenna element, and a combiner adapted to combine the phase shifted signals;
(ii) a signal chain adapted to convert the output of the subarray to baseband;
(b) a digital processing module, including;
(i) an angle of arrival estimation submodule adapted to estimate an angle of arrival of a signal at the antenna elements;
(ii) a phase control submodule adapted to control the phase shift applied by each phase shifter depending on the estimated angle of arrival; and
(iii) a digital beamformer adapted to combine the baseband signals from the digital branches using a weight vector to form an output signal.
1 Assignment
0 Petitions
Accused Products
Abstract
Disclosed is a hybrid antenna array (100) comprising a plurality of digital branches (145), each digital branch including an analogue beamforming subarray (e.g. 1101), each subarray having a plurality of antenna elements (120), a phase shifter (130) adapted to apply a phase shift to the signal from each antenna element, and a combiner (e.g. 1351) adapted to combine the phaseshifted signals. Each digital branch also includes a signal chain (e.g. 1401) adapted to convert the output of the subarray to baseband. The hybrid antenna array also comprises a digital processing module (150), including: an angle of arrival estimation submodule (155) adapted to estimate an angle of arrival of a signal at the antenna elements; a phase control submodule (170) adapted to control the phase shift applied by each phase shifter depending on the estimated angle of arrival; and a digital beamformer (165) adapted to combine the baseband signals from the digital branches using a weight vector to form an output signal (180).
49 Citations
View as Search Results
METHOD AND APPARATUS FOR DOUBLING THE CAPACITY OF A LENSBASED SWITCHED BEAM ANTENNA SYSTEM  
Patent #
US 20130162478A1
Filed 12/21/2011

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

METHOD AND APPARATUS FOR GENERATING REFERENCE SIGNAL IN ANALOG/DIGITAL MIXED BF SYSTEM  
Patent #
US 20130258972A1
Filed 03/29/2013

Current Assignee
Samsung Electronics Co. Ltd.

Sponsoring Entity
Samsung Electronics Co. Ltd.

METHOD AND SYSTEM FOR CALIBRATING A TRANSCEIVER ARRAY  
Patent #
US 20150124634A1
Filed 10/03/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Method and apparatus for doubling the capacity of a lensbased switched beam antenna system  
Patent #
US 9,041,603 B2
Filed 12/21/2011

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

Method and system for digital cancellation scheme with multibeam  
Patent #
US 9,100,968 B2
Filed 05/09/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Method and system for explicit APtoAP sounding in an 802.11 network  
Patent #
US 9,100,154 B1
Filed 08/01/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Method and apparatus for generating reference signal in analog/digital mixed BF system  
Patent #
US 9,125,070 B2
Filed 03/29/2013

Current Assignee
Samsung Electronics Co. Ltd.

Sponsoring Entity
Samsung Electronics Co. Ltd.

System and method for colocated and cochannel WiFi access points  
Patent #
US 9,155,110 B2
Filed 06/24/2013

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Implementing transmit RDN architectures in uplink MIMO systems  
Patent #
US 9,154,204 B2
Filed 02/25/2013

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Method and system for calibrating a transceiver array  
Patent #
US 9,172,454 B2
Filed 10/03/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Method and system for supporting sparse explicit sounding by implicit data  
Patent #
US 9,172,446 B2
Filed 09/09/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Transmitter and receiver calibration for obtaining the channel reciprocity for time division duplex MIMO systems  
Patent #
US 9,236,998 B2
Filed 11/17/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

System and method for transmit and receive antenna patterns calibration for time division duplex (TDD) systems  
Patent #
US 9,294,177 B2
Filed 08/25/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Implementing multi user multiple input multiple output (MU MIMO) base station using singleuser (SU) MIMO colocated base stations  
Patent #
US 9,300,378 B2
Filed 08/04/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

System and method for selective registration in a multibeam system  
Patent #
US 9,332,519 B2
Filed 10/17/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Multibeam MIMO time division duplex base station using subset of radios  
Patent #
US 9,343,808 B2
Filed 05/06/2013

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Beamformer phase optimization for a multilayer MIMO system augmented by radio distribution network  
Patent #
US 9,344,168 B2
Filed 07/01/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

System and method for colocated and cochannel WiFi access points  
Patent #
US 9,497,781 B2
Filed 02/05/2014

Current Assignee
Magnolia Broadband Incorporated

Sponsoring Entity
Magnolia Broadband Incorporated

Method and apparatus for generating reference signal in analog/digital mixed BF system  
Patent #
US 9,531,517 B2
Filed 08/27/2015

Current Assignee
Samsung Electronics Co. Ltd.

Sponsoring Entity
Samsung Electronics Co. Ltd.

SYSTEM INCLUDING A HYBRID ACTIVE ARRAY  
Patent #
US 20170040710A1
Filed 08/14/2015

Current Assignee
the united states of america as represented by the secretary of the navy

Sponsoring Entity
the united states of america as represented by the secretary of the navy

System including a hybrid active array  
Patent #
US 9,742,075 B2
Filed 08/14/2015

Current Assignee
the united states of america as represented by the secretary of the navy

Sponsoring Entity
the united states of america as represented by the secretary of the navy

Method and apparatus for generating reference signal in analog/digital mixed BF system  
Patent #
US 10,320,540 B2
Filed 12/24/2016

Current Assignee
Samsung Electronics Co. Ltd.

Sponsoring Entity
Samsung Electronics Co. Ltd.

Frequencyscaled ultrawide spectrum element  
Patent #
US 10,333,230 B2
Filed 05/22/2018

Current Assignee
The Government Of The United States Of America As Represented By The Secretary Of The Navy

Sponsoring Entity
Mitre Corporation

Frequencyscaled ultrawide spectrum element  
Patent #
US 10,340,606 B2
Filed 05/22/2018

Current Assignee
The Government Of The United States Of America As Represented By The Secretary Of The Navy

Sponsoring Entity
Mitre Corporation

Intelligent antenna metamaterial method and apparatus  
Patent #
US 10,446,927 B2
Filed 05/11/2018

Current Assignee
Metawave Corporation

Sponsoring Entity
Metawave Corporation

Transceiver using hybrid beamforming and performing an antenna calibration method  
Patent #
US 10,707,974 B1
Filed 10/14/2019

Current Assignee
Industrial Technology Research Institute

Sponsoring Entity
Industrial Technology Research Institute

Method and system for improving robustness of interference nulling for antenna arrays  
Patent #
US 20080039146A1
Filed 01/18/2007

Current Assignee
Cisco Technology Incorporated

Sponsoring Entity
Cisco Technology Incorporated

Beamforming for spatial sidelobe cancellation and AMR direction finding  
Patent #
US 20080068266A1
Filed 11/23/2005

Current Assignee
Northrop Grumman Corporation

Sponsoring Entity
Northrop Grumman Corporation

Multiantenna transmission method and system  
Patent #
US 7,403,748 B1
Filed 04/07/2000

Current Assignee
WSOU Investments LLC

Sponsoring Entity
Nokia Corporation

Enhanced Connection Acquisition Using an Array Antenna  
Patent #
US 20080194204A1
Filed 01/04/2007

Current Assignee
Bellsouth Intellectual Property Corporation

Sponsoring Entity
Bellsouth Intellectual Property Corporation

Method for efficiently computing the beamforming weights for a large antenna array  
Patent #
US 7,414,578 B1
Filed 09/27/2006

Current Assignee
Rockwell Collins Inc.

Sponsoring Entity
Rockwell Collins Inc.

Methods and apparatus for increasing the effective resolving power of array antennas  
Patent #
US 7,183,974 B1
Filed 05/21/2004

Current Assignee
Harris Global Communications Inc.

Sponsoring Entity
ITT Manufacturing Enterprises Inc.

Wireless network apparatus and adaptive digital beamforming method thereof  
Patent #
US 20070046538A1
Filed 08/29/2005

Current Assignee
Accton Technology Corporation

Sponsoring Entity
Accton Technology Corporation

Control of an Integrated Beamforming Array Using NearFieldCoupled or FarFieldCoupled Commands  
Patent #
US 20070080888A1
Filed 10/18/2006

Current Assignee
TiaLinx Inc.

Sponsoring Entity
TiaLinx Inc.

Beamforming using subset of antenna array  
Patent #
US 20070093269A1
Filed 08/03/2005

Current Assignee
Redline Communications Inc.

Sponsoring Entity
PureWave Networks

Apparatus and method for a system architecture for multiple antenna wireless communication systems using round robin channel estimation and transmit beam forming algorithms  
Patent #
US 20070093274A1
Filed 10/23/2006

Current Assignee
Regents of the University of California

Sponsoring Entity
Regents of the University of California

Method and apparatus for adaptive beamforming in an antenna array system for wireless communications  
Patent #
US 20070249404A1
Filed 04/04/2007

Current Assignee
Communication Components Antenna Inc.

Sponsoring Entity
Tenxc Wireless Inc.

Common antenna array using baseband adaptive beamforming and digital IF conversion  
Patent #
US 20070263748A1
Filed 05/12/2006

Current Assignee
Northrop Grumman Systems Corporation

Sponsoring Entity
Northrop Grumman Systems Corporation

Wireless communication method and antenna system for determining direction of arrival information to form a threedimensional beam used by a transceiver  
Patent #
US 6,992,622 B1
Filed 12/29/2004

Current Assignee
Interdigital Technology Corporation

Sponsoring Entity
Interdigital Technology Corporation

Method and apparatus for beamforming based on broadband antenna  
Patent #
US 20060181456A1
Filed 03/17/2004

Current Assignee
Uniloc 2017 LLC

Sponsoring Entity
Koninklijke Philips N.V.

Antennas array calibration arrangement and method  
Patent #
US 20060192710A1
Filed 06/21/2004

Current Assignee
Provenance Asset Group LLC

Sponsoring Entity
Siemens AG

Calibration method and radio apparatus  
Patent #
US 20050070333A1
Filed 09/28/2004

Current Assignee
Panasonic Intellectual Property Corporation of America

Sponsoring Entity
Sanyo Electric Company Limited

Beamforming Architecture For MultiBeam Phased Array Antennas  
Patent #
US 20050110681A1
Filed 11/26/2003

Current Assignee
The Boeing Co.

Sponsoring Entity
The Boeing Co.

CCE calibration with an array of calibration probes interleaved with the array antenna  
Patent #
US 20040061644A1
Filed 09/10/2003

Current Assignee
Lockheed Martin Corporation

Sponsoring Entity
Lockheed Martin Corporation

Adaptive array antenna  
Patent #
US 20030016173A1
Filed 09/23/2002

Current Assignee
Shuichi Obayashi

Sponsoring Entity
Shuichi Obayashi

Adaptive beamforming method for smart antenna system  
Patent #
US 20030128160A1
Filed 12/17/2002

Current Assignee
LG Electronics Inc.

Sponsoring Entity
LG Electronics Inc.

PN code acquisition with adaptive antenna array and adaptive threshold for CDMA wireless communications  
Patent #
US 6,404,803 B1
Filed 10/24/2000

Current Assignee
Rateze Remote Mgmt. L.L.C.

Sponsoring Entity
NeoReach Inc.

Direction of arrival angle tracking algorithm for smart antennas  
Patent #
US 6,483,459 B1
Filed 04/05/2001

Current Assignee
Rateze Remote Mgmt. L.L.C.

Sponsoring Entity
NeoReach Inc.

Adaptive antenna array  
Patent #
US 4,780,721 A
Filed 03/21/1986

Current Assignee
Commonwealth Scientific and Industrial Research Organization

Sponsoring Entity
Commonwealth Scientific and Industrial Research Organization

10 Claims
 1. A hybrid antenna array comprising:
(a) a plurality of digital branches, each digital branch including; (i) an analogue beamforming subarray, each subarray having a plurality of antenna elements, a phase shifter adapted to apply a phase shift to the signal from each antenna element, and a combiner adapted to combine the phase shifted signals; (ii) a signal chain adapted to convert the output of the subarray to baseband; (b) a digital processing module, including; (i) an angle of arrival estimation submodule adapted to estimate an angle of arrival of a signal at the antenna elements; (ii) a phase control submodule adapted to control the phase shift applied by each phase shifter depending on the estimated angle of arrival; and (iii) a digital beamformer adapted to combine the baseband signals from the digital branches using a weight vector to form an output signal.  View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
1 Specification
This application is a National Stage of International Application No. PCT/AU2010/000091 filed Jan. 29, 2010, and which claims the benefit of Australian Provisional Application No. 2009900371, filed Feb. 2, 2009, the disclosures of which are incorporated herein by reference.
The present invention relates generally to antenna design and, in particular, to architectures and methods for adaptive antenna arrays.
Adaptive antennas are important subsystems for long range, mobile and ad hoc wireless communications and sensing networks. The most powerful adaptive antenna architecture is the digital beamforming array, where each antenna element has an associated signal chain to convert the received signal from RF (radio frequency) to baseband, and the baseband signals are processed by a digital beamformer using adaptive filtering techniques. Digital beamforming antenna arrays have the capability of generating many antenna patterns simultaneously, performing high precision beamforming such as nulling, and producing output signals with the maximum signal to noise and interference ratio (SNIR). A further advantage of the digital beamforming array architecture is that online calibration of different branches can be handled automatically by the digital beamformer. However, purely digital beamforming antenna arrays do have a major disadvantage. Since the cost of processing digital data is proportional to the bandwidth and the computational power required for digital signal processing increases at least linearly with the number of elements, a large digitally beamformed antenna array for wideband operation is simply too costly and impractical for most applications. Another problem with digital beamforming arrays is that, due to the limitation of physical space defined by the array element spacing, a pure digital beamformer is impractical at millimeterwave frequencies beyond approximately 55 GHz. The element spacing required for suppression of grating lobes at scan angles within 60 degrees is limited to 2.9 and 1.7 mm at operating frequencies of 55 and 95 GHz respectively, thus making it extremely difficult to physically place the signal chains behind the antenna elements.
The other principal adaptive array architecture is the analogue beamformer, in which “smart beams” are formed using RF or IF phase shifters on each antenna element in the array. In purely analogue systems, cost is a very weak function of bandwidth. Thus, for wideband adaptive antenna systems with a large number of elements, an analogue adaptive antenna array is much more economical than a purely digital adaptive array. However, an analogue beamformer has certain disadvantages. First, the calibration of a large analogue array is labourintensive and online recalibration is extremely difficult. Second, special means must be employed for beam tracking, thus increasing the complexity and the cost. Third, there is no access to the baseband signals of individual analogue branches because the phaseshifted IF or RF signals are combined before conversion to baseband, so the capacity to form selective beam patterns is limited. These problems are particularly severe for large analogue arrays where the imperfection and ageing of components can cause serious degradation in antenna performance.
Disclosed are arrangements which seek to ameliorate the above disadvantages by means of a hybrid adaptive antenna array, which comprises an array of antenna elements partitioned into multiple analogue beamforming subarrays and a digital beamformer on the subarray outputs, with a control path back to the analogue phase shifters. This reduces the size of the digital beamformer by a factor equal to the average number of elements in each subarray. Also disclosed are calibration methods for both the analogue and the digital beamformers in the hybrid adaptive array. In addition, disclosed are search and tracking methods for the hybrid adaptive array. The disclosed hybrid array provides an efficient means of producing large, highgain adaptive antenna arrays at relatively low cost.
According to a first aspect of the present disclosure, there is provided a hybrid antenna array comprising:
(a) a plurality of digital branches, each digital branch including:
 (i) an analogue beamforming subarray, each subarray having a plurality of antenna elements, a phase shifter adapted to apply a phase shift to the signal from each antenna element, and a combiner adapted to combine the phaseshifted signals;
 (ii) a signal chain adapted to convert the output of the subarray to baseband; and
(b) a digital processing module, including:
 (i) an angle of arrival estimation submodule adapted to estimate an angle of arrival of a signal at the antenna elements;
 (ii) a phase control submodule adapted to control the phase shift applied by each phase shifter depending on the estimated angle of arrival; and
 (iii) a digital beamformer adapted to combine the baseband signals from the digital branches using a weight vector to form an output signal.
According to a second aspect of the present disclosure, there is provided a method of estimating the angle of arrival of a signal at an antenna array comprising a plurality of digital branches and a digital beamformer, the method comprising:
(a) crosscorrelating the baseband signals from adjacent digital branches;
(b) for a plurality of candidate beams,
 (i) setting the phase shifts using the arguments of the crosscorrelation values; and
 (ii) computing the power of the output signal of the digital beamformer;
(c) determining the candidate beam giving the maximum output power; and
(d) estimating the angle of arrival from the crosscorrelation values of the determined candidate beam.
According to a third aspect of the present disclosure, there is provided a method of calibrating an antenna array comprising a plurality of digital branches and a digital beamformer, the method comprising:
(a) estimating the expected phase difference between the baseband signals from adjacent digital branches;
(b) determining the phase deviation of each digital branch as the difference between the expected phase difference and an actual phase difference between the baseband signals from a digital branch and an adjacent previously calibrated digital branch; and
(c) applying each phase deviation to the baseband signal from the corresponding digital branch.
According to a fourth aspect of the present disclosure, there is provided a method of calibrating an antenna array comprising a plurality of analogue branches, each analogue branch comprising an antenna element and a phase shifter, the method comprising:
(a) optimising the phase shifts applied by the phase shifters to an input signal to the array so as to obtain a maximumpower output signal from the array;
(b) determining expected values of the phase shifts using an angle of arrival of the input signal;
(c) determining the phase deviation for each analogue branch in the array as the difference between the expected phase shift and the optimised phase shift for the analogue branch; and
(d) applying each phase deviation to adjust the phase shift applied by the corresponding phase shifter.
In the drawings:
Where reference is made in any one or more of the accompanying drawings to steps and/or features, which have the same reference numerals, those steps and/or features have for the purposes of this description the same function(s) or operation(s), unless the contrary intention appears.
A hybrid array is said to have a planar configuration (a planar array) if all the subarrays are located in a single plane. A linear configuration (a linear array) is a special case of a planar configuration in which all the subarrays are located in a single line.
Each subarray 110m in
A planar array is defined as uniformly spaced if the number M of subarrays can expressed as M=M_{x}×M_{y}, where M_{x }and M_{y }are the numbers of subarrays arranged along the xaxis and the yaxis respectively. The ith element in subarray m of a uniformly spaced planar array is located at coordinates (X_{i,m }Y_{i,m}) such that
X_{i,m}=X_{i}+m_{x}d_{x}^{(s)},m_{x}=0,1, . . . M_{x}−1
Y_{i,m}=Y_{i}+m_{y}d_{y}^{(s)},m_{y}=0,1, . . . M_{y}−1
where m=m_{y}M_{x}+m_{x}, d_{x}^{(s) }and d_{y}^{(s) }are the horizontal and vertical spacings of the subarrays respectively, and (X_{i}, Y_{i}) is the location of the ith element of the subarray numbered m=0. For example, the planar array 230 is uniformly spaced with M_{x}=M_{y}=2 and d_{x}^{(s)}=d_{y}^{(s)}.
A uniformly spaced linear array is a special case of a uniformly spaced planar array with M_{y}=1, so M_{x}=M, m_{y}=0, and Y_{i,m}=Y_{i}(d_{y}^{(s) }is irrelevant). For example, the linear array 200 is uniformly spaced with M_{x}=2 and M_{y}=1.
A rectangular subarray has N=N_{x}×N_{y }elements, where N_{x }and N_{y }are the numbers of elements arranged along the xaxis and the yaxis respectively. The location (X_{i}, Y_{i}) of the ith element of the rectangular subarray numbered m=0 is given by
X_{i}=X_{0}+i_{x}d_{x}^{(e)},i_{x}=0,1, . . . N_{x}−1
Y_{i}=Y_{0}+i_{y}d_{y}^{(e)}i_{y}=0,1, . . . N_{y}−1
where i=i_{y}N_{x}+i_{x}, d_{y}^{(e) }and d_{x}^{(e) }are the element spacings along the yaxis and xaxis respectively, and (X_{0}, Y_{0}) is the location of the element numbered 0. For example, each subarray 240m in the planar array 230 is rectangular, with N_{x}=N_{y}=4 and d_{x}^{(e)}=d_{y}^{(e)}.
A linear subarray is a special case of a rectangular subarray with N_{y}=1, so N_{x}=N, i_{y}=0, and Y_{i}=Y_{0 }for all i(d_{y}^{(e) }is irrelevant). For example, each subarray 210m in the linear array 200 is linear, with N_{x}=4.
The planar array 230 is a uniformly spaced planar array comprising rectangular subarrays with d_{x}^{(s)}=N_{x}d_{x}^{(e) }and d_{y}^{(s)}=N_{y}d_{y}^{(e)}. The linear array 200 is also uniformly spaced, comprising linear subarrays with d_{x}^{(s)}=N_{x}d_{x}^{(e)}.
Each antenna element 120 belongs to an analogue signal branch, e.g. 115, which also includes a phase shifter, e.g. 130, which is independently controllable by an external signal 175, as described below. The phase shifter 130 alters the phase of the received signal from the antenna element 120 by a controllable amount. Each analogue branch 115 may also contain a low noise amplifier, an attenuator, a filter, and a frequency downconverter, using a local oscillator, to intermediate frequency (not shown). The phase shifter 130 is preferably implemented as a controllable or ‘switched’ delay in the radio frequency (RF) signal, the downconverted intermediate frequency (IF) signal, or the local oscillator.
The outputs of each analogue branch 115 in the subarray 110m are combined by a combiner 135m. The phases of each phase shifter 130 are controlled as described below such that each subarray 110m generates a beam in a desired direction, for example the angle of arrival of the RF signal incident on the hybrid array 100. The combination of RF signals from different analogue branches by the combiner 135m is preferably implemented by a conventional corporate network.
In alternative embodiments, the analogue beamforming in a subarray 110m is performed by a Rotman lens or a Butler matrix rather than a phased array as described above.
The output of each subarray 110m is passed to a chain 140m, which converts the RF or IF signal to baseband. The combination, e.g. 145, of a subarray 110m and a chain 140m is referred to herein as a “digital branch”. The M baseband signals from the digital branches 145 are processed in a digital processing module 150 comprising several submodules: a digital beamformer 165, a calibration submodule 160, an angle of arrival estimation submodule 155, and a subarray phase control submodule 170. The operation of each submodule will be described below.
The digital processing module 150 serves several purposes. First, the digital beamformer 165 ensures that the output signal 180 of the hybrid array 100 has a highest signal to noise and interference ratio (SNIR). Second, the digital processing module 150, through the subarray phase control submodule 170, provides the control signal 175 for each phase shifter 130 in each subarray 110m. The angle of arrival estimation submodule 155 plays a role in this function, as described below. Third, the digital processing module 150, through the calibration submodule 160, provides calibration both for the digital branches 145 and for each analogue branch 115 in the hybrid array 100 as described below. The preferred implementations for the digital signal processing module 150 are embedded firmware on Field Programmable Gate Arrays (FPGAs) or a dedicated ASIC, which is more suitable for mass production.
The digital beamformer 165 generates an output signal z(n) 180 at time n by applying weights w(n)=[w_{0}(n) w_{1}(n) . . . w_{M−1}(n)]^{T }to the (complexvalued) baseband signal vector x(n)=[x_{0}(n) x_{1}(n) . . . x_{M−1}(n)]^{T }output by the digital branches 145 as follows:
z(n)=w^{H}(n)x(n) (1)
where ^{H }indicates the conjugate transpose of a vector.
Forming the Digital Beam
The weights w(n) vary or adapt with time to cope with changing signal and/or system characteristics. There are a number of adaptive algorithms such as the least mean squares (LMS) algorithm and the recursive least squares (RLS) algorithm which can be employed to update the weight vector in a conventional digital beamformer. The method used by the digital beamformer 165 according to one embodiment uses a reference signal s(n) that was known to have been transmitted to arrive at time n. This knowledge is available during a training period for the hybrid array 100. The reference signal could be the header portion of a data packet. According to this embodiment, the digital beamformer 165 iteratively adjusts the weight vector as follows:
w(n)=αw(n−1)+(1−α)s*(n)x(n) (2)
where α is a “forgetting factor” that controls the effective window size of the time averaging, and the * superscript indicates complex conjugation.
In an alternative embodiment, to reduce the sensitivity of the digital beamformer 165 to the magnitude of the input signal vector x(n), the input signal vector x(n) is normalized, resulting in the following iterative adjustment equation:
For slow moving targets, the weights w can be fixed after the training period until the next training period (periodic adaptation). For fast moving target, after the training period, the output signal z(n) can be used as the reference signal s(n) (known as decisionaided beamforming).
Estimating the Angle of Arrival from the Weights
If the subarrays 110m in the hybrid array 100 are uniformly spaced in a linear configuration (e.g. 200), the weights will converge to the following form:
w_{m}(n)=exp(jkmd_{x}^{(s) }sin θ)+v_{m}(n) (4)
where k is the wavenumber of the RF carrier signal, θ is the “zenith” angle of arrival as illustrated in
If the subarrays 110m in the hybrid array 100 are uniformly spaced in a planar configuration (e.g. 230), the weights will converge to the following form:
w_{m}(n)=exp(jk(m_{x}d_{x}^{(s) }sin θ cos φ+m_{y}d_{y}^{(s) }sin θ sin φ))+v_{m}(n) (5)
where φ is the “azimuth” angle of arrival as illustrated in
Once the digital beamformer 165 converges, the angle of arrival estimation submodule 155 analyses the weights w(n) according to the model in equation (4) or equation (5) to form an estimate ({circumflex over (θ)},{circumflex over (φ)}) of the angle of arrival, given values for k, d_{x}^{(s) }and d_{y}^{(s)}.
Search Strategies
If the angle of arrival is not initially known, the phase shifters 130 cannot be set to the correct values to give each subarray 110m anything like its maximum gain. The digital beamformer 165 may therefore be unable to form a reasonable beam, and thus the angle of arrival cannot be estimated by the submodule 155 by the method described above.
To form a coarse estimate of the angle of arrival, and thereby initialise the hybrid array 100, the digital processing module 150 takes one of several approaches (known as the “search strategy”). One search strategy is to divide the whole hybrid array 100 into multiple groups. The subarray phase control submodule 170 sets the phase shifters 130 in the analogue subarrays 110m using equation (19) or (20) below, such that each group monitors a different “look angle” spanning the field of view. The digital processing module 150 operates on the outputs from each separate group in sequence, keeping the phase shifters 130 fixed. The “look angle” monitored by the group that results in the highestpower output signal z(n) is a coarse estimate of the angle of arrival. The reduced digital beamforming gain caused by the use of fewer than M subarrays at a time can be compensated for by a longer convergence time in the beamforming module 165.
Another search strategy is for the digital processing module 150 to perform a coarse digital beamforming using only one element 120 in each analogue beamforming subarray 110m, making the phase of each element irrelevant. An equivalent approach is to set the initial phases of each analogue subarray such that the beam pattern of the subarray has low directivity (i.e. is close to omnidirectional). A coarse estimate of the angle of arrival may be then obtained using the method described above.
A third search strategy is for the subarray phase control submodule 170 to set all the phase shifters 130 in all the subarrays 110m so as to monitor a particular “look angle” using equation (19) or (20) below. The digital processing module 150 then estimates the angle of arrival using either of the methods described herein, and also records the corresponding output signal power. This process is repeated while the “look angle” is stepped through multiple look angles that cover the field of view. At the end of the search, the coarse estimate of the angle of arrival is the angle estimate corresponding to the highest output signal power.
All the search strategies are executed by the subarray phase control submodule 170 in collaboration with the beamforming module 165.
Estimating the Angle of Arrival by Correlation
An alternative method of estimating the angle of arrival is for the estimation submodule 155 to crosscorrelate the signals x_{m}(n) from adjacent digital branches 145 along the xaxis and the yaxis:
R_{x}=E{x_{m}_{y}_{M}_{x}_{+m}_{x}*(n)x_{m}_{y}_{M}_{x}_{+m}_{x}_{+1}(n)} (6)
R_{y}=E{x_{m}_{y}_{M}_{x}_{+m}_{x}*(n)x_{(m}_{y}_{+1)M}_{x}_{+m}_{x}(n)} (7)
Defining
where λ_{c }is the wavelength of the RF carrier signal, then provided d_{x}^{(s) }and d_{y}^{(s) }are less than or equal to λ_{c}/2, u_{x }and u_{y }will be in the range (−π, π) and are equal to arg{R_{x}} and arg{R_{y}} respectively without any ambiguity. The angle of arrival could then be estimated directly from u_{x }and u_{y}, given values for λ_{c}, d_{x}^{(s) }and d_{y}^{(s)}.
However, for a uniformly spaced planar configuration, d_{x}^{(s) }and d_{y}^{(s) }are typically greater than λ_{c}/2, so u_{x }and u_{y }can be outside the range (−π, π) and phase ambiguity will be present in R_{x }and R_{y}.
To handle this ambiguity, the estimation submodule 155 carries out the method 300 illustrated in the flow chart in
In step 320, the estimation submodule 155 estimates the crosscorrelations iteratively over time n using all available subarray outputs (to maximise the signaltonoise ratio) as follows:
where 0<μ<1 is an updating coefficient (typically set to 0.001).
The estimation submodule 155 then at step 330 computes u_{x }and u_{y }values from arg{R_{x}} and arg{R_{y}} at the current p and q respectively, as follows:
u_{x}(p)=2πp+arg{R_{x}}, (12)
u_{y}(q)=2πq+arg{R_{y}}, (13)
where [.] denotes the operation of taking the integer part of a value.
Each pair (μ_{x}(p),u_{y}(q)) represents a candidate beam. The phase control submodule 170 at step 340 sets the phase shifts of the phase shifters 130 within each subarray for the current (p, q) pair to
The beamforming submodule 165 then at step 350 forms the beam as described above to produce the output signal z_{p,q}(n), after which (at step 360) the estimation submodule 155 computes and records the output signal power for the current (p, q) pair iteratively as
P^{(n)}(p,q)=(1−β)P^{(n+1)}(p,q)+βz_{p,q}(n)^{2} (15)
where 0<β<1 is an updating coefficient (typically set to 0.25).
The estimation submodule 155 then at step 370 determines whether the double loop over p and q is complete; if not, p and q are updated at step 375 and the method 300 returns to step 320.
If the double loop is complete, the estimation submodule 155 at step 380 determines the (p, q) pair that yielded the largest output signal power computed at step 360:
(p_{max},q_{max})=_{p,q}^{arg max}P^{(n)}(p,q) (16)
Finally, at step 390 the estimation submodule 155 uses the maximising pair (p_{max}, q_{max}) to estimate the angle of arrival:
Due to the delay from the time when the phase shifts are loaded into the phase shifters 130 in subarrays 110m to the time when a change of the beamformed output signal is observed, several (typically 4) iterations per (p, q) pair are preferable for step 360. If the number of iterations for each (p, q) pair is chosen so that the output power P^{(n)}(p,q) of the corresponding beam can be calculated with sufficient accuracy, one pass through the complete double loop will be sufficient to determine the most likely beam and hence estimate the angle of arrival. If fewer than that number of iterations are used, the double loop is repeated until the power of each beam calculated across multiple double loops is obtained with sufficient accuracy.
Setting the Phase of the Analogue Branches
The subarray phase control submodule 170 uses the estimated angle of arrival ({circumflex over (θ)},{circumflex over (φ)}) from the estimation submodule 155 to set the phase of each phase shifter 130 in each analogue branch 115 via the control signals 175 as follows:
α_{i}=−k(X_{i }sin {circumflex over (θ)} cos {circumflex over (φ)}+Y_{i }sin {circumflex over (θ)} sin {circumflex over (φ)}) (19)
for the planar array, or
α_{i}=−kX_{i }sin {circumflex over (θ)} (20)
for the linear array.
In tracking mode, the subarray phase control submodule 170 periodically obtains an estimate of u_{x}(p_{max}) and u_{y}(q_{max}) from the estimation module 155 (which uses steps 320 and 330 of the method 300 with (p, q) set to (p_{max}, q_{max})), and updates the phase of each phase shifter 130 using equation (14), as in step 340 of the method 300.
Calibrating the Array
Equations (4) and (5) hold if the subarrays 110m are uniformly spaced and the signal chains 140m are identical. In practice, there will be small deviations from uniform spacing, and imperfections due to component ununiformity, temperature change, and component ageing in the signal chains 140m, which cause the weight vector to converge to values whose phase deviates slightly from the models in Equations (4) and (5). If these deviations are not corrected by calibration, the estimation of angle of arrival may be affected and the performance of the hybrid array 100 may be suboptimal.
If the subarrays 110m are uniformly spaced, the phases between the signals output from any two adjacent digital branches 145 should have the same difference regardless of the phase deviation. This fact can be used to calibrate the digital branches. The calibration submodule 160 analyses the signals x_{m}(n) from each digital branch 145 to determine the phase deviation of each digital branch 145. To do this, the calibration submodule 160 uses the method 300 to obtain an estimate of the “expected” phase difference, defined by equations (8) and (9), between any two adjacent digital branches 145. This estimate is then used to determine the phase deviation in each digital branch 145 as the difference between the expected phase difference and the actual phase difference between that digital branch 145 and an adjacent, previously calibrated digital branch. The calibration proceeds by induction in each direction from a designated reference digital branch. The phase deviation is used by the beamforming submodule 165 to correct the signals x_{i}(n) before forming the beam.
The calibration submodule 160 also calibrates each analogue branch 115. According to one embodiment, the subarray phase control submodule 170 optimises the phase shifts applied by the phase shifters 130 in a subarray 110m in order to obtain the maximumpower output signal x_{m}(n) from the corresponding digital branch 145. Since the angle of arrival is known from the angle of arrival estimation submodule 155, the expected phase shift values, accurate to the last bit of the phase shifters, can be calculated from equations (19) and (20). The difference between the expected phase shift values and the optimised phase shift values gives the phase deviation for each analogue signal branch. The deviation is applied by the phase control submodule 170 to adjust the phase shifts of the phase shifters 130 in the subarray. Calibration of analogue signal branches is preferably performed a few times a day.
The arrangements described are applicable to the data communication and sensing industries, and particularly to satellite communication and other high data rate communication (mobile, pointtopoint, and pointtomultipoint topologies). The disclosed arrangements are most beneficial for millimeterwave to terahertz adaptive antenna arrays where tight element spacing makes it very difficult or impossible to do conventional digital or analogue beamforming alone. The typical operating range is 10 to 100 GHz, but the arrangements are useful for frequencies as low as 400 MHz.
The foregoing describes only some embodiments of the present invention, and modifications and/or changes can be made thereto without departing from the scope and spirit of the invention, the embodiments being illustrative and not restrictive.