Beamforming scheme for phased-array antennas

Abstract

In a method for beamforming in communication a system, a plurality of different antenna weight vectors are generated. Each of the plurality of antenna weight vectors is a respective base vector of a unitary matrix. At least one training signal is transmitted a plurality of times via a plurality of antennas. A respective antenna weight vector from the plurality of different antenna weight vectors is applied each time the at least one training signal is transmitted. Feedback signals are received, where the feedback signals correspond to the at least one training signal transmitted the plurality of times and received at a receiver. A transmitter antenna weight vector is generated based on a mathematical combination of at least (i) the at least one training signal and (ii) the feedback signals.

Description

FIELD OF THE TECHNOLOGY

The disclosure relates generally to wireless communication systems and, more particularly, to techniques for generating beamforming vectors.

DESCRIPTION OF THE RELATED ART

An ever-increasing number of relatively inexpensive, low power wireless data communication services, networks and devices have been made available over the past number of years, promising near wire speed transmission and reliability. Various wireless technology is described in detail in several IEEE standards documents, including for example, the IEEE Standard 802.11b (1999) and its updates and amendments, the IEEE Standard 802.11n now in the process of being adopted, as well as the IEEE 802.15.3 Draft Standard (2003) and the IEEE 802.15.3c Draft D0.0 Standard, all of which are collectively incorporated herein fully by reference.

As one example, a type of a wireless network known as a wireless personal area network (WPAN) involves the interconnection of devices that are typically, but not necessarily, physically located closer together than wireless local area networks (WLANs) such as WLANs that conform to the IEEE Standard 802.11b. Recently, the interest and demand for particularly high data rates (e.g., in excess of 1 Gbps) in such networks has significantly increased. One approach to realizing high data rates in a WPAN is to use hundreds of MHz, or even several GHz, of bandwidth. For example, the unlicensed 60 GHz band provides one such possible range of operation.

Antennas and, accordingly, associated effective wireless channels are highly directional at frequencies near or above 60 GHz. In general, path loss on a wireless communication link may be partially determined by such operational parameters as carrier frequencies and distances between communicating devices, and may be further affected by shadowing effects along transmission paths, channel fading due to reflection, scattering, oxygen absorption, etc., and other environmental factors. As a result, link budget (i.e., the aggregate of gains and losses associated with a communication channel) is frequently subject to a significant path loss.

When multiple antennas are available at a transmitter, a receiver, or both, it is therefore important to apply efficient beam patterns to the antennas to better exploit spatial selectivity and improve the link budget of the corresponding wireless channel. Generally speaking, beamforming or beamsteering creates a spatial gain pattern having one or more high gain lobes or beams (as compared to the gain obtained by an omni-directional antenna) in one or more particular directions, with reduced the gain in other directions. If the gain pattern for multiple transmit antennas, for example, is configured to produce a high gain lobe in the direction of a receiver, better transmission reliability can be obtained over that obtained with an omni-directional transmission. In addition to providing better link reliability, beamforming can greatly reduce the amount of power dissipated by transmitting devices. More specifically, beamforming allows a transmitting device to focus the transmission power in a particular direction when transmitting data to one or several receiving devices.

Summary

In an embodiment, a method for beamforming in communication a system includes generating a plurality of different antenna weight vectors, wherein each of the plurality of antenna weight vectors is a respective base vector of a unitary matrix, and transmitting, a plurality of times, at least one training signal via a plurality of antennas, wherein a respective antenna weight vector from the plurality of different antenna weight vectors is applied each time the at least one training signal is transmitted. The method also includes receiving feedback signals corresponding to the at least one training signal transmitted the plurality of times and received at a receiver, and generating a transmitter antenna weight vector based on a mathematical combination of at least (i) the at least one training signal and (ii) the feedback signals.

In another embodiment, an apparatus comprises a base vector generator to provide a plurality of different antenna weight vectors, wherein each of the plurality of different antenna weight vectors is a respective base vector of a unitary matrix. Also, the apparatus comprises a beamforming network, and a beamforming controller configured to apply a respective antenna weight vector from the plurality of different antenna weight vectors each time at least one training signal is transmitted, wherein the at least one training signal is transmitted a plurality of times via a plurality of antennas. Additionally, the apparatus comprises a vector selection unit configured to generate a transmitter antenna weight vector based on a mathematical combination of at least (i) the at least one training signal and (ii) feedback signals corresponding to the at least one training signal transmitted the plurality of times and received at a receiver.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a communication system including a transmitting device and a receiving device, each having an antenna array, that may utilize efficient antenna weight vectors (AWVs) for beamforming techniques discussed herein;

FIG. 2 is a block diagram of an example architecture of the receiving device of FIG. 1;

FIG. 3 is a block diagram of an example architecture of the transmitting device of FIG. 1;

FIG. 4 is a block diagram of an example superframe that includes a receiver beamforming period and a transmitter beamforming period;

FIG. 5 is a flow diagram of an example method for generating an efficient receiver AWV which may be implemented by a receiving device such as the example receiving device of FIG. 2;

FIG. 6 is a flow diagram of an example method for generating an efficient transmitter AWV which may be implemented by a transmitting device such as the receiving device of FIG. 3;

FIG. 7 is a flow diagram of an example method for generating an efficient receiver or transmitter AWV for a channel that includes multiple carrier frequencies or frequency sub-carriers; and

FIG. 8 is a flow diagram of an example method for generating an efficient receiver or transmitter AWV using vectors generated at different instances of time.

DETAILED DESCRIPTION

FIG. 1 is a block diagram of an example wireless communication system 10 in which devices, such as a transmitting device 12 and a receiving device 14, use beamforming techniques discussed herein to significantly improve the quality and reliability of communications via a wireless communication channel 16. Each of the devices 12 and 14 may be, for example, a mobile station or a non-mobile station equipped with a set of one or more antennas 20-24 and 30-34, respectively. In general, the wireless communication system 10 may include any number of devices, each equipped with the same or a different number of antennas (e.g., 1, 2, 3, 4 antennas and so on). For beamforming, however, at least one of the devices 12, 14 generally should include more than one antenna. Further, the antennas 20-24 and 30-34 may define respective antenna arrays in at least some of the embodiments of the system 10.

During an Rx beamforming procedure, the receiving device 14 receives multiple training signals multiple times. For example, a same training signal, which may be known at the receiving device 14, may be received multiple times. As will be discussed in more detail below, the receiving device 14 can select several or all base vectors of a unitary matrix, and iteratively apply each of the selected vectors to the antenna array as an antenna weight vector (AWV) each time a training signal is received during the Rx beamforming procedure. As discussed in greater detail below, each element of an AWV assigns a particular weight parameter, or amplitude parameter and/or phase shift parameter, to a corresponding antenna. The multiple instances of the received training signal, each corresponding to the training signal received using a different AWV, define a received signal vector. Using the known training signals, the selected base vectors, and the received signal vector, the receiving device 14 generates an efficient receiver AWV to be used for subsequent reception of data via the channel 16. Further, the transmitting device 12 can apply a similar technique to generate an efficient transmitter AWV. To this end, the transmitting device 12 transmits training signals multiple times and applies a different AWV to the antenna array for each training signal, with each AWV selected from among the base vectors of a unitary matrix. The transmitting device 12 can then receive feedback from the receiving device 14 regarding the received signal vector (defined similarly to the Rx procedure discussed above), and use the training signals, the AWVs, and the received signal vector to generate an efficient transmitter AWV.

Although the wireless communication system 10 illustrated in FIG. 1 includes a transmitting device 12 and a receiving device 14, devices in the wireless communication system 10 can generally operate in multiple modes (e.g., a transmit mode and a receive mode). Accordingly, in some embodiments, antennas 20-24 and 30-34 support both transmission and reception, and each of the devices 12 and 14 operates as receiver and a transmitter. Alternatively or additionally, a given device may include separate transmit antennas and separate receive antennas.

Depending on the desired implementation, the antennas 20-24 and 30-34 of the transmitting device 12 and the receiving device 14, respectively, can operate as a part of a multiple input, multiple output (MIMO) system or as an antenna array. As is known, a channel between a transmitter and a receiver in MIMO wireless communication system includes a number of independent spatial channels. By utilizing the additional dimensionalities created by multiple spatial channels for the transmission of additional data, the MIMO system can provide improved spectral efficiency and link reliability. These benefits generally increase as the number of transmission and receive antennas within the MIMO system increases. However, in order to utilize multiple antennas in a MIMO mode, multiple radio frequency (RF) chains (e.g., amplifiers, mixers, analog-to-digital converters, etc.) are required, which increases costs.

By contrast, multiple antennas in an antenna array typically share a common RF chain, and generally require an architecture which is significantly less complex than the architecture of MIMO (or one of degenerate forms of MIMO such as SIMO or MISO). During beamforming, a communicating device (such as the transmitting device 12 or the receiving device 14) can vary the phase shift and/or amplitude of each antenna in the antenna array to control the radiation pattern and ultimately steer the beam. In the recent years, it has become common to control phase shifts and/or amplitudes of antennas digitally, i.e., by using a controller to generate signals determinative of specific phase shifts and/or amplitudes.

According to some implementations, transmitting or receiving devices equipped with multiple antennas control only the phase shift at each antenna. When an AWV includes only phase shift values, the antenna array to which the vector is applied is referred to as a phased antenna array. If, on other hand, each element in the AWV is a complex weight parameter that includes an amplitude and a phase shift component, the corresponding antenna array is referred to as a complex beamforming antenna array. It will be noted, however, that it is also common to use the shorter term “phased antenna array” to refer to both types of antenna arrays. As used herein, the term “phased antenna array” shall be construed broadly, and shall be understood to refer to antenna arrays in which one or both of the amplitude and phase shift of an antenna is controllable.

Next, FIG. 2 illustrates an example architecture of the receiving device 14 that includes an antenna array 102 connected to a shared analog/digital receiver data path 104. The antenna array 102 has N_rx, antennas including antennas 110, 112, and 114 coupled to respective delay lines 120, 122, and 124, which may be digitally controllable and which define a beamforming network 130. Each of the delay lines 120, 122, and 124 can shift the signal received from the corresponding antenna 110, 112, or 114 by a respective phase shifting angle θ. Together, the antenna array 102 and the delay lines 120, 122, and 124 define a phased antenna array of the receiving device 14.

For ease of illustration, FIG. 2 depicts the antenna array 102 as being controllable only by assigning respective phase shift values to the antennas 110, 112, and 114. However, in other embodiments, the antenna array 102 also includes a means for controlling the amplitude of the signal received at each antenna 110, 112, or 114 such as respective amplifiers. The amplifiers could be complex amplifiers to adjust both amplitude and phase, or scalar amplifiers for adjusting amplitude. It will be further noted that although some of the examples below focus on generating efficient phase-shift-only AWVs, the techniques discussed herein generally apply to AWVs that may also vary amplitudes.

An adder 132 may use any suitable technique to combine the signals from the antennas 110-114, shifted by the corresponding angles θ₁, θ₂, . . . θ_Nrx, prior to supplying the combined received signal y to the shared analog/digital receive data path 104. In general, the analog/digital receive data path 104 can include some or all of such components as an equalizer, a decoder, a de-interleaver, a demodulator, an A/D converter, a Fast Fourier Transform (FFT) processing block, etc.

Further, the receiving device 14 includes a beamforming controller 150 and a vector selection unit 154 communicatively coupled to a quality assessment unit 156, a base vector generator 158, a reference signal storage 160, and a past vector storage 162. In operation, the vector selection unit 154 supplies an AWV u or an indication of the AWV u to the beamforming controller 150. In the embodiment of FIG. 2, the AWV u specifies phase shifting angles θ₁, θ₂, . . . θ_Nrx for each of the N_rx antennas in the antenna array 102, and the beamforming controller 150 applies the angles θ₁, θ₂, . . . θ_Nrx to the respective delay lines 120, 122, and 124. In other embodiments, the AWV u may specify an amplitude for each antenna, and the beamforming controller 150 applies the amplitudes to amplifiers in the delay lines 120, 122, and 124.

The base vector generator 158 supplies N_rx AWVs u₁, u₂ . . . u_Nrx to the vector selection unit 154. More specifically, N_rx vectors u₁, u₂ . . . u_Nrx define a complete set of base vectors of a unitary matrix U having N_rx rows and N_rx columns (i.e., the N_rx vectors span the entire space of the matrix U). In some embodiments, the base vector generator 158 is a memory unit that stores the vectors u₁, u₂ . . . u_Nrx. During beamforming, the vector selection unit 154 selects R vectors from the set u₁, u₂ . . . u_Nrx and supplies the selected R vectors to the beamforming controller 150 which, in turn, applies a different one of the R vectors each time a training signal is received. In some embodiments, selection of R vectors from the set u₁, u₂ . . . u_Nrx can be implemented at the base vector generator 158 rather than at the vector selection unit 154. Further, in certain modes of operation of the receiving device 14, R=N_rx, and the vector selection unit 154 accordingly supplies every one of the Nr_x base vectors of the matrix U to the beamforming controller 150.

The reference signal storage 160 can include a memory component to store a known training signal s specified by the communication protocol or agreed upon during an earlier stage of communication, for example. The known training signal s may be a data unit or a data packet, or a portion, such as a payload, of the data unit or data packet, for example, associated with certain parameters such as spreading sequences, modulation scheme, etc. The reference signal storage 160 supplies the known training signal to the vector selection unit 154 which determines an efficient receiver AWV using the techniques discussed herein. If desired, the reference signal storage 160 may also store multiple training signals for use with different procedures or when a beamforming procedure involves instances of more than one training signal.

The analog/digital receive data path 104 supplies the received signal y to the quality assessment unit 108 and to the vector selection unit 154. During beamforming, each received signal y corresponds to an instance of the training signal s received via the antenna array 102. Thus, R instances of the known training signal s correspond to received signals y₁, y₂, . . . y_R defining a vector y. The analog/digital receive data path 104 may also supply the received signals y₁, y₂, . . . y_R to the quality assessment unit 108, which may use any desired technique to generate quality indicators or metrics for the signals y₁, y₂, . . . y_R. The quality assessment unit 156 supplies each calculated quality indicator to the vector selection unit 154. If desired, the quality assessment unit 156 can supply only some of the quality indicators (e.g., quality indicators exceeding a certain threshold value) to the vector selection unit 154. Using the received signals y₁, y₂, . . . y_R, R AWVs from the set u₁, u₂ . . . u_Nrx, and the known signal s, the vector selection unit 154 determines an efficient receiver AWV û. The receiving device 14 can use the AWV û to receive data units until the Rx beamforming procedure is repeated, for example. Additionally, the receiving device 14 can store the vector û and/or the vectors y₁, y₂, . . . y_R in the past vector storage 162 to improve subsequent beamforming procedures. Example techniques for generating the vector û are discussed in more detail below, following the discussion of an example transmitter architecture with reference to FIG. 3.

Referring now to FIGS. 2 and 3, in general, the architecture of the transmitting device 12 may be similar to the architecture of the receiving device 14. Of course, a device operating in the communication system 10 may be both a receiver and a transmitter (in respective operational states, for example). Thus, although FIGS. 2 and 3 illustrate receiver and transmitter architectures separately for the purposes of clarity, a device equipped with an antenna array and adapted to communicate via the channel 16 can have a common transceiver architecture with components illustrated in FIGS. 2 and 3. For example, a transceiver device can include a common base vector generator (such as the base vector generator 158) for generating base vectors for both receive beamforming training and transmit beamforming training, a common beamforming controller (such as the beamforming controller 150) for controlling the antenna array during both receive beamforming training and transmit beamforming training, etc., as well as components specific to transmitter and receiver operations such as the reference signal storage 160 or a feedback processing unit discussed below.

Referring now to FIG. 3, the transmitting device 12 includes an antenna array 202 coupled to a shared digital/analog transmitter data path 204. The antenna array 202 has N_tx antennas including antennas 210, 212, and 214 coupled to respective delay lines 220, 222, and 224, which can be digitally controllable and which define a beamforming network 230. Each of the delay lines 220, 222, and 224 can shift the signal received from the corresponding antenna 210, 212, or 214 by a respective phase shifting angle θ. Together, the antenna array 202 and the delay lines 220, 222, and 224 define a phased antenna array of the transmitting device 12.

A splitter or power divider 232 supplies the signal s from the shared digital/analog transmit data path 204 to the antennas 210-214, to be shifted by the corresponding angles θ₁, θ₂, . . . θ_Ntx prior to transmission to the receiving device 14. As in the example illustrated in FIG. 2, the digital/analog transmit data path 204 can include some or all of such components as an equalizer, a decoder, a de-interleaver, a demodulator, an A/D converter, a Fast Fourier Transform (FFT) processing block, etc.

A vector selection unit 256 supplies an AWV v to a beamforming controller 250. Similar to the AWV u, the AWV v specifies phase shifting angles θ₁, θ₂, . . . θ_Ntx for each antenna in the antenna array 202, and the beamforming controller 250 applies the angles θ₁, θ₂, . . . θ_Ntx to the respective delay lines 220, 222, and 224. Also similar to the example illustrated in FIG. 2, a base vector generator 254 supplies N_tx AWVs v₁, v₂ . . . V_Ntx, defining a complete set of base vectors of a unitary matrix V, to the vector selection unit 256. During beamforming, the vector selection unit 256 may select R vectors from the set v₁, v₂ . . . V_Ntx and supply the selected R vectors to the beamforming controller 250 which, in turn, applies a different one of the R vectors each time the training signal s is transmitted. In some embodiments, selection of R vectors from the set v₁, v₂ . . . V_Ntx can be implemented at the base vector generator 254 rather than at the vector selection unit 256. Further, in certain modes of operation of the transmitting device 12, R=N_rx, and the vector selection unit 256 accordingly supplies every one of the N_tx base vectors of the matrix V to the beamforming controller 250.

Each of a feedback processing unit 258 and a quality assessment unit 260 can receive a feedback signal descriptive of the signal y received at the receiving device 14 for a particular transmitted instance of the training signal s. During beamforming, R instances of the known training signal s correspond to received signals y₁, y₂, . . . y_R defining a vector y. Using the received feedback signals y₁, y₂, . . . y_R, R AWVs from the set v₁, v₂ . . . V_Ntx, and the training signal s, the vector selection unit 256 can determine an efficient transmitter AWV {circumflex over (v)}. As indicated above, the vector y can be also supplied to the quality assessment unit 260 to generate quality indicators or metrics for the signals y₁, y₂, . . . y_R. The quality assessment unit 260 can then supply each calculated quality indicator to the vector selection unit 256.

Referring to both FIG. 2 and FIG. 3, a signal y received at the receiving device 14 and corresponding to a training signal s transmitted from the transmitting device 12 is given byy=uHvs, where  (1)

u is a receiver AWV having one column and N_rx rows;

v is a receiver AWV having N_tx columns and one row; and

H is a channel gain matrix having dimensions N_rx×N_tx.

For ease of illustration, equation (1) does not include a noise term.

Of course, it is desirable that each value of y be as close to the transmitted signal s as possible, and an ideal value of the receiver AWV yields y=s. In other words, it is desirable that the communication system 10 alter s as little as possible. Referring to (1), it will be noted that y is equal to s when the receiver AWV u is û given by:û=(Hv)^H, where  (2)the superscript H denotes a Hermitian matrix, because theny=ûHvs=(Hv)^HHvs=s.

To consider beamforming at the receiving device 14 first, the transmitting device 12 can select an AWV v during an earlier beamforming training stage, which may be a coarse resolution beamforming training stage, for example, or using any other technique. The transmitter 12 can then transmit the training signal s N_rx times to the receiving device 14 which updates the AWV u to another one of u₁, u₂, . . . U_Nrx defining a unitary matrix U each time the training signal s is transmitted:U=[u₁,u₂, . . . U_Nrx]

As discussed above, the received signals y₁, y₂, . . . y_Nrx define the vector y:y=[y₁,y₂, . . . y_Nrx]_T so thaty=UHvs  (3)

Next, it will be observed thaty·s^H=UHv·s·s^H=UHv, U^Hy·s^H=U^HUHv, and thusU^Hy·s^H=Hv  (4)

Combining (4) with (2), the best estimate of the receiver AWV û is then given by:û=(Hv)^H=sy^HU  (5)

Referring to FIG. 2, the vector selection unit 154 can generate the vector û during the Rx beamforming procedure using the signal s supplied by the reference signal storage 160, the matrix U (or the base vectors of the matrix U) supplied by the base vector generator 158, and the receive vector y from the analog/digital Rx data path 104.

When implementing Rx beamforming according to the equation (5), the vector selection unit 154 can generate the vector û once every base vector of the unitary matrix U has been applied to the antenna array 102, and every respective signal y has been received. Because the unitary matrix U is necessarily a square matrix, the vector selection unit 154 requires that N_rx vectors be used. In some embodiments, particularly when the number of receive antennas N_rx is relatively low, a look-up table storing vectors u₁, u₂, . . . u_Nrx can be used to expedite processing. Alternatively, the vector selection unit 154 can reduce computational complexity during beamforming by selecting only a subset of R vectors from the set u₁, u₂, . . . u_Nrx and accordingly transmitting only R training units. In this case, the vector û′ is given by:û′=sy_R^HU_R,NRX, where  (6)

the vector y_R includes R elements and the matrix U_R,NRX accordingly includes R rows and N_rx columns of the unitary matrix U.

When using a reduced set of base vectors, the vector selection unit 154 can select the vectors from the set u₁, u₂, . . . u_Nrx based on past beamforming history, for example. To this end, the vector selection unit 154 can store a subset of AWVs used previously, along with the vector û, in the past vector storage 162. A certain iteration of the Rx beamforming procedure can generate the vector û based on the base vectors u₃, u₄, and u₅, for example, and a subsequent repetition of Rx beamforming may apply the vectors u₃, u₄, and u₅ first. In general, the vectors in a subset associated with the vector û need not be consecutive vectors in the set u₁, u₂, . . . u_Nrx. To assess the quality of a signal corresponding to a subset of R vectors, the quality assessment unit 156 can calculate average the quality indicators for each of the signals y corresponding to the R training units, for example. If the receiving device 14 cannot obtain an acceptable AWV û based on these vectors (i.e., if the quality assessment unit 156 generates a quality indicator for the set of R vectors below a certain threshold value), the vector selection unit 154 can proceed to apply vectors u₁, u₂ etc.

By way of a more specific example, the receiving device 14 can, during a certain iteration of Rx beamforming, receive multiple instances of the training signal s while applying different vectors u_i from the set u₁, u₂, . . . u_Nrx as antenna weight vectors to the antennas 110-114, and the quality assessment unit 156 may check the resulting quality of the corresponding received signal before applying the next vector u_i. Based on previously conducted coarse sweeping (or according to another selection principle), the receiving device 14 can select the vector u₃ first, and the quality assessment unit 156 may detect that an acceptable quality of the received signal has been reached after applying u₅.

As another alternative, the vector selection unit 154 can iteratively apply AWVs from the set u₁, u₂, . . . u_Nrx until a quality indicator obtained from the quality assessment unit 156 indicates than the respective signal y has an acceptable (although possibly not optimal) quality. The vector selection unit 154 can then skip the rest of the set u₁, u₂, . . . u_Nrx and generate the AWV û based on the available data. In this embodiment, the order in which the vector selection unit 154 steps through the sequence u₁, u₂, . . . u_Nrx may be made more efficient by utilizing past beamforming data stored in the past vector storage 162.

The transmitting device 12 can determine the transmitter antenna weight vector {circumflex over (v)} in a similar manner. In particular, if the AWVs v₁, v₂ . . . V_Ntx completely define the set of base vectors of a unitary matrix V, the vector {circumflex over (v)} can be calculated according to{circumflex over (v)}=sVy^H  (7)or, if only a subset of R vectors is being used,{circumflex over (v)}=sV_NTx,Ry_R^H  (8)where the vector y_R includes R elements and the matrix V_NTx,R includes N_tx rows and R columns of the unitary matrix V. As illustrated in FIG. 3, the transmitting device 12 can receive one or several feedback data units that specify the elements of the vector y (or y_R) from the receiving device 14. If desired, the receiving device 14 can transmit a feedback packet with only the dominant (i.e., best) set of signals y. As one example, the quality assessment unit 260 can generate a quality metric for each signal y and compare each quality metric to a threshold value, so that the receiving device 14 may include only those signals y that exceed the threshold value in the dominant set. As another example, the receiving device 14 can determine the dominant set by ranking the signals y and selecting the top N signals.

Generally with respect to Rx and Tx beamforming, the unitary matrices U and V can be phase-only matrices if the antenna arrays 102 or 202 are phased antenna arrays in which signal amplitude is not adjusted for individual antennas. In these embodiments, a discrete Fourier transform (DFT) engine can be efficiently used to generate the matrix U or V because the DFT engine similarly utilizes a unitary phase-only matrix.

In some embodiments, the vector selection unit 154 or 256 may generate an efficient AWV û or {circumflex over (v)} that specifies both an amplitude and a phase shift component for each antenna. To apply the vector û or {circumflex over (v)} to a phase-shift-only antenna array, the vector selection unit 154 or 256 can scale, or simply set, the amplitude of each element to unity.

Further, the vector selection unit 154 or 256 can generate an efficient AWV û or {circumflex over (v)} that does not precisely match the resolution of the antenna array 102 or 202. For example, an element of the vector û may specify the phase shift of 33.25° whereas the resolution of the antenna array may be 5°, so that an allowable phase shift at an antenna is 5°, 10°, 15°, etc. In this case, the vector selection unit 154 or 256 can additionally adjust the vector û or {circumflex over (v)} to match the desired resolution by setting each element to the closest allowable value (thus, 33.25° may be adjusted up to 35°). Such adjustments can be referred to as quantizing the phase shifts. The vector selection unit 154 or 256 can similarly quantize a gain-only vector û or {circumflex over (v)}, when applicable. In general, an efficient AWV û or {circumflex over (v)} may be scaled or adjusted (or quantized) in any suitable manner to make the vector û or {circumflex over (v)} compatible with a particular antenna array (which may be a phase-shift-only antenna array, a gain-only antenna array, a complex beamforming antenna array, etc., and which may be associated with any desired resolution and other parameters).

Still further, the Rx and Tx beamforming procedures discussed above can be carried out in frequency domain for each of a plurality of sub-carriers available in the communication channel 16. Upon generating several efficient AWVs {circumflex over (v)}_ci or {circumflex over (v)}_ci for each sub-carrier c_i=c₁, c₂, . . . c_n, the vector selection unit 154 or 256 can select the vector û_ci or {circumflex over (v)}_ci associated with the highest power, or apply any other suitable selection method to generate the vector û or {circumflex over (v)} based on multiple vectors û_ci or {circumflex over (v)}_ci.

Alternatively or additionally, an Rx and Tx beamforming procedure can include several iterations at distinct points in time, so that the vector selection unit 154 or 256 can generate multiple efficient AWVs û_ti or {circumflex over (v)}_ti at different times t_i=t₁, t₂, . . . , and select the vector û or {circumflex over (v)} from among the multiple vectors û_ti or {circumflex over (v)}_ti based on the power associated with each vector or another criterion.

In some embodiments, an ordered effective SINR or SNR vector is used instead of the y vector in generating the AWVs û or {circumflex over (v)}. It will be noted that the beamforming technique in these implementations becomes a gain-only weighing solution. However, using SINR or similar data may significantly simplify computation of the vector û or {circumflex over (v)}.

As an alternative to generating the transmitter AWV {circumflex over (v)} at the transmitting device 12, the transmitting device 12 can iteratively apply different base vectors of a unitary matrix V to the antennas 210-214 during Tx beamforming, and include these base vectors (or the corresponding identifiers) in the training signal s or some other signal sent to the receiving device 14. The receiving device 14 can then calculate the transmitter AWVs {circumflex over (v)} using the formula (7) or the formula (8), for example, and feed back the AWVs {circumflex over (v)} to the transmitting device 12. It is noted that in this embodiment, the receiving device 14 need not feed back the signals y to the transmitting device 12.

Generally with respect to the receiver architecture illustrated in FIGS. 2 and 3, it will be noted that various ones of the modules illustrated therein may be implemented using hardware, software instructions executed by a processor, firmware, instructions executed by a processor, or combinations thereof. Moreover, some of the components in FIGS. 2 and 3 may be custom integrated circuits, application-specific integration circuits (ASICs), etc., communicatively coupled by electrical busses, for example.

The transmitting device 12 and the receiving device 14 can carry out the respective Rx and Tx beamforming procedures in any order. For example, the transmitting device 12 may first determine an efficient transmitter AWV and use the efficient transmitter AWV when sending training signals to the receiving device 14. Alternatively, the receiving device 14 may determine an efficient receiver AWV prior to Tx beamforming. As yet another alternative, one or both of the transmitting device 12 and the receiving device 14 may carry out the respective Tx or Rx beamforming procedure multiple times. During a multi-stage sector Rx sweeping procedure, for example, the receiving device 14 applies different beamforming vectors (such as AWVs) to the antenna array 12 as each training signal is received, with the sector being swept at each stage being progressively narrower. Some of the examples of timing and ordering of Rx and Tx beamforming procedures which the devices 12 and 14 may implement are described in commonly-owned, co-pending U.S. patent application Ser. No. 12/548,393, filed on Aug. 26, 2009, and entitled “Beamforming by Sector Sweeping,” which is hereby expressly incorporated by reference herein, and in the U.S. Provisional Patent App. No. 61/091,914 entitled “Beamforming by Sector Sweeping,” filed Aug. 26, 2008, also expressly incorporated by reference herein.

In one particular embodiment, the receiving device 14 conducts Rx beamforming described above during a beam refinement stage of a multi-stage sector sweeping procedure. Prior to the beam refinement stage, the receiving device 14 conducts “coarse” sector sweeping stage to identify a (typically wide) sector in which a signal from the transmitting device 12 generates the highest power. The receiving device 14 then partitions the identified sector into multiple smaller sub-sectors in the beam refinement stage and sweeps through these sub-sectors to identify a more specific direction in which the antenna array 12 should be steered to maximize receive power. Of course, the receiving device 14 can further refine sector sweeping during as many stages as desired.

FIG. 4 illustrates one example of a superframe 300 in which an Rx beamforming training (TRN-R) period 304 occurs after a preamble period 302, and a Tx beamforming training (TRN-T) period 306 occurs several periods after the TRN-R period 304. The periods 304 and 306 may correspond to coarse sweeping stages, beam refinement stages, or both. In other words, coarse sweeping and/or beam refinement procedures for receive beamforming training may occur during the period 304, and coarse sweeping and/or beam refinement for transmit beamforming training may occur during the period 306.

FIGS. 5-8 illustrate flow diagrams of several example methods that the receiving device 14 and the transmitting device 12 may implement using hardware, firmware instructions implemented on a processor, software, or any combination thereof to support efficient digital beamforming techniques discussed herein. The various blocks illustrated in these diagrams may be implemented by various components illustrated in FIGS. 2 and 3 in any desired manner. Thus, it will be appreciated that FIGS. 2 and 3 illustrate specific arrangements of components that may implement particular blocks of FIGS. 5-8 by way of example only.

Referring to FIG. 5, the receiving device 14 (see FIGS. 1 and 2) can implement an example method 400 to generate an efficient receiver AWV. In block 402, the receiving device 14 obtains (e.g., by generating, retrieving from memory, receiving from a wired interface, etc.) N_rx base vectors of a unitary matrix having N_rx rows and N_rx columns, each specifying antenna weight parameters so that each of the base vectors is an AWV. Next, the receiving device 14 selects a subset of R vectors from the N_rx base vectors based on past beamforming data, in a sequential order, in a random order, or in any other suitable manner (block 404). In some cases, R may be equal to N_rx, i.e., the receiving device 14 may accordingly select as many vectors as there are receive antennas, or the block 404 may be omitted.

In blocks 405-410, the receiving device 12 iteratively steps through the subset of selected base vectors. In particular, the receiving device 12 selects a base vector from the subset of R vectors in block 405, applies the antenna weight factors specified by the selected base vector in block 406, receive a training signal via the antenna array 102 steered according to the selected base vector in block 408, and check whether the set has been exhausted in block 410. During this step, optionally, the receiving device 12 may check for other conditions such as whether a certain quality threshold has been met even though the set has been only partially exhausted, for example. If the set has not been exhausted, and if quality threshold has not been met (optional), flow proceeds back to block 405. Other flow proceeds to block 412. Finally, in block 412, the receiving device 12 generates a receiver AWV applying the formula (5) or using another suitable technique that relies on the received signals, the set of selected base vectors, and the training signal.

FIG. 6 illustrates an example method 430 which the transmitting device 12 can implement to generate a transmitter AWV. Similar to the method 400, the transmitting device 12 obtains N_tx base vectors of a unitary matrix in block 432 to be used as antenna weight vectors in block 432. Next, the transmitting device 12 selects a subset of R vectors from the N_tx base vectors in block 434, and iteratively applies the selected vectors to the antenna array 202 as a certain training signal is transmitted until the set is exhausted or another condition is satisfied (blocks 436-440). In particular, the transmitting device 12 selects a base vector from the subset or R vectors at block 435, applies the selected vector to the antennas 210-214 at block 436, transmits a training signal (which may be the same signal for all vectors applied at block 436) to the receiving device 14 at block 438, and checks whether the set of R vectors has been exhausted at block 440. If the set has not been exhausted, flow proceeds back to block 435. Otherwise, flow proceeds to block 442. It will be noted, however, that the transmitting device 12 may be unable to evaluate the quality of signals, received by the receiving device 14 and corresponding to the transmitted training signal and to a respective AWV, until after multiple instances of the training signal are transmitted and a feedback data unit is received in block 442. Of course, a communication protocol implemented by the devices 12 and 14 may alternatively specify that the feedback data unit can be transmitted after each training signal or according to any other schedule. In block 444, the transmitting device 12 generates an efficient transmitter AWV using the feedback data unit(s), the selected base vectors, and the training signal.

Referring to FIG. 7, one or both of the devices 12 and 14 may implement an example method 450 to generate an AWV for each of several sub-bands or carriers (blocks 452-454) and select an efficient AWV (block 456) based on these carrier-specific values. More specifically, the devices 12 or 14 may simply select the best AWV from among the set of carrier-specific AWVs, calculate an average and quantize or scale the average to meet the resolution requirements of the corresponding antenna array, or use any other suitable technique. Block 454 may include all or some of the blocks of the method 400 or method 430, for example.

Similarly, the device 12 or 14 can implement an example method 480 illustrated in FIG. 8 to generate AWVs at several time instances (block 482-484) and calculate an efficient transmitter or receiver AWV based on these multiple time-specific AWVs. As is known, a communication channel such as the channel 16 may change due to environmental conditions, interference, and other factors. Thus, the devices 12 and 14 may implement the method 480 to at least partially account for the variance of the channel 16 over time.

As least some of the various blocks, operations, and techniques described above can be implemented in hardware, firmware, software instructions implemented on a processor, or any combination thereof. When implemented in software instructions executable on a processor, the software may be stored in any computer readable memory such as on a magnetic disk, an optical disk, or other storage medium, in a RAM or ROM or flash memory of a computer, processor, hard disk drive, optical disk drive, tape drive, etc. Likewise, the software may be delivered to a user or a system via any known or desired delivery method including, for example, on a computer readable disk or other transportable computer storage mechanism or via communication media. Communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism. The term “modulated data signal” means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, radio frequency, infrared and other wireless media. Thus, the software may be delivered to a user or a system via a communication channel such as a telephone line, a DSL line, a cable television line, a wireless communication channel, the Internet, etc. (which are viewed as being the same as or interchangeable with providing such software via a transportable storage medium). When implemented in hardware, the hardware may comprise one or more of discrete components, an integrated circuit, an application-specific integrated circuit (ASIC), etc.

Moreover, while the beamforming techniques have been described with reference to specific examples, which are intended to be illustrative only and not to be limiting of the invention, it will be apparent to those of ordinary skill in the art that changes, additions and/or deletions may be made to the disclosed embodiments without departing from the spirit and scope of the invention.

Claims

1. A method for beamforming in a communication system, the method comprising: generating a plurality of different antenna weight vectors, wherein each of the plurality of antenna weight vectors is a respective base vector of a unitary matrix;transmitting, a plurality of times, at least one training signal via a plurality of antennas, wherein a respective antenna weight vector from the plurality of different antenna weight vectors is applied each time the at least one training signal is transmitted;receiving feedback signals corresponding to the at least one training signal transmitted the plurality of times and received at a receiver; andgenerating a transmitter antenna weight vector based on a mathematical combination of at least (i) the at least one training signal and (ii) the feedback signals.

2. The method of claim 1, wherein a number of antenna weight vectors in the plurality of different antenna weight vectors is equal to a number of antennas in the plurality of antennas, so that the plurality of different antenna weight vectors completely defines the unitary matrix.

3. The method of claim 1, wherein a number of antenna weight vectors in the plurality of different antenna weight vectors is less than a number of antennas in the plurality of antennas.

4. The method of claim 1, further comprising selecting the plurality of different antenna weight vectors from the unitary matrix based on past history of generating antenna weight vectors for transmit beamforming.

5. The method of claim 1, wherein: the plurality of antennas is associated with a resolution; andthe method further comprises quantizing the transmitter antenna weight vector to the resolution of the plurality of antennas.

6. The method of claim 1, wherein generating the transmitter antenna weight vector comprises generating the transmitter antenna weight vector based on a mathematical combination of (i) the at least one training signal, (ii) the feedback signals, and (iii) the plurality of different antenna weight vectors.

7. The method of claim 1, wherein each of the plurality of different antenna weight vectors is a phase-only vector.

8. The method of claim 1, wherein each of the plurality of different antenna weight vectors is an amplitude-only vector.

9. The method of claim 1, further comprising: generating a respective quality indicator for each antenna weight vector in the plurality of antenna weight vectors based on the feedback signals;comparing each quality indicator to a threshold quality value; anddetermining when to stop transmitting the at least one training signal based on the comparisons of each quality indicator to the threshold quality value.

10. The method of claim 1, further comprising setting an amplitude of the receiver antenna weight vector to unity.

11. The method of claim 1, further comprising: repeating transmitting the at least one training signal, receiving feedback signals, and generating the transmitter antenna weight vector for each of a plurality of sub-carriers to define a plurality of sub-carrier-specific transmitter antenna weight vectors,wherein generating the transmitter antenna weight vector is based on the plurality of sub-carrier-specific receiver antenna weight vectors.

12. The method of claim 1, further comprising: repeating transmitting the at least one training signal, receiving feedback signals, and generating the transmitter antenna weight vector at each of a plurality of time-instances to define a plurality of time-specific transmitter antenna weight vectors,wherein generating the transmitter antenna weight vector is based on the plurality of time-specific transmitter antenna weight vectors.

13. An apparatus comprising: a base vector generator to provide a plurality of different antenna weight vectors, wherein each of the plurality of different antenna weight vectors is a respective base vector of a unitary matrix;a beamforming network;a beamforming controller configured to apply a respective antenna weight vector from the plurality of different antenna weight vectors each time at least one training signal is transmitted, wherein the at least one training signal is transmitted a plurality of times via a plurality of antennas; anda vector selection unit configured to generate a transmitter antenna weight vector based on a mathematical combination of at least (i) the at least one training signal and (ii) feedback signals corresponding to the at least one training signal transmitted the plurality of times and received at a receiver.

14. The apparatus of claim 13, wherein the base vector generator is configured to select the plurality of different antenna weight vectors from the unitary matrix based on past history of generating antenna weight vectors for transmit beamforming.

15. The apparatus of claim 13, wherein: the plurality of antennas is associated with a certain resolution; andthe vector selection unit is configured to quantize the transmitter antenna weight vector to the resolution of the plurality of antennas.

16. The apparatus of claim 13, wherein the vector selection unit is configured to generate the transmitter antenna weight vector based on a mathematical combination of (i) the at least one training signal, (ii) the feedback signals, and (iii) the plurality of different antenna weight vectors.

17. The apparatus of claim 13, wherein each of the plurality of different antenna weight vectors is a phase-only vector.

18. The apparatus of claim 13, wherein each of the plurality of different antenna weight vectors is an amplitude-only vector.

19. The apparatus of claim 13, wherein the vector selection unit is configured to set an amplitude of the receiver antenna weight vector to unity.

20. The apparatus of claim 13, wherein the vector selection unit is configured to generate the transmitter antenna weight vector based on a plurality of sub-carrier-specific receiver antenna weight vectors.

21. The apparatus of claim 13, wherein the vector selection unit is configured to generate the transmitter antenna weight vector based on a plurality of time-specific transmitter antenna weight vectors.