[0001] Wireless communication permits two electronic devices to communicate with one another without having to be physically connected to each other via wires or cables. One type of wireless communication is multiple-input and multiple-output (MIMO) wireless communication. In MIMO wireless communication, each electronic device uses an antenna array of multiple antennas to communicate with the other device to achieve a higher transmission rate than if just a single antenna were used at each device.

BRIEF DESCRIPTION OF THE DRAWINGS

[0002] FIG. 1 is a diagram of an example multiple-input and multipleoutput (MIMO) wireless communication system.

[0003] FIG. 2 is a flowchart of an example method for MIMO wireless communication in which a phase-adjusted channel matrix can be used when there is high spatial correlation between a transmitter device and a receiver device.

[0004] FIG. 3 is a flowchart of an example method for effectively using a phase-adjusted channel matrix in MIMO wireless communication when there is high spatial correlation between a transmitter device and a receiver device.

[0005] FIGs. 4A, 4B, and 4C are diagrams of example electronic devices to phase-adjust a signal that is modulated using a channel matrix to effectively use a phase-adjusted channel matrix in MIMO wireless communication when there is high spatial correlation.

[0006] FIG. 5 is a diagram of an example non-transitory computer- readable data storage medium storing program code.

DETAILED DESCRIPTION

[0007] As noted in the background, in multiple-input and multiple-output (MIMO) wireless communication, a transmitter device and a receiver device each use an antenna array of multiple antennas to achieve a higher transmission rate than if each device used just a single antenna. The transmitter device modulates and transmits a signal to the receiver device over multiple antenna paths, and the receiver device demodulates the signal upon reception. There is an antenna path between each unique pair of one of the transmitter device’s antennas and one of the receiver device’s antennas. Therefore, if there are N antennas in the antenna array of the transmitter device and N antennas in the antenna array of the receiver device, there are NxN antenna paths, or channels.

[0008] MIMO wireless communication can achieve a higher transmission rate because it employs all the antenna paths between the antenna array at the transmitter device and the antenna array at the receiver device. However, MIMO wireless communication is predicated on there being low spatial correlation between the transmitter and receiver devices, and more specifically between their respective antenna arrays. Spatial correlation is the correlation between a signal’s spatial direction and its received signal gain, and depends on the relative position of the receiver device to the transmitter device. If there is no spatial correlation, then wireless communication can be achieved at a maximum transmission rate.

[0009] However, as spatial correlation increases, the transmission rate at which wireless communication can be achieved between the transmitter and receiver devices decreases. At maximum spatial correlation between the antenna arrays of the transmitter and receiver devices, MIMO wireless communication occurs at a minimum transmission rate. The benefit of having multiple antennas at each of the transmitter and receiver devices is therefore eliminated, with the resulting transmission rate being comparable to that when MIMO wireless communication is not employed. That is, the resulting transmission rate is comparable to as if the transmitter and receiver devices each had just a single antenna.

[0010] Techniques described herein improve the transmission rate at which MIMO wireless communication occurs when there is high spatial correlation between transmitter and receiver devices. For instance, the techniques can include phase-shifting a modulated signal to ameliorate the effects of high spatial correlation on transmission rate, and thus compensate for such high spatial correspondence. Specifically, at the transmitter device, the signal can be phase-shifted after having been modulated, prior to or at transmission. Similarly, at the receiver device, the signal can be phase-shifted before being demodulated, subsequent to or at reception.

[0011] FIG. 1 shows an example system 100 including electronic devices 102A and 102B, between which MIMO wireless communication can be performed. Examples of electronic devices 102A and 102B include laptop, notebook, and desktop computers, smartphones, tablet computing devices, routers, and access points, among other types of devices When the electronic device 102A wirelessly transmits a signal and the electronic device 102B wirelessly receives the signal, the device 102A is a transmitter device and the device 102B is a receiver device. When the electronic device 102B wirelessly transmits a signal and the electronic device 102A wirelessly receives the signal, the device 102B is a transmitter device and the device 102A is a receiver device. [0012] The electronic devices 102A and 102B respectively include antenna arrays 104A and 104B. The antenna arrays 104A and 104B respectively include multiple antennas 106A and 106B. There are at least two antennas 106A and there are at least two antennas 106B. There may be the same or a different number of antennas 106A as there are antennas 106B. There are antenna paths 108 between the antenna array 104A and 104B over which the devices 102A and 102B perform MIMO wireless communication. There is an antenna path 108 between each antenna 106A and each antenna 106B. Therefore, if there are N antennas 106A and N antennas 106B, there are NxN antenna paths 108.

[0013] The electronic devices 102A and 102B respectively include circuits 110A and 110B. The circuits 110A and 110B may be considered as the radio modules for their respective antenna arrays 104A and 104B. When the device 102A transmits a signal to the device 102B, the circuit 110A modulates the signal for transmission at the antennas 106A over the antenna paths 108, and the circuit 11 OB demodulates the modulated signal received at the antennas 106B. When the device 102B transmits a signal to the device 102A, the circuit 11 OB modulates the signal for transmission at the antennas 106B over the antenna paths 108, and the circuit 110A demodulates the modulated signal received at the antennas 106A.

[0014] The circuits 110A and 11 OB can detect and compensate for high spatial correlation between the electronic devices 102B and 102A, and more specifically for high spatial correlation between the antenna arrays 104A and 104B. When the device 102A transmits a signal to the device 102B and there is high spatial correlation, the circuit 110A phase-shifts the signal after modulation or otherwise causes the modulated signal to be phase-shifted prior to or at transmission by the antennas 106A. The circuit 11 OB then reverse phase-shifts the modulated signal before demodulation or otherwise causes the modulated signal to be phase-shifted subsequent to or at reception by the antennas 106B.

[0015] Similarly, when the electronic device 102B transmits a signal to the electronic device 102A and there is high spatial correlation, the circuit 110B phase-shifts the signal after modulation or otherwise causes the modulated signal to be phase-shifted prior to or at transmission by the antennas 106B. The circuit 110A then phase-shifts the modulated signal before demodulation or otherwise causes the modulated signal to be phase-shifted subsequent to or at reception by the antennas 106A. In this way, the circuits 110A and 110B provide for a higher transmission rate even when there is high spatial correlation between the devices 102A and 102B. [0016] The remainder of the detailed description is described in relation to the electronic device 102A being the transmitter device that wirelessly transmits a signal, and the electronic device 102B being the receiver device that wirelessly receives the signal. The described techniques are equally applicable to the reverse scenario when the device 102B transmits a signal and the device 102A receives the signal. The remainder of the detailed description is also described in relation to there being two antennas 106A in the antenna array 104A and two antennas 106B in the antenna array 104B. The described techniques are more generally applicable to there being more than two antennas 106A and/ore there being at more than two antennas 106B.

[0017] In MIMO wireless communication, the electronic device 102A wirelessly transmits a signal via multiple signal streams from its multiple antennas 106A through a matrix channel that includes all the antenna paths 108 between the antennas 106A and the antennas 106B. That is, the electronic device 102A modulates the signal into signal streams that are transmitted by respective antennas 106A. The electronic device 102B then receives the signal streams at respective of its antennas 106B, and recovers the signal by demodulating the signal streams.

[0018] The signal streams transmitted at the antennas 106A of the electronic device 102A constitute a transmit vector x, and the signal streams received at the antennas 106B of the electronic device 102B constitute a receive vector y. In MIMO wireless communication, the vectors x and y are related to one another by a channel matrix H and a noise vector n, as y = Hx + n. [0019] The channel matrix H governs how the signal to be transmitted at the device 102A is modulated to generate the transmit vector x, and how the receive vector y received at the device 102B is demodulated to recover the signal. The channel matrix H describes how the signal propagates from the device 102A to the device 102B and represents the combined effects of scattering, fading, and power decay with distance.

[0020] In the case in which there are two antennas 106A and two antennas 106B, the channel matrix H has four entries corresponding to the antenna paths 108, or channels, between the antenna arrays 104A and 104B of the electronic devices 102A and 102B. The channel matrix H can be specified as:

_{H =} pill .121 “ Iji21 h.22 '

[0021] In this equation, /ill corresponds to the path 108 between the first antenna 106B and the first antenna 106A, /il2 corresponds to the path 108 between the first antenna 106B and the second antenna 106A, /i21 corresponds to the path 108 between the second antenna 106B and the first antenna 106A, and h.22 corresponds to the path 108 between the second antenna 106B and the second antenna 106B. The entries of the channel matrix H are thus organized over rows corresponding to the antennas 106B of the (receiver) electronic device 102B and over columns corresponding to the antennas 106A of the (transmitter) electronic device 102A.

[0022] The noise vector n can be modeled in a number of different manners. For example, the noise vector n can be modeled as a circular symmetric complex normal, as n ~ CJV'CO, S). In this equation, CN specifies a complex normal distribution characterizing complex random variables having real and imaginary parts that are jointly normal. The mean value of the complex normal distribution is zero in the equation and the noise covariance matrix S is known.

[0023] FIG. 2 shows an example method 200 for MIMO wireless communication between the electronic devices 102A and 102B in which a higher transmission rate can be achieved even when there is high spatial correlation between the devices 102A and 102B. As noted above, the device 102A is described as the transmitter device that wirelessly transmits a signal and the device 102B is described as the receiver device that wirelessly receives a signal. However, the method 200 is equally applicable to the reverse scenario when the device 102B transmits a signal and the device 102A receives the signal.

[0024] For MIMO wireless communication to occur between the electronic devices 102A and 102B, both devices 102A and 102B have to be aware of the channel matrix H so that the device 102A can properly modulate the signal prior to transmission and the device 102B can properly demodulate the signal subsequent to reception. Therefore, the device 102A transmits a pilot signal 202 to the device 102B (202), which receives the pilot signal (204). The device 102B then empirically estimates the channel matrix H based on the pilot signal (206), and transmits it to the device 102A (208), which receives the channel matrix H (210). [0025] The electronic device 102B therefore effectively learns the state of the channel as the channel matrix H based on the pilot signal being known and the received signal corresponding to the pilot signal. Different techniques can be used to empirically estimate the channel matrix H. Examples of such techniques include zero-forcing, successive interference cancellation (i.e., V-blast), maximum likelihood estimation if the noise vector n is assumed to have a Gaussian distribution), neural network MIMO detection, and so on.

[0026] Both the electronic devices 102A and 102B may then detect whether there is high spatial correlation between them (212A, 212B). That is, each device 102A and 102B can detect whether there is spatial correlation between the devices 102A and 102B greater than a threshold. In general, the devices 102A and 102B can be considered as having high spatial correlation if the determinant of the channel matrix H is less than (another, different) threshold. [0027] The determinant of the channel matrix H is a scalar value that is a function of the matrix H. As the determinant approaches zero, the spatial correlation between the devices 102A and 102B approaches a maximum value. Therefore, when the determinant of the channel matrix H is less than a corresponding low threshold (such as close to zero), then this means that the spatial correlation between the devices 102A and 102B is greater than a corresponding high threshold (such as close to maximum correlation).

[0028] In the case in which there are two antennas 106A and two antennas 106B, the determinant of the channel matrix H is equal to /ill x /i22 - /il2 x /i21. Therefore, when /ill = /il2 and h22 = h21, the determinant is zero. As such, when /ill = /il2 and when h22 = h.21 (i.e., the absolute difference between /ill and /il2 and the absolute difference between h22 and /i21 are each less than a corresponding low threshold), this means that the spatial correlation between the devices 102A and 102B is greater than a corresponding high threshold. Stated another way, there is high spatial correlation when the entries of each row of the channel matrix H are identical to one another (by more than a threshold).

[0029] In response to detecting that there is not spatial correlation between the electronic devices 102A and 102B greater than a threshold, each device 102A and 102B uses a non-phase-adjusted channel matrix for wireless communication (214A, 214B). For instance, when demodulating a modulated signal received at the antenna array 104B, the device 102B uses the channel matrix H without any phase adjustment. Similarly, when modulating a signal for transmission at its antenna array 104A, the device 102A uses the inverse of the channel matrix H without any phase adjustment. The wireless communication is therefore modeled as y = Hx + n, as described above.

[0030] By comparison, in response to detecting that there is spatial correlation between the electronic devices 102A and 102B greater than a threshold, each device 102A and 102B uses a phase-adjusted channel matrix for wireless communication (216A, 216B). For instance, when demodulating a modulated signal received at the antenna array 104B, the device 102B uses the matrix-product of the channel matrix H and a specified phase-adjustment matrix Hpa. That is, the channel matrix H is effectively phase adjusted by being multiplied by the specified phase-adjustment matrix Hpa.

[0031] Similarly, when modulating a signal for transmission at its antenna array, the electronic device 102A uses the matrix-product of the inverse of the channel matrix H and the inverse of the specified phase-adjustment matrix Hpa. The phase-adjustment matrix Hpa is known ahead of time by each electronic device 102A and 102B. The wireless communication when there is spatial correlation between the devices 102A and 102B therefore can be modeled as y = Mx + n, where M = H x Hpa.

[0032] The matrix product M can be specified as:

.. [mil ml2] M = Lm2911 m2 oT 2 J

In this equation, mil corresponds to the path 108 between the first antenna 106B and the first antenna 106A, ml2 corresponds to the path 108 between the first antenna 106B and the second antenna 106A, m21 corresponds to the path 108 between the second antenna 106B and the first antenna 106A, and m22 corresponds to the path 108 between the second antenna 106B and the second antenna 106B.

[0033] The phase-adjustment matrix Hpa is selected so that when the determinant of the channel matrix H is zero (or below a low threshold), the determinant of the matrix product M is not. In the case where there are two antennas 106A and two antennas 106B, the phase-adjustment matrix Hpa is selected so that when /ill = /il2 and h22 = h21, mil ml2 and when m22 m21. That is, phase-adjustment matrix Hpa is selected so that when the absolute difference between /ill and /il2 and the absolute difference between h22 and h.21 are each less than a corresponding low threshold, the absolute difference between mil and ml2 and the absolute difference between m22 and m21 are not.

[0034] FIG. 3 shows an example method 300 for effectively using a phase- adjusted channel matrix in MIMO wireless communication when there is high spatial correlation between the electronic devices 102A and 102B. The device 102A is again described as the transmitter device that wirelessly transmits a signal and the device 102B is described as the receiver device that wirelessly receives a signal. The method 300 is also applicable to the reverse scenario when the device 102B transmits a signal and the device 102A receives the signal. The method 300 can be considered as a particular implementation of 212A, 212B, 214A, 214B, and 216A, and 216B of the method 200.

[0035] The electronic device 102A modulates a signal to be transmitted to the electronic device 102B using the inverse of the channel matrix H that the device 102B previously estimated and provided to the device 102A (302). If the device 102A has detected that there is spatial correlation greater than a threshold (304), then it further phase-shifts the modulated signal (306), which is then transmitted to the device 102B (308). By comparison, if the device 102A has detected that there is not spatial correlation greater than a threshold (304), then it transmits the modulated signal without first phase-shifting the modulated signal (308). [0036] The electronic device 102B then receives the modulated signal (310). If the device 102B has detected that there is spatial correlation greater than a threshold (312), then it first phase-shifts the modulated signal (314) before demodulation using the channel matrix H (316). By comparison, if the device 102B has detected that there is not spatial correlation greater than a threshold (312), then it demodulates the modulated signal using the channel matrix H without first phase-shifting the modulated signal (316).

[0037] When the modulated signal is phase-shifted at the electronic device 102A in 306 prior to transmission and is phased-shifted at the electronic device 102B in 314 prior to demodulation, wireless communication occurs according to the equation y = Mx + n, where M = H x Hpa. The phase-shifting in 306 and in 314 therefore in effect results in the performance of wireless communication using the channel matrix H as phase-adjusted by the phaseadjustment matrix Hpa.

[0038] The phase-shifting at the electronic device 102A in 306 and the phase-shifting in 308 at the electronic device 102B correspond to one another, in that the latter phase shifting in 308 undoes the former phase-shifting in 306 to recover the modulated signal so that it can be demodulated. The modulated signal is thus phase-shifted by an angle corresponding to the phase-adjustment matrix Hpa. The magnitude of the angle by which the modulated signal is phased-shifted in 306 is the same as the magnitude by which the modulated signal is phased-shifted in 314. However, the polarity of the angle by which the modulated signal is phase-shifted in 306 is the opposite of the polarity of the angle by which the modulated signal is phase-shifted in 314.

[0039] FIGs. 4A, 4B, and 4C different example implementations as to how an electronic device 102 can phase-adjust a modulated signal prior to transmission in 306 of the method 300 or prior to demodulation in 314 of the method 300. Each of the electronic devices 102A and 102B can be implemented as the device 102 of any of the example implementations. The devices 102A and 102B can each be implemented as a device 102 in the same or different example implementation. In FIGs. 4A, 4B, and 4C, the device 102 includes an antenna array 104 of multiple antennas 106, and a circuit 110.

[0040] In FIG. 4A, the circuit 110 is made up of two circuits, a phaseshifting circuit 402 and a modulation/demodulation circuit 404. In the case of signal transmission when there is high spatial correlation, the modulation/demodulation circuit 404 modulates a signal, which is then electronically phase-shifted by the phase-shifting circuit 402 prior to transmission. In the case of signal reception when there is high spatial correlation, the phaseshifting circuit 402 electronically phase-shifts the modulated signal that has been received, and then the modulation/demodulation circuit 404 demodulates the modulated signal. Therefore, phase-shifting and modulation/demodulation are performed by the same circuit 110.

[0041] In FIG. 4B, by comparison, the circuit 110 performs just modulation/demodulation, and thus can be considered the modulation/demodulation circuit 404. The phase-shifting circuit 402 is external to the circuit 110. In the case of signal transmission when there is high spatial correlation, the circuit 110 causes the phase-shifting circuit 402 to electronically phase-shift the signal after modulation by the circuit 110. In the case of signal reception when there is high spatial correlation, the circuit 110 causes the phaseshifting circuit 402 to electronically phase-shift the signal prior to demodulation by the circuit 110.

[0042] In FIG. 4C, the circuit 110 again performs just modulation/demodulation, and thus can be considered the modulation/demodulation circuit 404. The antenna array 104 is mechanically (i.e., physically) rotatable via a rotator 406. For example, the antennas 106 may be mounted to a platform that is rotatable using a motor, in which case the rotator 406 includes the motor. Rotation of the antenna array 104 effectively introduces a phase-shift within the modulated signal transmitted or received at the antennas 106.

[0043] When high spatial correlation is detected, the circuit 110 of the electronic device 102 thus controls the rotator 406 to rotate the antenna array 104 to in effect phase-shift the modulated signal. When high spatial correlation is no longer detected, the circuit 110 controls the rotator 406 to rotate the array 104 in the opposite direction so that the modulated signal is no effectively longer phase-shifted. Unlike in the examples of FIGs. 4A and 4B in which the modulated signal is electronically phase-shifted by a phase-shifting circuit 402, the modulated signal is effectively phased-shifted by physically rotating the array 104 in the example of FIG. 4C. [0044] FIG. 5 shows an example non-transitory computer-readable data storage medium 500 storing program code 502 executed by a processor, such as that of the circuit 110. The circuit 110 may thus be considered as including a processor and memory, which may more generally be considered as the data storage medium 500. The processor and memory may be integrated within an application-specific integrated circuit (ASIC), for instance, or the processor may be a general-purpose processor, in which case the memory may be a separate semiconductor or other type of memory. The circuit 110 can be implemented to realize the data storage medium 500 in another manner as well. The program code 502 is executed to perform processing.

[0045] The processing includes determining whether spatial correlation between the Ml MO antenna array of a transmitter device and the Ml MO antenna array of a receiver device is greater than a threshold (504). The processing includes, in response to determining that the spatial correlation is greater than the threshold (506), using a phase-adjusted channel matrix for MIMO wireless communication between the transmitter and receiver devices (508). The processing includes, in response to determining that the spatial correlation is not greater than the threshold (506), using a non-phase-adjusted channel matrix for MIMO wireless communication between the transmitter and receiver devices (510).

[0046] Techniques have been described for improving MIMO wireless communication performance even when there is high spatial correlation between the antenna arrays of the transmitter and receiver devices. Such spatial correlation is compensated for by introducing a phase-shift in the modulated signal prior to transmission at the transmitter device and prior to demodulation at the receiver device. As a result, wireless communication occurs in accordance with a phase-adjusted channel matrix.