BEAMFORMING IN MULTI- ANTENNA WIRELESS COMMUNICATION

FIELD OF THE INVENTION

The invention relates to the field of multi-antenna wireless communication with beam- forming techniques to focus transmission power towards an intended receiver. BACKGROUND OF THE INVENTION

With Transmit Beamforming (TxBF), a wireless transmitter with multiple antennas focuses its transmission towards an intended receiver, by aligning amplitude and phase of the radio signal from each antenna such that these signals interfere constructively in the direction of the intended receiver, as detailed e.g. in the standard IEEE 802.1 In. However, transmissions in other directions is not controlled and thus may lead to interference at further receivers in those directions. Particularly in dense wireless mesh networks, such interference degrades system capacity.

In the state-of-the-art TxBF, the transmitter uses knowledge of the channel coefficients hip from each transmit antenna i (i = 1 to NAM) to the intended receiver P. This channel information is obtained by the transmitter either implicitly from measuring hp in the reverse direction (from receiver at P to transmit antenna i), or explicitly from a cooperating receiver using a signalling protocol. An optimal selection of the transmit gains of the individual antennas, subject to a transmit power limit, ultimately maximizes the Signal to Noise Ratio SNR at the intended receiver ("beamformee", BF). In other words, transmissions from the NAM antennas are driven such that the resulting superposition is aligned in the direction §BF of the intended receiver.

For explanatory purposes, a line-of-sight case is considered in the following, with dipole antennas transmitting omni-directionally with respect to azimuth angle φ. The channel transfer function hip from a first antenna 1 to a reference point P in a plane perpendicular to the dipole and located at distance rip in the far field (rjp » wavelength λ) is

where ho is a constant proportional to \lrip, and where k = 2π/λ is the wavenumber of an unmodulated carrier wave.

The resulting overall antenna pattern resulting from N _AM antennas at point P depends on the azimuth angle φ and is obtained by superposition of the contributions from the individual antennas weighted with complex valued transmit gains a as follows:

With an optimal choice of the transmit gains a the NAM components in the sum are aligned for φ = BF, thus achieving the maximum gain G($RF) = NAM in the desired direction §BF of the intended receiver. However, sidelobes into other directions φ are also affected, and may thus generate unwanted, collateral interference at a susceptible receiver ("interferees", IF) in such direction

Fig.l and Fig.2 are a 2D antenna patterns generated by a transmitter with two antennas (depicted as small circles) one wavelength (e.g. 12 cm for a carrier frequency of 2.4 GHz) apart. The pattern is obtained by varying the reference point P = Ρ(φ) on a circle with azimuth 0≤ φ < 2π at large distance R (¾ np » λ). The directional power gain is depicted as a solid line and varies between zero and a maximum of two (dash-dot circle with maximum radius) equalling the number of antennas. The directional gain can be determined as

where the denominator is the average power,

\ \h ⁿ total I] ² = N ^l Ant \\h "0 I| ² ^2)

In Fig.l no Transmit Beamforming has been applied at all, resulting in a directional gain with two-fold symmetry and butterfly shape. In Fig.2 conventional Transmit Beamforming with correspondingly optimized transmit gains has been applied in order to maximize directional gain in the TxBF direction (broken line) of an intended receiver. As is readily apparent, maximum directional gain is also produced in other directions, which in a worst case may coincide with a direction of an exemplary collateral receiver which will thus suffer from unwanted interference. DESCRIPTION OF THE INVENTION

It is therefore an objective of the invention to minimize interference transmissions or unwanted radiation towards other susceptible receivers in multi-antenna wireless communication with beam-forming techniques. This objective is achieved by a method and a wireless transmitter according to the independent claims. Preferred embodiments are evident from the dependent patent claims.

According to the invention, the Signal to Noise Ratio SNR maximization criterion at an intended receiver is modified into an optimization criterion such that the unwanted interference power at unintended collateral receivers is also taken into account. The interference power is either itself minimized in the optimization process, or at least constrained, i.e. maintained below a maximum value. The invention increases capacity of interference- limited wireless mesh networks, by a computationally minor modification of the state-of-the-art Transmit Beam-Forming algorithm.

Specifically, in wireless multi-antenna transmission with beam- forming a radio signal is transmitted by a wireless transmitter with two or more antennas in a wireless mesh network with a first and a second receiver. The transmission comprises

- in a configuration step, determining different optimum transmit gains a for the two antennas such that a receive power PBF at the first receiver is optimized, specifically maximized, while, at the same time, a receive power PIF at the second receiver is minimized or constrained below an unintended receive power threshold, and

- in a transmission step, emitting, by each of the two antennas, the very same signal modified by, specifically amplified or multiplied with, the respective optimum transmit gain <¾.

In a preferred variant of the invention, the transmission comprises

- determining, in particular measuring, for each of the antennas towards, or in conjunction with, the first receiver, a respective first channel transfer function hp indicative of an electric field generated by the antenna i at a location P of the first receiver,

- determining, in particular measuring, for each of the antennas i towards, or together with, the second receiver at location Q, a respective second channel transfer function h%Q, - choosing the optimum transmit gains a optimizing, specifically maximizing or constraining above a receive power difference threshold, a power difference between the power PBF at the first receiver involving, or based on, the first channel transfer functions hp, and the power PIF at the second receiver involving, or based on, the second channel transfer functions hiQ, whilst keeping the total transmission power within the limit given by the wireless transmitter. This variant of optimizing a power difference is more flexible and provides a larger solution space than an approach constraining power at either receiver, and avoids meaningless solutions that may occur in a power ratio maximizing approach.

In an advantageous embodiment of the invention, the optimum transmit gains a are chosen to maximize a power difference between a weighted power at the first receiver and a differently weighted power at the second receiver. This embodiment allows favouring the power transmitted to one receiver over the other, preferably by means of a relative power weight that may be tuned in turn.

According to the invention, a wireless transmitter with two antennas for transmitting a signal to either of a first and a second receiver located at fixed, non-mobile receiver locations in a wireless mesh network is adapted to determine a first set of optimum transmit gains a for the two antennas such that a receive power PBF at the first receiver is maximized, while, at the same time, a receive power PIF at the second receiver is minimized or constrained, and to emit, by each of the two antennas, the signal modified by the respective optimum transmit gain <¾.

Preferably, the wireless transmitter is adapted to determine a second set of optimum transmit gains h such that a receive power PBF at the second receiver is maximized, while, at the same time, a receive power PIF at the first receiver is minimized or constrained. The transmitter is further adapted to switch from the first to the second set of transmit gains, and vice- versa, according to signal transmission focussing needs, preferably in a Time Division Multiplexing TDM mode for exchanging data between routers in a smart grid application.

The present invention also relates to a computer program product including computer program code for controlling a processor of a wireless transmitter with two antennas adapted to be connected to a wireless mesh network, particularly, a computer program product including a computer readable medium containing therein the computer program code.

BRIEF DESCRIPTION OF THE DRAWINGS

The subject matter of the invention will be explained in more detail in the following text with reference to preferred exemplary embodiments which are illustrated in the attached drawings depicting 2D antenna patterns with wireless multi antenna transmission and: Fig.l no Transmit Beamforming;

Fig.2 conventional Transmit Beamforming; Fig.3 conventional Transmit Beamforming with five antennas; and

Fig.4 interference minimizing Transmit Beamforming with weight γ = 0.4.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

Fig.3 and Fig.4 are a 2D antenna patterns generated by a transmitter with five antennas (depicted as small circles) arranged in the corners of a pentagon with a side length equivalent to 0.5/1, where λ is the typical wavelength of 12 cm at 2.4 GHz. In Fig.3, and similar to Fig.2, conventional Transmit Beamforming with correspondingly optimized transmit gains has been applied in order to maximize directional gain in the TxBF direction §BF (broken line) of an intended receiver. As is apparent, considerable directional gain is also produced in the IF direction /F (dot-dash line) of an exemplary collateral receiver, with a beamformee-to- interferee power ratio G{§BF)I G{§IF) of 4.12.

State-of-the-art Transmit Beamforming is performed by maximizing the normalized power PBF radiated towards the intended receiver, or beamformee BF located at point P:

where a is the beamforming vector comprising the transmit gains <¾ of the individual antennas, and where

is the channel matrix (vector) comprising the channel transfer functions (coefficients) hp from the NAM transmit antennas to the beamformee BF. Similarly, the normalized power PIF radiated to an unintended, collateral receiver, or interferee IF located at point Q is

_ a H _IF H _IF a

~ II II ²

Ml , (5)

where

According to an embodiment of this invention, the power ¾* to each of a plurality NIF of known interferees IF is minimized, whilst still maximizing the power PBF to the desired beamformee BF, or eventually even to a plurality of intended receivers in a multicast mode. The case of a single beamformee can be formalized as maximizing, over the beamforming vector a, the total weighted power

with appropriately normalized non-negative weights jk. Collecting (3) to (7), this yields

This is the Rayleigh coefficient of the matrix in the numerator. Hence, the maximum is obtained if a is the eigenvector pertaining to the largest eigenvalue of the matrix

which can be easily calculated numerically, once HBF and HIF are known.

While in general hp and hiQ are measured, e.g. using a measurement protocol exchange, in the line-of-sight example introduced above with a beamformee BF at P in direction §BF and with a single interferee IF (NIF = 1) at Q in direction the relevant hp and hiQ may be calculated. The transmit gain a is then the eigenvector to the (2x2) matrix

YH _BF H _BF - (\ - Y )H _{IF I} H _{IF I} where the relative power weight γ weights the importance of the beamformee power vs. interferee power (0≤ γ < 1); with 7 = 1 corresponding to the standard TxBF without interference minimization. Fig.4 depicts a numerical example for the line-of-sight case with calculated hp and hiQ, with a reduced beamformee weight of 7 = 0.4, for which the achieved beamformee / interferee power ratio increases to 23.36, as compared to a power ratio of 4.12 in the case of no interference minimization (Fig.3). Accordingly, it is evident that the power ratio may be greatly improved, with only a negligible reduction of the gain in the desired direction towards the beamformee.

The improvement of the beamformee / interferee power ratio depends on the number NAM and the mutual arrangement, including topology and inter-antenna distances dAnt, of the transmit antennas. Generally, the larger the number NAM of antennas per transmitter and the smaller their distances <¾ , the more pronounced the power ratio improvement, hence at least three antennas, preferably at least four antennas, and even more preferably at least five antennas may be provided per transmitter. With decreasing weight 7, the ratio G( S )/(J( / ) improves further, but at the cost of reduced G( S ). Attempting to directly maximize the ratio G( S )/G( /F) instead of the difference in equation (8) leads to a generalized eigenvalue problem. Due to the low rank of the matrices in (9), this may however lead to degenerate solutions with G( / )— 0 and some arbitrary small > 0, clearly an inefficient and meaningless solution.

The path losses determine the scaling of the channel matrices HBF and Hw,k. In practice, it may be preferable to exclude the effects of path loss. This is easily achieved by using the normalized matrices in equations (8) and (9). This re-scaling removes the distance- dependency of the channel matrices, such that only direction-dependent effects remain.

Alternatively to the weighted sum optimization criterion of equation (9), the minimization of the interferee power ¾* may be replaced by a constraint in the form of appropriately chosen interference power limits /%. The optimization is then simply formulated as max PBF, subject to PiF.k < pi .

The above relationships, in particular equation (3), are valid for any frequency, and Transmit Beamforming TxBF is applied separately for each OFDM subcarrier of an Orthogonal frequency-division multiplexing (OFDM) system. In the same manner, the principle of equation (8) of this invention can be applied in OFDM systems separately for each carrier frequency.

In standard Transmit Beamforming TxBF systems, the multiple transmit antennas are connected to, and controlled by, one transmitter at a given geographical location. With Coordinated Multipoint (CoMP) systems, the transmitting antennas are located at, and controlled by, base stations that are geographically separated by distances exceeding hundreds of meters. The present invention is applicable also to this case, where CoMP transmitters coordinate their transmissions to a given receiver whilst minimizing interference in the system.

Equation (5) above for PIF assumes that interferees have only one single receive antenna (MISO Multiple Input Single Output). Where the interferee has multiple receive antennas (MIMO Multiple Input Multiple Output) and employs a combining vector b, interference minimization may be performed by replacing the matrices H _{jF k} H _IFik in the optimization equation (8) by the matrices H^ _{F k} b _k b _k H _{IF k} . This implies knowledge of the combining vector bk at the interferee k, which may be difficult to obtain in practice.

HBF are dynamically obtained by implicit or explicit protocol exchanges between transmitter (beamformer) and receiver (beamformee) including suitable test, pilot, or training signal sequences. Basically, the channel matrices HIF may likewise be obtained via a protocol between beamformer and any interferee k to which interference is desired to be minimized. In general mobile systems, this may however not be practicable. In static systems however, such as wireless mesh networks for the Smart Grid, locations and directions of potential interferees are static and known. In such environments, it is realistic and practicable to determine and configure the required channel information HiF.k at system commissioning and maintenance.

The present invention further relates to a use of the method in a "smart grid" communication system for communicating electric power distribution grid data from sensor or source nodes and to actuator or destination nodes of the communication system. In this context, "Smart Grid" or "Distribution Automation" communication may include applications that consist of a city- wide network of smart meters and distribution monitoring sensors. Smart meters in an urban smart grid application will include electric, gas, and/or water meters typically administered by one or multiple utility companies. These meters will be capable of advanced sensing functionalities and provide status or cyclic sampled data as well as metering data, or even detect alarm conditions. In addition, they may be capable of advanced interactive functionalities, which may invoke an actuator component, such as remote service disconnect or remote demand reset. More advanced scenarios include demand response systems for managing peak load, and distribution automation systems to monitor the infrastructure that delivers energy.

While the invention has been described in detail in the drawings and foregoing description, such description is to be considered illustrative or exemplary and not restrictive. Variations to the disclosed embodiments can be understood and effected by those skilled in the art and practising the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. The mere fact that certain elements or steps are recited in distinct claims does not indicate that a combination of these elements or steps cannot be used to advantage, specifically, in addition to the actual claim dependency, any further meaningful claim combination shall be considered disclosed.