WO2018219475A1

WO2018219475A1 - Processing device, transmitting device and methods thereof

Info

Publication number: WO2018219475A1
Application number: PCT/EP2017/063480
Authority: WO
Inventors: Jocelyn Aulin
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2017-06-02
Filing date: 2017-06-02
Publication date: 2018-12-06
Also published as: CN110419171A

Abstract

A processing device (100) for a wireless communication system, the processing device (100) being configured to: obtain a first downlink precoder (V DL1 ) for a set of transmit antenna elements (302a, 302b,..., 302z) associated with a set of receiving devices (600a, 600b,..., 600z); obtain a first downlink transmit power allocation (P DL1 ) for a total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z); if a fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z) remains after having obtained the first downlink transmit power allocation (P DL1 ): compute at least one second downlink precoder (V DL2 ) for the set of transmit antenna elements (302a, 302b,..., 302z) based on a power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z); compute at least one second downlink transmit power allocation (P DL2 ) for the remaining fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z) based on the second downlink precoder (V DL2 ); compute a resultant downlink precoder and power allocation comprising a first product of the first downlink precoder (V DL1 ) and the first downlink transmit power allocation (P DL1 ), and a second product of the second downlink precoder (V DL2 ) and the second downlink transmit power allocation (p DL2 )- Furthermore, the invention also relates to a transmitting device, a corresponding method, and a computer program.

Description

PROCESSING DEVICE, TRANSMITTING DEVICE AND METHODS THEREOF

Technical Field

The invention relates to a processing device for a wireless communication system. Furthermore, the invention also relates to a transmitting device, a corresponding method, and a computer program.

Background

For multi-cell wireless communication systems, such as e.g. 3GPP Long Term Evolution (LTE) and LTE Advanced, inter-cell interference is a significant impairment to the downlink (DL) system performance. Multi-cell interference aware (MCA) precoding has therefore been introduced in multi-cell wireless communication systems. MCA precoding controls both the intra-cell interference and leakage of signals to other cells, i.e. inter-cell interference. The downlink system performance is thereby significantly improved. However, the determination of the MCA precoding matrices according to a corresponding cost function (i.e. minimization of the mean squared error of the received data symbols) has turned out to be very difficult.

An additional challenge in multi-cell wireless communication systems is to achieve a balanced user experience within a cell. Two problems have been identified:

(i) Poor user data rates at the cell edge.

(ii) Unbalanced user experience where about half of the users experience the lowest data rates and a very small percentage of the users experience the highest data rates. One challenge is to improve the spectral efficiency (bps/Hz) for users with the lowest data rates within a target cell while not sacrificing too much of the overall cell spectral efficiency (SE). User power allocation for the transmission antenna(s) is an approach to address this challenge. A user in this context may correspond to a receiving device. Consequently, there is a need to optimize both the downlink precoder and the corresponding downlink power allocation to improve user cell edge performance while not sacrificing too much aggregate cell spectral efficiency in the wireless communication system.

Summary

An objective of embodiments of the invention is to provide a solution which mitigates or solves the drawbacks and problems of conventional solutions. Another objective of embodiments of the invention is to provide a solution providing higher spectral efficiency compared to conventional solutions.

The above and further objectives are achieved by the subject matter of the independent claims. Further advantageous implementation forms of the invention are defined by the dependent claims.

According to a first aspect of the invention, the above mentioned and other objectives are achieved with a processing device for a wireless communication system, the processing device being configured to

obtain a first downlink precoder for a set of transmit antenna elements associated with a set of receiving devices;

obtain a first downlink transmit power allocation for a total transmit power allocated to the set of receiving devices;

if a fraction of the total transmit power allocated to the set of receiving devices remains after having obtained the first downlink transmit power allocation:

compute at least one second downlink precoder for the set of transmit antenna elements based on a power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices;

compute at least one second downlink transmit power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices based on the second downlink precoder;

compute a resultant downlink precoder and power allocation comprising a first product of the first downlink precoder and the first downlink transmit power allocation, and a second product of the second downlink precoder and the second downlink transmit power allocation.

The processing device may comprise a processor configured, for example, via stored instructions, to perform the actions specified above. In the case that there is not enough power left to be allocated for the second downlink precoder, compute a resultant downlink precoder and power allocation comprising only the first product of the first downlink precoder and the first downlink transmit power allocation. Hence, in this respect the processing device is further configured to, if no fraction of the total transmit power allocated to the set of receiving devices remains after having obtained the first downlink transmit power allocation:

compute a resultant downlink precoder and power allocation comprising only the first product of the first downlink precoder and the first downlink transmit power allocation. In an implementation form of a processing device according to the first aspect, the processing device is configured to compute the second downlink precoder for the set of transmit antenna elements based on an equal power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices. However, other power allocations (such as two-step, see below) for the remaining fraction of the total transmit power can also be used when computing the second downlink precoder for the set of transmit antenna elements.

According to embodiments of the invention, more than one second downlink precoder and second downlink transmit power allocation may be computed if a fraction of the total power remains. Therefore, the computation of further second downlink precoders and second downlink transmit power allocations can continue until no remaining power of the total transmit power is left. This also means that the resultant downlink precoder and power allocation can be computed based on the first product and one or more subsequent second products, wherein each subsequent second product is a product of a subsequent second downlink precoder and a subsequent second downlink transmit power allocation.

Advantages of computing more than one second downlink precoder and second downlink transmit power allocation are:

that the variance of the resultant power is decreased, including while the per antenna power constraint is fulfilled, and more of the available total transmit power is used and hence, the downlink rate for the cell is increased.

A fraction of the total transmit power could be seen as a scalar variable between 0 and 1 , where a 0 value denotes zero power of the total transmit power and value 1 denotes all of the total transmit power. The total transmit power can be understood as the total available power at a transmitting device (e.g. a base station) which in that case is a function of the maximum output power of each power amplifier at each transmit antenna port. A receiving device may comprise any mobile communication device configured to receive a communication signal transmitted by one or more transmit antenna elements i.e. having at least a transmit function. A set of receiving devices may comprise a plurality of receiving devices of various types. Transmit antenna elements associated with a set of receiving devices may mean that the transmit antenna elements are configured to transmit one or more communication signals to the set of receiving devices. For example, the transmissions from the transmit antenna elements to the set of receiving devices may relate to spatially multiplexed transmissions using precoders and/or beamforming. The set of transmit antenna elements may in one example be arranged in an antenna array of antenna elements, where each antenna element may be connected to a receive chain or a processing branch, and hence may be referred to as an antenna branch. The antenna elements may be arranged physically as a rectangular array of elements or according to any other physical arrangement. In a cellular system, such an antenna array may be configured or dedicated for transmission in a section of a cell or in a whole cell. The antenna array may be part of, or associated with a base station or a remote radio head or any other deployed radio network node. Further, the set of transmit antenna elements may support Multiple Input Multiple Output (MIMO) or any other multi-antenna transmission technique.

It should also be understood that the wording "obtain" in this context may also mean "compute" such that the downlink transmit power allocation is computed based on the combined power allocation in the final iteration. Other variables and parameters may also be considered when computing the downlink transmit power allocation.

The processing device according to the first aspect provides a number of advantages over conventional processing devices. Embodiments of the invention improve cell-edge user spectral efficiency without compromising cell spectral efficiency (SE) gains and are an enabler for application of per antenna power constraints. Furthermore, the idea of superposition of precoders with respective power allocations (as per the last shown step of the first aspect) can be applied to power allocation problems, such as per antenna power constraint (PAPC) problems.

In an implementation form of a processing device according to the first aspect, the processing device is configured to

compute the resultant downlink precoder and power allocation comprising a linear combination of the first product and the second product.

An advantage with this implementation form is that it permits making use of available remaining transmit power through the use of two downlink precoders and their respective downlink power allocations. In an implementation form of a processing device according to the first aspect, the processing device is configured to compute the resultant downlink precoder and power allocation comprising the linear combination of the first product and the second product constrained on a second power constraint relating to maximum transmit powers for each transmit antenna element in the set of transmit antenna elements.

The second constraint may be a PAPC and may be applied to the combination of the two above products.

An advantage with this implementation form is that the linear combination of the first product and the second product permits the resultant transmission of the precoder downlink data to fulfil the per antenna power constraint while achieving performance close to an unconstrained case.

re-compute the second downlink transmit power allocation so as to fulfil the second power constraint.

An advantage with this implementation form is that the second downlink power allocation can be optimized e.g. to improve the performance of receiving devices with low received signal- to-interference-plus-noise ratio (SINR).

compute the second downlink precoder for only non-saturated antenna elements in the set of transmit antenna elements.

This means that rows of antenna elements containing power saturated antenna elements are removed or not considered at all when computing the second downlink precoder.

An advantage with this implementation form is that the remaining transmit power can be used for the second downlink precoder leading to improved spectral efficiency performance since the saturated elements are removed. In an implementation form of a processing device according to the first aspect, the processing device is configured to obtain the first downlink precoder for a set of transmit antenna elements associated with a set of receiving devices, and

obtain the first downlink transmit power allocation for the total transmit power allocated to the set of receiving devices

by being configured to:

obtain an initial power allocation for the set of transmit antenna elements associated with the set of receiving devices;

a) compute a precoder or a corresponding uplink detector based on the initial power allocation or a combined power allocation from a previous iteration,

b) compute a first power allocation for a first fraction of the total transmit power allocated to the set of receiving devices based on the precoder or the corresponding uplink detector, c) compute a second power allocation for a remaining second fraction of the total transmit power allocated to the set of receiving devices based on the precoder or the corresponding uplink detector,

d) compute a combined power allocation based on the first power allocation and the second power allocation,

repeat a) to d) until the combined power allocation satisfies a convergence threshold in a final iteration;

obtain the first downlink precoder for the set of transmit antenna elements based on at least one of the precoder or the corresponding uplink detector, and the combined power allocation in the final iteration;

obtain the first downlink transmit power allocation for the set of receiving devices based on the combined power allocation in the final iteration. This implementation form can be considered as an inner iterative loop as a particular method for obtaining the first downlink precoder and the corresponding first downlink transmit power allocation. In this application, this implementation form may be referred to as "two step" power allocation, where for the first precoder its respective power allocation is comprised of two power allocations in a respective first step b) and second step c), each of which may be determined using different criteria.

The power allocation between the first and second fractions of the total transmit power can be seen as a trade-off between different power allocation algorithms employed for computing the first and second power allocations, respectively. The fractions in this implementation form are not necessarily the same fractions as in the first aspect. Hence, the values of the first and second fractions may depend on the network deployment or any other infrastructural parameters. Therefore, the first and second fractions in this implementation form may be considered as design parameters which may depend on the application and use. It is further noted that the combined power allocation converges to a convergence threshold in the final iteration. The combined power allocation is a vector which means that the corresponding convergence threshold is a vector of the same size as the combined power allocation. Moreover, the downlink precoder for the set of transmit antenna elements may be obtained based on the precoder or the corresponding uplink detector, or the precoder and the combined power allocation, or the corresponding uplink detector and the combined power allocation.

Furthermore, the downlink transmit power allocation for the set of receiving devices may be obtained by e.g. using a look-up table, which maps the combined power allocation in the final iteration to the downlink transmit power allocation.

The processing device according to this implementation form has a number of advantages. One such advantage is that the processing device provides an optimized downlink transmit power allocation and downlink precoder that offer improved performance for receiving devices having lower received SINR while not sacrificing too much the performance for receiving devices with higher received SINR.

In an implementation form of a processing device according to the first aspect, the initial power allocation is an equal power allocation for the set of transmit antenna elements.

This implementation form provides a simple, i.e. a lower complexity, initial power allocation.

compute the precoder or the corresponding uplink detector using a criterion of minimum mean squared error of received data symbols for an interference aware scheme.

The interference aware scheme may consider inter-cell, intra-cell, or inter-cell and intra-cell interference depending on the network deployment. The interference aware scheme may in one case be the previously mentioned MCA scheme.

This implementation form enables the wireless communication system to attain significantly improved performance since interference at the receiving devices is significantly reduced with MCA precoding. Moreover, a combined PAPC and MCA algorithm fulfills the PAPC and has performance close to that of the MCA with power allocation scheme without PAPC (i.e. the so-called unconstrained case). However, unconstrained MCA without PAPC put high requirements on hardware, such as power amplifiers, due to the need of supporting very high amplitude variations. Additionally, the cell-edge user performance of PAPC combined with MCA has a significant gain relative to the regularized zero forcing (RZF) precoder which is the benchmark reference precoder.

compute the first power allocation using a criterion of maximization of minimum spectral efficiency for the set of receiving devices.

This implementation form enables the received SINR at the receiving devices to satisfy a minimum SINR requirement with the first fraction of total transmit power. Herein, this implementation form may be referred to as max-min, as it uses a maximization of minimum spectral efficiency as an optimization criterion, which when applied gives the max min power allocation.

compute the second power allocation using a criterion of maximization of sum rate in the wireless communication system for the set of receiving devices.

This implementation form enables using a fraction of the total downlink transmit power to improve the aggregate cell spectral efficiency. Herein, this implementation form may be referred to as max-sum, for similar reasons as the above max-min method.

In an implementation form of a processing device according to the first aspect, the elements in the convergence threshold have a value in an interval between 0.005 - 0.100.

The convergence threshold can be a vector of values, where each element or component of the vector relates to a power element difference value between a combined power allocation vector in a previous iteration and a combined power allocation vector in a current iteration. This implementation form means that each vector element in the convergence threshold has a value in the mentioned interval. Each difference value relates to its corresponding respective threshold value. For convergence, each difference value is equal or less than the respective threshold value, thus satisfying the convergence threshold. This implementation form provides flexibility in assigning the convergence quality requirement for each power element in the convergence threshold and still provides good performance. Having flexibility in assigning convergence quality requirement means that the power allocation for some antenna elements have more stringent convergence requirements than others. However, if this relaxation of convergence quality does not sacrifice resultant spectral efficiency performance, which is close to that where all antenna elements have equally stringent convergence quality requirement in power allocation, then reduction in the complexity via reduction in the number of iterations is possible since the final resultant spectral efficiency performance is not significantly changed.

In an implementation form of a processing device according to the first aspect, the elements in the convergence threshold have the same threshold value.

This implementation form provides a uniform and low complexity assignment of threshold values.

compute the combined power allocation comprising a sum of the first power allocation and the second power allocation.

This implementation form enables both satisfying the minimum received SINR requirement for receiving devices and improving the aggregate cell spectral efficiency. In an implementation form of a processing device according to the first aspect, the processing device is configured to

compute the combined power allocation based on the sum of the first power allocation and the second power allocation constrained on a first power constraint related to maximum transmit powers for each transmit antenna element in the set of transmit antenna elements.

In one case the squared norm of the product of the first precoder multiplied with the square root of the vector sum of the first power allocation and the second power allocation is constrained on the first power constraint. An advantage with this implementation form is that the first downlink precoder with its corresponding first downlink power allocation provides a per antenna power constrained downlink transmission solution with good performance relative to the unconstrained case. In an implementation form of a processing device according to the first aspect, the processing device is configured to

obtain the initial power allocation as an initial uplink power allocation,

compute the corresponding uplink detector in a).

This implementation form permits solving for an optimized downlink precoder and downlink power allocation in the virtual uplink domain which in many cases may be simpler than solving these in the downlink domain.

In an implementation form of a processing device according to the first aspect, the first power allocation is a first uplink power allocation and the second power allocation is a second uplink power allocation such that the combined power allocation is a combined uplink power allocation.

This implementation form enables the combined use of two virtual uplink power allocation algorithms, each obtained using its respective optimization criterion. The sum of the two virtual uplink power allocations gives the combined virtual uplink power allocation. In an implementation form of a processing device according to the first aspect, the processing device is configured to

compute an uplink Signal-to-Noise plus Interference Ratio based on the combined uplink power allocation in the final iteration;

compute the first downlink transmit power allocation based on the combined uplink power allocation and the uplink Signal-to-Noise plus Interference Ratio.

A conversion function can be used to compute the first downlink transmit power allocation based on the combined uplink power allocation and the uplink Signal-to-Noise plus Interference Ratio.

This implementation form enables transforming the virtual uplink system to the equivalent downlink system. Further, by using the computed uplink Signal-to-Noise plus Interference Ratio, improved downlink transmit power allocation can be computed. In an implementation form of a processing device according to the first aspect, the processing device is configured to compute an uplink power allocation based on the combined downlink power allocation in d);

compute the corresponding uplink detector in a) based on the uplink power allocation. A conversion function can be used to compute an uplink power allocation based on the combined downlink power allocation in d).

This implementation form enables transforming the downlink system to its corresponding virtual uplink system where the obtained virtual uplink power allocation is used in obtaining the virtual uplink detector.

According to a second aspect of the invention, the above mentioned and other objectives are achieved with a transmitting device for a wireless communication system, the transmitting device being configured to

obtain a resultant downlink precoder and power allocation according to any of the preceding aspects or implementation forms;

perform downlink transmissions to a set of receiving devices using the resultant downlink precoder and power allocation. The downlink data transmission may be performed by multiplying the data intended for transmission with the downlink precoder and the downlink transmit power allocation.

According to a third aspect of the invention, the above mentioned and other objectives are achieved with a method for a wireless communication system, the method comprising:

obtaining a first downlink precoder for a set of transmit antenna elements associated with a set of receiving devices;

obtaining a first downlink transmit power allocation for a total transmit power allocated to the set of receiving devices;

computing at least one second downlink precoder for the set of transmit antenna elements based on a power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices;

computing at least one second downlink transmit power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices based on the second downlink precoder; computing a resultant downlink precoder and power allocation comprising a first product of the first downlink precoder and the first downlink transmit power allocation, and a second product of the second downlink precoder and the second downlink transmit power allocation.

In an implementation form of a method according to the third aspect, the method comprises computing the resultant downlink precoder and power allocation comprising a linear combination of the first product and the second product.

In an implementation form of a method according to the third aspect, the method comprises computing the resultant downlink precoder and power allocation comprising the linear combination of the first product and the second product constrained on a second power constraint relating to maximum transmit powers for each transmit antenna element in the set of transmit antenna elements.

In an implementation form of a method according to the third aspect, the method comprises re-computing the second downlink transmit power allocation so as to fulfil the second power constraint.

In an implementation form of a method according to the third aspect, the method comprises computing the second downlink precoder for only non-saturated antenna elements in the set of transmit antenna elements.

In an implementation form of a method according to the third aspect, the method comprises obtaining the first downlink precoder for a set of transmit antenna elements associated with a set of receiving devices, and

obtaining the first downlink transmit power allocation for the total transmit power allocated to the set of receiving devices, by:

obtaining an initial power allocation for the set of transmit antenna elements associated with the set of receiving devices;

a) computing a precoder or a corresponding uplink detector based on the initial power allocation or a combined power allocation from a previous iteration,

b) computing a first power allocation for a first fraction of the total transmit power allocated to the set of receiving devices based on the precoder or the corresponding uplink detector, c) computing a second power allocation for a remaining second fraction of the total transmit power allocated to the set of receiving devices based on the precoder or the corresponding uplink detector,

d) computing a combined power allocation based on the first power allocation and the second power allocation,

repeating a) to d) until the combined power allocation satisfies a convergence threshold in a final iteration;

obtaining the first downlink precoder for the set of transmit antenna elements based on at least one of the precoder or the corresponding uplink detector, and the combined power allocation in the final iteration;

obtaining the first downlink transmit power allocation for the set of receiving devices based on the combined power allocation in the final iteration.

In an implementation form of a method according to the third aspect, the initial power allocation is an equal power allocation for the set of transmit antenna elements.

In an implementation form of a method according to the third aspect, the method comprises computing the precoder or the corresponding uplink detector using a criterion of minimum mean squared error of received data symbols for an interference aware scheme.

In an implementation form of a method according to the third aspect, the method comprises computing the first power allocation using a criterion of maximization of minimum spectral efficiency for the set of receiving devices. In an implementation form of a method according to the third aspect, the method comprises computing the second uplink power allocation using a criterion of maximization of sum rate in the wireless communication system for the set of receiving devices.

In an implementation form of a method according to the third aspect, the elements in the convergence threshold have a value in an interval between 0.005 - 0.100.

In an implementation form of a method according to the third aspect, the elements in the convergence threshold have the same threshold value.

In an implementation form of a method according to the third aspect, the method comprises computing the combined power allocation comprising a sum of the first power allocation and the second power allocation. In an implementation form of a method according to the third aspect, the method comprises computing the combined power allocation based on the sum of the first power allocation and the second power allocation constrained on a first power constraint related to maximum transmit powers for each transmit antenna element in the set of transmit antenna elements.

In an implementation form of a method according to the third aspect, the method comprises obtaining the initial power allocation as an initial uplink power allocation,

computing the corresponding uplink detector in a).

In an implementation form of a method according to the third aspect, the first power allocation is a first uplink power allocation and the second power allocation is a second uplink power allocation such that the combined power allocation is a combined uplink power allocation. In an implementation form of a method according to the third aspect, the method comprises computing an uplink Signal-to-Noise plus Interference Ratio based on the combined uplink power allocation in the final iteration;

computing the downlink transmit power allocation based on the combined uplink power allocation and the uplink Signal-to-Noise plus Interference Ratio.

In an implementation form of a method according to the third aspect, the first power allocation is a first downlink power allocation and the second power allocation is a second downlink power allocation such that the combined power allocation is a combined downlink power allocation.

In an implementation form of a method according to the third aspect, the method comprises computing an uplink power allocation based on the combined downlink power allocation in d);

computing the corresponding uplink detector in a) based on the uplink power allocation.

The advantages of the methods according to the third aspect are the same as for the corresponding processing device according to the first aspect.

Embodiments of the invention also relate to a computer program, characterized in code means, which has instructions for processing means such as a processor to execute any method according to the invention. Thus, when the computer program is run on the processing means any of the said methods is performed. Further, the invention also relates to a computer program product which may comprise a computer readable medium and said mentioned computer program, wherein said computer program is included in the computer readable medium, which may comprise of one or more from the group: ROM (Read-Only Memory), PROM (Programmable ROM), EPROM (Erasable PROM), Flash memory, EEPROM (Electrically EPROM) and hard disk drive.

Further applications and advantages of the invention will be apparent from the following detailed description. Brief Description of the Drawings

The appended drawings are intended to clarify and explain different embodiments of the invention, in which:

- Fig. 1 shows a processing device according to an embodiment of the invention.

- Fig. 2 shows a corresponding method according to an embodiment of the invention. - Fig. 3 shows an exemplary implementation according to an embodiment of the invention in which the processing device is part of a transmitting device.

- Fig. 4 shows a wireless communication system according to an embodiment of the invention in which the processing device is part of a radio network node, such as a base station.

- Fig. 5 illustrates an embodiment of the invention in which the downlink-uplink duality is used in conjunction with an inner iterative loop.

- Fig. 6 illustrates a system architecture according to an embodiment of the invention.

Detailed Description

Fig. 1 shows a processing device 100 according to an embodiment of the invention. The processing device 100 may be a standalone device or may be part of another device. For example, the processing device 100 may be an integrated part of a radio network node, e.g. a base station, or of a central network node, e.g. a radio network controller. The processing device 100 comprises in an embodiment a processor 102 configured for processing data. The processor 102 may in an embodiment also comprise further parts, such as a transceiver, a memory, etc. In one such embodiment as shown the processor 102 may be coupled to a transceiver 104 and a memory 106 by means of suitable communication means 108 known in the art. The memory 106 may be configured for intermediate and/or final storage of data. By using the transceiver 104, the processing device 100 is configured to communicate with other network devices via suitable wired and/or wireless communication interfaces 1 10. In one embodiment, the processor 102 may be a dedicated processor for executing any of the methods and/or algorithms according to the invention. In some embodiments, the processor 102 of the processing device 100 may instead be shared with other functions.

The processing device 100 in Fig. 1 is configured to obtain a first downlink precoder V_DL1 for a set of transmit antenna elements 302a, 302b,..., 302z (as shown in Fig. 3) associated with a set of receiving devices 600a, 600b,..., 600z (as shown in Fig. 4). The processing device 100 is further configured to obtain a first downlink transmit power allocation p_DL1 for a total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z. The processing device 100 is further configured to, if a fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z remains after having obtained the first downlink transmit power allocation p_DL1:

compute at least one second downlink precoder V_DL2 for the set of transmit antenna elements 302a, 302b,..., 302z based on a power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z;

compute at least one second downlink transmit power allocation p_DL2 for the remaining fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z based on the second downlink precoder V_DL2;

compute a resultant downlink precoder and power allocation comprising a first product of the first downlink precoder V_DL1 and the first downlink transmit power allocation Pan, and a second product of the second downlink precoder V_DL2 and the second downlink transmit power allocation p_DL2. As previously mentioned, the second downlink precoder V_DL2 for the set of transmit antenna elements is one embodiment computed based on an equal power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z. However, also other power allocations for the remaining fraction of the total transmit power can also be used instead when computing the second downlink precoder V_DL2 for the set of transmit antenna elements.

The resultant downlink precoder and power allocation may be a matrix of size K by N and 1 by K, respectively, where K is the total number of antennas of receiving devices and N is the total number of antennas of the transmitting devices. Fig. 2 shows a corresponding method 200 for a processing device 100. The method may be executed in a processing device 100, such as the one shown in Fig. 1.

The method 200 in Fig. 2 comprises obtaining 202 a first downlink precoder V_DL1 for a set of transmit antenna elements 302a, 302b,..., 302z associated with a set of receiving devices 600a, 600b,..., 600z. The method 200 further comprises obtaining 204 a first downlink transmit power allocation p_DL1 for a total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z. The method 200 further comprises, if a fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z remains after having obtained the first downlink transmit power allocation p_DL1 , computing 206 at least one second downlink precoder V_DL2 for the set of transmit antenna elements 302a, 302b,..., 302z based on a power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z. The method 200 further comprises computing 208 at least one second downlink transmit power allocation p_DL2 for the remaining fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z based on the second downlink precoder V_DL2 . The method 200 further comprises computing 210 a resultant downlink precoder and power allocation comprising a first product of the first downlink precoder V_DL1 and the first downlink transmit power allocation p_DL1, and a second product of the second downlink precoder V_DL2 and the second downlink transmit power allocation p_DL2.

Fig. 3 shows an embodiment of a transmitting device 300 for a wireless communication system 500. The transmitting device 300 comprises in this embodiment the processing device 100, and further an antenna 302 coupled to a transceiver 304 of the transmitting device 300. The processing device 100, the transceiver 304 and the antenna 302 are coupled to each other as shown by means of communication means 306 known in the art. In the embodiment shown in Fig. 3, the antenna 302 is an antenna array comprising a set of transmit antenna elements 302a, 302b,..., 302z arranged in suitable rows and columns depending on application. Such an application may be MIMO transmissions in the downlink. The transmitting device 300 in Fig. 3 is configured to obtain a resultant downlink precoder and power allocation according to embodiments of the invention. Further, the transmitting device 300 is configured to perform downlink transmissions to a set of receiving devices 600a, 600b,..., 600z using the resultant downlink precoder and power allocation. In one example, the resultant downlink precoder V_DL and the resultant downlink transmit power allocation p_DL are multiplied with data addressed for the set of receiving devices 600a, 600b,..., 600z. Fig. 4 shows a wireless communication system 500 comprising a transmitting device 300 and a set of receiving devices 600a, 600b,..., 600z. It is illustrated in Fig. 4 how the transmitting device 300 is configured to perform downlink data transmissions to the set of receiving devices 600a, 600b,..., 600z utilizing the downlink precoder V_DL and the downlink transmit power allocation p_DL obtained from the processing device 100. The wireless communication system 500 may be a cellular system, such as LTE and LTE Advanced, but is not limited thereto. In such cellular systems, a transmitting device 300 may be a base station, and a receiving device may be a user equipment (UE). In the present disclosure, these expressions are interchangeably used.

In an embodiment, the processing device 100 is configured to compute the resultant downlink precoder and power allocation comprising a linear combination of the first product and the second product. In an embodiment, the processing device 100 is configured to compute the resultant downlink precoder and power allocation comprising the linear combination of the first product and the second product constrained on a second power constraint C2 relating to maximum transmit powers for each transmit antenna element in the set of transmit antenna elements 302a, 302b,..., 302z. The second power constraint C2 is explained in detail in the following disclosure.

In an embodiment, the processing device 100 is configured to re-compute the second downlink transmit power allocation p_DL2 so as to fulfil the second power constraint C2. In this case only the second power allocation is modified so that the second constraint C2 is satisfied.

In the following disclosure more detailed elaborations of the present invention is presented. In this respect LTE terminology and system architecture is used. However, as realised by the skilled person, embodiments of the invention are not limited thereto. In an embodiment, the present method employs what could be seen as an inner iterative loop for obtaining the first downlink precoder V_DL1 and the first downlink transmit power allocation PDLI- While the first power constraint C1 is fulfilled in the below presented inner loop method, there is still remaining power available after subtracting the power used for the first power allocation and the second power allocation constrained on the first power constraint C1 from the total available transmit power (see below). Hence, there performance is further improved by using the present method which fulfills both the first power constraint C1 and the second power constraint C2. The first and second power constraints C1 and C2, respectively, may be flexibly combined in the embodiments of the invention. Definitions of the first and second power constraints C1 and C2, as well as how to apply are given below. These are applicable individually to all embodiments. According to this embodiment, the method employed to make use of the remaining available power is to:

• Solve for a second downlink precoder V_DL2 and apply remaining power to this second set of precoding beams of the second downlink precoder V_DL2 by computing a second downlink transmit power allocation p_DL2- · Compute a resultant downlink precoder and power allocation as previously explained e.g. by employing linear superposition of the first downlink precoder V_DL1, first downlink transmit power allocation p_DL1 , second downlink precoder V_DL2 , and the second downlink transmit power allocation p_DL2. The inner iterative loop employs the processing device 100, which is configured to:

obtain an initial power allocation p_Init for the set of transmit antenna elements 302a, 302b,..., 302z associated with the set of receiving devices 600a, 600b,..., 600z;

a) compute a precoder V, or a corresponding uplink detector U, based on the initial power allocation p_Init or a combined power allocation from a previous iteration,

b) compute a first power allocation p-L for a first fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z based on the precoder V, or the corresponding uplink detector U,,

c) compute a second power allocation p₂ for a remaining second fraction of the total transmit power allocated to the set of receiving devices 600a, 600b,..., 600z based on the precoder V, or the corresponding uplink detector U,,

d) compute a combined power allocation based on the first power allocation p_t

and the second power allocation p₂.

The processing device 100 is configured to repeat i.e. iterate steps a) to d) in the sequence a) then b) then c) then d) until the combined power allocation

is smaller than or equal to a convergence threshold C_Th in a final iteration. The processing device 100 may compare the combined power allocation against the convergence threshold in each iteration. The processing device 100 is further configured to

obtain the first downlink precoder V_DL1 for the set of transmit antenna elements 302a, 302b,..., 302z based on at least one of the precoder V, or the corresponding uplink detector

U,, and the combined power allocation in the final iteration; obtain the first downlink transmit power allocation p_DL1 for the set of receiving devices 600a, 600b,..., 600z based on the combined power allocation in the final iteration.

The initial power allocation p_Init obtained by the processing device 100 may in one example be an equal power allocation for the set of receiving devices 600a, 600b,..., 600z. Equal power allocation means that the same amount of power is allocated to each receiving device in the set of receiving devices 600a, 600b,..., 600z. Also, other initial power allocation is possible, such as allocating more power to downlink user data streams whose channel links is known to have lower SINR at the outset. This may reduce the number of iterations by speeding up the convergence for e.g. the so called max-min optimization algorithm which is explained more in detail in the following disclosure. It is known that the base station may send one or more user data streams per receiving device, e.g. depending on channel conditions and the number of antennas a given receiving device has. User data streams usually refer to all downlink data streams for a receiving device (such as UE or user device) served by a base station.

In step d) above, the processing device 100 computes the combined power allocation

The computing of the combined power allocation

comprises, in an embodiment, to sum the first power allocation p-L and the second power allocation p₂ so as to compute the combined power allocation

According to an embodiment, the convergence threshold C_Th is a vector of elements. The elements in the convergence threshold C_Th may have a value in an interval between 0.005 - 0.100. In another embodiment, all the elements in the convergence threshold C_Th may have the same threshold value. However, the invention is not limited thereto. In other embodiments, the elements in the convergence threshold C_Th may instead have different threshold values.

In a multi-cell scenario with L number of cells, transforming a downlink minimum sum-mean- squared-error minimum mean squared error (MMSE) problem to the virtual uplink domain, the equivalent optimization problem to solve is the sum MMSE in the virtual uplink, given as:

subject to

where is the downlink channel matrix for links from the y^'th base station to user

equipments (UEs) in the Zth cell, is the precoder at the y^'th base station,

is a diagonal matrix specifying the virtual uplink transmit power allocation,

is the data sent by the UEs, is the additive receiver noise disturbance, G_; is an

estimate of the Hermitian transpose of the downlink channel to all UEs in the system. The diagonal matrix, , specifies the downlink power allocation, and the sum power

constraint at the base station is given by P_TX_BS- The notation tr(-) denotes taking the trace of the matrix argument.

The joint optimization of the precoder and power allocation in equations (1 ) and (2) is a very difficult problem to solve. Hence, an alternating optimization approach to the joint optimization in equations (1 ) and (2) for obtaining the downlink precoder V, and the downlink power allocation p_DL is applied according to an embodiment of the invention. The problem to solve is therefore to jointly determine an optimal downlink precoder V, and an optimal downlink power allocation p_DL, which minimizes the sum of the multiuser interference (i.e. intra-cell interference) and the signal leakage to other cells (i.e. inter-cell interference).

Therefore, in some embodiments of the invention, downlink-uplink duality is used for jointly obtaining the downlink precoder V_DL and the corresponding downlink transmit power allocation p_DL . Also, different interference aware schemes and power allocation criterions used for computing the first power allocation p-L and the second power allocation p₂ , respectively, are used in the various embodiments.

In one embodiment, the interference aware scheme is MCA which means that the precoder V, or the corresponding uplink detector U, are computed by using a criterion of minimum mean squared error of received data symbols at the UEs.

Downlink power allocation, allocated per user data stream, is determined using a power allocation criterion. Two different power allocation algorithms are used according to an embodiment. A first power allocation algorithm which maximizes the minimum user data rate (in short: max-min algorithm) and a second power allocation algorithm which maximize the sum of user data rates (in short: max-sum algorithm). These two different power allocation optimization algorithms have opposing objectives. The max-min algorithm maximizes the minimum user data rate in the system resulting in all UEs having similar data rates and channel conditions. The max-sum algorithm maximizes the sum of user data rates in the system. Maximizing the minimum user data rate will sacrifice the sum of user data rates since more power is allocated to UEs with poor channel conditions to increase their data rates. Alternatively, maximizing the sum of user data rates will sacrifice the minimum user data rate since more power is allocated to UEs with good channel conditions.

Therefore, the first power allocation p_t is in one embodiment computed according to a criterion of maximization of minimum spectral efficiency for the set of receiving devices 600a, 600b,..., 600z. The second power allocation p₂ is in this embodiment computed according to a criterion of maximization of sum rate, in the wireless communication system 500, for the set of receiving device 600a, 600b,..., 600z.

The joint precoder and power allocation optimization problem in the virtual uplink domain in equations (1 ) and (2) is equivalently re-written as

subject to

The joint optimization in equation (3) is a challenging problem to solve since the optimal detector and optimal power allocation are mutually dependent. As aforementioned, in an embodiment MCA is considered together with power allocation criteria of maximization of minimum spectral efficiency and maximization of sum rate. These considerations and respective solutions of the invention may be summarized as:

• Determine an optimal MCA detector, i.e. solve the sum MSE problem in equation (3) for a given virtual uplink power allocation. Therefore, compute the precoder V, or the corresponding uplink detector U, using a criterion of minimum mean squared error of received data symbols.

• Determine virtual uplink power allocation, i.e. given the MCA detector, solve for the virtual uplink power allocation according to selected power allocation criteria. Therefore, compute the first power allocation p-L using a criterion of maximization of minimum spectral efficiency for the set of receiving devices 600a, 600b,..., 600z. Further, compute the second power allocation p₂ using a criterion of maximization of sum rate for the set of receiving devices 600a, 600b,..., 600z.

In the second point above, the objectives for downlink power allocation is to achieve a balanced user experience where all user spectral efficiencies are above a desired minimum spectral efficiency, while simultaneously achieving a high cell spectral efficiency. Since an objective is to improve the minimum user rate in a cell and at the same time improve the cell spectral efficiency, the trade-off between two opposing power allocation optimization objectives discussed above has to be considered. This trade-off is considered by dividing the total available downlink transmit power between the two opposing objectives, or power allocations each using a different criterion. Thus, we select the first fraction of the total transmit power required to achieve a desired minimum user rate, and then use the remaining fraction, i.e. second fraction of the total transmit power for the sum rate maximization.

In further embodiments, the inventor has also realized that the fact that antenna elements are saturated or non-saturated should be considered when computing the second downlink precoder V_DL2. According to this embodiment, the processing device 100 is configured to compute the second downlink precoder V_DL2 for only non-saturated antenna elements in the set of transmit antenna elements 302a, 302b,..., 302z. Generally, this embodiment means it is determined which antenna elements that have saturated power and if an antenna element in a row has saturated power, exclude the row from consideration in the computation of the second downlink precoder V_DL2. This can e.g. be done by using rows of zeros in the matrix second downlink precoder V_DL2 , where the set of rows of zeros in second downlink precoder V_DL2 corresponds to excluding the row of antenna elements from consideration.

In this embodiment, first obtain the first downlink precoder V_DL1 and the first downlink transmit power allocation p_DL1 for respective UEs in a target cell. Next, determine resultant planar antenna array structure to be used for the second downlink precoder V_DL2 and the second downlink transmit power allocation p_DL2 for the UEs in the target cell, by removing or excluding the use of saturated antenna elements. As an example, assume an 8x8 dual polarized planar antenna array of a base station for the determination of the planar antenna array after solving for the first downlink precoder V_DL1 and the first downlink transmit power allocation PDH - For an antenna element with saturated transmit power, the entire row of antenna elements which contains the saturated antenna element is removed or excluded from use when computing the second downlink precoder V_DL2 and the second downlink transmit power allocation p_DL2. In one example, an entire row of antenna elements is removed rather than only removing the saturated antenna element(s) to avoid grating lobes in azimuth. If all antenna elements are saturated after solving for the first downlink precoder V_DL1 and the first downlink transmit power allocation p_DL1, no second downlink precoder V_DL2 or second downlink transmit power allocation p_DL2 is used. Hence, only one downlink precoder is used in that case. Next, make use of remaining available total transmit power to form the second downlink precoder V_DL2 with equal downlink power allocation to same respective UEs subject to the second power constraint C2, i.e.

where is the second downlink precoder V_DL2 at the y^' th base station

indicated by subscript s, and the diagonal matrix,

specifies the corresponding second downlink transmit power allocation p_DL2 at the y^'th base station indicated by subscript B , and m is the antenna index in the range from 1 to the maximum number of antenna elements at the base station. The scalar specifies the maximum output power of an

antenna branch or antenna element of an antenna array of a base station.

The second power constraint C2 is applied in the processing device or the method according to the following steps I to V. This occurs after convergence of the first downlink precoder V_DL1 and its corresponding first downlink power allocation P_DL1 :

I. Set a downlink power increment vector as:

(1, 1, 1), and the length of the vector v_ones is equal to K. The power increment factor can be set to as an example. Also, other values can be used for

II. Initialize a second downlink power allocation P_DL2 for a second downlink precoder V_Di2 for base station in cell j as

III. Set

IV. Determine val2_max using the second power constraint C2 wherein is the maximum

transmitted power on an antenna branch among all antenna branches.

Determine if the corresponding second downlink transmit power P_DL2 for the second downlink precoder V_Di2 can be incremented according to: if v then

increment the power vector according to

and repeat from step III,

else

set the final second downlink transmit power P_DL2 for the second downlink precoder V_DL2 to

Finally, transmit downlink data using the linear superposition of the two downlink precoders, i.e. the first V_DL1 and the second V_DL2 downlink precoders, with respective corresponding downlink power allocations, i.e. P_DL1 and P_DL2 , from the same base station site to the respective UE in the cell. The choice of equal downlink power allocation for the second downlink precoder V_DL2 is an example choice and is selected to minimize computational complexity. Obviously, other choices for the downlink power allocation for the second downlink precoder V_DL2 are possible.

Fig. 5 shows a flowchart of an embodiment in which superimposed precoders with respective downlink power allocations are used together with the inner iterative loop as previously described. Furthermore, Fig. 5 also uses the downlink-uplink duality in a multi-cell scenario. In embodiments of the invention downlink-uplink duality is therefore used. Hence, in processing according to the present solution, either the downlink or the corresponding uplink may be used since conversion functions between the downlink domain/system and the corresponding uplink domain/system (also known as the virtual uplink) are provided within the scope of the invention. Such conversion functions are described and explained in the detailed description herein. Generally, the transformation between the downlink (DL) and the virtual uplink (vUL) is performed by the use of a first conversion function f and a second conversion function f₂, respectively.

In the virtual uplink domain, the minimization of the minimum mean squared error cost function to solve is,

subject to a first power constraint C1 , i.e.

For the first downlink precoder, alternating optimization of the first downlink

precoder V_DL1 and the first downlink transmit power allocation P) = P_DL1 is employed, where the first downlink precoder V_DL1 is optimized in the virtual uplink domain and the two-step power allocation is optimized in the downlink domain. In the second step c of the two step power allocation for the first downlink precoder V_DL1, equal downlink power allocation can be used in place of using a max sum rate criterion to determine an optimized power allocation to reduce complexity. At the (n + l)^th iteration, is the downlink

transmit power allocation vector for user data streams in cell j obtained from the resultant sum of the two power allocation vectors in the downlink domain.

The first power constraint C1 is a per antenna power constraint (PAPC). This specifies the maximum output power of an antenna branch. It is applied to the resultant sum of the two downlink power allocation vectors according to the following steps I to III:

I. Determine: , i.e. the max output transmit power

among all antenna branches for the given downlink precoder Mf and downlink power allocation P_; .

II. Determine: j._e. a scaling factor for the combined power

allocation so that the resultant combined power allocation scaled by the factor meets the first power constraint C1 , i.e. where the updated

satisfies

III. Determine new PAPC constrained resultant power vector for the (n + l ^th iteration:

where each component q of the combined power allocation vector at the (n + l)'^h iteration is scaled by the factor UmitF ^'actor . At a given (n + l)'^h iteration, when the convergence threshold in power allocation condition is met, the method stops. Otherwise, the downlink transmit power allocation at the + iteration is transformed to its corresponding virtual uplink power allocation for the next iteration in the method.

The inner loop in which the first power constraint C1 is used involves Steps 1 to 6 in Fig. 5 whilst Steps 7 to 10 relate to the second power constraint C2. Whether the domain is the downlink or virtual uplink is indicated in the brackets.

Step 1 (vUL): Initialize for the first iteration by obtaining an initial power allocation p_Init for the virtual uplink which e.g. may be an initial equal power allocation.

Step 2 (vUL): Solve for the virtual uplink MCA detector for a given virtual uplink power allocation. The Hermitian transpose of the MCA detector gives the downlink MCA precoder. Step 3 (DL): Use a first fraction of a total available transmit power for a given cell allocated for maximization of minimum spectral efficiency for the set of receiving devices 600a, 600b,..., 600z. Compute the first power allocation p-L given the uplink detector U, from Step 2 and using the criterion of maximization of minimum spectral efficiency for the set of receiving devices 600a, 600b,..., 600z according to

where is the power allocation

vector for UEs in cell I = 1, with aggregate total number of UE antennas K. SINR_K is the signal-to-noise-plus interference ratio for the link between the UE with antenna indexed by κ and its respective serving base station. The power vector component is given by,

with superscript (n + 1) is used to signify the iteration number, and with subscript κ' is used to signify the link between UE with antenna indexed by κ' and its respective serving base station.

Step 4 (DL): Either use equal power allocation for the second remaining fraction of the total available transmit power for a given cell, or the following is performed in Step 4: use the remaining second fraction of the total available transmit power for a given cell allocated for a criterion of maximization of sum rate for the set of receiving device 600a, 600b,..., 600z. Compute the second power allocation p₂ given the uplink detector U, from Step 2 and using the criterion maximization of sum rate for the set of receiving device 600a, 600b,..., 600z according to

where is the second power

allocation vector for UEs in cell I = 1, with aggregate total number of UE antennas K. The quantity SINR_K is the signal-to-noise-plus interference ratio (SNIR) for the link between UE's antenna indexed by κ and its respective serving base station. The vector g_K is the virtual uplink channel between the UE's antenna indexed by κ and the serving base station. The vector u_K is the virtual uplink detector for the received signal at the serving base station, transmitted from UE's antenna indexed by k. F is hence the ratio between the leakage of virtual uplink signal from another UE with antenna index κ as received by detector u_K for UE with antenna index κ and the virtual uplink signal from desired UE with antenna index κ as received by desired detector u_K.

Step 5 (DL): Compute a combined power allocation based on the first power allocation

Pi computed in Step 3 and the second power allocation p₂ computed in Step 4. In one example, sum the first power allocation p-L with the second power allocation p₂ for a given cell to obtain the combined power allocation

as

If the first power constraint C1 described previously is met in Step 5, i.e. "YES" in Fig. 5, the algorithm continues to Step 7. However, if the first power constraint C1 is not met in Step 5, i.e. "NO" in Fig. 5, perform steps I to III related to the first power constraint C1 specified above, after the power constraint C1 equation. Thereafter go to Step 6 in the virtual uplink system.

Step 6 (vUL): Compute virtual uplink power allocation using parameters from the downlink system computed in Step 5 as inputs to the first conversion function f . Thereby, the power allocation in the uplink systems is obtained. The power allocation computed in Step 6 is used as input in Step 2 for computing a new virtual uplink MCA detector.

Iterate/repeat Steps 2 to 6 until the first power constraint C1 is met in Step 5. Thereafter, Steps 7 to 10 are executed in the downlink system as follows.

Step 7 (DL): Identify and remove rows of antenna elements which comprise saturated antenna elements.

Step 8 (DL): For the second downlink precoder V_DL2, employ MCA precoder obtained from the Hermitian transpose of the virtual uplink MCA detector with equal power allocation (EPA) and use equal downlink power allocation with the MCA precoder.

Step 9 (DL): Apply the second power constraint C2 to the resultant sum of first product and second product, where the first product is the product of the first downlink precoder V_DL1 and corresponding first downlink power allocation p_DL1, and the second product is a product of the second downlink precoder V_DL2 and corresponding second power allocation p_DL2 in Step 8 above. Apply the second power constraint C2 using the steps I to V related to the second power constraint C2 described above.

Step 10 (DL): Compute the resultant downlink precoder and power allocation comprising a product of the first downlink precoder V_DL1 and the first downlink transmit power allocation p_DL1, and a second product of the second downlink precoder V_DL2 and the second downlink transmit power allocation p_DL2-

The resultant downlink precoder and power allocation may be used for transmitting downlink data by a transmitting device 300 as described previously.

For providing a more thorough understanding of the downlink-uplink duality used herein the relation between the downlink domain/system parameters/variables and the virtual uplink or transformed domain parameters/variables will be discussed. The mapping between the downlink system variables/parameters and the variables/parameters in the virtual uplink system equivalent to the downlink system are given in Table 1 . The conversion direction from the downlink system to the virtual uplink system, or the reverse direction is indicated in Table 1 . Also, the conversion function or conversion operation is indicated.

Table 1

The first conversion function f_t provides the uplink power allocation Q₇,y = 1, ... , L as output in the transformed system given corresponding downlink input parameters

The first conversion function f is given by

where

The vector v_Jfc is the precoder for UE antenna k in cell j. The calculated downlink SINR, is used in equation (1 1 ) to achieve equivalence in SINR performance, where

for the two domains (downlink and uplink) is desired. The matrix \_KL

is an identity matrix of size KL.

Similarly, the second conversion function f₂ provides the downlink power allocation P) , j = 1, ... , L as output in the downlink system given corresponding uplink input parameters

Specifically, the second conversion function f₂ is given by

where

is a matrix with block diagonal entries

is a vector composed of the diagonal elements of

Equation (14) is obtained from re-arranging the terms in equation (15)

to solve for the downlink power allocation,

and using the calculated values of in place of The scalar, is the transmit power from base

station in cell assigned to UE in cell /with antenna indexed by k. The scalar, is the receiver

noise power at UE's antenna indexed by k.

Similarly, the expression for the first conversion function f above is obtained by solving equation (

for the uplink power allocation Q, and using the calculated values of in place of

The scalar, q_jk , is the virtual uplink transmit power from UE in cell j with

antenna indexed by k to base station in cell j. It is to be noted that the sum power constraint i.e. total transmit power in the downlink system is preserved in the virtual uplink system. Further, when the downlink-uplink duality is applied to a single cell, the sum power constraint is preserved for the given cell, specifically the relation:

When downlink-uplink duality is applied to a wireless communication system or corresponding network, the sum power constraint is preserved for the network, i.e.

where is the total transmit power for the network. For the MCA problem, the

sum power constraint is applied to the network comprising multiple cells. With application of a power allocation algorithm according to an optimization criteria, instead of using a fixed power allocation, a per cell sum power constraint is also introduced in the multi-cell context.

Fig. 6 shows a system architecture of a wireless communication system 500 according to an embodiment of the invention. The wireless communication system 500 comprises in Fig. 6 a plurality of transmitting devices 300a, 300b, 300c in the form of base stations deployed in a multi cell scenario i.e. each cell (not shown) is served by a single base station. Each base station comprises a processing unit 100 (not shown) which executes processing steps such as the one illustrated in Fig. 5. The base stations nodes 300 communicate with each other to exchange relevant scheduling and processing information. Also, a central coordination node 510 is comprised in the wireless communication system 500 and configured to communicate with the base stations using suitable communication interfaces illustrated with the arrows. The central coordination node 510 can be part of a base station or be an independent network control node, such as a radio network controller (RNC). The exchange of information 502 between the base stations via the coordination node 510 is illustrated in Fig. 6. It is also illustrated in Fig. 6 how a base station, such as base station 300a, is configured to obtain the downlink precoder and the downlink power allocation for its cell by using virtual uplink processing 322 and downlink processing 324 according to embodiments of the invention.

The system architecture of Fig. 6 shows that scheduling information regarding which UEs (not shown) are active in a given base station's cell is preferably exchanged with the base stations of other cells. This enables more effective MCA precoding. The user scheduling information relates to information about which UEs that are active or engage in communication sessions in the wireless communication system 500. Further, the transformation of transmit power allocation vectors from downlink to virtual uplink or from virtual uplink to downlink requires exchanging effective channel information (i.e. channel information plus downlink precoder) and downlink transmit power allocations using virtual uplink processing 302 and downlink processing 304 as explained previously.

A receiving device 600z herein may e.g. be any of a User Terminal (UT), a User Equipment (UE), mobile station (MS), wireless terminal or mobile terminal which is enabled to communicate wirelessly in a wireless communication system, sometimes also referred to as a cellular radio system. The UE may further be referred to as mobile telephones, cellular telephones, computer tablets or laptops with wireless capability. The UEs in the present context may be, for example, portable, pocket-storable, hand-held, computer-comprised, or vehicle-mounted mobile devices, enabled to communicate voice or data, via the radio access network, with another entity, such as another receiver or a server. The UE can be a Station (STA), which is any device that contains an IEEE 802.1 1 -conformant Media Access Control (MAC) and Physical Layer (PHY) interface to the Wireless Medium (WM).

A transmitting device 300 herein may be a radio network node or an access node or an access point or a Base Station (BS), e.g., a Radio Base Station (RBS), which in some networks may be referred to as transmitter, "eNB", "eNodeB", "NodeB" or "B node", depending on the technology and terminology used. The radio network nodes may be of different classes such as, e.g., macro eNodeB, home eNodeB or pico base station, based on transmission power and thereby also cell size. The radio network node can be a Station (STA), which is any device that contains an IEEE 802.1 1 -conformant Media Access Control (MAC) and Physical Layer (PHY) interface to the Wireless Medium (WM).

Furthermore, methods according to embodiments of the invention may be implemented in a computer program, having code means, which when run by processing means causes the processing means to execute the steps of the method. The computer program is included in a computer readable medium of a computer program product. The computer readable medium may comprise of essentially any memory, such as a ROM (Read-Only Memory), a PROM (Programmable Read-Only Memory), an EPROM (Erasable PROM), a Flash memory, an EEPROM (Electrically Erasable PROM), or a hard disk drive. Moreover, it is realized by the skilled person that the processing device 100 may comprise the necessary communication capabilities in the form of e.g., functions, means, units, elements, etc., for performing the present solution. Examples of other such means, units, elements and functions are: processors, memory, buffers, control logic, encoders, decoders, rate matchers, de-rate matchers, mapping units, multipliers, decision units, selecting units, switches, interleavers, de-interleavers, modulators, demodulators, inputs, outputs, antennas, amplifiers, receiver units, transmitter units, DSPs, MSDs, TCM encoder, TCM decoder, power supply units, power feeders, communication interfaces, communication protocols, etc. which are suitably arranged together for performing the present solution.

Especially, the processor(s) of the present processing device 100 may comprise, e.g., one or more instances of a Central Processing Unit (CPU), a processing unit, a processing circuit, a processor, an Application Specific Integrated Circuit (ASIC), a microprocessor, or other processing logic that may interpret and execute instructions. The expression "processor" may thus represent a processing circuitry comprising a plurality of processing circuits, such as, e.g., any, some or all of the ones mentioned above. The processing circuitry may further perform data processing functions for inputting, outputting, and processing of data comprising data buffering and device control functions, such as call processing control, user interface control, or the like.

Finally, it should be understood that the invention is not limited to the embodiments described above, but also relates to and incorporates all embodiments within the scope of the appended independent claims.

Claims

1 . A processing device (100) for a wireless communication system (500), the processing device (100) being configured to

obtain a first downlink precoder (V_DL1) for a set of transmit antenna elements (302a,

302b,..., 302z) associated with a set of receiving devices (600a, 600b,..., 600z);

obtain a first downlink transmit power allocation (p_DL1 ) for a total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z);

if a fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z) remains after having obtained the first downlink transmit power allocation

(PDLI):

compute at least one second downlink precoder (V_DL2) for the set of transmit antenna elements (302a, 302b,..., 302z) based on a power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z);

compute at least one second downlink transmit power allocation (p_DL2) for the remaining fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z) based on the second downlink precoder (V_DL2);

compute a resultant downlink precoder and power allocation comprising a first product of the first downlink precoder (V_DL1) and the first downlink transmit power allocation (p_DL1), and a second product of the second downlink precoder (V_DL2 ) and the second downlink transmit power allocation (p_DL2)-

2. The processing device (100) according to claim 1 , configured to

3. The processing device (100) according to claim 2, configured to

compute the resultant downlink precoder and power allocation comprising the linear combination of the first product and the second product constrained on a second power constraint (C2) relating to maximum transmit powers for each transmit antenna element in the set of transmit antenna elements (302a, 302b,..., 302z).

4. The processing device (100) according to claim 3, configured to

re-compute the second downlink transmit power allocation (p_DL2) so as to fulfil the second power constraint (C2).

5. The processing device (100) according to any of the preceding claims, configured to compute the second downlink precoder ( V_DL2 ) for only non-saturated antenna elements in the set of transmit antenna elements (302a, 302b,..., 302z).

6. The processing device (100) according to any of the preceding claims, configured to

obtain the first downlink precoder (V_DL1) for a set of transmit antenna elements (302a, 302b,..., 302z) associated with a set of receiving devices (600a, 600b,..., 600z), and

obtain the first downlink transmit power allocation (p_DL1) for the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z)

by being configured to:

obtain an initial power allocation (p_/nit) for the set of transmit antenna elements (302a, 302b,..., 302z) associated with the set of receiving devices (600a, 600b,..., 600z);

a) compute a precoder (V_; ) or a corresponding uplink detector (U_; ) based on the initial power allocation (p_/nit) or a combined power allocation from a previous iteration,

b) compute a first power allocation (p- for a first fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,... , 600z) based on the precoder (V_; ) or the corresponding uplink detector (U_; ),

c) compute a second power allocation (p₂ ) for a remaining second fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,... , 600z) based on the precoder (V_; ) or the corresponding uplink detector (U_; ),

d) compute a combined power allocation based on the first power allocation (p_t)

and the second power allocation (p₂ ),

repeat a) to d) until the combined power allocation satisfies a convergence threshold

(C_Th) in a final iteration;

obtain the first downlink precoder (V_DL1) for the set of transmit antenna elements (302a,

302b,..., 302z) based on at least one of the precoder (V_; ) or the corresponding uplink detector (U, ), and the combined power allocation in the final iteration;

obtain the first downlink transmit power allocation (p_DL1) for the set of receiving devices (600a, 600b,..., 600z) based on the combined power allocation in the final iteration.

7. The processing device (100) according to claim 6, wherein the initial power allocation

is an equal power allocation for the set of transmit antenna elements (302a, 302b,..., 302z).

8. The processing device (100) according to claim 6 or 7, configured to

compute the precoder (V_; ) or the corresponding uplink detector (U, ) using a criterion of minimum mean squared error of received data symbols for an interference aware scheme.

9. The processing device (100) according to any of claims 6 to 8, configured to compute the first power allocation (p- using a criterion of maximization of minimum spectral efficiency for the set of receiving devices (600a, 600b,..., 600z).

10. The processing device (100) according to any of claims 6 to 9, configured to

compute the second uplink power allocation (p₂) using a criterion of maximization of sum rate for the set of receiving device (600a, 600b,..., 600z).

1 1. The processing device (100) according to any of claims 6 to 10, configured to

compute the combined power allocation comprising a sum of the first power

allocation (p_t) and the second power allocation (p₂).

12. The processing device (100) according to claim 1 1 , configured to

compute the combined power allocation based on the sum of the first power

allocation (p- and the second power allocation (p₂) constrained on a first power constraint (C1 ) related to maximum transmit powers for each transmit antenna element in the set of transmit antenna elements (302a, 302b,..., 302z).

13. The processing device (100) according to any of claims 6 to 12, configured to

obtain the initial power allocation (p_/nit) as an initial uplink power allocation, compute the corresponding uplink detector (U,) in a).

14. The processing device (100) according to claim 13, wherein the first power allocation (p- is a first uplink power allocation and the second power allocation (p₂) is a second uplink power allocation such that the combined power allocation is a combined uplink power

allocation.

15. The processing device (100) according to claim 14, configured to

compute an uplink Signal-to-Noise plus Interference Ratio (SNIR_UL ) based on the combined uplink power allocation in the final iteration;

compute the downlink transmit power allocation (p_DL) based on the combined uplink power allocation and the uplink Signal-to-Noise plus Interference Ratio (SNIR_UL).

16. The processing device (100) according to claim 15, wherein the first power allocation (p- is a first downlink power allocation and the second power allocation (p- is a second downlink power allocation such that the combined power allocation is a combined downlink

power allocation.

17. The processing device (100) according to claim 16, configured to

compute an uplink power allocation based on the combined downlink power allocation in d);

compute the corresponding uplink detector (U, ) in a) based on the uplink power allocation.

18. A transmitting device (300) for a wireless communication system (500), the transmitting device (300) being configured to

obtain a resultant downlink precoder and power allocation according to any of preceding claims;

perform downlink transmissions to the set of receiving devices (600a, 600b,..., 600z) using the resultant downlink precoder and power allocation.

19. A method (200) for a wireless communication system (500), the method (200) comprising: obtaining (202) a first downlink precoder (V_DL1) for a set of transmit antenna elements (302a, 302b,..., 302z) associated with a set of receiving devices (600a, 600b,..., 600z); obtaining (204) a first downlink transmit power allocation (p_DL1) for a total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z);

computing (206) at least one second downlink precoder (V_DL2 ) for the set of transmit antenna elements (302a, 302b,..., 302z) based on a power allocation for the remaining fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z);

computing (208) at least one second downlink transmit power allocation (p_DL2) for the remaining fraction of the total transmit power allocated to the set of receiving devices (600a, 600b,..., 600z) based on the second downlink precoder (V_DL2 );

computing (210) a resultant downlink precoder and power allocation comprising a first product of the first downlink precoder (V_DL1) and the first downlink transmit power allocation

and a second product of the second downlink precoder (V_DL2 ) and the second downlink transmit power allocation (p_DL2)-

20. A computer program product comprising a computer program with program code for performing a method according to claim 19.