WO2017101999A1

WO2017101999A1 - Precoding device for cancelling asymmetrically known interference

Info

Publication number: WO2017101999A1
Application number: PCT/EP2015/080080
Authority: WO
Inventors: Abdellatif ZAIDI; Stefano Tomasin
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2015-12-16
Filing date: 2015-12-16
Publication date: 2017-06-22
Also published as: CN108476184A; CN108476184B

Abstract

The disclosure relates to a precoding device (300) for cancelling asymmetrically known interference, the precoding device (300) comprising: a first transmitter node (Tx1) configured to generate a first input signal (X₁) carrying a first message (c_W1) intended for transmission to a first reception node (Rx1); a second transmitter node (Tx2) configured to generate a second input signal (X₂) carrying a second message (C_W2) intended for transmission to the first reception node (Rx1) and an interference signal (X) intended for transmission to another reception node (Rx2), wherein the interference signal (X) is an asymmetrically known interference signal that is known to the second transmitter node (Tx2) and unknown to the first transmitter node (Tx1 ), wherein the second transmitter node (Tx2) comprises: an interference canceller (305) configured to cancel part of the interference signal (X) to generate a partially cancelled interference signal (X_c), and a precoder (303), configured to use a lattice-based precoding scheme to precode the second message (W₂) with the non-cancelled part (X_nc) of the interference signal (X) to generate a precoded signal (X_2W), wherein the second transmitter node (Tx2) is configured to generate the second input signal (X₂) based on the partially cancelled interference signal (X_c) and the precoded signal (X_2W).

Description

Precoding device for cancelling asymmetrically known interference

TECHNICAL FIELD The present disclosure relates to a precoding device and a method for cancelling

asymmetrically known interference. In particular, the disclosure relates to lattice-based precoding techniques for cancelling asymmetrically known interference.

BACKGROUND

Interference is one of the most limiting factors of communication networks. In certain cases, the effect of interference can be mitigated through some precoding techniques. For additive Gaussian interference, the publication "Max H. M. Costa, Writing on dirty paper, IEEE Transactions on Information Theory, Vol. 29, 1983, pp. 439-441 " showed that an instance of Gel'fand-Pinsker precoding according to "S. I. Gel'fand and M. S. Pinsker, "Coding for channel with random parameters," Problems of Control and Information Theory, vol. 9, no. 1 , pp. 19-31 , 1980", widely known as dirty paper coding (DPC), achieves the channel capacity of point-to-point channels with interference known beforehand, or non-causally, to only the transmitter, i.e., not to the receiver. Interestingly, in this case, DPC completely removes the effect of the interference allowing the same rates as if there were no interference at all.

The possibility of removing the effect of interference was later shown for other channel configurations, including Gaussian broadcast channels, Gaussian multi-access channels, physically degraded relay channels and physically degraded relay broadcast channels. For all these models, the key feature for complete interference removal is the symmetric availability of the interference at all the encoders. If the interference is only available to some encoders and not to others, i.e., the asymmetric case, complete mitigation is not possible and, in general, a rate penalty is incurred.

SUMMARY

It is the object of the invention to provide a concept for improving interference mitigation in communication networks in which interference is asymmetrically known.

This object is achieved by the features of the independent claims. Further implementation forms are apparent from the dependent claims, the description and the figures. As mentioned above, the present disclosure relates to the asymmetric scenario that is exemplary illustrated in Fig. 1 . In Figure 1 two transmitters (Tx1 and Tx2) and two receivers (Rx1 and Rx2) are depicted, where both transmitters communicate with the first receiver (Rx1 ), and only the second transmitter (Tx2) also sends information to the second receiver (Rx2). Thus, transmission from Tx2 to Rx2 yields interference to Rx1 . This interference is known, beforehand, asymmetrically only to Tx2 but not to Tx1 . A new lattice-based precoding technique is presented that partially cancels interference and then applies a suitable precoding. It is shown, see Figure 5, that this approach outperforms precoding-only solutions such as DPC and Tomlinson-Harashima precoding (THP) according to "M. Tomlinson, New automatic equalizer employing modulo arithmetic, IEEE Electronic Letter, Vol. 07, 1971 , pp. 138-139" and "H. Harashima and H. Miyakawa, Matched transmission technique for channels with intersymbol interference, IEEE Transactions on Communications, Vol. 20, 1972, pp. 774-780". The performance improvement is a remarkable result, since in the symmetric scenario (in which Tx1 as well knows the message to be transmitted by Tx2 to Rx2) it is well known that interference cancellation is suboptimal with respect to DPC.

A basic idea of the invention is to apply a novel lattice-based precoding strategy that better mitigates the effect of the interference at Rx1 in the asymmetric interference scenario compared to THP.

In order to describe the invention in detail, the following terms, abbreviations and notations will be used:

THP: Tomlinson-Harashima Precodi

DPC: Dirty Paper Coding

Tx: Transmitter

Rx: Receiver

Systems, devices and methods according to the disclosure may be applied in Multi-Input Multi-Output (MIMO) systems, e.g. MIMO systems as depicted in Fig. 1 and described in the following.

In the communication scenario shown in Fig. 1 two transmitter nodes (Tx1 and Tx2) communicate with a common receiver (Rx1 ). w₁ and W₂ are the messages to be transmitted respectively by Tx1 and Tx2 to Rx1 ; and X₁ and X₂ are the corresponding input signals. The second transmitter (Tx2) has also another message to send to a second receiver (Rx2), and communicates with it by transmitting some signal X. The channels among users are assumed to be flat fading Additive White Gaussian Noise (AWGN), although more general cases are possible and will be detailed in the patent application. The coefficients

h_lt, h₂₁ and h₂₂ are the channel gains of the Tx1 -Rx1 , Tx2-Rx1 and Tx2-Rx2 links, respectively. The channel coefficients are known, or can be estimated to high accuracy, by all nodes in the network model.

The transmission scheme is as follows: Node Tx1 transmits a signal X₁ (that carries a message W ) that is intended for Rx1 ; Node Tx2 transmits a signal X₂ (that carries a message W₂) that is intended for Rx1 ; and a signal X that is intended for Rx2. The inputs x_t , x₂ and X are subjected to respective average power constraints P₁ , P₂ and Q .

Here, as mentioned previously, the transmission to Rx2 creates interference to Rx1 . The channel input-output relationship is given by:

Y₁ = h X₁ + h₂₁ X₂ + h₂₁X + Z₁

where z₁ is some additive white noise, assumed to be zero-mean Gaussian with given variance N, and independent from all other signals.

Note that Fig. 1 relates to a simple model with single antenna terminals and no frequency selectivity. This is done for the ease of the description, but the systems, devices and methods according to the disclosure may also be applied to more elaborate scenarios including multiple (massive) MIMO systems, frequency selective channels (possibly using Orthogonal Frequency Division Multiplexing (OFDM) and other multi-carrier systems) and multiple access schemes.

Systems, devices and methods according to the disclosure may use pre-coders that apply THP and DPC precoding techniques as described in the following.

In the case in which both transmitters Tx1 and Tx2 know the interference signal, they can employ a joint DPC according to "Y.-H. Kim, A. Sutivong, and S. Sigurjonsson, "Multiple user writing on dirty paper," in Proceedings of the IEEE International Symposium on Information Theory, p. 534, Chicago, III, USA, June 2004" to cancel its effect completely. For the scenario that is described in this disclosure, the interference is known asymmetrically only to Tx2; and so only Tx2 can apply DPC or THP.

With THP, Tx2 generates X₂ in a manner that accounts for that interference as (see Fig. 1 ):

¾ = /¾ [c_W2 - «2^ - d ] mod Λ where c_w2 is a symbol or codeword that is associated through one-to-one mapping with the message W₂ from Tx2 to Rx2, d is a dither, the operation mod denotes the modulo reduction, Λ is a given lattice of dimension n (e.g., cubic lattice zⁿ, Hexagonal lattice A₂ , the

Checkerboard lattice D₄), β₂ is some scaling factor chosen so as to adjust the power at Tx2, and a₂ is some scaling factor whose choice can be unity (for ZF-THP) or the Wiener parameter (for MMSE-THP). The transmitter Tx1 sends its input as

X = βι [^cwi - d ] mod Λ

where c_wl is a symbol that is associated through one-to-one mapping with the message W₁ from Tx1 to Rx1 and β₁ is some scaling factor chosen so as to adjust the power at Tx1 . The receiver Rx1 decodes the messages from the two transmitters successively using standard modulo-lattice reduction operations.

Systems, devices and methods according to the disclosure may use pre-coders that apply Lattice-based precoding strategies as described in the following.

Lattice strategies for canceling known interference in a single user channel were studied by Erez et al., see "Uri Erez, Shlomo Shamai (Shitz) and Ram Zami, "Capacity and Lattice Strategies for Canceling Known Interference, IEEE Trans. On Info. Theory, Vol. 51 , No. 1 1 , Nov. 2005, pp. 3820-3833". The idea that informed encoders can help non-informed encoders through some generalization of DPC was studied in "A. Somekh-Baruch, S.

Shamai (Shitz), and S. Verdu, "Cooperative multiple access encoding with states available at one transmitter," IEEE Trans. Inf. Theory, vol. 54, no. 10, pp. 4448-4469, Oct. 2008" and "S. Kotagiri and J. N. Laneman, "Multiaccess channels with state known to some encoders and independent messages," EURASIP J. Wireless Commun. Netw., 2008, article ID 450680" for multiple access channel with degraded messages sets and in "A. Zaidi, S. P. Kotagiri, J. N. Laneman, and L. Vandendorpe "Cooperative Relaying with State Available Non-Causally at the Relay," IEEE Trans. Inf. Theory, vol. 5, no. 56, pp. 2272-2298, May 2010" and "A. Zaidi, S. Shamai (Shitz), P. Piantanida and L. Vandendorpe , "Bounds on the Capacity of the Relay Channel with Noncausal State at Source," IEEE Trans, on Inf. Theory, Vol. 59, No. 5, May 2013, pp. 2639-2672" for relay channels, all using non-feasible random codes.

A means of avoiding transmit power enhancement is to use non-linear precoding, or lattice- based precoding, where a modulo operation or vector quantization is used to reduce transmit power enhancement. The main idea is that an extended constellation is used at the transmitter with multiple equivalent points with the original points in the fundamental constellation boundary. The modulo operation finds a proper point in the fundamental boundary equivalent with a distorted point that the original point moves to in the extended region by power normalization. Tomlinson-Harashima MIMO precoding is one example of transmit precoding with a modulo operation. According to a first aspect, the invention relates to a precoding device for cancelling asymmetrically known interference, the precoding device comprising: a first transmitter node configured to generate a first input signal carrying a first message intended for transmission to a first reception node; a second transmitter node configured to generate a second input signal carrying a second message intended for transmission to the first reception node and an interference signal intended for transmission to another reception node, wherein the interference signal is an asymmetrically known interference signal that is known to the second transmitter node and unknown to the first transmitter node, wherein the second transmitter node comprises: an interference canceller configured to cancel part of the interference signal to generate a partially cancelled interference signal, and a precoder, configured to use a lattice-based precoding scheme to precode the second message with the non-cancelled part of the interference signal to generate a precoded signal, wherein the second transmitter node is configured to generate the second input signal based on the partially cancelled interference signal and the precoded signal. This provides the advantages of improved interference mitigation in communication networks in which interference is asymmetrically known. In particular the advantages of more efficient precoding against interference, better throughput in comparison to DPC and THP and low complexity solution using structured coding (lattices) can be implemented. In a first possible implementation form of the precoding device according to the first aspect, the precoder is configured to use a Tomlinson-Harashima precoding scheme for precoding the second message with the non-cancelled part of the interference signal.

By using only the non-cancelled part of the interference signal for message precoding, the average transmit power is not increased when a THP scheme is employed.

In a second possible implementation form of the device according to the first aspect as such or according to the first implementation form of the first aspect, the interference canceller comprises a multiplier unit configured to multiply the interference signal with a cancellation factor to generate the partially cancelled interference signal. Using a cancellation factor provides the advantage of a flexible choice between the amount of interference suppression and precoding using the interference.

In a third possible implementation form of the precoding device according to the second implementation form of the first aspect, the cancellation factor depends on a power of the second input signal.

This provides the advantage that the cancellation factor is adjustable depending on the power of the second input signal. Interference suppression can be flexibly adjusted depending on the power of the second input signal in order to provide an optimal transmission. Then, the power of the second input signal is split among canceling a part of the interference directly using a part of its input, and applying an appropriate THP with the remaining part of its input. In a fourth possible implementation form of the precoding device according to the third implementation form of the first aspect, the cancellation factor depends on a negative correlation coefficient indicating a correlation between the second input signal and the interference signal. A high correlation coefficient indicates a strong interference, which is heavily disturbing the message transmission. On the other hand, a low correlation coefficient indicates a weak interference signal, with less impact on the message transmission. By choosing the cancellation factor based on the correlation coefficient the amount of interference suppression can flexibly be varied. This provides the advantage that interference can be controlled by adjusting a high correlation coefficient, for example between -1 and -0.8 or by adjusting a low correlation coefficient, for example between -0.2 and 0.

In a fifth possible implementation form of the precoding device according to the fourth implementation form of the first aspect, the cancellation factor depends on a power of the interference signal.

An interference signal with a high power is heavily disturbing the message transmission, and can, by choosing the appropriate cancellation factor be suppressed effectively to a high degree. For example a threshold can be chosen as P2/Q < 0.5 corresponding to high interference power and P2/Q> 0.5 corresponding to low interference power. In a sixth possible implementation form of the precoding device according to the fifth implementation form of the first aspect, the cancellation factor depends on the relation ί \ (Q} I _' ^Where P '^{S the correlation coefficien p}2 is the power of the second input signal and Q is the power of the interference signal.

A part of the interference is cancelled directly by using a part of the input signal X₂.

This provides the advantage that the cancellation factor is adjustable depending on characteristics of the input signal and hence interference suppression can be flexibly adjusted in order to provide an optimal transmission.

In a seventh possible implementation form of the precoding device according to the sixth implementation form of the first aspect, the precoder is configured to generate the precoded signal depending on the rel where p is the correlation coefficient, P₂ is

the power of the second input signal and Q is the power of the interference signal.

This provides the advantage that the precoder is adjustable depending on the above relation and hence precoding can be flexibly adjusted in order to provide an optimal transmission. I.e. the precoder provides a more efficient precoding against interference, has a better throughput in comparison to DPC and THP and a low complexity due to using structured coding (lattices).

In an eighth possible implementation form of the device according to the first aspect as such or according to any of the preceding implementation forms of the first aspect, the second transmitter node comprises an adding unit configured to add the partially cancelled interference signal and the precoded signal to generate the second input signal.

This provides the advantage that one single signal can be generated that carries both, the precoded signal carrying a message intended for transmission to the first reception node and the partially cancelled interference signal carrying a message intended for transmission to another reception node. In a ninth possible implementation form of the device according to the first aspect as such or according to any of the preceding implementation forms of the first aspect, the second transmitter node comprises an optimizer configured to generate optimal parameters for adjusting the interference canceller and the precoder with respect to a maximum sum transmission rate of the first message and the second message.

This provides the advantage that using the optimal parameters results in a maximum sum transmission rate of both messages. Note that optimal parameters can be determined numerically.

In a tenth possible implementation form of the precoding device according to the ninth implementation form of the first aspect, the optimal parameters comprise: a negative correlation coefficient indicating a correlation between the second input signal and the interference signal, and a scaling factor for scaling the non-cancelled part of the interference signal.

This provides the advantage of high flexibility for increasing the sum transmission rate.

In an eleventh possible implementation form of the device according to the ninth or the tenth implementation form of the first aspect, the optimizer is configured to generate the optimal parameters based on the following relation :(p, a) = argmax R (P₁, P₂, Q, p, a) ,

aER,-l≤p≤0

where R(P₁, P₂, Q, p, ) is the sum rate obtained with given parameters, and p is the correlation coefficient, a is a precoding parameter, Pi is the power of the first input signal, P₂ is the power of the second input signal and Q is the power of the interference signal.

This provides the advantage that by using such optimization, the sum transmission rate can be maximized.

According to a second aspect, the invention relates to a method for cancelling

asymmetrically known interference, the method comprising: generating a first input signal carrying a first message intended for transmission to a first reception node; generating a second input signal carrying a second message intended for transmission to the first reception node; generating an interference signal, wherein the interference signal is an asymmetrically known interference signal that is known for generating the second input signal and unknown for generating the first input signal; cancelling part of the interference signal to generate a partially cancelled interference signal, and precoding the second message with the non-cancelled part of the interference signal by using a lattice-based precoding scheme to generate a precoded signal, wherein generating the second input signal is based on the partially cancelled interference signal and the precoded signal. This provides the advantages of improved interference mitigation in communication networks in which interference is asymmetrically known. In particular the advantages of more efficient precoding against interference, better throughput in comparison to DPC and THP and low complexity implementation using structured coding (lattices) can be realized by such method. In a first possible implementation form of the method according to the second aspect, the method comprises: using a Tomlinson-Harashima precoding scheme for precoding the second message with the non-cancelled part of the interference signal.

This provides the advantage that THP increases the average transmit power. No error propagation may occur since the feedback filter is located at the transmitter where the signals are perfectly known.

According to a third aspect, the invention relates to a lattice-based precoding method for mitigating the effect of asymmetrically known interferences in cellular networks.

Such a method provides the advantage of appropriately combining direct interference cancellation and precoding.

In a first possible implementation form of the method according to the third aspect, parameters used for the precoding are optimized with the aim of maximizing the sum rate of the system.

This provides the advantage of increased data throughput.

BRIEF DESCRIPTION OF THE DRAWINGS

Further embodiments of the invention will be described with respect to the following figures, in which:

Fig. 1 shows a schematic diagram of a MIMO communication scenario 100; Fig. 2 shows a block diagram of a pre-coder module 200 applying the Tomlinson-Harashima precoding (THP) scheme;

Fig. 3 shows a schematic diagram illustrating a precoding device 300 for cancelling asymmetrically known interference according to an implementation form;

Fig. 4 shows a schematic diagram illustrating a precoding device 400 for cancelling asymmetrically known interference according to an implementation form; Fig. 5 shows a performance diagram 500 illustrating an exemplary throughput of a precoding device according to the disclosure;

Fig. 6 shows a schematic diagram illustrating a wireless communication network 600 applying a multiple access with relaying technique according to an implementation form; and

Fig. 7 shows a schematic diagram illustrating a method 700 for cancelling asymmetrically known interference according to an implementation form.

DETAILED DESCRIPTION OF EMBODIMENTS

In the following detailed description, reference is made to the accompanying drawings, which form a part thereof, and in which is shown by way of illustration specific aspects in which the disclosure may be practiced. It is understood that other aspects may be utilized and structural or logical changes may be made without departing from the scope of the present disclosure. The following detailed description, therefore, is not to be taken in a limiting sense, and the scope of the present disclosure is defined by the appended claims.

It is understood that comments made in connection with a described method may also hold true for a corresponding device or system configured to perform the method and vice versa. For example, if a specific method step is described, a corresponding device may include a unit to perform the described method step, even if such unit is not explicitly described or illustrated in the figures. Further, it is understood that the features of the various exemplary aspects described herein may be combined with each other, unless specifically noted otherwise. Fig. 3 shows a schematic diagram illustrating a precoding device 300 for cancelling asymmetrically known interference according to an implementation form.

The precoding device 300 includes a first transmitter node Tx1 and a second transmitter node Tx2. The second transmitter node Tx2 includes an interference canceller 305 and a precoder 303. The first transmitter node Tx1 generates a first input signal Xi carrying a first message cwi intended for transmission to a first reception node Rx1 . The second transmitter node Tx2 generates a second input signal X₂ carrying a second message W₂ intended for transmission to the first reception node Rx1 and an interference signal X intended for transmission to another reception node, e.g. Rx2. The interference signal X is an

asymmetrically known interference signal that is known to the second transmitter node Tx2 and unknown to the first transmitter node Tx1 . The interference canceller 305 cancels part of the interference signal X to generate a partially cancelled interference signal X_c. The precoder 303 uses a lattice-based precoding scheme to precode the second message W₂ with the non-cancelled part X_nc of the interference signal X to generate a precoded signal X₂w. The second transmitter node Tx2 generates the second input signal X₂ based on the partially cancelled interference signal Xc and the precoded signal X₂w.

The precoder 303 may for example use a Tomlinson-Harashima precoding scheme (THP) for precoding the second message W₂ with the non-cancelled part X_nc of the interference signal X. Alternatively, the precoder 303 may use another DPC precoding scheme.

The interference canceller 305 may include a multiplier unit, e.g. a multiplier unit 407 as described below with respect to Fig. 4 to multiply the interference signal X with a cancellation factor, e.g. a cancellation factor 402 as described below with respect to Fig. 4 to generate the partially cancelled interference signal Xc.

The cancellation factor 402 may depend on a power P₂ of the second input signal X₂, e.g. as described below with respect to Fig. 4. The cancellation factor 402 may depend on a negative correlation coefficient p indicating a correlation between the second input signal X₂ and the interference signal X, e.g. as described below with respect to Fig. 4. The cancellation factor 402 may depend on a power Q of the interference signal X, e.g. as described below with respect to Fig. 4. The cancellation factor 402 may depend on the relation

where p is the correlation coefficient, P₂ is the power of the second input signal X₂ and Q is the power of the interference signal X as described below with respect to Fig. 4. The precoder 303 may generate the precoded signal X₂w depending on the relation , where p is the correlation coefficient, P₂ is the power of the second input

signal X₂ and Q is the power of the interference signal X, e.g. as described below with respect to Fig. 4.

The second transmitter node Tx2 may include an adding unit 307 for adding the partially cancelled interference signal X_c and the precoded signal X₂w to generate the second input signal X₂.

The precoder 303 may provide the precoded signal X₂w statistically independent of the interference signal X.

The second transmitter node Tx2 may include an optimizer, e.g. an optimizer 403 as described below with respect to Fig. 4 to generate optimal parameters for adjusting the interference canceller 305 and the precoder 303 with respect to a maximum sum

transmission rate of the first message cwi and the second message Cw₂.

These optimal parameters may include a negative correlation coefficient p indicating a correlation between the second input signal X₂ and the interference signal X, and a scaling factor a for scaling the non-cancelled part X_nc of the interference signal X, e.g. as described below with respect to Fig. 4.

The optimizer 403 may generate the optimal parameters based on optimizing the sum transmission rate depending on a power Pi of the first input signal Xi , a power P₂ of the second input signal X₂, a power Q of the interference signal X and the optimal parameters as a function of the optimal parameters, e.g. as described below with respect to Fig. 4.

Fig. 4 shows a schematic diagram illustrating a precoding device 400 for cancelling asymmetrically known interference according to an implementation form. The precoding device 400 may be implemented in a second transmission node Tx2, e.g. a second transmission node Tx2 as described above with respect to Fig. 3.

In the apparatus 400, Tx2 cancels part of the interference X and applies an appropriate precoding block 401 . The precoding block 401 computes

and -1 < p≤ 0 is a negative correlation coefficient. That is, by opposition to the strategy described above with respect to Figures 1 and 2, where the second transmitter Tx2 spends all its power to apply a THP, in the precoding device 400 the Tx2 power is split among: canceling a part of the interference directly, using the part p of its input _{2 >} ^witn

power p²P₂, and applying an appropriate THP with the remaining (1 - p²)P₂ of its power.

Then the second transmitter Tx2 sends

The part X_2w of the input of the second transmitter Tx2 is independent of the interference X and carries the message to Tx1. The rationale behind the above modified THP for obtaining X_2w can be seen by writing the output at Rx1 as

Y₁ = X + h₂₁ X_2w + h₂₁ [ 1 + p

where it can be found that the equivalent interference to precode against it is

+ ^x- ^Also' ^tne intuitive reason for which the partial interference cancellation

helps here can be seen from the above equation by noticing that now Tx1 faces a less strong interference, since ( fQ + p p^~ ) ≤ Q. Block a and p optimizer 403 in Fig. 4 computes the optimal values of parameters p and a to be used by the Precoding block 401 . Optimization (in order e.g., to maximize the sum rate) can be performed by numerical methods. For example, the optimization can be performed to maximize the sum rate, i.e.,

(p, a) = argmax R(P₁, P₂, Q,p, a),

aeR,-l≤p≤0

where P₂, Q, p, a) is the sum rate obtained with given parameters.

When comparing Fig. 4 with Fig 1 , it can be noted that the presented precoding scheme according to the disclosure is significantly different in the way in which the transmitted signal X₂ is generated. It comprises new blocks Precoding 401 and a and p optmizer 403 and different operations performed on X to obtain X₂. Note that had the second transmitter been alone (i.e., no Tx1 ) then it would have made no sense that this transmitter spends any part of its power to cancel the effect of the

interference X at Tx1 directly, since a standard THP would have been sufficient to cancel its effect completely. Here, it is precisely because Tx1 suffers from the interference X, which the concurrent transmitter Tx2 knows, that it is beneficial that the latter spends a non-zero part of its power to cancel a part of the interference directly, so as to reduce the interference seen by Tx1 . To see this, note that because Tx2 cancels a part of the interference directly (recall that p is non-positive), Tx1 sees an interference whose power is |/i₂i

which is strictly smaller than the stronger interference with power \h₂₁ \² Q which would be seen had Tx2 only applied a standard THP.

The maximizing p in the expression of the throughput is generally strictly negative (i.e., nonzero); and this explains the usefulness of canceling the interference partially by Tx2 so that Tx1 benefits from that cancellation.

Note that according to Costa's DPC principle, spending power on cancelling directly the interference known at Tx2 would have been completely useless if Tx1 also were aware of the interference. However, in the setting depicted in Fig. 4 this turns out to be useful precisely because Tx1 and Tx2 interfere and only Tx1 is not aware of the interference. That is, although Tx2 spends power on canceling the interference, the total sum rate is increased.

Fig. 5 shows a performance diagram 500 illustrating an exemplary throughput of a precoding device according to the disclosure. For the throughput validation sphere lattices with the following parameters were used: h_xx = 1, h₂₁ = 0.2, P₁ =1 ,P₂ = 1 - Q,N = 0.1. The solid lines illustrate the performance of a precoding device according to the disclosure while the dashed lines illustrate the performance of THP.

The close form expression of the sum transmission rates of the messages W₁ and W₂ to Rx1 can be derived as a function of the parameters a and p described above with respect to Fig. 4 for the considered scheme as

Note that the signal X acts as interference and does not carry any useful signal to Rx1 . The performance of the standard TH P is obtained by letting p = 0. From the above equation (**) it can be immediately seen that the presented solution is at least as good as TH P, since it includes it as a special case. In Fig. 5, the solid-line curve depicts the evolution of the sum-rate R_SUm (.^pi_> ^p2_> (?) ⁱⁿ the ^case of Sphere lattice (very large dimension n) as a function of the power Q of the interference, for some numerical values of the powers P₁ = l and P₂ = 1 - (? , as well as the channel gains h_l = 1, h₂₁ = 0.2 and h₂₂ = 1 and the noise variance N = 0.1. Also shown for comparison, the dashed-line curve is the sum-rate obtained using only TH P at Tx2, i.e., no partial interference cancellation, or, equivalently, p = 0. As shown from the figure, the precoding device according to the disclosure provides substantial sum-rate improvement over standard TH P for certain regimes of the interference Q.

Fig. 6 shows a schematic diagram illustrating a wireless communication network 600 applying a multiple access with relaying technique according to an implementation form.

With respect to Fig. 6, in the first embodiment of Multiple-access with relaying in cellular networks, Tx1 and Tx2 are two base stations, BS1 and BS2, communicating with a common station BS3 (Rx1 ) as in conventional uplink scenarios. In addition, BS2 also plays the role of a relay node to help a user communicate with its base station BS4 (Rx2). In this scenario, the transmission to BS4 causes interference to BS3, and this interference is known only by one of the transmitting nodes, BS2. Also, BS4 may be located relatively far away from the other base stations and so may not overhear the signals that are sent to BS3 (or it may overhear them and simply ignore them).

In this embodiment, an LTE scenario is considered with BSs equipped with 8 antennas each, a bandwidth of 20 MHz, links BS1 -BS3 and BS2-BS4 with average SNR of 0 dB and link BS2- BS3 with an average SNR of -20 dB, and an interference power Q of -3.7 dB, and power P₁ of 0 dB and P₂ of -9 dB. In this scenario an improvement of about 20 Mb/s of sum rate can be achieved over standard THP.

In the second embodiment of precoding against pilot contamination, Tx1 and Tx2 are two base stations BS1 and BS2 communicating with a mobile terminal MT1 (Rx1 ). BS2 is relaying the information of BS1 to MT1 , or is sending separate new information to MT1 . A new mobile terminal MT2 plays the role of Rx2 and is located in the cell of BS2. MT2 has just connected and needs to learn about its channel as well as the state of the surrounding network. BS2 sends a pilot signal that interferes with the transmission to MT1 . In this specific application, the interference caused at MT2 by transmission to MT1 is estimated by MT2 to know about current network conditions.

In the third embodiment of inter-carrier interference, Tx1 and Tx2 are two base stations BS1 and BS2 communicating with some mobile terminal MT1 (Rx1 ) over a certain frequency carrier f_x, in a collaborative or non-collaborative manner. The second base station BS2 also communicates with a second mobile terminal MT2 (Rx2) over another frequency carrier f₂. If the frequency carriers are not well apart, the transmission to MT2 causes some inter-carrier interference to MT1 . Also, this interference is known only asymmetrically, i.e., by only BS2, not BSL

Fig. 7 shows a schematic diagram illustrating a method 700 for cancelling asymmetrically known interference according to an implementation form. The method 700 includes a first block 701 of generating a first input signal Xi carrying a first message cwi intended for transmission to a first reception node Rx1 , e.g. as described above with respect to Figures 3 and 4. The method 700 includes a second block 702 of generating a second input signal X₂ carrying a second message Cw2 intended for transmission to the first reception node Rx1 , e.g. as described above with respect to Figures 3 and 4. The method includes a third block 703 of generating an interference signal X, wherein the interference signal X is an asymmetrically known interference signal that is known for generating the second input signal X₂ and unknown for generating the first input signal Xi , e.g. as described above with respect to Figures 3 and 4. The method 700 includes a fourth block 704 of cancelling part of the interference signal X to generate a partially cancelled interference signal Xc, e.g. as described above with respect to Figures 3 and 4. The method includes a fifth block 705 of precoding 705 the second message W₂ with the non-cancelled part X_nc of the interference signal X by using a lattice-based precoding scheme to generate a precoded signal X₂w, e.g. as described above with respect to Figures 3 and 4. The second block 702 further includes generating the second input signal X₂ based on the partially cancelled interference signal X_c and the precoded signal X_2W, e.g. as described above with respect to Figures 3 and 4.

The method 700 may further include: using a Tomlinson-Harashima precoding scheme (THP) for precoding the second message Cw₂ with the non-cancelled part X_nc of the interference signal X, e.g. as described above with respect to Figures 3 and 4. The present disclosure also supports a computer program product including computer executable code or computer executable instructions that, when executed, causes at least one computer to execute the performing and computing steps described herein, in particular the method 700 described above with respect to Fig. 7. Such a computer program product may include a readable storage medium storing program code thereon for use by a computer. The program code may perform the method 700 as described above with respect to Fig. 7.

While a particular feature or aspect of the disclosure may have been disclosed with respect to only one of several implementations, such feature or aspect may be combined with one or more other features or aspects of the other implementations as may be desired and advantageous for any given or particular application. Furthermore, to the extent that the terms "include", "have", "with", or other variants thereof are used in either the detailed description or the claims, such terms are intended to be inclusive in a manner similar to the term "comprise". Also, the terms "exemplary", "for example" and "e.g." are merely meant as an example, rather than the best or optimal. The terms "coupled" and "connected", along with derivatives may have been used. It should be understood that these terms may have been used to indicate that two elements cooperate or interact with each other regardless whether they are in direct physical or electrical contact, or they are not in direct contact with each other.

Although specific aspects have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that a variety of alternate and/or equivalent

implementations may be substituted for the specific aspects shown and described without departing from the scope of the present disclosure. This application is intended to cover any adaptations or variations of the specific aspects discussed herein.

Although the elements in the following claims are recited in a particular sequence with corresponding labeling, unless the claim recitations otherwise imply a particular sequence for implementing some or all of those elements, those elements are not necessarily intended to be limited to being implemented in that particular sequence.

Many alternatives, modifications, and variations will be apparent to those skilled in the art in light of the above teachings. Of course, those skilled in the art readily recognize that there are numerous applications of the invention beyond those described herein. While the present invention has been described with reference to one or more particular embodiments, those skilled in the art recognize that many changes may be made thereto without departing from the scope of the present invention. It is therefore to be understood that within the scope of the appended claims and their equivalents, the invention may be practiced otherwise than as specifically described herein.

Claims

CLAIMS:

1 . A precoding device (300) for cancelling asymmetrically known interference, the precoding device (300) comprising: a first transmitter node (Tx1 ) configured to generate a first input signal (Xi) carrying a first message (cwi) intended for transmission to a first reception node (Rx1 ); a second transmitter node (Tx2) configured to generate a second input signal (X₂) carrying a second message (Cw2) intended for transmission to the first reception node (Rx1 ) and an interference signal (X) intended for transmission to another reception node (Rx2), wherein the interference signal (X) is an asymmetrically known interference signal that is known to the second transmitter node (Tx2) and unknown to the first transmitter node (Tx1 ), wherein the second transmitter node (Tx2) comprises: an interference canceller (305) configured to cancel part of the interference signal (X) to generate a partially cancelled interference signal (Xc), and a precoder (303), configured to use a lattice-based precoding scheme to precode the second message (W₂) with the non-cancelled part (X_nc) of the interference signal (X) to generate a precoded signal (X₂w), wherein the second transmitter node (Tx2) is configured to generate the second input signal (X₂) based on the partially cancelled interference signal (Xc) and the precoded signal (X_2W).

2. The precoding device (300) of claim 1 , wherein the precoder (303) is configured to use a Tomlinson-Harashima precoding scheme (THP) for precoding the second message (W₂) with the non-cancelled part (X_nc) of the interference signal (X).

3. The precoding device (300) of claim 1 or 2, wherein the interference canceller (305) comprises a multiplier unit (407) configured to multiply the interference signal (X) with a cancellation factor (402) to generate the partially cancelled interference signal (X_c).

4. The precoding device (300) of claim 3, wherein the cancellation factor (402) depends on a power (P₂) of the second input signal (X₂).

5. The precoding device (300) of claim 4, wherein the cancellation factor (402) depends on a negative correlation coefficient ( p ) indicating a correlation between the second input signal (X₂) and the interference signal (X).

6. The precoding device (300) of claim 5, wherein the cancellation factor (402) depends on a power (Q) of the interference signal (X).

7. The precoding device (300) of claim 6, wherein the cancellation factor (402) depends on the relation

where p is the correlation coefficient, P₂ is the power of the second input signal (X₂) and Q is the power of the interference signal (X).

8. The precoding device (300) of claim 7, wherein the preco ) is configured to generate the precoded signal (X₂w) depending on the relation

9. The precoding device (300) of one of the preceding claims, wherein the second transmitter node (Tx2) comprises an adding unit (307, 405) configured to add the partially cancelled interference signal (Xc) and the precoded signal (X₂w) to generate the second input signal (X₂).

10. The precoding device (300) of one of the preceding claims, wherein the precoder (303) is configured to provide the precoded signal (X₂w) statistically independent of the interference signal (X).

1 1 . The precoding device (300) of one of the preceding claims, wherein the second transmitter node (Tx2) comprises an optimizer (403) configured to generate optimal parameters for adjusting the interference canceller (305) and the precoder (303, 401 ) with respect to a maximum sum transmission rate of the first message (cwi) and the second message (Cw2).

12. The precoding device (300) of claim 1 1 , wherein the optimal parameters comprise: a negative correlation coefficient ( p ) indicating a correlation between the second input signal (X₂) and the interference signal (X), and a scaling factor ( « ) for scaling the non-cancelled part (X_nc) of the interference signal (X).

13. The precoding device (300) of claim 1 1 or 1 2, wherein the optimizer (403) is configured to generate the optimal parameters based on the following relation :

(p, a) = argmax R (P_lt P₂, Q, p, a) ,

aER,-l≤p≤0

where P₂, Q, p, a) is the sum rate obtained with given parameters and p is the correlation coefficient, a is a precoding parameter, Pi is the power of the first input signal, P₂ is the power of the second input signal and Q is the power of the interference signal.

14. A method (700) for cancelling asymmetrically known interference, the method comprising: generating (701 ) a first input signal (Xi) carrying a first message (cwi) intended for transmission to a first reception node (Rx1 ) ; generating (702) a second input signal (X₂) carrying a second message (cw₂) intended for transmission to the first reception node (Rx1 ); generating (703) an interference signal (X), wherein the interference signal (X) is an asymmetrically known interference signal that is known for generating the second input signal (X₂) and unknown for generating the first input signal (Xi ); cancelling (704) part of the interference signal (X) to generate a partially cancelled interference signal (Xc), and precoding (705) the second message (W₂) with the non-cancelled part (X_nc) of the interference signal (X) by using a lattice-based precoding scheme to generate a precoded signal (X₂w), wherein generating (702) the second input signal (X₂) is based on the partially cancelled interference signal (Xc) and the precoded signal (X₂w).

15. The method of claim 14, comprising: using a Tomlinson-Harashima precoding scheme (THP) for precoding the second message (cw₂) with the non-cancelled part (X_nc) of the interference signal (X).