WO2022185050A1

WO2022185050A1 - Methods and apparatus for transmitting and receiving data

Info

Publication number: WO2022185050A1
Application number: PCT/GB2022/050546
Authority: WO
Inventors: Christos MASOUROS; Abdelhamid SALEM
Original assignee: Ucl Business Ltd
Priority date: 2021-03-02
Filing date: 2022-03-02
Publication date: 2022-09-09
Also published as: GB202102913D0

Abstract

Aspect of the present disclosure provide a data transmission method. The method comprises receiving first data to be transmitted on a first NOMA, channel to a first receiver, and second data to be transmitted on a second NOMA channel to a second receiver. First channel strength information corresponds to the first NOMA channel and second channel strength information corresponds to the second NOMA channel. The first NOMA channel is stronger than the second NOMA channel. A first precoding, to be applied to the first data, and a second precoding, to be applied to the second data, is determined such that transmission of the second data is expected to constructively interfere with transmission of the first data. The first and second precodings are applied, and the data is transmitted on the first and second NOMA channels.

Description

METHODS AND APPARATUS FOR TRANSMITTING AND

RECEIVING DATA

BACKGROUND

The present technique relates to the field of data transmission, and more specifically to techniques for non-orthogonal multiple access (NOMA). NOMA is a family of techniques for encoding digital data onto carrier signals. Unlike orthogonal multiple access, NOMA techniques can use non-orthogonal carriers or subcarriers. This has various advantages over orthogonal multiple access techniques, for example the ability to multiplex data streams in the same frequency-time-space resources and thereby improving spectral efficiency. However, the lack of orthogonality in a NOMA system means that communication channels or subchannels can interfere with each other, which increases the complexity of a receiver which is to extract its data from a plurality of interfering channels.

There is thus a desire for improved NOMA techniques which improve the efficiency and effectiveness of decoding.

SUMMARY

At least some examples provide a data transmission method comprising: receiving: first data to be transmitted on a first non-orthogonal multiple access, NOMA, channel to a first receiver; and second data to be transmitted on a second non-orthogonal multiple access, NOMA, channel to a second receiver; and first channel strength information corresponding to the first NOMA channel and second channel strength information corresponding to the second NOMA channel, wherein the first channel strength indicates that the first NOMA channel is stronger than the second NOMA channel; based on at least the first channel strength information, determining a first precoding, to be applied to the first data, and a second precoding, to be applied to the second data, such that transmission of the second data is expected to constructively interfere with transmission of the first data to the first receiver; applying the first precoding to the first data to produce first transmission data and applying the second precoding to the second data to produce second transmission data; and transmitting the first transmission data on the first NOMA channel and the second transmission data on the second NOMA channel. Further examples provide a data transmission apparatus comprising: interface circuitry configured to receive: first data to be transmitted on a first non-orthogonal multiple access, NOMA, channel to a first receiver; second data to be transmitted on a second non-orthogonal multiple access, NOMA, channel to a second receiver; and first channel strength information corresponding to the first NOMA channel and second channel strength information corresponding to the second NOMA channel, wherein the first channel strength indicates that the first NOMA channel is stronger than the second NOMA channel; precoding circuitry configured to: based on at least the first channel strength information, determine a first precoding, to be applied to the first data, and a second precoding, to be applied to the second data, such that transmission of the second data is expected to constructively interfere with transmission of the first data to the first receiver; and apply the first precoding to the first data to produce first transmission data and applying the second precoding to the second data to produce second transmission data, and transmission circuitry configured to transmit the first transmission data on the first NOMA channel and the second transmission data on the second NOMA channel.

Further examples provide an apparatus comprising: receiver circuitry configured to receive, from a transmission device as set out above, the first transmission data and the second transmission data; and decoding circuitry configured to extract one of: the first transmission data; and the second transmission data.

Further examples provide a system comprising a transmission device and at least one receiver as set out above.

Further aspects, features and advantages of the present technique will be apparent from the following description of examples, which is to be read in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 schematically depicts a system according to an example of the present disclosure.

Figure 2 schematically depicts and transmitter device and a receiver device, according to an example. Figures 3A and 3B depict operational flows of a comparative example, and the present disclosure, respectively.

Figure 4 depicts a method for determining precodings using the present disclosure, according to an example.

Figure 5 depicts a method using the present disclosure, according to an example.

DESCRIPTION OF EXAMPLES

As set out above, some examples relate to a data transmission method. The method may for example be performed by a transmission device, such as a wireless transmission device, for example circuitry within a base station for mobile telecommunication, a wireless router, or other device configured to transmit data to receiver devices. First data is to be transmitted on a first non-orthogonal (NOMA) channel to a first receiver, and second data is to be transmitted on a second NOMA channel to a second receiver. This first and second data is received at the transmission device, for example from a processing component of the device, or from another device communicatively coupled to the device. The first and second data may for example comprise modulated NOMA data symbols.

The method further comprises receiving first channel strength information corresponding to the first NOMA channel and second channel strength information corresponding to the second NOMA channel. In this example, the first channel strength indicates that the first NOMA channel is stronger than the second NOMA channel. For example, the first receiver may be closer to the transmission device than the second receiver, and/or may have a geographical location which is more conducive to NOMA transmissions.

Based on at least the first channel strength information, a first precoding is determined, to be applied to the first data. Similarly, a second precoding is determined, to be applied to the second data. The first and second precoding are determined such that transmission of the second data is expected to constructively interfere with transmission of the first data to the first receiver. Examples of how such constructive interference can be implemented will be described in more detail below.

The first precoding is applied to the first data, to produce first transmission data, and the second precoding is applied to the second data, to produce second transmission data. In order to produce the first and second transmission data, the precoded first and second data may be modulated onto NOMA carrier symbols.

Finally, the first transmission data is transmitted on the first NOMA channel and the second transmission data is transmitted on the second NOMA channel.

This method allows more efficient and effective decoding at the receivers. In some comparative example NOMA systems, the first receiver (with a stronger channel) is required to initially detect and determine the second channel (i.e. the second receiver’s signal) and subtract this from the received superposition of the first and second channels, and then perform channel equalisation. This is to correct for the non-orthogonal nature of the first and second channels, which leads to interference between the channel. Only then can the first receiver, in this comparative system, extract the first channel and thereby decode the first data. In this comparative example, the second receiver does not need to subtract the first channel, because the second channel is assigned a higher transmission power (to compensate for the weaker channel).

In the present example, in contrast, the precoding using the present disclosure is designed such that the second channel interferes constructively with the first channel, at the first receiver. The first receiver thus effectively receives a boosted version of the first channel, without a destructive superposition with the second channel. The first receiver therefore does not need to subtract the second channel, and thus does not need to detect or determine the second channel. Instead, the first receiver can directly detect the first channel from the received signal. This significantly reduces the complexity of the decoding process and thus can reduce processing and power resource usage in the receivers, as well as reducing latency. There is also a significant improvement in effective transmission power (as received at the first receiver) and error rate performance, as a consequence of the “boosting” that the first channel receives.

This method can be extended to accommodate additional receivers by treating each receiver, other than the receiver with the weakest channel, in the same manner as the above- described first receiver (i.e. determining precodings to cause constructive interference at the receivers).

The first and second precoding can be applied in various ways. In one example, the first precoding is a first precoding weight vector and the second precoding is a second precoding weight vector. Applying each precoding then comprises weighting the data with the corresponding precoding weight vector. This provides an effective way of applying the first and second precoding.

The first and second precoding (e.g. the values of the aforementioned weight vectors) can also be determined in various ways. In some examples, a plurality of potential first and second precodings is determined (for example randomly, or according to an iterative algorithm). An expected level of constructive interference is then determined, for example by calculating an expected combined signal that will be received at the first receiver. The pair of potential first and second precodings having the highest expected level of constructive interference (or, alternatively, a pair having an expected level of constructive interference above a threshold) is then selected as the first and second precoding. This provides an effective way to identify a combination of first and second precoding which will produce constructive interference. In one such example, the above-described process for determining the first and second precoding (i.e. the determining of the potential precodings, and the determining of the associated expected levels of constructive interference) comprises performing a numerical optimisation of the expected level of constructive interference. For example, the plurality of potential precodings may be iterated through, until a local maximum of constructive interference is found. This is an effective way of identifying a suitable first and second precoding, within a reasonable amount of time to allow practical use of the technique for example on real-time communication channels. Alternatively, the present disclosure could operate on a block of data symbols to be transmitted, to reduce the latency of determining the precodings. As another alternative, a look-up-table can be formulated based on the channel information, and for all possible data symbol combinations, to allow an effective selection of precodings.

As mentioned above, the expected interference for each pair of potential first and second precodings can be determined by calculating an expected combined signal at the first receiver. In one example, this comprises applying the potential first precoding to the first data to produce potential precoded first data, and similarly applying the potential second precoding to the second data to produce potential precoded second data. An expected received signal at the first receiver is then determined, based on the potential precoded first and second data and the first channel strength information. For example, a potential combined signal may be produced by adding the potential precoded first and second data, and then the first channel strength information may be used to estimate the potential combined signal as received at the first receiver. The expected level of constructive interference is then determined based on an extent to which the potential second precoded data constructively interferes with the potential first precoded data in the expected signal at the first receiver. This thereby provides an effective way to determine the expected interference for each pair of potential first and second precodings.

In an example, which may be implemented in addition to the examples described above, the second precoding is determined as a precoding that gives the second transmission a transmission power above a power threshold. For example, this may be a constraint applied to the above-described examples, in order to ensure that the second transmission has sufficient power to be successfully received and decoded by the second receiver.

The threshold may be pre-defined and independent of the strength of the first and second channels. Alternatively, the threshold may be adaptively determined based on at least one of the first and second channel strength information. For example, the threshold may be based on the second channel strength information, thereby allowing the system to relax the constraint if the second channel strength increases (which may for example open up more potential precodings), and to tighten the constraint if the second channel strength decreases (to ensure, or at least improve the chance of, successful transmission).

For example, determining the second precoding may comprise determining a plurality of potential second precodings and determining, for each potential second precoding of said plurality, a corresponding transmission power of the second transmission data. A potential second precoding with transmission power above the power threshold, or the potential second precoding with highest transmission power, may then be selected as the second precoding.

Alternatively or additionally, the second precoding may be determined based on the second channel strength data such that the second transmission data has an expected signal quality, at the second receiver, above a threshold. All of these threshold examples allow assurance that the second transmitted data will have sufficient received power to be received and decoded by the second receiver.

In one such example, determining the second precoding comprises determining a plurality of potential second precodings and determining, for each potential second precoding or said plurality, an expected signal quality (for example as described above). A potential second precoding with an expected signal quality above a signal quality threshold is then selected as the second precoding. In an example, the potential second precoding having the highest expected signal quality may be selected. Acceptable signal quality at the second receiver is thus assured.

In all of these methods for determining the second precoding, the determination may be performed simultaneously with the determination of the first precoding such that signal power/received signal quality at the second receiver is achieved simultaneously with the above-described constructive interference at the first receiver. This may for example be performed as a simultaneous optimisation of the first precoding and the second precoding. Acceptable and decodable signals are thus received at each of the first and second receivers.

As mentioned above, aspects of the present disclosure relate to a system comprising a transmitter operating according to one of the above examples, and a first and second receiver. Each of the first and second receiver is configured to receive a combination of the first and second transmission data, and to extract its corresponding one of the first and second transmission data. The extracted transmission data can then be decoded to retrieve the underlying first or second data.

In an example, the first receiver apparatus is configured to extract the first transmission data without performing a subtraction of an expected second transmission data. As described above, the presently described techniques render such subtraction unnecessary, and thus power and processing resource consumption can be reduced by omitting the subtraction. Examples of the present disclosure will now be described with reference to the drawings.

Figure 1 schematically shows a system according to an example of the present disclosure. The system comprises a transmitter 105 and two receivers 110a, 110b. The transmitter 105 transmits data to the first receiver 110a on a first communication channel, and transmits data to the second receiver 110b on a second communication channel. The first receiver 110a is relatively close to the transmitter 105, and the second receiver 110b is relatively far from the transmitter.

A spectrum graph 115 shows how power and frequency are allocated to the channels. Specifically, a higher power is allocated to the second channel, because of the greater distance of the second receiver 110b from the transmitter 105. Both channels share the same frequency range, and are transmitted simultaneously. The channels can be separated because the transmitted data is modulated onto NOMA symbols, such that a receiver can extract its associated channel from the transmitted sum of multiple channels.

The first receiver 110a extracts, from the received combined signal, the data transmitted on the first channel. As a consequence of the higher transmission power of the second channel, the first receiver 110a receives the second channel more strongly than the first channel. In comparative examples of NOMA systems which do not implement examples of the present disclosure, because of the non-orthogonal nature of the transmissions, the second channel can interfere with the first signal and such interference may be destructive. Thus, in such comparative examples, the first receiver 110a must estimate the second channel and subtract it from the combined signal, in order to extract the first channel. This estimation and subtraction is taxing on processing and power resources, and the estimation is also vulnerable to error (leading to inaccurate extraction of the first channel). As described in more detail below, the present disclosure can eliminate the aforementioned destructive interference and cause the second channel to interfere constructively with the first channel at the first receiver 110a, thereby boosting the first channel and eliminating the need to estimate and subtract the second channel: instead, the first receiver 110a can directly extract the first channel.

The second receiver 110b directly extracts, from the received signal, the data transmitted on the second channel. Because of the higher transmission power of the second channel, the second channel dominates the received signal at the second receiver 110b and thus the above-described consideration are less relevant for the second receiver 110b.

The system may be expanded to include additional receivers. The methods described herein can be extended to support such multiple receivers by configuring each receiver other than the furthest (or the receiver with the weakest received signal) to behave analogously with the above-described first receiver 110a, and the furthest receiver to behave analogously with the above-described second receiver 110b.

Figure 2 schematically depicts a transmitter device 205 and a receiver device 210, according to an example of the present disclosure. The transmitter device 205 may form part of the larger apparatus, such as a base station or wireless router. Similarly, the receiver device 210 may form part of a larger apparatus, such as a mobile telephone or other user equipment. The described circuitry elements may be implemented in dedicated hardware or, alternatively, some or all of the described functionality may be performed by general-purpose circuitry such as a processor.

The transmitter device 205 has an interface 215. The interface 215 is configured to receive data to be transmitted to a plurality of receivers, including the receiver 210, with the data for each receiver being transmitted on a different channel. The data is received in the form of data modulated onto NOMA symbols. The interface 215 also receives channel strength information, such as a channel quality indicator, for each receiver. The data and channel strength information may be received from another component of the device, such as a processor.

The transmitter device 205 has a precoder 220. The precoder 220 determines a set of precodings, one for each receiver, such that channels directed to distant receivers are expected to constructively interfere at a given receiver with the channel directed to that receiver. The precoder 220 applies each precoding to its associated data.

The transmitter device 205 has a transmitter 225, which may for example comprise an antenna or other transmission hardware. The transmitter device receives the precoded data and transmits it, simultaneously, to the receivers.

The receiver device 210 has a receiver 230, which may for example comprise an antenna or other receiving hardware. The receiver 230 receives the combined signal from the transmitter device 205.

The receiver device 210 comprises a decoder 235. The decoder decodes the received combined signal in order to extract the channel that is intended for the receiver device 210, and to decode the channel to recover the transmitted data. Because of the aforementioned constructive interference, the received signal can be directly decoded and the channel extracted, without estimating or subtracting the channels directed at other receivers.

Figures 3A and 3B depict operational flows. Figure 3A represents a comparative example which does not implement the present disclosure, and Figure 3B represents an example of the present disclosure. The top row of each figure represents operations in a transmitter, and the bottom row represents operations in a receiver.

In the comparative example of Figure 3A, data, to be transmitted to a receiver, is modulated with NOMA symbols. The modulated data is then precoded, based on channel estimation information which represents the transmission conditions, such as channel quality, of each receiver. For example, the precoding may be such as to assign a higher transmission power to a receiver with a worse channel quality.

The precoded, modulated data for all channels is then sent to one or more antennas, for transmission to the receiver.

At the receiver, a combined signal comprising multiple transmitted channels is received at one or more antennas. In this example, the receiver has a stronger channel than one or more weaker channels directed to other receivers (e.g. because it is closer to the transmitter). The receiver estimates the signal corresponding to the weaker channel or channels and, based on this estimation, performs a successive interference cancelling (SIC) process to subtract the signals of the weaker channel or channels (which have a higher received power at the present receiver as a consequence of their higher transmission power, as described above). The receiver then performs channel equalization, after which it can demodulate its own channel and thereby extract the transmitted data.

In the example of Figure 3B, data to be transmitted to the receiver is modulated with NOMA symbols, similarly to the comparative example of Figure 3A. However, in contrast to Figure 3A, the modulated data is precoded (based on channel estimation information) such that the channel or channels directed to a more distant/weaker receiver are expected to constructively interfere, at a nearer/stronger receiver, with the channel directed to that nearer/stronger receiver. Examples of such precoding are described elsewhere in this disclosure.

At the receiver, a combined signal comprising multiple transmitted channels is received at one or more antennas. As a consequence of the aforementioned constructive interference, this is effectively received as a boosted version of the channel directed to the first receiver, without destructive interference from other channels. Thus, no estimation of other channels, nor any subtraction of such channels, is performed. Instead, the receiver directly demodulates the received signal, to extract the transmitted data.

It can therefore be seen that the example of Figure 3B is significantly more computationally efficient than the comparative example of Figure 3A, as no channel estimation or subtraction is performed. Furthermore, the “boosting” effect improves the channel quality as received at the receiver.

Figure 4 depicts an example method for determining precodings which are expected to constructively interfere. At block 405, data to be transmitted to receivers (which may already be modulated onto NOMA symbols), and channel quality information associated with each receiver, is received.

At block 410, a set of precodings (one for each channel) is generated, for example as described in more detail below.

At block 415, an iterative method is performed to generate an improved set of precodings. An example of such an iterative method is described in more detail below. Each iteration of the iterative method may for example produce a set of precodings with improved expected constructive interference properties.

At block 420, a decision is made as to whether to terminate the iterative process. This may be based on whether one or more constraints are satisfied, for example that the channels will exhibit sufficiently favourable constructive interference policies, and optionally that each precoded channel will have an acceptable transmission power. Alternatively, a predefined number of iterations may be performed prior to termination, on the assumption that the predefined number is sufficient to generate an acceptable set of precodings. If the decision is not to terminate, flow returns to block 415 and another iteration is performed. If the decision is to terminate, flow proceeds to block 420 where the iteratively generated precodings are output (and subsequently used to precode each channel as described above).

An example of determining and applying precoding will now be described. In this example, there is a first user receiver (near) and a second user receiver (far), and the aim of the precoding is to align the superimposed symbols that can be known to the transmitter to increase the useful signal power. That is, with the knowledge of all channel and users' data symbols at the serving transmitter, the interference to a given channel (from other channels) can be classified as constructive if it pushes the received symbols away from the decision thresholds towards the direction of the desired symbol. Accordingly, the transmission precoding is designed to impose constructive interference to the desired symbol.

Precodings with such desirable properties are generated by an optimisation process, subject to interference constraints, by optimising instantaneous interference at the first receiver, to contribute to the received signal power. The received signal-to-noise ratio at the first and second user can be expressed as:

where:

|x| denotes the absolute value of the number x; s_t = b^fί is the data symbol of user I, with f; being the angle of the complex number s,; h_su, is the channel vector from the transmitter to user /; w, is the precoding weight vector of user /; and s„ expresses the power spectral density of estimated additive white Gaussian noise at user /

Based on the above SNR expressions and the concept of constructive signal superposition, the precodings can be determined by solving the below convex optimization:

Where:

||x|| is the norm of argument x;

Re(x) denotes the real part of complex number x; lm(x) denotes the imaginary part of complex number x; n is a target SNR level for user /; tan(x) denotes the tangent of x;

0 __ and ^"

where M is the order of the digital modulation used to modulate the information symbols

This optimization can be used to iteratively determine a suitable set of precoding weight vectors, by way of the following algorithm:

1. Set the maximum number of iterations T, set t=0, and randomly generate w,

2. Repeat

3. Calculate the solutions to the above optimization as w*

4. Update w, = w*

5. t = t+1

6. Repeat until t = T

7. Output w* for all /.

This provides an effective, and mathematically efficient, way of generating a set of precoding vectors that are expected to cause desirable constructing interference. It should be noted that this is one example, and other methods may be used to generate precoding vectors according to the present disclosure. Figure 5 depicts a method according to an example of the present disclosure. The method is performed in a transmitter, such as one of the transmitters described above.

At block 505, first and second data are received. The first and second data are to be transmitted to respective first and second receivers. The first receiver has more favourable communication conditions with the transmitter, for example because it is nearer to the transmitter.

At block 510, first and second channel strength information are received, which describe the strength of communication channels between the transmitter and each receiver.

At block 515, a first precoding is determined (to be applied to the first data) and a second precoding is determined (to be applied to the second data). The precodings are determined such that the transmission of the second data is expected to interfere constructively with the transmission of the first data, at the first receiver.

At block 520, the first precoding is applied to the first data and the second precoding is applied to the second data, to produce first and second transmission data.

At block 525, the precoded first and second data are transmitted, simultaneously, to the first and second receivers.

Apparatuses and methods are thus provided for implementing an improved NOMA system, in which the higher-power signal directed to a user with a weaker channel constructively interferes with the lower-power signal directed to a user with a stronger channel. From the above description it will be seen that the techniques described herein provides a number of significant benefits. In particular, receiver complexity can be reduced, leading to reduced consumption of processing and power resources. Received signal quality is also improved.

In the present application, the words “configured to...” are used to mean that an element of an apparatus has a configuration able to carry out the defined operation. In this context, a “configuration” means an arrangement or manner of interconnection of hardware or software. For example, the apparatus may have dedicated hardware which provides the defined operation, or a processor or other processing device may be programmed to perform the function. “Configured to” does not imply that the apparatus element needs to be changed in any way in order to provide the defined operation.

Although illustrative embodiments of the invention have been described in detail herein with reference to the accompanying drawings, it is to be understood that the invention is not limited to those precise embodiments, and that various changes and modifications can be effected therein by one skilled in the art without departing from the scope of the invention as defined by the appended claims.

Claims

WE CLAIM:

1. A data transmission method comprising: receiving: first data to be transmitted on a first non-orthogonal multiple access, NOMA, channel to a first receiver; and second data to be transmitted on a second non-orthogonal multiple access, NOMA, channel to a second receiver; and first channel strength information corresponding to the first NOMA channel and second channel strength information corresponding to the second NOMA channel, wherein the first channel strength indicates that the first NOMA channel is stronger than the second NOMA channel; based on at least the first channel strength information, determining a first precoding, to be applied to the first data, and a second precoding, to be applied to the second data, such that transmission of the second data is expected to constructively interfere with transmission of the first data to the first receiver; applying the first precoding to the first data to produce first transmission data and applying the second precoding to the second data to produce second transmission data; and transmitting the first transmission data on the first NOMA channel and the second transmission data on the second NOMA channel.

2. A method according to claim 1 , wherein the first and second data comprise modulated NOMA data symbols.

3. A method according to claim 1 or claim 2, wherein: the first precoding is a first precoding weight vector; the second precoding is a second precoding weight vector; applying the first precoding to the first data comprises weighting the first data with the first precoding weight vector; and applying the second precoding to the second data comprises weighting the second data with the second precoding weight vector.

4. A method according to any preceding claim, wherein determining the first precoding and the second precoding comprises: determining a plurality of potential first and second precodings; determining, for each of said plurality, an expected level of constructive interference; and selecting, as the first and second precoding, the potential first and second precoding having a highest expected level of constructive interference.

5. A method according to claim 4, wherein the determining the plurality of potential first and second precodings, and the determining the expected levels of constructive interference, comprises performing a numerical optimisation of the expected level of constructive interference.

6. A method according to claim 4 or claim 5, wherein determining the expected level of constructive interference for each potential first and second precoding comprises: applying the potential first precoding to the first data to produce potential precoded first data; applying the potential second precoding to the second data to produce potential precoded second data; based on the first channel strength information, and the potential precoded first and second data, determining an expected received signal at the first receiver; and determining the expected level of constructive interference based on an extent to which the potential second precoded data constructively interferes with the potential first precoded data in the expected received signal at the first receiver.

7. A method according to any preceding claim comprising determining the second precoding as a precoding giving the second transmission data a transmission power above a power threshold.

8. A method according to any preceding claim, comprising determining at least one of the first precoding and second precoding based additionally on the second channel strength information.

9. A method according to claim 8, comprising determining the second precoding based on the second channel strength data such that the second transmission data has an expected signal quality, at the second receiver, above a threshold.

10. A method according to claim 8 or claim 9, comprising determining the second precoding such that the second transmission data has a transmission power above a power threshold, said power threshold being based on the second channel strength information.

11. A method according to claim 9, wherein determining the second precoding comprises: determining a plurality of potential second precodings; determining, for each potential second precoding of said plurality, an expected signal quality; and selecting, as the second precoding, the potential second precoding having an expected signal quality above the threshold, and optionally selecting the potential second precoding having the highest expected signal quality.

12. A method according to claim 7 or claim 10, wherein determining the second precoding comprises: determining a plurality of potential second precodings; determining, for each potential second precoding of said plurality, a corresponding transmission power of the second transmission data; and selecting, as the second precoding, the potential second precoding having corresponding transmission power above the threshold, and optionally selecting the potential second precoding having the highest corresponding transmission power.

13. A method according to any preceding claim, comprising simultaneously determining the first precoding and the second precoding.

14. A method according to claim 13, wherein the simultaneously determining the first precoding and the second precoding comprises performing a simultaneous optimisation of the first precoding and the second precoding.

15. A data transmission apparatus comprising: interface circuitry configured to receive: first data to be transmitted on a first non-orthogonal multiple access, NOMA, channel to a first receiver; second data to be transmitted on a second non-orthogonal multiple access, NOMA, channel to a second receiver; and first channel strength information corresponding to the first NOMA channel and second channel strength information corresponding to the second NOMA channel, wherein the first channel strength indicates that the first NOMA channel is stronger than the second NOMA channel; precoding circuitry configured to: