Transmitter and method for probabilistic shaping

Abstract

The invention relates to a transmitter and method for efficient signal shaping in communication systems. The transmitter incorporates a distribution matcher that maps a message to a list of first sequences based on a target distribution criterion. The transmitter converts these first sequences into second sequences. A selector evaluates the second sequences using a second criterion such as peak-to-average power ratio or cubic metric to select a sequence for transmission. In this way, both average transmit power and peak-to-average power ratio (or similar metrics) can be reduced.

Description

[0001] TRANSMITTER AND METHOD FOR PROBABILISTIC SHAPING

[0002] TECHNICAL FIELD

[0003] The present disclosure relates to the field of telecommunication. For instance, the disclosure relates to a transmitter and a device for probabilistic shaping.

[0004] BACKGROUND

[0005] Energy efficiency has become a critical key performance indicator (KPI) in the development of next-generation wireless communication systems, such as in 6G networks and so on. This is especially significant for battery-powered user equipment (UEs). Future data-intensive applications such as real-time video streaming from mobile devices, including vehicles, drones, and robots, as well as data aggregation for machine learning, are expected to increase the demand for uplink transmissions.

[0006] To achieve higher spectral efficiency, high-order modulation schemes are increasingly necessary. This is particularly relevant in low-frequency bands with limited spectrum availability, which are essential for wide-area coverage. High spectral efficiency is partly facilitated by dense network deployments and enhanced beamforming gains enabled by large antenna arrays. However, achieving these benefits often necessitates relatively high transmit power, which poses a challenge to energy efficiency.

[0007] A significant portion of a UE's total power consumption, and consequently its battery life, is consumed by the power amplifier during high transmit power operations. Therefore, reducing the average transmit power while maintaining a target spectral efficiency is of paramount importance.

[0008] SUMMARY

[0009] The efficiency of a power amplifier depends heavily on its operating point. Maximum efficiency is achieved at the amplifier's peak output power. However, for high-order modulation schemes, the amplifier shall operate within a linear regime to prevent nonlinear distortions. A large crest factor or peak-to-average power ratio (PAPR) could significantly decrease the amplifier efficiency. This issue is particularly prominent in low-cost power amplifiers commonly used in UEs, which typically lack advanced features like envelope tracking.

[0010] The present invention addresses these challenges by providing a solution to simultaneously reduce the average transmit power and PAPR (or similar measures), thereby enhancing energy efficiency without compromising performance.

[0011] These and other objectives are achieved by the subject matter of the independent claims. Further implementation forms are apparent from the dependent claims, the description, and the drawings.

[0012] A first aspect of the present disclosure provides a transmitter comprising a distribution matcher and a selector. The distribution matcher is configured to transform a message into a list of first sequences based on a first criterion. The transmitter is configured to transform the list of first sequences into a list of second sequences. The selector is configured to select a to-be-transmitted sequence from the list of second sequences based on a second criterion. The transmitter is configured to transmit the selected sequence.

[0013] Notably, the first criterion and the second criterion may be determined based on various aspects in optimizing signal quality, efficiency, or transmission requirements. For instance, the first criterion may be based on a signal property (e.g., closeness toa target statistical distribution) or energy efficiency (minimizing average transmit power). The second criterion may refer to a measure of how well each candidate second sequence meets one or more optimization objectives, such as improving signal quality or reducing peak levels. In general, the first criterion is used to ensure the generated first sequences meet a target statistical or performance property. The second criterion is used to evaluate the multiple second sequences to identify the one that satisfies the optimization goal for the final transmitted signal.

[0014] In this way, a simultaneous optimization of signal quality and energy efficiency can be achieved in the transmitter.

[0015] In an implementation form of the first aspect, the first criterion comprises an average transmit power, or a Kullback-Leibler distance to a target distribution. The target distribution may be a Gaussian distribution.

[0016] In this way, it can be ensured that the signal is energy-efficient while maintaining spectral efficiency. This contributes to extending battery life in user equipment (UE) and reducing power consumption in communication systems, making the transmitter particularly beneficial for power-constrained devices.

[0017] In a further implementation form of the first aspect, the second criterion comprises a peak-to-average power ratio, or a cubic metric.

[0018] That is, the selector may be configured to select a sequence with the lowest PAPR (or cubic metric) from the list of second sequences for transmission.

[0019] In a further implementation form of the first aspect, each second sequence represents a superposition of symbols of a respective first sequence.

[0020] The second sequences generated through superposition are designed to be compatible with power amplifier constraints, ensuring linearity and reducing distortions.

[0021] In a further implementation form of the first aspect, the distribution matcher comprises one or more polar list decoders configured to generate a list of candidate codewords based on the message.

[0022] The polar list decoder leverages the inherent flexibility of probabilistic shaping to approximate a desired target distribution (e.g., Gaussian) by controlling the likelihood of specific codewords being selected. The probabilistic shaping introduced by the polar list decoder minimizes the average transmit power while maintaining reliable communication, directly benefiting energy-constrained devices.

[0023] In a further implementation form of the first aspect, the distribution matcher comprises one or more symbol mappers configured to map the list of candidate codewords into the list of first sequences according to a preset modulation scheme.

[0024] Optionally, the one or more symbol mappers may comprise one or more of: an amplitude shift keying (ASK) mapper, or binary phase shift keying (BPSK) mapper. The ASK mapper is configured to generate amplitude-modulated symbols for the first sequences. The BPSK mapper is configured to generate sign-bit-modulated symbols for the first sequences.

[0025] The ASK mapper can map candidate codewords into ASK symbols, encoding amplitude information into the first sequences. This approach is particularly suited for probabilistic amplitude shaping, where the amplitudes are adjusted to approximate a target distribution (e.g., Gaussian).The BPSK mapper can map candidate codewords into BPSK symbols, encoding sign-bit information into the first sequences. BPSK mapping is useful for systems employing sign-bit shaping, ensuring efficient encoding of polar-decoded sign bits.

[0026] The symbol mapper may comprise a combination of ASK and BPSK mapping. The amplitude components (e.g., ASK symbols) and sign-bit components (e.g., BPSK symbols) are mapped independently and combined to form the first sequences. This approach enables compatibility with high-order modulation schemes, such as quadrature amplitude modulation (QAM), where real and imaginary parts of the symbol carry distinct information.

[0027] The preset modulation scheme may be configurable based on system requirements or transmission conditions. For low-complexity scenarios, simple modulation schemes like ASK or BPSK may be used. For high-throughput applications, more advanced schemes like multi-level amplitude and phase shift keying (APSK) or QAM can be employed, providing greater flexibility in adapting to different system goals.

[0028] In a further implementation form of the first aspect, the list of first sequences comprises sign bits and / or amplitude bits.

[0029] Amplitude bits are obtained by amplitude shaping, which is to approximate a target statistical distribution (e.g., Gaussian), minimizing average transmit power and improving spectral efficiency. Sign bits are obtained by sign-bit shaping (e.g., through polar decoding or other suitable techniques).

[0030] The sign and amplitude bits are mapped to modulation symbols through processes such as:

[0031] combining amplitude bits in the real and imaginary parts of a QAM symbol, and embedding sign bits as polarity markers for these symbols.

[0032] By including both sign and amplitude bits in the first sequences, the distribution matcher ensures that the statistical shaping properties of the generated sequences are preserved throughout the signal generation process.

[0033] In a further implementation form of the first aspect, for transforming the message into the list of first sequences, the distribution matcher is configured to combine multiple lists of subsequences generated based on the message.

[0034] That is, the input message is split into multiple smaller sub-messages, each processed independently to generate lists of subsequences. This allows for parallel processing, reducing computational complexity and delay (compared to processing the entire message in a single step).

[0035] The multiple lists of subsequences are combined to form the list of first sequences, ensuring that the resulting sequences inherit the shaping properties applied to each subsequence. The combination process may involve concatenation of subsequences to form longer sequences.

[0036] In a further implementation form of the first aspect, for transforming the list of first sequences into the list of second sequences, the transmitter is configured to apply inverse Fourier transformation on the list of first sequences, so as to obtain the list of second sequences.

[0037] In a further implementation form of the first aspect, the transmitter is configured to apply spatial precoding on the list of first sequences, so as to obtain the list of second sequences.In a further implementation form of the first aspect, the transmitter is configured to receive (e.g., from the network side) one or more parameters referencing to one or more modulation and coding scheme tables that specify a mapping of shaping bits and / or at least one parameter related to a symbol distribution.

[0038] A second aspect of the present disclosure provides a method applied to a transmitter comprising a distribution matcher and a selector. The method comprises:

[0039] transforming, by the distribution matcher, a message into a list of first sequences based on a first criterion; transforming, by the transmitter, the list of first sequences into a list of second sequences;

[0040] selecting, by the selector, a to-be-transmitted sequence from the list of second sequences based on a second criterion; and

[0041] transmitting, by the transmitter, the selected sequence.

[0042] In an implementation form of the second aspect, the first criterion comprises an average transmit power, or a Kullback-Leibler distance to a target distribution. The target distribution may be a Gaussian distribution.

[0043] In a further implementation form of the second aspect, the second criterion comprises a peak-to-average power ratio, or a cubic metric.

[0044] In a further implementation form of the second aspect, each second sequence represents a superposition of symbols of a respective first sequence.

[0045] In a further implementation form of the second aspect, the distribution matcher comprises one or more polar list decoders and the method comprises generating, by the one or more polar list decoders, a list of candidate codewords based on the message.

[0046] In a further implementation form of the first aspect, the distribution matcher comprises one or more symbol mappers and the method comprises mapping, by the one or more symbol mappers, the list of candidate codewords into the list of first sequences according to a preset modulation scheme.

[0047] In a further implementation form of the second aspect, the list of first sequences comprises sign bits and / or amplitude bits.

[0048] In a further implementation form of the second aspect, for transforming the message into the list of first sequences, the method comprises combining multiple lists of subsequences generated based on the message.

[0049] In a further implementation form of the second aspect, for transforming the list of first sequences into the list of second sequences, the method comprises applying inverse Fourier transformation on the list of first sequences, so as to obtain the list of second sequences.In a further implementation form of the second aspect, the method comprises applying spatial precoding on the list of first sequences, so as to obtain the list of second sequences.

[0050] In a further implementation form of the second aspect, the method comprises receive (e.g., from the network side) one or more parameters referencing to one or more modulation and coding scheme tables that specify a mapping of shaping bits and / or at least one parameter related to a symbol distribution.

[0051] The method of the second aspect may share the same optional features and advantages as the transmitter of the first aspect.

[0052] A third aspect of the present disclosure provides a system comprising a transmitter according to the first aspect or any implementation form thereof. The system further comprises one or more receivers configured to receive a sequence from the transmitter and decode a message from the received sequence.

[0053] A fourth aspect of the present disclosure provides a computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method according to the second aspect or any implementation form thereof.

[0054] A fifth aspect of the present disclosure provides a computer-readable medium comprising instructions which, when executed by a computer, cause the computer to carry out the method according to the second aspect or any implementation form thereof.

[0055] A sixth aspect of the present disclosure provides a chipset comprising instructions which, when executed by the chipset, cause the chipset to carry out the method according to the second aspect or any implementation form thereof.

[0056] It has to be noted that all devices, terminals, elements, units, and means described in the present disclosure could be implemented in software or hardware elements or any kind of combination thereof. All steps which are performed by the various entities described in the present application as well as the functionalities described to be performed by the various entities are intended to mean that the respective entity is adapted to or configured to perform the respective steps and functionalities. Even if, in the following description, a specific functionality or step to be performed by external entities is not reflected in the description of a specific detailed element of that entity, which performs that specific step or functionality, it should be clear for a skilled person that these methods and functionalities can be implemented in respective software or hardware elements, or any kind of combination thereof.

[0057] BRIEF DESCRIPTION OF DRAWINGS

[0058] The above-described aspects and implementation forms will be explained in the following description in relation to the enclosed drawings, in which:

[0059] FIG. 1 shows a general structure of part of a transmitter;

[0060] FIG. 2 shows an example of a distribution matcher;

[0061] FIG. 3 shows a further example of a distribution matcher;FIG. 4 shows an example of sequence transformation;

[0062] FIG. 5 shows a further example of sequence transformation; and

[0063] FIG. 6 shows a diagram of a method.

[0064] DETAILED DESCRIPTION OF EMBODIMENTS

[0065] In FIGs. 1-6 below, corresponding elements may share the same features and function likewise.

[0066] FIG. 1 shows a general structure of part of a transmitter 100 according to the present disclosure. The transmitter 100 comprises a distribution matcher 110, which is configured to process an input message b 101. The distribution matcher 110 is configured to transform the message 101 into a list of first sequences s' . ... , sL105 according to a first criterion, such as minimizing the average transmit power or achieving a distribution closer to a target distribution (e.g., Gaussian distribution).

[0067] For instance, the distribution matcher 110 may be based on sign-bit shaping. The distribution matcher 110 may comprise an ASK mapper and a BPSK mapper. The ASK mapper may be a bipolar ASK. In this case, the distribution matcher 110 is configured to select sign-bits that are BPSK modulated by the BPSK mapper and added to the ASK symbols generated by the ASK mapper, in order to approximate a Gaussian signal distribution. Hence, the ASK symbols are shifted to the left or right depending on the sign-bits. Alternatively, the ASK mapper may be a unipolar ASK. In this case, the distribution matcher 110 is configured to generate unipolar ASK symbols corresponding to amplitudes, which are multiplied with uniformly distributed BPSK modulated sign-bits. Hence, the amplitude ASK symbols are flipped to the left or right depending on the sign-bits. In general, a plurality of first sequences 105 are output by the distribution matcher 110.

[0068] The list of first sequences 105 is then transformed into a list of second sequences x1, xL107. This operation may be performed by a transformation unit 120 comprised in the transmitter. In general, the transformation unit 120 may be configured to apply a function including additional processing steps on the first sequences, e.g., generating parity bits, combining the parity bits with the first sequences, performing non-linear pre-distortion, etc. As an example, each second sequence may be formed by superimposing symbols of its respective first sequence. That is, each second sequence represents a superposition of symbols of a respective first sequence. This superposition may occur through various transformations, such as frequency-time domain transformation and spatial precoding.

[0069] A selector 130 of the transmitter is configured to receive the list of second sequences 107 and select a to-be-transmitted sequence x 109 from the list of second sequences 107 according to a second criterion. For this purpose, the selector 130 may be configured to evaluate each second sequence according to the second criterion, such as PAPR or cubic metric. From this evaluation, the selector 130 can identify and select the to-be-transmitted 109 sequence that balances the competing requirements of low average power and low PAPR (or similar metrics).

[0070] Once selected, the transmitter 100 is configured to transmit the selected sequence 109.

[0071] FIG. 2 shows an example of a distribution matcher 210. The distribution matcher 210 of FIG. 2 may be applied to the distribution matcher 110 of FIG. 1. In general, a distribution matcher of this disclosure may be polar-based and may be configured to generate a list of amplitude sequences with corresponding bit labels. For each amplitude sequence, a systematicchannel encoder may be used to generate parity bits, that are mapped to BPSK symbols and added / multiplied with the amplitude sequences.

[0072] The transmitter 100 may implement sign-bit shaping with polar codes. In the example of FIG. 2, the distribution matcher 210 may comprise at least one polar list decoder 211 and at least one symbol mapper.

[0073] > >

[0074] >

[0075]

[0076] FIG. 3 shows a further example of a distribution matcher, which may be built based on the distribution matcher 210 in FIG. 2 and can be applied to the distribution matcher 110 in FIG. 1.

[0077] In the example of FIG. 3, the overall modulation process is divided into smaller, independently processed segments, enabling efficient parallel processing and scalable list generation, in order to reduce delay and computational complexity during the generation of candidate sequences.

[0078]

[0079] FIG. 4 shows an example of sequence transformation for transforming a list of first sequences into a list of second sequences. In this example, a transformation unit 420 is configured to perform frequency-time domain transformation. Operations like inverse Fourier transformation (IFFT) may be applied to convert the frequency-domain representation of each first sequence into a time-domain sequence as a respective second sequence suitable for transmission.

[0080] FIG. 5 shows a further example sequence transformation. In this example, a transformation unit 520 is configured to perform spatial precoding, which is used to combine symbols from the first sequence across spatial streams, resulting in second sequences that are optimized for spatial diversity or multiplexing gains.

[0081] The examples of FIGs. 4 and 5 may be combined with the examples shown in FIG. 2 or FIG. 3, and are applied to the transmitter 100 in FIG. 1.FIG. 6 shows a diagram of a method 600. The method 600 is applied to a transmitter and comprises the following steps: step 601: transforming, by a distribution matcher of the transmitter, a message into a list of first sequences based on a first criterion;

[0082] step 602: transforming, by the transmitter, the list of first sequences into a list of second sequences; step 603: selecting, by a selector of the transmitter, a to-be-transmitted sequence from the list of second sequences based on a second criterion; and

[0083] step 604: transmitting, by the transmitter, the selected sequence.

[0084] It is noted that the steps of method 600 may share the same functions and details from the perspective of FIGs. 1-5 described above.

[0085] A core idea of this disclosure is to generate multiple second sequences and select one with a smallest PAPR (or based on other metrics), while the average transmit power can be minimized during the generation of the multiple first sequences. In this way, the average transmit power and PAPR (or other metrics like cubic metric) may be simultaneously reduced without introducing nonlinear distortions or additional overhead. Only minor modifications are required at a receiver side to decode the transmitted message, as shaping bits can be treated by a decoder like message bits and then be discarded.

[0086] For instance, the mapping of message bits to transmitted symbols may be specified and signaled to the receiver. In particular, the allocation of shaping bits before or within the distribution matcher (e.g., to certain polar subchannels) needs to be specified.

[0087] Using shaping bits for PAPR reduction changes the distribution of the transmit signal, which needs to be known at the receiver for correct symbol demapping and decoding. To this end, a corresponding modulation and coding scheme (MCS) table may be defined, which is similar to that of transform precoding (= DFT-s-OFDM) in related 3GPP standards (e.g., 3GPP TS 38.211). Furthermore, PAPR reduction and signal shaping may be enabled or disabled by RRC signaling parameters. An example of RRC parameters is as follows:

[0088] PUSCH-Config : -SEQUENCE {

[0089] msc-T ablePaprReduction ENUMERATED {qam256, ShapedQam256}

[0090] PaprReduction ENUMERATED {enabled, disabled}

[0091] The present disclosure may be applied to uplink transmission. In this case, the transmitter corresponds to a UE. In this case, the transmitter may be configured to obtain the modulation and coding parameters from the network side or a base station. In general, the transmitter of the present disclosure may be configured to receive one or more parameters referencing to one or more modulation and coding scheme tables that specify a mapping of shaping bits and / or at least one parameter related to a symbol distribution.

[0092] The transmitter in the present disclosure may comprise processing circuitry configured to perform, conduct or initiate the various operations of the transmitter described herein, respectively. The processing circuitry may comprise hardware and software. The hardware may comprise analog circuitry or digital circuitry, or both analog and digital circuitry. The digitalcircuitry may comprise components such as application-specific integrated circuits (ASICs), field-programmable arrays (FPGAs), digital signal processors (DSPs), or multi-purpose processors. Optionally, the processing circuitry comprises one or more processors and a non-transitory memory connected to the one or more processors. The non-transitory memory may carry executable program code which, when executed by the one or more processors, causes the transmitter to perform, conduct or initiate the operations or methods described herein, respectively.

[0093] The present disclosure has been described in conjunction with various aspects as examples as well as implementations. However, other variations can be understood and effected by those persons skilled in the art and practicing the claimed subject matter, from the studies of the drawings, this disclosure and the independent claims. In the claims as well as in the description the word “comprising” does not exclude other elements or steps and the indefinite article “a” or “an” does not exclude a plurality. A single element or another unit may fulfill the functions of several entities or items recited in the claims. The mere fact that certain measures are recited in the mutual different dependent claims does not indicate that a combination of these measures cannot be used in an advantageous implementation.

Claims

CLAIMS1. A transmitter (100) comprising a distribution matcher (110) and a selector (130), whereinthe distribution matcher (110) is configured to transform a message (101) into a list of first sequences (105) based on a first criterion;the transmitter (100) is configured to transform the list of first sequences (105) into a list of second sequences (107); the selector (130) is configured to select a to-be-transmitted sequence (109) from the list of second sequences (107) based on a second criterion; andthe transmitter (100) is configured to transmit the selected sequence (109).

2. The transmitter (100) according to claim 1, wherein the first criterion comprises an average transmit power, or a Kullback-Leibler distance to a target distribution.

3. The transmitter (100) according to claim 1 or 2, wherein the second criterion comprises a peak-to-average power ratio, or a cubic metric.

4. The transmitter (100) according to any one of claims 1 to 3, wherein each second sequence represents a superposition of symbols of a respective first sequence.

5. The transmitter (100) according to any one of claims 1 to 4, wherein the distribution matcher (110) comprises one or more polar list decoders (211) configured to generate a list of candidate codewords based on the message.

6. The transmitter (100) according to claim 5, wherein the distribution matcher (110) comprises one or more symbol mappers configured to map the list of candidate codewords into the list of first sequences according to a preset modulation scheme.

7. The transmitter (100) according to any one of claims 1 to 6, wherein the list of first sequences (105) comprises sign bits and / or amplitude bits.

8. The transmitter (100) according to any one of claims 1 to 7, wherein for transforming the message (101) into the list of first sequences (105), the distribution matcher (110) is configured to combine multiple lists of subsequences generated based on the message (101).

9. The transmitter (100) according to any one of claims 1 to 8, wherein for transforming the list of first sequences (105) into the list of second sequences (107), the transmitter (100) is configured to apply inverse Fourier transformation on the list of first sequences (105), so as to obtain the list of second sequences (107).

10. The transmitter (100) according to any one of claims 1 to 9, wherein the transmitter (100) is configured to apply spatial precoding on the list of first sequences (105), so as to obtain the list of second sequences (107).

11. The transmitter (100) according to any one of claims 1 to 10, configured to receive one or more parameters referencing to one or more modulation and coding scheme tables that specify a mapping of shaping bits and / or at least one parameter related to a symbol distribution.

12. A method (600) applied to a transmitter comprising a distribution matcher and a selector, wherein the method comprises:transforming (601), by the distribution matcher, a message into a list of first sequences based on a first criterion; transforming (602), by the transmitter, the list of first sequences into a list of second sequences;selecting (603), by the selector, a to-be-transmitted sequence from the list of second sequences based on a second criterion; andtransmitting (604), by the transmitter, the selected sequence.

13. The method (600) according to claim 12, wherein the first criterion comprises an average transmit power, or aKullback-Leibler distance to a target distribution.

14. The method (600) according to claim 12 or 13, wherein the second criterion comprises a peak-to-average power ratio, or a cubic metric.

15. A computer program comprising instructions which, when the program is executed by a computer, cause the computer to perform the method according to any one of claims 12 to 14.

Images