CN117395115A - Communication method and device - Google Patents

Communication method and device Download PDF

Info

Publication number
CN117395115A
CN117395115A CN202210785094.2A CN202210785094A CN117395115A CN 117395115 A CN117395115 A CN 117395115A CN 202210785094 A CN202210785094 A CN 202210785094A CN 117395115 A CN117395115 A CN 117395115A
Authority
CN
China
Prior art keywords
bit sequence
bit
value
neural network
network model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210785094.2A
Other languages
Chinese (zh)
Inventor
刘永
孙荣朝
李润华
戴刚
董蕾
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN202210785094.2A priority Critical patent/CN117395115A/en
Publication of CN117395115A publication Critical patent/CN117395115A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/18—Phase-modulated carrier systems, i.e. using phase-shift keying
    • H04L27/20—Modulator circuits; Transmitter circuits
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/18—Phase-modulated carrier systems, i.e. using phase-shift keying
    • H04L27/22—Demodulator circuits; Receiver circuits
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/32—Carrier systems characterised by combinations of two or more of the types covered by groups H04L27/02, H04L27/10, H04L27/18 or H04L27/26
    • H04L27/34—Amplitude- and phase-modulated carrier systems, e.g. quadrature-amplitude modulated carrier systems
    • H04L27/36—Modulator circuits; Transmitter circuits
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/32—Carrier systems characterised by combinations of two or more of the types covered by groups H04L27/02, H04L27/10, H04L27/18 or H04L27/26
    • H04L27/34—Amplitude- and phase-modulated carrier systems, e.g. quadrature-amplitude modulated carrier systems
    • H04L27/38—Demodulator circuits; Receiver circuits

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The application relates to the field of communication technology. A communication method and apparatus for flexibly and simply performing modulation and demodulation are provided. And the transmitting end adjusts bits with the value of 0 in the first bit sequence to be transmitted to be integers which are not 0 and not 1, so as to obtain a second bit sequence. Adding 0 into the second bit sequence based on the number of neurons of an input layer of the neural network model to obtain a third bit sequence; the length of the third bit sequence is equal to the number of input layer neurons of the neural network model. Inputting the third bit sequence into a neural network model to obtain a first signal; the neural network model is used to modulate an input bit sequence. The number of neurons of an input layer of the neural network model is used as a fixed bit input length, and 0 is adopted to complement effective bits to be modulated so as to achieve the fixed bit input length, so that the modulation of processing the effective bits with different lengths by adopting a fixed neural network model is realized, and the method is simple and flexible.

Description

Communication method and device
Technical Field
The embodiment of the application relates to the fields of communication and the like, in particular to a communication method and device.
Background
With the increasing maturity of communication technology, the industry in the academic world is looking for higher rate, lower latency, greater coverage, safer, and more intelligent mobile communication technology. Terahertz communication, space-world integration, artificial intelligence and the like are frequently mentioned, wherein the deep learning algorithm of the artificial intelligence is highly attractive in the fields of computer vision, natural language processing and the like. With the improvement of the performance of hardware facilities, the application of the deep learning model in various industrial fields is also more and more extensive, and meanwhile, the combination research of the deep learning model and mobile communication is also gradually developed. The mobile communication includes coding, modulating, channel, equalizing, demodulating and decoding. When a neural network is used to realize modulation of signals, training is usually required under a fixed neural network structure, for example, different neural network models are designed for different modulation orders (or understood as different modulation modes), and the receiving end/transmitting end uses the neural network model with the corresponding order to perform modulation/demodulation, which is very complex and inflexible.
Disclosure of Invention
The embodiment of the application provides a communication method and device for flexibly and simply modulating and demodulating.
In a first aspect, a communication method is provided, where the execution body of the method may be a transmitting end, or may be a component applied in the transmitting end, such as a chip, a processor, or the like. The following describes an example in which the execution body is a transmitting end. The transmitting end adjusts a bit with a value of 0 in a first bit sequence to be transmitted to a first preset value to obtain a second bit sequence; wherein the first preset value is an integer which is not 0 and not 1. Then, based on the number of neurons of an input layer of the first neural network model, adding at least one bit with a value of 0 into the second bit sequence to obtain a third bit sequence; wherein the length of the third bit sequence is equal to the number of neurons of the input layer of the first neural network model. Inputting the third bit sequence into the first neural network model to obtain a first signal; the first neural network model is used for modulating an input bit sequence. And then, the first signal is sent.
And adjusting 0 in the original bit sequence to be transmitted to be a non-0 and non-1 value, taking the number of neurons of an input layer of the first neural network model as a fixed bit input length, and supplementing the effective bits (namely the first bit sequence) to be modulated by 0 to reach the fixed bit input length, wherein the length of the bit sequence input into the first neural network model is fixed, the length of the effective bits (namely the non-0 value) is variable, and the neuron nodes with 0 supplementing bits are not activated in the operation process of the neural network and cannot influence the processing of other neuron nodes. By the 0-filling operation, a fixed neural network model can be used to handle the modulation of the effective bits (i.e. non-0 values) of different lengths (i.e. i.s. i.s) and is simple and flexible.
In one possible implementation, the transmitting end sends first information to the receiving end, where the first information is used to indicate the number of bits with a value of 0 or the length of the first bit sequence. The transmitting end determines the first length of the first bit sequence or the number of added 0, and informs the receiving end of the first length or the number of added 0, so that the receiving end can accurately determine the effective bit sequence from the demodulated bit sequence.
In one possible implementation, the transmitting end may determine the number of bits with the value of 0 based on the signal quality parameter of the first channel; determining the length of the first bit sequence based on the number of neurons of an input layer of the first neural network model and the number of bits with the value of 0, wherein the first channel is a channel between the transmitting end and the receiving end; alternatively, the transmitting end may determine the length of the first bit sequence based on the signal quality parameter of the first channel; and determining the number of the at least one bit with a value of 0 based on the number of neurons of an input layer of the first neural network model and the length of the first bit sequence, wherein the first channel is a channel between the transmitting end and the receiving end. For example, the signal quality parameter includes at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. Determining the length of the effective bits (i.e., the first bit sequence) that need to be modulated based on the quality of the channel may improve the quality of the communication.
In one possible implementation, the transmitting end may receive second information from the receiving end, where the second information is used to indicate the number of bits with the value of 0; and then, adding at least one bit with the value of 0 into the second bit sequence according to the second information to obtain a third bit sequence. The receiving end determines the number of additions 0 and informs the transmitting end of the number of additions 0 so that the transmitting end can add 0 in the second bit sequence in an appropriate manner.
In one possible implementation, before adding at least one bit with a value of 0 to the second bit sequence according to the second information to obtain a third bit sequence, the transmitting end may further determine a length of the first bit sequence according to the number of neurons of the input layer of the first neural network model and the second information; and further determining the first bit sequence according to the length of the first bit sequence. The receiving end determines the number of added 0 and notifies the sending end of the number of added 0, and the sending end can determine the length of the effective bit sequence based on the number of added 0, thereby determining the effective bit sequence.
In one possible implementation, the transmitting end receives second information from the receiving end, where the second information is used to indicate a length of the first bit sequence. The transmitting end determines the number of the at least one bit with the value of 0 according to the number of the neurons of the input layer of the first neural network model and the second information; and the transmitting end adds at least one bit with the value of 0 into the second bit sequence based on the number of the at least one bit with the value of 0 to obtain a third bit sequence. The receiving end determines the length of the effective bit sequence and informs the length of the effective bit sequence to the transmitting end, so that the transmitting end determines the quantity of 0 added based on the length of the effective bit sequence, and then 0 can be added in the second bit sequence according to a proper mode.
In one possible implementation, before adding at least one bit with a value of 0 to the second bit sequence based on the number of the at least one bit with a value of 0 to obtain a third bit sequence, the transmitting end may further determine the first bit sequence according to the second information. The receiving end determines the length of the valid bit sequence and informs the transmitting end of the length of the valid bit sequence, so that the transmitting end determines the valid bit sequence based on the length of the valid bit sequence.
In one possible implementation, the transmitting end sends third information to the receiving end, where the third information indicates a position of the at least one bit with a value of 0 in the third bit sequence. The transmitting end determines the position of the added 0 in the third bit sequence and informs the receiving end of the position information, so that the receiving end can accurately determine the effective bit sequence from the demodulated bit sequence.
In one possible implementation, adding at least one bit with a value of 0 to the second bit sequence to obtain a third bit sequence, where the third bit sequence is obtained by receiving fourth information from a receiving end, where the fourth information indicates a position of the at least one bit with a value of 0 in the third bit sequence; and then adding at least one bit with the value of 0 into the second bit sequence according to the fourth information to obtain a third bit sequence. The receiving end determines the position of the added 0 in the third bit sequence and informs the transmitting end of the position information so that the transmitting end can add 0 in an appropriate manner.
In one possible implementation, the first preset value is-1.
In a second aspect, a communication method is provided, where the execution subject of the method may be a receiving end, or may be a component applied in the receiving end, such as a chip, a processor, or the like. The following describes an example in which the execution body is a receiving end. The receiving end receives a second signal, and then inputs the second signal into a second neural network model to obtain a fourth bit sequence; the second neural network model is used for demodulating the input signal, and the length of the fourth bit sequence is equal to the number of neurons of an output layer of the second neural network model. And then deleting at least one bit in the fourth bit sequence according to the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end, so as to obtain a fifth bit sequence. And then, adjusting the value of each bit in the fifth bit sequence to be a numerical value in a first set to obtain a sixth bit sequence, wherein the first set comprises a first preset value and 1. Then, the bit with the value of the first preset value in the sixth bit sequence is adjusted to be 0, and a first bit sequence is obtained; wherein the first preset value is an integer which is not 0 and not 1.
The transmitting end adjusts 0 in an original bit sequence to be transmitted to be a non-0 and non-1 numerical value, takes the number of neurons of an input layer of the first neural network model as a fixed bit input length, supplements effective bits (namely a first bit sequence) to be modulated by 0 to achieve the fixed bit input length, so that the length of the bit sequence input into the first neural network model is fixed, the length of the effective bits (namely a non-0 value) is variable, and a neuron node with the 0 supplemented bit is not activated in the operation process of the neural network and does not influence the processing of other neuron nodes. By the 0-filling operation, a fixed neural network model can be used to handle modulation of significant bits (i.e., non-0 values) of different lengths (i.e., variable lengths). Correspondingly, the receiving end operates the bit sequence in a mode corresponding to the transmitting end, namely, bit positions corresponding to the 0 supplementing position in the fourth bit sequence obtained through demodulation are deleted, and the rest bit positions are restored to original bit values, so that the original bit sequence is obtained.
In one possible implementation, the receiving end may receive first information from the transmitting end, where the first information indicates the number of bits with a value of 0 in a bit sequence carried in a signal sent by the transmitting end; the receiving end can determine the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end based on the first information. For example, the receiving end may determine, based on the first information and a predetermined start position/end position, a position of a bit having a value of 0 in a bit sequence carried in a signal sent by the transmitting end in the bit sequence carried in the signal sent by the transmitting end.
In one possible implementation, a receiving end may receive first information from a transmitting end, where the first information indicates a length of an original bit sequence to be transmitted, which is carried in a signal sent by the transmitting end; then, according to the number of neurons of an output layer of the second neural network model and the first information, determining the number of bits with the value of 0 in a bit sequence carried in a signal sent by the sending end; and then the receiving end determines the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end according to the number of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end. For example, the receiving end may determine, based on the number of bits with a value of 0 and a predetermined start position/end position, a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the transmitting end.
In one possible implementation, the receiving end sends second information to the sending end, where the second information indicates the number of bits with a value of 0 or the length of an original bit sequence to be sent (the original bit sequence to be sent may be understood as a first bit sequence on the sending end side) carried in a signal sent by the sending end. The receiving end determines the first length of the first bit sequence or the number of added 0, and informs the transmitting end of the first length or the number of added 0, so that the transmitting end can perform the operations of splitting the bit sequence and adding 0 in a proper manner.
In one possible implementation, the receiving end may determine, based on a signal quality parameter of a first channel, a number of bits with a value of 0 in a bit sequence carried in a signal sent by the sending end, and determine, based on a number of neurons of an output layer of the second neural network model and a number of bits with a value of 0 in a bit sequence carried in a signal sent by the sending end, a length of an original bit sequence to be sent carried in a signal sent by the sending end, where the first channel is a channel between the sending end and the receiving end. Or, the receiving end may determine the length of the original bit sequence to be sent, which is carried in the signal sent by the sending end, based on the signal quality parameter of the first channel, and determine the number of bits with a value of 0 in the bit sequence carried in the signal sent by the sending end, based on the number of neurons of the output layer of the second neural network model and the length of the original bit sequence to be sent, which is carried in the signal sent by the sending end, where the first channel is a channel between the sending end and the receiving end. For example, the signal quality parameter includes at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. Determining the length of the effective bits (i.e., the first bit sequence) that need to be modulated based on the quality of the channel may improve the quality of the communication.
In one possible implementation, the receiving end may receive third information from the transmitting end, where the third information indicates a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the transmitting end (where the position indicated by the third information may be understood as a position of a bit with a value of 0 in the third bit sequence by the transmitting end). And the receiving end can delete the bit at the corresponding position in the fourth bit sequence based on the third information to obtain a fifth bit sequence. The transmitting end determines the position of the added 0 in the third bit sequence and informs the receiving end of the position information, so that the receiving end can accurately determine the effective bit sequence from the demodulated bit sequence.
In one possible implementation, the receiving end may send fourth information to the sending end, where the fourth information indicates a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the sending end (where the position indicated by the fourth information may be understood as a position of a bit with a value of 0 in the sending end in a third bit sequence). The receiving end determines the position of the added 0 in the third bit sequence and informs the transmitting end of the position information so that the transmitting end can add 0 in an appropriate manner.
In one possible implementation, the first preset value is-1.
In a third aspect, a communication method is provided, where the execution subject of the method may be a transmitting end, or may be a component applied in the transmitting end, such as a chip, a processor, or the like. The following describes an example in which the execution body is a transmitting end. The transmitting end selects a first sub-neural network model from a third neural network model, the number of neurons of an input layer of the first sub-neural network model is the same as the value of the first length, and the third neural network model is used for modulating an input bit sequence. And then, inputting the first bit sequence with the first length into the first sub-neural network model to obtain a third signal. The third signal is then transmitted.
The sub-neural network model is obtained by dynamically selecting the neural network model to match the length of the effective bit sequence, and the modulation of the effective bit sequences with different lengths (namely, different lengths) can be processed by adopting one fixed neural network model, so that the method is simple and flexible.
In one possible implementation, the transmitting end may send fifth information to the receiving end, where the fifth information is used to indicate the length of the first bit sequence. The transmitting end determines the length of the bit sequence to be modulated and transmits the length to the receiving end, so that the receiving end can accurately select a corresponding sub-neural network model from the original neural network models for demodulation to demodulate.
In one possible implementation, the transmitting end may determine the length of the first bit sequence based on a signal quality parameter of a first channel, where the first channel is a channel between the transmitting end and the receiving end. For example, the signal quality parameter includes at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. Determining the length of the bit sequence that needs to be modulated, i.e. the first bit sequence, based on the quality of the channel may improve the quality of the communication.
In one possible implementation, the transmitting end may receive sixth information from the receiving end, where the sixth information is used to indicate a length of the first bit sequence; then, a first bit sequence is determined based on the length of the first bit sequence, and a first sub-neural network model is selected from the third neural network models. The receiving end determines the length of the bit sequence to be modulated and informs the transmitting end of the length, so that the transmitting end can determine the bit sequence to be modulated based on the length and execute a first sub-neural network model of the modulation operation.
In one possible implementation, the transmitting end may send seventh information to the receiving end, where the seventh information is used to indicate a location of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model. The transmitting end determines the position of the input layer neuron of the sub-neural network model in the input layer neuron of the original neural network model for modulation and notifies the receiving end of the position, so that the receiving end can accurately select the corresponding sub-neural network model from the original neural network model for demodulation based on the position for demodulation.
In one possible implementation, when the transmitting end selects the first sub-neural network model in the third neural network model, it may be that the transmitting end receives eighth information from the receiving end, where the eighth information is used to indicate a location of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model; then, the transmitting end may select the first sub-neural network model from the third neural network models according to the sixth information. The receiving end determines the position of the input layer neuron of the sub-neural network model in the input layer neuron of the original neural network model for modulation, and notifies the transmitting end of the position, so that the transmitting end can accurately select the corresponding sub-neural network model from the original neural network model for modulation based on the position to perform modulation.
In a fourth aspect, a communication method is provided, where the execution subject of the method may be a receiving end, or may be a component applied in the receiving end, such as a chip, a processor, or the like. The following describes an example in which the execution body is a receiving end. The receiving end receives the fourth signal. Then, selecting a second sub-neural network model from the fourth neural network model, wherein the number of neurons of an output layer of the second sub-neural network model is the same as the value of a first length, and the first length is the length of a bit sequence carried in a signal sent by a sending end; the fourth neural network model is used for demodulating an input signal. And inputting the fourth signal into the second sub-neural network model to obtain a seventh bit sequence with the first length.
The sub-neural network model is obtained by dynamically selecting the neural network model to match the length of the effective bit sequence, and the demodulation of the effective bit sequences with different lengths (namely, variable lengths) can be processed by adopting a fixed neural network model, so that the demodulation method is simple and flexible.
In one possible implementation, the receiving end receives fifth information from the transmitting end, where the fifth information is used to indicate a length of a bit sequence carried in a signal sent by the transmitting end. And then, the receiving end selects a second sub-neural network model from the fourth neural network models according to the fifth information. The transmitting end determines the length of the bit sequence to be modulated and transmits the length to the receiving end, so that the receiving end can accurately select a corresponding sub-neural network model from the original neural network models for demodulation to demodulate.
In one possible implementation, the receiving end sends sixth information to the sending end, where the sixth information is used to indicate a length of a bit sequence carried in a signal sent by the sending end. The receiving end determines the length of the bit sequence to be modulated and informs the transmitting end of the length, so that the transmitting end can determine the bit sequence to be modulated based on the length and execute a first sub-neural network model of the modulation operation.
In one possible implementation, the receiving end may determine a length of a bit sequence carried in a signal sent by the sending end based on a signal quality parameter of a first channel, where the first channel is a channel between the sending end and the receiving end. For example, the signal quality parameter includes at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. Determining the length of the bit sequence that needs to be modulated, i.e. the first bit sequence, based on the quality of the channel may improve the quality of the communication.
In one possible implementation, when the second sub-neural network model is selected from the fourth neural network models, the seventh information may be received from the transmitting end, where the seventh information is used to indicate the position of the sub-neural network model used by the transmitting end to perform the modulation in the original neural network model (i.e., the position of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model); then, a second sub-neural network model is selected among the fourth neural network models according to the seventh information. The output layer neurons of the second sub-neural network model are located in the same positions in the output layer neurons of the fourth neural network model as the input layer neurons of the first sub-neural network model are located in the input layer neurons of the third neural network model. The transmitting end determines the position of the input layer neuron of the sub-neural network model in the input layer neuron of the original neural network model for modulation and notifies the receiving end of the position, so that the receiving end can accurately select the corresponding sub-neural network model from the original neural network model for demodulation based on the position for demodulation.
In one possible implementation, the receiving end sends eighth information to the sending end, where the eighth information is used to indicate a location of the sub-neural network model used by the sending end to perform the modulation in the original neural network model (i.e., a location of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model). The receiving end determines the position of the input layer neuron of the sub-neural network model in the input layer neuron of the original neural network model for modulation, and notifies the transmitting end of the position, so that the transmitting end can accurately select the corresponding sub-neural network model from the original neural network model for modulation based on the position to perform modulation.
In a fifth aspect, there is provided a communications device having functionality to implement any of the above aspects and any possible implementation of any of the aspects. These functions may be implemented by hardware, or may be implemented by hardware executing corresponding software. The hardware or software includes one or more functional modules corresponding to the functions described above.
In a sixth aspect, a communications apparatus is provided that includes a processor, optionally, a memory; the processor and the memory are coupled; the memory is used for storing a computer program or instructions; the processor is configured to execute part or all of the computer program or instructions in the memory, which when executed, is configured to implement the functions in the method of any one of the above aspects and any one of the possible implementations of any one of the above aspects.
In one possible implementation, the apparatus may further include a transceiver for transmitting the signal processed by the processor or receiving a signal input to the processor. The transceiver may perform the transmitting or receiving actions of any aspect and any possible implementation of any aspect.
In a seventh aspect, the present application provides a chip system comprising one or more processors (which may also be referred to as processing circuits) electrically coupled between the processors and a memory (which may also be referred to as storage medium); the memory may or may not be located in the chip system; the memory is used for storing a computer program or instructions; the processor is configured to execute part or all of the computer program or instructions in the memory, which when executed, is configured to implement the functions in the method of any one of the above aspects and any one of the possible implementations of any one of the above aspects.
In one possible implementation, the chip system may further include an input/output interface (may also be referred to as a communication interface), which is configured to output a signal processed by the processor or receive a signal input to the processor. The input-output interface may perform the sending or receiving actions of any aspect and any possible implementation of any aspect. Specifically, the output interface performs a transmission action, and the input interface performs a reception action.
In one possible implementation, the chip system may be formed of a chip, or may include a chip and other discrete devices.
In an eighth aspect, there is provided a computer readable storage medium storing a computer program comprising instructions for implementing the functions in any aspect and any possible implementation of any aspect.
Alternatively, a computer readable storage medium storing a computer program which, when executed by a computer, may cause the computer to perform any of the above aspects and any possible implementation of the method of any of the above aspects.
In a ninth aspect, there is provided a computer program product comprising: computer program code which, when run on a computer, causes the computer to perform the method of any one of the above aspects and any one of the possible implementations of any one of the above aspects.
In a tenth aspect, a communication system is provided, the communication system comprising a transmitting end performing the method in any of the possible implementations of the first aspect and a receiving end performing the method in any of the possible implementations of the second aspect and the second aspect. Or, a transmitting end performing the method in any one of the possible implementations of the third aspect and a receiving end performing the method in any one of the possible implementations of the fourth aspect and the fourth aspect are included.
Technical effects of the fifth to tenth aspects described above may be referred to the descriptions in the first to fourth aspects, and the repetition is omitted.
Drawings
Fig. 1a is a schematic diagram of a neural network model according to an embodiment of the present application;
fig. 1b is a schematic diagram of a communication system architecture according to an embodiment of the present application;
fig. 1c is a schematic diagram of a communication model under an additive white gaussian noise (additive white gaussian noise, AWGN) channel according to an embodiment of the present application;
fig. 2a is a schematic diagram of binary phase shift keying (binary phase shift keying, BPSK) modulation according to an embodiment of the present application;
fig. 2b is a schematic diagram of quadrature phase shift keying (quadrature phase shift keying, QPSK) modulation according to an embodiment of the present application;
fig. 2c is a schematic diagram of 16 quadrature amplitude modulation (quadrature amplitude modulation, QAM) provided in an embodiment of the present application;
FIG. 2d is a diagram of a 64 quadrature amplitude modulation QAM provided by an embodiment of the present application;
fig. 2e is a schematic diagram of 256 QAM provided in an embodiment of the present application;
fig. 3a is a schematic diagram of a communication flow provided in an embodiment of the present application;
fig. 3b is a schematic diagram of a communication flow provided in an embodiment of the present application;
Fig. 4 is a schematic diagram of a communication flow provided in an embodiment of the present application;
fig. 5a is a schematic diagram of a communication flow provided in an embodiment of the present application;
fig. 5b is a schematic diagram of a communication flow provided in an embodiment of the present application;
fig. 5c is a schematic diagram of a high-level network and a low-level network according to an embodiment of the present application;
fig. 6 is a block diagram of a communication device according to an embodiment of the present application;
fig. 7 is a block diagram of a communication device according to an embodiment of the present application.
Detailed Description
To facilitate understanding of embodiments of the present application, some of the terms of embodiments of the present application are explained below to facilitate understanding by those skilled in the art.
1) A terminal device, also called User Equipment (UE), mobile Station (MS), mobile Terminal (MT), terminal, etc., is a device that provides voice and/or data connectivity to a user. For example, the terminal device includes a handheld device, an in-vehicle device, an on-board device, and the like having a wireless connection function. Currently, the terminal device may be: a mobile phone, a tablet, a notebook, a palm, a mobile internet device (mobile internet device, MID), a wearable device, a Virtual Reality (VR) device, an augmented reality (augmented reality, AR) device, a wireless terminal in industrial control (industrial control) (e.g., a sensor, etc.), a wireless terminal in unmanned (self-driving), a wireless terminal in teleoperation (remote medical surgery), a wireless terminal in smart grid (smart grid), a wireless terminal in transportation security (transportation safety), a wireless terminal in smart city (smart city), or a wireless terminal in smart home (smart home), or a wireless terminal with Vehicle-to-Vehicle (V2V), or a wireless terminal with Vehicle-to-Vehicle (Vehicle to everything, V2X), or long term evolution technology of workshop communication (long term evolution Vehicle) LTE-V functions, etc. But also subscriber units (subscriber units), cellular phones (cellphones), smart phones (smart phones), wireless data cards, personal digital assistants (personal digital assistant, PDA) computers, wireless modems (modems), hand-held devices (handsets), laptop computers (lap computers), machine type communication (machine type communication, MTC) terminals, etc.
2) A (radio) access network device ((R) AN), which is a device that provides a wireless communication function for a terminal device, is also called AN access network device. RAN devices in this application include, but are not limited to: next generation base stations (gnodebs, gnbs) in 5G, evolved node bs (enbs), radio network controllers (radio network controller, RNC), node bs (node bs, NB), base station controllers (base station controller, BSC), base transceiver stations (base transceiver station, BTS), home base stations (e.g., home evolved nodeB, or home node B, HNB), baseBand units (BBUs), access Points (APs) in wireless fidelity (wireless fidelity, WIFI) systems, wireless relay nodes, wireless backhaul nodes, transmission points (transmitting and receiving point, TRP), transmission points (transmitting point, TP), mobile switching centers, and the like. In systems employing different radio access technologies, the names of base station capable devices may vary, for example, in fifth generation (5th generation,5G) systems, referred to as RAN or gNB (5G NodeB); in the LTE system, it is called evolved NodeB (eNB or eNodeB); in the third generation (3rd generation,3G) system, it is called a Node B (Node B) or the like. The relay node may be in the form of a small station, an integrated access and backhaul (integrated access and backhauling, IAB) node, a Distributed Unit (DU), a terminal device, a transceiver point (transmitter and receiver point, TRP), a relay transmission reception point (rTRP), an IAB node (IAB node), and the like.
In addition, the "network element" referred to herein may be referred to as a "device", or an "entity", etc.
3) The signal-to-noise ratio (SNR or S/N), which is the ratio of the power of the output signal of an amplifier to the noise power of the simultaneous output, is often expressed in decibels.
4) The signal-to-interference-plus-noise ratio (signal to interference plus noise ratio, SINR), which refers to the ratio of the strength of the received useful signal to the strength of the received interfering signal (noise and interference), can be understood as the "signal-to-noise ratio".
5) Channel quality indication (channel quality indication, CQI), measured by the UE, is generally referred to as downlink channel quality. In general, the higher the coding mode is, the higher the channel condition requirement is, because the downlink scheduling is determined by the access network device, and the access network device is used as the transmitting end and does not know the quality of the channel quality, the access network device determines what coding mode is to be adopted, and the terminal device is required to feed back the quality of the channel.
6) A neural network model comprising n layers of neurons, each of the n layers of neurons comprising one or more neurons, all neurons of each layer being connected with all neurons of a next layer. Taking the neural network model 100 in fig. 1a as an example, the layer 1 includes two neurons, each of the layers 2 to n-1 includes three neurons, and the layer n includes one neuron, where n is a positive integer not less than 2, and i in fig. 1a is a positive integer not greater than n and not less than 1. The first layer neurons may be referred to as input layer neurons for inputting training data. The last layer of neurons may be referred to as output layer neurons for outputting the prediction result.
The technical solution of the embodiment of the application can be applied to various communication systems, for example: satellite communication system, conventional mobile communication system. Wherein the satellite communication system may be integrated with a conventional mobile communication system, i.e. a terrestrial communication system. A communication system such as: wireless local area network (wireless local area network, WLAN) communication systems, wireless fidelity (wireless fidelity, wiFi) systems, long term evolution (long term evolution, LTE) systems, LTE frequency division duplex (frequency division duplex, FDD) systems, LTE time division duplex (time division duplex, TDD), fifth generation (5th generation,5G) systems or New Radio (NR), sixth generation (6th generation,6G) systems, and other future communication systems, and the like, and also support communication systems in which multiple wireless technologies are integrated, for example, systems in which non-terrestrial networks (non-terrestrial network, NTN) such as unmanned aerial vehicles, satellite communication systems, high altitude platform (high altitude platform station, HAPS) communication are integrated.
In order to facilitate understanding of the embodiments of the present application, the application scenario of the present application is described next, where the network architecture and the service scenario described in the embodiments of the present application are for more clearly describing the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application, and as a person of ordinary skill in the art can know that, with the appearance of a new service scenario, the technical solutions provided by the embodiments of the present application are applicable to similar technical problems.
The present application may be applicable to any two communication devices in a communication system that may have a wireless communication connection. As shown in fig. 1b, the application may be applied to a communication system of an access network device and a terminal device, or may be applied to a sidelink communication system in which a terminal device directly communicates with a terminal device. The method and the device can be applied to communication scenes with or without network coverage, and can be in the coverage of network equipment or outside the coverage of the network equipment.
In a mobile communication scenario, the impact of the channel is not negligible. Many techniques have emerged to overcome the effects of channels, and the basic idea is generally to roll back the communication problem of complex scenarios to simple additive white gaussian noise (additive white gaussian noise, AWGN) scenarios and to achieve distortion-free communication via source channel coding. As shown in fig. 1c, the communication model under AWGN channel.
The mobile communication includes coding, modulating, channel, equalizing, demodulating and decoding. The modulation is to adjust a certain number of bits according to a constellation diagram determined in advance, and the modulation order is how many bits are adjusted to be one symbol. And the distance between constellation points is increased as much as possible under the condition of meeting the limitation of average energy, so that misjudgment is reduced, and a theoretical basis is provided for demodulation and decoding of a receiver side. The modulation mode is related to the channel quality, the transmitting end can modulate the bit sequence based on the channel quality selection and the proper adjustment mode, and the receiving end needs to select different demodulation modes according to different modulation modes.
The current constellation modulation mode is to modulate the bit sequence according to the modulation rule of the uniform constellation. Fig. 2a corresponds to binary phase shift keying (binary phase shift keying, BPSK) modulation, where one bit is adjusted to one symbol; fig. 2b corresponds to quadrature phase shift keying (quadrature phase shift keying, QPSK) modulation, which is to adjust 2 bits to one symbol; fig. 2c corresponds to 16 quadrature amplitude modulation (quadrature amplitude modulation, QAM) in which 4 bits are modulated into one symbol; fig. 2d corresponds to 64 quadrature amplitude modulation QAM, which adjusts 6 bits to one symbol; fig. 2e corresponds to 256QAM modulation, which is to adjust 8 bits to one symbol. At present, only the modulation mode of the even power of 2 can be processed, and the dynamically adjustable range is limited.
In addition, it is possible to consider an artificial intelligence (artificial intelligence, AI) method to implement the modulation, for example, using a reinforcement learning algorithm to obtain rewards Reward (rewards such as throughput, block error rate (block error ratio, BLER)) for various Action actions (actions such as correction values) in each State (different states are determined by different parameters), generalizing the Reward for all actions in various states by a neural network based on training samples, and selecting the optimal value as the final scheduling result.
The modulation mode implemented by using the neural network generally needs to be trained under a fixed neural network structure, for example, different neural network structures are respectively designed for different modulation orders (or understood as different modulation modes), and the receiving end and the transmitting end both adopt the neural network with the corresponding order to perform modulation and demodulation, which is very complex and cumbersome.
Based on this, the present application proposes a method that can accommodate modulation of a plurality of different effective bits.
The detailed description will be given next with reference to the accompanying drawings. Features or content identified in the drawings by dashed lines may be understood as optional operations or optional structures of embodiments of the present application. The contents of the embodiments/examples of the present application may refer to each other, and each embodiment/example may be taken as a separate embodiment, or may be combined as one embodiment.
Example 1:
as shown in fig. 3a, the present application provides a schematic communication flow, where the execution body of the process may be a transmitting end (note that the transmitting end may also have a function of receiving a signal), for example, a terminal, an access network device, and so on.
The method at least comprises the following steps:
Step 301: and the transmitting end adjusts the bit with the value of 0 in the first bit sequence to be transmitted to be a first preset value, so as to obtain a second bit sequence.
The transmitting end can split the bit sequence to be transmitted into a plurality of bit sequences, and select a first bit sequence from the plurality of bit sequences; alternatively, the transmitting end may split the bit sequence to be transmitted into a plurality of first bit sequences. The length of the first bit sequence is set to a first length. The first length is set to m, m being an integer greater than or equal to 1. The length of the bit sequence to be transmitted is set to be n, n is an integer greater than or equal to 1, and n can be an integer multiple of m, and can be a non-integer multiple of m. The transmitting end can divide the n bits into a plurality of groups according to one group of m bits in turn, and each group is a first bit sequence with a first length. For example, n is 128 and m is 4, then it can be divided into 32 groups; for example, n is 120 and m is 6, and can be divided into 20 groups.
Typically, any bit in the bit sequence has a value of 0 or 1. The transmitting end adjusts a bit with a value of 0 in a first bit sequence to be transmitted to a first preset value, wherein the first preset value meets the requirements of non-0 and non-1. In an alternative example, if the first preset value is-1, the transmitting end may adjust the bit with the value of 0 in the first bit sequence to be-1. Alternatively, the first preset value may be an integer value of-2, -3, or the like. In addition, the bit with the value of 1 in the first bit sequence is still 1.
In another alternative example, the transmitting end may modulate a bit with a value of 0 in the bit sequence to be transmitted to a first preset value, so as to obtain a first bit sequence. And the transmitting end sequentially selects a second bit sequence with the first length from the first bit sequences. That is, the order of adjusting 0 to the first preset value and selecting the bit sequence of the first length may not be limited.
Step 302: and the transmitting end adds at least one bit with the value of 0 into the second bit sequence based on the number of neurons of an input layer of the first neural network model to obtain a third bit sequence.
The length of the third bit sequence is equal to the number of neurons of the input layer of the first neural network model. The first neural network model is used for modulating an input bit sequence, and a third bit sequence can be input to the first neural network model for modulation.
The number of bits added to the second bit sequence having a value of 0 is set to a first number (the first number may be referred to as Padding size), the first number is set to s, and the first number is an integer greater than or equal to 1. The number of input layer neurons (i.e. the length of the third bit sequence) is fixed, and the first length and the first number are adjustable. The third bit sequence has a length of m + s. The length of the third bit sequence may be an even power of 2, or an odd power of 2, or any power of any number. Compared with a modulation mode which can only process the even power of 2, the dynamic adjustable range is enlarged.
As shown in fig. 4, the number of neurons in the input layer of the neural network model is 6. If the first length of the first bit sequence is 2, 4 values of 0 bits can be complemented, and since the neural network model is a fully connected network, 0 is not functional at all as an input, and an input of total length 6 can be equivalent to an input of length 2. If the first bit sequence has a first length of 4, 2 bits of value 0 may be complemented and an input of total length 6 may be equivalent to an input of length 4.
Step 303: and the transmitting end modulates the third bit sequence to obtain a first signal.
In one example, the third bit sequence is input into the first neural network model to obtain a first signal. The neural network model is used for modulating an input signal, and the length of the third bit sequence is equal to the number of neurons of the input layer of the first neural network model.
After the processing of step 301, one or more second bit sequences may be obtained; accordingly, after the processing of step 302, one or more third bit sequences may be obtained; accordingly, after the modulation process of step 303, one or more first signals may be obtained.
Step 304: and the transmitting end transmits the first signal.
According to the method, 0 in an original bit sequence to be transmitted is adjusted to be a non-0 and non-1 numerical value, the number of neurons of an input layer of a first neural network model is used as a fixed bit input length, 0 is adopted to supplement effective bits (namely, a first bit sequence) to be modulated so as to achieve the fixed bit input length, the length of the bit sequence input into the first neural network model is fixed, the length of the effective bits (namely, a non-0 value) is variable, and a neuron node with the 0 supplement bit is not activated in the operation process of the neural network and does not influence the processing of other neuron nodes. By the 0-filling operation, a fixed neural network model can be used to handle the modulation of the effective bits (i.e. non-0 values) of different lengths (i.e. i.s. i.s) and is simple and flexible.
In one example, the sender may determine a first length and/or a first number.
In this example, the transmitting end transmits the first information to the receiving end, and accordingly, the receiving end receives the first information from the transmitting end. Wherein the first information is used for indicating the number (i.e. the first number) of added bits with a value of 0, and/or the first information is used for indicating the first length of the first bit sequence. The first information may include a specific value of the first number (i.e., the number of bits added to a value of 0) and/or a specific value of the first length, or other information capable of identifying the first number and/or the first length. The transmitting end determines the first length of the first bit sequence or the number of added 0, and informs the receiving end of the first length or the number of added 0, so that the receiving end can accurately determine the effective bit sequence from the demodulated bit sequence.
The transmitting end may first determine the first number, and then determine the first length according to the number of neurons of the input layer and the first number of neurons of the neural network model. For example, the difference between the number of neurons in the input layer and the first number is a value of the first length. For example, the first number may be related to a signal quality parameter, and the transmitting end may determine the first number based on the signal quality parameter of a first channel, which is a channel between the transmitting end and the receiving end. Of course, the transmitting end may determine the first number in other manners, which is not limited in this application. The signal quality parameter of the first channel may be determined by the transmitting end itself, or may be transmitted to the transmitting end by the receiving end.
The transmitting end may determine the first length of the first bit sequence first, and then determine the first number according to the number of neurons of the input layer and the first length of the first neural network model. For example, the difference between the number of neurons in the input layer and the value of the first length is the first number. For example, the first length may be related to a signal quality parameter, and the transmitting end may determine the first length based on the signal quality parameter of the first channel. Of course, the transmitting end may determine the first length in other manners, which is not limited in this application. The signal quality parameter of the first channel may be determined by the transmitting end itself, or may be transmitted to the transmitting end by the receiving end.
If the signal quality parameter of the first channel is sent by the receiving end to the sending end, the sending end may also receive the signal quality parameter of the first channel from the receiving end before sending the first information to the receiving end. That is, before the receiving end receives the first information, the receiving end may also send the signal quality parameter of the first channel to the transmitting end.
The signal quality parameters include, but are not limited to, at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. The signal quality parameter may also be determined based on the number and/or frequency of ACK or NACK replies to the peer. For example, a NACK may be recovered always in the case of a decoding error, indicating that the signal quality parameter is not good.
Determining the length of the effective bits (i.e., the first bit sequence) that need to be modulated based on the quality of the channel may improve the quality of the communication.
In another example, the receiving end may determine the first length and/or the first number.
For example, the receiving end sends second information to the sending end, and correspondingly, the sending end receives the second information from the receiving end, where the second information is used to indicate the number (i.e. the first number) of at least one bit with a value of 0. And the transmitting end can add at least one bit with the value of 0 into the second bit sequence according to the second information to obtain a third bit sequence. The receiving end determines the number of additions 0 and informs the transmitting end of the number of additions 0 so that the transmitting end can add 0 in the second bit sequence in an appropriate manner. In addition, before the transmitting end adds at least one bit with a value of 0 in the second bit sequence according to the second information to obtain a third bit sequence, the transmitting end can also determine a first length of the first bit sequence according to the number of neurons of an input layer of the first neural network model and the second information, and then determine the first bit sequence in a bit stream to be transmitted based on the first length. The receiving end determines the number of added 0 and notifies the sending end of the number of added 0, and the sending end can determine the length of the effective bit sequence based on the number of added 0, thereby determining the effective bit sequence. The second information may include a specific value of the first number (i.e., the number of bits added to a value of 0) or other information capable of identifying the first number.
For another example, the receiving end sends second information to the sending end, and correspondingly, the sending end receives the second information from the receiving end, where the second information is used to indicate the first length of the first bit sequence. The transmitting end may determine the first number according to the number of neurons of the input layer of the first neural network model and the second information, and further may add a bit with a value of 0 to the second bit sequence based on the first number (which may be understood as adding the first number of bits with a value of 0 to the second bit sequence) to obtain a third bit sequence. The receiving end determines the length of the effective bit sequence and informs the length of the effective bit sequence to the transmitting end, so that the transmitting end determines the quantity of 0 added based on the length of the effective bit sequence, and then 0 can be added in the second bit sequence according to a proper mode. In addition, the transmitting end may further determine the first bit sequence according to the second information before adding at least one bit with a value of 0 to the second bit sequence based on the number of the at least one bit with a value of 0 to obtain a third bit sequence. I.e. a first bit sequence of a first length is determined in the bit stream to be transmitted. The receiving end determines the length of the valid bit sequence and informs the transmitting end of the length of the valid bit sequence, so that the transmitting end determines the valid bit sequence based on the length of the valid bit sequence. The second information may include a specific value of the first length or other information capable of identifying the first length.
For another example, the receiving end sends second information to the sending end, and correspondingly, the sending end receives the second information from the receiving end, where the second information is used to indicate the number of at least one bit with a value of 0 and the first length of the first bit sequence. The transmitting end may determine a first length of the first bit sequence according to the number of neurons of the input layer of the first neural network model and the second information, and further determine the first bit sequence in the bit stream to be transmitted based on the first length. The transmitting end may further add a bit with a value of 0 to the second bit sequence based on the first number (which may be understood as adding a first number of bits with a value of 0 to the second bit sequence) to obtain a third bit sequence.
The receiving end may determine the first length and/or the first number based on the signal quality parameter of the first channel, or may determine the first length and/or the first number by other manners, which is not limited in this application. For the receiving end, the signal quality parameter of the first channel may be determined by the receiving end itself, or may be sent by the sending end to the receiving end.
In an alternative example, the transmitting end may further transmit the signal quality parameter of the first channel to the receiving end before receiving the second information. So that the receiving end determines the second information based on the signal quality parameter of the first channel. For example, the transmitting end may determine the signal quality parameter, but may not support determining the first number or the first length based on the signal quality parameter (e.g., the mapping relationship between the value of the first length (or the value of the first number) and the signal quality parameter is not configured); the receiving end cannot determine the signal quality parameter, but may support determining the first number or the first length based on the signal quality parameter. The transmitting end may transmit the signal quality parameter to the receiving end, and the receiving end determines the first number or the first length based on the signal quality parameter and transmits the first number or the first length to the transmitting end.
In another example, both the transmitting end and the receiving end support determining the first number or the first length based on the signal quality parameters in the same manner (e.g. the same mapping relation (i.e. the value of the first length (or the mapping relation of the first number of values) and the signal quality parameters) is configured), but only the receiving end supports measuring the signal quality parameter(s), and the transmitting end does not support measuring the signal quality parameter(s). The receiving end may determine the first length and/or the first number based on the mapping relation and the determined signal quality parameter of the first channel. The receiving end may also send the signal quality parameter of the first channel to the transmitting end, so that the transmitting end determines the first length and/or the first number based on the mapping relation and the received signal quality parameter of the first channel.
In another example, both the transmitting end and the receiving end support determining the first number or the first length based on the signal quality parameters in the same manner (e.g. the same mapping relation (i.e. the value of the first length (or the mapping relation of the first number of values) and the signal quality parameters) is configured), but only the transmitting end supports measuring the signal quality parameter(s) and the receiving end does not support measuring the signal quality parameter(s). The transmitting end may determine the first length and/or the first number based on the mapping relation and the determined signal quality parameter of the first channel. The transmitting end may also transmit the signal quality parameter of the first channel to the receiving end, so that the receiving end determines the first length and/or the first number based on the mapping relation and the received signal quality parameter of the first channel.
In yet another example, both the transmitting end and the receiving end support determining the first number or the first length based on the signal quality parameter in the same manner (e.g. the same mapping relation (i.e. the value of the first length (or the mapping relation of the first number of values)) is configured), and both support measuring one or some signal quality parameter (e.g. the signal quality parameter has uplink and downlink reciprocity, e.g. SNR). In this way, the receiving end and the transmitting end may determine the first number or the first length based on the mapping relationship and the signal quality parameter of the first channel determined by themselves. The process does not need to interact, and signaling can be saved.
In one example, the transmitting end may determine the position of the bit having a value of 0 in the third bit sequence. For example, the transmitting end sends third information to the receiving end, and correspondingly, the receiving end receives the third information from the transmitting end, where the third information indicates a position of a bit with a value of 0 in the third bit sequence. The transmitting end determines the position of the added 0 in the third bit sequence and informs the receiving end of the position information, so that the receiving end can accurately determine the effective bit sequence from the demodulated bit sequence.
In one example, the receiving end may determine the position of the bit having a value of 0 in the third bit sequence. For example, the receiving end sends fourth information to the sending end, and correspondingly, the sending end receives fourth information from the receiving end, where the fourth information indicates a position of a bit with a value of 0 in the third bit sequence. And the transmitting end can add at least one bit with the value of 0 into the second bit sequence according to the fourth information to obtain a third bit sequence. The receiving end determines the position of the added 0 in the third bit sequence and informs the transmitting end of the position information so that the transmitting end can add 0 in an appropriate manner.
The third information or fourth information may include, but is not limited to, one or more of the following: for indicating fetchingBitmap of a specific position of a bit having a value of 0 in the third bit sequence, bit of a specific position of a bit having a value of 0 in the third bit sequence (e.g., kth j Bit position, kth j+s A bit position), a start position of a bit having a value of 0 in the third bit sequence, an end position of a bit having a value of 0 in the third bit sequence, a continuation position of a bit having a value of 0 in the third bit sequence (e.g. kth j To the kth j+s A number of bit positions), the number of bits with a value of 0, etc.
Several examples of the positions of the 0 valued bits in the third bit sequence are presented below.
The first number of 0 s may be located at the first s bit positions of the third bit sequence, i.e. the first number of 0 s is added before the second bit sequence, resulting in the third bit sequence. For example, the first number is 3, the second bit sequence is 1, -1, the third bit sequence may be 0, 1, -1.
The first number of 0 s may be located at s bit positions after the third bit sequence, i.e. the first number of 0 s is added after the second bit sequence, resulting in the third bit sequence. For example, the first number is 3, the second bit sequence is 1, -1, the third bit sequence may be 1, -1, 0.
The first number of 0 s may be located in the middle s bit positions of the third bit sequence, i.e. the first number of 0 s is added in the middle of the second bit sequence, resulting in the third bit sequence. For example, the first number is 3, the second bit sequence is 1, -1, the third bit sequence may be 1, 0, 1, -1-1, or 1, 0, -1.
The first number of 0 s may be uniformly located in the third bit sequence. For example, the first number is 3, the second bit sequence is 1, -1, the third bit sequence may be 1, 0, -1-1, or 1, 0, -1.
The first number of 0 s may be located at the kth of the third bit sequence j To the kth j+s At bit positions. For example, the first number is 3 and the second bit sequence is 1, -1, k 1 For 5, the third bit sequence may be 1, -1, 0, -1.
The positions of the bits with the value 0 added in the second bit sequence in the third bit sequence may be continuous or not continuous, may be uniform (for example, 0 is added to every other bit or 0 is added to every 2 bits), or may not be uniform.
In one example, the transmitting end and/or the receiving end may determine the position of the bit having a value of 0 in the third bit sequence based on the number of bits having a value of 0.
For example, a start position or an end position of adding a bit having a value of 0 to the second bit sequence may be defined in advance, and the predetermined may be arranged only on the receiving side, only on the transmitting side, or both the receiving side and the transmitting side. Alternatively, both the receiving end and the transmitting end define a start position or an end position at which a bit having a value of 0 is added to the second bit sequence.
For example, the addition of a bit having a value of 0 is started before the second bit sequence, or the addition of a bit having a value of 0 is started after the second bit sequence, or the addition of a bit having a value of 0 is started at the kth of the second bit sequence j-1 Bit position and kth j The addition of bits with a value of 0 starts between the bit positions. k (k) j-1 Is an integer greater than or equal to 1, e.g. k j-1 Is an integer of m/2.
Alternatively, the bits added to the second bit sequence with a value of 0 may be consecutive or uniform (for example, 0 is added to every other bit, or 0 is added to every 2 bits), and the predetermined may be configured only on the receiving side, only on the transmitting side, or both the receiving side and the transmitting side. Alternatively, both the receiving end and the transmitting end define that the bits added to the second bit sequence, which have a value of 0, are consecutive or uniform.
Alternatively, the correspondence between the first number of bits having a value of 0 and the start position/end position where bits having a value of 0 are added to the second bit sequence (i.e., the correspondence between Padding size and the cut-off position (i.e., the start position/end position where bits having a value of 0 are added to the second bit sequence) may be specified in advance, and may be configured only at the receiving end, or may be configured only at the transmitting end, or may be configured at both the receiving end and the transmitting end. Or the receiving end and the transmitting end may mutually agree on a correspondence between the first number of bits with a value of 0 and a starting position of adding the bit with the value of 0 in the second bit sequence. For example, when the first number is 2 or less, adding a bit having a value of 0 starts before the second bit sequence; when the first number is greater than 2, adding bits with a value of 0 starts after the second bit sequence.
The transmitting end and/or the receiving end can determine the position of the bit with the value of 0 in the third bit sequence based on the number of the bit with the value of 0 and combining with the preset or two-party agreed starting position or ending position of the bit with the value of 0 added in the second bit sequence. Optionally, the transmitting end indicates to the receiving end the position of the bit with the value of 0 in the third bit sequence (for example, the transmitting end sends the third information to the receiving end), or the receiving end indicates to the transmitting end the position of the bit with the value of 0 in the third bit sequence (for example, the receiving end sends the fourth information to the transmitting end).
Alternatively, the transmitting end and/or the receiving end may determine, based on the number of bits with a value of 0, a starting position or an ending position of adding bits with a value of 0 to the second bit sequence in combination with a predetermined or mutually agreed first number of bits with a value of 0 and a starting position/ending position of adding bits with a value of 0 to the second bit sequence, and further determine a position of adding bits with a value of 0 to the third bit sequence. Optionally, the transmitting end indicates to the receiving end the position of the bit with the value of 0 in the third bit sequence (for example, the transmitting end sends the third information to the receiving end), or the receiving end indicates to the transmitting end the position of the bit with the value of 0 in the third bit sequence (for example, the receiving end sends the fourth information to the transmitting end).
How the transmitting end and the receiving end determine the number of bits with a value of 0 is described above, and the description is not repeated. Through the preset regulation or the agreement of the two parties, the interaction times of the two parties can be reduced, and the signaling overhead can be saved.
One or more of the first information, the second information, the third information, the fourth information, etc. may be carried in any of the following: radio resource control (radio resource control, RRC), downlink control information (downlink control information, DCI), data link control cells (MAC control element, MAC-CE). In addition, the signaling used to carry one or more of the first information, the second information, the third information, the fourth information may be periodically triggered, or semi-permanently triggered, or aperiodically triggered.
When the transmitting end is UE, the UE may also transmit, to the receiving end, a modulation order currently supported by the UE (modulation order may also be referred to as a modulation rate) or a modulation order desired by the UE. So that the receiving end can determine the length of the first bit sequence and/or the position of the added bit with a value of 0 based on the capability supported by the UE or the expectation of the UE.
As shown in fig. 3b, the present application provides a communication method, where the execution body of the method may be a receiving end (note that the receiving end may also have a function of sending a signal), for example, a terminal, an access network device, and so on.
The method at least comprises the following steps:
step 305: the receiving end receives the second signal.
Step 306: the receiving end demodulates the second signal to obtain a fourth bit sequence.
In one example, the receiving end inputs the second signal into a second neural network model to obtain a fourth bit sequence. The second neural network model is used for demodulating the input signal, and the length of the fourth bit sequence is equal to the number of neurons of an output layer of the second neural network model.
It can be understood that the third bit sequence is modulated by the transmitting end to obtain a first signal, the first signal is received by the receiving end after being transmitted by the channel, the first signal may be affected by factors such as noise in the channel in the transmission process, the signal received by the receiving end is called a second signal, and the second signal is obtained by interference transformation of the first signal by factors such as noise in the channel. In an ideal case, the second signal is the first signal. It will be appreciated that the fourth bit sequence demodulated from the second signal corresponds to the third bit sequence before modulation, being the same length, and having the same or a smaller difference in bit values. The second neural network model is co-trained with the first neural network model, which ensures that the second neural network model can identify the output of the first neural network model.
The receiving end may receive one or more second signals, and correspondingly, after the demodulation processing in step 306, one or more fourth bit sequences may be obtained.
Step 307: and the receiving end deletes at least one bit in the fourth bit sequence according to the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end, so as to obtain a fifth bit sequence.
The bit sequence carried in the signal sent by the sending end corresponds to the third bit sequence, and the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end is the position of the bit with the value of 0 in the third bit sequence. Under ideal conditions (no interference of the channel), the bit sequence carried in the signal sent by the sending end is the third bit sequence.
The fourth bit sequence corresponds to the third bit sequence (both of length m+s, bit values are the same or differ less), and the fifth bit sequence corresponds to the second bit sequence (both of length m, bit values are the same or differ less). Adding a bit with a value of 0 into the second bit sequence to obtain a third bit sequence; correspondingly, deleting the bit at the corresponding position in the fourth bit sequence can obtain a fifth bit sequence.
Next, several ways of determining the position of the bit with the value 0 in the bit sequence carried in the signal sent by the sender (i.e. the position of the bit with the value 0 in the third bit sequence) in the bit sequence carried in the signal sent by the sender are described.
In one example, the transmitting end determines a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the transmitting end (i.e., a position of a bit with a value of 0 in a third bit sequence) in the bit sequence carried in the signal sent by the transmitting end, and sends the position to the receiving end. For example, the transmitting end sends third information to the receiving end, and the receiving end receives the third information from the transmitting end, where the third information indicates a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the transmitting end in the bit sequence carried in the signal sent by the transmitting end (i.e., a position of a bit with a value of 0 in the third bit sequence).
In one example, the receiving end determines the position of the bit with the value 0 in the bit sequence carried in the signal sent by the sending end (i.e. the position of the bit with the value 0 in the third bit sequence). The receiving end may further send fourth information to the sending end, where the fourth information indicates a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the sending end in the bit sequence carried in the signal sent by the sending end (i.e. a position of a bit with a value of 0 in the third bit sequence).
In one example, the transmitting end and/or the receiving end may determine, based on the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e., the first number of bits with a value of 0 in the third bit sequence), a position of the bit with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e., a position of the bit with a value of 0 in the third bit sequence).
For example, the transmitting end and/or the receiving end may determine, based on the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e., the first number of bits with a value of 0 in the third bit sequence), the position of the bit with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e., the position of the bit with a value of 0 in the third bit sequence) in combination with a predetermined or both-agreed starting position or ending position of the addition of the bit with a value of 0 in the second bit sequence. Optionally, the transmitting end indicates to the receiving end the position of the bit with the value of 0 in the third bit sequence (for example, the transmitting end sends the third information to the receiving end), or the receiving end indicates to the transmitting end the position of the bit with the value of 0 in the third bit sequence (for example, the receiving end sends the fourth information to the transmitting end). The specific details may refer to the description of the transmitting end side, and the detailed description will not be repeated.
For another example, the transmitting end and/or the receiving end may determine, based on the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e., the first number of bits with a value of 0 in the third bit sequence), a correspondence between a predetermined first number of bits with a value of 0 or both sides mutually agreed with a starting position/an ending position of adding bits with a value of 0 in the second bit sequence, and determine a starting position or an ending position of adding bits with a value of 0 in the second bit sequence, and further determine a position of the bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e., a position of the bits with a value of 0 in the third bit sequence). Optionally, the transmitting end indicates to the receiving end the position of the bit with the value of 0 in the third bit sequence (for example, the transmitting end sends the third information to the receiving end), or the receiving end indicates to the transmitting end the position of the bit with the value of 0 in the third bit sequence (for example, the receiving end sends the fourth information to the transmitting end). The specific details may refer to the description of the transmitting end side, and the detailed description will not be repeated.
In one example, the transmitting end may determine the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e., the first number of bits with a value of 0 in the third bit sequence), and send the determined number of bits to the receiving end. For example, the transmitting end sends first information to the receiving end, and correspondingly, the receiving end receives the first information from the transmitting end, where the first information is used to indicate the number of bits with a value of 0 (i.e. the number of bits with a value of 0) in a bit sequence carried in a signal sent by the transmitting end. The receiving end can determine, based on the first information, the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end.
In an example, the transmitting end may determine the length of the original bit sequence to be transmitted (which may be understood as the first length of the first bit sequence) carried in the signal sent by the transmitting end, and send the length of the original bit sequence to the receiving end, so that the receiving end determines the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e. the number of bits with a value of 0 in the third bit sequence). For example, a transmitting end sends first information to a receiving end, and correspondingly, the receiving end receives the first information from the transmitting end, where the first information is used to indicate the length of an original bit sequence to be transmitted, which is carried in a signal sent by the transmitting end. The receiving end can determine the number of bits with the value of 0 in the bit sequence carried in the signal sent by the sending end according to the number of neurons of the output layer of the second neural network model and the first information. And the receiving end further determines the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end according to the number of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end.
In one example, the receiving end may determine the number of bits with a value of 0 in the bit sequence carried in the signal sent by the sending end.
For example, the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end (i.e. the number of bits with a value of 0 in the third bit sequence) may be related to the signal quality parameter, and the receiving end may determine the number based on the signal quality parameter of the first channel, where the first channel is a channel between the transmitting end and the receiving end. Of course, the receiving end may determine the number in other ways, which is not limited in this application. The signal quality parameter of the first channel may be determined by the receiving end itself, or may be sent by the sending end to the receiving end. Optionally, the receiving end may further determine the length of the original bit sequence to be sent, which is carried in the signal sent by the sending end, based on the number of neurons of the output layer of the second neural network model and the number of bits with a value of 0 in the bit sequence carried in the signal sent by the sending end.
For another example, the receiving end may first determine the length of the original bit sequence to be transmitted (which may be understood as the first length of the first bit sequence) carried in the signal sent by the transmitting end, and then determine the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end according to the number of neurons of the output layer of the second neural network model and the length of the original bit sequence to be transmitted (which may be understood as the first length of the first bit sequence) carried in the signal sent by the transmitting end. For example, the difference between the number of output layer neurons and the value of the length of the bit sequence to be originally transmitted is the number of bits having a value of 0. For example, the length of the original bit sequence to be transmitted may be related to a signal quality parameter, and the receiving end may determine the length of the original bit sequence to be transmitted based on the signal quality parameter of the first channel. Of course, the receiving end may determine the length of the original bit sequence to be sent in other manners, which is not limited in this application. The signal quality parameter of the first channel may be determined by the receiving end itself, or may be sent by the sending end to the receiving end.
The signal quality parameters include, but are not limited to, at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. The signal quality parameter may also be determined based on the number and/or frequency of ACK or NACK replies to the peer. For example, a NACK may be recovered always in the case of a decoding error, indicating that the signal quality parameter is not good.
Determining the length of the effective bits (i.e., the first bit sequence) that need to be modulated based on the quality of the channel may improve the quality of the communication.
Optionally, the receiving end may further send second information to the sending end, where the second information indicates the number of bits with a value of 0 in a bit sequence carried in a signal sent by the sending end or the length of an original bit sequence to be sent carried in a signal sent by the sending end.
Step 308: the receiving end adjusts the value of each bit in the fifth bit sequence to be a numerical value in a first set to obtain a sixth bit sequence, wherein the first set comprises a first preset value and 1.
The receiving end adjusts the value of each bit in the fifth bit sequence to be the value closest to the value of the bit to which the rain belongs in the first set, and a sixth bit sequence is obtained. For example, if the absolute value of the difference between the bit and 1 is smaller than the absolute value of the difference between the bit and the first preset value, the bit is adjusted to 1. For example, if the absolute value of the difference between the bit and 1 is greater than the absolute value of the difference between the bit and the first preset value, the bit is adjusted to the first preset value.
In another adjustment method, the receiving end adjusts the value of each bit in the fifth bit sequence to be the value closest to the value of the bit belonging to rain in the first set based on the channel information, so as to obtain a sixth bit sequence. For example, based on the channel information, a probability that the value of each bit in the fifth bit sequence is the value in the first set is determined, and the value in the first set with the highest probability is taken as the value after the bit is adjusted. For example, if the probability that the bit is 1 is smaller than the probability that the bit is a first preset value, the bit is adjusted to be 1. For example, if the probability that the bit is 1 is greater than the probability that the bit is a first preset value, the bit is adjusted to the first preset value.
Step 309: and the receiving end adjusts the bit with the value of the first preset value in the sixth bit sequence to be 0 to obtain a seventh bit sequence.
It is understood that the first bit sequence corresponds to the seventh bit sequence, and the lengths of the first bit sequence and the seventh bit sequence are the same, and are both the first length. In an ideal case, the seventh bit sequence is the first bit sequence.
The first preset value is an integer which is not 0 and not 1. In an alternative example, if the first preset value is-1, the receiving end may adjust the bit with the value-1 in the sixth bit sequence to be 0. Alternatively, the first preset value may be a value of-2, -3, or the like. In addition, the bit having a value of 1 in the sixth bit sequence is still 1.
As shown in fig. 4, for the 6 bits recovered by the receiving end, the bit corresponding to the complementary 0 position is deleted, only the bit corresponding to the original input with length of 2 or the input with length of 4 is reserved, and the reserved bit is recovered to the original valid bit value. Thus, a neural network is used to achieve any variable length input that is less than the maximum length contracted.
The foregoing describes that a first number of bits with a value of 0 added in the second bit sequence, a first length of the first bit sequence, may be related to a signal quality parameter of the first channel; the number of the at least one bit and/or the length of the first bit sequence satisfying the first condition is determined according to the signal quality parameter of the first channel; wherein, the first channel is a channel between the sending end and the receiving end. Signal quality parameters such as signal-to-noise ratio SNR, signal-to-interference plus noise ratio SINR, channel quality indication CQI.
Subsequently, the receiving end may combine the multiple bit sequences obtained based on the processing procedures from step 305 to step 309 to obtain the transmitted bit sequence. The bit sequence obtained based on the processing procedure of steps 305 to 309 is set as a seventh bit sequence. For example, the receiving end may sequentially combine the second signals according to the receiving order corresponding to the seventh bit sequence, to obtain the transmitted bit sequence. For another example, the receiving end may sequentially combine the seventh bit sequences according to the numbering sequence in the second signal, to obtain the transmitted bit sequence.
In another alternative example, the receiving end may first combine one or more sixth bit sequences to obtain a seventh bit sequence. And then mapping each bit in the seventh bit sequence into a numerical value in the first set, and adjusting the bit with the numerical value of the first preset value to 0 to obtain a transmitted bit sequence. That is, the order in which each bit in the bit sequence is mapped to a value in the first set, and the bit sequence is combined may not be limited.
The transmitting end adjusts 0 in an original bit sequence to be transmitted to be a non-0 and non-1 numerical value, takes the number of neurons of an input layer of the first neural network model as a fixed bit input length, supplements effective bits (namely a first bit sequence) to be modulated by 0 to achieve the fixed bit input length, so that the length of the bit sequence input into the first neural network model is fixed, the length of the effective bits (namely a non-0 value) is variable, and a neuron node with the 0 supplemented bit is not activated in the operation process of the neural network and does not influence the processing of other neuron nodes. By the 0-filling operation, a fixed neural network model can be used to handle modulation of significant bits (i.e., non-0 values) of different lengths (i.e., variable lengths). Correspondingly, the receiving end operates the bit sequence in a mode corresponding to the transmitting end, namely, bit positions corresponding to the 0 supplementing position in the fourth bit sequence obtained through demodulation are deleted, and the rest bit positions are restored to original bit values, so that the original bit sequence is obtained.
A detailed example is presented below:
the method comprises the steps that a sending end determines the number of bits with a first length and a value of 0 based on the number of neurons of an input layer of a first neural network model;
the transmitting end adjusts a bit with a value of 0 in a first bit sequence to be transmitted to a first preset value based on a first length to obtain a second bit sequence; adding at least one bit with a value of 0 into the second bit sequence based on the number of neurons of an input layer of the first neural network model to obtain a third bit sequence; wherein the length of the third bit sequence is equal to the number of neurons of the input layer of the first neural network model; inputting the third bit sequence into the first neural network model to obtain a first signal; the first neural network model is used for modulating an input bit sequence; and transmitting the first signal.
Further, the transmitting end transmits third information to the receiving end, wherein the third information indicates the position of the at least one bit with the value of 0 in the third bit sequence.
The receiving end receives a second signal (the second signal is identical to the first signal in ideal condition), and the second signal is input into a second neural network model to obtain a fourth bit sequence; the second neural network model is used for demodulating the input signal, and the value of the length of the fourth bit sequence is equal to the number of neurons of the output layer of the second neural network model;
The receiving end receives the third information; and the receiving end determines the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end according to the third information.
And the receiving end deletes at least one bit in the fourth bit sequence according to the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end, so as to obtain a fifth bit sequence. Then, adjusting the value of each bit in the fifth bit sequence to be a numerical value in a first set to obtain a sixth bit sequence, wherein the first set comprises a first preset value and 1; the bit with the value of the first preset value in the sixth bit sequence is adjusted to be 0, and a seventh bit sequence is obtained; wherein the first preset value is an integer which is not 0 and not 1.
A detailed example is described below:
the method comprises the steps that a sending end determines the number of bits with a first length and a value of 0 based on the number of neurons of an input layer of a first neural network model;
the transmitting end adjusts a bit with a value of 0 in a first bit sequence to be transmitted to a first preset value based on a first length to obtain a second bit sequence; adding at least one bit with a value of 0 into the second bit sequence based on the number of neurons of an input layer of the first neural network model to obtain a third bit sequence; wherein the length of the third bit sequence is equal to the number of neurons of the input layer of the first neural network model; inputting the third bit sequence into the first neural network model to obtain a first signal; the first neural network model is used for modulating an input bit sequence; and transmitting the first signal.
Further, the transmitting end transmits first information to the receiving end, wherein the first information indicates the number of at least one bit with a value of 0 or the length of the first bit sequence.
The receiving end receives a second signal (the second signal is identical to the first signal in ideal condition), and the second signal is input into a second neural network model to obtain a fourth bit sequence; the second neural network model is used for demodulating the input signal, and the value of the length of the fourth bit sequence is equal to the number of neurons of the output layer of the second neural network model;
the receiving terminal receives first information; and the receiving end determines the position of a bit with the value of 0 in a bit sequence carried in a signal sent by the sending end in the bit sequence carried in the signal sent by the sending end according to the first information. For example, based on the first information and a predetermined mapping relationship, determining the position of a bit with a value of 0 in a bit sequence carried in a signal sent by a sending end in the bit sequence carried in the signal sent by the sending end.
And the receiving end deletes at least one bit in the fourth bit sequence according to the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end, so as to obtain a fifth bit sequence. Then, adjusting the value of each bit in the fifth bit sequence to be a numerical value in a first set to obtain a sixth bit sequence, wherein the first set comprises a first preset value and 1; the bit with the value of the first preset value in the sixth bit sequence is adjusted to be 0, and a seventh bit sequence is obtained; wherein the first preset value is an integer which is not 0 and not 1.
Other examples may be described with reference to the foregoing and will not be described in detail.
Example 2:
as shown in fig. 5a, a schematic communication flow diagram is provided, and the execution body of the process may be a transmitting end (note that the transmitting end may also have a function of receiving signals), such as a terminal, an access network device, and the like.
At least comprises the following steps:
step 501: the transmitting end selects a first sub-neural network model from the third neural network models.
The number of neurons of the input layer of the first sub-neural network model is the same as the value of the first length, and the third neural network model is used for modulating the input bit sequence. The first sub-neural network model is part or all of the third neural network model.
The third neural network model may be a neural network model of Nested Nested structure. The third neural network model may be understood as the highest order modulation network, and may be determined based on channel prior knowledge. The third neural network model (high order modulation network) has a nested relationship with the first sub-neural network model (low order modulation network). As shown in fig. 5c, a lower order modulation network may be obtained by pruning a portion of the higher order modulation network. The low order modulation networks are a subset of the high order modulation networks. The transmitting end can select a proper low-order modulation network according to the nesting relation between the high-order modulation network and the low-order modulation network so as to realize modulation.
Step 502: and the transmitting end inputs the first bit sequence with the first length into the first sub-neural network model to obtain a third signal.
The transmitting end can split the bit sequence to be transmitted into a plurality of bit sequences, and select a first bit sequence from the plurality of bit sequences; alternatively, the transmitting end may split the bit sequence to be transmitted into a plurality of first bit sequences. The length of the first bit sequence is set to a first length. The first length is set to m, m being an integer greater than or equal to 1. The length of the bit sequence to be transmitted is set to be n, n is an integer greater than or equal to 1, and n can be an integer multiple of m, and can be a non-integer multiple of m. The transmitting end can divide the n bits into a plurality of groups according to one group of m bits in turn, and each group is a first bit sequence with a first length. For example, n is 128 and m is 4, then it can be divided into 32 groups; for example, n is 120 and m is 6, and can be divided into 20 groups.
The length of the first bit sequence may be an even power of 2, or an odd power of 2, or any power of any number. Compared with a modulation mode which can only process the even power of 2, the dynamic adjustable range is enlarged. The bit sequence to be modulated can be cut off arbitrarily according to the channel condition, for example, the maximum length is 8 bits, the bit sequence can be cut off into any length of 1bit to 8 bits, the more flexible actual modulation order is supported, and the constellation point can be dynamically variable through the scheme.
Step 503: and the transmitting end transmits the third signal.
The sub-neural network model is obtained by dynamically selecting the neural network model to match the length of the effective bit sequence, and the modulation of the effective bit sequences with different lengths (namely, different lengths) can be processed by adopting one fixed neural network model, so that the method is simple and flexible.
In one example, the transmitting end may determine the length of the bit sequence that needs to be modulated, i.e., the first length.
For example, the transmitting end sends fifth information to the receiving end, and correspondingly, the receiving end receives the fifth information from the transmitting end, where the fifth information is used to indicate the first length of the first bit sequence. The fifth information may include a specific value of the first length or other information capable of identifying the first length. The transmitting end determines the length of the bit sequence to be modulated and informs the receiving end of the length, so that the receiving end can accurately select a corresponding sub-neural network model from the original neural network models for demodulation to demodulate.
For example, the first length may be related to a signal quality parameter, and the transmitting end may determine the first length based on the signal quality parameter of a first channel, which is a channel between the transmitting end and the receiving end. Of course, the transmitting end may determine the first length in other manners, which is not limited in this application. The signal quality parameter of the first channel may be determined by the transmitting end itself, or may be transmitted to the transmitting end by the receiving end.
If the signal quality parameter of the first channel is sent by the receiving end to the sending end, the sending end may also receive the signal quality parameter of the first channel from the receiving end before sending the first information to the receiving end. That is, before the receiving end receives the first information, the receiving end may also send the signal quality parameter of the first channel to the transmitting end.
The signal quality parameters include, but are not limited to, at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. The signal quality parameter may also be determined based on the number and/or frequency of ACK or NACK replies to the peer. For example, a NACK may be recovered always in the case of a decoding error, indicating that the signal quality parameter is not good.
Determining the length of the bit sequence to be modulated based on the quality of the channel may improve the quality of the communication.
In another example, the receiving end may determine the length of the bit sequence that needs to be modulated, i.e., the first length.
For example, the receiving end sends sixth information to the sending end, and correspondingly, the sending end receives the sixth information from the receiving end, where the sixth information is used to indicate the first length of the first bit sequence. The sixth information may include a specific value of the first length or other information capable of identifying the first length. The receiving end determines the length of the bit sequence to be modulated and informs the transmitting end of the length, so that the transmitting end can determine the bit sequence to be modulated based on the length and execute a first sub-neural network model of the modulation operation.
The receiving end may determine the first length based on the signal quality parameter of the first channel, or may determine the first length by other manners, which is not limited in this application. For the receiving end, the signal quality parameter of the first channel may be determined by the receiving end itself, or may be sent by the sending end to the receiving end.
In an alternative example, the transmitting end may further transmit the signal quality parameter of the first channel to the receiving end before receiving the sixth information. So that the receiving end determines the sixth information based on the signal quality parameter of the first channel. For example, the transmitting end may determine the signal quality parameter, but does not support determining the first length based on the signal quality parameter (e.g., the mapping relationship between the value of the first length and the signal quality parameter is not configured); the receiving end cannot determine the signal quality parameter, but may support determining the first length based on the signal quality parameter. The transmitting end can transmit the signal quality parameter to the receiving end, and the receiving end determines the first length based on the signal quality parameter and transmits the first length to the transmitting end.
In another example, both the transmitting end and the receiving end support determining the first length based on the signal quality parameter in the same manner (e.g. the same mapping relation (i.e. the mapping relation between the value of the first length and the signal quality parameter) is configured), but only the receiving end supports measuring one or some signal quality parameter, and the transmitting end does not support measuring one or some signal quality parameter. The receiving end may determine the first length based on the mapping relationship and the determined signal quality parameter of the first channel. The receiving end may also send the signal quality parameter of the first channel to the transmitting end, so that the transmitting end determines the first length based on the mapping relationship and the received signal quality parameter of the first channel.
In another example, both the transmitting end and the receiving end support determining the first length based on the signal quality parameter in the same manner (e.g. the same mapping relation (i.e. the mapping relation between the value of the first length and the signal quality parameter) is configured), but only the transmitting end supports measuring one or some signal quality parameter, and the receiving end does not support measuring one or some signal quality parameter. The transmitting end may determine the first length based on the mapping relationship and the determined signal quality parameter of the first channel. The transmitting end may also transmit the signal quality parameter of the first channel to the receiving end, so that the receiving end determines the first length based on the mapping relationship and the received signal quality parameter of the first channel.
In yet another example, both the transmitting end and the receiving end support determining the first length based on the signal quality parameter in the same manner (e.g., the same mapping relation (i.e., the mapping relation between the value of the first length and the signal quality parameter) is configured), and both support measuring one or some signal quality parameter (e.g., the signal quality parameter has uplink and downlink reciprocity, such as SNR). In this way, the receiving end and the transmitting end can determine the first length based on the mapping relationship and the signal quality parameter of the first channel determined by themselves. The process does not need to interact, and signaling can be saved.
In one example, a sender may determine to determine a location of an input layer neuron of a sub-neural network model in an input layer neuron of an original neural network model for modulation. For example, the transmitting end sends seventh information to the receiving end, and correspondingly, the receiving end receives the seventh information from the transmitting end, where the seventh information is used to indicate a position of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model. The transmitting end determines the position of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model, and informs the receiving end of the position information, so that the receiving end can accurately select a corresponding sub-neural network model from the original neural network models for demodulation based on the position to demodulate.
In another example, the receiving end may determine to determine the location of the input layer neurons of the sub-neural network model in the input layer neurons of the original neural network model for modulation. For example, the receiving end sends eighth information to the sending end, and correspondingly, the sending end receives the eighth information from the receiving end, where the eighth information is used to indicate a position of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model. And the transmitting end can select the first sub-neural network model from the third neural network model according to the eighth information. The receiving end determines the position of the input layer neuron of the sub-neural network model in the input layer neuron of the original neural network model for modulation, and notifies the transmitting end of the position, so that the transmitting end can accurately select the corresponding sub-neural network model from the original neural network model for modulation based on the position to perform modulation.
The seventh information or eighth information may include, but is not limited to, one or more of the following: bitmap for representing a specific location of an input layer neuron of a first sub-neural network model in an input layer neuron of the third neural network model, a specific location of an input layer neuron of a first sub-neural network model in an input layer neuron of the third neural network model (e.g., kth j Position, kth j+s A location), a starting location of an input layer neuron of a first sub-neural network model in an input layer neuron of the third neural network model, a terminating location of an input layer neuron of a first sub-neural network model in an input layer neuron of the third neural network model, a persistent location of an input layer neuron of a first sub-neural network model in an input layer neuron of the third neural network model (e.g., kth j To the kth j+s A location), the number of input layer neurons of the first sub-neural network model, etc.
An example of the locations of the input layer neurons of several first sub-neural network models in the input layer neurons of the third neural network model is presented below.
The input layer neurons of the first sub-neural network model are front neurons, or rear neurons, or intermediate neurons, of the input layer neurons of the third neural network model.
The positions of the input layer neurons of the first sub-neural network model in the input layer neurons of the third neural network model may or may not be continuous, may or may not be uniform, and the positions of the input layer neurons of the first sub-neural network model in the input layer neurons of the third neural network model are not limited in the present application.
In one example, the sender and/or receiver may determine a location of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model based on the first length.
For example, the starting position or the ending position of the selected neuron in the original neural network model for modulation (i.e., the third neural network model) may be specified in advance, and the specification may be configured only on the receiving side, only on the transmitting side, or both the receiving side and the transmitting side. Alternatively, both the receiving end and the transmitting end define a start position or an end position of a selected neuron in the original neural network model (i.e., the third neural network model) for modulation.
For example, the neuron is selected starting at the position of the 1 st neuron in the third neural network model, i.e. the starting position of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model is the position of the 1 st neuron in the third neural network model.
Alternatively, the positions of the input layer neurons of the first sub-neural network model in the input layer neurons of the third neural network model may be continuous or uniform, and the pre-specification may be configured only on the receiving side, only on the transmitting side, or both the receiving side and the transmitting side. Alternatively, the positions of the input layer neurons defined in the first sub-neural network model in the input layer neurons of the third neural network model are continuous or uniform at both the receiving end and the transmitting end.
Alternatively, the correspondence between the first length (the number of neurons included in the first sub-neural network model) and the starting position or the ending position of the selected neuron in the original neural network model for modulation (i.e., the third neural network model) may be specified in advance, and the specification may be configured only on the receiving side, may be configured only on the transmitting side, or may be configured on both the receiving side and the transmitting side. Or the receiving end and the transmitting end may mutually agree on a correspondence between the first length (the number of neurons included in the first sub-neural network model) and the starting position or the ending position of the selected neuron in the original neural network model (i.e., the third neural network model) for modulation. For example, when the value of the first length is less than or equal to 2, selecting the neuron from the 1 st neuron position in the third neural network model; when the first number is greater than 2, the selection of neurons begins at the location of the 5 th neuron in the third neural network model.
The transmitting end and/or the receiving end may determine, based on the first length, a position of an input layer neuron of the first sub-neural network model in an input layer neuron of a third neural network model in combination with a predetermined or both-agreed starting position or ending position of a neuron selected in the original neural network model for modulation (i.e., the third neural network model). Optionally, the transmitting end indicates to the receiving end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the transmitting end sends seventh information to the receiving end), or the receiving end indicates to the transmitting end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the receiving end sends eighth information to the transmitting end).
Alternatively, the transmitting end and/or the receiving end may determine, based on the first length, a starting position or an ending position of the neuron selected in the original neural network model (i.e., the third neural network model) for modulation, in combination with a predetermined, or mutually agreed, first length and a corresponding relation between the starting position or the ending position of the neuron selected in the original neural network model (i.e., the third neural network model) for modulation, thereby determining a position of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model. Optionally, the transmitting end indicates to the receiving end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the transmitting end sends seventh information to the receiving end), or the receiving end indicates to the transmitting end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the receiving end sends eighth information to the transmitting end).
How the transmitting end and the receiving end determine the first length is described above, and a detailed description is not repeated. Through the preset regulation or the agreement of the two parties, the interaction times of the two parties can be reduced, and the signaling overhead can be saved.
One or more of the fifth information, sixth information, seventh information, eighth information, etc. may be carried in any of the following: radio resource control, RRC, downlink control information, DCI, data link control cells, MAC-CE. In addition, the signaling for carrying one or more of the fifth information, the sixth information, the seventh information, the eighth information may be periodically triggered, or semi-permanently triggered, or aperiodically triggered.
When the transmitting end is UE, the UE may also transmit to the receiving end, a modulation order (modulation order may also be referred to as a modulation rate) currently supported by the UE or a modulation order desired by the UE. So that the receiving end can determine the position of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model based on the capability supported by the UE, or the expectation of the UE.
As shown in fig. 5b, a schematic communication flow diagram is provided, and the execution body of the process may be a receiving end (note that the receiving end may also have a function of sending a signal), such as a terminal, an access network device, and the like.
At least comprises the following steps:
step 504: the receiving end receives the fourth signal.
It can be understood that the fourth signal is received by the receiving end after being transmitted by the channel, and the fourth signal may be affected by factors such as noise in the channel during the transmission process, and the signal received by the receiving end is called a fourth signal, and the fourth signal is obtained by interference transformation of the third signal by factors such as noise in the channel. In an ideal case, the fourth signal is the third signal.
Step 505: the receiving end selects a second sub-neural network model from the fourth neural network model, wherein the number of neurons of an output layer of the second sub-neural network model is the same as the value of the first length; the fourth neural network model is configured to demodulate an input signal, where the first length is a length of a bit sequence carried in a signal sent by the sending end.
The bit sequence carried in the signal sent by the sending end can be understood as a first bit sequence, and the first length is the length of the first bit sequence.
The fourth neural network model is co-trained with the third neural network model, and the fourth neural network model can identify an output of the third neural network model.
The fourth neural network model may be a neural network model of Nested Nested structure. The fourth neural network model may be understood as the highest order demodulation network, and may be determined based on channel prior knowledge. The fourth neural network model (higher order demodulation network) has a nested relationship with the second sub-neural network model (lower order demodulation network). As shown in fig. 5c, a low-order demodulation network can be obtained by deleting a part of the network of the high-order demodulation network. The low order demodulation networks are a subset of the high order demodulation networks. The receiving end can select a proper low-order demodulation network according to the nesting relation of the high-order demodulation network and the low-order demodulation network so as to realize demodulation.
Step 506: and the receiving end inputs the fourth signal into the second sub-neural network model to obtain a second bit sequence with the first length.
The second bit sequence corresponds to the first bit sequence, and ideally, the second bit sequence is the first bit sequence.
It will be appreciated that the location of the output layer neurons of the second sub-neural network model in the output layer neurons of the fourth neural network model is depicted as the same location as the location of the input layer neurons of the first sub-neural network model in the input layer neurons of the third neural network model.
In one example, the transmitting end may determine a location of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model and transmit the location to the receiving end. For example, the transmitting end transmits seventh information to the receiving end, and the receiving end receives the seventh information from the transmitting end; the seventh information is used for indicating the position of the input layer neuron of the sub-neural network model for signal modulation in the transmitting end in the original neural network model (i.e. the position of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model). The receiving end can determine the position of the output layer neuron of the second sub-neural network model in the output layer neuron of the fourth neural network model according to the fifth information, and then select the second sub-neural network model in the fourth neural network model. The positions of the output layer neurons of the second sub-neural network model in the output layer neurons of the fourth neural network model are the same as the positions of the input layer neurons of the first sub-neural network model in the input layer neurons of the third neural network model. The transmitting end determines the position of the input layer neuron of the sub-neural network model in the input layer neuron of the original neural network model for modulation and notifies the receiving end of the position, so that the receiving end can accurately select the corresponding sub-neural network model from the original neural network model for demodulation based on the position for demodulation.
In one example, the receiving end determines a location of an input layer neuron of a sub-neural network model in which the signal is modulated in the transmitting end in the original neural network model (i.e., a location of an input layer neuron of a first sub-neural network model in an input layer neuron of a third neural network model), and then selects a second sub-neural network model in the fourth neural network model according to the location.
In one example, the transmitting end and/or the receiving end may determine a location of an input layer neuron of a sub-neural network model in which the signal is modulated in the transmitting end in the original neural network model (i.e., a location of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model) based on a length of a bit sequence (i.e., a first length of the first bit sequence) carried in a signal transmitted by the transmitting end.
For example, the transmitting end and/or the receiving end may determine, based on the first length, a position of an input layer neuron of a sub-neural network model in which signal modulation is performed in the transmitting end in the original neural network model (i.e., a position of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model) in combination with a predetermined or both-agreed starting position or ending position of a neuron selected in the original neural network model for modulation (i.e., the third neural network model). Optionally, the transmitting end indicates to the receiving end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the transmitting end sends seventh information to the receiving end), or the receiving end indicates to the transmitting end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the receiving end sends eighth information to the transmitting end).
For another example, the transmitting end and/or the receiving end may determine, based on the first length and in combination with a predetermined or mutually agreed first length and a correspondence between a start position or an end position of a neuron selected in the original neural network model (i.e., the third neural network model) for modulation, a start position or an end position of the neuron selected in the original neural network model (i.e., the third neural network model) for modulation, and further determine a position of an input layer neuron of the sub-neural network model for signal modulation in the transmitting end in the original neural network model (i.e., a position of an input layer neuron of the first sub-neural network model in an input layer neuron of the third neural network model). Optionally, the transmitting end indicates to the receiving end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the transmitting end sends seventh information to the receiving end), or the receiving end indicates to the transmitting end the location of the input layer neuron of the first sub-neural network model in the input layer neuron of the third neural network model (e.g., the receiving end sends eighth information to the transmitting end).
In one example, the transmitting end may determine the length of the bit sequence (which may be understood as the first length of the first bit sequence) carried in the signal transmitted by the transmitting end, and transmit it to the receiving end. For example, the transmitting end sends fifth information to the receiving end, and correspondingly, the receiving end receives the fifth information from the transmitting end, where the fifth information is used to indicate the length of the bit sequence carried in the signal sent by the transmitting end.
In one example, the receiving end may determine the length of a bit sequence carried in a signal transmitted by the transmitting end.
For example, the length of the bit sequence carried in the signal sent by the transmitting end may be related to the signal quality parameter, and the receiving end may determine the length of the bit sequence carried in the signal sent by the transmitting end based on the signal quality parameter of the first channel. Of course, the receiving end may determine the length of the bit sequence carried in the signal sent by the sending end in other manners, which is not limited in this application. The signal quality parameter of the first channel may be determined by the receiving end itself, or may be sent by the sending end to the receiving end.
The signal quality parameters include, but are not limited to, at least one of: signal to noise ratio SNR, signal to interference plus noise ratio SINR, channel quality indicator CQI. The signal quality parameter may also be determined based on the number and/or frequency of ACK or NACK replies to the peer. For example, a NACK may be recovered always in the case of a decoding error, indicating that the signal quality parameter is not good.
Determining the length of the effective bits (i.e., the first bit sequence) that need to be modulated based on the quality of the channel may improve the quality of the communication.
Optionally, the receiving end may further send sixth information to the sending end, where the sixth information indicates a length of a bit sequence carried in a signal sent by the sending end.
Subsequently, the receiving end may combine the multiple bit sequences obtained based on the processing procedures from step 504 to step 506 to obtain the transmitted bit sequence. The bit sequence obtained based on the processing procedure of steps 504 to 506 is set to be referred to as a second bit sequence. For example, the receiving end may sequentially combine the second signals according to the receiving order of the second bit sequence, to obtain the transmitted bit sequence. For another example, the receiving end may sequentially combine the second bit sequences according to the numbering sequence in the second signal, to obtain the transmitted bit sequence.
The sub-neural network model is obtained by dynamically selecting the neural network model to match the length of the effective bit sequence, and the demodulation of the effective bit sequences with different lengths (namely, variable lengths) can be processed by adopting a fixed neural network model, so that the demodulation method is simple and flexible.
The foregoing describes the method of embodiments of the present application, and the apparatus of embodiments of the present application will be described hereinafter. The method and the device are based on the same technical conception, and because the principles of solving the problems by the method and the device are similar, the implementation of the device and the method can be mutually referred to, and the repeated parts are not repeated.
The embodiment of the present application may divide the functional modules of the apparatus according to the above method example, for example, each function may be divided into each functional module, or two or more functions may be integrated into one module. These modules may be implemented in hardware or in software functional modules. It should be noted that, in the embodiment of the present application, the division of the modules is schematic, which is merely a logic function division, and other division manners may be used in specific implementation.
Based on the same technical concept as the above method, referring to fig. 6, there is provided a schematic structural diagram of a communication apparatus 600, and the communication apparatus 600 may include: the processing module 610, optionally, further includes a receiving module 620a, a sending module 620b, and a storage module 630. The processing module 610 may be connected to the storage module 630 and the receiving module 620a and the transmitting module 620b, respectively, and the storage module 630 may also be connected to the receiving module 620a and the transmitting module 620 b.
In one example, the receiving module 620a and the transmitting module 620b may be integrated together, and defined as a transceiver module.
In one example, the communication apparatus 600 may be a transmitting end, or may be a chip or a functional unit applied in the transmitting end. The communication device 600 has any function of the transmitting end in the above-described method, for example, the communication device 600 can perform each step performed by the transmitting end in the above-described methods of fig. 3a, 4, and 5 a.
The receiving module 620a may perform the receiving action performed by the transmitting end in the above method embodiment.
The sending module 620b may perform the sending action performed by the sending end in the above method embodiment.
The processing module 610 may perform actions other than the sending action and the receiving action in the actions performed by the sending end in the above method embodiment.
In an example, the processing module 610 is configured to adjust a bit with a value of 0 in the first bit sequence to be transmitted to a first preset value, so as to obtain a second bit sequence; wherein the first preset value is an integer which is not 0 and not 1; adding at least one bit with a value of 0 into the second bit sequence based on the number of neurons of an input layer of the first neural network model to obtain a third bit sequence; wherein the length of the third bit sequence is equal to the number of neurons of the input layer of the first neural network model; inputting the third bit sequence into the first neural network model to obtain a first signal; the first neural network model is used for modulating an input bit sequence; the sending module 620b is configured to send the first signal.
In one example, the sending module 620b is further configured to send first information to a receiving end, where the first information indicates the number of bits with a value of 0 or the length of the first bit sequence.
In one example, the processing module 610 is further configured to determine the number of bits with the at least one value of 0 based on a signal quality parameter of the first channel; and determining the length of the first bit sequence based on the number of neurons of the input layer of the first neural network model and the number of the at least one bit with a value of 0, wherein the first channel is a channel between the transmitting end and the receiving end; or determining the length of the first bit sequence based on the signal quality parameter of the first channel; and determining the number of the at least one bit with a value of 0 based on the number of neurons of the input layer of the first neural network model and the length of the first bit sequence, wherein the first channel is a channel between the transmitting end and the receiving end.
In one example, the processing module 610 is specifically configured to receive second information from the receiving end, where the second information indicates the number of bits with the value of 0; and adding at least one bit with a value of 0 into the second bit sequence according to the second information to obtain a third bit sequence.
In one example, before adding at least one bit with a value of 0 to the second bit sequence according to the second information to obtain a third bit sequence, the processing module 610 is further configured to determine, according to the number of neurons of the input layer of the first neural network model and the second information, a length of the first bit sequence; and determining the first bit sequence according to the length of the first bit sequence.
In one example, the processing module 610 is specifically configured to receive second information from a receiving end, where the second information indicates a length of the first bit sequence; determining the number of the at least one bit with the value of 0 according to the number of the neurons of the input layer of the first neural network model and the second information; and adding at least one bit with the value of 0 into the second bit sequence based on the number of the at least one bit with the value of 0, so as to obtain a third bit sequence.
In one example, before adding at least one bit with a value of 0 to the second bit sequence based on the number of the at least one bit with a value of 0 to obtain a third bit sequence, the processing module 610 is further configured to determine the first bit sequence according to the second information.
In one example, the sending module 620b is further configured to send third information to the receiving end, where the third information indicates a position of the at least one bit with a value of 0 in the third bit sequence.
In one example, the processing module 610 is specifically configured to receive fourth information from the receiving end, where the fourth information indicates a position of the at least one bit with a value of 0 in the third bit sequence; and according to the fourth information, adding at least one bit with a value of 0 into the second bit sequence to obtain a third bit sequence.
In one example, the first preset value is-1.
In one example, the storage module 630 may store computer-executable instructions of the method performed by the sender, so that the processing module 610 and the receiving module 620a and the sending module 620b perform the method performed by the sender in the above example.
By way of example, a memory module may include one or more memories, which may be one or more devices, circuits, or means for storing programs or data. The memory module may be a register, a cache, a RAM, etc., and may be integrated with the processing module. The memory module may be a ROM or other type of static storage device that may store static information and instructions, and may be independent of the processing module.
The transceiver module may be an input or output interface, a pin or circuit, etc.
In one example, the communication device 600 may be a receiving end, or may be a chip or a functional unit applied in the receiving end. The communication device 600 has any function of the receiving end in the above method, for example, the communication device 600 can perform the steps performed by the receiving end in the above methods of fig. 3b, fig. 4, fig. 5 b.
The receiving module 620a may perform the receiving action performed by the receiving end in the method embodiment.
The sending module 620b may perform the sending action performed by the receiving end in the method embodiment.
The processing module 610 may perform actions other than the sending action and the receiving action in the actions performed by the receiving end in the above method embodiments.
In one example, the receiving module 620a is configured to receive a second signal; the processing module 610 is configured to input the second signal into a second neural network model to obtain a fourth bit sequence; the second neural network model is used for demodulating the input signal, and the value of the length of the fourth bit sequence is equal to the number of neurons of the output layer of the second neural network model; deleting at least one bit in the fourth bit sequence according to the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end, so as to obtain a fifth bit sequence; adjusting the value of each bit in the fifth bit sequence to be a numerical value in a first set to obtain a sixth bit sequence, wherein the first set comprises a first preset value and 1; the bit with the value of the first preset value in the sixth bit sequence is adjusted to be 0, and a seventh bit sequence is obtained; wherein the first preset value is an integer which is not 0 and not 1.
In one example, the receiving module 620a is further configured to receive first information from the transmitting end, where the first information indicates a number of bits with a value of 0 in a bit sequence carried in a signal sent by the transmitting end; the processing module 610 is specifically configured to determine, based on the first information, a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the sending end in the bit sequence carried in the signal sent by the sending end.
In one example, the receiving module 620a is further configured to receive first information from the transmitting end, where the first information indicates a length of an original bit sequence to be transmitted, where the length is carried in a signal sent by the transmitting end; the processing module 610 is specifically configured to determine, according to the number of neurons of an output layer of the second neural network model and the first information, the number of bits with a value of 0 in a bit sequence carried in a signal sent by the sending end; and determining the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end according to the number of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end.
In an example, the sending module 620b is further configured to send second information to the sending end, where the second information indicates the number of bits with a value of 0 in a bit sequence carried in a signal sent by the sending end or the length of an original bit sequence to be sent carried in a signal sent by the sending end.
In an example, the processing module 610 is further configured to determine, based on a signal quality parameter of a first channel, a number of bits with a value of 0 in a bit sequence carried in a signal sent by the sending end, and determine, based on a number of neurons of an output layer of the second neural network model and a number of bits with a value of 0 in a bit sequence carried in a signal sent by the sending end, a length of an original bit sequence to be sent carried in a signal sent by the sending end, where the first channel is a channel between the sending end and the receiving end; or determining the length of an original bit sequence to be transmitted, which is carried in a signal sent by a transmitting end, based on a signal quality parameter of a first channel, and determining the number of bits with a value of 0 in the bit sequence carried in the signal sent by the transmitting end, which is a channel between the transmitting end and the receiving end, based on the number of neurons of an output layer of the second neural network model and the length of the original bit sequence to be transmitted, which is carried in the signal sent by the transmitting end.
In an example, the receiving module 620a is further configured to receive third information from the transmitting end, where the third information indicates a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the transmitting end in the bit sequence carried in the signal sent by the transmitting end.
In an example, the sending module 620b is further configured to send fourth information to the sending end, where the fourth information indicates a position of a bit with a value of 0 in a bit sequence carried in a signal sent by the sending end in the bit sequence carried in the signal sent by the sending end.
In one example, the first preset value is-1.
In one example, the storage module 630 may store computer-executable instructions of a method performed by the receiving end, so that the processing module 610 and the receiving module 620a and the sending module 620b perform the method performed by the receiving end in the above example.
By way of example, a memory module may include one or more memories, which may be one or more devices, circuits, or means for storing programs or data. The memory module may be a register, a cache, a RAM, etc., and may be integrated with the processing module. The memory module may be a ROM or other type of static storage device that may store static information and instructions, and may be independent of the processing module.
The transceiver module may be an input or output interface, a pin or circuit, etc.
As one possible product form, the apparatus may be implemented by a general bus architecture.
As shown in fig. 7, a schematic block diagram of a communication device 700 is provided.
The communication device 700 may include: processor 710, optionally, also includes a transceiver 720, a memory 730. The transceiver 720 may be configured to receive a program or instructions and transmit the program or instructions to the processor 710, or the transceiver 720 may be configured to interact with other communication devices, such as interaction control signaling and/or traffic data, etc., by the communication apparatus 700. The transceiver 720 may be a code and/or data read-write transceiver, or the transceiver 720 may be a signal transmission transceiver between a processor and a transceiver. The processor 710 is electrically coupled to the memory 730.
In one example, the communication device 700 may be a transmitting end or a chip applied to the transmitting end. It should be appreciated that the apparatus has any of the functions of the transmitting end in the above method, for example, the communication apparatus 700 can perform the steps performed by the transmitting end in the above methods of fig. 3a, 4 and 5 a. By way of example, the memory 730 is used to store a computer program; the processor 710 may be configured to invoke a computer program or instructions stored in the memory 730 to perform the method performed by the sender in the above example or to perform the method performed by the sender in the above example through the transceiver 720.
In one example, the communication device 700 may be a receiving end or a chip applied to the receiving end. It should be appreciated that the apparatus has any of the functions of the receiving end in the above method, for example, the communication apparatus 700 can perform the steps performed by the receiving end in the above methods of fig. 3b, fig. 4, fig. 5 b. By way of example, the memory 730 is used to store a computer program; the processor 710 may be configured to invoke a computer program or instructions stored in the memory 730 to perform the method performed by the receiving end in the above example or to perform the method performed by the receiving end in the above example through the transceiver 720.
The processing module 610 in fig. 6 may be implemented by the processor 710.
The receiving module 620a and the transmitting module 620b in fig. 6 may be implemented through the transceiver 720. Alternatively, the transceiver 720 is divided into a receiver that performs the function of a receiving module and a transmitter that performs the function of a transmitting module.
The storage module 630 in fig. 6 may be implemented by the memory 730.
As one possible product form, the apparatus may be implemented by a general-purpose processor (a general-purpose processor may also be referred to as a chip or a system-on-chip).
In a possible implementation manner, a general purpose processor implementing a device applied to a transmitting end or a device of a receiving end includes: processing circuitry (processing circuitry may also be referred to as a processor); optionally, the method further comprises: an input/output interface in communication with the processing circuit, and a storage medium (the storage medium may also be referred to as a memory) for storing instructions executed by the processing circuit to perform the method performed by the transmitting end or the receiving end in the above example.
The processing module 610 in fig. 6 may be implemented by a processing circuit.
The receiving module 620a and the transmitting module 620b in fig. 6 may be implemented through input-output interfaces. Or the input/output interface is divided into an input interface and an output interface, the input interface performs the function of the receiving module, and the output interface performs the function of the transmitting module.
The storage module 630 in fig. 6 may be implemented by a storage medium.
As one possible product form, the apparatus of the embodiments of the present application may also be implemented using: one or more FPGAs (field programmable gate arrays), PLDs (programmable logic devices), controllers, state machines, gate logic, discrete hardware components, any other suitable circuitry, or any combination of circuitry capable of performing the various functions described throughout this application.
The embodiment of the application also provides a computer readable storage medium, which stores a computer program, and the computer program can enable the computer to execute the communication method when being executed by the computer. Or the following: the computer program comprises instructions for implementing the above-described communication method.
The embodiment of the application also provides a computer program product, which comprises: computer program code for enabling a computer to carry out the communication method provided above when said computer program code is run on the computer.
The embodiment of the application also provides a communication system, which comprises: a transmitting end and a receiving end for executing the communication method.
In addition, the processor mentioned in the embodiments of the present application may be a central processor (central processing unit, CPU), a baseband processor, and a CPU may be integrated together or separated, or may be a network processor (network processor, NP) or a combination of a CPU and an NP. The processor may further comprise a hardware chip or other general purpose processor. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (programmable logic device, PLD), or a combination thereof. The PLD may be a complex programmable logic device (complex programmable logic device, CPLD), a field-programmable gate array field-programmable gate array (FPGA), general-purpose array logic (generic array logic, GAL), and other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like, or any combination thereof. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory referred to in the embodiments of the present application may be volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory. The nonvolatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable EPROM (EEPROM), or a flash Memory. The volatile memory may be random access memory (Random Access Memory, RAM) which acts as an external cache. By way of example, and not limitation, many forms of RAM are available, such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (Double Data Rate SDRAM), enhanced SDRAM (ESDRAM), synchronous DRAM (SLDRAM), and Direct RAM (DR RAM). It should be noted that the memory described herein is intended to comprise, without being limited to, these and any other suitable types of memory.
The transceiver mentioned in the embodiments of the present application may include a separate transmitter and/or a separate receiver, or the transmitter and the receiver may be integrated. The transceiver may operate under the direction of a corresponding processor. Alternatively, the transmitter may correspond to a transmitter in a physical device and the receiver may correspond to a receiver in the physical device.
Those of ordinary skill in the art will appreciate that the various method steps and elements described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the steps and components of the various embodiments have been described generally in terms of functionality in the foregoing description to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Those of ordinary skill in the art may implement the described functionality using different approaches for each particular application, but such implementation is not to be considered as beyond the scope of the present application.
In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. In addition, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices, or elements, or may be an electrical, mechanical, or other form of connection.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purposes of the embodiments of the present application.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application is essentially or a part contributing to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In the present application, "and/or" describing the association relationship of the association object, it means that there may be three relationships, for example, a and/or B may mean: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship. The term "plurality" as used herein refers to two or more. In addition, it should be understood that in the description of this application, the words "first," "second," and the like are used merely for distinguishing between the descriptions and not for indicating or implying any relative importance or order.
While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various modifications and variations can be made to the embodiments of the present application without departing from the spirit and scope of the embodiments of the present application. Thus, if such modifications and variations of the embodiments of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to include such modifications and variations.

Claims (25)

1. A communication method applied to a transmitting end, comprising:
adjusting a bit with a value of 0 in a first bit sequence to be transmitted to a first preset value to obtain a second bit sequence; wherein the first preset value is an integer which is not 0 and not 1;
adding at least one bit with a value of 0 into the second bit sequence based on the number of neurons of an input layer of the first neural network model to obtain a third bit sequence; wherein the length of the third bit sequence is equal to the number of neurons of the input layer of the first neural network model;
inputting the third bit sequence into the first neural network model to obtain a first signal; the first neural network model is used for modulating an input bit sequence;
and transmitting the first signal.
2. The method as recited in claim 1, further comprising:
and sending first information to a receiving end, wherein the first information indicates the number of the at least one bit with the value of 0 or the length of the first bit sequence.
3. The method as recited in claim 2, further comprising:
determining the number of the at least one bit with a value of 0 based on the signal quality parameter of the first channel; and determining the length of the first bit sequence based on the number of neurons of the input layer of the first neural network model and the number of the at least one bit with a value of 0, wherein the first channel is a channel between the transmitting end and the receiving end; or,
Determining a length of the first bit sequence based on a signal quality parameter of a first channel; and determining the number of the at least one bit with a value of 0 based on the number of neurons of the input layer of the first neural network model and the length of the first bit sequence, wherein the first channel is a channel between the transmitting end and the receiving end.
4. The method of claim 1, wherein adding at least one bit having a value of 0 to the second bit sequence to obtain a third bit sequence comprises:
receiving second information from a receiving end, wherein the second information indicates the number of the at least one bit with the value of 0;
and adding at least one bit with a value of 0 into the second bit sequence according to the second information to obtain a third bit sequence.
5. The method of claim 4, further comprising, before adding at least one bit having a value of 0 to the second bit sequence according to the second information, obtaining a third bit sequence:
determining the length of the first bit sequence according to the number of neurons of an input layer of the first neural network model and the second information;
And determining the first bit sequence according to the length of the first bit sequence.
6. The method of claim 1, wherein adding at least one bit having a value of 0 to the second bit sequence to obtain a third bit sequence comprises:
receiving second information from a receiving end, wherein the second information indicates the length of the first bit sequence;
determining the number of the at least one bit with the value of 0 according to the number of the neurons of the input layer of the first neural network model and the second information;
and adding at least one bit with the value of 0 into the second bit sequence based on the number of the at least one bit with the value of 0, so as to obtain a third bit sequence.
7. The method of claim 6, further comprising, prior to adding at least one bit of value 0 to the second bit sequence based on the number of the at least one bit of value 0 to obtain a third bit sequence:
and determining the first bit sequence according to the second information.
8. The method of any one of claims 1-7, further comprising:
and sending third information to a receiving end, wherein the third information indicates the position of the at least one bit with the value of 0 in the third bit sequence.
9. The method according to any of claims 1-7, wherein adding at least one bit with a value of 0 to the second bit sequence to obtain a third bit sequence, comprising:
receiving fourth information from a receiving end, wherein the fourth information indicates the position of the at least one bit with the value of 0 in the third bit sequence;
and according to the fourth information, adding at least one bit with a value of 0 into the second bit sequence to obtain a third bit sequence.
10. The method according to any one of claims 1-9, wherein the first preset value is-1.
11. A communication method applied to a receiving end, comprising:
receiving a second signal;
inputting the second signal into a second neural network model to obtain a fourth bit sequence; the second neural network model is used for demodulating the input signal, and the value of the length of the fourth bit sequence is equal to the number of neurons of the output layer of the second neural network model;
deleting at least one bit in the fourth bit sequence according to the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end, so as to obtain a fifth bit sequence;
Adjusting the value of each bit in the fifth bit sequence to be a numerical value in a first set to obtain a sixth bit sequence, wherein the first set comprises a first preset value and 1;
the bit with the value of the first preset value in the sixth bit sequence is adjusted to be 0, and a seventh bit sequence is obtained; wherein the first preset value is an integer which is not 0 and not 1.
12. The method as recited in claim 11, further comprising:
receiving first information from the transmitting end, wherein the first information indicates the number of bits with the value of 0 in a bit sequence carried in a signal transmitted by the transmitting end;
and determining the position of a bit with a value of 0 in a bit sequence carried in a signal sent by the sending end in the bit sequence carried in the signal sent by the sending end based on the first information.
13. The method as recited in claim 11, further comprising:
receiving first information from the transmitting end, wherein the first information indicates the length of an original bit sequence to be transmitted, which is carried in a signal transmitted by the transmitting end;
determining the number of bits with the value of 0 in a bit sequence carried in a signal sent by the sending end according to the number of neurons of an output layer of the second neural network model and the first information;
And determining the position of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end according to the number of the bit with the value of 0 in the bit sequence carried in the signal sent by the sending end.
14. The method as recited in claim 11, further comprising:
and sending second information to the sending end, wherein the second information indicates the number of bits with the value of 0 in a bit sequence carried in a signal sent by the sending end or the length of an original bit sequence to be sent carried in the signal sent by the sending end.
15. The method as recited in claim 14, further comprising:
determining the number of bits with the value of 0 in a bit sequence carried in a signal sent by a sending end based on a signal quality parameter of a first channel, and determining the length of an original bit sequence to be sent carried in the signal sent by the sending end based on the number of output layer neurons of the second neural network model and the number of bits with the value of 0 in the bit sequence carried in the signal sent by the sending end, wherein the first channel is a channel between the sending end and the receiving end; or,
Determining the length of an original bit sequence to be transmitted, which is carried in a signal transmitted by a transmitting end, based on a signal quality parameter of a first channel, and determining the number of bits with a value of 0 in the bit sequence carried in the signal transmitted by the transmitting end, which is a channel between the transmitting end and the receiving end, based on the number of neurons of an output layer of the second neural network model and the length of the original bit sequence to be transmitted, which is carried in the signal transmitted by the transmitting end.
16. The method of any one of claims 11-15, further comprising:
and receiving third information from the transmitting end, wherein the third information indicates the position of a bit with a value of 0 in a bit sequence carried in a signal transmitted by the transmitting end in the bit sequence carried in the signal transmitted by the transmitting end.
17. The method of any one of claims 11-15, further comprising:
and sending fourth information to the sending end, wherein the fourth information indicates the position of a bit with a value of 0 in a bit sequence carried in a signal sent by the sending end in the bit sequence carried in the signal sent by the sending end.
18. The method of any one of claims 11-17, wherein the first preset value is-1.
19. A communication device, comprising: functional module for implementing the method according to any of claims 1-18.
20. A communications apparatus comprising a processor coupled to a memory;
the memory is used for storing a computer program or instructions;
the processor being configured to execute part or all of the computer program or instructions in the memory, which, when executed, is configured to implement the method of any one of claims 1-18.
21. A communication device comprising a processor and a memory;
the memory is used for storing a computer program or instructions;
the processor being configured to execute part or all of the computer program or instructions in the memory, which, when executed, is configured to implement the method of any one of claims 1-18.
22. A chip system, the chip system comprising: a processing circuit; the processing circuit is coupled with a storage medium;
the processing circuitry being adapted to execute part or all of the computer program or instructions in the storage medium, which, when executed, is adapted to carry out the method of any one of claims 1-18.
23. A computer readable storage medium storing a computer program comprising instructions for implementing the method of any one of claims 1-18.
24. A computer program product, the computer program product comprising: computer program code which, when run on a computer, causes the computer to perform the method of any of claims 1-18.
25. A communication system comprising a transmitting end performing the method of any of claims 1-10 and a receiving end performing the method of any of claims 11-18.
CN202210785094.2A 2022-06-29 2022-06-29 Communication method and device Pending CN117395115A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210785094.2A CN117395115A (en) 2022-06-29 2022-06-29 Communication method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210785094.2A CN117395115A (en) 2022-06-29 2022-06-29 Communication method and device

Publications (1)

Publication Number Publication Date
CN117395115A true CN117395115A (en) 2024-01-12

Family

ID=89435091

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210785094.2A Pending CN117395115A (en) 2022-06-29 2022-06-29 Communication method and device

Country Status (1)

Country Link
CN (1) CN117395115A (en)

Similar Documents

Publication Publication Date Title
EP3579481B1 (en) Method and device for transmitting data
US10887067B2 (en) Coding scheme determining method and apparatus
US10484137B2 (en) Polar code hybrid automatic repeat request method and apparatus
US11601167B2 (en) Method and device for power adjustment in UE and base station
CN111953459B (en) Communication method and device
JP2021512509A (en) Methods used for uplink data transmission, terminal equipment and network equipment
CN115777183A (en) Method and apparatus for determining uplink/downlink transport block size and modulation coding scheme
CN106922206B (en) Method, device and equipment for determining modulation coding order
US11431541B2 (en) Data sending method and apparatus
KR102184074B1 (en) Method and apparatus of interference alignment in cellular network
CN111245560B (en) Data transmission method and device
US20220377707A1 (en) Communication Method and Communication Apparatus
US10171207B2 (en) Methods and apparatus for control bit detection
CN117395115A (en) Communication method and device
CN115842601A (en) Data processing method and device
CN116458236A (en) Method and apparatus for repeating transmission of data channels
WO2020107039A2 (en) Methods and apparatus for communications with finite blocklength coded modulation
CN111511025A (en) Power control method and terminal equipment
US20230421286A1 (en) Communication method and apparatus
CN111262647A (en) Data transmission method and device
WO2024011554A1 (en) Techniques for joint probabilistic shaping of multiple bits per modulation constellation
WO2024065566A1 (en) Methods and apparatus for communication of updates for machine-learning model
CN109923925A (en) A kind of terminal that be used to wirelessly communicate, the method and apparatus in base station
CN116073942A (en) Communication method and communication device
CN117176294A (en) Communication method and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication