CN114499605A - Signal transmission method, signal transmission device, electronic equipment and storage medium - Google Patents

Signal transmission method, signal transmission device, electronic equipment and storage medium Download PDF

Info

Publication number
CN114499605A
CN114499605A CN202210182057.2A CN202210182057A CN114499605A CN 114499605 A CN114499605 A CN 114499605A CN 202210182057 A CN202210182057 A CN 202210182057A CN 114499605 A CN114499605 A CN 114499605A
Authority
CN
China
Prior art keywords
signal
precoding
strength indication
signal strength
analog
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.)
Granted
Application number
CN202210182057.2A
Other languages
Chinese (zh)
Other versions
CN114499605B (en
Inventor
张志锋
刘鹤
郭亮
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.)
BOE Technology Group Co Ltd
Beijing BOE Sensor Technology Co Ltd
Original Assignee
BOE Technology Group Co Ltd
Beijing BOE Sensor Technology 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 BOE Technology Group Co Ltd, Beijing BOE Sensor Technology Co Ltd filed Critical BOE Technology Group Co Ltd
Priority to CN202210182057.2A priority Critical patent/CN114499605B/en
Publication of CN114499605A publication Critical patent/CN114499605A/en
Application granted granted Critical
Publication of CN114499605B publication Critical patent/CN114499605B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H04B7/0426—Power distribution
    • H04B7/043—Power distribution using best eigenmode, e.g. beam forming or beam steering
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • 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
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H04B7/0456—Selection of precoding matrices or codebooks, e.g. using matrices antenna weighting
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

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

Abstract

The disclosure provides a signal transmission method, a signal transmission device, electronic equipment and a storage medium, and belongs to the technical field of communication. The method comprises the following steps: a receiving end measures the signal strength indication sent by a synchronous signal broadcasted by a base station; inputting the signal strength indication measurement into an unsupervised neural network model to obtain digital precoding and analog precoding, wherein the unsupervised neural network model is obtained by training by adopting sample signal strength indication measurement and optimizing based on a throughput loss function; and coding transmission data by adopting a mixed beam forming mode based on the digital precoding and the analog precoding, and transmitting the output signal to the terminal. The scheme improves the efficiency of signal transmission while ensuring the accuracy of signal transmission. Therefore, the throughput of signal transmission is ensured, large-scale channel information feedback is not needed, and the resource consumption of signal transmission adopting mixed beam forming is reduced.

Description

Signal transmission method, signal transmission device, electronic equipment and storage medium
Technical Field
The present disclosure relates to the field of computers, and in particular, to a signal transmission method and apparatus, an electronic device, and a storage medium.
Background
Hybrid Beamforming (HBF) is one of Beamforming technologies, and combines Digital Beamforming (DBF) and Analog Beamforming (ABF) to adjust the weight coefficients of an antenna array to generate directional beams, so as to effectively reduce the number of radio frequency links in a millimeter wave communication system and reduce hardware cost.
However, the hybrid beamforming technique needs to obtain the signal state information between each transmitting antenna and each receiving antenna, which occupies a large amount of spectrum resources, and as the number of antennas in millimeter wave band increases, the amount and complexity of channel information estimation increase greatly.
Disclosure of Invention
The disclosure provides a signal transmission method, a signal transmission device, electronic equipment and a storage medium.
A method of signal transmission, the method comprising:
a receiving end measures the signal strength indication sent by a synchronous signal broadcasted by a base station;
inputting the signal strength indication measurement into an unsupervised neural network model to obtain digital precoding and analog precoding, wherein the unsupervised neural network model is obtained by training by adopting sample signal strength indication measurement and optimizing based on a throughput loss function;
and coding transmission data by adopting a mixed beam forming mode based on the digital precoding and the analog precoding, and transmitting the output signal to the terminal.
Optionally, the inputting the signal strength indication measurement to an unsupervised neural network model to obtain digital precoding and analog precoding includes:
performing linear quantization operation on the signal strength indication measurement to obtain signal strength indication linear measurement;
and inputting the signal strength indication linear measurement to the unsupervised neural network model to obtain digital precoding and analog precoding.
Optionally, the performing a linear quantization operation on the signal strength indication measurement to obtain a signal strength indication linear measurement includes:
performing a linear quantization operation on the signal strength indication measurement by:
Figure BDA0003521948670000021
wherein, the
Figure BDA0003521948670000022
Represents the signal strength indication measurement sent by the receiving end u, said auRepresents the signal strength indication sent by the receiving end u, NbRepresenting the quantization bit width.
Optionally, the unsupervised neural network model is obtained by:
obtaining a sample signal strength indication measurement;
inputting the sample signal strength indication measurement to an unsupervised neural network model to be trained to obtain a prediction digital precoding and a prediction analog precoding;
calculating the spectral efficiency of the predictive digital precoding and the predictive analog precoding;
calculating a value of a throughput loss function based on the spectral efficiency;
and when the value of the throughput loss function does not meet the optimization requirement, regulating the weight value and the bias value of the unsupervised raw clean network model according to the value of the throughput loss function, and then retraining until the value of the throughput loss function meets the optimization requirement, and confirming that the unsupervised neural played model is trained.
Optionally, the calculating a value of a throughput loss function based on the spectral efficiency comprises:
inputting the spectral efficiency into the following formula to calculate the value of the throughput loss function:
Figure BDA0003521948670000023
wherein, L isHBFA value representing a throughput loss function, said
Figure BDA0003521948670000024
Representing the spectral efficiency of a predictive analog precoding of the l-th signal
Figure BDA0003521948670000025
Weight values representing the analog precoding, said A(l)Representing the phase and amplitude values of the predictive analog precoding of the l-th signal, Pa(l)Represents the predictive digital precoding of the L-th signal, said L representing the total number of signals.
Optionally, the calculating the spectral efficiency of the predictive digital precoding and the predictive analog precoding includes:
inputting the predicted digital precoding and the predicted analog precoding into the following formula to obtain the spectrum efficiency:
Figure BDA0003521948670000031
wherein, the
Figure BDA0003521948670000032
Representing the maximum spectral efficiency of the signal, A representing the phase and amplitude values of the predictive analog precoding, W representing the weight values corresponding to the predictive digital precoding, and SINR (A, W)u) Represents the corresponding signal-to-interference-and-noise ratio of the receiving end under the combined action of the mixed pre-coding, and the wuIdentifying the u-th predictive digital precoding, said u representing the number of signals, said PmaxAnd A represents.
Optionally, the encoding the transmission data by using a hybrid beamforming method based on the digital precoding and the analog precoding to obtain an output signal includes:
coding the transmission signal based on the digital pre-coding to obtain a corresponding digital signal;
and taking the analog signal obtained by assigning values to the antenna on the radio frequency link according to the digital signal and the analog precoding as an output signal.
Some embodiments of the present disclosure provide a signal transmission apparatus, the apparatus comprising:
a receiving module configured to receive a signal strength indication measurement transmitted by a receiving end for a synchronization signal broadcasted by a base station;
a prediction module configured to input the signal strength indication measurements to an unsupervised neural network model, resulting in digital precoding and analog precoding, wherein the unsupervised neural network model is trained using sample signal strength indication measurements and optimized based on a throughput loss function;
and the output module is configured to encode transmission data by adopting a hybrid beam forming mode based on the digital precoding and the analog precoding and transmit the output signal to the terminal.
Optionally, the prediction module is further configured to:
performing linear quantization operation on the signal strength indication measurement to obtain signal strength indication linear measurement;
and inputting the signal strength indication linear measurement to the unsupervised neural network model to obtain digital precoding and analog precoding.
Optionally, the prediction module is further configured to:
performing a linear quantization operation on the signal strength indication measurement by:
Figure BDA0003521948670000041
wherein, the
Figure BDA0003521948670000042
Represents the signal strength indication measurement sent by the receiving end u, said auRepresents the signal strength indication sent by the receiving end u, NbRepresenting the quantization bit width.
Optionally, the apparatus further comprises: a training module configured to:
obtaining a sample signal strength indication measurement;
inputting the sample signal strength indication measurement to an unsupervised neural network model to be trained to obtain a prediction digital precoding and a prediction analog precoding;
calculating the spectral efficiency of the predictive digital precoding and the predictive analog precoding;
calculating a value of a throughput loss function based on the spectral efficiency;
and when the value of the throughput loss function does not meet the optimization requirement, regulating the weight value and the bias value of the unsupervised raw clean network model according to the value of the throughput loss function, and then retraining until the value of the throughput loss function meets the optimization requirement, and confirming that the unsupervised neural played model is trained.
The training module further configured to:
inputting the spectral efficiency into the following formula to calculate the value of the throughput loss function:
Figure BDA0003521948670000043
wherein, L isHBFA value representing a throughput loss function, said
Figure BDA0003521948670000044
Representing the spectral efficiency of predictive analog precoding of the l-th signal
Figure BDA0003521948670000045
Weight values representing the analog precoding, said A(l)Representing the phase and amplitude values of the predictive analog precoding of the l-th signal, Pa(l)Represents the predictive digital precoding of the L-th signal, said L representing the total number of signals.
The training module further configured to:
inputting the predicted digital precoding and the predicted analog precoding into the following formula to obtain the spectrum efficiency:
Figure BDA0003521948670000051
wherein, the
Figure BDA0003521948670000052
Representing the maximum spectral efficiency of the signal, A representing the phase and amplitude values of the predictive analog precoding, W representing the weight values corresponding to the predictive digital precoding, and SINR (A, W)u) Represents the corresponding signal-to-interference-and-noise ratio of the receiving end under the combined action of the mixed pre-coding, and the wuIdentifying the u-th predictive digital precoding, said u representing the number of signals, said PmaxRepresents the maximum transmit power, and a represents the analog-side precoding.
Optionally, the output module is further configured to:
coding the transmission signal based on the digital pre-coding to obtain a corresponding digital signal;
and taking the analog signal obtained by assigning values to the antenna on the radio frequency link according to the digital signal and the analog precoding as an output signal.
Some embodiments of the present disclosure provide a computing processing device comprising:
a memory having computer readable code stored therein;
one or more processors that, when the computer readable code is executed by the one or more processors, the computing processing device performs the signal transmission method as described above.
Some embodiments of the present disclosure provide a computer program comprising computer readable code which, when run on a computing processing device, causes the computing processing device to perform a signal transmission method as described above.
Some embodiments of the present disclosure provide a non-transitory computer readable medium in which the signal transmission method as described above is stored.
According to the signal transmission method, the signal strength indication measurement at the receiving end is used for representing the channel state, the unsupervised neural network model obtained by taking the throughput as the optimization target is used for predicting the digital precoding and the analog precoding according to the signal strength indication measurement so as to be used by the subsequent mixed beam forming transmitting signal, the throughput of signal transmission is ensured, meanwhile, large-scale channel information feedback is not needed, and the resource consumption of signal transmission adopting mixed beam forming is reduced.
The foregoing description is only an overview of the technical solutions of the present disclosure, and the embodiments of the present disclosure are described below in order to make the technical means of the present disclosure more clearly understood and to make the above and other objects, features, and advantages of the present disclosure more clearly understandable.
Drawings
In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present disclosure, and other drawings can be obtained according to the drawings without creative efforts for those skilled in the art.
Fig. 1 schematically illustrates a flow chart of a signal transmission method provided by some embodiments of the present disclosure;
fig. 2 is a system diagram schematically illustrating a hybrid beam forming transceiver system in the related art;
fig. 3 schematically illustrates a system diagram of a signal transmission method according to some embodiments of the present disclosure;
fig. 4 schematically illustrates one of the flow diagrams of another signal transmission method provided by some embodiments of the present disclosure;
FIG. 5 schematically illustrates one of the flow diagrams of a method for training an unsupervised neural network model provided by some embodiments of the present disclosure;
FIG. 6 schematically illustrates a schematic diagram of a method for obtaining a data set according to some embodiments of the present disclosure;
FIG. 7 schematically illustrates a second flowchart of a method for unsupervised neural network model training according to some embodiments of the present disclosure;
fig. 8 schematically illustrates a third flowchart of a method for training an unsupervised neural network model according to some embodiments of the present disclosure;
fig. 9 schematically illustrates a second flow chart of another signal transmission method provided in some embodiments of the present disclosure;
fig. 10 schematically illustrates a structural diagram of a signal transmission device according to some embodiments of the present disclosure;
FIG. 11 schematically illustrates a block diagram of a computing processing device for performing methods according to some embodiments of the present disclosure;
fig. 12 schematically illustrates a memory unit for holding or carrying program code implementing methods according to some embodiments of the disclosure.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present disclosure more clear, the technical solutions of the embodiments of the present disclosure will be described clearly and completely with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are some, but not all embodiments of the present disclosure. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without inventive step, are intended to be within the scope of the present disclosure.
Fig. 1 schematically shows a flow chart of a signal transmission method provided by the present disclosure, where the method includes:
step 101, the receiving end measures the signal strength indication sent by the synchronization signal broadcasted by the base station.
The main implementation body of the embodiment of the present disclosure is a base station for sending data signals to a receiving end, where the base station broadcasts Synchronization Signals (SSBs) outwards according to a specific time period, and each SSB Signal is composed of analog precodes of a plurality of bits (bits), so that a user u at the receiving endthReceived kththThe SSB signal can be expressed as the following equation (1):
Figure BDA0003521948670000071
wherein the content of the first and second substances,
Figure BDA0003521948670000072
indicating that the kth user receives all the kth signal,
Figure BDA0003521948670000073
represented by user uthThe noise at the time of reception of the signal,
Figure BDA0003521948670000074
the channel vector at this time is represented.
After receiving the SSB signal, the receiving end feeds back a signal strength indication signal for indicating the strength of the SSB signal to the base station, and the receiving end can measure the SSB signal strength, and the magnitude of the signal strength indication signal can be obtained by measuring the average value of the SSB signal received by the receiving end, which can specifically represent the following formula (2):
Figure BDA0003521948670000075
wherein, therein
Figure BDA0003521948670000081
Representative is the signal strength indicator measurement, σ2Representing the noise level. The receiving end can feed back the signal strength indication measurement to the base through an error-free channelAnd (4) a station. Since the signal strength indication measurement is direct strength measurement of the signal at the receiving end, large-scale CSI (Channel State Information) feedback is not required, and the overhead of signal transmission can be reduced.
And 102, inputting the signal strength indication measurement into an unsupervised neural network model to obtain digital precoding and analog precoding, wherein the unsupervised neural network model is obtained by training by adopting sample signal strength indication measurement and optimizing based on a throughput loss function.
In this embodiment of the present disclosure, since the information obtained by the base station to reflect the channel state is not large-scale channel information in an ideal state, but signal strength indication measurement is used to characterize the signal state characteristics, and the base station cannot clearly measure the noise optical signal level contained in the entire signal, based on this disclosure, the unsupervised neural network obtained by deep learning training is used to predict and obtain hybrid precoding, and the sir expression of the receiving end under the effect of hybrid precoding may be the following formula (3):
Figure BDA0003521948670000082
wherein, SINR (A, w)u) Representing the signal to interference and noise ratio, A representing the phase and amplitude values of the analog signal, wuIndicating that the u-th user received the digital signal, huThe virtual channel matrices corresponding to different users are shown. It can be known from the above equation (3) that the remaining analog precoding can be optimized in addition to the digital precoding, so the present disclosure optimizes the analog precoding part as the following equation (4):
Figure BDA0003521948670000083
wherein, betaiRepresenting the eigenvalue corresponding to the channel matrix, and the coefficient μ representing the coefficient of the water filling algorithm, the following relation of formula (5) needs to be satisfied:
Figure BDA0003521948670000091
wherein ρ represents a Signal-to-Noise Ratio (SNR), a loss value is calculated according to a loss function representing throughput obtained by using the above optimization method, and an optimal candidate weight is continuously searched by using a genetic algorithm in an iterative manner according to the magnitude of the loss value, so that a mixed precoding composed of a digital precoding and an analog precoding output by the unsupervised neural network model obtained by training can achieve the maximum throughput of an output Signal obtained in the mixed beam shaping process.
And 103, coding transmission data by adopting a hybrid beam forming mode based on the digital precoding and the analog precoding, and transmitting the output signal to the terminal.
In the embodiment of the present disclosure, a system diagram of a hybrid beamforming transceiving system in the related art is shown with reference to fig. 2, where base and digital precoding (Baseband digital precoder) needs to perform amplitude and phase joint modulation with RF analog precoding (Radio Frequency analog precoder) based on received signal information Ns to realize signal transmission based on hybrid beamforming.
Fig. 3 is a system diagram illustrating a signal transmission method provided by the present disclosure, wherein a Digital precoder (Digital precoding) in a base station may directly encode transmission data nS according to a precoder output by an unsupervised neural network model provided by the present disclosure to obtain a Digital signal, and then assigns a value to a connected antenna through a radio frequency link (RF Chain) based on the Digital signal and analog precoding to obtain an analog signal, and sends the analog signal to a receiving end to send an output signal.
According to the method and the device, the channel state is represented through the signal strength indication measurement of the receiving end, the non-supervision neural network model obtained by taking the throughput as an optimization target is used for predicting the digital pre-coding and the analog pre-coding according to the signal strength indication measurement so as to be used by the subsequent mixed beam forming transmitting signals, the throughput of signal transmission is guaranteed, meanwhile, large-scale channel information feedback is not needed, and the resource consumption of signal transmission adopting mixed beam forming is reduced.
Optionally, referring to fig. 4, the step 102 includes:
step 1021, performing a linear quantization operation on the signal strength indication measurement to obtain a signal strength indication linear measurement.
In the disclosed embodiment, in order to apply the signal strength indication measurement to the actual use process of the unsupervised neural network model, the present disclosure performs a linear quantization operation on the signal strength indication measurement.
And 1022, inputting the signal strength indication linear measurement to the unsupervised neural network model to obtain digital precoding and analog precoding.
According to the embodiment of the model training and prediction method, the signal intensity obtained by linear quantization operation is used for indicating linear measurement to train and predict the model, the normalization of the model input data is ensured, and the accuracy of the unsupervised neural network model training and prediction is improved.
Optionally, the step 1022 includes:
performing a linear quantization operation on the signal strength indication measurement by:
Figure BDA0003521948670000101
wherein, the
Figure BDA0003521948670000102
Represents the signal strength indication measurement sent by the receiving end u, said auRepresents the signal strength indication sent by the receiving end u, NbRepresenting the quantization bit width.
In bookIn the disclosed embodiment, wherein NbThe quantization bit width is represented, and the difference caused by the quantization bit width is small, so that the quantization bit width is not considered in the system. The base station may perform digital precoding operation through a pure digital domain, and the corresponding weight may be expressed as the following formula (6):
Figure BDA0003521948670000103
wherein
Figure BDA0003521948670000104
Representative is a specific digital precoding for a certain user, whereas analog precoding is an operation of assigning values to all antennas on different RF radio links. After codebook design is carried out on analog precoding, a basic signal model y obtained at a receiving end can be further obtaineduThe following equation (7):
Figure BDA0003521948670000105
wherein x isuRepresented are symbol data transmitted by different users,
Figure BDA0003521948670000106
representing the channel vector between different transmit and receive antennas.
Fig. 5 schematically shows a flowchart of a method for training an unsupervised neural network model provided by the present disclosure, where the method includes:
step 201, a sample signal strength indication measurement is obtained.
In the embodiment of the present disclosure, the sample signal strength indication measurement may be obtained by counting the signal strength indication measurement fed back by the receiving end aiming at the base station in different signal environments, and because the unsupervised deep learning manner is adopted in the present disclosure, the sample signal strength indication measurement does not need to be labeled.
Illustratively, referring to FIG. 6, the model data set may be generated by:
a1, generating MIMO (Multiple Input Multiple Output) channel vectors;
a2, randomly generating a channel vector according to the position of the UE and the channel matrix;
a3, designing an SSB signal;
a4, calculating the RSSI (Received Signal Strength Indication) of the user;
a5, designing precoding weights of an analog domain and a digital domain with maximized throughput;
and A6, generating the weight of the simulation end codebook.
And A7, dividing the calculated values into a test set and a training set according to a specific proportion.
Step 202, inputting the sample signal strength indication measurement to the unsupervised neural network model to be trained, and obtaining the predictive digital precoding and the predictive analog precoding.
In the embodiment of the present disclosure, a structural composition diagram of the unsupervised neural network model to be trained may refer to fig. 7, where the corresponding output dimension is 2 × NU×NRFThe same size operation block is used in the network for all convolutional layers, and the normalization operation of Batch is performed after each convolutional layer, so as to reduce the reduction of internal variables and the fast convergence of the training process, and simultaneously resist the interference problem caused by overfitting and noise. At the same time, Dropout operation is adopted to further reduce the potential risk of network overfitting.
Step 203, calculating the spectral efficiency of the predictive digital precoding and the predictive analog precoding.
In the embodiment of the present disclosure, the spectral efficiency is calculated by inputting the values of the digital precoding and the analog precoding output by the model into a preset calculation formula of the spectral efficiency.
A value of a throughput loss function is calculated based on the spectral efficiency, step 204.
In the present embodiment, in consideration of the correlation between the frequency efficiency and the throughput, the present disclosure exploits the loss function representing the throughput calculated by the maximum spectral efficiency, and it should be noted that the throughput loss function herein may be applicable to the present embodiment by representing the size of the throughput rather than calculating the specific throughput.
Step 205, when the value of the throughput loss function does not meet the optimization requirement, adjusting the weight value and the bias value of the unsupervised raw and clean network model according to the value of the throughput loss function, and then retraining until the value of the throughput loss function meets the optimization requirement, and confirming that the training of the unsupervised neural played model is finished.
In the embodiment of the present disclosure, referring to fig. 8, comparing the proposed HBF-NET scheme with the optimization scheme with the conventional Zero Forcing (ZF) and OMP scheme, from the result of numerical simulation, the unsupervised learning based neural network solution can obtain an approximately optimal performance boundary. The scheme provides a design method of hybrid beam forming, which uses an unsupervised deep learning method. This reduces training time and cost by using unsupervised learning methods for training. The deep learning approach selects APs from the codebook in order to reduce complexity while ensuring that the computational complexity remains within a reasonable range. In addition, the method can be used for deploying application in a real-time system due to small calculation amount and low complexity.
Optionally, the step 204 includes:
inputting the spectral efficiency into the following formula to calculate the value of the throughput loss function:
Figure BDA0003521948670000121
wherein, L isHBFA value representing a throughput loss function, said
Figure BDA0003521948670000122
Representing the spectral efficiency of a predictive analog precoding of the l-th signal
Figure BDA0003521948670000123
Weight values representing the analog precoding, said A(l)Representing the phase and amplitude values of the predictive analog precoding of the l-th signal, Pa(l)Represents the predictive digital precoding of the L-th signal, said L representing the total number of signals.
Optionally, the step 203 includes:
inputting the predictive digital precoding and the predictive analog precoding into the following formulas to obtain the spectrum efficiency:
Figure BDA0003521948670000124
wherein, the
Figure BDA0003521948670000125
Representing the maximum spectral efficiency of the signal, A representing the phase and amplitude values of the predictive analog precoding, W representing the weight values corresponding to the predictive digital precoding, and SINR (A, W)u) Represents the corresponding signal-to-interference-and-noise ratio of the receiving end under the combined action of the mixed pre-coding, and the wuIdentifying the u-th predictive digital precoding, said u representing the number of signals, said PmaxRepresents the maximum transmission power, said
Figure BDA0003521948670000131
Representing analog side precoding.
Optionally, referring to fig. 9, the step 104 includes:
step 1041, encoding the transmission signal based on the digital precoding to obtain a corresponding digital signal.
And 1042, assigning the antenna on the radio frequency link according to the digital signal and the analog precoding to obtain an analog signal, and using the analog signal as an output signal.
In the embodiment of the present disclosure, after a Digital precoder (Digital precoding) in a base station directly encodes transmission data nS according to a precoder output by an unsupervised neural network model provided by the present disclosure to obtain a Digital signal, an antenna connected to the Digital signal and analog precoding is assigned through a radio frequency link (RF Chain) to obtain an analog signal, and the analog signal is transmitted to a receiving end to obtain an output signal, unsupervised learning may maximally reduce the computational complexity of hybrid beamforming, and at the same time, the whole model network structure is more convenient for deployment and use in an actual system, and performance consumption of the base station in a signal transmission process is reduced.
Fig. 10 schematically shows a structural schematic diagram of a signal transmission device 30 provided by the present disclosure, where the device includes:
a receiving module 301 configured to receive a signal strength indication measurement transmitted for a synchronization signal broadcasted by a base station;
a prediction module 302 configured to input the signal strength indication measurements to an unsupervised neural network model, resulting in digital precoding and analog precoding, wherein the unsupervised neural network model is trained using sample signal strength indication measurements and optimized based on a throughput loss function;
and an output module 303, configured to encode transmission data by using a hybrid beamforming method based on the digital precoding and the analog precoding, and transmit the output signal to the terminal.
Optionally, the prediction module 302 is further configured to:
performing linear quantization operation on the signal strength indication measurement to obtain signal strength indication linear measurement;
and inputting the signal strength indication linear measurement to the unsupervised neural network model to obtain digital precoding and analog precoding.
Optionally, the prediction module 302 is further configured to:
performing a linear quantization operation on the signal strength indication measurement by:
Figure BDA0003521948670000141
wherein, the
Figure BDA0003521948670000142
Represents the signal strength indication measurement sent by the receiving end u, said auRepresents the signal strength indication sent by the receiving end u, NbRepresenting the quantization bit width.
Optionally, the apparatus further comprises: a training module configured to:
obtaining a sample signal strength indication measurement;
inputting the sample signal strength indication measurement to an unsupervised neural network model to be trained to obtain a prediction digital precoding and a prediction analog precoding;
calculating the spectral efficiency of the predictive digital precoding and the predictive analog precoding;
calculating a value of a throughput loss function based on the spectral efficiency;
and when the value of the throughput loss function does not meet the optimization requirement, regulating the weight value and the bias value of the unsupervised raw clean network model according to the value of the throughput loss function, and then retraining until the value of the throughput loss function meets the optimization requirement, and confirming that the unsupervised neural played model is trained.
The training module further configured to:
inputting the spectral efficiency into the following formula to calculate the value of the throughput loss function:
Figure BDA0003521948670000143
wherein, L isHBFA value representing a throughput loss function, said
Figure BDA0003521948670000144
Representing the spectral efficiency of a predictive analog precoding of the l-th signal
Figure BDA0003521948670000145
Weight values representing the analog precoding, said A(l)Representing the phase and amplitude values of the predictive analog precoding of the l-th signal, Pa(l)Represents the predictive digital precoding of the L-th signal, said L representing the total number of signals.
The training module further configured to:
inputting the predicted digital precoding and the predicted analog precoding into the following formula to obtain the spectrum efficiency:
Figure BDA0003521948670000151
wherein, the
Figure BDA0003521948670000152
Representing the maximum spectral efficiency of the signal, A representing the phase and amplitude values of the predictive analog precoding, W representing the weight values corresponding to the predictive digital precoding, and SINR (A, W)u) Represents the corresponding signal-to-interference-and-noise ratio of the receiving end under the combined action of the mixed pre-coding, and the wuIdentifying the u-th predictive digital precoding, said u representing the number of signals, said PmaxRepresents the maximum transmission power, said
Figure BDA0003521948670000153
Analog side precoding is shown.
Optionally, the output module 303 is further configured to:
coding the transmission signal based on the digital pre-coding to obtain a corresponding digital signal;
and taking the analog signal obtained by assigning values to the antenna on the radio frequency link according to the digital signal and the analog precoding as an output signal.
According to the method and the device, the channel state is represented through the signal strength indication measurement of the receiving end, the non-supervision neural network model obtained by taking the throughput as an optimization target is used for predicting the digital pre-coding and the analog pre-coding according to the signal strength indication measurement so as to be used by the subsequent mixed beam forming transmitting signals, the throughput of signal transmission is guaranteed, meanwhile, large-scale channel information feedback is not needed, and the resource consumption of signal transmission adopting mixed beam forming is reduced.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed 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 modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Various component embodiments of the disclosure may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or Digital Signal Processor (DSP) may be used in practice to implement some or all of the functionality of some or all of the components in a computing processing device according to embodiments of the disclosure. The present disclosure may also be embodied as apparatus or device programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present disclosure may be stored on a non-transitory computer readable medium or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.
For example, FIG. 11 illustrates a computing processing device that may implement methods in accordance with the present disclosure. The computing processing device conventionally includes a processor 310 and a computer program product or non-transitory computer-readable medium in the form of a memory 320. The memory 320 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read only memory), an EPROM, a hard disk, or a ROM. The memory 320 has a storage space 330 for program code 331 for performing any of the method steps of the above-described method. For example, the storage space 330 for the program code may include respective program codes 331 respectively for implementing various steps in the above method. The program code can be read from or written to one or more computer program products. These computer program products comprise a program code carrier such as a hard disk, a Compact Disc (CD), a memory card or a floppy disk. Such a computer program product is typically a portable or fixed storage unit as described with reference to fig. 12. The memory unit may have memory segments, memory spaces, etc. arranged similarly to the memory 320 in the computing processing device of fig. 11. The program code may be compressed, for example, in a suitable form. Typically, the memory unit comprises computer readable code 331', i.e. code that can be read by a processor, such as 310, for example, which when executed by a computing processing device causes the computing processing device to perform the steps of the method described above.
It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless explicitly stated herein. Moreover, at least a portion of the steps in the flow chart of the figure may include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed alternately or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
Reference herein to "one embodiment," "an embodiment," or "one or more embodiments" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. Moreover, it is noted that instances of the word "in one embodiment" are not necessarily all referring to the same embodiment.
In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the disclosure may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The disclosure may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The usage of the words first, second and third, etcetera do not indicate any ordering. These words may be interpreted as names.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solutions of the present disclosure, not to limit them; although the present disclosure has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present disclosure.

Claims (10)

1. A method of signal transmission, the method comprising:
a receiving end measures the signal strength indication sent by a synchronous signal broadcasted by a base station;
inputting the signal strength indication measurement into an unsupervised neural network model to obtain digital precoding and analog precoding, wherein the unsupervised neural network model is obtained by training by adopting sample signal strength indication measurement and optimizing based on a throughput loss function;
and coding transmission data by adopting a mixed beam forming mode based on the digital precoding and the analog precoding, and transmitting the output signal to the terminal.
2. The method of claim 1, wherein inputting the signal strength indication measurements to an unsupervised neural network model results in digital precoding and analog precoding, comprising:
performing linear quantization operation on the signal strength indication measurement to obtain signal strength indication linear measurement;
and inputting the signal strength indication linear measurement to the unsupervised neural network model to obtain digital precoding and analog precoding.
3. The method of claim 1, wherein performing a linear quantization operation on the signal strength indication measurements resulting in signal strength indication linear measurements comprises:
performing a linear quantization operation on the signal strength indication measurement by:
Figure FDA0003521948660000011
wherein, the
Figure FDA0003521948660000012
Represents the signal strength indication measurement sent by the receiving end u, said auRepresents the signal strength indication sent by the receiving end u, NbRepresenting the quantization bit width.
4. The method of claim 1, wherein the unsupervised neural network model is obtained by:
obtaining a sample signal strength indication measurement;
inputting the sample signal strength indication measurement to an unsupervised neural network model to be trained to obtain a prediction digital precoding and a prediction analog precoding;
calculating the spectral efficiency of the predictive digital precoding and the predictive analog precoding;
calculating a value of a throughput loss function based on the spectral efficiency;
and when the value of the throughput loss function does not meet the optimization requirement, regulating the weight value and the bias value of the unsupervised raw clean network model according to the value of the throughput loss function, and then retraining until the value of the throughput loss function meets the optimization requirement, and confirming that the unsupervised neural played model is trained.
5. The method of claim 4, wherein the calculating a value of a throughput loss function based on the spectral efficiency comprises:
calculating a value of a throughput loss function by inputting the spectral efficiency into the following equation:
Figure FDA0003521948660000021
wherein, L isHBFA value representing a throughput loss function, said
Figure FDA0003521948660000022
Representing the spectral efficiency of a predictive analog precoding of the l-th signal
Figure FDA0003521948660000023
Weight values representing the analog precoding, said A(l)Representing the phase and amplitude values of the predictive analog precoding of the l-th signal, Pa(l)Represents the predictive digital precoding of the L-th signal, said L representing the total number of signals.
6. The method of claim 4, wherein calculating the spectral efficiency of the predictive digital precoding and the predictive analog precoding comprises:
inputting the predicted digital precoding and the predicted analog precoding into the following formula to obtain the spectrum efficiency:
Figure FDA0003521948660000024
wherein, the
Figure FDA0003521948660000025
Representing the maximum spectral efficiency of the signal, A representing the phase and amplitude values of the predictive analog precoding, W representing the weight values corresponding to the predictive digital precoding, and SINR (A, W)u) Represents the corresponding signal-to-interference-and-noise ratio of the receiving end under the combined action of the mixed pre-coding, and the wuIdentifying the u-th predictive digital precoding, said u representing the number of signals, said PmaxRepresents the maximum transmit power, and said a represents the analog end precoding.
7. The method of claim 1, wherein the encoding the transmission data by hybrid beamforming based on the digital precoding and the analog precoding to obtain an output signal comprises:
coding the transmission signal based on the digital pre-coding to obtain a corresponding digital signal;
and taking the analog signal obtained by assigning values to the antenna on the radio frequency link according to the digital signal and the analog precoding as an output signal.
8. A signal transmission apparatus, characterized in that the apparatus comprises:
a receiving module configured to receive a signal strength indication measurement transmitted by a receiving end for a synchronization signal broadcasted by a base station;
a prediction module configured to input the signal strength indication measurements to an unsupervised neural network model, resulting in digital precoding and analog precoding, wherein the unsupervised neural network model is trained using sample signal strength indication measurements and optimized based on a throughput loss function;
and the output module is configured to encode transmission data by adopting a hybrid beam forming mode based on the digital precoding and the analog precoding and transmit the output signal to the terminal.
9. A computing processing device, comprising:
a memory having computer readable code stored therein;
one or more processors that when the computer readable code is executed by the one or more processors, the computing processing device performs the signal transmission method of any of claims 1-7.
10. A non-transitory computer-readable medium in which a computer program of the signal transmission method according to any one of claims 1 to 7 is stored.
CN202210182057.2A 2022-02-25 2022-02-25 Signal transmission method, signal transmission device, electronic equipment and storage medium Active CN114499605B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210182057.2A CN114499605B (en) 2022-02-25 2022-02-25 Signal transmission method, signal transmission device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210182057.2A CN114499605B (en) 2022-02-25 2022-02-25 Signal transmission method, signal transmission device, electronic equipment and storage medium

Publications (2)

Publication Number Publication Date
CN114499605A true CN114499605A (en) 2022-05-13
CN114499605B CN114499605B (en) 2023-07-04

Family

ID=81483578

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210182057.2A Active CN114499605B (en) 2022-02-25 2022-02-25 Signal transmission method, signal transmission device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN114499605B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108462517A (en) * 2018-03-06 2018-08-28 东南大学 A kind of MIMO link self-adaption transmission methods based on machine learning
WO2019119442A1 (en) * 2017-12-22 2019-06-27 Telefonaktiebolaget Lm Ericsson (Publ) A wireless communications system, a radio network node, a machine learning unt and methods therein for transmission of a downlink signal in a wireless communications network supporting beamforming
CN110557177A (en) * 2019-09-05 2019-12-10 重庆邮电大学 DenseNet-based hybrid precoding method in millimeter wave large-scale MIMO system
CN111092641A (en) * 2019-12-18 2020-05-01 重庆邮电大学 Hybrid precoding design method based on millimeter wave MIMO system deep learning
CN112910520A (en) * 2021-02-03 2021-06-04 广州市埃特斯通讯设备有限公司 Convolutional neural network-based MIMO system beam training method
CN113344187A (en) * 2021-06-18 2021-09-03 东南大学 Machine learning precoding method for single-cell multi-user MIMO system
CN113411110A (en) * 2021-06-04 2021-09-17 东南大学 Millimeter wave communication beam training method based on deep reinforcement learning

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019119442A1 (en) * 2017-12-22 2019-06-27 Telefonaktiebolaget Lm Ericsson (Publ) A wireless communications system, a radio network node, a machine learning unt and methods therein for transmission of a downlink signal in a wireless communications network supporting beamforming
CN108462517A (en) * 2018-03-06 2018-08-28 东南大学 A kind of MIMO link self-adaption transmission methods based on machine learning
CN110557177A (en) * 2019-09-05 2019-12-10 重庆邮电大学 DenseNet-based hybrid precoding method in millimeter wave large-scale MIMO system
CN111092641A (en) * 2019-12-18 2020-05-01 重庆邮电大学 Hybrid precoding design method based on millimeter wave MIMO system deep learning
CN112910520A (en) * 2021-02-03 2021-06-04 广州市埃特斯通讯设备有限公司 Convolutional neural network-based MIMO system beam training method
CN113411110A (en) * 2021-06-04 2021-09-17 东南大学 Millimeter wave communication beam training method based on deep reinforcement learning
CN113344187A (en) * 2021-06-18 2021-09-03 东南大学 Machine learning precoding method for single-cell multi-user MIMO system

Also Published As

Publication number Publication date
CN114499605B (en) 2023-07-04

Similar Documents

Publication Publication Date Title
CN103139117B (en) Use the generalized reference signaling schemes of the MU MIMO of any precoded
US7336727B2 (en) Generalized m-rank beamformers for MIMO systems using successive quantization
RU2518177C2 (en) Method and device for determining precoding vector
WO2021142631A1 (en) Method, device and computer readable medium of communication
CN101507141A (en) Transform-domain feedback signaling for MIMO communication
EP3915200B1 (en) Design and adaptation of hierarchical codebooks
KR20160118086A (en) Apparatus and method for feeding back channel information in wireless communication system
CN113179109A (en) Honeycomb-removing large-scale MIMO uplink spectrum efficiency optimization method
KR101813601B1 (en) Method and apparatus for precoding mode selection with joint processing/transmission based on linited feedback in mobile communication system
US8625560B2 (en) Method and apparatus for feeding back channel quality information in multi-user multi-input multi-output communication system
JP5049587B2 (en) Method, apparatus, and signal for reporting information related to interference component, method and apparatus for controlling signal transfer, and computer program
CN108631830B (en) Method for determining transmitted beam, transmitting end and receiving end
EP3429256A1 (en) Apparatus and method for wireless communications, and parameter optimization apparatus and method
KR20160066665A (en) Method for Feedback and Scheduling for Massive MIMO System and Apparatus Therefor
CN114499605B (en) Signal transmission method, signal transmission device, electronic equipment and storage medium
EP2112771A1 (en) Adaptive MIMO mode selection method and apparatus thereof
CN104253639A (en) Channel quality indicator acquisition method and device
JPWO2009107635A1 (en) Wireless communication system, transmission apparatus, and communication control method
CN117318774A (en) Channel matrix processing method, device, terminal and network side equipment
CN116996142A (en) Wireless channel parameter prediction method, device, electronic equipment and storage medium
KR20230138538A (en) Information reporting method, apparatus, first apparatus and second apparatus
Kumar et al. Multi-user mmWave massive-MIMO hybrid beamforming: A quantize deep learning approach
CN115604824A (en) User scheduling method and system
EP1962539A1 (en) Method for providing channel information in a radio communications system and mobile station thereof
CN114448478B (en) Signal transmission method, signal transmission device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant