Transmission method, device, equipment and readable storage medium

Abstract

The present application discloses a transmission method, apparatus, device and readable storage medium, the method comprising: obtaining a first signal and a first information; inputting the first signal and the first information into a first model, and encoding or modulating the first signal according to the first information by the first model to obtain a second signal; and sending the second signal to a receiving end. In an embodiment of the present application, the first signal is encoded or modulated by the first information to achieve transmission under adaptive channel conditions, and various channel conditions can be adapted without changing network parameters, thereby reducing the storage capacity of model parameters.

Description

Technical Field

[0001] The present application belongs to the field of communication technology, and specifically relates to a transmission method, device, equipment and readable storage medium. Background Art

[0002] With the continuous emergence of various application scenarios based on extended reality (XR) and holographic communication technology, and the rapid development of intelligent interconnection of all things and data fusion in vertical industries such as transportation and manufacturing, the importance of data transmission services has become increasingly prominent. Better data transmission performance is reflected in the use of fewer channel resources to obtain less information distortion. The encoding process of image and video data in modern communication systems is usually divided into two steps:

[0003] (1) Use source coding (e.g., JPEG, JPEG2000) to eliminate the inherent redundancy of the source and reduce the amount of information to be transmitted;

[0004] (2) Use channel coding (e.g., Low Density Parity Check Code (LDPC), Turbo) to perform error checking coding on the compressed bit stream, increasing the number of transmitted bit streams to combat the interference of channel noise.

[0005] Based on Shannon's separation law, the above steps are theoretically optimal in the limit of asymptotic infinity. However, many emerging applications from the Internet of Things to autonomous driving and the tactile Internet require the transmission of images or videos under extremely low latency, bandwidth, and energy constraints, which limits the use of separate source channel coding techniques that rely on longer codes in computation. Joint source channel coding optimizes the source coding and channel coding of the communication system as a whole, pursuing end-to-end optimal performance. The emergence of deep learning makes it possible to design end-to-end joint source channel coding networks.

[0006] Existing joint source channel coding network models based on deep learning are as follows: Figure 1 As shown in the figure, the encoder of the network consists of 5 convolutional layers + a module with a parametric linear rectified unit (PRELU) activation function, and the decoder consists of 5 transposed convolutional layers + a module with a PRELU / Sigmoid activation function. The channel layer is located between the encoder and the decoder as a non-trainable layer. Compared with modern communication systems, the encoder of this model completes the functions of source coding + channel coding + modulation, and the decoder completes the functions of demodulation + channel decoding + source decoding.

[0007] Figure 2The performance of the deep joint source-channel coding algorithm and JPEG / JPEG200+channel capacity is compared on the CIFAR10 dataset. The performance simulation shows that under poor channel conditions (SNR = 0dB), the maximum transmission rate of the separate coding scheme is less than the channel capacity, and error-free transmission is not possible, resulting in decoding failure, while DJSCC coding can be transmitted with reasonable and good performance. Under medium (SNR = 10dB) and high (SNR = 20dB) signal-to-noise ratios and limited channel resources k / n < 0.3, even assuming reliable transmission under channel capacity, the performance of the proposed DJSCC coding is much higher than that of JPEG and JPEG2000. This comparison reflects the advantages of the deep joint source-channel coding scheme over the separate coding scheme in small bandwidths.

[0008] Figure 2 The performance of the deep joint source-channel codec and JPEG / JPEG200+ channel capacity is evaluated on CIFAR10 data for compression ratio k / n. For each case, the same SNR value is used in training and evaluation.

[0009] The performance of the above joint source channel coding network model when trained and evaluated using different SNR values ​​is shown in the figure below. Figure 3 As shown. It can be seen that when SNR train With SNR test When not matched, except when SNR train =SNR test To achieve the best performance, the SNR train ≠SNR test When , there is a large gap between the optimal performance. As a result, if the communication system works at SNR∈[0,20]db, the transmitter and receiver of the system need to store different SNRs at a certain SNR interval. train The system parameters trained under the condition are approximately the best performance of the system. The smaller the SNR interval, the closer the system is to the best performance, but the cost is the increase in the storage capacity of the transceiver occupied by the parameters. Assuming the model size is S, the storage capacity required is 6*S when stored at an SNR interval of 4dB, and the storage capacity required is 20*S when stored at an SNR interval of 1dB. When the model size of single-point training is large, the exponential storage capacity requirement increases the cost of communication equipment and reduces the practicality of deep joint source channel coding. Summary of the invention

[0010] The purpose of the embodiments of the present application is to provide a transmission method, apparatus, device and readable storage medium to solve the problem of large storage capacity of model parameters.

[0011] In a first aspect, an embodiment of the present application provides a transmission method, applied to a transmitting end, comprising:

[0012] Acquiring a first signal and first information;

[0013] Inputting the first signal and the first information into a first model, and having the first model encode or modulate the first signal according to the first information to obtain a second signal;

[0014] sending the second signal to a receiving end;

[0015] The first information includes one or more of the following combinations: channel state information of the communication channel between the sending end and the receiving end, information collected by the sensor of the sending end, and information collected by the sensor of the receiving end.

[0016] In a second aspect, an embodiment of the present application provides a transmission method, applied to a receiving end, comprising:

[0017] Acquiring first information and receiving a second signal from a transmitting end;

[0018] Inputting the second signal and the first information into a second model, and having the second model decode or demodulate the second signal according to the first information to obtain a first signal;

[0019] The first information includes one or more of the following combinations: channel state information of the communication channel between the sending end and the receiving end, information collected by the sensor of the sending end, and information collected by the sensor of the receiving end.

[0020] In a third aspect, an embodiment of the present application provides a transmission device, applied to a transmitting end, including:

[0021] A first acquisition module, used to acquire a first signal and first information;

[0022] A first processing module, configured to input the first signal and the first information into a first model, and the first model encodes or modulates the first signal according to the first information to obtain a second signal;

[0023] A sending module, used for sending the second signal to a receiving end;

[0024] The first information includes one or more of the following combinations: channel state information of the communication channel between the sending end and the receiving end, information collected by the sensor of the sending end, and information collected by the sensor of the receiving end.

[0025] In a fourth aspect, an embodiment of the present application provides a transmission device, applied to a receiving end, including:

[0026] A second acquisition module, used to acquire the first information and receive a second signal from the transmitting end;

[0027] A second processing module, used for inputting the second signal and the first information into a second model, and the second model decodes or demodulates the second signal according to the first information to obtain a first signal;

[0028] The first information includes one or more of the following combinations: channel state information of the communication channel between the sending end and the receiving end, information collected by the sensor of the sending end, and information collected by the sensor of the receiving end.

[0029] In a fifth aspect, an embodiment of the present application provides a communication device, comprising: a processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the transmission method described in the first aspect or the second aspect.

[0030] In a sixth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect or the second aspect are implemented.

[0031] According to a seventh aspect, a program product is provided. The program product is stored in a non-volatile storage medium and is executed by at least one processor to implement the steps of the processing method described in the first aspect or the second aspect.

[0032] In an eighth aspect, a chip is provided, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the processing method described in the first aspect or the second aspect.

[0033] In an embodiment of the present application, the first signal is encoded or modulated by the first information to achieve adaptive channel condition transmission, which can adapt to various channel conditions without changing the network parameters, thereby reducing the storage capacity of the model parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the existing joint source channel coding network model based on deep learning;

[0035] Figure 2 is a schematic diagram of the performance of the joint source channel coding network model when trained and evaluated using different compression ratios;

[0036] Figure 3 is a schematic diagram of the performance of the joint source channel coding network model when trained and evaluated using different SNR values;

[0037] Figure 4 is a block diagram of a wireless communication system to which the embodiments of the present application can be applied;

[0038] Figure 5 It is one of the flow charts of the transmission method in the embodiment of the present application;

[0039] Figure 6 This is the second flowchart of the transmission method in the embodiment of the present application;

[0040] Figure 7 is a schematic diagram of a deep joint source channel coding network based on first information in an embodiment of the present application;

[0041] Figure 8 It is a schematic diagram of a deep joint source channel coding network performance simulation based on SNR feedback in an embodiment of the present application;

[0042] Fig. 9 is one of the block diagrams of the transmission device according to the embodiment of the present application;

[0043] Fig.10 is a schematic diagram of a transmitting end of an embodiment of the present application;

[0044] Fig.11 This is the second block diagram of the transmission device according to the embodiment of the present application;

[0045] Fig.12 It is a block diagram of the receiving end of an embodiment of the present application. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0047] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specified order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first" and "second" are generally of the same type, and the number of objects is not limited. For example, the first object can be one or more. In addition, the "and" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0048] It is worth noting that the technology described in the embodiments of the present application is not limited to the Long Term Evolution (LTE) / LTE-Advanced (LTE-A) system, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA) and other systems. The terms "system" and "network" in the embodiments of the present application are often used interchangeably, and the described technology can be used for the systems and radio technologies mentioned above as well as for other systems and radio technologies. However, the following description describes a New Radio (NR) system for illustrative purposes, and the NR terminology is used in most of the following descriptions, although these technologies can also be applied to applications other than NR system applications, such as the 6th generation (6 th Generation, 6G) communication system.

[0049] Figure 4 A block diagram of a wireless communication system applicable to the embodiments of the present application is shown. The wireless communication system includes a transmitting end 41 and a receiving end 42, and can be applied to the compressed transmission of non-error-sensitive data such as images and videos, for example, it can be used for application scenarios such as augmented reality (AR), virtual reality (VR) transmission, and drone backhaul to the ground.

[0050] The transmitting end 41 may be a terminal, and the receiving end 42 may be a network side device, or the transmitting end 41 may be a network side device, and the receiving end 42 may be a terminal, or both the transmitting end 41 and the receiving end 42 may be terminals. The terminal may also be referred to as a terminal device or a user terminal (User Equipment, UE), and the terminal may be a mobile phone, a tablet computer (Tablet Personal Computer), a laptop computer (Laptop Computer) or a notebook computer, a personal digital assistant (Personal Digital Assistant, PDA), a handheld computer, a netbook, an ultra-mobile personal computer (ultra-mobile personal computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), a wearable device (Wearable Device) or a vehicle-mounted device (VUE), a pedestrian terminal (PUE) and other terminal side devices, and the wearable device includes: a bracelet, a headset, glasses, etc. It should be noted that the specific type of the terminal is not limited in the embodiments of the present application. The network side device can be a base station or a core network, wherein the base station can be called a node B, an evolved node B, an access point, a base transceiver station (Base Transceiver Station, BTS), a radio base station, a radio transceiver, a basic service set (Basic Service Set, BSS), an extended service set (Extended Service Set, ESS), a B node, an evolved B node (eNB), a home B node, a home evolved B node, a WLAN access point, a WiFi node, a transmitting and receiving point (Transmitting Receiving Point, TRP) or some other suitable term in the field. As long as the same technical effect is achieved, the base station is not limited to the specified technical vocabulary. It should be noted that the specific type of the base station is not limited.

[0051] The transmission method, apparatus, device and readable storage medium provided in the embodiments of the present application are described in detail below with reference to the accompanying drawings through specific embodiments and their application scenarios.

[0052] See also Figure 5 The embodiment of the present application provides a transmission method, the execution subject of the method may be a sending end, and the specific steps include:

[0053] Step 501: Acquire a first signal and first information;

[0054] In the embodiment of the present application, the first signal may include one or more of the following combinations: a text signal, an image signal, a video signal, an audio signal, etc.

[0055] Step 502: inputting the first signal and the first information into the first model, and the first model encodes or modulates the first signal according to the first information to obtain a second signal;

[0056] Step 503: Send a second signal to the receiving end;

[0057] The first information includes one or more of the following combinations:

[0058] (1) Channel State Information (CSI) of the communication channel at the transmitter and receiver;

[0059] Optionally, the CSI information includes one or more of the following combinations: channel path loss, noise, reference signal received power (RSRP), reference signal received quality (RSRQ), interference (e.g., intra-cell interference, inter-cell interference), received signal-to-noise ratio, received signal-to-interference-plus-noise ratio, frequency selection characteristics, time variability, Doppler, etc.

[0060] (2) Information collected by sensors at the sending end;

[0061] (3) Information collected by sensors at the receiving end;

[0062] Optionally, the sensor includes one or more of the following: a positioning module, a temperature sensor, a humidity sensor, a gyroscope, an acceleration sensor, a camera, a microphone, etc.

[0063] Optionally, the information collected by the sensor at the sending end or the receiving end includes one or more of the following combinations: location information (such as latitude and longitude, altitude, etc.), usage scenario information (such as indicating the usage status of the sending end or the receiving end), time information, temperature information, humidity information, audio information, video information, and image information.

[0064] Among them, the functions of the first model include one or more combinations of the following: source coding, channel coding, modulation, filtering, interleaving, scrambling, denoising, equalization, and multi-antenna signal processing.

[0065] In the embodiment of the present application, in step 501, the first information is obtained by one or more combinations of the following:

[0066] (1) obtaining first information through a reference signal;

[0067] For example, the reference signal includes one or more of the following combinations: Channel State Information Reference Signals (CSI-RS), Synchronization Signal and PBCH block (SSB), and Sounding Reference Signal (SRS).

[0068] (2) obtaining first information collected by a sensor at the sending end;

[0069] (3) obtaining first information collected by a sensor at the receiving end;

[0070] (4) Obtaining the first information through the business indicator requirements of the application layer.

[0071] Optionally, the service indicator requirement includes one or more of the following: bandwidth, resources, required received signal quality, etc.

[0072] In the embodiment of the present application, the transmitting end is a network side device, and the receiving end is a terminal. The network side device can obtain the first information reported by the terminal and collected by the sensor of the terminal through the physical layer or the high layer.

[0073] In an embodiment of the present application, the sending end is a terminal, the receiving end is a network side device, and the terminal obtains the first information sent by the network side device and collected by the sensor of the network side device through a physical layer or a high layer.

[0074] In the embodiment of the present application, in step 502, inputting the first signal and the first information into the first model includes any one of the following:

[0075] (1) directly inputting the first signal and the first information into the first model;

[0076] (2) inputting the first information and the first signal after feature extraction processing into the first model;

[0077] (3) Inputting the information of the first signal after dimension matching and the first information into the first model.

[0078] In an embodiment of the present application, the first model includes: one or more first modules, the first module is used to reallocate weights to characteristic information of the first signal, and the input of the first module includes at least: the first information and characteristic information of the first signal;

[0079] The structure of the first module includes one or more of the following:

[0080] (1) Fully connected network;

[0081] (2) Convolutional networks;

[0082] (3) Recurrent networks;

[0083] (4) Residual Network.

[0084] That is, the characteristic information and the first information are input into the first module, the characteristic information is extracted in the first module, dimension matching is performed with the first information, a weight distribution coefficient of the characteristic information is generated, and the weight of the characteristic information is reallocated.

[0085] In the embodiment of the present application, the parameters (or referred to as operating parameters) of the multiple first modules are the same, or the parameters of the multiple first modules are different.

[0086] In the embodiment of the present application, the first model further includes: one or more second modules (other modules in the first model except the first module), the input of the second module is at least part of the output of the first module, and / or the output of the second module is at least part of the input of the first module;

[0087] The structure of the second module includes one or more of the following:

[0088] (1) Fully connected network;

[0089] (2) Convolutional networks;

[0090] (3) Recurrent networks;

[0091] (4) Residual Network.

[0092] In the embodiment of the present application, the parameters (or referred to as operating parameters) of the multiple second modules are the same (ie, the multiple second modules share a set of parameters), or the parameters of the multiple second modules are different.

[0093] It is to be understood that the combination module of the first module and the second module can be reused in the first model.

[0094] In an embodiment of the present application, the functions of the first model include one or more combinations of the following: source coding, channel coding, modulation, filtering, interleaving, scrambling, denoising, equalization, and multi-antenna signal processing.

[0095] In an embodiment of the present application, the first signal is encoded or modulated by the first information to achieve adaptive channel condition transmission, which can adapt to various channel conditions without changing the network parameters, thereby reducing the storage capacity of the model parameters. Relative to the existing deep joint source channel coding technology, the embodiment of the present application constructs a deep joint source channel coding based on feedback adaptive channel state at the cost of slightly increasing the number of parameters and the amount of calculation. In engineering applications, the network performance is close to the optimal performance of deep source channel joint coding, while exponentially reducing the storage capacity required for the network and reducing the application cost of deep source channel joint coding.

[0096] See also Figure 6 The embodiment of the present application provides a transmission method, the execution subject of the method is a receiving end, and the specific steps include:

[0097] Step 601: Acquire first information and receive a second signal from a transmitting end;

[0098] Step 602: input the second signal and the first information into a second model, and decode or demodulate the second signal according to the first information through the second model to obtain a first signal;

[0099] The first information includes one or more of the following combinations:

[0100] (1) CSI information of the communication channel at the transmitter and receiver;

[0101] Optionally, the CSI information includes one or more of the following combinations: channel path loss, noise, RSRP, RSRQ, interference (intra-cell interference, inter-cell interference), received signal-to-noise ratio, received signal-to-interference-plus-noise ratio, frequency selection characteristics, time variability, Doppler, etc.

[0102] (2) Information collected by sensors at the sending end;

[0103] (3) Information collected by sensors at the receiving end.

[0104] In the embodiment of the present application, in step 601, the first information is obtained by one or more combinations of the following:

[0105] (1) obtaining the first information through a reference signal;

[0106] (2) acquiring the first information through a sensor at the sending end;

[0107] (3) acquiring the first information through a sensor at the receiving end;

[0108] (4) Obtaining the first information through the business indicator requirements of the application layer.

[0109] In an embodiment of the present application, the second model includes: one or more third modules, the third modules are used to reallocate weights to feature information of the second signal, and the input of the third modules includes at least: the first information;

[0110] The structure of the third module includes one or more of the following:

[0111] (1) Fully connected network;

[0112] (2) Convolutional networks;

[0113] (3) Recurrent networks;

[0114] (4) Residual Network.

[0115] In the embodiment of the present application, the parameters of the multiple third modules are the same, or the parameters of the multiple third modules are different.

[0116] In the embodiment of the present application, the second model further includes: one or more fourth modules, the input of the fourth module is the output of the third module, or the output of the fourth module is at least part of the input of the third module;

[0117] The structure of the fourth module includes one or more of the following:

[0118] (1) Fully connected network;

[0119] (2) Convolutional networks;

[0120] (3) Recurrent networks;

[0121] (4) Residual Network.

[0122] In the embodiment of the present application, the parameters of the multiple fourth modules are the same, or the parameters of the multiple fourth modules are different.

[0123] In an embodiment of the present application, the functions of the second model include one or more combinations of the following: source decoding, channel decoding, demodulation, filtering, interleaving, scrambling, denoising, equalization, and multi-antenna signal processing.

[0124] It is understandable that the combination module of the third module and the fourth module can be reused in the second model.

[0125] In an embodiment of the present application, the first signal is obtained by decoding or demodulating the second signal through the first information, so that the transmitting end can adapt to various channel conditions to transmit the first signal without changing the network parameters, thereby reducing the storage capacity of the model parameters.

[0126] In the embodiment of the present application, the training process of the first model and the second model is as follows:

[0127] Step 1: Select the training data set;

[0128] Step 2: Build the network structure of the encoding end, that is, build the first model and the second model;

[0129] Step 3: construct the first information;

[0130] (1) Use channel simulation data or real channel measurement data for channel modeling;

[0131] (2) collecting background information of the transmitter and / or receiver corresponding to the channel data (such as location information, temperature information, humidity information, time information, audio information, video information, image information, etc.);

[0132] Step 4: construct a transmitting end coding network and a receiving end decoding network, wherein the transmitting end coding network includes a first module and a second module, and the receiving end decoding network includes a third module and a fourth module;

[0133] Step 5: Train the network using the training data set and the first information;

[0134] Step 6: Set the training termination condition and stop training when the condition is met.

[0135] In the following, it is taken as an example that the first signal is an image signal and the first information includes: CSI and background information (such as location information, temperature information, humidity information, time information, audio information, video information, image information, etc.).

[0136] See also Figure 7 , the figure illustrates an implementation method of deep joint source-channel coding and decoding based on feedback information, the image signal is input into the convolutional network layer of the encoder (equivalent to the second module), the convolutional network layer extracts feature map information, and then the feature map information, CSI and background information are used as inputs of the feature map weight allocation layer (equivalent to the first module), and the feature map information obtained by convolution is re-weighted through the feature map weight allocation layer. It can be understood that the above steps can be repeated n times, the image is encoded to obtain a second signal, and the second signal is sent to the receiving end through the communication channel.

[0137] After the receiving end receives the second signal, the second signal is decoded by a decoder to output an image. Specifically, the convolutional network layer in the decoder (equivalent to the fourth module) extracts feature map information, and then uses the feature map information, CSI and background information as input to the feature map weight allocation layer (equivalent to the third module). The feature map information obtained by convolution is re-weighted through the feature map weight allocation layer. It can be understood that the above steps can be repeated n times, and the second signal is decoded to obtain an image signal.

[0138] See also Figure 8 , the performance of the deep joint source channel coding network based on SNR feedback and the joint source channel coding network model trained and evaluated at different SNR values ​​on the CIFAR10 dataset under the condition of additive white Gaussian noise (AWGN) channel. feedback ) represents the performance of deep joint source channel coding based on SNR feedback; DJSCC(SNR train = x dB) means that the SNR train = x dB, at SNR test ∈[0,20]dB; DJSCC(SNR singlebest ) is the SNR train = x dB training, the line connecting the performance points evaluated under the same SNR represents the best performance of the joint source channel coding network model. By comparison, DJSCC (SNR feedback ) performance in addition to SNR train =SNR test point, the performance is slightly worse than DJSCC (SNR train = xdB), and is better than DJSCC (SNR train = xdB) solution 1~10dB. DJSCC (SNR feedback ) is better than DJSCC (SNR singlebest ) performance is slightly worse by 0.1 to 0.3 dB. From the perspective of storage capacity, DJSCC (SNR feedback ) is slightly larger than DJSCC (SNR train = x dB) storage capacity required, assuming DJSCC (SNR singlebest ) saves n single-point training models, then DJSCC (SNR feedback )The storage capacity required is approximately DJSCC(SNR singlebest )1 / n of the storage capacity.

[0139] See also Fig. 9 , the embodiment of the present application provides a transmission device, applied to a transmitting end, the device 900 includes:

[0140] A first acquisition module 901, configured to acquire a first signal and first information;

[0141] A first processing module 902 is used to input the first signal and the first information into a first model, and the first model encodes or modulates the first signal according to the first information to obtain a second signal;

[0142] A sending module 903, configured to send the second signal to a receiving end;

[0143] The first information includes one or more of the following combinations: channel state information of the communication channel between the sending end and the receiving end, information collected by the sensor of the sending end, and information collected by the sensor of the receiving end.

[0144] In the embodiment of the present application, the first acquisition module 901 is further configured to acquire the first information by one or more combinations of the following:

[0145] (1) obtaining the first information through a reference signal;

[0146] (2) acquiring the first information collected by the sensor of the sending end;

[0147] (3) acquiring the first information collected by the sensor of the receiving end;

[0148] (4) Obtaining the first information through the business indicator requirements of the application layer.

[0149] In an embodiment of the present application, the first processing module 902 is further used to: directly input the first signal and the first information into the first model; or, input the first information and the first signal after feature extraction processing into the first model; or, input the information of the first signal after dimension matching and the first information into the first model.

[0150] In an embodiment of the present application, the first model includes: one or more first modules, the first modules are used to reallocate weights to the feature graphs, and the input of the first modules includes at least: the first information and feature information of the first signal;

[0151] The structure of the first module includes one or more of the following:

[0152] (1) Fully connected network;

[0153] (2) Convolutional networks;

[0154] (3) Recurrent networks;

[0155] (4) Residual Network.

[0156] In the embodiment of the present application, the parameters of the multiple first modules are the same, or the parameters of the multiple first modules are different.

[0157] In the embodiment of the present application, the first model further includes: one or more second modules, the input of the second module is at least part of the output of the first module, and / or the output of the second module is at least part of the input of the first module;

[0158] Wherein, the structure of the second module includes one or more of the following:

[0159] (1) Fully connected network;

[0160] (2) Convolutional networks;

[0161] (3) Recurrent networks;

[0162] (4) Residual Network.

[0163] In the embodiment of the present application, the parameters of the multiple second modules are the same, or the parameters of the multiple second modules are different.

[0164] In an embodiment of the present application, the functions of the first model include one or more combinations of the following: source coding, channel coding, modulation, filtering, interleaving, scrambling, denoising, equalization, and multi-antenna signal processing.

[0165] In an embodiment of the present application, the background information includes one or more of the following combinations: location information of the receiving end, location information of the sending end, usage scenario information, time information, temperature information, humidity information, audio information, video information, and image information.

[0166] The transmission device provided in the embodiment of the present application can achieve Figure 5 The various processes implemented by the method embodiment shown achieve the same technical effect and will not be described again here to avoid repetition.

[0167] like Fig.10 As shown, the embodiment of the present application also provides a transmitting end, the transmitting end 1000 includes a processor 1001, a memory 1002, a program or instruction stored in the memory 1002 and executable on the processor 1001, and the program or instruction is executed by the processor 1001 to implement the above Figure 5 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0168] The transmitting end provided in the embodiment of the present application can realize Figure 5 The various processes implemented by the method embodiment shown achieve the same technical effect and will not be described again here to avoid repetition.

[0169] See also Fig.11 , an embodiment of the present application provides a transmission device, applied to a receiving end, the device 1100 includes:

[0170] The second acquisition module 1101 is used to acquire the first information and receive the second signal from the transmitting end;

[0171] A second processing module 1102 is used to input the second signal and the first information into a second model, and the second model decodes or demodulates the second signal according to the first information to obtain a first signal;

[0172] The first information includes one or more of the following combinations: channel state information of the communication channel between the sending end and the receiving end, information collected by the sensor of the sending end, and information collected by the sensor of the receiving end.

[0173] In the embodiment of the present application, the second acquisition module 1101 acquires the first information by one or more of the following combinations:

[0174] (1) obtaining the first information through a reference signal;

[0175] (2) acquiring the first information through a sensor at the sending end;

[0176] (3) acquiring the first information through a sensor at the receiving end;

[0177] (4) Obtaining the first information through the business indicator requirements of the application layer.

[0178] In an embodiment of the present application, the second model includes: one or more third modules, the third module is used to reallocate weights to feature information of the second signal, and the input of the third module includes at least: the first information;

[0179] The structure of the third module includes one or more of the following:

[0180] (1) Fully connected network;

[0181] (2) Convolutional networks;

[0182] (3) Recurrent networks;

[0183] (4) Residual Network.

[0184] In the embodiment of the present application, the parameters of the multiple third modules are the same, or the parameters of the multiple third modules are different.

[0185] In the embodiment of the present application, the second model further includes: one or more fourth modules, the input of the fourth module is the output of the third module, or the output of the fourth module is at least part of the input of the third module;

[0186] The structure of the fourth module includes one or more of the following:

[0187] (1) Fully connected network;

[0188] (2) Convolutional networks;

[0189] (3) Recurrent networks;

[0190] (4) Residual Network.

[0191] In the embodiment of the present application, the parameters of the multiple fourth modules are the same, or the parameters of the multiple fourth modules are different.

[0192] In an embodiment of the present application, the functions of the second model include one or more combinations of the following: source decoding, channel decoding, demodulation, filtering, interleaving, scrambling, denoising, equalization, and multi-antenna signal processing.

[0193] In an embodiment of the present application, the background information includes one or more of the following combinations: location information of the sending end, location information of the receiving end, usage scenario information, time information, temperature information, humidity information, audio information, video information, and image information.

[0194] The transmission device provided in the embodiment of the present application can achieve Figure 6 The various processes implemented by the method embodiment shown achieve the same technical effect and will not be described again here to avoid repetition.

[0195] like Fig.12 As shown, the embodiment of the present application also provides a receiving end 1200, including a processor 1201, a memory 1202, a program or instruction stored in the memory 1202 and executable on the processor 1201, and the program or instruction is executed by the processor 1201 to implement the above Figure 6 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0196] The receiving end provided in the embodiment of the present application can realize Figure 6 The various processes implemented by the method embodiment shown achieve the same technical effect and will not be described again here to avoid repetition.

[0197] The embodiment of the present application also provides a program product, which is stored in a non-volatile storage medium and is executed by at least one processor to implement the following Figure 5 or Figure 6 The steps of the processing method.

[0198] The embodiment of the present application also provides a readable storage medium on which a program or instruction is stored. When the program or instruction is executed by a processor, the above Figure 5 or Figure 6 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0199] The readable storage medium includes a computer-readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0200] The present application also provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a network side device program or instruction to implement the above Figure 5 or Figure 6 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0201] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0202] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0203] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0204] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims

1. A transmission method, applied to a sending end, It is characterized in that include: Acquiring a first signal and first information; Inputting the first signal and the first information into a first model, and having the first model encode or modulate the first signal according to the first information to obtain a second signal; sending the second signal to a receiving end; The first information includes one or more of the following combinations: channel state information of the communication channel between the transmitting end and the receiving end, information collected by the sensor of the transmitting end, and information collected by the sensor of the receiving end; The first model includes: one or more first modules, the first modules are used to reallocate weights to the feature information of the first signal, and the input of the first module includes at least: the first information and the feature information of the first signal; The structure of the first module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

2. The method according to claim 1, It is characterized in that The obtaining of the first information includes one or more of the following combinations: Acquire the first information through a reference signal; Acquire the first information collected by the sensor of the sending end; Acquire the first information collected by the sensor of the receiving end; The first information is obtained through the business indicator requirements of the application layer.

3. The method according to claim 2, It is characterized in that The sending end is a network side device, the receiving end is a terminal, and the acquiring of the first information collected by the sensor of the receiving end includes: Acquire, through a physical layer or a higher layer, first information reported by the terminal and collected by a sensor of the terminal; or, The sending end is a terminal, the receiving end is a network side device, and the obtaining of the first information collected by the sensor of the receiving end includes: obtaining the first information sent by the network side device and collected by the sensor of the network side device through a physical layer or a high layer.

4. The method according to claim 1, It is characterized in that Inputting the first signal and the first information into a first model comprises: directly inputting the first signal and the first information into the first model; or, Inputting the first information and the first signal after feature extraction processing into the first model; or, The dimensionally matched information of the first signal and the first information are input into the first model.

5. The method according to claim 1, It is characterized in that The parameters of the multiple first modules are the same, or the parameters of the multiple first modules are different.

6. The method according to claim 1, It is characterized in that The first model further includes: one or more second modules, the input of the second module is at least part of the output of the first module, and / or the output of the second module is at least part of the input of the first module; Wherein, the structure of the second module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

7. The method according to claim 6, It is characterized in that The parameters of the multiple second modules are the same, or the parameters of the multiple second modules are different.

8. The method according to claim 1, It is characterized in that The functions of the first model include one or more of the following combinations: source coding, channel coding, modulation, filtering, interleaving, scrambling, denoising, equalization, and multi-antenna signal processing.

9. The method according to claim 1, Features ,The information collected by the sensor at the sending end or the receiving end includes one or more combinations of the following: location information, usage scenario information, time information, temperature information, humidity information, audio information, video information, and image information.

10. A transmission method, applied to a receiving end, It is characterized in that include: Acquiring first information and receiving a second signal from a transmitting end; Inputting the second signal and the first information into a second model, and having the second model decode or demodulate the second signal according to the first information to obtain a first signal; The first information includes one or more of the following combinations: channel state information of the communication channel between the transmitting end and the receiving end, information collected by the sensor of the transmitting end, and information collected by the sensor of the receiving end; The second model includes: one or more third modules, the third modules are used to reallocate weights to the feature information of the second signal, and the input of the third modules includes at least: the first information; The structure of the third module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

11. The method according to claim 10, It is characterized in that The obtaining of the first information includes one or more of the following combinations: Acquire the first information through a reference signal; Acquire the first information through the sensor of the sending end; Acquire the first information through a sensor at the receiving end; The first information is obtained through the business indicator requirements of the application layer.

12. The method according to claim 10, It is characterized in that The parameters of the plurality of third modules are the same, or the parameters of the plurality of third modules are different.

13. The method according to claim 10, It is characterized in that The second model further includes: one or more fourth modules, the input of the fourth module is at least part of the output of the third module, or the output of the fourth module is at least part of the input of the third module; The structure of the fourth module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

14. The method according to claim 13, It is characterized in that The parameters of the plurality of fourth modules are the same, or the parameters of the plurality of fourth modules are different.

15. The method according to claim 10, It is characterized in that The functions of the second model include one or more of the following combinations: source decoding, channel decoding, demodulation, filtering, interleaving, scrambling, denoising, equalization, and multi-antenna signal processing.

16. The method according to claim 10, It is characterized in that The information collected by the sensor at the transmitting end or the receiving end includes one or more of the following combinations: location information, usage scenario information, time information, temperature information, humidity information, audio information, video information, and image information.

17. A transmission device, applied to a transmitting end, It is characterized in that include: A first acquisition module, used to acquire a first signal and first information; A first processing module, configured to input the first signal and the first information into a first model, and the first model encodes or modulates the first signal according to the first information to obtain a second signal; A sending module, used for sending the second signal to a receiving end; The first information includes one or more of the following combinations: channel state information of the communication channel between the transmitting end and the receiving end, information collected by the sensor of the transmitting end, and information collected by the sensor of the receiving end; The first model includes: one or more first modules, the first modules are used to reallocate weights to the feature information of the first signal, and the input of the first module includes at least: the first information and the feature information of the first signal; The structure of the first module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

18. The transmission device according to claim 17, It is characterized in that The first acquisition module is further configured to acquire the first information by one or more combinations of the following: Acquire the first information through a reference signal; Acquire the first information collected by the sensor of the sending end; Acquire the first information collected by the sensor of the receiving end; The first information is obtained through the business indicator requirements of the application layer.

19. The transmission device according to claim 17, It is characterized in that The first processing module is further used to: directly input the first signal and the first information into the first model; or, input the first information and the first signal after feature extraction processing into the first model; or, input the information of the first signal after dimension matching and the first information into the first model.

20. The transmission device according to claim 17, It is characterized in that The first model further comprises: one or more second modules, the input of the second module is at least part of the output of the first module, and / or the output of the second module is at least part of the input of the first module; Wherein, the structure of the second module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

21. A transmission device, applied to a receiving end, It is characterized in that include: A second acquisition module, used to acquire the first information and receive a second signal from the transmitting end; A second processing module, used for inputting the second signal and the first information into a second model, and the second model decodes or demodulates the second signal according to the first information to obtain a first signal; The first information includes one or more of the following combinations: channel state information of the communication channel between the transmitting end and the receiving end, information collected by the sensor of the transmitting end, and information collected by the sensor of the receiving end; The second model includes: one or more third modules, the third modules are used to reallocate weights to the feature information of the second signal, and the input of the third modules includes at least: the first information; The structure of the third module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

22. The transmission device according to claim 21, It is characterized in that The second acquisition module acquires the first information by one or more of the following combinations: Acquire the first information through a reference signal; Acquire the first information through the sensor of the sending end; Acquire the first information through a sensor at the receiving end; The first information is obtained through the business indicator requirements of the application layer.

23. The transmission device according to claim 21, It is characterized in that The second model further includes: one or more fourth modules, the input of the fourth module is the output of the third module, or the output of the fourth module is at least part of the input of the third module; The structure of the fourth module includes one or more of the following: Fully connected network; Convolutional networks; Recurrent networks; Residual Network.

24. A communication device, It is characterized in that include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of the transmission method according to any one of claims 1 to 16 are implemented.

25. A readable storage medium, It is characterized in that The readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the steps of the transmission method according to any one of claims 1 to 16 are implemented.

Images