Data transmission method and communication apparatus

Abstract

The present application provides a data transmission method and a communication apparatus. The method comprises: a sending end device determines data decompression precision corresponding to a first data compression result, wherein the first data compression result is obtained by compressing first service data, the data decompression precision is obtained on the basis of the first service data and first decompressed data, and the first decompressed data is obtained on the basis of the first data compression result; when the data decompression precision is greater than or equal to a first precision threshold corresponding to the first service data, the sending end device sends the first data compression result; and when the data decompression precision is less than the first precision threshold, the sending end device sends the first data compression result and a first residual compression result corresponding to the first data compression result. By means of such a method, the flexibility of the data decompression precision of a compression solution is improved, thereby improving the adaptation degree between the data decompression precision of the compression solution and the requirements of a current actual service scenario.

Description

A data transmission method and communication device

[0001] This application claims priority to Chinese Patent Application No. 202410974014.7, filed on July 19, 2024, entitled "A Data Transmission Method and Communication Device", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of communication technology, and in particular to a data transmission method and communication device. Background Technology

[0003] In communication systems, the raw data acquired by the transmitting device (which can be understood as the data to be transmitted) is typically characterized by large data volume and redundancy / correlation. To reduce the data transmission volume between the transmitting and receiving devices, data compression technology can be introduced during communication. The communication process with data compression technology can be roughly described as including the following steps: S1, the transmitting device inputs the raw data into an encoder (hereinafter referred to as a data compression model) to obtain the data characteristics of the raw data; S2, after performing operations such as quantization, channel coding, and modulation on the data characteristics, the transmitting device sends a wireless signal carrying the data characteristics to the receiving device; S3, after receiving the wireless signal, the receiving device performs operations such as demodulation, channel decoding, and inverse quantization on the wireless signal to obtain the data characteristics; S4, the receiving device inputs the data characteristics into a decoder (hereinafter referred to as a data decompression model) to obtain decompressed data. It can be understood that the closer the decompressed data in S4 is to the raw data in S1 (or the more similar it is understood), the higher the data decompression accuracy of the compression scheme (or the better the compression reconstruction performance is understood) of the compression scheme (or the better the compression reconstruction performance is understood).

[0004] It is important to understand that changes in communication service scenarios or the environment in which native data is acquired may cause fluctuations in the data decompression accuracy of the compression scheme, which may further prevent the compression scheme from meeting the data decompression accuracy requirements of actual business scenarios. Summary of the Invention

[0005] This application provides a data transmission method and communication device, which is beneficial to improving the flexibility of data decompression accuracy of compression schemes, thereby improving the adaptability of data decompression accuracy of compression schemes to the needs of current actual business scenarios.

[0006] In a first aspect, this application provides a data transmission method. This method is applied to a transmitting device or a module (e.g., a chip or chip system) within the transmitting device. Taking the application to a transmitting device as an example, the method includes: the transmitting device determining a data decompression precision corresponding to a first data compression result, wherein the first data compression result is obtained based on the compression of first service data, and the data decompression precision is obtained based on the first service data and the first decompressed data, wherein the first decompressed data is obtained based on the decompression of the first data compression result. Further, when the data decompression precision is greater than or equal to a first precision threshold, the first data compression result is transmitted, wherein the first precision threshold is associated with the service type corresponding to the first service data; when the data decompression precision is less than the first precision threshold, the first data compression result and a first residual compression result are transmitted, wherein the first residual compression result is obtained based on compression of first residual information, and the first residual information is obtained based on the first service data and the first decompressed data.

[0007] Based on the method described in the first aspect, after the sending device acquires the raw data of the actual business scenario (i.e., the first business data mentioned in this application), it compresses the first business data to obtain a first data compression result, and determines the data decompression accuracy based on the first data compression result. Further, the sending device makes a judgment based on the data decompression accuracy requirements of the actual business scenario (i.e., the first accuracy threshold mentioned in this application). If the data decompression accuracy corresponding to the first data compression result meets the data decompression accuracy requirements of the actual business scenario, the sending device sends the first data compression result to the receiving device. If the data decompression accuracy corresponding to the first data compression result does not meet the data decompression accuracy requirements of the actual business scenario, the sending device sends the first data compression result and the corresponding first residual compression result to the receiving device. It should be understood that the data decompression accuracy decompressed by the receiving device using the first data compression result and the first residual compression result will be higher than the data decompression accuracy decompressed by the receiving device using only the first data compression result. The method provided in the first aspect can be used to adjust the data decompression accuracy of the compression scheme, thereby improving the adaptability of the data decompression accuracy of the compression scheme to the needs of the current actual business scenario.

[0008] In one possible implementation, the transmitting device receives a first precision threshold corresponding to the first service data.

[0009] By implementing this possible implementation, the receiving device (which can be understood as the user of the first business data) determines the actual business scenario's requirements for data decompression accuracy, which helps improve the adaptability of the compression scheme's data decompression accuracy to the current actual business scenario's requirements. Furthermore, it also helps save computing resources on the sending device.

[0010] In one possible implementation, the first service data corresponds to P precision thresholds, where P is a positive integer, and the first precision threshold is the maximum value among the P precision thresholds. Each of the P precision thresholds corresponds one-to-one with a compression ratio. Specifically, precision threshold #1 and precision threshold #2 are any two of the P precision thresholds, where precision threshold #1 is greater than precision threshold #2, and the compression ratio of precision threshold #1 is higher than that of precision threshold #2.

[0011] It is important to understand that the compression ratio mentioned in this application refers to the ratio between the data volume of the first business data and the data volume of the compressed result (i.e., the sum of the data volume of the first data compression result and the data volume of the first residual compression result). With the data volume of the first data compression result remaining constant, the data volume of the first residual compression result can be adjusted by adjusting the compression ratio, and the minimum data volume of the first residual compression result is 0. A business can correspond to multiple precision thresholds; a higher precision threshold corresponds to a higher compression ratio, and a higher compression ratio results in a smaller data volume of the compressed result.

[0012] In one possible implementation, the P precision thresholds correspond to P precision intervals, and each of the P precision intervals corresponds one-to-one with one of the P compression rates. In this case, the transmitting device compresses the first residual information based on the compression parameters corresponding to the P compression rates, respectively, to obtain P candidate residual compression results. Further, the transmitting device determines the first residual compression result from the P candidate residual compression results based on the data decompression precision. The first residual compression result is obtained by compression based on the compression rate corresponding to the first precision interval, where the first precision interval is the interval in which the data decompression precision falls within the P precision intervals.

[0013] By implementing this possible implementation, a business scenario can correspond to multiple precision thresholds, and multiple precision ranges can be determined based on these multiple precision thresholds. Further, the sending device can determine a target compression ratio based on the precision range where the data decompression precision of the first data compression result falls, and determine the candidate residual compression result obtained based on the target compression ratio as the first residual compression result. It is understood that when the data decompression precision of the first data compression result does not meet the actual business scenario's requirements for data decompression precision (i.e., it is less than the first precision threshold), if the gap between the data compression precision of the first data compression result and the actual business scenario's requirements for data decompression precision is larger (i.e., the difference between the data decompression precision of the first data compression result and the first precision threshold is larger), then the compression ratio of the first residual result corresponding to the first data compression result will be smaller, meaning the first residual result contains more information, thereby improving the adaptability of the compression scheme's data decompression precision to the requirements of the current actual business scenario.

[0014] In one possible implementation, the first residual information includes N residual data, where N is a positive integer. Candidate residual compression result #1 is the candidate residual compression result corresponding to the precision threshold #1, and candidate residual compression result #2 is the candidate residual compression result corresponding to the precision threshold #2. In this case, the quantization bit width of the N residual data included in candidate residual compression result #1 is smaller than the quantization bit width of the N residual data included in candidate residual compression result #2; or, the number of residual data included in candidate residual compression result #1 is smaller than the number of residual data included in candidate residual compression result #2. By implementing this possible implementation, a larger precision threshold corresponds to a higher compression ratio, and the corresponding candidate residual compression result has a smaller quantization bit width or a smaller number of residual data. When the data compression precision of the first data compression result is small compared to the actual business scenario's requirement for data decompression precision (i.e., the difference between the data decompression precision of the first data compression result and the first precision threshold is small), the data volume of the first residual compression result is small, which helps save communication resources.

[0015] In one possible implementation, the first residual information includes N residual data points, where N is a positive integer. In this case, the transmitting device filters the first residual information to obtain second residual information, which includes N1 residual data points from the N residual data points, where N1 is a positive integer less than or equal to N. Further, the transmitting device compresses the second residual information to obtain the first residual compression result. By implementing this possible implementation, after obtaining the first residual information, the transmitting device compresses and transmits a portion of the residual data in the first residual information, thereby reducing the data volume of the first residual compression result and saving communication resources.

[0016] In one possible implementation, the transmitting device filters the first residual information according to a residual threshold to obtain the second residual information, wherein all N1 residual data included in the second residual information are greater than or equal to the residual threshold. The residual threshold is one of the N residual data, or it can be a statistical representation of the N residual data. By implementing this possible implementation, after obtaining the first residual information, the transmitting device filters out the larger portion of residual data from the residual data included in the first residual information for compression and transmission. This helps reduce the amount of data in the first residual compression result while also improving the data decompression accuracy based on the first residual compression result and the first data compression result.

[0017] In one possible implementation, the N residual data points correspond one-to-one with N index information points. The transmitting device sends first indication information, which indicates the index information of the N1 residual data points. By implementing this possible implementation, the transmitting and receiving devices align the residual data included and / or not included in the first residual compression result, which helps to improve the data decompression accuracy.

[0018] In one possible implementation, the transmitting device performs data recombination processing on the residual data included in the first residual information to obtain the third residual information. The data recombination processing includes one or more of matrix straightening processing, splicing processing, and vector segmentation processing. Further, the transmitting device compresses the third residual information to obtain the first residual compression result.

[0019] In one possible implementation, the transmitting device sends a second indication message, which indicates the operating parameters corresponding to the data reassembly process. By implementing this possible implementation, the processing procedures of the transmitting and receiving devices for the first residual compression result are aligned, which helps to improve the data decompression accuracy.

[0020] In one possible implementation, the transmitting device inputs the first residual information and the first decompressed data into an artificial intelligence (AI) residual encoder to obtain the first residual compression result.

[0021] Secondly, this application provides a data transmission method, which is applied to a receiving device or a module (e.g., a chip or chip system) within the receiving device. Taking the application to a receiving device as an example, the method includes: the receiving device receiving a first data compression result and a first residual compression result, wherein the first data compression result is obtained based on the compression of first service data, the first residual compression result is obtained based on the compression of first residual information, the first residual information is obtained based on the first service data and first decompressed data, and the first decompressed data is obtained by decompressing based on the first data compression result. Further, the receiving device decompresses the data based on the first data compression result and the first residual compression result to obtain second decompressed data.

[0022] Compared to the method of obtaining decompressed data solely based on the first data compression result, the method described in the second aspect, where the receiving device obtains decompressed data based on the first data compression result and the corresponding first residual compression result, is beneficial for improving data decompression accuracy. It should be noted that for the beneficial effects of other embodiments of the second aspect, please refer to the relevant descriptions in the first aspect.

[0023] In one possible implementation, the first service data corresponds to P precision thresholds, where P is a positive integer, the first precision threshold is the maximum value among the P precision thresholds, and the P precision thresholds correspond one-to-one with P compression rates; wherein, precision threshold #1 and precision threshold #2 are any two precision thresholds among the P precision thresholds, precision threshold #1 is greater than precision threshold #2, and the compression rate of precision threshold #1 is less than the compression rate of precision threshold #2.

[0024] In one possible implementation, the first residual information includes N residual data points, and the first data compression result is obtained based on N1 residual data points out of the N residual data points. In this case, the receiving device receives first indication information, which is used to indicate the index information of the N1th residual data point.

[0025] In one possible implementation, the first data compression result is obtained by reassembling the residual data included in the first residual information. In this case, the receiving device receives second indication information, which indicates the operating parameters corresponding to the data reassembly process.

[0026] Thirdly, this application provides a communication device, which can be a transmitting device, a device within a transmitting device, or a device compatible with a transmitting device. The communication device can also be a chip system. The communication device can execute the method described in the first aspect. The functions of the communication device can be implemented in hardware or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the above functions. The unit or module can be software and / or hardware. The operations performed by the communication device and its beneficial effects can be found in the method described in the first aspect and its beneficial effects.

[0027] Fourthly, this application provides a communication device, which can be a receiving device, a device within a transmitting device, or a device compatible with a transmitting device. The communication device can also be a chip system. The communication device can execute the method described in the second aspect. The functions of the communication device can be implemented in hardware or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the above functions. These units or modules can be software and / or hardware. The operations performed by the communication device and its beneficial effects can be found in the method described in the second aspect above.

[0028] Fifthly, this application provides a communication device including a processor and an interface circuit. The interface circuit is configured to receive signals from other communication devices outside the communication device and transmit them to the processor, or to send signals from the processor to other communication devices outside the communication device. The processor is configured to implement the method described in the first aspect through logic circuits or executable code instructions, or the processor is configured to implement the method described in the second aspect through logic circuits or executable code instructions.

[0029] Sixthly, this application provides a communication device including a processor connected to a memory for calling computer programs or instructions stored in the memory to execute the methods described in the first or second aspect. The memory may be located within or outside the network device or the first terminal device. The processor may include one or more processors.

[0030] In a seventh aspect, this application provides a computer-readable storage medium storing a computer program or instructions that, when executed by a communication device, implement the method described in the first aspect or the method described in the second aspect.

[0031] Eighthly, this application provides a computer program product including a computer program or instructions, which, when read and executed by a communication device, causes the communication device to perform the method described in the first aspect, or causes the communication device to perform the method described in the second aspect.

[0032] Ninthly, this application provides a communication system, including a communication device for performing the method described in the first aspect and a communication device for performing the method described in the second aspect. Attached Figure Description

[0033] Figure 1 is a schematic diagram of the architecture of a communication system provided in an embodiment of this application;

[0034] Figure 2 is another schematic diagram of a wireless communication system applicable to an embodiment of this application;

[0035] Figure 3 is a schematic diagram of a compression scheme provided in an embodiment of this application;

[0036] Figure 4 is a flowchart illustrating a data transmission method provided in an embodiment of this application;

[0037] Figure 5 is a schematic diagram of a data compression model and a data decompression model provided in an embodiment of this application;

[0038] Figure 6 is a schematic diagram of obtaining residual compression results according to an embodiment of this application;

[0039] Figure 7 is a flowchart illustrating another data transmission method provided in an embodiment of this application;

[0040] Figure 8 is a schematic diagram of obtaining candidate residual compression results according to an embodiment of this application;

[0041] Figure 9 is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0042] Figure 10 is a schematic diagram of another communication device provided in an embodiment of this application. Detailed Implementation

[0043] To facilitate a detailed understanding of the embodiments of this application, the system architecture involved in the embodiments of this application will be described below.

[0044] Figure 1 is a schematic diagram of the architecture of the communication system applied in the embodiments of this application. As shown in Figure 1, the communication system includes a radio access network (RAN) 100 and a core network 200. Optionally, the communication system may also include an Internet 300. RAN 100 includes at least one RAN node (110a and 110b in Figure 1, collectively referred to as 110) and may also include at least one terminal (120a-120j in Figure 1, collectively referred to as 120). RAN 100 may also include other RAN nodes, such as wireless relay devices and / or wireless backhaul devices (not shown in Figure 1). Terminal 120 is wirelessly connected to RAN node 110, and RAN node 110 is wirelessly or wiredly connected to the core network 200. The core network equipment in the core network 200 and the RAN node 110 in the RAN 100 can be independent and different physical devices, or they can be the same physical device integrating the logical functions of the core network equipment and the logical functions of the RAN node. Terminals and RAN nodes can be interconnected via wired or wireless means.

[0045] RAN100 can be an evolved universal terrestrial radio access (E-UTRA) system, a new radio (NR) system, or a future radio access system as defined in the 3rd generation partnership project (3GPP). RAN100 can also include two or more of the above-mentioned different radio access systems. RAN100 can also be an open RAN (O-RAN).

[0046] RAN nodes, also known as radio access network devices, RAN entities, or access nodes (hereinafter referred to as network devices), are used to help terminals access communication systems wirelessly. In one application scenario, an RAN node can be a base station, an evolved NodeB (eNodeB), a transmission reception point (TRP), a next-generation NodeB (gNB) in a 5G mobile communication system, a next-generation base station in a 6G mobile communication system, or a base station in a future mobile communication system. RAN nodes can be macro base stations (as shown in Figure 1, 110a), micro base stations or indoor stations (as shown in Figure 1, 110b), relay nodes, or donor nodes.

[0047] In another application scenario, multiple RAN nodes can collaborate to help terminals achieve wireless access, with different RAN nodes implementing different functions of the base station. For example, a RAN node can be a central unit (CU), a distributed unit (DU), or a radio unit (RU). Here, the CU performs the functions of the base station's Radio Resource Control (RRC) and Packet Data Convergence Protocol (PDCP), and can also perform the functions of the Service Data Adaptation Protocol (SDAP). The DU performs the functions of the base station's Radio Link Control (RANC) and Medium Access Control (MAC) layers, and can also perform some or all of the physical layer functions. For specific descriptions of these protocol layers, refer to the relevant 3GPP technical specifications. The RU can be used to implement radio frequency signal transmission and reception. The CU and DU can be two independent RAN nodes or integrated into the same RAN node, such as within a baseband unit (BBU). The RU can be included in radio frequency equipment, such as in a remote radio unit (RRU) or an active antenna unit (AAU). The CU can be further divided into two types of RAN nodes: CU-control plane and CU-user plane.

[0048] RAN nodes can support one or more types of fronthaul interfaces. Different fronthaul interfaces correspond to DUs and RUs with different functions. If the fronthaul interface between the DU and RU is a common public radio interface (CPRI), the DU is configured to implement one or more baseband functions, and the RU is configured to implement one or more radio frequency functions. If the fronthaul interface between the DU and RU is an enhanced common public radio interface (eCPRI), compared to CPRI, some downlink and / or uplink baseband functions are moved from the DU to the RU. Different splitting methods between DUs and RUs correspond to different types (category, Cat) of eCPRI, such as eCPRI Cat A, B, C, D, E, F.

[0049] Taking eCPRI Cat A as an example, for downlink transmission, the DU is configured to implement one or more functions before and after layer mapping (i.e., coding, rate matching, scrambling, modulation, and layer mapping), while other functions after layer mapping (e.g., resource element (RE) mapping, digital beamforming (BF), or one or more functions of inverse fast Fourier transform (IFFT) / adding cyclic prefix (CP)) are moved to the RU. For uplink transmission, the DU is configured to implement one or more functions before and after de-mapping (i.e., decoding, rate matching de-mapping, descrambling, demodulation, inverse discrete Fourier transform (IDFT), channel equalization, and de-RE mapping), while other functions after de-mapping (e.g., digital BF or one or more functions of fast Fourier transform (FFT) / removing CP) are moved to the RU. It is understandable that the functional descriptions of the DU and RU corresponding to various types of eCPRI can be found in the eCPRI protocol, and will not be elaborated here.

[0050] In one possible design, the processing unit in the BBU used to implement baseband functions is called the baseband high (BBH) unit, and the processing unit in the RRU / AAU / RRH used to implement baseband functions is called the baseband low (BBL) unit.

[0051] In different systems, RAN nodes may have different names. For example, in an O-RAN system, a CU can be called an open CU (O-CU), a DU can be called an open DU (O-DU), and an RU can be called an open RU (O-RU). The RAN nodes in the embodiments of this application can be implemented through software modules, hardware modules, or a combination of software and hardware modules. For example, a RAN node can be a server loaded with the corresponding software modules. The embodiments of this application do not limit the specific technology or device form used in the RAN nodes. For ease of description, a base station is used as an example of a RAN node in the following description.

[0052] A terminal is a device with wireless transceiver capabilities, capable of sending signals to or receiving signals from a base station. Terminals can also be called terminal equipment, user equipment (UE), mobile station, mobile terminal, etc. Terminals can be widely used in various scenarios, such as device-to-device (D2D), vehicle-to-everything (V2X) communication, machine-type communication (MTC), Internet of Things (IoT), virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grids, smart furniture, smart offices, smart wearables, smart transportation, smart cities, etc. Terminals can be mobile phones, tablets, computers with wireless transceiver capabilities, wearable devices, vehicles, airplanes, ships, robots, robotic arms, smart home devices, etc. The embodiments of this application do not limit the specific technology or device form used in the terminal.

[0053] Base stations and terminals can be fixed or mobile. They can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; they can also be deployed on water; and they can be deployed on aircraft, balloons, and satellites. The embodiments of this application do not limit the application scenarios of the base stations and terminals.

[0054] The roles of base stations and terminals can be relative. For example, the helicopter or drone 120i in Figure 1 can be configured as a mobile base station. For terminals 120j that access the wireless access network 100 through 120i, terminal 120i is a base station; however, for base station 110a, 120i is a terminal, meaning that 110a and 120i communicate via a wireless air interface protocol. Of course, 110a and 120i can also communicate via a base station-to-base station interface protocol. In this case, relative to 110a, 120i is also a base station. Therefore, both base stations and terminals can be collectively referred to as communication devices. 110a and 110b in Figure 1 can be called communication devices with base station functions, and 120a-120j in Figure 1 can be called communication devices with terminal functions.

[0055] Communication between base stations and terminals, between base stations, and between terminals can be conducted using licensed spectrum, unlicensed spectrum, or both simultaneously. Communication can be conducted using spectrum below 6 GHz, spectrum above 6 GHz, or both simultaneously. The embodiments of this application do not limit the spectrum resources used for wireless communication.

[0056] In the embodiments of this application, the functions of the base station can be executed by modules (such as chips) within the base station, or by a control subsystem that includes base station functions. This control subsystem, including base station functions, can be a control center in the aforementioned application scenarios such as smart grids, industrial control, intelligent transportation, and smart cities. Similarly, the functions of the terminal can be executed by modules (such as chips or modems) within the terminal, or by a device that includes terminal functions.

[0057] Please refer to Figure 2, which is another schematic diagram of a wireless communication system applicable to embodiments of this application.

[0058] As shown in Figure 2, the wireless communication system includes a RAN intelligent controller (RIC). As an example, the RIC can be used to implement artificial intelligence (AI) related functions. As an example, the RIC includes near-real-time (near-RT) RICs and non-real-time (non-RT) RICs. Non-real-time RICs primarily process non-real-time information, such as data that is not sensitive to latency, with latency in the order of seconds. Real-time RICs primarily process near-real-time information, such as data that is relatively sensitive to latency, with latency in the order of tens of milliseconds.

[0059] The near real-time RIC is used for model training and inference. For example, it can be used to train an AI model and then use that AI model for inference. The near real-time RIC can obtain network-side and / or terminal-side information from RAN nodes (e.g., CU, CU-CP, CU-UP, DU, and / or RU) and / or terminals. This information can be used as training data or inference data. Optionally, the near real-time RIC can deliver inference results to RAN nodes and / or terminals. Optionally, inference results can be exchanged between CU and DU, and / or between DU and RU. For example, the near real-time RIC delivers the inference result to the DU, and the DU sends it to the RU.

[0060] The non-real-time RIC is also used for model training and inference. For example, it can be used to train an AI model and then use that model for inference. The non-real-time RIC can obtain network-side and / or terminal-side information from RAN nodes (e.g., CU, CU-CP, CU-UP, DU, and / or RU) and / or terminals. This information can be used as training data or inference data, and the inference results can be delivered to the RAN nodes and / or terminals. Optionally, inference results can be exchanged between CU and DU, and / or between DU and RU. For example, the non-real-time RIC delivers the inference results to the DU, which then forwards them to the RU.

[0061] The near real-time RIC and non-real-time RIC can also be configured as separate network elements. Optionally, the near real-time RIC and non-real-time RIC can also be part of other devices. For example, the near real-time RIC can be set in the RAN node (e.g., in CU, DU), while the non-real-time RIC can be set in the network management (OAM) system, cloud server, core network device, or other network device.

[0062] In practical applications, this wireless communication system can include multiple network devices (also known as access network devices) and multiple terminal devices simultaneously, without limitation. A network device can serve one or more terminal devices simultaneously. A terminal device can also access one or more network devices simultaneously. This application embodiment does not limit the number of terminal devices and network devices included in the wireless communication system.

[0063] To facilitate understanding of the relevant content of the embodiments of this application, some terms involved in the embodiments of this application will be explained below. This part is only for the purpose of understanding and should not be regarded as a disclosure or specific limitation of the technical solution of this application.

[0064] 1. AI Model

[0065] AI models, also known as neural networks, can be composed of neural units. A neural unit can refer to a unit represented by x. s The input is the arithmetic unit, and the output of the arithmetic unit can be as shown in formula (1).

[0066] Where s = 1, 2, ..., n, n is a natural number greater than 1, W s For x s The weights are denoted by b, where b is the bias of the neural unit. f is the activation function of the neural unit, used to introduce nonlinear characteristics into the neural network to convert the input signal in the neural unit into an output signal. The output signal of this activation function can be used as the input to the next convolutional layer. The activation function can be the sigmoid function. A neural network is a network formed by connecting many of the above-mentioned individual neural units together, that is, the output of one neural unit can be the input of another neural unit. The input of each neural unit can be connected to the local receptive field of the previous layer to extract the features of the local receptive field, which can be a region composed of several neural units.

[0067] It should be noted that the AI ​​model mentioned in this application may be one or more of the following: neural network model, deep neural network (DNN) network model, convolutional neural network (CNN) network model, recurrent neural network (RNN) network model, generative adversarial network model, residual network (ResNet) network model, or a combination thereof. This application does not impose any specific limitations on this.

[0068] 2. Compression scheme

[0069] A compression scheme includes a data compression model for compressing data and a data decompression model for decompressing the compressed data (referred to as the compression result in this application). As shown in Figure 3(a), the transmitting device inputs the data to be transmitted into the data compression model to obtain data features (also referred to as the compression result in this application). Then, the transmitting device quantizes the compression result to obtain a binary code stream. After channel coding and modulation of the binary code stream, the transmitting device transmits it to the receiving device through a wireless channel, that is, the transmitting device sends a wireless signal carrying the data features to the receiving device. After receiving the wireless signal, the receiving device demodulates and decodes the wireless signal to obtain a binary code stream, and then performs inverse quantization on the binary code stream to obtain the compression result. Further, the receiving device inputs the compression result into the data decompression model to obtain decompressed data.

[0070] It should be noted that the data compression model (or data decompression model) mentioned in this application can be an AI model or an encoder other than an AI model, and this application does not make any specific limitations on it.

[0071] The following explanation uses an AI model, specifically a CNN or ResNet network model, as an example to illustrate the data compression process. As shown in Figure 3(b), the data to be sent undergoes multi-layer CNN or ResNet computation to obtain the result. This result is then flattened to obtain a vector. Finally, it is compressed to a specific dimension using a fully connected neural network (FCN) to obtain the data feature. This data feature is typically a floating-point number.

[0072] Due to changes in communication service scenarios or the environment in which the sending device acquires native data, the data decompression accuracy of the same compression scheme may fluctuate across different communication service scenarios. This could lead to the compression scheme's data decompression accuracy failing to meet the requirements of actual business scenarios. Furthermore, different communication service scenarios may have different requirements for the data decompression accuracy of the communication system, which could also cause the compression scheme's data decompression accuracy to fail to meet the requirements of actual business scenarios.

[0073] To improve the data decompression accuracy of the compression scheme and its adaptability to actual business scenarios, and to ensure that the data decompression accuracy of the compression scheme meets the needs of actual business scenarios, this application provides several data transmission methods and communication devices. The data transmission methods and communication devices provided in this application will be further described below with reference to the accompanying drawings.

[0074] Please refer to Figure 4, which is a flowchart illustrating a data transmission method provided in an embodiment of this application. As shown in Figure 4, the data transmission method includes the following steps S401 to S405. The method execution entities shown in Figure 4 can be a transmitting device and a receiving device, or the method execution entities shown in Figure 4 can be modules in the transmitting device and modules in the receiving device, or the method execution entities shown in Figure 4 can be chips in the transmitting device and chips in the receiving device. Figure 4 uses the transmitting device and the receiving device as examples of the method execution entities. The transmitting device can be the RAN node shown in Figure 1 or Figure 2, or the terminal shown in Figure 1 or Figure 2; the receiving device mentioned in this application can be the RAN node shown in Figure 1 or Figure 2, or the terminal shown in Figure 1 or Figure 2; this application does not impose specific limitations.

[0075] S401. The sending device determines the data decompression accuracy corresponding to the first data compression result.

[0076] The first data compression result is obtained based on the compression of the first business data, the data decompression accuracy is obtained based on the first business data and the first decompressed data, and the first decompressed data is obtained based on the decompression of the first data compression result.

[0077] In other words, the sending device acquires first service data in a real-world business scenario and inputs this first service data into its data compression model to obtain a first data compression result. Further, the sending device decompresses this first data compression result using its data decompression model to obtain first decompressed data. The sending device then calculates the data decompression accuracy based on the first service data and the first decompressed data. This data decompression accuracy indicates the similarity or distance between the first decompressed data and the first service data. The closer the first decompressed data is to the first service data, the higher the data decompression accuracy; conversely, the further apart they are, the lower the data decompression accuracy.

[0078] It should be noted that the data compression model mentioned in this application can be an AI model or other encoders (such as entropy encoders or differential encoders), and the data decompression model mentioned in this application can also be an AI model or other decoders (such as entropy decoders or differential decoders). For ease of understanding, this application also provides an implementation where both the data compression model and the data decompression model are AI models, as shown in Figure 5.

[0079] Below, taking the first business data with dimensions n1×n2×4 as an example, we will explain the data compression model and data decompression model in Figure 5. When this first business data is input into the data compression model, it will be processed through the first layer of the ResNet network (i.e., ResNet1 in Figure 5), resulting in a data with dimensions n1×n2×4. The data features are then processed through a second-layer ResNet network (i.e., ResNet2 in Figure 5) to obtain a dimension of... The data characteristics; then straighten the data characteristics into The vector, and after The FCN process yields a first data compression result of length N. Further, this first data compression result of length N is input into a data decompression model, which then undergoes... The FCN processing yields a length of The vector is then reshaped to obtain a vector with dimensions of... The data features are then upsampled through the first layer of the ResNet upsampling network (i.e., ResNet upsampling 1 in Figure 5) to obtain a dimension of The data features are then processed by the second layer ResNet upsampling network (i.e., ResNet upsampling 2 in Figure 5) to obtain the first decompressed data with dimensions n1×n2×4.

[0080] It should be noted that the two ResNet layers in the data compression model have the same structure. Taking ResNet2 in the data compression model in Figure 5 as an example, the specific structure of the ResNet network will be explained. ResNet2 in this data compression model includes three CNN modules (i.e., CNN 1 to CNN 3). In this CNN, CNN 1 is followed by a batch normalization (BN) layer and a parametric rectified linear unit (PReLU) activation function. CNN 1 has a 3×3 kernel size, 4 channels, and a stride of 2. CNN 2 is also followed by a BN layer and a PReLU activation function. CNN 2 has a 3×3 kernel size, 4 channels, and a stride of 1. CNN 3 processes the shortened input signal, mainly for aligning dimensions and channels. CNN 3 has a 1×1 kernel size, 4 channels, and a stride of 2.

[0081] It should also be noted that the two ResNet upsampling networks in the data decompression model have the same structure. Taking ResNet upsampling 1 in the data decompression model in Figure 5 as an example, the specific structure of the ResNet upsampling network will be explained. ResNet upsampling 1 includes 3 sets of deconvolutional neural network (DCNN) modules, namely DCNN 1 to DCNN 3. DCNN 1 is followed by a Batch Normalization (BN) layer and a PReLU activation function. DCNN 1 has a 3×3 kernel size, 4 channels, and an upsampling factor of 2. DCNN 2 also has a BN layer and a PReLU activation function, with a 3×3 kernel size, 4 channels, and an upsampling factor of 1. DCNN 3 processes the shortened input signal, primarily for aligning dimensions and channels. It has a 1×1 kernel size, 4 channels, and an upsampling factor of 2.

[0082] It should be understood that after the transmitting device in this application obtains the first data compression result, it simulates the data decompression process of the receiving device using its data decompression model. In other words, the first decompressed data obtained by the transmitting device's data decompression model based on the first data compression result is approximately the same as the first decompressed data obtained by the receiving device's data decompression model based on the first data compression result. That is to say, for ease of description and understanding, this application uses the example of ignoring signal transmission losses and the computational differences between the transmitting and receiving device's data decompression models, and should not be considered a specific limitation of this application.

[0083] It should also be noted that this application does not limit the first service data. For example, the service type of the first service data can be any of the following: environmental reflection points, environmental patches, environmental imaging data, environmental reconstruction maps, radio frequency maps, or positioning data in perception data; training data of models, model parameters, model gradients, model inference results, feature data extracted by models, or model performance data in AI data; H matrix or channel state information (CSI) fed back by devices in a multi-antenna system in channel data.

[0084] It should also be noted that the calculation indicators of data decompression accuracy are related to one or more of the following information: the data type of the first business data, the business type of the first business data, the quality of service (QoS) requirements corresponding to the first business data, or the purpose of the first business data, etc.

[0085] For example, when the calculation metric for data decompression accuracy is related to the business type of the first business data, if the first business data is CSI (Computer-Sensitive Information), the mean square error (MSE) or generalized cosine similarity (GCS) can be used as the calculation metric. That is, the data decompression accuracy is obtained by calculating the MSE or GCS between the first business data and the first decompressed data. If the first business data is sensor data, the chamfer distance or normal deviation can be used as the calculation metric. That is, the data decompression accuracy is obtained by calculating the chamfer distance or normal deviation between the first business data and the first decompressed data. If the first business data is AI data, the reconstruction feature MSE or downstream task accuracy can be used as the calculation metric. That is, the data decompression accuracy is obtained by the MSE between the extreme first business data and the first decompressed data, or by determining the downstream task accuracy.

[0086] For example, when the calculation metric for data decompression accuracy is related to the data type of the first service, if the first service data is a radio frequency map, the calculation metric for data decompression accuracy is related to the radio frequency map format. If the first service data is in multi-path component (MPC) format, then the multi-path parameter MSE is used as the calculation metric to obtain the data decompression accuracy. If the first service data is in grid map format, then the MSE value of the grid map is used as the calculation metric to obtain the data decompression accuracy.

[0087] For example, when the calculation metric for data decompression accuracy is related to the purpose of the first service data, if the first service data is a CSI used for precoding, then GCS can be used as the calculation metric to obtain the data decompression accuracy. If the first service data is a CSI used for channel estimation, then MSE can be used as the calculation metric to obtain the data decompression accuracy.

[0088] For example, when the calculation metric for data decompression accuracy is related to QoS requirements, if the first service data is CSI and the QoS requirements take network throughput into account, then GCS can be used as the calculation metric to obtain the data decompression accuracy, which is related to network throughput. If the first service data is AI data and the QoS requirements take network accuracy loss into account, then the downstream task accuracy can be used as the calculation metric to determine the data decompression accuracy, which is related to network accuracy loss.

[0089] Furthermore, after the sending device determines the data decompression accuracy corresponding to the first data compression result according to S401, the sending device combines the first accuracy threshold to determine whether to execute branch one or branch two. If the data decompression accuracy is greater than or equal to the first accuracy threshold, branch one is executed; a detailed description of branch one can be found in the subsequent descriptions of S402 and S403. If the data decompression accuracy is less than the first accuracy threshold, branch two is executed; a detailed description of branch two can be found in the subsequent descriptions of S404 and S405. The specific values ​​of the thresholds mentioned in this application (including the first accuracy threshold, other accuracy thresholds mentioned later, and threshold #1, etc.) can be adjusted according to the actual application scenario, and this application does not impose specific limitations on them.

[0090] It should be noted that this application only uses data decompression accuracy as an evaluation index for data decompression and should not be regarded as a specific limitation of this application. For example, this application can also use data decompression error as an evaluation index for data decompression. The greater the data decompression accuracy between the first decompressed data and the first service data, the smaller the data decompression error between the first decompressed data and the first service data can be regarded as. That is to say, after the sending device obtains the first decompressed data, it determines the data decompression error based on the first decompressed data and the first service data. When the data decompression error is small (for example, the data decompression error is less than or equal to threshold #1), the sending device executes branch one; when the data decompression error is large (for example, the data decompression error is greater than threshold #1), the sending device executes branch two.

[0091] The following section will provide a detailed explanation of the first precision threshold and the method by which the transmitting device determines the first precision threshold.

[0092] In the examples provided in this application, different service types of data can correspond to different precision thresholds. This application uses the precision threshold corresponding to the service type of the first service data as an example for illustration. The sending device determines the first precision threshold in two ways, namely, method 11 and method 12.

[0093] Method 11: The transmitting device determines the first precision threshold on its own.

[0094] The accuracy threshold is related to the service type. After acquiring (or generating) the first service data, the transmitting device determines the first accuracy threshold based on the service type of the first service data.

[0095] For example, the sending and receiving devices may have agreed on precision thresholds for different service types, or the communication protocol may specify precision thresholds for different services. In this case, after acquiring the first service data, the sending device can determine the first precision threshold based on the service type of the first service data.

[0096] Method 12: The transmitting device receives the first precision threshold from the receiving device.

[0097] In other words, as the user of the first business data, the receiving device can determine the first precision threshold based on the actual business scenario's requirements for data decompression accuracy. Furthermore, the receiving device indicates this first precision threshold to the sending device.

[0098] S402 (Branch 1): When the data decompression accuracy is greater than or equal to the first accuracy threshold, the sending device sends the first data compression result to the receiving device. The first accuracy threshold is associated with the service type corresponding to the first service data.

[0099] In this first branch, the sending device can be considered to believe that the first decompressed data obtained by the receiving device based on the first data compression result is sufficient to meet the needs of the actual business scenario. Therefore, the sending device sends the first data compression result to the receiving device.

[0100] It should be noted that the transmission of the first data compression result mentioned in this application can be understood as processing the first data compression result (including one or more of quantization, channel coding or modulation) and then transmitting a wireless signal to the receiving end to carry the first data compression result.

[0101] In one possible implementation, to facilitate data decompression by the receiving device, when executing branch one, the sending device also sends a third indication message to the receiving device. This third indication message indicates that only the first data compression result was sent to the receiving device, and the receiving device can obtain the decompressed data by decompressing the data based on the first data compression result.

[0102] S403 (Branch 1): The receiving device decompresses the first data based on the first data compression result to obtain the first decompressed data.

[0103] The receiving device receives a wireless signal from the transmitting device carrying the first data compression result, and processes the wireless signal (including one or more of demodulation, channel decoding, or inverse quantization) to obtain the first data compression result. Further, the receiving device decompresses the first data compression result to obtain first decompressed data.

[0104] S404 (Branch 2): When the data decompression accuracy is less than the first accuracy threshold, the sending device sends the first data compression result and the first residual compression result to the receiving device. The first residual compression result is obtained by compression based on the first residual information, and the first residual information is obtained based on the first service data and the first decompressed data.

[0105] In other words, after obtaining the data decompression accuracy in S401, if the data decompression accuracy is less than a first accuracy threshold, the sending device obtains first residual information based on the first service data and the first decompressed data. Further, the sending device performs compression processing based on the first residual information to obtain a first residual compression result. When sending the first data compression result to the receiving device, the sending device also sends the first residual compression result to the receiving device.

[0106] It should be noted that this application does not specifically limit the process by which the transmitting device obtains the first residual information. For example, as shown in Figure 6(a), the transmitting device can input the first service data and the first decompression data into the differential encoder to obtain the first residual information.

[0107] In one possible implementation, to facilitate data decompression by the receiving device, when executing branch two, the sending device also sends a fourth indication message to the receiving device. This fourth indication message indicates that a first data compression result and a first residual compression result have been sent to the receiving device. The receiving device can then decompress the data based on the first data compression result and the first residual compression result to obtain the decompressed data.

[0108] The following describes several methods provided in this application for determining the first residual compression result by the transmitting end device. Please refer to methods 21 to 23 below for details.

[0109] Method 21: The first residual information includes N residual data points, where N is a positive integer. In this case, the transmitting device can filter the first residual information to obtain second residual information, which includes N1 residual data points from the N residual data points, where N1 is a positive integer less than or equal to N. Further, the transmitting device compresses the second residual information to obtain the compressed first residual information.

[0110] Specifically, as shown in Figure 6(b), after the transmitting device obtains the first residual information, it inputs the first residual information into the data filtering module to obtain the second residual information. Further, the transmitting device performs quantization and entropy encoding processing on the second residual information to obtain the first residual compression result. In this way, the first residual compression result is obtained by compressing a portion of the data in the first residual information, which helps reduce the data volume of the first residual compression result, thereby saving communication resources.

[0111] In one possible implementation of method 21, the transmitting device obtains second residual information by filtering the first residual information according to a residual threshold. The second residual information includes N1 residual data points, all of which are greater than or equal to the residual threshold. That is, the transmitting device selects the larger N1 residual data points from the N residual data points included in the first residual information to form the second residual information. Alternatively, it can be understood that any residual data point (or energy) included in the second residual information is greater than any residual data point in the first residual information other than the second residual data point itself.

[0112] The residual threshold can be one of the N residual data points. For example, by sorting the N residual data points from largest to smallest and using the N1th residual data point as the residual threshold, the N1 largest residual data points can be selected from the N residual data points to form the second residual information. Alternatively, the residual threshold can be statistical information of the N residual data points, such as the median, mode, or mean of the N residual data points.

[0113] Understandably, in order for the receiving device to determine which residual data among the N residual data included in the first residual information are compressed and / or which are not compressed after receiving the first residual compression result, in one possible implementation, the N residual data in the first residual information correspond one-to-one with N index information. In this case, the sending device also sends first indication information to the receiving device, which is used to indicate the index information of the N1 compressed residual values ​​in the first residual information.

[0114] For example, the transmitting device calculates a first residual information, which is {4, 3, 2, 5}, based on the first decompressed data and the first service data. Further, the transmitting device determines a second residual information, {4, 5}, from this first residual information and compresses it to obtain the first residual compression result. In this case, the transmitting device can also send a first indication information to the receiving device, indicating that the residual data with index values ​​of 0 and 3 are compressed, while the residual data with index values ​​of 1 and 2 are not compressed. Alternatively, the four residual data in the first residual information correspond one-to-one with four bits, and the transmitting device can also send a first indication information, which is 1001, to the receiving device.

[0115] Method 22: The transmitting device performs data reconstruction processing on the residual data included in the first residual information to obtain the third residual information. This data reconstruction processing includes one or more of matrix straightening, splicing, and vector segmentation. Further, the transmitting device compresses the third residual information to obtain the compressed first residual information.

[0116] Specifically, as shown in Figure 6(c), after the transmitting device obtains the first residual information, it inputs the first residual information into the data reconstruction module to obtain the third residual information. Further, the transmitting device then sequentially performs transformation processing, quantization processing, and entropy coding processing on the third residual information to obtain the first residual compression result. The transformation processing includes, but is not limited to, discrete Fourier transform processing and discrete cosine transform processing. It should be understood that, compared to the third residual information without transformation processing, the third residual information after transformation processing is more conducive to subsequent entropy coding processing, thus improving compression efficiency.

[0117] It is understandable that, in order for the receiving device to determine which data reassembly processes were used to compress the first residual information after receiving the first residual compression result, in one possible implementation, the sending device further sends second indication information to the receiving device. This second indication information is used to indicate the operation parameters corresponding to the data reassembly processes. That is, the second indication information indicates which data reassembly processes were used to obtain the third residual information, and the operation parameters of each data reassembly process. Specifically, the operation parameters for matrix straightening include the direction of matrix straightening, such as straightening by row or by column; the matrix concatenation process includes the dimension of the matrix concatenation; and the vector segmentation process includes the length of the vector segments.

[0118] For example, after the transmitting device performs matrix straightening on the first residual information column-by-column, it then performs vector segmentation of length L to obtain the third residual information. The transmitting device then compresses this third residual information to obtain the compressed first residual. In this case, the transmitting end can also send a second indication to the receiving end, indicating that the third residual information was obtained by first performing matrix straightening and then vector segmentation. The operation parameters for matrix straightening are row-by-row straightening and the length of the vector segment is length L.

[0119] Method 23: The sending device inputs the first residual information and the first decompression data into the AI ​​residual encoder to obtain the first residual compression result.

[0120] Specifically, as shown in Figure 6(d), after the sending device obtains the first residual information, it inputs the first residual information and the first decompression data into the AI ​​residual encoder (or can be understood as an AI model used to output the residual compression result) to obtain the first residual compression result.

[0121] S405 (Branch 2): The receiving device decompresses the data based on the first data compression result and the first residual compression result to obtain the second decompressed data.

[0122] The receiving device receives a wireless signal from the transmitting device carrying the first data compression result and the first residual compression result, and processes the wireless signal (including one or more of demodulation, channel decoding, or inverse quantization) to obtain the first data compression result and the first residual compression result. Further, the receiving device decompresses the first data compression result to obtain first decompressed data, and decompresses the first residual compression result to obtain first residual decompressed information. The receiving device performs data processing (e.g., data recovery processing or data reshaping processing) based on the first residual decompressed information and the first decompressed data to obtain second decompressed data.

[0123] Understandably, in branch two, the decompression accuracy of the second decompressed data obtained based on the first data compression result and the first residual compression result is higher than the decompression accuracy of the first decompressed data obtained based on the first data compression result. That is, the second decompressed data is closer to the first business data than the first decompressed data.

[0124] The process by which the receiving device obtains the first decompressed data based on the first data compression result can be found in the relevant description of S403 above. The process by which the receiving device obtains the first residual decompressed information based on the first residual compression result is explained below with reference to methods 21 to 23 in S404 above.

[0125] 1. In conjunction with the aforementioned method 21, the receiving device decompresses the first residual compression result to obtain second residual decompression information, which includes N1 residual data. Further, the receiving device performs data recovery processing based on the second residual decompression information to obtain first residual decompression information, which includes N residual data.

[0126] For example, the receiving device receives a first data compression result and a first residual compression result, and decompresses the first residual compression result to obtain second residual decompression information, which is {4,5}. In this case, the receiving device also receives first indication information from the sending device, which indicates that residual data with index values ​​of 0 and 3 are compressed, and / or indicates that residual data with index values ​​of 1 and 3 are not compressed. Further, based on the second residual decompression information and the first indication information, the receiving device determines that the residual data with index value of 0 in the first residual decompression information is 4, the residual data with index value of 1 is 0, the residual data with index value of 2 is 0, and the residual data with index value of 3 is 5, that is, the first residual decompression information is {4, 0, 0, 5}.

[0127] 2. In conjunction with the aforementioned method 22, the receiving device decompresses the first residual compression result to obtain the third residual decompression information. Further, the receiving device performs data reshaping processing based on this third residual decompression information to obtain the first residual decompression information. It can be understood that this data reshaping processing is the reverse of the data reconstruction processing in the aforementioned method 22.

[0128] For example, the receiving device receives a first data compression result and a first residual compression result, and decompresses the first residual compression result to obtain third residual decompression information, which is {4, 3, 9, 7}. In this case, the receiving device also receives second indication information from the sending device, which indicates that the data reconstruction process for obtaining the third residual information includes matrix straightening, the operation parameter of which is row-wise straightening, with each row including 2 elements. Further, the receiving device performs data reshaping processing on the third residual decompression information to obtain the first residual decompression information.

[0129] 3. In conjunction with method 23 described above, the receiving device decompresses the first data compression result to obtain the first decompressed data. Then, the receiving device inputs the first decompressed data and the first residual compression result into the AI ​​residual decoder to obtain the first residual decompression information.

[0130] In summary, by implementing the data transmission method described in Figure 4 of this application, when the data decompression accuracy corresponding to the first data compression result meets the actual business scenario's requirements for data decompression accuracy, the sending device sends the first data compression result to the receiving device. When the data decompression accuracy corresponding to the first data compression result does not meet the actual business scenario's requirements for data decompression accuracy, the sending device sends the first data compression result and the corresponding first residual compression result to the receiving device. This method adjusts the data decompression accuracy of the compression scheme, which helps improve the adaptability of the compression scheme's data decompression accuracy to the current actual business scenario's requirements.

[0131] To ensure that the compression scheme meets the needs of actual business scenarios while also saving data transmission volume, this application also provides a possible implementation method. In this possible implementation method, the first business data corresponds to P precision thresholds, where P is a positive integer, and the first precision threshold described in Figure 4 is the maximum value among the P precision thresholds. Each of the P precision thresholds corresponds one-to-one with a different compression ratio; the smaller the precision threshold, the smaller the compression ratio. For example, precision threshold #1 and precision threshold #2 are any two of the P precision thresholds, and precision threshold #1 is greater than precision threshold #2; therefore, the compression ratio of precision threshold #1 is higher than that of precision threshold #2.

[0132] Based on this, this application also provides a data transmission method as shown in Figure 7. The method execution entities shown in Figure 7 can be a transmitting device and a receiving device, or they can be modules in the transmitting device and modules in the receiving device, or they can be chips in the transmitting device and chips in the receiving device. Figure 7 uses the transmitting device and the receiving device as examples of method execution entities. The transmitting device can be the RAN node shown in Figure 1 or Figure 2, or a terminal shown in Figure 1 or Figure 2; the receiving device mentioned in this application can be the RAN node shown in Figure 1 or Figure 2, or a terminal shown in Figure 1 or Figure 2; this application does not impose specific limitations.

[0133] S701 and the transmitting device compress the first residual information based on the compression parameters corresponding to P compression rates to obtain P candidate residual compression results.

[0134] The transmitting device obtains first residual information based on the first service data and the first decompressed data. Further, the transmitting device compresses the first residual information based on P compression rates corresponding to P precision thresholds of the first service data, obtaining P candidate residual compression results. The specific method by which the transmitting device determines the first residual information can be referred to the relevant description of obtaining the first residual information in S404 above; the method by which the transmitting device determines the P precision thresholds of the first service can be referred to the method by which the transmitting device determines the first precision thresholds in S401 above (i.e., methods 11 and 12), and will not be repeated here.

[0135] It should be noted that the compression ratio refers to the ratio between the amount of data in the first business data and the amount of data in the compressed result (i.e., the sum of the amount of data in the first data compression result and the amount of data in the candidate residual compression result).

[0136] It's important to understand that, assuming the total amount of data in the first data compression result remains constant, a higher compression ratio results in a smaller amount of data in the corresponding candidate residual compression result; conversely, a lower compression ratio results in a larger amount of data in the corresponding candidate residual compression result. For example, if the compression ratio of precision threshold #1 is higher than the compression ratio of precision threshold #2, the amount of data in the candidate residual compression result obtained based on the compression ratio of precision threshold #1 is less than the amount of data in the candidate residual compression result obtained based on the compression ratio of precision threshold #2.

[0137] It is also necessary to understand that the first precision threshold is the maximum value among the P precision thresholds, the compression rate corresponding to the first precision threshold is the maximum value among the P compression rates corresponding to the P precision thresholds, the compression rate of the first precision threshold is the ratio of the data volume of the first business data to the data volume of the first data compression result, and the data volume of the candidate residual compression result obtained based on the compression rate of the first precision threshold is empty (or 0).

[0138] For ease of understanding, the candidate residual compression result obtained based on the compression ratio of precision threshold #1 is denoted as candidate residual compression result #1, and the candidate residual compression result obtained based on the compression ratio of precision threshold #2 is denoted as candidate residual compression result #2. This application also provides the following two examples (i.e., Example 1 and Example 2) to illustrate that the data volume of candidate residual compression result #1 is smaller than the data volume of candidate residual compression result #2.

[0139] Example 1: The first residual information includes N residual data points, where N is a positive integer. In this case, the quantization bit width of the N residual data points included in candidate residual compression result #1 is smaller than the quantization bit width of the N residual data points included in candidate residual compression result #2. It should be noted that the quantization bit width of a candidate residual compression result mentioned in this application refers to the number of bits represented by each floating-point number in the candidate residual compression result after quantization.

[0140] For example, the first service data corresponds to four precision thresholds, which are P1 (i.e., the first precision threshold), P2, P3, and P4 in descending order. The transmitting device obtains first residual information, including 10 residual data points, based on the first service data and the first decompressed data. This first residual information is {19,7,4,5,12,9,7,17,12,11}. In this case, the candidate residual compression result obtained by the transmitting device based on the compression ratio corresponding to P1 is empty (i.e., the quantization bit width is 0). The candidate residual compression results obtained by the transmitting device based on the compression ratios corresponding to P2 to P4 are shown in Figure 8(a), namely candidate residual compression result 2, candidate residual compression result 3, and candidate residual compression result 4, respectively. The quantization bit width of candidate residual compression result 4 is 5, the quantization bit width of candidate residual compression result 3 is 4, and the quantization bit width of candidate residual compression result 2 is 2. In other words, the quantization bit width of candidate residual compression result 4 is greater than the quantization bit width of candidate residual compression result 3, and greater than the quantization bit width of candidate residual compression result 2.

[0141] Example 2: The first residual information includes N residual data points, where N is a positive integer. In this case, the number of residual data points included in candidate residual compression result #1 is less than the number of residual data points included in candidate residual compression result #2.

[0142] For example, the first service data corresponds to four precision thresholds, which are P1 (i.e., the first precision threshold), P2, P3, and P4 in descending order. The transmitting device obtains first residual information, including 10 residual data points, based on the first service data and the first decompressed data. This first residual information is {19,7,4,5,12,9,7,17,12,11}. In this case, after the transmitting device sorts the 10 residual data points in the first residual information from largest to smallest, the candidate residual compression result obtained by the transmitting device based on the compression ratio corresponding to P1 is empty. The candidate residual compression results obtained based on the compression ratios corresponding to P2 to P4 are shown in Figure 8(b), namely candidate residual compression result 2, candidate residual compression result 3, and candidate residual compression result 4, respectively. Among them, candidate residual compression result 4 includes 10 residual data points, candidate residual compression result 3 includes 7 residual data points, and candidate residual compression result 2 includes 4 residual data points. In other words, the number of residual data included in candidate residual compression result 4 is greater than the number of residual data included in candidate residual compression result 3, which is greater than the number of residual data included in candidate residual compression result 2.

[0143] It should also be noted that the method of obtaining the candidate residual compression result based on the compression ratio corresponding to a certain precision threshold in S701 of this application can also refer to methods 21 to 23 in S404 above. For example, in method 21, the smaller the compression ratio, the more residual data are included in the candidate residual result corresponding to that compression ratio, or the smaller the residual threshold used to determine the candidate residual result.

[0144] S702. The transmitting device determines the data decompression accuracy corresponding to the first data compression result. The first data compression result is obtained based on the compression of the first service data, the data decompression accuracy is obtained based on the first service data and the first decompressed data, and the first decompressed data is obtained by decompressing based on the first data compression result.

[0145] For details on the specific implementation of S702, please refer to the foregoing description of the specific implementation of S401, which will not be repeated here.

[0146] It should be noted that this application does not limit the execution order of S701 and S702. S701 can be executed before S702, S701 can be executed simultaneously with S702, and S701 can be executed after S702.

[0147] S703. The transmitting device determines the first residual compression result from the P candidate residual compression results based on the data decompression accuracy.

[0148] The sending device determines the first residual compression result from the P candidate residual compression results based on the difference between the data decompression accuracy determined in S702 and the P accuracy thresholds.

[0149] In one possible implementation, the P precision thresholds correspond to P precision intervals, and each of the P precision intervals corresponds one-to-one with a P compression ratio. In this case, the transmitting device determines a first residual compression result from the P candidate residual compression results. This first residual compression result is obtained by compression based on the compression ratio corresponding to the first precision interval, where the first precision interval is the interval in which the data decompression precision determined in S702 falls within the P precision intervals.

[0150] For example, the first business data corresponds to four precision thresholds, which are P1 (i.e., the first precision threshold), P2, P3, and P4 in descending order. The correspondence between these four precision thresholds and precision intervals, the correspondence between each precision interval and compression ratio, and the candidate residual compression results obtained based on various compression ratios are shown in Table 1.

[0151] Table 1

[0152] When the data decompression accuracy determined by S702 is within the accuracy range [P1, P0), the first residual compression result is candidate residual compression result 1, and the candidate residual compression result is empty; when the data decompression accuracy determined by S702 is within the accuracy range [P2, P1), the first residual compression result is candidate residual compression result 2; when the data decompression accuracy determined by S702 is within the accuracy range [P3, P2), the first residual compression result is candidate residual compression result 3; when the data decompression accuracy determined by S702 is within the accuracy range [P4, P3), the first residual compression result is candidate residual compression result 4.

[0153] S704. The transmitting device sends the first data compression result and the first residual compression result to the receiving device.

[0154] When the data decompression accuracy is greater than or equal to the first accuracy threshold (i.e., the maximum value among the P accuracy thresholds), the first residual compression result is empty, and the sending device sends the first data compression result to the receiving device. In this case, the specific implementation method can be found in the relevant description in the aforementioned S402.

[0155] When the data decompression accuracy is less than the first accuracy threshold, the first residual compression result is not empty, and the sending device sends the first data compression result and the first residual compression result to the receiving device. For specific implementation details in this case, please refer to the relevant description in the aforementioned S404.

[0156] S705, The receiving device decompresses the data based on the first data compression result and the first residual compression result to obtain decompressed data.

[0157] When the first residual compression result is empty, the transmitting device decompresses the data based on the first data compression result to obtain the first decompressed data. For specific implementation details in this case, please refer to the relevant description in S403 above.

[0158] When the first residual result is not empty, the sending device decompresses the data based on the first data compression result and the first residual compression result to obtain the second decompressed data. For specific implementation details in this case, please refer to the relevant description in S405 above.

[0159] In summary, by implementing the data transmission method described in Figure 7 of this application, when the data decompression accuracy corresponding to the first data compression result meets the requirements of the actual business scenario for data decompression accuracy, the sending device sends the first data compression result to the receiving device. When the data decompression accuracy corresponding to the first data compression result does not meet the requirements of the actual business scenario for data decompression accuracy, the sending device determines the data volume corresponding to the first residual compression result based on the difference between the data decompression accuracy and the accuracy required to meet the requirements of the actual business scenario (i.e., the first accuracy threshold), and sends the first data compression result and the corresponding first residual compression result to the receiving device. In this way, by adjusting the data volume of the first residual compression result, it is beneficial to meet the needs of the current actual business scenario while also reducing the amount of data transmitted during communication, thus saving communication resources.

[0160] It is understood that, in order to achieve the aforementioned functions, the device includes corresponding hardware structures and / or software modules for performing each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0161] This application embodiment can divide the transmitting or receiving device into functional modules according to the above method examples. For example, each function can be divided into its own functional modules, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0162] Please refer to Figure 9, which shows a schematic diagram of the structure of a communication device 900 according to an embodiment of this application. The communication device shown in Figure 9 can be a transmitting device, a device within a transmitting device, or a device that can be used in conjunction with a transmitting device. The communication device shown in Figure 9 can include a communication unit 901 and a processing unit 902; the communication device shown in Figure 9 can be a receiving device, a device within a receiving device, or a device that can be used in conjunction with a receiving device. The communication device shown in Figure 8 can include a communication unit 901 and a processing unit 902. Specifically, the processing unit 902 is used to process data, which can be data received by the communication unit 901, and the processed data can also be sent by the communication unit 901; the communication unit 901 can be understood as a transceiver unit, including a receiving module and / or a sending module. The receiving module is used to perform the receiving action of the device (i.e., the transmitting device or the receiving device) in any embodiment of Figure 4 or Figure 7, and the sending module is used to perform the sending action of the device (i.e., the transmitting device or the receiving device) in any embodiment of Figure 4 or Figure 7.

[0163] In one embodiment, when the communication device 900 is a transmitting device, a device within the transmitting device (e.g., a chip or chip system within the transmitting device), or a device compatible with the transmitting device, wherein:

[0164] Processing unit 902 is configured to determine the data decompression accuracy corresponding to the first data compression result, wherein the first data compression result is obtained based on the compression of the first service data, and the data decompression accuracy is obtained based on the first service data and the first decompressed data, wherein the first decompressed data is obtained based on the decompression of the first data compression result; when the data decompression accuracy is greater than or equal to a first accuracy threshold, communication unit 901 is configured to send the first data compression result, wherein the first accuracy threshold is associated with the service type corresponding to the first service data; when the data decompression accuracy is less than the first accuracy threshold, communication unit 901 is configured to send the first data compression result and a first residual compression result, wherein the first residual compression result is obtained based on the compression of first residual information, wherein the first residual information is obtained based on the first service data and the first decompressed data.

[0165] In one possible implementation, the communication unit 901 is further configured to receive a first precision threshold corresponding to the first service data.

[0166] In one possible implementation, the first service data corresponds to P precision thresholds, where P is a positive integer, the first precision threshold is the maximum value among the P precision thresholds, and the P precision thresholds correspond one-to-one with P compression rates; wherein, precision threshold #1 and precision threshold #2 are any two precision thresholds among the P precision thresholds, precision threshold #1 is greater than precision threshold #2, and the compression rate of precision threshold #1 is higher than the compression rate of precision threshold #2.

[0167] In one possible implementation, the P precision thresholds correspond to P precision intervals, and the P precision intervals correspond one-to-one with the P compression ratios; the processing unit 902 is further configured to compress the first residual information based on the compression parameters corresponding to the P compression ratios respectively, to obtain P candidate residual compression results; the processing unit 902 is further configured to determine the first residual compression result from the P candidate residual compression results based on the data decompression precision, the first residual compression result being obtained by compression based on the compression ratio corresponding to the first precision interval, and the first precision interval being the interval in which the data decompression precision is located among the P precision intervals.

[0168] In one possible implementation, the first residual information includes N residual data, where N is a positive integer; the quantization bit width of the N residual data included in candidate residual compression result #1 is smaller than the quantization bit width of the N residual data included in candidate residual compression result #2; or, the number of residual data included in candidate residual compression result #1 is smaller than the number of residual data included in candidate residual compression result #2; wherein, candidate residual compression result #1 is the candidate residual compression result corresponding to the precision threshold #1, and candidate residual compression result #2 is the candidate residual compression result corresponding to the precision threshold #2.

[0169] In one possible implementation, the first residual information includes N residual data, where N is a positive integer; the processing unit 902 is further configured to perform filtering processing on the first residual information to obtain second residual information, the second residual information including N1 residual data from the N residual data, where N1 is a positive integer less than or equal to N; the processing unit 902 is further configured to compress the second residual information to obtain the first residual compression result.

[0170] In one possible implementation, the processing unit 902 is further configured to filter the first residual information according to a residual threshold to obtain the second residual information, wherein the second residual information includes N1 residual data that are all greater than or equal to the residual threshold; wherein the residual threshold is one of the N residual data, or the residual threshold is statistical information of the N residual data.

[0171] In one possible implementation, the N residual data correspond one-to-one with N index information, and the communication unit 901 is also used to send first indication information, which is used to indicate the index information of the N1 residual data.

[0172] In one possible implementation, the processing unit 902 is further configured to perform data recombination processing on the residual data included in the first residual information to obtain the third residual information. The data recombination processing includes one or more of matrix straightening processing, splicing processing, and vector segmentation processing. The processing unit 902 is further configured to compress the third residual information to obtain the first residual compression result.

[0173] In one possible implementation, the communication unit 901 is further configured to send a second indication message, which indicates the operation parameters corresponding to the data reassembly process.

[0174] In one possible implementation, the processing unit 902 is further configured to input the first residual information and the first decompression data into the artificial intelligence (AI) residual encoder to obtain the first residual compression result.

[0175] For a more detailed description of the communication unit 901 and the processing unit 902 described above, please refer to the relevant description of the transmitting device in the method embodiment shown in Figure 4 or Figure 7.

[0176] In one embodiment, when the communication device 900 is a receiving device, a device within a receiving device, or a device compatible with a receiving device, wherein:

[0177] The communication unit 901 is used to receive a first data compression result and a first residual compression result, wherein the first data compression result is obtained based on the compression of the first service data, the first residual compression result is obtained based on the compression of the first residual information, the first residual information is obtained based on the first service data and the first decompressed data, and the first decompressed data is obtained by decompressing based on the first data compression result; the processing unit 902 is used to decompress based on the first data compression result and the first residual compression result to obtain second decompressed data.

[0178] In one possible implementation, the communication unit 901 is further configured to transmit a first precision threshold corresponding to the first service data.

[0179] In one possible implementation, the first service data corresponds to P precision thresholds, where P is a positive integer, the first precision threshold is the maximum value among the P precision thresholds, and the P precision thresholds correspond one-to-one with P compression rates; wherein, precision threshold #1 and precision threshold #2 are any two precision thresholds among the P precision thresholds, precision threshold #1 is greater than precision threshold #2, and the compression rate of precision threshold #1 is less than the compression rate of precision threshold #2.

[0180] In one possible implementation, the first residual information includes N residual data, and the first data compression result is obtained based on N1 residual data among the N residual data; the communication unit 901 is also used to receive first indication information, which is used to indicate the index information of the N1th residual data.

[0181] In one possible implementation, the first data compression result is obtained by performing data recombination processing on the residual data included in the first residual information; the communication unit 901 is further configured to receive second indication information, which is used to indicate the operation parameters corresponding to the data recombination processing.

[0182] For a more detailed description of the communication unit 901 and the processing unit 902 described above, please refer to the relevant description of the receiving device in the method embodiment shown in Figure 4 or Figure 7.

[0183] In one possible implementation, when the communication device 900 is a chip, the communication unit 901 can be a communication interface, pins, or circuits. The communication interface can be used to input data to be processed to the processor and can output the processor's processing results. In a specific implementation, the communication interface can be a general purpose input / output (GPIO) interface, which can be connected to multiple peripheral devices (such as displays (LCDs), cameras, radio frequency (RF) modules, antennas, etc.). The communication interface is connected to the processor via a bus.

[0184] The processing unit 902 may be a processor, which can execute computer programs or instructions stored in the storage module to cause the chip to execute the methods involved in any of the embodiments shown in Figure 4 or Figure 7. Further, the processor may include a controller, an arithmetic logic unit (ALU), and registers. For example, the controller is mainly responsible for decoding the computer program or instructions and issuing control signals for the operations corresponding to the computer program or instructions. The ALU is mainly responsible for performing fixed-point or floating-point arithmetic operations, shift operations, and logical operations, and can also perform address operations and conversions. The registers are mainly responsible for storing register operands and intermediate operation results temporarily stored during the execution of the computer program or instructions. In specific implementations, the processor's hardware architecture may be an application-specific integrated circuit (ASIC) architecture, a microprocessor without interlocked piped stages architecture (MIPS) architecture, an advanced reduced instruction set machine (RISC) machine (ARM) architecture, or a network processor (NP) architecture, etc. The processor may be single-core or multi-core. The storage module may be an internal storage module of the chip, such as a register or cache. Storage modules can also be external to the chip, such as read-only memory (ROM) or other types of static storage devices that can store static information and computer programs or instructions, random access memory (RAM), etc.

[0185] It should be noted that the functions of the processor and interface can be implemented through hardware design, software design, or a combination of both; no restrictions are imposed here.

[0186] Figure 10 is a schematic diagram of another communication device provided in an embodiment of this application. It is understood that the communication device 1000 includes necessary means such as modules, units, elements, circuits, or interfaces, appropriately configured together to execute this solution. The communication device 1000 can be the aforementioned transmitting or receiving device, or a component (e.g., a chip) within these devices, used to implement the methods described in the above method embodiments.

[0187] In one possible design, as shown in Figure 10, the communication device 1000 includes a processor 1010 and an interface circuit 1020. The processor 1010 and the interface circuit 1020 are coupled to each other.

[0188] Optionally, the communication device 1000 may include one or more processors 1010. The processor 1010 may be a general-purpose processor or a dedicated processor, such as a baseband processor or a central processing unit (CPU). The baseband processor can be used to process communication protocols and communication data, while the CPU can be used to control the communication device (e.g., a terminal device, a transmitting device, or a chip), execute computer programs or instructions, and process the data of the computer programs or instructions.

[0189] It is understood that the interface circuit 1020 can be a transceiver or an input / output interface. When the communication device 1000 is a transmitting or receiving device, the interface circuit 1020 is a transceiver, including a transmitter and / or a receiver. The transmitter can be referred to as a transmitting unit, transmitter, or transmitting circuit, etc., and is used to implement the transmitting function. The receiver can be referred to as a receiving unit, receiver, or receiving circuit, etc., and is used to implement the receiving function. When the communication device 1000 is a chip in the transmitting or receiving device, the interface circuit 1020 is the input / output interface of that chip. Optionally, the communication device 1000 may also include an antenna (not shown in the figure). The interface circuit 1020 may sometimes be referred to as a transceiver unit, transceiver, transceiver circuit, or transceiver, etc., and is used to realize the transmitting and receiving functions of the communication device through the antenna.

[0190] Optionally, the communication device 1000 may further include a memory 1030 for storing computer programs or instructions executed by the processor 1010, or storing input data required by the processor 1010 to run the computer programs or instructions, or storing data generated by the processor 1010 after running the computer programs or instructions. Optionally, the processor 1010 and the memory 1030 may be provided separately or integrated together.

[0191] When the communication device 1000 is used to implement the method shown in FIG4 or FIG7, the processor 1010 is used to implement the function of the processing unit 902, and the interface circuit 1020 is used to implement the function of the communication unit 901.

[0192] When the aforementioned communication device is a chip applied to a transmitting device, the terminal chip implements the functions of the transmitting device in the above method embodiments. The transmitting device chip receives information from the receiving device, which can be understood as the information being first received by other modules (such as an RF module or antenna) in the transmitting device, and then sent to the transmitting device chip by these modules. The transmitting device chip sends information to the receiving device, which can be understood as the information being first sent to other modules (such as an RF module or antenna) in the transmitting device, and then sent to the receiving device by these modules.

[0193] When the aforementioned communication device is a chip applied to a receiving device, the receiving device chip implements the functions of the receiving device in the above method embodiments. The receiving device chip receives information from the transmitting device, which can be understood as the information being first received by other modules (such as an RF module or antenna) in the receiving device, and then sent to the receiving device chip by these modules. The receiving device chip sends information to the transmitting device, which can be understood as the information being sent down to other modules (such as an RF module or antenna) in the receiving device, and then sent back to the transmitting device by these modules.

[0194] This application also provides a computer-readable storage medium storing a computer program or instructions that, when executed, cause a computer to perform the method described in any of the embodiments shown in FIG4 or FIG7.

[0195] This application also provides a computer program product, which includes computer program code. When the computer program code is run, it causes the computer to perform the method described in any one of the embodiments shown in FIG4 or FIG7.

[0196] In this application, entity A sends information to entity B, either directly or indirectly through other entities. Similarly, entity B receives information from entity A, either directly or indirectly through other entities. Entities A and B can be RAN nodes or terminals, or modules within RAN nodes or terminals. Information transmission and reception can be between RAN nodes and terminals, such as between a base station and a terminal; between two RAN nodes, such as between a CU and a DU; or between different modules within a single device, such as between a terminal chip and other modules of the terminal, or between a base station chip and other modules of the base station.

[0197] It is understood that the processor in the embodiments of this application can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0198] The method steps in the embodiments of this application can be implemented in hardware or in software instructions executable by a processor. The software instructions can consist of corresponding software modules, which can be stored in random access memory, flash memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. The storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. Alternatively, the ASIC can reside in a base station or terminal. The processor and storage medium can also exist as discrete components in a base station or terminal.

[0199] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this application are performed entirely or partially. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a network device, a user equipment, or other programmable device. The computer program or instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, the computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; it can also be an optical medium, such as a digital video optical disc; or it can be a semiconductor medium, such as a solid-state drive. The computer-readable storage medium may be a volatile or non-volatile storage medium, or may include both types of storage media.

[0200] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0201] In the various embodiments of this application, unless otherwise specified or in case of logical conflict, the terminology and / or descriptions of different embodiments are consistent and can be referenced by each other. The technical features of different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0202] In this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. In the textual description of this application, the character " / " generally indicates an "or" relationship between the preceding and following related objects; in the formulas of this application, the character " / " indicates a "division" relationship between the preceding and following related objects. "Including at least one of A, B, and C" can mean: including A; including B; including C; including A and B; including A and C; including B and C; including A, B, and C.

[0203] The terms "first" and "second," etc., used in the specification, claims, and drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of operations or units is not limited to the listed operations or units, but may optionally include operations or units not listed, or may optionally include other operations or units inherent to these processes, methods, products, or apparatuses.

[0204] In this application, "send" and "receive" refer to the direction of signal transmission. For example, "send information to XX" can be understood as the destination of the information being XX, which can include direct transmission via the air interface or indirect transmission via the air interface from other units or modules. "Receive information from YY" can be understood as the source of the information being YY, which can include direct reception from YY via the air interface or indirect reception from YY via the air interface from other units or modules. "Send" can also be understood as the "output" of a chip interface, and "receive" can also be understood as the "input" of a chip interface. In other words, sending and receiving can occur between devices, such as between network devices and terminal devices, or within a device, such as between components, modules, chips, software modules, or hardware modules within the device via a bus, wiring, or interface. It is understood that information may undergo necessary processing, such as encoding and modulation, between the source and destination of the information transmission, but the destination can understand the valid information from the source. Similar expressions in this application can be understood in a similar way and will not be elaborated further.

[0205] In this application, "instruction" can include direct and indirect instructions, as well as explicit and implicit instructions. The information indicated by a certain piece of information (as described below, the instruction information) is called the information to be instructed. In specific implementation, there are many ways to instruct the information to be instructed, such as, but not limited to, directly instructing the information to be instructed, such as the information to be instructed itself or its index. It can also indirectly instruct the information to be instructed by instructing other information, where there is a relationship between the other information and the information to be instructed; or it can only instruct a part of the information to be instructed, while the other parts are known or pre-agreed upon. For example, the instruction can be implemented by using a pre-agreed (e.g., protocol predefined) arrangement of various information, thereby reducing the instruction overhead to some extent. This application does not limit the specific method of instruction. It is understood that for the sender of the instruction information, the instruction information can be used to instruct the information to be instructed; for the receiver of the instruction information, the instruction information can be used to determine the information to be instructed.

[0206] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application. The order of the process numbers described above does not imply the order of execution; the execution order of each process should be determined by its function and internal logic.

Claims

1. A data transmission method, characterized by, The method comprises: determining a data decompression accuracy corresponding to a first compression result, the first data compression result being obtained based on first service data compression, the data decompression accuracy being obtained based on the first service data and first decompressed data, the first decompressed data being obtained based on first data compression result decompression; in the case where the data decompression accuracy is greater than or equal to a first accuracy threshold, sending the first data compression result, the first accuracy threshold being associated with a service type corresponding to the first service data; in the case where the data decompression accuracy is less than the first accuracy threshold, sending the first data compression result and a first residual compression result, the first residual compression result being obtained based on first residual information compression, the first residual information being obtained based on the first service data and the first decompressed data.

2. The method of claim 1, wherein, The method further comprises: receiving the first accuracy threshold corresponding to the first service data.

3. The method of claim 1 or 2, wherein, The first service data corresponds to P accuracy thresholds, P being a positive integer, the first accuracy threshold being the maximum of the P accuracy thresholds, the P accuracy thresholds corresponding to P compression rates one by one; wherein accuracy threshold #1 and accuracy threshold #2 are any two of the P accuracy thresholds, the accuracy threshold #1 being greater than the accuracy threshold #2, the compression rate of the accuracy threshold #1 being higher than that of the accuracy threshold #2.

4. The method of claim 3, wherein, The P accuracy thresholds correspond to P accuracy intervals, the P accuracy intervals corresponding to the P compression rates one by one; The method further comprises: compressing the first residual information based on the compression parameters corresponding to the P compression rates respectively, to obtain P candidate residual compression results; determining the first residual compression result from the P candidate residual compression results based on the data decompression accuracy, the first residual compression result being obtained based on the compression rate corresponding to the first accuracy interval, the first accuracy interval being the interval in which the data decompression accuracy is located among the P accuracy intervals.

5. The method of claim 3 or 4, wherein, The first residual information comprises N residual data, N being a positive integer; The quantization bit width of the N residual data included in the candidate residual compression result #1 is less than that of the N residual data included in the candidate residual compression result #2; Or, the number of residual data included in the candidate residual compression result #1 is less than that of residual data included in the candidate residual compression result #2; wherein the candidate residual compression result #1 is the candidate residual compression result corresponding to the accuracy threshold #1, and the candidate residual compression result #2 is the candidate residual compression result corresponding to the accuracy threshold #2.

6. The method according to any one of claims 1 to 5, characterized in that, The first residual information comprises N residual data, N being a positive integer; The method further comprises: performing screening processing on the first residual information to obtain second residual information, the second residual information comprising N1 residual data in the N residual data, N1 being a positive integer less than or equal to N; compressing the second residual information to obtain the first residual compression result.

7. The method of claim 6, wherein, The screening processing on the first residual information obtains second residual information, and the screening processing on the first residual information comprises: The screening processing on the first residual information obtains the second residual information according to a residual threshold, and the N1 residual data included in the second residual information are all greater than or equal to the residual threshold. The residual threshold is one of the N residual data, or the residual threshold is statistical information of the N residual data.

8. The method of claim 6 or 7, wherein, The N residual data correspond to N index information one by one, and the method further comprises: Sending first indication information, the first indication information being used for indicating index information of the N1 residual data.

9. The method of any one of claims 1-5, wherein, The method further comprises: Performing data reorganization processing on residual data included in the first residual information to obtain third residual information, the data reorganization processing comprising one or more of matrix straightening processing, splicing processing, and vector segmentation processing; Compressing the third residual information to obtain the first residual compression result.

10. The method of claim 9, wherein, The method further comprises: Sending second indication information, the second indication information being used for indicating an operation parameter corresponding to the data reorganization processing.

11. The method of any one of claims 1-5, wherein, The processing on the first residual information to obtain the residual compression result comprises: Inputting the first residual information and the first decompression data into an artificial intelligence (AI) residual encoder to obtain the first residual compression result.

12. A data transmission method, characterized by, The method comprises: Receiving a first data compression result and a first residual compression result, the first data compression result being obtained based on first service data compression, the first residual compression result being obtained based on first residual information compression, the first residual information being obtained based on the first service data and first decompression data, and the first decompression data being obtained based on decompression of the first data compression result; Performing decompression based on the first data compression result and the first residual compression result to obtain second decompression data.

13. The method of claim 12, wherein, The method further comprises: Sending a first precision threshold corresponding to the first service data.

14. The method of claim 12 or 13, wherein, The first service data corresponds to P precision thresholds, the P being a positive integer, the first precision threshold being a maximum value of the P precision thresholds, and the P precision thresholds corresponding to P compression rates one by one; wherein precision threshold #1 and precision threshold #2 are any two of the P precision thresholds, the precision threshold #1 being greater than the precision threshold #2, and a compression rate of the precision threshold #1 being less than a compression rate of the precision threshold #2.

15. The method according to any one of claims 12-14, characterized in that, The first residual information comprises N residual data, and the first data compression result is obtained based on N1 residual data of the N residual data. The method further comprises: Receiving first indication information, the first indication information being used for indicating index information of the N1 residual data.

16. The method according to any one of claims 12-15, characterized in that, The first data compression result is obtained by performing data reorganization processing on residual data included in the first residual information. The method further comprises: Receiving second indication information, the second indication information being used for indicating an operation parameter corresponding to the data reorganization processing.

17. A communications device, characterized by The method comprises modules for performing the method as claimed in any one of claims 1-16.

18. A communications device, characterized by The communication device further comprises a memory for storing the computer program or instructions, which, when executed by the processor, cause the method of any of claims 1-16 to be implemented.

19. The communication apparatus of claim 18, wherein, The communication device further comprises a memory for storing the computer program or instructions, which, when executed by the processor, cause the method of any of claims 1-16 to be implemented.

20. A computer-readable storage medium, characterized in that, The storage medium has stored therein computer programs or instructions, which, when executed, cause the method of any of claims 1-16 to be implemented.

21. A computer program product, characterised in that, The computer program product comprises computer programs or instructions, which, when executed, cause the method of any of claims 1-16 to be implemented.

22. A communication system, characterized by The communication system comprises a communication device for performing the method of any of claims 1-11, and a communication device for performing the method of any of claims 12-16.

Images