Data transmission method and apparatus

Abstract

Disclosed in the embodiments of the present application are a data transmission method and apparatus, which are used for improving the efficiency of adjusting data transmission bit rates. The method in the embodiments of the present application comprises: a network device determines a target traffic service rate, the target traffic service rate being a predicted data transmission bit rate that can be provided by the network device for a user equipment at a first moment; on the basis of the target traffic service rate and a neural network model, the network device determines a target quality-of-service parameter, the target quality-of-service parameter being used for indicating the traffic service quality corresponding to the user equipment at the first moment, an input of the neural network model comprising a quality-of-service parameter, and an output of the neural network model comprising a traffic service rate; and, on the basis of the target quality-of-service parameter, the network device adjusts a transmission behavior at a second moment, such that a server or the user equipment senses a traffic quality-of-service change and adjusts a data transmission bit rate, the second moment being earlier than the first moment.

Description

Data transmission method and device

[0001] The present application claims priority from the Chinese patent application No. 202410874418.9 filed on June 29, 2024, and entitled "A data transmission method and device", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] Embodiments of the present application relate to the field of communication, in particular to a data transmission method and device. BACKGROUND

[0003] In recent years, with the continuous development of the fifth generation (5G) communication system, the data transmission delay is continuously reduced, and the transmission capacity is increasingly large. The 5G communication system gradually penetrates into some multimedia services with strong real-time performance and large data capacity requirements, such as video transmission, cloud gaming (CG) and extended reality (XR), etc. These services have gradually become one of the core services in the current network.

[0004] In the current data transmission process of multimedia services, the data transmission has a relatively high delay requirement. For example, in the XR service, the video frame needs to be transmitted from the server to the user equipment within a certain time. However, due to noise interference, multiple users arriving at the same time and other reasons in the actual network, different degrees of congestion will be caused to the data transmission. Therefore, in order to match the changes of the network state, the server needs to adjust the service rate of the service according to the service quality of the probe transmission data, so as to ensure that the user equipment receives smooth video data.

[0005] However, since the server first probes and then adjusts the channel capacity of the actual network, and the server has periodic statistics on the service quality, there will be a certain response time between the server perceiving the decrease of the service quality and the decrease of the channel capacity of the actual network. Therefore, the adjustment of the service rate of the service by the server often lags behind the change of the actual network state, resulting in the user equipment receiving video transmission data playing with lag. SUMMARY

[0006] Embodiments of the present application provide a data transmission method for improving the adjustment efficiency of the service rate, and further improving the video playing fluency. The present application also provides a data transmission device, a network device, a terminal device, a computer readable storage medium and a computer program product corresponding to the data transmission method.

[0007] In a first aspect, an embodiment of the present application provides a reasoning method of a large language model, which can be executed by a network device, or by a component of the network device, such as a processor, a chip or a chip system of the network device, or by a logic module or software capable of realizing all or part of the functions of the network device. The method provided in the first aspect comprises: determining, by the network device, a target service rate, the target service rate being a predicted data transmission bit rate that the network device can provide for a user equipment at a first time point. Determining, by the network device, a target quality of service parameter based on the target service rate and a neural network model, the target quality of service parameter being used to indicate a service quality corresponding to the user equipment at the first time point, the first time point being a future time point, the neural network model being used to establish a correlation between a quality of service parameter and a service rate, an input of the neural network model comprising the quality of service parameter, and an output of the neural network model comprising the service rate. Adjusting, by the network device, a transmission behavior at a second time point based on the target quality of service parameter, so that the server or the user equipment can perceive a change in the quality of service and adjust the data transmission bit rate, wherein the second time point is earlier than the first time point.

[0008] In the embodiment of the present application, after the network device predicts the target service rate in the future, the network device can determine the target quality of service parameter corresponding to the target service rate based on the neural network model, and adjust the transmission behavior according to the target quality of service parameter, so that the server or the user equipment can perceive the change in the quality of service parameter and adjust the service rate. Compared with the prior art in which the server or the user equipment first detects and then adjusts the service rate, the data transmission method in the embodiment of the present application can adjust the service rate in advance, thereby improving the efficiency of adjusting the service rate and further improving the smoothness of video playing of the user equipment.

[0009] In a possible implementation, before the network device determines the target quality of service parameter based on the target service rate and the neural network model, the network device needs to train the neural network model. The network device receives a first message sent by the user equipment, wherein the first message carries training data, the training data being used to train the neural network model, and the training data comprising the quality of service parameter and the corresponding service rate collected by the user equipment.

[0010] In the embodiment of the present application, before the network device determines the target quality of service parameter based on the target service rate and the neural network model, the network device needs to train the neural network model based on the training data, and use the trained neural network model to infer the service rate, thereby improving the accuracy of the neural network model in calculating the service rate.

[0011] In a possible implementation, the training data includes one or more of the following: total number of data packets, total number of packet bytes, packet loss rate, time delay, jitter time, time stamp, and bit rate. The network device can determine the quality of service parameters, such as packet loss rate and delay, and the corresponding service rate, such as code rate, based on the training data.

[0012] In the embodiments of the present application, the training data includes multiple types of parameters, and therefore, the trained neural network model can infer multiple quality of service parameters, thereby improving the accuracy of the neural network model in calculating the service efficiency.

[0013] In a possible implementation, in the process of receiving the training data sent by the user equipment, the network device receives the training data sent by the user equipment based on one or more of the following: based on a medium access control layer control element (MAC CE), based on user assistance information (UAI) in a radio resource control (RRC) message, and based on pre-configured grant-free resources of the network device.

[0014] In the embodiments of the present application, the user equipment can send the training data to the network device based on multiple modes, thereby improving the implementability of the user equipment in reporting the training data to the network device.

[0015] In a possible implementation, before receiving the training data sent by the user equipment, the network device sends a second message to the user equipment, where the second message is used to configure the data type and upload period of the training data uploaded by the user equipment, and the data type includes, for example, packet loss rate, time delay, and bandwidth. The second message includes a radio resource control (RRC) configuration message, or a MAC CE, or information for configuring periodic transmission of data in future network evolution forms.

[0016] In the embodiments of the present application, the network device configures the data type and upload period of the training data uploaded by the user equipment by sending the second message, thereby improving the implementability of reporting the training data.

[0017] In a possible implementation, the network device can generate a service graph based on the neural network model, and in the process of determining the target quality of service parameter, the network device can also determine the target quality of service parameter according to the service graph. For example, the network device queries the service graph based on the candidate quality of service parameter, determines the corresponding service rate, and then selects the candidate quality of service parameter as the target quality of service parameter according to the target service rate.

[0018] In the embodiments of the present application, the network device can generate a service graph based on the neural network model, and query the service rate corresponding to the quality of service parameter according to the service graph, thereby improving the output efficiency of the service rate.

[0019] In a possible implementation, in the process of determining the target service rate of the service, the network device predicts the target service rate of the service based on scheduling awareness information, the scheduling awareness information including one or more of the following: user equipment location information, user equipment historical data, future possible user arrival and network load information. The user equipment location information indicates a location where the user equipment is located, and the historical data of the user equipment can include network usage patterns, service request frequencies and data transmission amounts of the user equipment in the past period of time. The future user arrival can be how many user equipments the network device estimates to join the network in a future time period. The network load information can be, for example, total bandwidth utilization, traffic conditions of specific applications, congestion degrees and the like.

[0020] In the embodiments of the present application, the network device can predict the target service rate of the service at a future time based on the scheduling awareness information, so as to adjust the transmission behavior, and also enable the server or the user equipment to perceive the change of the service quality parameter and adjust the service rate, thereby improving the adjustment efficiency of the service rate.

[0021] In a possible implementation, the neural network model is a non-memory network model, and in the process of determining the target service quality parameter based on the target service rate of the service and the neural network model, the network device determines one or more candidate target service quality parameters based on the current service quality parameter. The network device inputs the one or more candidate target service quality parameters into the non-memory network model. When a difference between an output result of the non-memory network model and the target service rate of the service is less than a threshold, the network device determines the corresponding candidate target service parameter as the target service quality parameter.

[0022] The neural network model in the embodiments of the present application can be a non-memory network model, and the network device inputs the candidate target service quality parameter determined based on the current service quality parameter into the non-memory network model, so as to calculate the corresponding service rate, thereby improving the calculation efficiency of the service rate.

[0023] In a possible implementation, the neural network model is a non-memory network model, and in the process of determining the target service quality parameter based on the target service rate of the service and the neural network model, the network device determines one or more candidate target service quality parameters based on the current service quality parameter. The network device inputs the one or more candidate target service quality parameters into the non-memory network model. When a difference between an output result of the non-memory network model and the target service rate of the service is less than a threshold, the network device determines the corresponding candidate target service parameter as the target service quality parameter.

[0024] The neural network model in the embodiments of the present application can also be a memory network model. The network device inputs the candidate target service quality parameter determined based on the current service service parameter and the service quality parameters of multiple historical periods into the memory network model, and calculates a corresponding service service rate, thereby improving the calculation accuracy of the service service rate.

[0025] In a possible implementation, the network device sends a third message to the user equipment, the third message being used to instruct the user equipment to send one or more pieces of inference data, including the current service quality parameter and the historical service quality parameters of the historical period, to the network device, and the third message includes downlink control information (DCI).

[0026] In the embodiments of the present application, the network device can send a third message to the user equipment to instruct the user equipment to send inference data, thereby improving the implementability of the user equipment sending inference data to the network device.

[0027] In a possible implementation, the network device receives a fourth message sent by the user equipment, the fourth message being used to carry inference data, and the fourth message includes a message sent based on a medium access control (MAC) control element (CE).

[0028] In the embodiments of the present application, the user equipment can carry inference data through the fourth message, thereby improving the implementability of the user equipment sending inference data to the network device.

[0029] In a possible implementation, in the process of adjusting the transmission behavior of the network device at the second time based on the target service quality parameter, the network device performs one or more operations, including packet loss processing and increasing latency, on the transmission data in the transmission period.

[0030] In the embodiments of the present application, the network device adjusts the transmission behavior by performing packet loss processing or increasing latency on the transmission data, thereby improving the implementability of the network device adjusting the transmission behavior.

[0031] In a second aspect, the embodiments of the present application provide a data transmission apparatus, which includes a transceiver unit and a processing unit. The processing unit is configured to determine a target service rate, the target service rate being a predicted data transmission bit rate that the network device can provide for the user equipment at a first time. The processing unit is further configured to determine a target service quality parameter based on the target service rate and a neural network model, the target service quality parameter being used to indicate a corresponding service quality of the user equipment at the first time, an input of the neural network model including a service quality parameter, and an output of the neural network model including a service rate. The processing unit is further configured to adjust a transmission behavior at a second time based on the target service quality parameter, so that the server or the user equipment perceives a change in the service quality and adjusts a data transmission bit rate, the second time being earlier than the first time.

[0032] In a possible implementation, the transceiver is configured to receive a first message sent by the user equipment, the first message carrying training data, the training data being used for training the neural network model, and the training data including quality of service parameters and corresponding service rates collected by the user equipment.

[0033] In a possible implementation, the training data includes one or more of the following: total number of data packets, total number of packet bytes, packet loss rate, time delay, jitter time, time stamp, and bit rate.

[0034] In a possible implementation, the transceiver is further configured to send a second message to the user equipment, the second message being used for configuring data types and uploading periods of the training data uploaded by the user equipment.

[0035] In a possible implementation, the processing unit is specifically configured to predict the target service rate based on scheduling awareness information, the scheduling awareness information including one or more of the following: user equipment location information, user equipment historical data, future possible user arrival, and network load information.

[0036] In a possible implementation, the neural network model is a non-memory network model, and the processing unit is specifically configured to determine one or more candidate target service quality parameters based on the current quality of service parameters, input the one or more candidate target service quality parameters into the non-memory network model, and determine the corresponding candidate target service rate as the target service quality parameter when a difference between an output result of the non-memory network model and the target service rate is less than a threshold.

[0037] In a possible implementation, the neural network model is a memory network model, and the processing unit is specifically configured to determine one or more candidate target service quality parameters based on the current quality of service parameters, input the one or more candidate target service quality parameters and historical quality of service parameters of a historical period into the memory network model, and determine the corresponding candidate target service rate as the target service quality parameter when a difference between an output result of the memory network model and the target service rate is less than a threshold.

[0038] In a possible implementation, the transceiver is further configured to send a third message to the user equipment, the third message being used for instructing the user equipment to send one or more of the following inference data to the network device: the current quality of service parameters and the historical quality of service parameters of the historical period.

[0039] In a possible implementation, the transceiver is further configured to receive a fourth message sent by the user equipment, the fourth message being used for carrying the inference data.

[0040] In a third aspect, an embodiment of the present application provides a network device, the network device comprising a processor and a memory coupled to the processor, the memory being configured to store instructions which, when executed by the processor, cause the network device to perform the method in the first aspect or any possible implementation of the first aspect.

[0041] In a fourth aspect, an embodiment of the present application provides a terminal device, the terminal device comprising a processor and a memory coupled to the processor, the memory being configured to store instructions which, when executed by the processor, cause the terminal device to perform the method performed by the user equipment in the first aspect or any possible implementation of the first aspect.

[0042] In a fifth aspect, an embodiment of the present application provides a communication apparatus, the communication apparatus being a chip system, the chip system comprising a processor, a memory, a transceiver and an antenna, the memory being configured to store instructions which, when executed by the processor, cause the communication apparatus to perform the method in the first aspect or any possible implementation of the first aspect.

[0043] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium, the computer-readable storage medium storing instructions, the instructions being executed to cause a computer to perform the method in the first aspect or any possible implementation of the first aspect.

[0044] In a seventh aspect, an embodiment of the present application provides a computer program product, the computer program product comprising instructions, the instructions being executed to cause a computer to implement the method in the first aspect or any possible implementation of the first aspect.

[0045] It can be understood that the beneficial effects that can be achieved by any one of the data transmission apparatus, the network device, the terminal device, the communication apparatus, the computer-readable medium or the computer program product provided above can refer to the beneficial effects in the corresponding method, which will not be described herein again. BRIEF DESCRIPTION OF DRAWINGS

[0046] FIG. 1a is a schematic diagram of a system architecture of a data transmission system according to an embodiment of the present application;

[0047] FIG. 1b is a schematic diagram of a system architecture of another data transmission system according to an embodiment of the present application;

[0048] FIG. 2 is a schematic diagram of a flow of a data transmission method according to an embodiment of the present application;

[0049] FIG. 3 is a schematic diagram of a flow of another data transmission method according to an embodiment of the present application;

[0050] FIG. 4 is a schematic diagram of a load content of a first message according to an embodiment of the present application;

[0051] FIG. 5 is a structural schematic diagram of a neural network model according to an embodiment of the present application;

[0052] FIG. 6 is a flowchart of another data transmission method according to an embodiment of the present application;

[0053] FIG. 7 is a structural schematic diagram of a data transmission apparatus according to an embodiment of the present application;

[0054] FIG. 8 is a structural schematic diagram of a communication apparatus according to an embodiment of the present application;

[0055] FIG. 9 is a structural schematic diagram of a communication device according to an embodiment of the present application. DETAILED DESCRIPTION

[0056] The embodiments of the present application provide a data transmission method and apparatus, which are used for improving the adjustment efficiency of service rate adjustment.

[0057] The terms "first", "second", "third", "fourth" and the like in the description and the claims of the present application and the above-mentioned drawings (if any) are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product or device.

[0058] In the embodiments of the present application, the words "exemplary" or "for example" are used to mean serving as an example, instance, or illustration. Any embodiment or design presented as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of "exemplary" or "for example" is intended to present concepts in a concrete manner.

[0059] First, some terms involved in the embodiments of the present application are introduced, so as to facilitate the understanding of the technical solutions by those skilled in the art.

[0060] The radio resource control (RRC) layer is a key component in wireless communication networks, primarily responsible for managing, controlling, and scheduling wireless resources. Through intelligent scheduling algorithms, RRC dynamically allocates and adjusts wireless resources based on network load, user demand, and other factors. RRC also has error detection and correction capabilities, real-time monitoring of data transmission quality, and provides core functions such as broadcast system information and RRC connection control, which are important means to ensure network service quality, improve spectrum utilization, and system reliability.

[0061] The media access control (MAC) layer is a sublayer in the data link layer, responsible for processing data frame encapsulation, decapsulation, and controlling access to shared transmission media by multiple devices. The MAC layer defines the transmission rules of data frames on the medium, controls the sending and receiving process of data, and realizes data exchange between devices. The MAC layer also has functions such as flow control and error detection to ensure reliable data transmission. In wireless communication systems, the MAC layer closely cooperates with the physical layer to jointly realize wireless data transmission and management.

[0062] The real-time transport control protocol (RTCP) is a companion protocol of the real-time transport protocol (RTP), used to provide statistical information and control functions for data transmission. RTCP sends control packets regularly to collect key indicators such as packet loss rate and transmission delay during data transmission, and feeds back these information to the sender and receiver. The sender can adjust the data transmission strategy according to the RTCP feedback information to improve the performance and stability of real-time communication.

[0063] User assistant information (UAI) is additional information used to assist user operation or improve user experience in specific applications or systems. These information may include user's personal preferences, historical behavior records, context data, etc. By collecting and analyzing UAI, the system can better understand user needs and provide more personalized and intelligent services. For example, in a recommendation system, UAI can be used to analyze user's interests and preferences to recommend more suitable content to the user.

[0064] In order to make the technical solutions of the present application more clear and easy to understand, the system architecture of the present application is introduced below in conjunction with the drawings.

[0065] Please refer to FIG. 1a, which is a schematic diagram of a system architecture of a data transmission system according to an embodiment of the present application. In the example shown in FIG. 1a, the data transmission system 10 includes a server 101, a network device 102, and a terminal device 103, wherein the server 101 can also be referred to as a source server or a data network, and the network device 102 can be one or more network elements in a wireless communication network, including core network devices and access network devices. The functions of the various modules in the data transmission system 10 are described below.

[0066] The server 101 is configured to generate transmission data and transmit the transmission data to the terminal device 103 through the network device 102, i.e., downlink transmission, wherein the transmission data includes video data in extended reality (XR) services and video transmission services. The server 101 is also configured to receive data sent by the terminal device 103, i.e., uplink transmission. In the downlink transmission, when the transmission data is video data in XR services, the server 101 encodes and renders the original video data and transmits the generated video data to the terminal device 103 through the core network device and the access network device. Since there is a strict delay requirement for video data transmission in XR services, the data transmission system 10 needs to transmit data from the server 101 to the terminal device 103 within a certain time, for example, the delay from the server 101 to the terminal device 103 is not more than 70 ms.

[0067] The network device 102 is configured to perform data transmission between the server 101 and the terminal device 103, and the network device 102 includes core network devices and access network devices. The core network device refers to a device in the core network (CN) that provides service support for the terminal device 103, and is configured to implement three major functions of registration, connection, and session management. For example, the core network device can be a network exposure function (NEF) network element, a policy control function (PCF) network element, an application function (AF) network element, an access and mobility management function (AMF) network element, a session management function (SMF) network element, a user plane function (UPF) network element, and the like.

[0068] An access network (AN) device can be any kind of device with wireless transceiver function, which can be used to be responsible for air interface related functions, such as wireless link maintenance function, wireless resource management function, part of mobility management function. The access network device includes but is not limited to: evolved NodeB (NodeB or eNB) in LTE, base station (gNodeB or gNB) in NR, transmission teception point (TRP), base station of subsequent evolution of 3GPP, access node in WiFi system, wireless relay node and wireless backhaul node, etc. Among them, the base station can be a macro base station, a micro base station, a pico base station, a small station, a relay station and a balloon station, etc.

[0069] The terminal device 103 includes a device that provides voice and data connectivity to a user. For example, the terminal device 103 can include a handheld device having a wireless connection function or a processing device connected to a wireless modem. The terminal device 103 can communicate with the core network device through the access network device and exchange voice and data through the access network device.

[0070] The terminal device 103 can also be referred to as a terminal, user equipment (UE), wireless terminal device, mobile terminal (MT) device, subscriber unit, subscriber station, mobile station (MS), mobile, remote station, access point (AP), remote terminal, access terminal, user terminal, user agent and user device, etc., without limitation.

[0071] In addition, the terminal device 103 can be a mobile phone, a tablet computer (Pad), a computer with wireless transceiver function, a virtual reality (VR) device, an augmented reality (AR) device, an extended reality (XR) service terminal, a cloud gaming (CG) service terminal, a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in remote medical surgery, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city, a wireless terminal in smart home, and the like.

[0072] The data method provided in the application can be applied to a 4th generation mobile communication technology (4G) system, a 5th generation mobile communication technology (5G) system, a 6th generation mobile communication technology (6G) system, and subsequent evolution systems, and the application is not limited in this regard. In the following, the 5th generation mobile communication technology system is taken as an example for introduction.

[0073] Please refer to FIG. 1b, which is a schematic diagram of another data transmission system provided in an embodiment of the application. In the example shown in FIG. 1b, the data transmission system is a 5th generation mobile communication technology system, and the network elements involved in the data transmission system include a server, a data network DN, a core network, an access network AN, and a user equipment UE. The core network includes a network exposure function (NEF) network element, a policy control function (PCF) network element, an application function (AF) network element, an access and mobility management function (AMF) network element, a session management function (SMF) network element, and a user plane function (UPF) network element. The functions of the network elements in the core network are introduced in detail below.

[0074] The network exposure function (NEF) network element is used to expose the services and capabilities of the 3GPP network functions to the application function (AF) network element, and also allows the application function (AF) network element to provide information to the 3GPP network functions.

[0075] The policy control function entity PCF network element is used for policy management of charging policy and QoS policy, and contains user subscription data management function, policy control function, charging policy control function, quality of service (QoS) control, etc. It should be pointed out that in actual network, PCF can also be divided into session management PCF (SM-PCF) and access management PCF (AM-PCF).

[0076] The application function AF network element mainly transmits the demand of the application side to the network side.

[0077] The access and mobility management AMF network element is used for mobility management, access authentication, authorization, etc. In addition, it is also responsible for transmitting user policy between UE and PCF. The access and mobility management AMF network element terminates non access stratum (NAS) messages, completes registration management, connection management, and reachability management, allocates a track area list (TA list) and mobility management, etc., and transparently routes session management (SM) messages to the session management network element.

[0078] The session management function module SMF network element is mainly used for session management in a mobile network, such as session establishment, modification, and release. Specifically, the session management function module SMF network element is used to complete UE IP address allocation, UPF selection, charging and QoS policy control, and other session management functions.

[0079] The user plane function UPF network element is an interface with a data network, also known as a user plane entity of a core network. The user plane function UPF network element is used to perform functions related to the user plane of the core network. For example, the UPF network element is mainly responsible for connecting external networks and transmitting service data, including user plane data forwarding, session-based charging statistics, bandwidth limitation, etc.

[0080] Based on the data transmission system 10 shown in FIG. 1a, the present application also provides a data transmission method. The data transmission method provided by the embodiments of the present application will be introduced below in combination with embodiments.

[0081] Please refer to FIG. 2, which is a flowchart of a data transmission method provided by an embodiment of the present application. In the example shown in FIG. 2, the method includes the following steps:

[0082] Step 201. The network device determines a target service rate, which is a predicted data transmission bit rate that the network device can provide for the user equipment at a first time.

[0083] The network device 102 first determines a target service rate before determining the target service quality parameter, the target service rate being a data transmission bit rate that the network device 102 can provide for the terminal device 103 at a first time, the target service rate reflecting a service capability of the network device 102, and since the target service rate is a predicted data transmission bit rate, the first time can be a future time.

[0084] Specifically, in the process of determining the target service rate, the network device 102 predicts the target service rate based on scheduling awareness information, for example, the network device 102 predicts the target service rate at t time based on the scheduling awareness information, the target service rate being a predicted target service rate at t+1 time. The scheduling awareness information includes one or more of the following: user device location information, user device historical data, future possible user arrival and network load information.

[0085] The user device location information indicates a location where the user device is located, and the historical data of the user device can include a network usage mode, a service request frequency and a data transmission amount of the user device in a past period of time, and the like. The future user arrival can be how many user devices the network device estimates to join the network in a future time period. The network load information is, for example, a total bandwidth utilization rate, a traffic condition of a specific application, a congestion degree, and the like.

[0086] It should be noted that the service rate in the embodiments of the present application can also be referred to as a transmission rate, a transmission bit rate and a transmission code rate in different descriptions, and the specific distinction is not made.

[0087] Step 202. The network device determines a target service quality parameter based on the target service rate and a neural network model, the target service quality parameter being used to indicate a service quality of the user device at the first time, an input of the neural network model including a service quality parameter, and an output of the neural network model including a service rate.

[0088] After the network device 102 predicts the target service rate, the network device 102 determines a target service quality parameter based on the target service rate and the neural network model, where the target service quality parameter is used to indicate a service quality of the terminal device 103 or the server 101 at the first time, and the server 101 or the terminal device 103 adjusts the service rate with the target service quality parameter as a target, so that an actual service quality parameter reaches the target service quality parameter at the first time. In the embodiments of the present application, the neural network model is used to establish a correlation between the service quality parameter and the service rate, and the input of the neural network model includes the service quality parameter, and the output of the neural network model includes the service rate.

[0089] Specifically, in the process of determining the target service quality parameter by the network device 102, first, a candidate service quality parameter is determined based on the current target service quality parameter, and the candidate service quality parameter is input into the neural network model, and the service rate corresponding to the candidate service quality parameter can be output by the neural network model, and when the calculated service rate meets the target service rate, the candidate service quality parameter is determined as the target service quality parameter. The service quality parameter includes a packet loss rate and a time delay.

[0090] It should be noted that the network device 102 can generate a service graph based on the neural network model, and the service graph includes a corresponding relationship between the service quality parameter and the service rate. In the process of determining the target service quality parameter by the network device 102, the network device 102 can also determine the target service quality parameter according to the service graph, for example, the network device 102 queries the service graph based on the candidate service quality parameter to determine the corresponding service rate, and then selects the candidate service quality parameter as the target service quality parameter according to the target service rate.

[0091] Before the network device 102 infers the target service quality parameter by using the neural network model, the network device 102 needs to train the neural network model, and the network device 102 receives a first message sent by a user device, where the first message carries training data, the training data is used to train the neural network model, and the training data includes service quality parameters and corresponding service rates collected by the user device. Specifically, the training data includes one or more of the following: a total number of data packets, a total number of packet bytes, a packet loss rate, a time delay, a jitter time, a time stamp, and a bit rate.

[0092] In a possible implementation, before the network device 102 receives the training data sent by the terminal device 103, the network device 102 sends a second message to the terminal device, the second message being used to instruct the terminal device 103 to report the training data, and the second message further being used to configure a data type and a reporting period of the training data reported by the terminal device 103, where the second message includes a radio resource control (RRC) configuration message and a medium access control (MAC) control element (CE) configuration message.

[0093] The RRC configuration refers to that the network device 102 configures the terminal device 103 to report the training data through an RRC message. For example, after the terminal device 103 receives the RRC message, the terminal device 103 needs to parse and store the RRC message in a high layer. Then, when the training data needs to be reported, the terminal device 103 transmits the training data to a base station of the network device 102 through a physical uplink shared channel (PUSCH).

[0094] The MAC CE configuration refers to that the network device 102 defines a new MAC CE to instruct or activate the terminal device 103 to report the training data. When the terminal device 103 needs to report the training data, the network device 102 sends the MAC CE to the terminal device 103. After receiving the MAC CE, the terminal device 103 directly triggers a reporting process in a MAC layer.

[0095] Referring to FIG. 3, FIG. 3 is a flowchart of a data transmission method provided in an embodiment of the present application. In the example shown in FIG. 3, the data transmission method includes a training phase and an inference phase. In the training phase, the network device 102 trains a neural network model based on training data. In the inference phase, the network device 102 outputs a service rate of a service corresponding to a candidate target service quality parameter based on the trained neural network model, and determines a target service quality parameter.

[0096] In step 1 of the example shown in FIG. 3, the network device 102 sends a second message to the terminal device 103, the second message being used to instruct the terminal device 103 to report training data, the training data being data used by the network device 102 to train the neural network model. For example, in this example, the training data includes a total number of data packets, a total number of packet bytes, a packet loss rate, a jitter time, and a timestamp in a real-time transport control protocol (RTCP) layer. The network device 102 can determine a service quality parameter and a corresponding service rate of a service according to the information.

[0097] Please continue to refer to FIG. 3, in step 1 of the example shown in FIG. 3, the second message sent by the network device 102 to the terminal device 103 can also configure the terminal device 103 to report the data type and upload period of the training data, for example, the data type such as packet loss rate, delay, and bandwidth, etc. For example, the network device 102 configures the terminal device 103 to report the data type and upload period of the training data by sending a radio resource control (RRC) message, or the network device 102 can also trigger or instruct the terminal device 103 to report the training data by a medium access control (MAC) control element (CE), and the network device 102 can periodically report the training data after triggering once.

[0098] In steps 2 to 3 of the example shown in FIG. 3, after the network device 102 sends the second message to the terminal device 103, the terminal device 103 collects the training data, which includes the quality of service (QoS) parameter and the corresponding service rate, for example, the terminal device 103 reads the QoS parameter in the real-time transport control protocol (RTCP) and calculates the corresponding service rate. After the terminal device 103 collects the training data, the terminal device 103 reports the training data to the network device 102 according to the period configured by the network device 102, that is, the terminal device 103 sends the first message to the network device 102.

[0099] In a possible implementation, in the process of the terminal device 103 sending the training data to the network device 102, the network device 102 receives the training data sent by the terminal device 103 based on one or more of the following manners: based on a medium access control (MAC) control element (CE), based on user assistance information (UAI) in a radio resource control (RRC) message, and based on pre-configured grant-free resources of the network device.

[0100] When the terminal device 103 sends the training data based on the MAC CE, the training data can carry the payload part of the first message, wherein the data precision of the payload part can be the same as that in the real-time transport control (RTCP) protocol, for example, can be INT8 or FLOAT16.

[0101] Please refer to FIG. 4, which is a schematic diagram of the payload content of the first message provided by an embodiment of the present application. In the example shown in FIG. 4, the payload content of the first message sent by the terminal device 103 to the network device 102 is the payload in the newly added MAC CE, and the payload content is the training data, which includes packet loss, delay, bit rate, and timestamp information.

[0102] When the terminal device 103 transmits training data based on the preconfigured grant-free resource of the network device 102, the network device 102 can configure a resource for the terminal device 103 to transmit the training data, for example, the network device 102 can configure the resource in the form of configured grant.

[0103] When the terminal device 103 transmits training data based on the radio resource control (RRC) message, the terminal device 103 can report the combination of the packet loss, the delay, the bit rate, etc. through the user assistance information (UAI) in the RRC message.

[0104] After the network device 102 trains the neural network model, the network device 102 performs inference based on the neural network model and determines the target service quality parameter based on the inference result. The neural network model in the embodiment of the present application includes a memoryless network model and a memory network model. When the neural network model is a memoryless network model, the input of the neural network model includes a candidate target service quality parameter determined based on the current service quality parameter, and the output of the neural network model is the service rate. When the neural network model is a memory network model, the input of the neural network model includes the service quality parameters of a plurality of historical periods in addition to the candidate target service quality parameter determined based on the current service quality parameter.

[0105] Please refer to FIG. 5, which is a schematic diagram of a neural network model provided by an embodiment of the present application. In the example shown in FIG. 5, the input of the neural network model includes the service quality parameter, for example, the service quality parameter includes the packet loss, the delay, the delay interval, and the throughput, etc. The output of the neural network model includes the service rate, for example, the service rate is the bit rate.

[0106] In the example shown in FIG. 5, the neural network model is composed of a multi-layer structure, each layer is composed of a plurality of neurons, and the multi-layer structure includes an input layer, a hidden layer, and an output layer. The input layer is used to receive external input data. Each neuron in the input layer corresponds to a feature of the input data.

[0107] In the example shown in FIG. 5, the hidden layer is located between the input layer and the output layer, and the hidden layer can contain multiple layers, each layer having a plurality of neurons. These neurons are connected to the neurons of the previous layer through weights and biases. The hidden layer is used to extract and convert features from the input data to capture complex correlation relationships in the data. The output layer is the last layer of the neural network model, and the output layer is used to generate the prediction result of the model.

[0108] The following describes the manner in which the network device 102 determines the target service quality parameter under two different neural network models.

[0109] In one possible implementation, the neural network model is a non-memory network model. In determining the target service quality parameter based on the target service rate and the neural network model, the network device 102 determines one or more candidate target service quality parameters based on the current service quality parameter. The network device 102 inputs the one or more candidate target service quality parameters into the non-memory network model. When the difference between the output of the non-memory network model and the target service rate is less than a threshold, the network device 102 determines the corresponding candidate target service parameter as the target service quality parameter.

[0110] It should be noted that in determining the candidate target service quality parameter, the network device 102 can also not be based on the current service quality parameter, but can search or exhaustively determine the candidate target service quality parameter within the instruction range, which is not limited.

[0111] Please continue to refer to FIG. 3. In the inference stage of the data transmission method shown in FIG. 3, in step 4, in determining the target service quality parameter based on the target service rate and the neural network model, the network device 102 first determines one or more candidate target service quality parameters based on the current service quality parameter. For example, the network device 102 determines that the current service quality parameter of the service includes a packet loss rate and a delay, where the packet loss rate is 0% and the delay is 10 ms. Further, the network device 102 can determine one or more candidate target quality parameters within a range near the current service quality parameter. For example, the network device 102 determines that the candidate target service quality parameter 1 is a packet loss rate of 1% and a delay of 10 ms, and determines that the candidate target service quality parameter 2 is a packet loss rate of 0% and a delay of 15 ms.

[0112] In step 4 of the example shown in FIG. 3, the network device 102 determines one or more candidate target quality parameters. The network device 102 inputs the one or more candidate target service quality parameters into the non-memory neural network model. For example, the network device 102 inputs the candidate target service quality parameter 1 and the candidate target service quality parameter 2 into the non-memory neural network model. The non-memory neural network model outputs that the service rate corresponding to the candidate target service quality parameter 1 is 10 Mbps, and the service rate corresponding to the candidate target service quality parameter 2 is 10 Mbps.

[0113] In step 4 of the example shown in FIG. 3, when the difference between the output result of the network model without memory and the target service rate is less than a threshold value, for example, 1 Mbps or 2 Mbps, the network device 102 determines the corresponding candidate target service parameter as the target service quality parameter. For example, the current service rate of the network device 102 is 30 Mbps, and the prediction result is that the target service rate at the first future time is 10 Mbps. Since the service rate corresponding to the candidate target service quality parameter 1 and the service rate corresponding to the candidate target service quality parameter 2 are both 10 Mbps, the candidate target service quality parameter 1 and the candidate target service quality parameter 2 can both be the target service quality parameter.

[0114] In a possible implementation, when the neural network model is a network model with memory, in the process of determining the target service quality parameter based on the target service rate and the neural network model, the network device 102 determines one or more candidate target service quality parameters based on the current service quality parameter. The network device 102 inputs the one or more candidate target service quality parameters and the historical service quality parameters of the historical period into the network model with memory. When the difference between the output result of the network model with memory and the target service rate is less than a threshold value, the corresponding candidate target service parameter is determined as the target service quality parameter.

[0115] Referring to FIG. 6, FIG. 6 is a flowchart of another data transmission method provided by an embodiment of the present application. In steps 1 to 3 of the example shown in FIG. 6, the process of training the neural network model by the network device 102 is similar to the process of training the neural network model by the network device 102 in the example shown in FIG. 3, which will not be described herein again.

[0116] In step 4 of the example shown in FIG. 6, in the process of determining the target service quality parameter based on the target service rate and the neural network model, since the neural network model is a network model with memory, after the network device 102 determines the candidate target service quality parameter, the network device 102 inputs the one or more candidate target service quality parameters and the historical service quality parameters of the historical period, for example, the service quality parameters of N periods, into the neural network model with memory, and the neural network model outputs the service rate.

[0117] In a possible implementation, the network device 102 sends a third message to the terminal device 103, where the third message is used to instruct the terminal device 103 to send one or more pieces of inference data, including a current service quality parameter and a historical service quality parameter of a historical period, to the network device 102, and the third message includes a downlink control information (DCI) or a MAC CE. The network device 102 receives a fourth message sent by the terminal device 103, where the fourth message is used to carry the inference data, and the fourth message includes a message sent based on a media access control (MAC) control element (CE) or uplink control information (UCI).

[0118] Please continue to refer to FIG. 6. In the example shown in FIG. 6, because the neural network model is a memory network model, when the network device 102 calculates the service rate based on the memory neural network model, the historical service quality parameter of the historical period needs to be input into the memory neural network model. Therefore, the network device 102 needs to configure the terminal device 103 to upload inference data, where the inference data includes the historical service quality parameter of the historical period.

[0119] In steps 5 to 6 of the example shown in FIG. 6, the network device 102 sends a third message to the terminal device 103, where the third message is used to instruct the terminal device 103 to report inference data, and the inference data includes one or more of a current service quality parameter and a historical service quality parameter of a historical period. For example, the third message can be a downlink control information (DCI). Then, the terminal device 103 sends a fourth message to the network device, where the fourth message carries the inference data. For example, the fourth message can be a message sent based on a media access control (MAC) control element (CE).

[0120] Step 203. The network device adjusts a transmission behavior at a second time based on the target service quality parameter, so that the server or the user equipment perceives a change in the service quality and adjusts a data transmission bit rate, and the second time is earlier than the first time.

[0121] After the network device 102 determines the target service quality parameter based on the neural network model, the network device 102 adjusts a transmission behavior at a second time according to the target service quality parameter, so that the server 101 or the terminal device 103 perceives a change in the service quality and adjusts a data transmission bit rate, and the second time is earlier than the first time. For example, the second time can be the current time, and the first time can be a future time.

[0122] After the network device 102 adjusts the transmission behavior at the second time according to the target service quality parameter, the server 101 or the terminal device 103 can perceive a change in the service quality before the first time, so as to adjust the service rate, so that the service rate approaches the target service rate at the first time.

[0123] It can be understood that, in the downlink data transmission scenario, the network device 102 adjusts the transmission behavior at the second time according to the target service quality parameter, and the server 101 perceives the change of the service quality parameter and adjusts the service service rate. In the uplink data transmission scenario, the network device 102 adjusts the transmission behavior at the second time according to the target service quality parameter, and the terminal device 103 perceives the change of the service quality parameter and adjusts the service service rate.

[0124] In a possible implementation, in the process in which the network device 102 adjusts the transmission behavior at the second time based on the target service quality parameter, the network device 102 performs one or more of the following operations on the transmission data in the transmission period: packet loss processing and increasing latency.

[0125] Please continue to refer to FIG. 3, in step 5 of the example shown in FIG. 3, the network device 102 adjusts the transmission behavior at the second time based on the target service quality parameter, for example, the target service quality parameter determined by the network device 102 is a packet loss rate of 1% and a delay of 10 ms, and a packet loss rate of 0% and a delay of 15 ms, assuming that the current service quality parameter of the service is a packet loss rate of 0% and a delay of 10 ms, then the network device 102 discards 1% of the data packets in a statistical period, or the network device 102 increases the delay by 5 ms in a statistical period.

[0126] Please continue to refer to FIG. 6, in steps 7 to 8 of the example shown in FIG. 6, after the network device 102 determines the target service quality parameter based on the target service rate of the service and the recurrent neural network, the network device 102 adjusts the transmission behavior at the second time based on the target service quality parameter, which is similar to step 5 of the example shown in FIG. 3, and details are not repeated here.

[0127] As can be seen from the above embodiments, after predicting the target service rate of the service in the future, the network device in the embodiments of the present application can determine the target service quality parameter corresponding to the target service rate of the service based on the neural network model, and adjust the transmission behavior according to the target service quality parameter, so that the server can perceive the change of the service quality parameter and adjust the service service rate, thereby improving the service rate adjustment efficiency.

[0128] Based on the above method embodiments, the embodiments of the present application also provide a data transmission device, and the data transmission device provided by the embodiments of the present application is specifically introduced as follows.

[0129] Please refer to FIG. 7, which is a structural schematic diagram of a data transmission device provided by an embodiment of the present application. In the example shown in FIG. 7, the data transmission device 700 is used to implement each step performed by the network device in each embodiment described above, and the data transmission device 700 includes a transceiver unit 701 and a processing unit 702.

[0130] The processing unit 702 is configured to determine a target service rate, the target service rate being a predicted data transmission bit rate that the network device is able to provide for the user equipment at a first time point. The processing unit 702 is further configured to determine a target service quality parameter based on the target service rate and a neural network model, the target service quality parameter being used to indicate a service quality corresponding to the user equipment at the first time point, an input of the neural network model comprising a service quality parameter, and an output of the neural network model comprising a service rate. The processing unit 702 is further configured to adjust a transmission behavior at a second time point based on the target service quality parameter, so that a server or the user equipment is aware of a change in the service quality and adjusts a data transmission bit rate, the second time point being earlier than the first time point.

[0131] In a possible implementation, the transceiver 701 is configured to receive a first message sent by the user equipment, the first message carrying training data, the training data being used to train the neural network model, and the training data comprising a service quality parameter collected by the user equipment and a corresponding service rate.

[0132] In a possible implementation, the training data comprises one or more of the following: a total number of data packets, a total number of packet bytes, a packet loss rate, a time delay, a jitter time, a time stamp, and a bit rate.

[0133] In a possible implementation, the transceiver 701 is further configured to send a second message to the user equipment, the second message being used to configure a data type and an upload period of the training data uploaded by the user equipment.

[0134] In a possible implementation, the processing unit 702 is specifically configured to predict the target service rate based on scheduling awareness information, the scheduling awareness information comprising one or more of the following: user equipment location information, user equipment history data, future possible user arrival, and network load information.

[0135] In a possible implementation, the neural network model is a non-memory network model, and the processing unit 702 is specifically configured to determine one or more candidate target service quality parameters based on a current service quality parameter, input the one or more candidate target service quality parameters into the non-memory network model, and determine a corresponding candidate target service quality parameter as the target service quality parameter when a difference between an output result of the non-memory network model and the target service rate is less than a threshold.

[0136] In a possible implementation, the neural network model is a memory network model, and the processing unit 702 is specifically configured to determine one or more candidate target service quality parameters based on the current service quality parameter, and input the one or more candidate target service quality parameters and the historical service quality parameters of the historical period into the memory network model. When the difference between the output result of the memory network model and the target service rate is less than a threshold, the corresponding candidate target service parameter is determined as the target service quality parameter.

[0137] In a possible implementation, the transceiver 701 is further configured to send a third message to the user equipment, where the third message is used to instruct the user equipment to send one or more pieces of inference data, including the current service quality parameter and the historical service quality parameters of the historical period, to the network device.

[0138] In a possible implementation, the transceiver 701 is further configured to receive a fourth message sent by the user equipment, where the fourth message carries the inference data.

[0139] [Corrected according to Rule 91 on 13.05.2025] It can be understood that the transceiver 701 and the processing unit 702 in the data transmission apparatus 700 can be mapped to the various modules in the data transmission system 10 in FIG. 1a as functional modules, thereby realizing the functions of the various modules in the data transmission system.

[0140] It should be understood that the division of the units in the above apparatus is only a logical functional division, and in actual implementation, all or part of them can be integrated into one physical entity, or can be physically separated. The units in the apparatus can all be implemented in the form of software called by a processing element; or all be implemented in the form of hardware; or part of the units are implemented in the form of software called by a processing element, and part of the units are implemented in the form of hardware. For example, each unit can be a separately established processing element, or can be integrated in a chip of the apparatus, in addition, the unit can also be stored in the form of a program in a memory, and the function of the unit is called and executed by a processing element of the apparatus. In addition, all or part of the units can be integrated together, or can be independently implemented. The processing element described herein can be a processor, which can be an integrated circuit with a signal processing capability. In the implementation process, each step of the above method or each unit can be implemented by an integrated logic circuit of hardware in the processing element, or in the form of software called by the processing element.

[0141] It is to be understood that all the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. It is also to be understood that the use of "comprising" or "including" or "having" one or more elements or steps does not exclude the presence of one or more additional elements or steps. It is further to be understood that the use of "a" or "an" in the description refers to "one or more" of something unless context clearly indicates otherwise.

[0142] Any other reasonable combination of steps, which can be conceived by those skilled in the art based on the above description, is also within the scope of the present application. It is also to be understood that the embodiments described in the specification are preferred embodiments and the steps involved are not necessarily essential to the application.

[0143] Referring to FIG. 8, FIG. 8 is a schematic diagram of a system structure of a communication apparatus according to an embodiment of the present application. In the example shown in FIG. 8, the communication apparatus 800 can be a chip system. The chip system can be composed of a chip or can include a chip and other discrete devices. The communication apparatus 800 includes one or more processors 801 for implementing or supporting implementation of the functions of the method of the present application by the communication apparatus 800.

[0144] The processor 801 can also be referred to as a processing unit or a processing module, and can implement certain control functions. The processor can be a general purpose processor or a special purpose processor. For example, it includes a central processing unit, an application processor, a modem processor, a graphics processor, an image signal processor, a digital signal processor, a video coding and decoding processor, a controller, a memory, and / or a neural network processor, etc. The central processing unit can be used to control the communication apparatus 800, execute software programs and / or process data. Different processors can be independent devices or can be integrated into one or more processors, for example, integrated into one or more application specific integrated circuits. It can be understood that the processor in the embodiments of the present application can be a central processing unit (CPU), and can also be 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. The general purpose processor can be a microprocessor or any conventional processor.

[0145] Optionally, the communication apparatus 800 comprises one or more memories 802 storing instructions that are executable on the processor. The memory and the processor are coupled, and the coupling between the various devices, units or modules in the present application is indirect coupling or communication connection between devices, units or modules, which can be electrical, mechanical or other form, for information interaction between devices, units or modules.

[0146] Optionally, the memory can also store data. The processor and the memory can be separately arranged, or integrated together. The memory can be a non-volatile memory such as a hard disk drive (HDD) or a solid-state drive (SSD), etc., and can also be a volatile memory such as a random-access memory (RAM). In the embodiments of the present application, the processor can also be a flash memory, a read-only memory (ROM), a programmable ROM (PROM), an erasable PROM (EPROM), an electrically EPROM (EEPROM), a register, a hard disk, a mobile hard disk, a CD-ROM, or any other form of storage medium well known in the art.

[0147] Optionally, the communication apparatus 800 can comprise instructions (which can also be referred to as code or programs) that are executable on the processor.

[0148] Optionally, the communication apparatus 800 can further comprise a transceiver 803 and an antenna 804. The transceiver can be referred to as a transceiving unit, a transceiving module, a transceiver, a transceiving circuit, a transceiver, an input / output interface, etc., and is used to realize the transceiving function of the communication apparatus 800 through the antenna.

[0149] When the communication apparatus 800 is used to implement the method performed by the network device in the method embodiments described above, the processor 801 needs to generate instructions indicating that the user equipment uploads training data or instructions configuring the user equipment to periodically upload training data, such as RRC messages. The processor 801 needs to generate signaling indicating that the user equipment reports inference data, such as DCI, MAC CE, etc. For processing the training and inference data reported by the UE, the processor 801 needs to train a neural network model and use the neural network model to obtain a target service quality parameter. The transceiver 803 and the antenna 804 need to send instructions indicating that the user equipment uploads training and inference data.

[0150] When the communication device 800 is used to implement the method performed by the terminal device in the above method embodiments, the processor 801 processes instructions indicating that the user equipment uploads training data or instructions configuring the user equipment to periodically upload training data, such as RRC signaling. The processor 801 is configured to process signaling indicating that the user equipment reports inference data, such as DCI, MAC CE, etc. The processor 801 is configured to package and transmit the quality of service parameters of the application layer. The transceiver 803 and the antenna 804 are configured to receive instructions for training and inference data.

[0151] Please refer to FIG. 9, which is a structural schematic diagram of a communication device provided by an embodiment of the present application. As shown in FIG. 9, the communication device 900 includes a processor 901, a memory 902, a communication interface 903, and a bus 904. The processor 901, the memory 902, and the communication interface 903 are coupled through the bus (not labeled in the figure). The memory 902 stores instructions, and when the instructions stored in the memory 902 are executed, the communication device 900 performs the method performed by the terminal device or the network device in the above method embodiments.

[0152] The communication device 900 can be one or more integrated circuits configured to implement the above method, such as one or more application specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms. For another example, when the units in the device can be implemented in the form of a processing element scheduler, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call a program. For another example, these units can be integrated together in the form of a system-on-a-chip (SOC).

[0153] The processor 901 can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. The general-purpose processor can be a microprocessor or any conventional processor.

[0154] The memory 902 can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically EPROM (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM) used as an external cache. By way of example but not limitation, many forms of RAM can be used, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM) and direct rambus RAM (DR RAM).

[0155] The memory 902 stores executable program code, and the processor 901 executes the executable program code to respectively implement the functions of the foregoing units or modules, thereby implementing the domain name resolution method described above. That is, the memory 902 has instructions for executing the domain name resolution method described above.

[0156] The communication interface 903 uses a transceiver module such as but not limited to a network interface card, a transceiver, to realize the communication between the communication device 900 and other devices or communication networks.

[0157] The bus 904 can include, in addition to a data bus, a power bus, a control bus, and a state signal bus, etc. The bus can be a peripheral component interconnect express (PCIe) bus, or an extended industry standard architecture (EISA) bus, a unified bus (Ubus or UB), a compute express link (CXL), a cache coherent interconnect for accelerators (CCIX), etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0158] In another embodiment of the present application, a computer readable storage medium is also provided, and the computer readable storage medium stores computer execution instructions. When the processor of the device executes the computer execution instructions, the device executes the method performed by the data transmission system in the above method embodiments.

[0159] In another embodiment of the present application, a computer program product is also provided, and the computer program product includes computer execution instructions stored in a computer readable storage medium. When the processor of the device executes the computer execution instructions, the device executes the method performed by the data transmission system in the above method embodiments.

[0160] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0161] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented by other ways. For example, the above-described device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units or components shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0162] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0163] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present alone, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0164] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical scheme of the present application essentially or the part that contributes to the prior art or the whole or part of the technical scheme can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, read-only memory), a random access memory (RAM, random access memory), a magnetic disk or an optical disk, and various program code storage media.

Claims

1. A data transmission method, characterized by, The method comprises the following steps: The network device determines a target service rate, which is a predicted data transmission bit rate that the network device can provide for a user equipment at a first time point; The network device determines a target service quality parameter based on the target service rate and a neural network model, the target service quality parameter being used to indicate a service quality at the first time point, an input of the neural network model comprising a service quality parameter, and an output of the neural network model comprising a service rate; The network device adjusts a transmission behavior at a second time point based on the target service quality parameter, so that a server or the user equipment perceives a change in the service quality and adjusts a data transmission bit rate, the second time point being earlier than the first time point.

2. The method of claim 1, wherein, Before the network device determines the target service quality parameter based on the target service rate and the neural network model, the method further comprises the following steps: The network device receives a first message sent by the user equipment, the first message carrying training data used to train the neural network model, the training data comprising a service quality parameter and a corresponding service rate collected by the user equipment.

3. The method of claim 2, wherein, The training data comprises one or more of the following: a total number of data packets, a total number of packet bytes, a packet loss rate, a time delay, a jitter time, a time stamp, and a bit rate.

4. The method according to any one of claims 1 to 3, characterized in that, Before the network device receives the training data sent by the user equipment, the method further comprises the following steps: The network device sends a second message to the user equipment, the second message being used to configure a data type and an upload period of the training data uploaded by the user equipment.

5. The method according to any one of claims 1 to 4, characterized in that, The network device determines the target service rate comprises the following steps: The network device predicts the target service rate based on scheduling awareness information, the scheduling awareness information comprising one or more of the following: user equipment location information, user equipment historical data, future possible user arrival, and network load information.

6. The method according to any one of claims 1 to 5, characterized in that, The neural network model is a non-memory network model, and the network device determines the target service quality parameter based on the target service rate and the neural network model comprises the following steps: The network device determines one or more candidate target service quality parameters based on a current service quality parameter; The network device inputs the one or more candidate target service quality parameters into the non-memory network model; When a difference between an output result of the non-memory network model and the target service rate is less than a threshold value, the network device determines a corresponding candidate target service parameter as the target service quality parameter.

7. The method according to any one of claims 1 to 5, characterized in that, The neural network model is a memory network model, and the network device determines the target service quality parameter based on the target service rate and the neural network model comprises the following steps: The network device determines one or more candidate target service quality parameters based on a current service quality parameter; The network device inputs the one or more candidate target service quality parameters and historical service quality parameters of a historical period into the memory network model; When a difference between an output result of the memory network model and the target service rate is less than a threshold, a corresponding candidate target service parameter is determined as the target service quality parameter.

8. The method according to claim 6 or 7, characterized in that, The method further includes: The network device sends a third message to the user device, the third message being used to instruct the user device to send one or more pieces of inference data to the network device, the inference data including the current service quality parameter and the historical service quality parameter of the historical period.

9. The method of claim 8, wherein, The method further includes: The network device receives a fourth message sent by the user device, the fourth message being used to carry the inference data.

10. A data transmission apparatus, characterized by comprising: Comprise: A processing unit is configured to determine a target service rate, the target service rate being a predicted data transmission bit rate that the network device can provide for a user device at a first time point; The processing unit is further configured to determine a target service quality parameter based on the target service rate and a neural network model, the target service quality parameter being used to indicate a service quality corresponding to the first time point, an input of the neural network model including a service quality parameter, and an output of the neural network model including a service rate; The processing unit is further configured to adjust a transmission behavior at a second time point based on the target service quality parameter, so that a server or the user device perceives a change in service quality and adjusts a data transmission bit rate, the second time point being earlier than the first time point.

11. The apparatus of claim 10, wherein, The apparatus further comprises: A transceiver is configured to receive a first message sent by the user device, the first message carrying the training data, the training data being used to train the neural network model, and the training data including a service quality parameter and a corresponding service rate collected by the user device.

12. The apparatus of claim 11, wherein, The training data includes one or more of the following: a total number of data packets, a total number of packet bytes, a packet loss rate, a time delay, a jitter time, a timestamp, and a bit rate.

13. The apparatus of any one of claims 10-12, wherein, The transceiver is further configured to: Send a second message to the user device, the second message being used to configure a data type and an upload period of the training data uploaded by the user device.

14. The apparatus of any one of claims 10-13, wherein, The processing unit is specifically configured to: Predict a target service rate based on scheduling awareness information, the scheduling awareness information including one or more of the following: user device location information, user device historical data, future possible user arrival, and network load information.

15. The apparatus of any one of claims 10 to 14, wherein, The neural network model is a memory network model, and the processing unit is specifically configured to: Determine one or more candidate target service quality parameters based on the current service quality parameter; Input the one or more candidate target service quality parameters into the memory network model; When a difference between an output result of the memory network model and the target service rate is less than a threshold, a corresponding candidate target service parameter is determined as the target service quality parameter.

16. The apparatus of any one of claims 10-14, wherein, The neural network model is a memory network model, and the processing unit is specifically configured to: Determine one or more candidate target service quality parameters based on the current service quality parameter; Input the one or more candidate target service quality parameters and a historical service quality parameter of a historical period into the memory network model; When a difference between an output result of the memory network model and the target service rate is less than a threshold value, a corresponding candidate target service parameter is determined as a target service quality parameter.

17. The apparatus of claim 15 or 16, wherein, The transceiver is further configured to: send, to the user equipment, a third message, the third message being used to instruct the user equipment to send one or more pieces of inference data to the network device, the inference data comprising the current service quality parameter and the historical service quality parameter of the historical period.

18. The apparatus of claim 17, wherein, The transceiver is further configured to: receive a fourth message sent by the user equipment, the fourth message being used to carry the inference data.

19. A network device, comprising: A processor coupled with a memory, the memory being used to store instructions, when the instructions are executed by the processor, to cause the network device to perform the method in any one of claims 1 to 9.

20. A terminal device, comprising: A processor coupled with a memory, the memory being used to store instructions, when the instructions are executed by the processor, to cause the terminal device to perform the method performed by the user equipment in any one of claims 1 to 9.

21. A computer-readable storage medium having stored thereon instructions, The instructions, when executed, cause a computer to perform the method in any one of claims 1 to 9.

22. A computer program product, comprising instructions therein, characterised in that, The instructions, when executed, cause a computer to implement the method in any one of claims 1 to 9.

Images