Network traffic prediction method and device, and storage medium
By collecting and analyzing user traffic data, determining the traffic usage behavior pattern, and using the LSTM model to predict, the problem of low accuracy of network traffic prediction in the prior art is solved, and higher prediction accuracy is achieved.
Patent Information
- Application Number
- CN202411653727.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-19
- Publication Date
- 2025-03-07
AI Technical Summary
The existing network traffic prediction methods adopt machine learning methods, requiring complex parameter selection processes, resulting in low accuracy of network traffic prediction.
By collecting user traffic data within the preset time period, extracting traffic data characteristics, determining the user's traffic usage behavior pattern based on these characteristics, and using the LSTM model for model training, obtaining a traffic prediction model. This model can accurately predict the usage of user traffic data in the next cycle.
Without the need for a complex parameter selection process, it can effectively avoid the reduction in prediction accuracy caused by parameter selection errors, thereby improving the accuracy of network traffic prediction.
Smart Images
Figure CN119583369A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of traffic monitoring, and in particular to a network traffic prediction method, device and storage medium. Background Art
[0002] In the context of intelligent and connected vehicles, TBOX, as a core component of the vehicle communication system, plays an important role in the communication between vehicles and the outside world. With the popularization of 5G technology and the upgrade and iteration of smart cockpit technology, TBOX not only needs to handle traditional vehicle monitoring and remote diagnosis functions, but also needs to support a variety of vehicle application services, such as real-time navigation, online entertainment, and Internet of Vehicles services. The increase in these services has led to an explosive growth in vehicle network traffic, which has put forward higher requirements for TBOX's data processing capabilities and network resource management.
[0003] Existing network traffic prediction methods usually use machine learning methods to predict network traffic, which requires a complex parameter selection process, resulting in low accuracy of network traffic prediction. Summary of the invention
[0004] The present invention provides a network traffic prediction method, device and storage medium to solve the technical problem that the existing network traffic prediction method uses a machine learning method to predict network traffic, which requires a complex parameter selection process and leads to low accuracy of network traffic prediction.
[0005] The present invention provides a network traffic prediction method, comprising:
[0006] Collecting user traffic data within a preset time period; wherein the user traffic data includes traffic start timestamp, traffic end timestamp, application type, application IP address, traffic transmission size and traffic peak rate;
[0007] Extracting traffic data features of the user traffic data, and determining the user's traffic usage behavior pattern based on the traffic data features;
[0008] Performing model training based on the user traffic data and the traffic usage behavior pattern to obtain a traffic prediction model;
[0009] The user traffic data to be predicted in a preset period is input into the traffic prediction model to obtain the user traffic data usage of the next period.
[0010] Furthermore, the traffic data characteristics include at least one of functional characteristics, application characteristics, date characteristics, time characteristics and traffic size characteristics.
[0011] Furthermore, the determining of the user's traffic usage behavior pattern based on the traffic data characteristics includes:
[0012] The traffic data features are normalized, and a decision tree is used to determine the user's traffic usage behavior pattern based on the normalized traffic data features.
[0013] Furthermore, the model training is performed based on the user traffic data and the traffic usage behavior pattern to obtain a traffic prediction model, including:
[0014] A training set is constructed based on the user traffic data, the traffic usage behavior pattern is used as a label, and a preset LSTM model is used for training to obtain a traffic prediction model.
[0015] Furthermore, the LSTM model includes an input layer, a first hidden layer, a regularization layer, a second hidden layer and an output layer connected in sequence, wherein the first hidden layer is provided with 64 neurons and the second hidden layer is provided with 128 neurons.
[0016] Furthermore, after inputting the user traffic data to be predicted in a preset period into the traffic prediction model to obtain the user traffic data usage of the next period, the method further includes:
[0017] The user traffic usage is sent to a communication controller, so that the communication controller allocates network bandwidth to the IP address of the application based on the user traffic usage.
[0018] The present invention also provides a network traffic prediction device, comprising:
[0019] A flow data collection module is used to collect user flow data within a preset time period; wherein the user flow data includes flow start timestamp, flow end timestamp, application type, application IP address, flow transmission size and flow peak rate;
[0020] A behavior model determination module is used to extract traffic data features of the user traffic data and determine the user's traffic usage behavior pattern based on the traffic data features;
[0021] A traffic prediction model building module, used to perform model training based on the user traffic data and the traffic usage behavior pattern to obtain a traffic prediction model;
[0022] The traffic usage prediction module is used to input the user traffic data to be predicted in a preset period into the traffic prediction model to obtain the user traffic data usage of the next period.
[0023] Furthermore, the usage behavior model determination module is also used to:
[0024] The traffic data features are normalized, and a decision tree is used to determine the user's traffic usage behavior pattern based on the normalized traffic data features.
[0025] The present invention also provides a terminal device, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the network traffic prediction method as described above when executing the computer program.
[0026] The present invention also provides a computer-readable storage medium, which includes a stored computer program; wherein, when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the network traffic prediction method as described above.
[0027] The embodiment of the present invention collects user traffic data in a preset time period, determines the corresponding user traffic usage behavior pattern, further performs model training to obtain a traffic prediction model, and inputs the user traffic data to be predicted into the model. It can accurately obtain the user traffic data usage of the next cycle without the need for a complicated parameter selection process, and can effectively avoid the reduction in prediction accuracy caused by incorrect parameter selection, thereby effectively improving the accuracy of network traffic prediction.
[0028] Furthermore, the first hidden layer in the LSTM model of the embodiment of the present invention is equipped with 64 neurons, which pass the processed data to the second hidden layer after being screened by the Dropout layer. The number of neurons in the second hidden layer is increased to 128, so that the features of the traffic time series can be further accurately extracted by matching the number of neurons at different levels to enhance the accuracy of feature extraction, thereby further improving the accuracy of network traffic prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] Figure 1 It is a flowchart of a network traffic prediction method provided by an embodiment of the present invention;
[0030] Figure 2 It is a schematic diagram of traffic division provided by an embodiment of the present invention;
[0031] Figure 3 It is a schematic diagram of the LSTM model structure provided by an embodiment of the present invention;
[0032] Figure 4 is a schematic diagram of the structure of a gating mechanism provided by an embodiment of the present invention;
[0033] Figure 5 is a schematic diagram of executing a network optimization strategy provided by an embodiment of the present invention;
[0034] Figure 6It is a schematic diagram of the structure of a network traffic prediction device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0035] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0036] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the feature. In the description of this application, unless otherwise specified, "plurality" means two or more.
[0037] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.
[0038] See also Figure 1 The present invention provides a network traffic prediction method, comprising:
[0039] S1. Collect user traffic data within a preset time period; wherein the user traffic data includes traffic start timestamp, traffic end timestamp, application type, application IP address, traffic transmission size and traffic peak rate;
[0040] In an embodiment of the present invention, the network traffic prediction method can be executed in the vehicle controller local area network architecture, wherein the vehicle controller local area network architecture includes a communication controller (which can be TBOX), a vehicle computer, a gateway and multiple electronic control units (ECUs), and the gateway is connected to the communication controller, the vehicle computer and other electronic control units through the CAN bus as a central node for data conversion and forwarding to ensure smooth communication and interaction between various nodes in the vehicle. TBOX provides network and serial communication services to the vehicle computer through a USB interface, collects data on the CAN bus in real time, and uploads the data to the cloud big data platform after reasonable packaging, thereby realizing remote monitoring and data analysis of the vehicle's operating status.
[0041] In an embodiment of the present invention, the preset time period may be within one month, and the vehicle computer may collect user flow data once every 1 minute within one month. The format of the user flow data is shown in Table 1 below:
[0042] Table 1: User traffic data format
[0043]
[0044] S2. extracting traffic data features of the user's traffic data, and determining the user's traffic usage behavior pattern based on the traffic data features;
[0045] In the embodiment of the present invention, the traffic data feature includes at least one of a functional feature, an application feature, a date feature, a time feature, and a traffic size feature. The traffic usage behavior pattern is the traffic usage behavior of a user at a certain node or a certain scene. For example, on a highway, the user mainly uses navigation traffic, and in a traffic jam section, the user mainly uses music traffic. After determining the user's traffic usage behavior pattern, a traffic usage report can be generated in combination with the user's traffic usage size.
[0046] In an embodiment of the present invention, a decision tree algorithm can be used to extract five important types of traffic data features from user traffic data, including functional features, application features, date features, time features, and traffic size features, wherein the functional features are divided into five categories in total, using numbers [1-5] as markers; application features are independently distinguished according to each type of function, using numbers [1-20] as markers; date features only distinguish between weekdays (1) and weekends (2); time features use hours and minutes as markers [0.00-23.99]; and traffic size features use the actual traffic transmission size data (in MBYTE) of each application within a 1-minute time unit. The distinguishing features are shown in the figure below. For example, at 17:30 on Friday, iQiyi transmitted 200MB of traffic, and the data obtained is [3, 1, 1, 17.5, 200].
[0047] See also Figure 2 , which is a flow division diagram provided in an embodiment of the present invention, the flow can be gradually divided into functional characteristics, practical application characteristics, date characteristics, time characteristics and flow size characteristics according to the order of priority. In an embodiment of the present invention, in order to ensure the stability, convergence speed and prediction accuracy of the training process, the minimum and maximum normalization (Min-MaxSca ling) is used to linearly change the five-element vector data (i.e., five types of flow data feature data) extracted by the decision tree, and the five-element vector data is mapped to [0,1] to keep the data distribution consistent with the original. Among them, the linear change formula is as follows:
[0048]
[0049] In an embodiment of the present invention, a decision tree algorithm can be combined to conduct an in-depth analysis of user traffic data, so that the user's behavior pattern during the use of in-vehicle applications can be determined in a tree model, including: summarizing the user's behavior habits based on the branches of the decision tree, and determining the user's application usage habits in the in-vehicle environment by refining the core path of the decision tree, and combining the traffic size at each time node to accurately determine the user's traffic usage behavior pattern. For example, during peak hours, users may be more inclined to use navigation and music functions.
[0050] S3, training a model based on user traffic data and traffic usage behavior patterns to obtain a traffic prediction model;
[0051] In an embodiment of the present invention, one month's worth of data may be collected for model training, wherein 70% of the data is used as input for training the model, 20% is used for verification, and 10% is used for testing.
[0052] S4. Input the user traffic data to be predicted in a preset period into the traffic prediction model to obtain the user traffic data usage of the next period.
[0053] In an embodiment of the present invention, the preset period may be one week, and the traffic prediction model may be used to predict the user traffic data usage of the next week, wherein the user traffic data usage includes traffic usage size, traffic usage type, traffic usage time, and the function corresponding to the traffic usage.
[0054] The embodiment of the present invention collects user traffic data in a preset time period, determines the corresponding user traffic usage behavior pattern, further performs model training to obtain a traffic prediction model, and inputs the user traffic data to be predicted into the model. It can accurately obtain the user traffic data usage of the next cycle without the need for a complicated parameter selection process, and can effectively avoid the reduction in prediction accuracy caused by incorrect parameter selection, thereby effectively improving the accuracy of network traffic prediction.
[0055] In one embodiment, determining the user's traffic usage behavior pattern based on traffic data characteristics includes:
[0056] The traffic data features are normalized, and a decision tree is used to determine the user's traffic usage behavior pattern based on the normalized traffic data features.
[0057] The embodiment of the present invention normalizes the extracted traffic data feature data to ensure stability, convergence speed and prediction accuracy during the training process.
[0058] In one embodiment, model training is performed based on user traffic data and traffic usage behavior patterns to obtain a traffic prediction model, including:
[0059] A training set is constructed based on user traffic data, and traffic usage behavior patterns are used as labels. The preset LSTM model is used for training to obtain a traffic prediction model.
[0060] The embodiment of the present invention adopts the LSTM model to conduct in-depth learning of user traffic data by processing and predicting long-term dependencies in time series data, thereby effectively improving the accuracy of traffic prediction.
[0061] In one embodiment, the LSTM model includes an input layer, a first hidden layer, a regularization layer, a second hidden layer and an output layer connected in sequence, wherein the first hidden layer is provided with 64 neurons and the second hidden layer is provided with 128 neurons.
[0062] See also Figure 3 In the embodiment of the present invention, the hidden layer of the LSTM model is only two layers, namely the first hidden layer and the second hidden layer.
[0063] In an embodiment of the present invention, the input layer of the LSTM model has the ability to process two-dimensional arrays x of arbitrary sizes, so that a unified data set can be cyclically trained and predicted. The LSTM network structure adopted by the LSTM model of the embodiment of the present invention includes two layers, which mainly focus on the accurate prediction of short-term traffic time series. In this structure, the Dropout layer plays a key role, which can randomly discard 50% of the neurons during the training process (the parameter is set to 0.5), thereby effectively preventing the problem of overfitting of the neural network and improving the generalization ability of the prediction.
[0064] In the architecture of the LSTM model, the first hidden layer is equipped with 64 neurons, which pass the processed data to the second hidden layer after being filtered by the Dropout layer. The number of neurons in the second hidden layer is increased to 128, so that the characteristics of the traffic time series can be further accurately extracted by matching the number of neurons at different levels to enhance the accuracy of feature extraction, thereby further improving the accuracy of prediction.
[0065] It can be understood that the last part of the LSTM model is a fully connected layer, which is used to output the final prediction result to complete the closed loop of the entire prediction process.
[0066] See also Figure 4 In the embodiment of the present invention, the LSTM model introduces a gating mechanism to manage the flow of information and realize the influence of long-term memory on the output. t-1 , c t-1 Represent the hidden state and cell state at time t-1, respectively, x tis the input at the current time t. Then the calculation formula of the attenuation coefficient of the forget gate is:
[0067] f t =σ(w f [h t-1 ,x t ]+b f )
[0068] The input gate calculation formula is:
[0069] i t =σ(w i [h t-1 ,x t ]+b i )
[0070]
[0071] The output gate calculation formula is:
[0072] o t =σ(w o +[h t-1 ,x t ]+b o )
[0073] Final output calculation formula:
[0074]
[0075] where f, i, o and represents the three gate vectors and candidate vectors involved in the gating mechanism; tanh represents the hyperbolic tangent function; w f , w i , w o , w c represents the corresponding weight vector; Represents a vector dot product operation; b f , b i , b o , b c Represents the corresponding offset; σ represents the Sigmoid function.
[0076] In one embodiment, after inputting the user traffic data to be predicted in a preset period into the traffic prediction model to obtain the user traffic data usage of the next period, the method further includes:
[0077] The user traffic usage is sent to the communication controller, so that the communication controller allocates network bandwidth to the IP address of the application based on the user traffic usage.
[0078] In an embodiment of the present invention, the traffic prediction model can be deployed to a cloud big data platform, which has periodic automatic processing capabilities and can perform independent calculations on the user traffic data of each user to predict the user traffic usage at various time points in the next week. On this basis, the user traffic usage can be sent to TBOX, which can generate a network optimization strategy based on the predicted user traffic usage, specific time points, and functional classification, that is, allocate network bandwidth to the IP address of the application, thereby realizing intelligent management and optimization of network resources.
[0079] For further information, see Figure 5 The embodiment of the present invention can synchronize the network optimization strategy to the vehicle computer, so that the user can browse and confirm the network optimization strategy through the interactive page of the vehicle computer. For example, after receiving the network optimization strategy, the vehicle computer generates user confirmation information. After the user agrees to the network optimization strategy, the network optimization strategy is executed. If the user disagrees with the network optimization strategy, the current original strategy is maintained.
[0080] The implementation of the embodiments of the present invention has the following beneficial effects:
[0081] The embodiment of the present invention collects user traffic data in a preset time period, determines the corresponding user traffic usage behavior pattern, further performs model training to obtain a traffic prediction model, and inputs the user traffic data to be predicted into the model. It can accurately obtain the user traffic data usage of the next cycle without the need for a complicated parameter selection process, and can effectively avoid the reduction in prediction accuracy caused by incorrect parameter selection, thereby effectively improving the accuracy of network traffic prediction.
[0082] Furthermore, the first hidden layer in the LSTM model of the embodiment of the present invention is equipped with 64 neurons, which pass the processed data to the second hidden layer after being screened by the Dropout layer. The number of neurons in the second hidden layer is increased to 128, so that the features of the traffic time series can be further accurately extracted by matching the number of neurons at different levels to enhance the accuracy of feature extraction, thereby further improving the accuracy of network traffic prediction.
[0083] See also Figure 6 Based on the same inventive concept as the above embodiment, the present invention also provides a network traffic prediction device, comprising:
[0084] The flow data collection module 10 is used to collect user flow data within a preset time period; wherein the user flow data includes flow start timestamp, flow end timestamp, application type, application IP address, flow transmission size and flow peak rate;
[0085] A behavior model determination module 20 is used to extract traffic data features of user traffic data and determine the user's traffic usage behavior pattern based on the traffic data features;
[0086] The traffic prediction model building module 30 is used to perform model training based on user traffic data and traffic usage behavior patterns to obtain a traffic prediction model;
[0087] The traffic usage prediction module 40 is used to input the user traffic data to be predicted in a preset period into the traffic prediction model to obtain the user traffic data usage of the next period.
[0088] In one embodiment, the traffic data feature includes at least one of a functional feature, an application feature, a date feature, a time feature, and a traffic size feature.
[0089] In one embodiment, the behavior model determination module 20 is used to:
[0090] The traffic data features are normalized, and a decision tree is used to determine the user's traffic usage behavior pattern based on the normalized traffic data features.
[0091] In one embodiment, model training is performed based on user traffic data and traffic usage behavior patterns to obtain a traffic prediction model, including:
[0092] A training set is constructed based on user traffic data, and traffic usage behavior patterns are used as labels. The preset LSTM model is used for training to obtain a traffic prediction model.
[0093] In one embodiment, the LSTM model includes an input layer, a first hidden layer, a regularization layer, a second hidden layer and an output layer connected in sequence, wherein the first hidden layer is provided with 64 neurons and the second hidden layer is provided with 128 neurons.
[0094] Accordingly, an embodiment of the present invention further provides a terminal device, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the network traffic prediction method of any one of the above embodiments is implemented.
[0095] The terminal device of this embodiment includes: a processor, a memory, and a computer program and computer instructions stored in the memory and executable on the processor. When the processor executes the computer program, each step in the above embodiment 1 is implemented, for example Figure 1 Alternatively, when the processor executes the computer program, the functions of each module / unit in the above-mentioned device embodiment are implemented, for example, the behavior model determination module 20 is used.
[0096] Exemplarily, the computer program may be divided into one or more modules / units, one or more modules / units are stored in a memory and executed by a processor to complete the present invention. One or more modules / units may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in a terminal device. For example, a behavior model determination module 20 is used to extract traffic data features of user traffic data, and determine the user's traffic usage behavior pattern based on the traffic data features.
[0097] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the schematic diagram is merely an example of a terminal device and does not constitute a limitation on the terminal device. The terminal device may include more or fewer components than shown in the diagram, or may combine certain components, or different components. For example, the terminal device may also include an input / output device, a network access device, a bus, etc.
[0098] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.
[0099] The memory can be used to store computer programs and / or modules. The processor realizes various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile terminal, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Med ia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (F l ash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.
[0100] Among them, if the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electric carrier signals and telecommunication signals.
[0101] Correspondingly, an embodiment of the present invention also provides a computer-readable storage medium, the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the network traffic prediction method of any one of the above embodiments.
[0102] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.
Claims
1. A network traffic prediction method, characterized in that: include: Collecting user traffic data within a preset time period; wherein the user traffic data includes traffic start timestamp, traffic end timestamp, application type, application IP address, traffic transmission size and traffic peak rate; Extracting traffic data features of the user traffic data, and determining the user's traffic usage behavior pattern based on the traffic data features; Performing model training based on the user traffic data and the traffic usage behavior pattern to obtain a traffic prediction model; The user traffic data to be predicted in a preset period is input into the traffic prediction model to obtain the user traffic data usage of the next period.
2. The network traffic prediction method according to claim 1, characterized in that: The traffic data feature includes at least one of a functional feature, an application feature, a date feature, a time feature, and a traffic size feature.
3. The network traffic prediction method according to claim 1, characterized in that: The determining the user's traffic usage behavior pattern based on the traffic data feature includes: The traffic data features are normalized, and a decision tree is used to determine the user's traffic usage behavior pattern based on the normalized traffic data features.
4. The network traffic prediction method according to claim 1, characterized in that: The performing model training based on the user traffic data and the traffic usage behavior pattern to obtain a traffic prediction model includes: A training set is constructed based on the user traffic data, the traffic usage behavior pattern is used as a label, and a preset LSTM model is used for training to obtain a traffic prediction model.
5. The network traffic prediction method according to claim 4, characterized in that: The LSTM model includes an input layer, a first hidden layer, a regularization layer, a second hidden layer and an output layer connected in sequence, wherein the first hidden layer is provided with 64 neurons, and the second hidden layer is provided with 128 neurons.
6. The network traffic prediction method according to claim 1, characterized in that: After inputting the user traffic data to be predicted in a preset period into the traffic prediction model to obtain the user traffic data usage of the next period, the method further includes: The user traffic usage is sent to a communication controller, so that the communication controller allocates network bandwidth to the IP address of the application based on the user traffic usage.
7. A network traffic prediction device, characterized in that: include: A flow data collection module is used to collect user flow data within a preset time period; wherein the user flow data includes flow start timestamp, flow end timestamp, application type, application IP address, flow transmission size and flow peak rate; A behavior model determination module is used to extract traffic data features of the user traffic data and determine the user's traffic usage behavior pattern based on the traffic data features; A traffic prediction model building module, used to perform model training based on the user traffic data and the traffic usage behavior pattern to obtain a traffic prediction model; The traffic usage prediction module is used to input the user traffic data to be predicted in a preset period into the traffic prediction model to obtain the user traffic data usage of the next period.
8. The network traffic prediction device according to claim 7, characterized in that: The usage behavior model determination module is also used for: The traffic data features are normalized, and a decision tree is used to determine the user's traffic usage behavior pattern based on the normalized traffic data features.
9. A terminal device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the network traffic prediction method according to any one of claims 1 to 6 is implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program; wherein, when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the network traffic prediction method according to any one of claims 1-6.