Remaining early warning method and device for shared traffic packet, and product

Abstract

The invention provides a shared traffic packet margin early warning method, device and product, and the method comprises the steps: taking first historical data of a plurality of first IPs using shared traffic resources within a first preset time period as input, inputting the first historical data into a traffic analysis and prediction model obtained in advance, obtaining a first predicted flow consumption value of the first IP within a second preset duration after the current time; according to the first predicted consumed traffic value, performing traffic simulation deduction in the current traffic of at least one shared traffic packet in the shared traffic resources, and obtaining the residual traffic of the current traffic after the traffic simulation deduction; and when the residual flow is smaller than the first threshold value, carrying out flow alarm prompt. The shared traffic packets of different resource pools are subjected to unified simulation deduction, and an alarm is given when the current traffic is not deducted enough, so that a user can be prompted to supplement a new shared traffic packet in advance, and the problems of traffic packet margin alarm failure, alarm skipping and direct traffic cost generation are avoided.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method, device and product for warning the remaining amount of a shared traffic package. Background Art

[0002] Currently, a large number of public clouds do not have a surplus warning solution for shared traffic packages. So far, only some shared traffic packages have surplus warning functions. One type of shared traffic package does not distinguish between resource pools. It is divided into Asia Pacific and Europe and America by geographical region, and divided into full-time and off-peak by time. The surplus warning function of shared traffic packages only needs to be set by combining partitions and time divisions. By calculating the total amount of shared traffic packages held and the surplus ratio, an alarm is triggered when it is lower than the threshold. For other public cloud shared traffic packages, the surplus of shared traffic packages that can be deducted from different resource pools in the same period is not exactly the same. There is a situation where the surplus alarm of the traffic package fails, the alarm is skipped, and traffic charges are directly incurred. Summary of the invention

[0003] The purpose of the present invention is to provide a shared traffic package balance warning method, device and product, which is used to solve the problem in the prior art that the shared traffic package balances that can be deducted from different resource pools in the same period in public cloud shared traffic packages are not completely the same, and there is a problem that the traffic package balance alarm is invalid, the alarm is skipped, and traffic charges are directly generated.

[0004] To achieve the above object, an embodiment of the present invention provides a method for warning the remaining amount of a shared traffic package, which includes:

[0005] Taking first historical data of the use of shared traffic resources by multiple first IPs within a first preset time as input, inputting it into a pre-acquired traffic analysis prediction model, and obtaining a first predicted traffic consumption value of the first IP within a second preset time after the current time;

[0006] According to the first predicted consumed traffic value, a traffic simulation deduction is performed on the current traffic of at least one shared traffic package in the shared traffic resource to obtain the remaining traffic of the current traffic after the traffic simulation deduction;

[0007] When the remaining flow is less than the first threshold, a flow alarm prompt is issued.

[0008] Optionally, in the remaining warning method, when the shared traffic package is a first type shared traffic package, the method further includes:

[0009] Obtaining a remaining ratio of the shared traffic package; wherein the remaining ratio is a ratio of the current traffic of the shared traffic package to the total traffic value of the shared traffic package;

[0010] When the surplus ratio is less than a second threshold, the first historical data is input into the flow analysis prediction model.

[0011] Optionally, in the surplus warning method, when the at least one shared traffic packet is a designated IP shared traffic packet of the second IP, the method further comprises:

[0012] The second IP is determined as the first IP.

[0013] Optionally, the remaining warning method further includes:

[0014] Acquire second historical data of a first time period to be trained and third historical data of a second time period of a plurality of the first IPs; wherein the second time period is after the first time period;

[0015] Acquire a network topology map of the first IP according to the second historical data;

[0016] Inputting the network topology graph into a spatiotemporal graph convolutional network model integrating an attention mechanism to obtain a prediction result; wherein the prediction result is prediction data for the third historical data;

[0017] The spatiotemporal graph convolutional network model integrating the attention mechanism is adjusted according to the prediction result and the third historical data to obtain the traffic analysis prediction model.

[0018] Optionally, in the surplus warning method, the step of acquiring a network topology diagram of the first IP according to the second historical data includes:

[0019] Abstract each of the first IPs into a network node, and obtain a network node set;

[0020] Abstracting the data transmission between the network nodes as edges, and obtaining an edge set;

[0021] Acquire an adjacency matrix according to the network nodes and the edges;

[0022] The network topology graph is obtained according to the network node set, the edge set and the adjacency matrix.

[0023] Optionally, in the margin warning method, the spatiotemporal graph convolutional network model integrating the attention mechanism includes a first fully connected layer, a first asynchronous spatiotemporal convolutional module, a first channel attention module, a second asynchronous spatiotemporal convolutional module, a second channel attention module and a second fully connected layer connected in sequence; wherein the first asynchronous spatiotemporal convolutional module and the second asynchronous spatiotemporal convolutional module are both composed of a plurality of spatiotemporal convolutional layers including graph convolution modules and spatiotemporal convolutional layers including asynchronous time convolution modules, which are alternately stacked.

[0024] Optionally, in the remaining warning method, the step of providing a flow alarm prompt includes:

[0025] Simulate the flow deduction process of the at least one shared flow package within each third preset duration according to the first predicted flow consumption value, and obtain a flow consumption diagram of the at least one shared flow package; wherein the second preset duration includes a plurality of the third preset durations;

[0026] The flow consumption diagram is sent to the customer as the content of the flow alarm prompt.

[0027] Optionally, the remaining traffic warning method, wherein the step of performing a traffic simulation deduction in the current traffic of at least one shared traffic package in the shared traffic resource according to the first predicted traffic consumption value, comprises:

[0028] Obtain a list of currently valid shared traffic packages in the at least one shared traffic package; wherein the shared traffic packages in the list are sorted from early to late according to expiration dates;

[0029] Perform traffic simulation deduction on each of the multiple first IPs in the current traffic one by one in the order of the list, deduct the idle time traffic first, and then deduct the full-time traffic after the idle time traffic is deducted; wherein the current traffic includes the idle time traffic and the full-time traffic.

[0030] In order to achieve the above object, an embodiment of the present invention further provides a remaining warning device for a shared traffic package, which includes:

[0031] A first acquisition module is used to input first historical data of the use of shared traffic resources by multiple first IPs within a first preset time period into a pre-acquired traffic analysis prediction model to obtain a first predicted traffic consumption value of the first IP within a second preset time period after the current time;

[0032] A second acquisition module is used to perform a flow simulation deduction on the current flow of at least one shared flow package in the shared flow resource according to the first predicted consumed flow value, and obtain the remaining flow of the current flow after the flow simulation deduction;

[0033] The first processing module is used to issue a flow alarm prompt when the remaining flow is less than a first threshold.

[0034] In order to achieve the above-mentioned purpose, an embodiment of the present invention also provides an electronic device, including: a processor, a memory, and a program or instruction stored in the memory and executable on the processor; wherein, when the processor executes the program or instruction, the method for warning the remaining capacity of the shared traffic package as described above is implemented.

[0035] In order to achieve the above-mentioned purpose, an embodiment of the present invention also provides a readable storage medium, on which a program or instruction is stored, wherein when the program or instruction is executed by a processor, the steps in the shared traffic package balance warning method as described above are implemented.

[0036] In order to achieve the above object, an embodiment of the present invention further provides a computer program product, which includes computer instructions, and when the computer instructions are executed by a processor, the steps of the above-mentioned method for warning the remaining amount of shared traffic packages are implemented. The beneficial effects of the above-mentioned technical solution of the present invention are as follows:

[0037] The embodiment of the present invention inputs the historical data of the shared traffic package into the traffic analysis and prediction model to obtain the predicted consumption traffic value, and performs traffic simulation deduction in the current traffic of all shared traffic packages until the current traffic is insufficient for the simulated deduction, and issues a traffic alarm prompt. The shared traffic packages of different resource pools are uniformly simulated for deduction and an alarm is issued when the current traffic is insufficient for the deduction. This can prompt the user to replenish a new shared traffic package in advance, thereby avoiding the problem of the traffic package balance alarm being invalid, skipping the alarm, and directly incurring traffic charges. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of a method for early warning of a remaining amount of a shared traffic package according to an embodiment of the present invention;

[0039] Figure 2 A schematic diagram of a shared traffic package deduction in the remaining warning method of a shared traffic package according to an embodiment of the present invention;

[0040] Figure 3 A schematic diagram of deduction of a first type of shared traffic package in a remaining warning method for a shared traffic package according to an embodiment of the present invention;

[0041] Figure 4 A schematic diagram of deduction of a designated IP shared traffic package in the method for warning the remaining amount of a shared traffic package according to an embodiment of the present invention;

[0042] Figure 5 A schematic diagram of the structure of a traffic analysis model of a method for early warning of a remaining amount of a shared traffic package according to an embodiment of the present invention;

[0043] Figure 6 A schematic diagram of the structure of a channel attention module of the remaining warning method of a shared traffic package according to an embodiment of the present invention;

[0044] Figure 7 A schematic diagram of a first type of shared traffic package warning of a remaining shared traffic package warning method according to an embodiment of the present invention;

[0045] Figure 8 A schematic diagram of a shared traffic package warning for a designated IP in a method for warning the remaining amount of a shared traffic package according to an embodiment of the present invention;

[0046] Fig. 9 A schematic diagram of a shared traffic package remainder warning device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0048] It should be understood that the references to "one embodiment" or "an embodiment" throughout the specification mean that the specific features, structures, or characteristics associated with the embodiment are included in at least one embodiment of the present invention. Therefore, the references to "in one embodiment" or "in an embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0049] In various embodiments of the present invention, it should be understood that the size of the serial numbers of the following processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0050] Additionally, the terms "system" and "network" are often used interchangeably herein.

[0051] In the embodiments provided in the present application, it should be understood that "B corresponding to Aβ" means that B is associated with Aβ, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B based only on A, and B can also be determined based on A and / or other information.

[0052] For ease of understanding, some contents involved in the embodiments of the present invention are described below:

[0053] like Figure 1 As shown, a method for warning the remaining amount of a shared traffic package according to an embodiment of the present invention includes:

[0054] S10, taking first historical data of the use of shared traffic resources by multiple first IPs within a first preset time as input, and inputting it into a pre-acquired traffic analysis prediction model, and obtaining a first predicted traffic consumption value of the first IP within a second preset time after the current time;

[0055] It should be noted that after obtaining the traffic analysis and prediction model in advance, the first preset time period, for example, the first historical data of seven days, is used to make a data set according to the traffic data generated by each of the first IPs every hour. The first historical data is input into the traffic analysis and prediction model in batches, and the prediction is carried out in stages. The prediction result is the first predicted consumption traffic value of the first IP within the second preset time period after the current time, usually the traffic usage of each of the first IPs in the next natural day.

[0056] S20, performing a flow simulation deduction on the current flow of at least one shared flow package in the shared flow resource according to the first predicted consumed flow value, and obtaining the remaining flow of the current flow after the flow simulation deduction;

[0057] It should be noted that if Figure 2 As shown, when deducting traffic from shared traffic packages, they are arranged in reverse order of expiration time. After obtaining the first predicted traffic consumption value (usually the traffic usage of each of the first IP in the next natural day), the traffic deduction process is simulated every hour in the current traffic of at least one shared traffic package in the shared traffic resources (that is, the total traffic of the shared traffic packages currently arranged in reverse order of expiration time) to obtain the remaining traffic.

[0058] S30, when the remaining flow is less than the first threshold, issuing a flow alarm prompt;

[0059] It should be noted that, when the remaining traffic is less than the first threshold (usually 0), that is, the current traffic is not enough to offset the deduction, the page will be marked in a striking color to remind the customer and issue a traffic alarm.

[0060] In this embodiment, the historical data of the shared traffic package is input into the traffic analysis prediction model to obtain the predicted traffic consumption value, and the traffic simulation deduction is performed in the current traffic of all shared traffic packages until the current traffic is not enough for the simulated deduction, and a traffic alarm is issued. The shared traffic packages of different resource pools are uniformly simulated and deducted, and an alarm is issued when the current traffic is not enough for the deduction. This can prompt users to replenish new shared traffic packages in advance, thereby avoiding the problem of traffic package balance alarm failing, skipping the alarm, and directly incurring traffic charges.

[0061] Optionally, in the remaining warning method, when the shared traffic package is a first type shared traffic package, the method further includes:

[0062] Obtaining a remaining ratio of the shared traffic package; wherein the remaining ratio is a ratio of the current traffic of the shared traffic package to the total traffic value of the shared traffic package;

[0063] When the surplus ratio is less than a second threshold, the first historical data is input into the flow analysis prediction model.

[0064] In this embodiment, the first type of shared traffic packet is a common shared traffic packet without IP restriction, such as Figure 3 As shown, the original traffic of IP A is deducted in sequence from the customer's current valid shared traffic package resource list and the deduction record is stored until all ordinary shared traffic packages are deducted. The balance warning of ordinary shared traffic packages is to monitor and manage the balance of ordinary shared traffic packages currently held by customers. Without new orders, the balance of customers' ordinary shared traffic packages will be deducted less and less within a month. At the beginning of next month, the total amount of traffic packages may increase because the customer has a designated shared traffic package that is effective. The traffic generated by an ordinary IP (i.e., the first IP) can only be offset by an ordinary shared traffic package. Therefore, warning the balance of ordinary shared traffic packages can provide balance reminders and warnings for all traffic packages consumed by IPs that are charged by traffic. Therefore, the following method is adopted Figure 7 In the manner shown, the currently valid ordinary shared traffic package balance (remainA) and the total traffic value (total) of the shared traffic package are obtained before deduction, and the currently valid ordinary shared traffic package balance (remainB) is calculated after deduction. When the balance ratio after deduction is less than the second threshold (Alert Num), the predicted traffic for the next natural day is considered to be simulated for deduction. Before this, if the balance ratio is not less than the second threshold, there is no need to perform the simulated deduction process, thereby avoiding waste of computing resources.

[0065] Optionally, in the surplus warning method, when the at least one shared traffic packet is a designated IP shared traffic packet of the second IP, the method further comprises:

[0066] The second IP is determined as the first IP.

[0067] In this embodiment, if Figure 4 As shown, when traffic package TRAFFIC B is a designated IP shared traffic package of IP A and IP C, during the deduction process of IP B, TRAFFIC A and TRAFFIC C can be deducted normally, but the designated IP shared traffic package is skipped. The traffic generated by the IP (i.e., the second IP) in the shared traffic package subscribed by the designated IP can be deducted from both the designated traffic package and the ordinary traffic package. The designated IP deduction function is generally used for users with special business needs. Some IPs can deduct part of the traffic package resources, and other IPs do not use the traffic package resources. For example Figure 8As shown in the figure, IP A can deduct TRAFFIC A and TRAFFIC C of unlimited IP, and can also deduct TRAFFIC B of limited IP A and IP C. Before the deduction, the remaining amount (IP AremainA) and the total amount (IP Atotal) of the valid common shared traffic package (all shared traffic packages of the current IP) that the current IP can deduct are calculated. After the deduction, the remaining amount (IPAremain B) of the valid common shared traffic package that the current IP can deduct is calculated. If the balance ratio before deduction (IP Aremain A / IP Atotal) is greater than the second threshold (Alert Num), and the balance ratio after deduction (IP Aremain B / IP Atotal) is less than the second threshold (Alert Num), the second IP is taken as the first IP, and its historical traffic is input into the traffic analysis and prediction model for prediction. If it is predicted that the current traffic of the ordinary shared traffic package (including traffic packages with limited IP and traffic packages with no limited IP) for the IP within the next natural day is not enough for simulated deduction, a balance warning for the IP is sent to the user, thereby avoiding the situation where the unused balance of the shared traffic package is occupied due to the failure to set a specific IP for the traffic package of the specified IP, resulting in the shared traffic package balance not reaching the warning value and skipping the alarm, thereby directly incurring traffic charges.

[0068] Optionally, the remaining warning method further includes:

[0069] Acquire second historical data of a first time period to be trained and third historical data of a second time period of a plurality of the first IPs; wherein the second time period is after the first time period;

[0070] Acquire a network topology map of the first IP according to the second historical data;

[0071] Inputting the network topology graph into a spatiotemporal graph convolutional network model integrating an attention mechanism to obtain a prediction result; wherein the prediction result is prediction data for the third historical data;

[0072] The spatiotemporal graph convolutional network model integrating the attention mechanism is adjusted according to the prediction result and the third historical data to obtain the traffic analysis prediction model.

[0073] It should be noted that for network traffic data, the data is time-series, that is, the traffic at the previous moment and the traffic at the current moment will affect the traffic at the next moment. Deep learning models can be effectively applied to such scenarios. In cloud-based networks, customer resources are all in the cloud. The process of network data from source to destination is actually manifested as a data flow process in the network structure. The traffic of different IPs has a certain spatial correlation due to data flow and affects each other.

[0074] In this embodiment, the network traffic trends of the next IPs are predicted by observing the network traffic data generated by the first IPs, and the IPs held by the customer and their interactions are abstracted into a network topology graph G = (V, E, A), which is input into the spatiotemporal graph convolutional network model (CA-ASTGCN) with fusion attention mechanism for traffic analysis and prediction. After the model is built, a data set is made according to the second historical data of the first time period of the first IPs of the user and the traffic data generated every hour, and the traffic data is input into the spatiotemporal graph convolutional network model with fusion attention mechanism in batches, and predicted in stages. The prediction results are compared with the third historical data of the second time period, so as to adjust the spatiotemporal graph convolutional network model with fusion attention mechanism and obtain the traffic analysis prediction model.

[0075] Optionally, in the surplus warning method, the step of acquiring a network topology diagram of the first IP according to the second historical data includes:

[0076] Abstract each of the first IPs into a network node, and obtain a network node set;

[0077] Abstracting the data transmission between the network nodes as edges, and obtaining an edge set;

[0078] Acquire an adjacency matrix according to the network nodes and the edges;

[0079] The network topology graph is obtained according to the network node set, the edge set and the adjacency matrix.

[0080] In this embodiment, each of the first IPs is abstracted as a network node V, and the data forwarding between network nodes constitutes the edge E of the network topology graph. A is the adjacency matrix of the network topology graph, and the network topology graph G = (V, E, A) is obtained, which reflects the connectivity between the nodes.

[0081] Optionally, in the margin warning method, the spatiotemporal graph convolutional network model integrating the attention mechanism includes a first fully connected layer, a first asynchronous spatiotemporal convolutional module, a first channel attention module, a second asynchronous spatiotemporal convolutional module, a second channel attention module and a second fully connected layer connected in sequence; wherein the first and the second asynchronous spatiotemporal convolutional modules are each composed of a plurality of spatiotemporal convolutional layers including graph convolution modules and spatiotemporal convolutional layers including asynchronous time convolution modules, alternately stacked.

[0082] In this embodiment, the spatiotemporal graph convolutional network model integrating the attention mechanism is as follows: Figure 5 As shown, it includes a first fully connected layer, a first asynchronous spatiotemporal convolution module, a first channel attention module, a second asynchronous spatiotemporal convolution module, a second channel attention module and a second fully connected layer connected in sequence. Among them, the two asynchronous spatiotemporal convolution modules first use the graph convolution module to extract the node spatial information, and then use the asynchronous time convolution module to extract the node time information. The two modules are alternately used to transmit the node information to the neighboring nodes in time and space. As the number of convolution layers increases, a large amount of redundant information will be stored in the model, and the model performance will decrease accordingly. In order to quickly obtain effective information, after the completion of multiple time convolution modules, the two asynchronous spatiotemporal convolution modules respectively use the first channel attention module and the second channel attention module to improve the model performance. The first fully connected layer converts the input data features into high-dimensional space features to improve the representation ability of the network. The output layer uses the second fully connected layer to convert the high-dimensional space data into prediction results.

[0083] The asynchronous spatiotemporal graph convolution operation (inputting the network topology graph into the asynchronous spatiotemporal convolution module) can be expressed as follows:

[0084]

[0085] in, represents the weighted asynchronous spatiotemporal correlation matrix, represents the weight matrix, represents the asynchronous spatiotemporal correlation matrix, ⊙ represents element-by-element multiplication, is the input of the convolution operation, W and b are learnable parameters, h = σ(A ′ XW+b) represents the graph convolution operation, and σ represents the activation function. The asynchronous spatiotemporal graph convolution operation aggregates a large amount of information from the previous time step (a time step is usually one hour, but can also be other time lengths) and the next time step. Its structure is as follows Figure 6As shown in the figure, the feature map after the asynchronous spatiotemporal convolution operation is firstly pooled to obtain a feature vector of 1*C (number of channels), and then sent to a two-layer neural network for training. The first layer reduces the number of parameters by using a small number of neurons. The number of neurons in the second layer is the same as that in the input layer. The output feature vector is activated using sigmoid (activation function), and the activated weight vector is weighted with the input feature map.

[0086] Optionally, in the remaining battery warning method, step S30 includes:

[0087] Simulate the flow deduction process of the at least one shared flow package within each third preset duration according to the first predicted flow consumption value, and obtain a flow consumption diagram of the at least one shared flow package; wherein the second preset duration includes a plurality of the third preset durations;

[0088] The flow consumption diagram is sent to the customer as the content of the flow alarm prompt.

[0089] In this embodiment, the hourly shared traffic package deduction process is simulated based on the first predicted traffic consumption value of the user in a future natural day, and the traffic consumption diagram is drawn to indicate the shared traffic package balance before and after each deduction, which is sent to the customer as the content of the traffic alarm prompt. If the traffic package is not enough for the deduction, it will be marked in a striking color on the page to achieve a warning effect.

[0090] Optionally, in the remaining battery warning method, step S20 includes:

[0091] Obtain a list of currently valid shared traffic packages in the at least one shared traffic package; wherein the shared traffic packages in the list are sorted from early to late according to expiration dates;

[0092] Perform traffic simulation deduction on each of the multiple first IPs in the current traffic one by one in the order of the list, deduct the idle time traffic first, and then deduct the full-time traffic after the idle time traffic is deducted; wherein the current traffic includes the idle time traffic and the full-time traffic.

[0093] In this embodiment, the order of deduction of shared traffic packages is as follows: deduction for each IP one by one, with priority given to deduction during off-peak hours, and then deduction for full-time hours. A list of currently valid shared traffic packages is obtained, and priority is given to deducting the shared traffic packages that expire first. If there is any surplus after the deduction of the single resource pool traffic package, the global shared traffic package will be deducted. The deduction logic is consistent with the deduction order of the single resource pool. The global shared traffic package can deduct the traffic generated by the IP and IPv6 bandwidth of all the customer's resource pools that are billed by traffic.

[0094] like Fig. 9As shown, in order to achieve the above-mentioned purpose, an embodiment of the present invention further provides a remaining warning device for a shared traffic package, which includes:

[0095] The first acquisition module 901 is used to input the first historical data of the use of shared traffic resources by multiple first IPs within a first preset time period into a pre-acquired traffic analysis prediction model to obtain a first predicted traffic consumption value of the first IP within a second preset time period after the current time;

[0096] A second acquisition module 902 is used to perform a flow simulation deduction on the current flow of at least one shared flow package in the shared flow resource according to the first predicted consumed flow value, and obtain the remaining flow of the current flow after the flow simulation deduction;

[0097] The first processing module 903 is used to issue a flow alarm prompt when the remaining flow is less than a first threshold.

[0098] Optionally, the remaining warning device further includes:

[0099] A third acquisition module is used to obtain the remaining ratio of the shared traffic package; wherein the remaining ratio is the ratio of the current traffic of the shared traffic package to the total traffic value of the shared traffic package;

[0100] The second processing module is used to input the first historical data into the flow analysis prediction model when the surplus ratio is less than a second threshold.

[0101] Optionally, the remaining warning device further includes:

[0102] A first determining module is used to determine the second IP as the first IP.

[0103] Optionally, the remaining warning device further includes:

[0104] A fourth acquisition module, used to acquire second historical data of a first time period to be trained and third historical data of a second time period of a plurality of the first IPs; wherein the second time period is after the first time period;

[0105] A fifth acquisition module, configured to acquire a network topology map of the first IP according to the second historical data;

[0106] A sixth acquisition module, used for inputting the network topology graph into the spatiotemporal graph convolutional network model integrating the attention mechanism to obtain a prediction result; wherein the prediction result is the prediction data of the third historical data;

[0107] The seventh acquisition module is used to adjust the spatiotemporal graph convolutional network model of the fusion attention mechanism according to the prediction result and the third historical data to obtain the traffic analysis prediction model.

[0108] Optionally, in the remaining warning device, the fifth acquisition module includes:

[0109] A first acquisition unit, configured to abstract each of the first IPs into a network node and acquire a network node set;

[0110] A second acquisition unit is used to abstract the data transmission between the network nodes into edges and acquire an edge set;

[0111] A third acquisition unit, configured to acquire an adjacency matrix according to the network nodes and the edges;

[0112] The fourth acquisition unit is used to acquire the network topology graph according to the network node set, the edge set and the adjacency matrix.

[0113] Optionally, the remaining warning device, wherein the spatiotemporal graph convolutional network model integrating the attention mechanism includes a first fully connected layer, a first asynchronous spatiotemporal convolutional module, a first channel attention module, a second asynchronous spatiotemporal convolutional module, a second channel attention module and a second fully connected layer connected in sequence; wherein the first asynchronous spatiotemporal convolutional module and the second asynchronous spatiotemporal convolutional module are both composed of a plurality of spatiotemporal convolutional layers including graph convolution modules and spatiotemporal convolutional layers including asynchronous time convolution modules, which are alternately stacked.

[0114] Optionally, in the remaining warning device, the first processing module 903 includes:

[0115] A fifth acquisition unit, configured to simulate the flow deduction process of the at least one shared flow package within each third preset duration according to the first predicted flow consumption value, and acquire a flow consumption diagram of the at least one shared flow package; wherein the second preset duration includes a plurality of the third preset durations;

[0116] The first sending unit is used to send the flow consumption diagram as the content of the flow alarm prompt to the customer.

[0117] Optionally, in the remaining warning method, the second obtaining module 902 includes:

[0118] A sixth acquisition unit, configured to acquire a list of currently valid shared traffic packages in the at least one shared traffic package; wherein the shared traffic packages in the list are sorted from early to late according to expiration dates;

[0119] The first processing unit is used to perform traffic simulation deduction on each of the multiple first IPs in the current traffic one by one in the order of the list, first deducting the off-peak traffic, and then deducting the full-time traffic after the off-peak traffic is deducted; wherein the current traffic includes the off-peak traffic and the full-time traffic.

[0120] It should be noted here that the above-mentioned device provided in the embodiment of the present invention can implement all the method steps implemented in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects that are the same as the method embodiment in this embodiment will not be described in detail here.

[0121] In order to achieve the above-mentioned purpose, an embodiment of the present invention also provides an electronic device, including: a processor, a memory, and a program or instruction stored in the memory and executable on the processor; wherein, when the processor executes the program or instruction, the method for warning the remaining capacity of the shared traffic package as described above is implemented.

[0122] In order to achieve the above-mentioned purpose, an embodiment of the present invention also provides a readable storage medium, on which a program or instruction is stored, wherein when the program or instruction is executed by a processor, the steps in the shared traffic package balance warning method as described above are implemented.

[0123] It should be further explained that the terminals described in this specification include but are not limited to smart phones, tablet computers, etc., and many of the functional components described are called modules in order to more particularly emphasize the independence of their implementation methods.

[0124] In the embodiment of the present invention, module can be implemented with software so that it can be executed by various types of processors. For example, an executable code module of an identification can include one or more physical or logical blocks of computer instructions, for example, it can be constructed as an object, process or function. Nevertheless, the executable code of the identified module does not need to be physically located together, but can include different instructions stored in different positions, and when these instructions are logically combined together, it constitutes a module and realizes the specified purpose of the module.

[0125] In fact, executable code module can be a single instruction or many instructions, and can even be distributed on a plurality of different code segments, distributed among different programs, and distributed across a plurality of memory devices. Similarly, operating data can be identified in the module, and can be implemented and organized in the data structure of any appropriate type according to any appropriate form. The operating data can be collected as a single data set, or can be distributed in different locations (including on different storage devices), and can only be present on a system or network as an electronic signal at least in part.

[0126] When a module can be implemented by software, considering the level of existing hardware technology, a person skilled in the art can build a corresponding hardware circuit to implement the corresponding function of the module that can be implemented by software without considering the cost. The hardware circuit includes a conventional very large scale integration (VLSI) circuit or gate array and existing semiconductors such as logic chips, transistors, or other discrete components. The module can also be implemented by a programmable hardware device, such as a field programmable gate array, a programmable array logic, a programmable logic device, etc.

[0127] In order to achieve the above-mentioned purpose, an embodiment of the present invention further provides a computer program product, which includes computer instructions, and when the computer instructions are executed by a processor, the steps of the shared traffic package remaining warning method as described above are implemented.

[0128] The above exemplary embodiments are described with reference to the accompanying drawings, and many different forms and embodiments are feasible without departing from the spirit and teachings of the present invention. Therefore, the present invention should not be constructed as a limitation of the exemplary embodiments proposed herein. More specifically, these exemplary embodiments are provided so that the present invention will be perfect and complete, and the scope of the present invention will be conveyed to those who are familiar with the technology. In these figures, the component sizes and relative sizes may be exaggerated for clarity. The terms used here are only based on the purpose of describing specific exemplary embodiments and are not intended to be limiting. As used herein, unless the text clearly indicates otherwise, the singular forms "one", "an" and "the" are intended to include these multiple forms. It will be further understood that the terms "including" and / or "comprising" when used in this specification indicate the presence of the features, integers, steps, operations, components and / or components, but do not exclude the presence or increase of one or more other features, integers, steps, operations, components, components and / or their groups. Unless otherwise indicated, when stated, a range of values ​​includes the upper and lower limits of that range and any subranges therebetween.

[0129] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for early warning of the remaining amount of a shared traffic package, characterized in that: include: Taking first historical data of the use of shared traffic resources by multiple first IPs within a first preset time as input, inputting it into a pre-acquired traffic analysis prediction model, and obtaining a first predicted traffic consumption value of the first IP within a second preset time after the current time; According to the first predicted consumed traffic value, a traffic simulation deduction is performed on the current traffic of at least one shared traffic package in the shared traffic resource to obtain the remaining traffic of the current traffic after the traffic simulation deduction; When the remaining flow is less than the first threshold, a flow alarm prompt is issued.

2. The remaining quantity warning method according to claim 1 is characterized in that: In the case where the shared traffic packet is a first type shared traffic packet, the method further includes: Obtaining a remaining ratio of the shared traffic package; wherein the remaining ratio is a ratio of the current traffic of the shared traffic package to the total traffic value of the shared traffic package; When the surplus ratio is less than a second threshold, the first historical data is input into the flow analysis prediction model.

3. The remaining quantity warning method according to claim 1 is characterized in that: In the case where the at least one shared traffic packet is a designated IP shared traffic packet of the second IP, the method further includes: The second IP is determined as the first IP.

4. The remaining quantity warning method according to claim 1 is characterized in that: The method further comprises: Acquire second historical data of a first time period to be trained and third historical data of a second time period of a plurality of the first IPs; wherein the second time period is after the first time period; Acquire a network topology map of the first IP according to the second historical data; Inputting the network topology graph into a spatiotemporal graph convolutional network model integrating an attention mechanism to obtain a prediction result; wherein the prediction result is prediction data for the third historical data; The spatiotemporal graph convolutional network model integrating the attention mechanism is adjusted according to the prediction result and the third historical data to obtain the traffic analysis prediction model.

5. The remaining quantity warning method according to claim 4 is characterized in that: The acquiring the network topology diagram of the first IP according to the second historical data includes: Abstract each of the first IPs into a network node, and obtain a network node set; Abstracting the data transmission between the network nodes as edges, and obtaining an edge set; Acquire an adjacency matrix according to the network nodes and the edges; The network topology graph is obtained according to the network node set, the edge set and the adjacency matrix.

6. The remaining quantity warning method according to claim 4 is characterized in that: The spatiotemporal graph convolutional network model integrating the attention mechanism includes a first fully connected layer, a first asynchronous spatiotemporal convolution module, a first channel attention module, a second asynchronous spatiotemporal convolution module, a second channel attention module and a second fully connected layer connected in sequence; wherein, the first asynchronous spatiotemporal convolution module and the second asynchronous spatiotemporal convolution module are both composed of a plurality of spatiotemporal convolution layers including graph convolution modules and spatiotemporal convolution layers including asynchronous time convolution modules, which are alternately stacked.

7. The remaining quantity warning method according to claim 1, characterized in that: The flow alarm prompt includes: Simulate the flow deduction process of the at least one shared flow package within each third preset duration according to the first predicted flow consumption value, and obtain a flow consumption diagram of the at least one shared flow package; wherein the second preset duration includes a plurality of the third preset durations; The flow consumption diagram is sent to the customer as the content of the flow alarm prompt.

8. The remaining quantity warning method according to claim 1 is characterized in that: The performing flow simulation deduction in the current flow of at least one shared flow package in the shared flow resource according to the first predicted consumed flow value comprises: Obtain a list of currently valid shared traffic packages in the at least one shared traffic package; wherein the shared traffic packages in the list are sorted from early to late according to expiration dates; Perform traffic simulation deduction on each of the multiple first IPs in the current traffic one by one in the order of the list, deduct the idle time traffic first, and then deduct the full-time traffic after the idle time traffic is deducted; wherein the current traffic includes the idle time traffic and the full-time traffic.

9. A shared traffic package surplus warning device, characterized in that: include: A first acquisition module is used to input first historical data of the use of shared traffic resources by multiple first IPs within a first preset time period into a pre-acquired traffic analysis prediction model to obtain a first predicted traffic consumption value of the first IP within a second preset time period after the current time; A second acquisition module is used to perform a flow simulation deduction on the current flow of at least one shared flow package in the shared flow resource according to the first predicted consumed flow value, and obtain the remaining flow of the current flow after the flow simulation deduction; The first processing module is used to issue a flow alarm prompt when the remaining flow is less than a first threshold.

10. An electronic device comprising: A processor, a memory, and a program or instruction stored in the memory and executable on the processor; characterized in that when the processor executes the program or instruction, the method for warning the remaining amount of a shared traffic package as described in any one of claims 1-8 is implemented.

11. A readable storage medium having a program or instruction stored thereon, characterized in that: When the program or instruction is executed by the processor, the steps in the shared traffic package remainder warning method as described in any one of claims 1-8 are implemented.

12. A computer program product, characterized in that It includes computer instructions, which, when executed by a processor, implement the steps of the shared traffic package remainder warning method as described in any one of claims 1-8.

Images