A distributed storage method and system based on AI traffic identification

Abstract

The application discloses a distributed storage method and system based on AI traffic identification, and comprises the following steps: acquiring current CDN network node state information and simulating a first GNN node in a GNN model, and constructing adjacent nodes of the GNN model according to the communication relationship between the current CDN network node and different CDN network nodes; acquiring communication traffic information between the first GNN node and the corresponding adjacent nodes, identifying traffic feature data in the communication traffic information by using an AI model, and constructing a dynamic adjacency matrix between the first GNN node and the adjacent nodes; updating the state of the first GNN node and the adjacent nodes by using the dynamic adjacency matrix, executing a cache strategy or a traffic processing strategy of the corresponding CDN network node according to the updated state of the first GNN node and the adjacent nodes; and routing the nodes of the current CDN network to the nodes of the optimal CDN network according to the cache strategy or the traffic processing strategy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of distributed storage, and particularly relates to a distributed storage method and system based on AI traffic identification. BACKGROUND

[0002] The existing traditional distributed storage method is to store data on multiple physical location servers to realize data redundancy and backup. Even if a server fails, the data copy stored on other servers is still available, thereby ensuring data persistence and system stability. The above-mentioned traditional distributed storage does not comprehensively analyze the traffic, such as data traffic with network threats or data traffic with illegal content, which are stored according to the method of distributed data sharding, thereby resulting in low safety and cleanliness of the stored data, and malicious occupation of server storage resources in the state of server node attack. At the same time, the traditional distributed storage does not effectively analyze the data traffic in the storage resource scheduling process, thereby resulting in low distributed storage efficiency. SUMMARY

[0003] One of the purposes of the present application is to provide a distributed storage method and system based on AI traffic identification, which utilizes an AI model to identify the traffic data of different storage nodes. The present application utilizes an AI model to identify illegal content data that may exist in the traffic data, and based on the identified illegal data, the traffic data is intercepted or scheduled, thereby realizing directional processing of illegal content data traffic, improving the data safety of distributed storage, reducing the occupation of distributed storage resources by illegal content data, and improving the effect of distributed data storage.

[0004] Another purpose of the present application is to provide a distributed storage method and system based on AI traffic identification, which utilizes a CDN network (content distribution network) as a distributed storage system, utilizes an AI model to identify and analyze the traffic of different nodes of the CDN network, obtains illegal content data of the traffic of different nodes of the CDN network, and based on the illegal content data, carries out corresponding CDN network node caching strategy. In the present application, efficient nearby access is utilized based on the CDN network node caching, while the safety of access is also taken into account, and the compliance effect of access data content is improved, and the occupation of bandwidth resources or illegal content data caching resources by illegal access is reduced.

[0005] Another object of the present application is to provide a distributed storage method and system based on AI traffic identification, which simulates the behavior of nodes including but not limited to traffic data of CDN network nodes using a GNN model (graph neural network model), wherein the CDN network node related data is used as the node feature in the GNN model, and the results of AI analysis of traffic data between different CDN network nodes are used as the connection edge feature (adjacency matrix) between different nodes in the GNN model, and the GNN model is used to construct a dynamic adjacency matrix of traffic features based on AI model identification to perform data update or cache strategy of the corresponding CDN network node. The different node relationship description method between the graph nodes in the GNN model can effectively solve the description method of traffic data between different CDN network nodes, thereby efficiently realizing the technical effect of safe and compliant distributed storage.

[0006] To achieve at least one of the above-mentioned objects, the present application further provides a distributed storage method based on AI traffic identification, comprising:

[0007] Obtaining current CDN network node state information, simulating and constructing a first GNN node in the GNN model according to the current CDN network node state information, and constructing an adjacent node of the GNN model according to the communication relationship between the current CDN network node and different CDN network nodes;

[0008] Obtaining the communication traffic information between the first GNN node and the corresponding adjacent node, identifying the traffic feature data in the communication traffic information using an AI model, and constructing a dynamic adjacency matrix between the first GNN node and the adjacent node according to the traffic feature data;

[0009] Using the dynamic adjacency matrix to update the state of the first GNN node and the adjacent node, and executing the cache strategy or traffic processing strategy of the corresponding CDN network node according to the updated state of the first GNN node and the adjacent node;

[0010] Obtaining the current CDN network node routing link table, and routing the nodes of the current CDN network to the nodes of the optimal CDN network according to the cache strategy or traffic processing strategy.

[0011] According to a preferred embodiment of the present application, the current CDN network node state information includes hardware resource state, network resource state and cache state, wherein the hardware resource state includes CPU usage, memory usage, and read-write performance data of cache disk; the network resource state includes uplink and downlink bandwidth usage of the current CDN network node, packet loss rate of the current CDN network node, network request rate, concurrent connection number and URL number; and the cache state includes cache hit rate, cache eviction rate, cache warm-up state and cache content type; the hardware resource state, network resource state and cache state are respectively preprocessed to obtain the feature matrix of the first GNN node, and the dynamic adjacency matrix is constructed according to the traffic feature data between the first GNN node and the corresponding adjacent node identified by the AI model.

[0012] According to another preferred embodiment of the present application, the construction method of the dynamic adjacency matrix includes: identifying and obtaining network traffic feature data between the first GNN node and the corresponding adjacent node by using an image recognition AI model, a character recognition AI model and a voice recognition AI model; wherein the traffic feature data includes illegal video frame data, illegal voice data, illegal character data and quantity; and different traffic feature data are respectively preprocessed and converted into the dynamic adjacency matrix, and the state of the first GNN node and the corresponding adjacent node is updated according to the dynamic adjacency matrix.

[0013] According to another preferred embodiment of the present application, the state updating method of the first GNN node and the corresponding adjacent node includes: constructing a multi-head attention coefficient wherein i represents the current first GNN node, j represents an adjacent node having a communication relationship with the first GNN node, h i represents the first GNN node feature, a and W represent different learnable parameters, W is a weight matrix, a is an action parameter between nodes converted based on traffic feature data, T represents a transposed matrix identifier, LeakyReLU represents an activation function, Softmax represents a linear classification function, and the attention coefficient ij represents the similar feature weight of the first GNN node and the corresponding adjacent node.

[0014] According to another preferred embodiment of the present application, the state updating method of the first GNN node and the corresponding adjacent node further includes: performing parallel calculation on the multi-head attention coefficient ij for capturing the relationship type between the first GNN node i and the adjacent node j, and updating the content of the first GNN node by using the multi-head attention coefficient ij .ij constructing a dynamic adjacency matrix A dynamic [i,j] k = , wherein k represents a corresponding attention head, and replacing the original static adjacency matrix A dynamic [i,j] k with the dynamic adjacency matrix A static , and performing an update of the first GNN node according to the following formula after the replacement , wherein represents an updated predicted first GNN node state, N represents a node relationship function, K represents a total number of attention heads, and σ is an activation function, and performing a node cache policy or a traffic processing policy of a corresponding current CDN network according to the updated predicted first GNN node state .

[0015] According to another preferred embodiment of the present application, the cache policy or the traffic processing policy comprises: allowing incoming cache of corresponding adjacent node data according to the updated predicted first GNN node state of the node of the current CDN network; and predicting a cache state of the node of the current CDN network from corresponding adjacent node violation data according to the updated predicted first GNN node state .

[0016] According to another preferred embodiment of the present application, the CDN network node comprises a central CDN network node and an edge CDN network node, and after obtaining a routing link table of a current CDN network node, the central CDN network node is used to determine a closest edge CDN network node of the current CDN network node according to the routing link table of the current CDN network node, and data of adjacent nodes are transmitted to the closest edge CDN network node through routing conversion according to the cache policy, so that efficient processing of cache data is realized.

[0017] According to another preferred embodiment of the present application, a distance L between the closest edge CDN network nodes of the current CDN network node is calculated by: obtaining a round-trip time RTT between the current CDN network node and any one of the edge CDN network nodes, calculating a hop number Hops to the edge CDN network node, and calculating a bandwidth utilization rate P of the current CDN network node, and a calculation formula of the distance L between the closest edge CDN network nodes of the current CDN network node is as follows: , wherein β, λ and γ are weight coefficients.

[0018] To achieve at least one of the above-mentioned objects, the present application further provides a distributed storage system based on AI traffic identification, which executes the above-mentioned distributed storage method based on AI traffic identification.

[0019] The present application further provides a computer readable storage medium storing a computer program, which is executed by a processor to implement the above-mentioned distributed storage method based on AI traffic identification. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The figure shows a flowchart of the distributed storage method based on AI traffic identification of the present application. DETAILED DESCRIPTION

[0021] The following description is provided to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only examples, and other obvious modifications can be made by those skilled in the art. The basic principles defined in the following description can be applied to other embodiments, modifications, improvements, equivalents and other technical solutions without departing from the spirit and scope of the present application.

[0022] It can be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of one element can be one, and in another embodiment, the number of the element can be multiple, and the term "one" cannot be understood as a limitation on the number.

[0023] Please refer to Figure 1This invention discloses a distributed storage method and system based on AI traffic identification. The method includes: firstly, acquiring CDN network node status information, and then performing feature transformation to simulate and construct GNN nodes (Graph Neural Network nodes) based on the CDN network node status information. The CDN network nodes are content delivery network nodes, including central CDN network nodes and edge CDN network nodes. The central CDN network node acts as a master control node, monitoring multiple edge CDN network nodes and performing content distribution caching on these edge CDN network nodes. The central CDN network node can also perform traffic control between multiple edge network nodes, or direct the distribution of content from a corresponding edge CDN network node to a specific CDN network node. This invention utilizes the CDN network nodes for distributed storage while implementing different caching and traffic control strategies based on the traffic characteristics between different CDN network nodes, particularly targeting caching and traffic control strategies for illegal data. This invention utilizes a GNN (Graph Neural Network) model as the driving model for distributed storage strategies. The GNN model simulates the states of CDN network nodes and the relationships between different CDN network nodes. Adjacency relationships based on the GNN model are constructed according to traffic characteristic data between different CDN network nodes (equivalent to connecting edges between different GNN model nodes). State changes of GNN model nodes are determined based on these adjacency relationships, and corresponding caching and traffic control strategies for the CDN network nodes are executed based on the updated and predicted GNN model node states. Therefore, this invention can effectively improve the security performance of distributed caching, effectively reduce malicious occupation of distributed storage resources caused by abnormal traffic, and improve the resource utilization rate of distributed storage.

[0024] Specifically, the method for acquiring the CDN network node state information for GNN model node construction comprises: acquiring the CDN network node state information, including hardware resource state, network resource state and cache state, wherein the hardware resource state includes but is not limited to CPU usage rate, GPU usage rate, memory usage rate, read-write performance data of cache disk and the like of the current CDN network node, and the hardware resource state is used to represent the overall performance of the edge device where the CDN network node is located. Generally, the CDN network node of the edge device is usually used as a distributed cache hardware resource, and the hardware devices deployed in different edge devices have different performances. The network resource state includes but is not limited to uplink and downlink bandwidth usage rate, packet loss rate, network request rate, concurrent connection quantity and URL quantity of the current CDN network node, and the network resource state of the CDN network node is automatically identified according to the network environment change of the current CDN network node. For example, in the process of high concurrency interaction of multiple CDN network nodes, there may be a situation of insufficient network bandwidth and decreased network request speed. The identification of the above network resource state can provide reference data for CDN network node cache strategy and traffic limitation strategy to a certain extent, thereby improving the distributed storage efficiency of the CDN network node as a whole. The cache state includes but is not limited to cache hit rate, cache eviction rate, cache warm-up state and cache content type and the like. The cache hit rate is used to describe the performance of the cache system, especially in a specific business scenario. If the cache hit performance is high, it indicates that the cache system of the CDN network node has good distribution effect on the current business content request, and high adaptability of the edge device CDN network node and the corresponding business content can be realized. Therefore, the cache hit rate is used as the state parameter of the GNN graph node, which can better describe the adaptability of different CDN network nodes to related business traffic. Similarly, based on the cache eviction rate, the cache warm-up state and the cache content type, the state parameters of the GNN graph node can describe the adaptability effect of the corresponding edge device CDN network node and the corresponding business content, which will not be described in detail herein.

[0025] Further, after obtaining the relevant state data of the hardware resource state, network resource state and cache state of the current CDN network node, the application adopts data preprocessing including but not limited to normalization, standardization and feature transformation, so that the relevant state data of the current CDN network node can be more standardized, wherein the application adopts label encoding (labelEncoder) method to classify and construct different features for the standardized relevant state data. The above label encoding (labelEncoder) method is an existing classification feature construction method, and the application will not be described in detail. The relevant state data of the hardware resource state, network resource state and cache state of the current CDN network node is used to construct the GNN graph node feature matrix by using the label encoding (labelEncoder) method, and the feature matrix is defined as the feature matrix of the first GNN node, which is used to describe the relevant state data of the current CDN network node.

[0026] It is worth mentioning that one of the core technical points of the application is to construct the adjacency matrix of the first GNN node, wherein the adjacency matrix of the first GNN node is a matrix describing the relationship between the first GNN node and the adjacent node, wherein the adjacent node and the first GNN node in the application are defined as CDN network nodes having a communication relationship, especially CDN network nodes having a content distribution scheduling relationship. The adjacency matrix contains the traffic feature data between the first GNN node and the adjacent node, and the adjacency matrix constructed by the traffic feature data is used to update and predict the state change of the first GNN node and the adjacent node itself. It should be noted that the traffic data can be from the first GNN node to the adjacent node, or from the adjacent node to the first GNN node, and the above definition name can be converted according to the actual traffic transmission situation, and the application will not be limited specifically.

[0027] The present application needs to identify the traffic feature data present in the traffic data. Since there may be traffic data including but not limited to illegal video, audio and text in the traffic data, the present application needs to use the existing image recognition AI model, text recognition AI model and speech recognition AI model to identify and obtain the network traffic feature data between the first GNN node and the corresponding adjacent node, and use the traffic feature output label identified by the different AI models to further construct the traffic feature matrix based on the traffic feature label (label). The present application uses the traffic feature matrix as the adjacency matrix between the first GNN node and the adjacent node to drive the state update of the first GNN node and the adjacent node. The traffic feature label includes but is not limited to illegal video frame data, illegal voice data, illegal text data and quantity.

[0028] It should be noted that the adjacency matrix of the GNN network node in the traditional GNN model is generally a static matrix used for message passing to update the GNN network node state. For example, the formula of the adjacency matrix message passing mechanism of the traditional GNN model is: , wherein A Static is a conventional static adjacency matrix, is the GNN node state before update, represents the predicted GNN node state after update, and all are in the form of node state feature matrix; N represents a node relationship function; i and j represent different nodes; W is a learnable parameter, which is generally a weight matrix corresponding to the node state feature.

[0029] In order to adapt to the dynamic update of the GNN network node state driven by the dynamic traffic feature, the present application replaces the conventional static adjacency matrix A with a dynamic adjacency matrix A dynamic [i, j] k based on traffic feature data, dynamic [i, j] k The implementation method of the dynamic adjacency matrix A , wherein i represents the current first GNN node, j represents the adjacent node having a communication relationship with the first GNN node, h i represents the first GNN node feature, a and W represent different learnable parameters, wherein W is a weight matrix and a is a node action parameter based on traffic feature data conversion, T represents a transpose matrix identifier, LeakyReLU represents an activation function, and Softmax represents a linear classification function. The attention coefficient a ijdenotes the similar feature weight of the first GNN node and the corresponding adjacent node, and further constructs the multi-head attention coefficient α based on the traffic feature data ij constructs the dynamic adjacency matrix A dynamic [i,j] k =α ij , wherein the multi-head attention coefficient α ij represents the influence of different types of traffic features on the update of the first GNN node and the corresponding adjacent node, so that the dynamic adjacency matrix A can be directly constructed by parameter conversion dynamic [i,j] k .

[0030] Further, the multi-head attention coefficient α ij needs to be calculated in parallel in the present application, which is used to capture the relationship type between the first GNN node i and the adjacent node j, and the updated first GNN node is calculated by the multi-head attention coefficient α ij The method for calculating the updated first GNN node comprises: constructing a dynamic adjacency matrix A ij using the multi-head attention coefficient α dynamic [i,j] k = , wherein k represents the corresponding attention head, and the dynamic adjacency matrix A dynamic [i,j] k replaces the original static adjacency matrix A static , and after replacement, the update of the first GNN node is performed according to the following formula , wherein represents the updated predicted first GNN node state, N represents the node relationship function, K represents the total number of attention heads, and σ is an activation function, and according to the updated predicted first GNN node state The node cache strategy or traffic processing strategy of the corresponding current CDN network is executed.

[0031] The cache strategy or traffic processing strategy comprises: according to the updated predicted first GNN node state , the incoming cache of the corresponding adjacent node data is allowed to update and predict the cache state of the node of the current CDN network; and according to the updated predicted first GNN node state , the access and cache strategy including flow limiting or prohibition is performed on the predicted illegal data from the corresponding adjacent node of the node of the current CDN network. The above cache strategy and traffic control strategy are examples, and the present application does not make specific limitations thereon.

[0032] In one preferred embodiment of the present application, in order to find the CDN network node closest to the current CDN network node, after obtaining the routing link table of the current CDN network node, the central CDN network node determines the edge CDN network node closest to the current CDN network node according to the routing link table of the current CDN network node, and transmits the data of the adjacent node to the closest edge CDN network node through routing conversion according to the cache strategy, thereby realizing efficient processing of the cache data. The distance L between the edge CDN network node closest to the current CDN network node is calculated by obtaining the round-trip time RTT between the current CDN network node and any edge CDN network node, calculating the hop number Hops to the edge CDN network node, and calculating the bandwidth utilization P of the current CDN network node. The calculation formula of the distance L between the edge CDN network node closest to the current CDN network node is: where β, λ and γ are weight coefficients, respectively.

[0033] The processes described above with reference to the flowcharts can be implemented as computer software programs in accordance with embodiments of the present disclosure. Embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for performing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication section, and / or installed from a detachable medium. When the computer program is executed by a central processing unit (CPU), the above-described functions defined in the methods of the present application are performed. It should be noted that the computer readable medium of the present application can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but not limited to, be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus or device. In the present application, the computer readable signal medium can include a data signal carried in a baseband or as part of a carrier wave, in which the computer readable program code is carried. Such a propagated data signal can take many forms, including but not limited to, an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium that can send, propagate or transfer a program for use by or in connection with an instruction execution system, apparatus or device. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to, wireless, wire, optical cable, RF, or any suitable combination of the above.

[0034] The computer program product of the present application can be a computer program product comprising a computer-readable medium bearing computer program code embodied therein for use with a computer. The computer program code can be code defining and / or implementing the present application. The computer program code can be organized into one or more computer program components, including but not limited to, program components for implementing the features of the present application. The one or more computer program components can each include the code implementing the features of the present application. The one or more computer program components can each be embodied on the computer-readable medium. The computer-readable medium can be a tangible computer-readable medium. In operation, the computer program code is provided on the computer-readable medium to a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the code can be executed to implement the present application.

[0035] It will be appreciated by persons skilled in the art that the present application is not limited to what has been described above and illustrated in the drawings. The application is defined by the claims and many modifications and variations of the illustrative embodiments can be made by those skilled in the art without departing from the scope of the present application, which is defined by the following claims.

Claims

1. A distributed storage method based on AI traffic recognition, characterized in that, The method includes: Obtain the current CDN network node status information, simulate and construct the first GNN node in the GNN model based on the current CDN network node status information, and construct the adjacent nodes of the GNN model based on the communication relationship between the current CDN network node and different CDN network nodes. The communication traffic information between the first GNN node and its corresponding neighboring nodes is obtained, and the traffic feature data in the communication traffic information is identified using an AI model. A dynamic adjacency matrix between the first GNN node and its neighboring nodes is constructed based on the traffic feature data. The dynamic adjacency matrix is ​​used to update the state of the first GNN node and its neighboring nodes. Based on the updated state of the first GNN node and its neighboring nodes, the corresponding CDN network node's caching strategy or traffic processing strategy is executed. Obtain the current CDN network node routing link table, and route the current CDN network node to the optimal CDN network node according to the caching strategy or traffic processing strategy; The current CDN network node status information includes hardware resource status, network resource status, and cache status. The hardware resource status includes CPU utilization, memory utilization, and cache disk read / write performance data. The network resource status includes the current CDN network node's uplink / downlink bandwidth utilization, packet loss rate, network request rate, concurrent connections, and number of URLs. The cache status includes cache hit rate, cache eviction rate, cache warm-up status, and cache content type. The hardware resource status, network resource status, and cache status are preprocessed to obtain the feature matrix of the first GNN node. The dynamic adjacency matrix is ​​then constructed based on the traffic feature data between the first GNN node and its corresponding neighboring nodes identified by the AI ​​model. The method for constructing the dynamic adjacency matrix includes: using an image recognition AI model, a text recognition AI model, and a speech recognition AI model to obtain network traffic feature data between the first GNN node and its corresponding neighboring nodes; wherein the traffic feature data includes: illegal video frame data, illegal voice data, illegal text data, and their quantities; preprocessing the different traffic feature data to convert them into the dynamic adjacency matrix, and updating the state of the first GNN node and its corresponding neighboring nodes according to the dynamic adjacency matrix; The state update method for the first GNN node and its corresponding neighboring nodes includes: constructing multi-head attention coefficients based on traffic feature data obtained from traffic identification by the AI ​​model. Where i represents the current first GNN node, j represents the neighboring node that has a communication relationship with the first GNN node, and h i The first GNN node features are represented by α and W, where W is the weight matrix, α is the action parameter between nodes based on traffic feature data transformation, T represents the transpose matrix identifier, LeakyReLU represents the activation function, and Softmax represents the linear classification function. The attention coefficient α... ij This represents the similarity feature weights between the first GNN node and its corresponding neighboring nodes.

2. The distributed storage method based on AI traffic identification according to claim 1, characterized in that, The state update method for the first GNN node and its corresponding neighboring nodes further includes: updating the multi-head attention coefficient α. ij Parallel computation is performed to capture the relationship type between the first GNN node i and its neighboring node j, using the multi-head attention coefficient α. ij The calculation of the updated content of the first GNN node includes: utilizing the multi-head attention coefficient α ij Construct a dynamic adjacency matrix A dynamic [i, j] k = Where k represents the corresponding attention head, and the dynamic adjacency matrix A is... dynamic [i, j] k Replace the original static adjacency matrix A static After replacement, the first GNN node is updated according to the following formula. ,in This represents the updated predicted state of the first GNN node, where N represents the node relationship function, K represents the total number of attention heads, and σ is the activation function. Based on the updated predicted state of the first GNN node... Execute the node caching policy or traffic processing policy corresponding to the current CDN network.

3. The distributed storage method based on AI traffic identification according to claim 2, characterized in that, The caching strategy or traffic processing strategy mentioned above includes: based on the updated predicted state of the first GNN node. The cache state of the current CDN network node update prediction allows the input of data from the corresponding neighboring nodes into the cache; based on the updated predicted first GNN node state... The system processes the illegal data predicted from the corresponding neighboring nodes of the current CDN network using strategies including rate limiting, access blocking, and caching.

4. The distributed storage method based on AI traffic identification according to claim 1, characterized in that, CDN network nodes include central CDN network nodes and edge CDN network nodes. After obtaining the routing link table of the current CDN network node, the central CDN network node determines the nearest edge CDN network node based on the routing link table of the current CDN network node, and transmits the data of the adjacent node to the nearest edge CDN network node through routing transformation according to the caching policy, thereby achieving efficient processing of cached data.

5. A distributed storage method based on AI traffic identification according to claim 4, characterized in that, The method for calculating the distance L between the current CDN network node and the nearest edge CDN network node includes: obtaining the round-trip time (RTT) from the current CDN network node to any edge CDN network node, calculating the hop count (Hops) to the edge CDN network node, and calculating the bandwidth utilization (P) of the current CDN network node. The formula for calculating the distance L between the current CDN network node and the nearest edge CDN network node is: , where β, λ and γ are the weighting coefficients, respectively.

6. A distributed storage system based on AI traffic recognition, characterized in that, The system executes a distributed storage method based on AI traffic identification as described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is executed by a processor to implement a distributed storage method based on AI traffic identification as described in any one of claims 1-5.

Images