Iot edge server service caching method and apparatus

Abstract

The application provides an Internet of Things edge server service caching method and device, the method comprising: obtaining caching information of each edge server corresponding to each user equipment in the Internet of Things, obtaining link states between each edge server, taking the caching information of each edge server as each node and taking each link state as each edge to construct a topology relationship graph; obtaining first task information of a current time step of each user equipment and second task information of at least part of time steps before the current time step, generating a task sequence based on the first task information and the second task information, inputting the topology relationship graph and the task sequence into a trained service caching prediction model to predict third task information of a next time step of each user equipment; and determining a service caching strategy based on the third task information of each user equipment and the caching information of each edge server. The method improves service caching efficiency and system performance.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) device technology, and in particular to a method and apparatus for caching services on an IoT edge server. Background Technology

[0002] The Internet of Things (IoT) is a network architecture that connects various devices via the internet to enable data collection and intelligent management. In recent years, with advancements in sensor and communication technologies, IoT has rapidly gained popularity in areas such as smart homes, smart cities, and industrial automation. In industrial automation, real-time data acquisition and control using sensors and communication technologies can monitor equipment status, promptly detect faults, and thus ensure the coordinated operation of production line equipment, reducing downtime. This approach can improve system response speed, enhance operational safety, and optimize resource allocation, ultimately leading to greater economic benefits and a better user experience.

[0003] With the expansion of industrial automation, the number of sensors and other devices in production environments is increasing, leading to a surge in real-time data acquisition and decision-making tasks. In this context, cloud computing is becoming increasingly inadequate. First, cloud computing often suffers from high latency; data transmission from sensors to the cloud for processing and then feedback to the devices may not meet real-time response requirements. Second, the proliferation of devices puts immense pressure on network bandwidth, and frequent data uploads can create bandwidth bottlenecks, increasing data transmission costs. In this context, edge computing is emerging as a crucial technology for addressing the real-time processing needs of massive amounts of data. Edge computing offloads some computation to edge servers, processing data near the data source and reducing data transmission distance, thus significantly reducing response time. Simultaneously, data processing and analysis at the edge reduces the amount of data that needs to be sent to the cloud, saving bandwidth and transmission costs. The combination of edge computing and cloud computing supports a distributed architecture, better addressing the needs of devices and data in different geographical locations. Therefore, by offloading computational tasks from cloud servers to network edge devices, edge computing reduces data transmission latency, improves real-time performance and efficiency, and is particularly suitable for tasks with high real-time response requirements.

[0004] Currently, many scenarios (such as autonomous driving, industrial automation, and smart healthcare) have increasingly higher requirements for real-time response. Meanwhile, edge nodes typically have limited computing and storage resources; if resource utilization is low, it will reduce overall system performance and increase operating costs. In recent years, service caching technology has become a research hotspot to minimize task offloading completion time and maximize edge node resource utilization. Service caching technology is an indispensable module for high-performance scenarios, and an excellent service caching technology can significantly improve the efficiency of the entire system. Service caching technology can alleviate some of the pressure on task offloading strategies by pre-unloading computing services, reducing the total latency of task completion; it can also avoid duplicate uploads of cached services by continuously optimizing cache hits, effectively reducing network congestion and the total system cost.

[0005] While existing service caching methods can achieve service caching, they suffer from a number of problems. For example, in publicly available methods for deploying voice command tasks in IoT smart homes, voice commands are first classified using federated learning. Then, to minimize the computational load during command inference, LSTM is used to predict the voice commands, and the corresponding tasks are proactively stored at the edge device. This method uses a neural network for prediction on a single node and pre-caches the computational services of the prediction tasks; however, when there are multiple user devices, this method becomes significantly more complex and imposes additional CPU and storage burdens on the system, leading to reduced system performance.

[0006] In another technical solution, addressing the shortcomings of the multi-region collaborative caching method mentioned above, a novel collaborative caching algorithm is proposed. This caching method calculates the popularity of tasks based on their occurrence frequency, and the popularity is updated as tasks arrive. The system caches the most popular task each time, thus achieving service caching based on this method. Although this service caching method considers the collaborative caching of computing services by nodes within a certain region, it lacks flexibility, cannot utilize the temporal patterns of task arrival, and suffers from low service caching efficiency.

[0007] In summary, existing service caching methods suffer from low service caching efficiency and low system performance; therefore, improving service caching efficiency and system performance is an urgent technical problem to be solved. Summary of the Invention

[0008] In view of this, embodiments of the present invention provide an IoT edge server service caching method and apparatus to eliminate or improve one or more defects existing in the prior art.

[0009] One aspect of the present invention provides a method for caching services on an Internet of Things (IoT) edge server, the method comprising:

[0010] Obtain the cache information of each edge server corresponding to each user device in the Internet of Things, obtain the link status between each edge server, and construct a topology graph with the cache information of each edge server as each node and the link status as each edge.

[0011] Obtain the first task information of each user device at the current time step and the second task information of at least some time steps before the current time step. Generate a task sequence based on the first task information and the second task information. Input the topology graph and the task sequence into the trained service cache prediction model to predict the third task information of each user device at the next time step.

[0012] The service caching strategy is determined based on the third task information corresponding to each user device and the cache information of each edge server.

[0013] In some embodiments of the present invention, a service caching strategy is determined based on the third task information corresponding to each user equipment and the caching information of each edge server, including:

[0014] Based on the third task information corresponding to each user device and the cache information of each edge server, it is determined whether the edge server caches a service corresponding to the third task information.

[0015] If the service corresponding to the third task information is not cached in the edge server, the service corresponding to the third task information will be cached in the corresponding edge server.

[0016] In some embodiments of the present invention, when the service corresponding to the third task information is not cached in the edge server, the service corresponding to the third task information is cached in the corresponding edge server, including:

[0017] If no service corresponding to the third task information is cached in the edge server, determine whether the total number of services cached in the edge server is less than the threshold.

[0018] When the threshold is not less than 1, identify and delete the infrequently used services that are cached in the edge server.

[0019] In some embodiments of the present invention, infrequently used services cached in the edge server are identified and deleted based on the LRU algorithm.

[0020] In some embodiments of the present invention, the method further includes:

[0021] Construct a sample dataset and a loss function. The sample data in the sample dataset includes topological relationship graph sample data, task sequence sample data, and user equipment task information sample data.

[0022] Based on the sample dataset and loss function, the initial service cache prediction model is pre-trained to obtain a trained service cache prediction model.

[0023] In some embodiments of the present invention, the loss function is:

[0024]

[0025] Among them, v l This represents the actual task information of the user device at time step l, v l+1 This represents the actual task information of the user device at time step l+1. W represents the user device task information predicted by the model. θ This represents the model parameters, and M represents the total number of time steps.

[0026] In some embodiments of the present invention, the service cache prediction model is the ST-GCN model.

[0027] According to another aspect of the present invention, an Internet of Things edge server service caching system is also disclosed. The system includes a processor, a memory, and a computer program stored in the memory. The processor is used to execute the computer program. When the computer program is executed, the system implements the steps of the method as described in any of the above embodiments.

[0028] Accordingly, the present invention also discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method as described in any of the above embodiments.

[0029] According to another aspect of the present invention, a computer program product is also disclosed, comprising a computer program that, when executed by a processor, implements the steps of the method as described in any of the above embodiments.

[0030] The IoT edge server service caching method and apparatus disclosed in the above embodiments of the present invention use edge nodes and incoming node tasks as input information for a service caching prediction model. The service caching prediction model predicts the subsequent arrival of tasks for all nodes in the system. Based on the caching prediction model, this method can predict the task information of each user device in the next time step. Then, by pre-caching, the services corresponding to each user device are cached in the corresponding edge server. This method not only improves the efficiency of service caching and reduces task completion latency, but also enhances system performance.

[0031] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the description, or may be learned by practice of the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures specifically pointed out in the description and drawings.

[0032] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description

[0033] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, are not intended to limit the scope of the invention. The components in the drawings are not drawn to scale but are merely illustrative of the principles of the invention. For ease of illustration and description of certain parts of the invention, corresponding portions in the drawings may be enlarged, i.e., may appear larger relative to other components in an exemplary device actually manufactured according to the invention. In the drawings:

[0034] Figure 1 This is a flowchart illustrating an embodiment of an IoT edge server service caching method according to this application.

[0035] Figure 2 This is a flowchart illustrating an IoT edge server service caching method according to another embodiment of this application. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0037] It should also be noted that, in order to avoid obscuring the invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.

[0038] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.

[0039] It should also be noted that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection with an intermediary, and can refer not only to a wired connection, but also to a wireless connection. The specific meaning can be changed based on the actual application scenario.

[0040] In this application, we consider an IoT edge network consisting of n user devices, m edge devices, and a cloud server. The set of user devices is denoted by N = {1, 2, 3, ..., n}. These devices have relatively weak processing capabilities and can only generate and transmit data and tasks. The number of edge servers is equal to the number of user devices, and the edge servers are located close to their respective user devices, resulting in low communication latency. The cloud server is denoted by C, possessing powerful processing capabilities and what is considered unlimited storage capacity, caching all services and executing all tasks. However, the cloud server is located far from the user devices, meaning that communication between the user devices and the cloud server will incur significant latency. Furthermore, user devices must communicate with the cloud server through the edge servers.

[0041] Furthermore, there are q types of tasks in this scheme, generated by user equipment. Due to resource limitations and low computing power, user equipment cannot process these tasks and needs to offload them to edge devices for processing. To improve task processing speed, edge servers can use caching services to handle tasks. If the edge server has the corresponding computing service for a task, when the user equipment offloads the task to that edge device, it can only transmit the data required by the task, without transmitting the computing service required by the task. The task can only be executed on the computing node that caches the corresponding service. In this application, a binary variable δ can be used. k σ k Indicates the deployment of the service. Let q represent the type of service k; This indicates that service k is deployed on node m; otherwise, it equals 0. Furthermore, the storage and computing resources of edge servers are also limited. Assume that each edge server m has processing resources (CPU cycles) of C(m) and storage size of S. m Meanwhile, when task v is offloaded to server m for execution, it consumes L per unit of time. v CPU cycles, simultaneously occupy Storage; when service k is cached on server m, it occupies storage. Storage.

[0042] In the following description, embodiments of the invention will be illustrated with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar parts, or the same or similar steps.

[0043] Figure 1 This is a flowchart illustrating an embodiment of an IoT edge server service caching method according to this application, as follows: Figure 1 As shown, the method includes at least steps S10 to S30.

[0044] Step S10: Obtain the cache information of each edge server corresponding to each user device in the Internet of Things, obtain the link status between each edge server, and construct a topology graph with the cache information of each edge server as each node and the link status as each edge.

[0045] In this step, the edge device topology is viewed as a topological graph. Therefore, this application, over M time steps, can be considered a dataset composed of M topological graphs. For example, let G represent this spatiotemporal graph. The graph at time step t can be mapped to a directed or undirected graph, which can be defined as G. t =(V t (E, W); where V t Let E be the set of points in the topology graph, E be the set of edges, and W be the adjacency matrix. In this topology graph, the cached information of each edge server serves as each node, and the links between edge servers determine each edge. In this embodiment, the number of edge servers is equal to the number of user devices in the Internet of Things, and the relationships between nodes and edges in the topology graph remain unchanged at different time steps, as does the adjacency matrix W. However, the characteristics of the nodes change over time, and these characteristics refer to the cached information of the nodes at the corresponding time steps.

[0046] Step S20: Obtain the first task information of each user device at the current time step and the second task information of at least some time steps before the current time step; generate a task sequence based on the first task information and the second task information; input the topology graph and the task sequence into the trained service cache prediction model to predict the third task information of each user device at the next time step.

[0047] After obtaining the topology graph in step S10, the topology graph is further input into the trained service cache prediction model. In addition to the topology graph, the input information of the service cache prediction model also includes a task sequence consisting of task information generated by the user device at the current time step and task information generated at least some time steps between the current time step. The output information of the service cache prediction model is the predicted task information to be generated by the user device in the next time step. It can be understood that "task information generated at least some time steps between the current time step" can be task information generated in all time steps before the current time step, or task information generated in some of the time steps, specifically determined based on the length of the task sequence.

[0048] For example, the service cache prediction model can specifically be the ST-GCN model, a deep learning model specifically designed for processing spatiotemporal data, particularly graph-structured data such as traffic networks and social networks. The ST-GCN model combines the characteristics of Graph Convolutional Networks (GCN) and Temporal Convolutional Networks (TCN), enabling it to simultaneously capture the spatial and temporal dependencies of data. The ST-GCN model structure includes two spatiotemporal graph convolutional blocks (ST-Conv Blocks) and an output fully connected layer, where each spatiotemporal graph convolutional block consists of two temporally gated convolutions and one spatially gated convolution. In this embodiment, the input information of the ST-GCN model is the feature vectors of the graph corresponding to M time steps. The adjacency matrix W∈R corresponding to the graph n×n Where M is the time step, n is the number of nodes (the number of edge servers), and c i The number of features per channel; in the ST-GCN model, the input information passes through two spatiotemporal graph convolutional blocks and one fully connected layer, and the output... The output is the feature of each node corresponding to the predicted M+1 time step. It can be understood that in the topology graph, the feature of the node represents the cache information of the edge server.

[0049] In some embodiments of the present invention, in order to obtain a trained service cache prediction model, the IoT edge server service cache method further includes the following steps: constructing a sample dataset and a loss function, wherein the sample data in the sample dataset includes topology graph sample data, task sequence sample data, and user equipment task information sample data; and pre-training the initial service cache prediction model based on the sample dataset and the loss function to obtain a trained service cache prediction model.

[0050] For example, when using the ST-GCN model as the service cache prediction model, the ST-GCN model is trained to accurately predict task arrival patterns in the scenario of this invention. As shown in Algorithm 1, the input data in the ST-GCN training algorithm is the training data. The output data is the model parameters θ. During training, the parameters of the neural network and the value of the counter are first initialized (line 1 of Algorithm 1). Then, a batch of data is taken from the training data, and the predicted values ​​of this batch of data are predicted using ST-GCN. The loss function is then calculated, and the gradient is used to calculate and update the neural network parameters, while the counter is incremented (lines 3 to 7 of Algorithm 1). All steps are repeated to update the parameters until the counter reaches the set value. In Algorithm 1, the training data is the sample data, N represents the number of sample data in the sample dataset, and X... (i) This can be understood as data including topological relationship diagram sample data and task sequence sample data, while Y (i)This can be understood as sample data of user device task information.

[0051]

[0052] Furthermore, the service cache prediction model uses L2 loss as the loss function. Specifically, the model loss function can be:

[0053]

[0054] Among them, v l This represents the actual task information of the user device at time step l, v l+1 This represents the actual task information of the user device at time step l+1. W represents the user device task information predicted by the model. θ This represents the trainable model parameters, and M represents the total number of time steps.

[0055] Step S30: Determine the service caching strategy based on the third task information corresponding to each user device and the caching information of each edge server.

[0056] In this step, the caching strategy of the edge server is further determined based on the third task information corresponding to each user device predicted by the service cache prediction model.

[0057] In one embodiment, determining a service caching strategy based on the third task information corresponding to each user device and the cache information of each edge server includes the following steps: determining whether the edge server caches a service corresponding to the third task information based on the third task information corresponding to each user device and the cache information of each edge server; if the edge server does not cache a service corresponding to the third task information, the service corresponding to the third task information is cached on the corresponding edge server. In this embodiment, the predicted third task information corresponding to each user device is compared with the cache information of each edge server corresponding to the current time step to determine whether the edge server caches a service corresponding to the task generated by the user device in the next time step; if the edge server caches a service corresponding to the task generated by the user device in the next time step, then the edge server does not need to re-cache the predicted service; if the edge server does not cache a service corresponding to the task generated by the user device in the next time step, then the predicted service needs to be cached on the corresponding edge server.

[0058] Furthermore, when no service corresponding to the third task information is cached in the edge server, the service corresponding to the third task information is cached in the corresponding edge server. This includes: when no service corresponding to the third task information is cached in the edge server, determining whether the total number of services cached in the edge server is less than a threshold; if it is not less than the threshold, identifying and deleting the cached infrequently used services in the edge server. In this embodiment, for an edge server that needs to cache services, it is determined whether it has cache space. If cache space exists, the corresponding service can be directly cached in the edge server. If no cache space exists in the cache server, cache space can be provided for the services that need to be cached by deleting infrequently used services. Optionally, when deleting infrequently used services in the edge server, the LRU algorithm can be used to identify and delete the cached infrequently used services in the edge server. The LRU algorithm improves the cache hit rate by retaining the most recently used data and eliminating the least recently used data.

[0059] Figure 2 This is a flowchart illustrating an IoT edge server service caching method according to another embodiment of this application, as shown below. Figure 2 As shown, this embodiment predicts the third task information of each user device in the next time step based on a trained ST-GCN model, and performs service caching based on the prediction results of the ST-GCN model. Specifically, this embodiment includes the following steps: S01: When each time step arrives, if the input sequence length is l, firstly, the task information D of the current time step l is stored... t The task information from the previous l-1 time steps is merged into the input sequence d. t Step S02: d t Input into ST-GCN to obtain the predicted value of the task data at time step l+1; Step S03: Based on the predicted value of node i at time step l+1, task r i And the cache status of the node at time step l, c i Perform computation service caching. If the current node already has task r cached... i If the current node does not have a computing service for the task and its cache is full, the least frequently used computing service is replaced using the LRU rule; if the cache is not full, the task is directly assigned to the next node. i The computing service is added to the cache.

[0060] For example, Algorithm 2 is a computation service caching algorithm based on ST-GCN. In Algorithm 2, the input data are model M (specifically, a pre-trained ST-GCN model), time series data D, and node caching information {c1, c2, ..., c...}. k}, node cache size C, current time step l; output data is the updated node cache information {c1, c2, ..., c k It is understandable that the input time series data D in this algorithm is derived from the task information D at the current time step l. t The node cache information {c1, c2, ..., c1} is obtained by merging the task information from the previous l-1 time steps with the task information from the previous l-1 time steps. k} represents the cache information of the edge server corresponding to the current time step, and the node cache size C is the cache space C(m) of each edge server. The second line in Algorithm 2 represents the construction of the input sequence d. t The third line indicates that d t Inputting the data into ST-GCN yields the predicted values ​​of the task data at time step l+1; lines 4 to 6 indicate whether the current node has already cached task r. i The computation service for the current task is to process the next node; lines 8 and 9 indicate that the current node does not have a computation service for this task and the cache is full, so the least frequently used computation service is replaced using the LRU rule; line 10 indicates that the cache is not full, and the task r is to be processed. i The computing service is added to the cache.

[0061]

[0062] In the IoT edge server service caching method described in the above embodiment, edge nodes and incoming node tasks are abstracted and used as input to ST-GCN. Then, ST-GCN is used to derive the prediction results for all subsequent incoming tasks of all nodes. In this way, a neural network model can predict the task status of all nodes, and by pre-caching the computational services of the predicted tasks, task completion latency is reduced, thus improving the overall system performance. Furthermore, this service caching method, by coordinating cloud servers and edge servers, not only effectively reduces the load on cloud servers but also lowers network latency.

[0063] Accordingly, the present invention also provides an IoT edge server service caching system, the system including a processor, a memory and a computer program stored in the memory, the processor being used to execute the computer program, and when the computer program is executed, the system implementing the steps of the method as described in any of the above embodiments.

[0064] As can be seen from the above embodiments, this application combines edge computing and cloud computing, introduces computing service caching under the premise of limited node resources, and the multiple edge servers of this application can predict the tasks of all nodes in the system using only one neural network model, saving CPU and storage resources of nodes in the system. At the same time, the corresponding computing services can be cached in advance according to the prediction results, reducing the delay of task completion.

[0065] In summary, the IoT edge server service caching method of this application, combining the characteristics of edge computing and cloud computing scenarios, considers caching the computing services for real-time data acquisition and decision-making tasks in a network structure with three types of devices: IoT devices, edge servers, and cloud servers. This application does not simply divide tasks into edge and cloud deployments, but rather divides the task itself into two parts: computing services and corresponding data. It reduces task completion latency by pre-caching computing services, making subsequent task unloading more flexible and better suited to actual production environments. Specifically, the ST-GCN-based service caching method of this application abstracts the node topology and arriving tasks into a spatiotemporal graph during prediction. A single ST-GCN model can predict the arrival of tasks for all nodes, overcoming the deficiency of current research that requires deploying a neural network for each node for prediction.

[0066] This invention also provides a computer-readable storage medium and a computer program product having a computer program stored thereon, which, when executed by a processor, performs the steps of the method described in any of the above embodiments. The computer-readable storage medium may be a tangible storage medium, such as random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, floppy disk, hard disk, removable storage disk, CD-ROM, or any other form of storage medium known in the art.

[0067] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the desired tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave.

[0068] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0069] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0070] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. For those skilled in the art, various modifications and variations can be made to the embodiments of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for caching services on an IoT edge server, characterized in that, The method includes: Obtain the cache information of each edge server corresponding to each user device in the Internet of Things, obtain the link status between each edge server, and construct a topology graph with the cache information of each edge server as each node and the link status as each edge. Obtain first task information of each user device at the current time step and second task information of at least some time steps before the current time step, and generate a task sequence based on the first task information and the second task information. A sample dataset and a loss function are constructed. The sample data in the sample dataset includes topology graph sample data, task sequence sample data, and user device task information sample data. Based on the sample dataset and the loss function, the initial service cache prediction model is pre-trained to obtain a trained service cache prediction model. The topology graph and the task sequence are input into the trained service cache prediction model to predict the third task information of each user device at the next time step. The service caching strategy is determined based on the third task information corresponding to each user device and the cache information of each edge server.

2. The IoT edge server service caching method according to claim 1, characterized in that, The service caching strategy is determined based on the third task information corresponding to each user device and the cache information of each edge server, including: Based on the third task information corresponding to each user device and the cache information of each edge server, it is determined whether the edge server caches a service corresponding to the third task information. If the service corresponding to the third task information is not cached in the edge server, the service corresponding to the third task information will be cached in the corresponding edge server.

3. The IoT edge server service caching method according to claim 2, characterized in that, When the service corresponding to the third task information is not cached in the edge server, the service corresponding to the third task information is cached in the corresponding edge server, including: If no service corresponding to the third task information is cached in the edge server, determine whether the total number of services cached in the edge server is less than the threshold. When the threshold is not less than 1, identify and delete the infrequently used services that are cached in the edge server.

4. The IoT edge server service caching method according to claim 3, characterized in that, The LRU algorithm is used to identify and delete infrequently used services that have been cached in the edge server.

5. The IoT edge server service caching method according to claim 1, characterized in that, The loss function is: in, This represents the actual task information of the user device at time step l. This represents the actual task information of the user device at time step l+1. This represents the user device task information predicted by the model. This represents the model parameters, and M represents the total number of time steps of the input information.

6. The IoT edge server service caching method according to any one of claims 1 to 5, characterized in that, The service cache prediction model is the ST-GCN model.

7. An IoT edge server service caching system, the system comprising a processor, a memory, and a computer program stored in the memory, characterized in that, The processor is configured to execute the computer program, and when the computer program is executed, the system implements the steps of the method as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method as described in any one of claims 1 to 6.

9. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 6.

Images