An edge server, an edge computing system and a communication method thereof

Abstract

The application discloses an edge server, an edge computing system and a communication method thereof. The method comprises the following steps: periodically acquiring, by the edge server, communication state information of each edge computing node in a topological network structure and edge computing nodes within a preset distance range around the edge computing node, inputting the acquired communication state information into a pre-established node connection decision model to obtain a real-time decision result, dynamically adjusting, based on the real-time decision result, a connection relationship of each edge computing node in the topological network structure and edge computing nodes within the preset distance range around the edge computing node to form an adaptive topological network structure, and performing data transmission by the edge server based on the adaptive topological network structure and each edge computing node. The topological network structure can be dynamically adjusted, and data can be stably and efficiently transmitted in a dynamically changing network environment.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to an edge server, an edge computing system, and a communication method thereof. Background Technology

[0002] Edge computing refers to an open platform that integrates network, computing, storage, and application capabilities, located close to the source of objects or data, to provide services at the nearest edge. With the rapid development of the Internet of Things (IoT) and mobile internet, mobile computing has shifted from centralized mobile cloud computing to mobile edge computing. Edge computing is increasingly widely used in mobile communications. Breakthroughs in edge computing technology mean that much control will be implemented through local devices without being handed over to the cloud. Processing will be completed at the local edge computing layer, greatly improving processing efficiency, reducing latency, increasing response speed, and alleviating the load on the cloud.

[0003] Currently, traditional edge computing communication methods use a fixed topology network structure to achieve information transmission between edge computing nodes. Summary of the Invention

[0004] In actual communication, the communication status information between edge computing nodes is constantly changing, and therefore the network environment is also dynamically changing. Traditional edge computing systems suffer from problems such as fixed network topology, inability to adapt to dynamically changing network environments, and low communication efficiency.

[0005] In view of the above problems, the present invention is proposed to provide an edge computing server, an edge computing system and a communication method thereof that overcomes or at least partially solves the above problems.

[0006] In a first aspect, embodiments of the present invention provide a communication method for an edge computing system, the edge computing system including at least one edge server and multiple edge computing nodes, the method comprising:

[0007] The edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it. The acquired communication status information is then input into a pre-established node connection decision model to obtain real-time decision results. Based on the real-time decision results, the connection relationship between each edge computing node in the topology network structure and the edge computing nodes within the preset distance range around it is dynamically adjusted in real time to form an adaptive topology network structure. The real-time decision results include the edge computing nodes that need to be connected and the edge computing nodes that need to be disconnected in the topology network structure.

[0008] The edge server transmits data with each edge computing node based on an adaptive topology network structure.

[0009] In some optional embodiments, before the edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, an initial topology network structure is formed, including:

[0010] The edge server selects an edge computing node as the topology starting point based on user choice, and obtains the communication status information of the topology starting point and other edge computing nodes within a preset distance range. The obtained communication status information is input into a pre-established node connection decision model to obtain the decision result. Based on the decision result, the topology connection between the topology starting point and the edge computing nodes within the preset distance range is established to form the initial topology network structure.

[0011] In some optional embodiments, the edge server periodically acquires communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, including:

[0012] Each edge computing node in the topology network periodically collects communication status information of the surrounding edge computing nodes within a preset distance range, and sends its own communication status information and the collected communication status information to the edge server.

[0013] The edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it.

[0014] In some optional embodiments, the acquired communication status information is input into a pre-established node connection decision model to obtain real-time decision results. Based on the real-time decision results, the connection relationships between each edge computing node and its surrounding edge computing nodes in the topology network structure are dynamically adjusted in real time to form an adaptive topology network structure, including:

[0015] The edge server inputs the periodically acquired communication status information into the node connection decision model and outputs the edge computing nodes that each edge computing node in the topology network structure needs to connect to and the edge computing nodes that need to be disconnected.

[0016] The edge server dynamically adjusts the connection between each edge computing node and the edge computing nodes that need to be connected in the real-time topology network structure, and also dynamically adjusts the disconnection between each edge computing node and the edge computing nodes that need to be disconnected in the real-time topology network structure, forming an adaptive topology network structure.

[0017] The edge computing nodes that need to be connected and the edge computing nodes that need to be disconnected are determined from the edge computing nodes within a preset distance range around each edge computing node.

[0018] In some alternative embodiments, the communication status information includes at least one of signal strength, data transmission rate, packet loss rate, latency, power consumption, mobility speed, and historical connection quality.

[0019] In some optional embodiments, the edge server inputs the periodically acquired communication status information into the node connection decision model, including:

[0020] Extract decision features from the acquired communication status information; the decision features include at least one of the following: standard deviation of signal strength, maximum signal strength, minimum signal strength, average signal strength, standard deviation of data transmission rate, maximum signal transmission rate, minimum data transmission rate, average data transmission rate, standard deviation of packet loss rate, maximum packet loss rate, minimum packet loss rate, standard deviation of delay, standard deviation of energy consumption, average energy consumption, standard deviation of mobile speed, maximum mobile speed, historical connection quality average over the past time period, and trend features of historical data;

[0021] The extracted decision features are input into the node-connected decision model.

[0022] In some optional embodiments, a node connection decision model is established, including:

[0023] The sample data included in the training set is input into the machine learning model for iterative training until the loss function of the machine learning model meets the requirements, and the trained machine learning model is used as a node to connect the decision model.

[0024] The sample data includes the historical communication status information of each edge computing node and its surrounding edge computing nodes within a preset distance range in a known topology network structure, as well as the historical connection relationship between each edge computing node and its surrounding edge computing nodes within a preset distance range under different historical communication states.

[0025] In some optional embodiments, the edge server transmits data with each edge computing node based on an adaptive topology network structure, including:

[0026] The edge server receives data sent by each edge computing node in the adaptive topology network structure and processes the data;

[0027] The processed data is then fed back to the corresponding edge computing nodes.

[0028] In some optional embodiments, preprocessing is performed on the data before processing, including:

[0029] The edge server preprocesses the received data, including at least one of the following: data cleaning, data integration, data transformation, data reduction, data discretization, data encoding and transformation, and data quality checking, to obtain preprocessed data.

[0030] In some optional embodiments, the communication method of the above-described edge computing system further includes:

[0031] Edge servers are used to periodically optimize the adaptive network topology based on network conditions.

[0032] In some optional embodiments, the edge server optimizes the adaptive topology network structure based on network conditions, including:

[0033] The edge server collects the status information of each edge computing node in the current adaptive topology network structure. The status information includes the connection status and communication quality parameters of each edge computing node.

[0034] Based on state information and optimization algorithms, the current adaptive topology network structure is evaluated and optimized; the optimization algorithms include genetic algorithms and particle swarm optimization algorithms.

[0035] The edge server sends the optimized adaptive topology network structure to each edge computing node, guiding the edge computing nodes to adjust the adaptive topology network structure.

[0036] In a second aspect, embodiments of the present invention provide an edge computing server, comprising:

[0037] The decision generation module is used to periodically acquire the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, and input the acquired communication status information into the pre-established node connection decision model to obtain real-time decision results.

[0038] The topology network generation module is used to dynamically adjust the connection relationship between each edge computing node and the surrounding edge computing nodes within a preset distance range in the topology network structure based on real-time decision results, so as to form an adaptive topology network structure.

[0039] The communication module is used for data transmission with each edge computing node based on an adaptive topology network structure.

[0040] Thirdly, embodiments of the present invention provide an edge computing system, including: an edge computing node and an edge computing server;

[0041] Edge computing nodes are used to periodically send their own communication status information and collect the communication status information of terminal nodes within a preset distance range.

[0042] The edge computing server is used to periodically acquire the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it. The acquired communication status information is input into a pre-established node connection decision model to obtain real-time decision results. Based on the real-time decision results, the connection relationship between each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it is dynamically adjusted in real time to form an adaptive topology network structure. Data transmission is also carried out between the edge computing server and each edge computing node based on the adaptive topology network structure.

[0043] This invention provides a computer storage medium storing computer-executable instructions, which, when executed by a processor, implement the communication method of the aforementioned edge computing system.

[0044] This invention provides an edge computing device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the communication method of the edge computing system described above.

[0045] The beneficial effects of the above-described technical solutions provided in the embodiments of the present invention include at least the following:

[0046] This invention utilizes an edge server to periodically acquire communication status information of each edge computing node in a topological network structure and its surrounding edge computing nodes within a preset distance range. This acquired communication status information is input into a pre-established node connection decision model to obtain real-time decision results. Based on these results, the connection relationships between each edge computing node and its surrounding edge computing nodes within the preset distance range are dynamically adjusted in real time, forming an adaptive topological network structure. The edge server then transmits data with each edge computing node based on this adaptive topological network structure. Compared to existing edge computing systems that use fixed topological network structures for communication, this method can acquire the communication status information of edge computing nodes in the topological network structure in real time, predict the connection relationships of each edge computing node in the next cycle, and adjust the connections between edge computing nodes in real time to form an adaptive network topology. By dynamically adjusting the topological network structure, efficient and stable communication is achieved in complex and ever-changing network environments. Furthermore, the optimization of the topological network structure using the edge server further improves the overall network communication performance. This invention is applicable to large-scale edge computing environments and has high practical value.

[0047] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0048] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0049] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0050] Figure 1 This is a flowchart of the communication method of the edge computing system in Embodiment 1 of the present invention;

[0051] Figure 2 This is a schematic diagram of the edge server structure in Embodiment 1 of the present invention. Detailed Implementation

[0052] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0053] Traditional edge computing systems suffer from problems such as fixed topology networks, inability to adapt to dynamically changing environments, and low communication efficiency. To address these issues, this invention provides a communication method for edge computing systems that overcomes the low efficiency caused by using fixed topology networks in existing edge computing systems.

[0054] Example 1

[0055] Embodiment 1 of the present invention provides a communication method for an edge computing system. The edge computing system includes at least one edge server and multiple edge computing nodes. Wireless or wired communication is possible between edge computing nodes and between edge computing nodes and the edge server. This edge computing system can dynamically adjust the connection relationships between edge computing nodes and between edge computing nodes and the edge server to form an adaptive topology network structure. Communication is performed based on this adaptive topology network structure. The flow of the communication method of this edge computing system is as follows: Figure 1 As shown, it includes the following steps:

[0056] Step S101: The edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, and inputs the acquired communication status information into the pre-established node connection decision model to obtain real-time decision results.

[0057] Step S102: Based on the real-time decision results, dynamically adjust the connection relationship between each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it to form an adaptive topology network structure.

[0058] Step S103: The edge server transmits data with each edge computing node based on the adaptive topology network structure.

[0059] Optionally, before step S101 above, an initial topology network structure can be formed first, and then dynamically adjusted based on the initial topology network structure to form an adaptive topology network structure. The process of forming the initial topology network structure includes:

[0060] S201: The edge server determines an edge computing node as the topology starting point based on the user's selection, and obtains the communication status information of the topology starting point and other edge computing nodes within a preset distance range around it. The obtained communication status information is input into a pre-established node connection decision model to obtain the decision result.

[0061] S202: Based on the decision results, establish topological connections between the topological starting point and edge computing nodes within a preset distance range to form an initial topological network structure. After establishing the initial topological network structure, the edge server transmits data with each edge computing node based on the initial topological network structure, and then enters the subsequent process of dynamically adjusting the connection relationships between edge computing nodes and between edge computing nodes and edge servers to form an adaptive topological network structure.

[0062] In the initial edge computing system, users can select a topology starting point among the edge computing nodes. Based on the communication status information of the topology starting point and the communication status information of the edge computing nodes within a preset distance range around the topology starting point, the system determines the edge computing nodes that need to be connected to the topology starting point and the edge computing nodes that need to be disconnected from the topology starting point. The server connects the topology starting point and the edge computing nodes that need to be connected to form the initial topology network structure. For the edge computing nodes that need to be disconnected, no connection has been established between the edge computing nodes when the initial topology network structure is formed, so there is no need to operate on these edge computing nodes that need to be disconnected.

[0063] Optionally, in step S101 above, the edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, including:

[0064] Each edge computing node in the topology network periodically collects the communication status information of the surrounding edge computing nodes within a preset distance range, and sends its own communication status information and the collected communication status information to the edge server.

[0065] The edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it.

[0066] Each edge computing node in an edge computing system periodically sends its own communication status information so that surrounding edge computing nodes can obtain this information in real time. Simultaneously, each edge computing node in the topology network periodically acquires the communication status information of edge computing nodes within a preset distance range. Each edge computing node in the topology network then sends its own communication status information and the collected communication status information to the server. This allows the server to determine the next network connection adjustment for each edge computing node in the topology network. It is important to note that the edge computing nodes in the topology network and the edge computing nodes in the edge computing system may be different. Edge computing nodes in the edge computing system include those in the topology network, but the edge computing nodes in the topology network may not include all edge computing nodes in the edge computing system.

[0067] The edge server periodically acquires communication status information and inputs it into the node connection decision model, outputting real-time decision results for each edge computing node in the topology network structure. The real-time decision results include which edge computing nodes in the topology network structure need to be connected and which need to be disconnected. For example, in a topology network structure with 3 edge computing nodes, and each edge computing node having 5 edge computing nodes within a preset distance range, the edge server periodically acquires the communication status information of each edge computing node and its 5 surrounding edge computing nodes, and inputs this information into the node connection decision model to predict which edge computing nodes within the preset distance range of the 3 edge computing nodes in the topology network structure need to be connected and which need to be disconnected.

[0068] The communication status information of edge computing nodes includes at least one of the following: signal strength, data transmission rate, packet loss rate, latency, energy consumption, movement speed, and historical connection quality.

[0069] Optionally, in step S102 above, based on the real-time decision results, the connection relationships between each edge computing node and the surrounding edge computing nodes in the topology network structure are dynamically adjusted in real time to form an adaptive topology network structure, including:

[0070] Based on real-time decision results, the edge server dynamically adjusts the connection between each edge computing node in the topology network structure and the edge computing nodes that need to be connected according to the real-time decision results, and also dynamically adjusts the disconnection between each edge computing node in the topology network structure and the edge computing nodes that need to be disconnected according to the real-time decision results, forming an adaptive topology network structure.

[0071] The edge computing nodes that need to be connected and the edge computing nodes that need to be disconnected are determined from the edge computing nodes within a preset distance range around each edge computing node.

[0072] In a network topology, an edge computing node needs to connect to other edge computing nodes, including those already connected and those that need to be connected but are not currently connected. If an edge computing node in the current topology is already connected to the node it needs to connect to, no adjustments are needed for either node. If an edge computing node in the current topology is not currently connected to the node it needs to connect to, then a connection is established between the two.

[0073] In a network topology, an edge computing node that needs to be disconnected includes surrounding edge computing nodes that are currently connected but need to be disconnected, as well as surrounding edge computing nodes that are not connected to this node. For an edge computing node that needs to be disconnected, if an edge computing node in the current network topology is already connected to the node it needs to disconnect from, then the connection between the two nodes is disconnected; if an edge computing node in the current network topology is not connected to the node it needs to disconnect from, then no adjustments to either edge computing node are required.

[0074] In traditional edge computing systems, the fixed topology network does not support adjustments to the topology network structure based on complex and ever-changing network environments, thus hindering efficient and stable data transmission. This method, deployed on an edge server with a node connection decision model, can acquire in real-time the communication status information of each edge computing node in the current topology network structure and its surrounding edge computing nodes within a preset range. This communication status information is then input into the node connection decision model to predict the connection relationship of each edge computing node in the topology network structure for the next cycle. Based on the prediction results, the method guides the edge computing nodes to adjust their connections with other surrounding edge computing nodes. For example, if the node connection decision model predicts that connecting an edge computing node in the topology network structure to an edge computing node within a preset distance range will improve overall network performance, then it attempts to establish a connection; if it predicts that connecting to an edge computing node within a preset distance range will degrade performance, then it disconnects the connection.

[0075] Adjusting the connections of edge computing nodes results in a new topology network structure. This adjustment generates new communication status information, which the edge server acquires in real time. Based on the node connection decision model, it generates new decision results to guide the next cycle's topology network connection adjustments. These decision results refer to the edge computing nodes within a preset distance range surrounding each edge computing node in the current topology network structure that need to be connected to their respective corresponding edge computing nodes and those that need to be disconnected. By acquiring the communication status information of each edge computing node and its surrounding edge computing nodes in real time and generating new decision results, the topology network structure can be dynamically adjusted in real time, forming an adaptive topology network structure. This dynamic adjustment method can continuously optimize the topology network structure according to complex and ever-changing network environments to achieve efficient and stable data transmission. It is applicable to large-scale edge computing environments and has high practical value.

[0076] Optionally, the edge server inputs the periodically acquired communication status information into the node connection decision model, including:

[0077] Extract decision features from the acquired communication status information; the decision features include at least one of the following: standard deviation of signal strength, maximum signal strength, minimum signal strength, average signal strength, standard deviation of data transmission rate, maximum signal transmission rate, minimum data transmission rate, average data transmission rate, standard deviation of packet loss rate, maximum packet loss rate, minimum packet loss rate, standard deviation of delay, standard deviation of energy consumption, average energy consumption, standard deviation of mobile speed, maximum mobile speed, historical connection quality average over the past time period, and trend features of historical data;

[0078] The extracted decision features are input into the node-connected decision model.

[0079] Extracting decision features to better capture the inherent characteristics of communication state information helps machine learning models focus only on features that are highly correlated with real-time decision results during learning, thereby improving the model's learning performance and making the model more applicable.

[0080] Optionally, a node connection decision model is established, including:

[0081] The sample data included in the training set is input into the machine learning model for iterative training until the loss function of the machine learning model meets the requirements, and the trained machine learning model is used as a node to connect the decision model.

[0082] The sample data includes the historical communication status information of each edge computing node and its surrounding edge computing nodes within a preset distance range in a known topology network structure, as well as the historical connection relationship between each edge computing node and its surrounding edge computing nodes within a preset distance range under different historical communication states.

[0083] Machine learning models can be classifiers, determining whether or not edge computing nodes around each edge computing node in a topological network connect or not, or regression models, predicting the quality or utility of each edge computing node connection. The machine learning algorithms used in these models can include decision trees, support vector machines (SVM), neural networks, random forests, gradient boosting machines (GBM), deep learning, and more. The choice of algorithm depends on the specific application scenario, the characteristics of the data, and the available computing resources. In practical applications, the real-time performance of the algorithm must also be considered; that is, the model's inference speed must be fast enough to adapt to dynamically changing network environments.

[0084] Optionally, in step S103 above, the edge server transmits data with each edge computing node based on the adaptive topology network structure, including:

[0085] The edge server receives data sent by each edge computing node in the adaptive topology network structure and processes the data;

[0086] The processed data is then fed back to the corresponding edge computing nodes.

[0087] Optionally, preprocessing may be performed on the data before processing, including:

[0088] The edge server performs at least one of the following on the received data: data cleaning, data integration, data transformation, data reduction, data discretization, data encoding and transformation, and data quality inspection, to obtain preprocessed data.

[0089] Edge servers should process the received data differently based on different user needs. Specific methods for data preprocessing include:

[0090] Data cleaning is performed on the data received from the edge server. Data cleaning methods can include handling missing values, such as filling in missing values, deleting rows or columns containing missing values, or using interpolation methods; processing noisy data can involve identifying and handling outliers or exceptions in the data, and statistical methods or model-based methods can be used to process noisy data; in addition, when there are data entry errors or inconsistencies, data errors are corrected.

[0091] Data integration is performed on data received from edge servers. Data integration can include merging data, integrating data from different sources into a consistent dataset; entity identification, identifying and merging data about the same entity from different sources; and unifying the units of measurement and formats in different data sources to resolve data conflicts.

[0092] Data transformation can be performed on the data received from the edge server. Data transformation can include scaling the data to a specific range or distribution, such as [0,1] or a standard normal distribution to normalize the data; converting continuous variables into discrete categories through binning to discretize the data; and creating new features to construct data features, which may be a combination or transformation of existing features.

[0093] Data reduction of data received from edge servers can include using methods such as PCA (Principal Component Analysis) and LDA (Linear Discriminant Analysis) to reduce the dimensionality of the data; selecting the most representative or informative features from the collected data and deleting redundant or irrelevant features; and summarizing the data, such as grouping by category or time period and performing calculations and statistics to achieve data aggregation.

[0094] For data received from the edge server, the values ​​of continuous attributes of the data are converted into a series of intervals to achieve data discretization, which may be necessary for certain types of algorithms, such as rule-based algorithms.

[0095] The data received from the edge server is encoded and transformed. Category encoding is used to convert the non-numerical category features of the collected data into numerical representations, such as one-hot encoding or label encoding. Time series processing is performed on the data collected at different periods to extract time-related features and improve the accuracy of data transmission.

[0096] Perform data quality checks on the data received from the edge server to verify whether the data conforms to the expected distribution or attributes, and ensure the correctness of the data after transformation.

[0097] In summary, data preprocessing can be performed according to actual needs, selecting one or more processing methods. Edge servers can reduce the burden of data transmission by preprocessing data before transmission.

[0098] Optionally, the communication method of the above-mentioned edge computing system further includes:

[0099] Edge servers periodically optimize the adaptive network topology based on network conditions.

[0100] The edge server periodically optimizes the adaptive network topology based on network conditions, specifically including:

[0101] The edge server collects the status information of each edge computing node in the current adaptive topology network structure. The status information includes the connection status and communication quality parameters of each edge computing node.

[0102] Based on state information and optimization algorithms, the current adaptive topology network structure is evaluated and optimized; the optimization algorithms include genetic algorithms and particle swarm optimization algorithms.

[0103] The edge server sends the optimized adaptive topology network structure to each edge computing node, guiding the edge computing nodes to adjust the adaptive topology network structure.

[0104] When generating an adaptive topology network structure, the edge server can also further optimize the current topology network structure based on the current topology network status information, such as the connection status and communication quality parameters of each edge computing node, using optimization algorithms. This allows the topology network structure to be dynamically adjusted in real time to adapt to the complex and ever-changing network environment. Optimization algorithms can include genetic algorithms, particle swarm optimization algorithms, etc.

[0105] Based on the same inventive concept, embodiments of the present invention also provide an edge computing server, which can be set up in an edge computing system, and the structure of the server is as follows. Figure 2 As shown, it includes:

[0106] The decision generation module 21 is used to periodically acquire the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, and input the acquired communication status information into the pre-established node connection decision model to obtain real-time decision results.

[0107] The topology network generation module 22 is used to dynamically adjust the connection relationship between each edge computing node and the surrounding edge computing nodes within a preset distance range in the topology network structure based on real-time decision results, so as to form an adaptive topology network structure.

[0108] The communication module 23 is used to transmit data with each edge computing node based on the adaptive topology network structure.

[0109] This invention also provides a computer storage medium storing computer-executable instructions, which, when executed by a processor, enable a communication method for an edge computing system.

[0110] This invention also provides an edge computing device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the communication method of the aforementioned edge computing system.

[0111] Regarding the edge server in the above embodiments, the specific ways in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0112] The method and edge server described in this embodiment of the invention involve the edge server periodically acquiring communication status information of each edge computing node in the topological network structure and its surrounding edge computing nodes within a preset distance range. This acquired communication status information is input into a pre-established node connection decision model to obtain real-time decision results. Based on these real-time decision results, the connection relationships between each edge computing node and its surrounding edge computing nodes within the preset distance range are dynamically adjusted in real time to form an adaptive topological network structure. The edge server then transmits data with each edge computing node based on this adaptive topological network structure. Compared to existing edge computing systems that use a fixed topological network structure for communication, this method can acquire the communication status of edge computing nodes in the topological network structure in real time, predict the connection relationships of each edge computing node in the next cycle, and adjust the connections between edge computing nodes in real time to form an adaptive network topology. By dynamically adjusting the topological network structure, efficient and stable communication is achieved in complex and ever-changing network environments. Furthermore, the optimization of the topological network structure using the edge server further improves the overall network communication performance. This method is applicable to large-scale edge computing environments and has high practical value.

[0113] Example 2

[0114] Embodiment 2 of the present invention provides an edge computing system, including: an edge computing node and an edge computing server;

[0115] Edge computing nodes periodically send their own communication status information and collect the communication status information of terminal nodes within a preset distance range. Each edge computing node needs to periodically send its own communication status information. Edge computing nodes included in the topology network structure not only need to periodically send their own communication status, but also need to receive the communication status of edge computing nodes within a preset distance range to determine their connection relationship with surrounding edge computing nodes in the next cycle.

[0116] The edge computing server is used to periodically acquire the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it. The acquired communication status information is input into a pre-established node connection decision model to obtain real-time decision results. Based on the real-time decision results, the connection relationship between each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it is dynamically adjusted in real time to form an adaptive topology network structure. Data transmission is also carried out between the edge computing server and each edge computing node based on the adaptive topology network structure.

[0117] The aforementioned edge computing system, based on the communication status information of each edge computing node in the topological network structure and the edge computing nodes within a preset distance range around it, and the node connection decision model, determines the network connection relationship of each edge computing node in the topological network structure for the next cycle and dynamically adjusts the topological network structure in real time. It can adapt to the dynamically changing network environment and use the topological network structure to efficiently and stably transmit data.

[0118] Unless otherwise specifically stated, terms such as processing, calculation, operation, determination, display, etc., may refer to the actions and / or processes of one or more processing or computing systems or similar devices that represent the manipulation and conversion of data representing physical (e.g., electronic) quantities within the registers or memory of the processing system into other data similarly representing physical quantities within the memory, registers, or other such information storage, transmission, or display devices of the processing system. Information and signals can be represented using any of a variety of different techniques and methods. For example, data, instructions, commands, information, signals, bits, symbols, and chips mentioned throughout the above description can be represented by voltage, current, electromagnetic waves, magnetic fields or particles, light fields or particles, or any combination thereof.

[0119] It should be understood that the specific order or hierarchy of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process may be rearranged without departing from the scope of this disclosure. The appended method claims provide elements of various steps in an exemplary order and are not intended to limit the scope to the specific order or hierarchy described.

[0120] In the detailed description above, various features are combined together in a single embodiment to simplify this disclosure. This approach to disclosure should not be construed as reflecting an intention that embodiments of the claimed subject matter require more features than are explicitly stated in each claim. Rather, as reflected in the appended claims, the invention is presented with fewer features than all of the features in a single disclosed embodiment. Therefore, the appended claims are hereby explicitly incorporated into the detailed description, with each claim representing a separate preferred embodiment of the invention.

[0121] Those skilled in the art will also understand that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments herein can be implemented as electronic hardware, computer software, or a combination thereof. To clearly illustrate the interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps described above are generally described in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art can implement the described functionality in alternative ways for each specific application; however, such implementation decisions should not be construed as departing from the scope of this disclosure.

[0122] The steps of the methods or algorithms described in conjunction with the embodiments herein can be directly embodied in hardware, software modules executed by a processor, or a combination thereof. The software modules can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium well known in the art. An exemplary storage medium is connected to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. The ASIC can reside in a user terminal. Alternatively, the processor and storage medium can exist as discrete components in the user terminal.

[0123] For software implementation, the techniques described in this application can be implemented using modules (e.g., procedures, functions, etc.) that perform the functions described in this application. This software code can be stored in memory units and executed by a processor. The memory units can be implemented within the processor or outside the processor; in the latter case, they are communicatively coupled to the processor via various means, as is well known in the art.

[0124] The foregoing description includes examples of one or more embodiments. It is certainly impossible to describe all possible combinations of components or methods in order to describe the above embodiments, but those skilled in the art will recognize that further combinations and arrangements of the various embodiments are possible. Therefore, the embodiments described herein are intended to cover all such changes, modifications, and variations that fall within the scope of the appended claims. Furthermore, the term "comprising" as used in the specification or claims is interpreted in a manner similar to the term "including," as interpreted when used as a conjunction in the claims. Additionally, the use of any term "or" in the specification of the claims is intended to mean "non-exclusive or."

Claims

1. A communication method for an edge computing system, characterized in that, The edge computing system includes at least one edge server and multiple edge computing nodes, and the method includes: Each edge computing node in the edge computing system periodically sends its own communication status information so that surrounding edge computing nodes can obtain its own communication status information in real time. The edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it. This acquired communication status information is then input into a pre-established node connection decision model to predict the connection relationship of each edge computing node in the topology network structure for the next period, yielding real-time decision results. Based on these real-time decision results, the connection relationships of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it are dynamically adjusted in real time, forming an adaptive topology network structure. The real-time decision results include which edge computing nodes in the topology network structure need to be connected and which need to be disconnected. Each edge computing node in the topological network structure periodically acquires the communication status information of the surrounding edge computing nodes within a preset distance range, and sends its own communication status information and the collected communication status information to the edge server. In the topology network structure, the edge computing nodes that an edge computing node needs to connect to include the surrounding edge computing nodes that are currently connected, as well as the surrounding edge computing nodes that need to be connected to this edge computing node but are not currently connected. In the topology network structure, the edge computing nodes that need to be disconnected include the surrounding edge computing nodes that are currently connected but need to be disconnected, as well as the surrounding edge computing nodes that are not connected to this edge computing node. The process of establishing the node connection decision model includes: inputting the sample data included in the training set into the machine learning model for iterative training until the loss function of the machine learning model meets the requirements, and obtaining the trained machine learning model as the node connection decision model. The sample data includes historical communication status information of each edge computing node and its surrounding edge computing nodes within a preset distance range in a known topology network structure, as well as historical connection relationships between each edge computing node and its surrounding edge computing nodes within a preset distance range under different historical communication states. The edge server transmits data with each edge computing node based on an adaptive topology network structure; The communication method further includes: periodically optimizing the adaptive topology network structure based on network conditions using the edge server.

2. The method as described in claim 1, characterized in that, Before the edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, an initial topology network structure is formed, including: The edge server selects an edge computing node as the topology starting point based on user selection, and obtains the communication status information of the topology starting point and other edge computing nodes within a preset distance range around it. The obtained communication status information is input into a pre-established node connection decision model to obtain a decision result. Based on the decision result, a topology connection is established between the topology starting point and the edge computing nodes within a preset distance range around it to form an initial topology network structure.

3. The method as described in claim 1, characterized in that, The edge server periodically acquires communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, including: Each edge computing node in the topological network structure periodically collects the communication status information of the surrounding edge computing nodes within a preset distance range, and sends its own communication status information and the collected communication status information to the edge server. The edge server periodically acquires the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it.

4. The method as described in claim 1, characterized in that, The acquired communication status information is input into a pre-established node connection decision model to obtain real-time decision results. Based on these results, the connection relationships between each edge computing node and its surrounding edge computing nodes in the topology are dynamically adjusted in real time to form an adaptive topology, including: The edge server inputs the periodically acquired communication status information into the node connection decision model and outputs the edge computing nodes that each edge computing node in the topology network structure needs to connect to and the edge computing nodes that need to be disconnected. The edge server dynamically adjusts the connection between each edge computing node and the edge computing nodes that need to be connected in the real-time topology network structure, and also dynamically adjusts the disconnection between each edge computing node and the edge computing nodes that need to be disconnected in the real-time topology network structure, forming an adaptive topology network structure. The edge computing nodes that need to be connected and the edge computing nodes that need to be disconnected are determined from the edge computing nodes within a preset distance range around each edge computing node.

5. The method as described in claim 4, characterized in that, The communication status information includes at least one of the following: signal strength, data transmission rate, packet loss rate, latency, power consumption, mobile speed, and historical connection quality.

6. The method as described in claim 4, characterized in that, The edge server inputs the periodically acquired communication status information into the node connection decision model, including: Decision features are extracted from the acquired communication status information; the decision features include at least one of the following: standard deviation of signal strength, maximum signal strength, minimum signal strength, average signal strength, standard deviation of data transmission rate, maximum signal transmission rate, minimum data transmission rate, average data transmission rate, standard deviation of packet loss rate, maximum packet loss rate, minimum packet loss rate, standard deviation of delay, standard deviation of energy consumption, average energy consumption, standard deviation of mobile speed, maximum mobile speed, historical connection quality average over a past time period, and trend features of historical data; The extracted decision features are input into the node-connected decision model.

7. The method as described in claim 1, characterized in that, The edge server transmits data with each edge computing node based on an adaptive topology network structure, including: The edge server receives data sent by each edge computing node in the adaptive topology network structure and processes the data; The processed data is then fed back to the corresponding edge computing nodes.

8. The method as described in claim 7, characterized in that, Preprocessing the data before processing includes: The edge server preprocesses the received data, including at least one of the following: data cleaning, data integration, data transformation, data reduction, data discretization, data encoding and transformation, and data quality checking, to obtain preprocessed data.

9. The method as described in claim 1, characterized in that, The edge server optimizes the adaptive topology network structure based on network conditions, including: The edge server collects the status information of each edge computing node in the current adaptive topology network structure. The status information includes the connection status and communication quality parameters of each edge computing node. Based on the state information and optimization algorithm, the current adaptive topology network structure is evaluated and optimized; the optimization algorithm includes genetic algorithm and particle swarm optimization algorithm. The edge server sends the optimized adaptive topology network structure to each edge computing node, guiding the edge computing nodes to adjust the adaptive topology network structure.

10. An edge computing server, characterized in that, include: The decision generation module is used to periodically acquire the communication status information of each edge computing node in the topology network structure and the edge computing nodes within a preset distance range around it, and input the acquired communication status information into the pre-established node connection decision model to predict the connection relationship of each edge computing node in the topology network structure in the next cycle and obtain real-time decision results. The real-time decision results include the edge computing nodes that need to be connected and disconnected in the topology network structure; The process of establishing the node connection decision model includes: inputting the sample data included in the training set into the machine learning model for iterative training until the loss function of the machine learning model meets the requirements, and obtaining the trained machine learning model as the node connection decision model; the sample data includes the historical communication status information of each edge computing node and its surrounding edge computing nodes within a preset distance range in a known topological network structure, as well as the historical connection relationship between each edge computing node and its surrounding edge computing nodes within a preset distance range under different historical communication states. The topology network generation module is used to dynamically adjust the connection relationships between each edge computing node and the surrounding edge computing nodes within a preset distance range in the topology network structure based on real-time decision results, forming an adaptive topology network structure; and to periodically optimize the adaptive topology network structure according to network conditions; the edge computing nodes that an edge computing node in the topology network structure needs to connect to include the surrounding edge computing nodes that are currently connected, and the surrounding edge computing nodes that need to be connected to this edge computing node but are not currently connected; the edge computing nodes that an edge computing node in the topology network structure needs to disconnect from include the surrounding edge computing nodes that are currently connected but need to be disconnected, and the surrounding edge computing nodes that are not connected to this edge computing node. The communication module is used for data transmission with each edge computing node based on an adaptive topology network structure.

11. An edge computing system, characterized in that, include: Edge computing nodes and edge servers; Edge computing nodes are used to periodically send their own communication status information and collect the communication status information of edge computing nodes within a preset distance range, and then send their own communication status information and the collected communication status information to the edge server. An edge server is used to periodically acquire communication status information of each edge computing node in the topological network structure and its surrounding edge computing nodes within a preset distance range. This acquired communication status information is then input into a pre-established node connection decision model to predict the connection relationship of each edge computing node in the topological network structure for the next period, obtaining real-time decision results. Based on these real-time decision results, the connection relationships of each edge computing node and its surrounding edge computing nodes within a preset distance range in the topological network structure are dynamically adjusted in real time to form an adaptive topological network structure. Data transmission is also performed between the adaptive topological network structure and each edge computing node. The establishment process of the node connection decision model includes: inputting sample data included in the training set into a machine learning model for iterative training until the loss function of the machine learning model meets the requirements, obtaining a trained machine learning model as the node connection decision model; the sample data includes historical communication status information of each edge computing node and its surrounding edge computing nodes within a preset distance range in the known topological network structure, as well as the historical connection relationships between each edge computing node and its surrounding edge computing nodes within a preset distance range under different historical communication states. In the topology network structure, the edge computing nodes that an edge computing node needs to connect to include the surrounding edge computing nodes that are currently connected, as well as the surrounding edge computing nodes that need to be connected to this edge computing node but are not currently connected. In the topology network structure, the edge computing nodes that an edge computing node needs to disconnect from include the surrounding edge computing nodes that are currently connected but need to be disconnected, as well as the surrounding edge computing nodes that are not connected to this edge computing node. The edge server is also used to periodically optimize the adaptive topology network structure based on network conditions.

12. A computer storage medium, characterized in that, The computer storage medium stores computer-executable instructions, which, when executed by a processor, implement the communication method of the edge computing system according to any one of claims 1-9.

13. An edge computing device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the communication method of the edge computing system according to any one of claims 1-9.

Images