Cloud mobile phone instance selection method and device based on load balancing, equipment and medium

Abstract

The invention provides a cloud mobile phone instance selection method and device based on load balancing, equipment and a medium, and relates to the technical field of computers. The method comprises the following steps: collecting network state data and user position information; dynamic coefficients of a plurality of candidate cloud mobile phone instances are generated based on the network state data through a pre-trained long and short-term memory network model; based on the dynamic coefficient, the network state data and the user position information, generating a dynamic weight load model corresponding to each candidate cloud mobile phone instance; respectively calculating weight scores of the plurality of candidate cloud mobile phone instances through a dynamic weight load model; and determining an optimal cloud mobile phone instance based on the weight score. According to the embodiment, the influence of the network state data of each candidate cloud mobile phone instance on the load balancing can be dynamically evaluated in real time based on the load balancing, so that the advantages and disadvantages of the candidate cloud mobile phone instances can be more comprehensively, efficiently and accurately evaluated, and the resource allocation in the cloud environment is optimized.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, specifically to the field of cloud phone processing technology, and in particular to a method, device, equipment and medium for selecting a cloud phone instance based on load balancing. Background Art

[0002] The existing load balancing solutions for cloud phones based on Domain Name System (DNS) resolution have a lower quality of service (QoS) when users move due to the rigid rules. The measured data shows that the delay fluctuation of users switching across domains is more than 300ms. In addition, the traditional polling detection mechanism used by such solutions only relies on the current network status, and the probability of sudden delay increase when accessing across regions is 45%.

[0003] Therefore, there is an urgent need for a load balancing solution for cross-domain cloud phone instances with more stable latency. Summary of the invention

[0004] The embodiments of the present application propose a method, device, equipment and medium for selecting a cloud phone instance based on load balancing.

[0005] In the first aspect, an embodiment of the present application proposes a cloud phone instance selection method based on load balancing, including: collecting network status data and user location information; generating dynamic coefficients of multiple candidate cloud phone instances based on the network status data through a pre-trained long short-term memory network model; generating a dynamic weight load model corresponding to each of the candidate cloud phone instances based on the dynamic coefficient, network status data and user location information; calculating the weight scores of the multiple candidate cloud phone instances through the dynamic weight load model; and determining the optimal cloud phone instance based on the weight scores.

[0006] Furthermore, the dynamic coefficients include: a delay weight coefficient, a packet loss rate weight coefficient and a geographic distance weight coefficient.

[0007] The generating of the dynamic weight load model corresponding to each of the candidate cloud phone instances based on the dynamic coefficient, the network status data and the user location information includes:

[0008] Determine network delay data and packet loss rate data based on the network status data;

[0009] Based on the network delay data, packet loss rate data, user location information and the delay weight coefficient, packet loss rate weight coefficient and geographic distance weight coefficient, the dynamic weight load model corresponding to each candidate cloud phone instance is constructed respectively.

[0010] Furthermore, it also includes:

[0011] Calculate the user's moving speed based on the user location information;

[0012] Calculate the historical delay standard deviation based on the network delay data;

[0013] In response to the user's moving speed or the historical delay standard deviation being greater than the corresponding preset threshold, generate a regional network delay heat map result based on the user location information, network status data, and a pre-trained graph neural network model;

[0014] Generate a delay prediction result within a preset time period based on the regional network delay heat map result.

[0015] Further, it also includes:

[0016] In response to a change in the user's network status, perform cloud mobile phone instance migration based on the change in the network status through an edge computing node; or,

[0017] In response to the network delay data or the delay prediction result exceeding a preset delay threshold, perform cloud mobile phone instance migration based on the network delay data or the delay prediction result through the edge computing node; or,

[0018] In response to the packet loss rate data exceeding a preset packet loss rate threshold, perform cloud mobile phone instance migration based on the packet loss rate data through the edge computing node.

[0019] Further, it also includes:

[0020] Record the cloud mobile phone instance migration event through a blockchain.

[0021] Further, for generating the dynamic weight load model corresponding to each of the candidate cloud mobile phone instances based on the dynamic coefficient, network status data, and user location information, it also includes:

[0022] Update the dynamic weight load model corresponding to each of the candidate cloud mobile phone instances based on the delay prediction result, packet loss rate data, user location information, and the delay weight coefficient, packet loss rate weight coefficient, and geographical distance weight coefficient.

[0023] Further, for generating the regional network delay heat map result based on the user location information, network status data, and a pre-trained graph neural network model, it includes:

[0024] Use the user location information as node features and the network topology structure obtained based on the network status data as edge features;

[0025] Construct a graph structure based on the node features and edge features, and the features of the graph structure include: the network delay data, packet loss rate data, and bandwidth of the node features;

[0026] Predict the node features of the graph structure through the pre-trained graph neural network model to obtain the regional network delay heat map result.

[0027] In a second aspect, an embodiment of the present application provides a cloud phone instance selection device based on load balancing, including: a data acquisition module configured to collect network status data and user location information; a dynamic coefficient generation module configured to generate dynamic coefficients of multiple candidate cloud phone instances respectively based on the network status data through a pre-trained long short-term memory network model; a dynamic weight load model generation module configured to generate a dynamic weight load model corresponding to each of the candidate cloud phone instances respectively based on the dynamic coefficients, network status data, and user location information; a weight score calculation module configured to calculate the weight scores of the multiple candidate cloud phone instances respectively through the dynamic weight load models; and a cloud phone instance determination module configured to determine an optimal cloud phone instance based on the weight scores.

[0028] In a third aspect, an embodiment of the present invention provides a computer device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the method according to the first aspect or any corresponding implementation manner thereof.

[0029] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to cause a computer to execute the method according to the first aspect or any corresponding implementation manner thereof.

[0030] In a fifth aspect, an embodiment of the present application provides a computer program product including a computer program, and the computer program can implement the load balancing-based cloud phone instance selection method described in any implementation manner in the first aspect when executed by a processor.

[0031] The load balancing-based cloud phone instance selection method, device, equipment, and medium provided by the embodiments of the present application can, based on the changes in the network environment where the cloud phone instance is located and the location of the user, evaluate in real time and dynamically the impact of the network status data of each candidate cloud phone instance on load balancing based on load balancing, so as to be able to evaluate the advantages and disadvantages of the candidate cloud phone instances more comprehensively, efficiently, and accurately, allocate better cloud phone instances for users to provide better services, and thus optimize the resource allocation in the cloud environment.

[0032] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. Brief Description of the Drawings

[0033] Other features, objects, and advantages of the present disclosure will become more apparent from the following detailed description of non-limiting embodiments read with reference to the accompanying drawings:

[0034] Figure 1 is an exemplary system architecture to which the present disclosure can be applied;

[0035] Figure 2 is a flowchart of a method for selecting cloud phone instances based on load balancing provided by an embodiment of the present application;

[0036] Figure 3 is a flowchart of another method for selecting cloud phone instances based on load balancing provided by an embodiment of the present application;

[0037] Figure 4 is a flowchart of another method for selecting cloud phone instances based on load balancing provided by an embodiment of the present application;

[0038] Figure 5 is a structural block diagram of a device for selecting cloud phone instances based on load balancing provided by an embodiment of the present application;

[0039] Figure 6 is a schematic structural diagram of an electronic device suitable for executing the method for selecting cloud phone instances based on load balancing provided by an embodiment of the present application. Detailed Embodiments

[0040] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present application are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for clarity and conciseness, descriptions of well-known functions and structures are omitted below. It should be noted that, without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0041] In the technical solution of the present disclosure, the processing of collection, storage, use, processing, transmission, provision, and disclosure of user personal information complies with the provisions of relevant laws and regulations and does not violate public order and good customs.

[0042] Figure 1 Illustrates an exemplary system architecture 100 of embodiments of the method, device, electronic device, and computer-readable storage medium for selecting cloud phone instances based on load balancing to which the present disclosure can be applied.

[0043] As Figure 1As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0044] Users can use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various applications for implementing information communication between the two can be installed on the terminal devices 101, 102, 103 and the server 105, such as instant messaging applications, etc.

[0045] The terminal devices 101, 102, 103 and the server 105 can be either hardware or software. When the terminal devices 101, 102, 103 are hardware, they can be various electronic devices with a display screen, including but not limited to smartphones, tablets, laptop computers, and desktop computers, etc.; when the terminal devices 101, 102, 103 are software, they can be installed in the above-listed electronic devices, and can be implemented as multiple software or software modules, or as a single software or software module, without specific limitation here. When the server 105 is hardware, it can be implemented as a distributed server cluster composed of multiple servers, or as a single server; when the server is software, it can be implemented as multiple software or software modules, or as a single software or software module, without specific limitation here.

[0046] The server 105 can provide various services through various built-in applications. It should be noted that the data or information required to provide various services can be obtained from the terminal devices 101, 102, 103 via the network 104, or can also be pre-stored locally in the server 105 in various ways. Therefore, when the server 105 detects that these data have been stored locally, it can choose to directly obtain these data from the local. In this case, the exemplary system architecture 100 may not include the terminal devices 101, 102, 103 and the network 104.

[0047] Since the process of analyzing based on data may require a large amount of computing resources and strong computing power, the method for selecting cloud mobile phone instances based on load balancing provided in subsequent embodiments of the present disclosure is generally executed by the server 105 with strong computing power and a large amount of computing resources. Correspondingly, the device for selecting cloud mobile phone instances based on load balancing is generally also set in the server 105. However, it should also be noted that when the terminal devices 101, 102, and 103 also have computing power and computing resources that meet the requirements, the terminal devices 101, 102, and 103 can also complete the various operations originally performed by the server 105 through relevant applications installed thereon, and then output the same results as the server 105. Especially when there are multiple terminal devices with different computing capabilities at the same time, but when the relevant application determines that the terminal device where it is located has strong computing power and a large amount of remaining computing resources, the terminal device can be allowed to perform the above operations, thereby appropriately reducing the computing pressure on the server 105. Correspondingly, the device for selecting cloud mobile phone instances based on load balancing can also be set in the terminal devices 101, 102, and 103. In this case, the exemplary system architecture 100 may not include the server 105 and the network 104 either.

[0048] It should be understood that Figure 1 the numbers of terminal devices, networks, and servers in

[0049] Please refer to Figure 2 , Figure 2 which is a flowchart of a method for selecting cloud mobile phone instances based on load balancing provided in an embodiment of the present application. The process 200 includes the following steps:

[0050] Step S201: Collect network status data and user location information.

[0051] In an embodiment of the present application, the current network status data and user location information are collected by the (such as Figure 1 the server 105 shown) method for selecting cloud mobile phone instances based on load balancing. Among them, the network status data may include: latency, bandwidth, jitter, packet loss rate, throughput, etc. The user location information may include geographical location information, moving speed and direction information, etc.

[0052] Step S202: Generate dynamic coefficients of multiple candidate cloud mobile phone instances respectively based on the network status data through a pre-trained long short-term memory network model.

[0053] In the embodiment of the present application, the above-mentioned pre-trained long short-term memory network model (LSTM) generates respective corresponding dynamic coefficients based on the network state data corresponding to each candidate cloud mobile phone instance. The dynamic coefficient is used to represent the real-time weight parameter, and is used to quantify the relative importance of different network or service metrics in the load balancing decision. These coefficients are usually represented as α, β, γ, etc., corresponding to key metrics such as latency, packet loss, and geodistance. The training process of the LSTM model uses historical network state data (including bandwidth, latency, packet loss, etc.) as input, uses the mean square error (MSE) as the loss function, and the optimization algorithm is Adam. The training data set includes at least 1 million historical network state records, and the data preprocessing includes normalization and sliding window sampling.

[0054] Step S203: Generate a dynamic weight load model corresponding to each candidate cloud mobile phone instance based on the dynamic coefficient, network state data, and user location information.

[0055] In the embodiment of the present application, a dynamic weight load model is constructed based on the dynamic coefficient, network state data, and user location information, and the candidate cloud mobile phone instances are comprehensively evaluated in multiple dimensions through the dynamic coefficient.

[0056] Step S204: Calculate the weight scores of multiple candidate cloud mobile phone instances through the dynamic weight load model respectively.

[0057] In the embodiment of the present application, the weight scores of each candidate cloud mobile phone instance can be calculated through the dynamic weight load model, and the weight score is used to represent the comprehensive evaluation score of the network environment where each candidate cloud mobile phone is located.

[0058] Step S205: Determine the optimal cloud mobile phone instance based on the weight score.

[0059] In the embodiment of the present application, based on the weight score, the optimal cloud mobile phone instance can be determined. Usually, the cloud mobile phone instance with the lowest weight score is used as the optimal cloud mobile phone instance to provide the corresponding cloud mobile phone service for the user.

[0060] The method for selecting cloud mobile phone instances based on load balancing provided by the embodiment of the present application can, based on the changes in the network environment where the cloud mobile phone instances are located and the location of the user, evaluate the impact of the network state data of each candidate cloud mobile phone instance on load balancing in real time and dynamically based on load balancing, so as to be able to more comprehensively, efficiently, and accurately evaluate the advantages and disadvantages of the candidate cloud mobile phone instances, allocate better cloud mobile phone instances for the user to provide better services, and thus optimize the resource allocation in the cloud environment.

[0061] In some alternative embodiments of the present embodiment, the above dynamic coefficients mainly include: a delay weight coefficient, a packet loss rate weight coefficient, and a geographical distance weight coefficient. The delay weight coefficient reflects the impact of network delay on load balancing; the packet loss rate weight coefficient reflects the impact of network packet loss on load balancing; the geographical distance weight coefficient reflects the impact of geographical distance on load balancing. Correspondingly, the above step S203, the process of generating a dynamic weight load model corresponding to each candidate cloud mobile phone instance based on the dynamic coefficients, network status data, and user location information mainly includes:

[0062] Step 1: Determine network delay data and packet loss rate data based on network status data.

[0063] In the embodiments of the present application, network status data is first collected from multiple data sources, and these data sources may include log records of network devices, real-time data of network monitoring software, etc. After collecting the data, it is preliminarily sorted and filtered to remove invalid or incorrect data to ensure the accuracy and availability of the data. Then, specific algorithms are used to extract information related to network delay from the sorted data. For example, by analyzing the difference between the sending time and receiving time of data packets, the network delay data is calculated.

[0064] For the determination of the packet loss rate data, the total number of data packets sent within a certain time period and the number of data packets that were not successfully received are counted, and then the number of data packets that were not successfully received is divided by the total number of data packets sent to obtain the packet loss rate data. To ensure the reliability of the data, it is calculated multiple times and the average value is taken to finally obtain relatively accurate network delay data and packet loss rate data.

[0065] Step 2: Construct a dynamic weight load model corresponding to each candidate cloud mobile phone instance based on the network delay data, packet loss rate data, user location information, delay weight coefficient, packet loss rate weight coefficient, and geographical distance weight coefficient.

[0066] The user location information is obtained according to the user's device information or IP address, etc., and at the same time, the pre-set delay weight coefficient, packet loss rate weight coefficient, and geographical distance weight coefficient are clarified. For each candidate cloud mobile phone instance, the execution entity calculates the geographical distance between the instance and the user, which can be completed by means of geographic information system (GIS) technology.

[0067] Then, the previously obtained network delay data, packet loss rate data, and the calculated geographical distance data are multiplied by the corresponding weight coefficients respectively. Then, these products are added together to obtain a comprehensive score for the candidate cloud mobile phone instance. This comprehensive score is the core index in the dynamic weight load model, which reflects the load situation of the candidate cloud mobile phone instance after considering factors such as network delay, packet loss rate, and geographical distance.

[0068] By performing such calculations on each candidate cloud phone instance, a dynamic weight load model corresponding to each instance can be constructed, providing a basis for subsequent cloud phone selection and resource allocation.

[0069] In the embodiment of the present application, it is intended to construct a dynamic weight load model corresponding to each candidate cloud phone instance based on the obtained network latency data and packet loss rate data, combined with the user location information and the corresponding latency weight coefficient, packet loss rate weight coefficient, and geographical distance weight coefficient. Among them, the latency weight coefficient corresponds to the network latency data, the packet loss rate weight coefficient corresponds to the packet loss rate data, and the geographical distance weight coefficient corresponds to the user location information.

[0070] Exemplarily, the dynamic weight load model can be expressed as the following formula (1):

[0071] W = α·Latency + β·PacketLoss + γ·Geodistance (1)

[0072] Where α is the latency weight coefficient, Latency is the network latency data; β is the packet loss rate weight coefficient, PacketLoss is the packet loss rate data; γ is the geographical distance weight coefficient, and Geodistance is the user location information.

[0073] In practical applications, when the user's location changes frequently or has a large span, the network conditions in different regions may vary, resulting in large network fluctuations in the area where the user is located. Relatively speaking, the network latency may also be high. For such a situation, in order to ensure network stability during the user's usage process, in this embodiment, network latency prediction can also be performed based on the user's location change or network latency situation. In some alternative implementation manners of this embodiment, as Figure 3 shown, the method for selecting a cloud phone instance based on load balancing further includes:

[0074] Step S301: Calculate the user's moving speed based on the user location information.

[0075] In the embodiment of the present application, continuously collect the user's location information, and the precise location of the user at different time points can be obtained by means of technologies such as GPS positioning and base station positioning. Record these location information and the corresponding timestamps, and then select two time points and their corresponding location data, and calculate the distance difference between these two locations to obtain the distance traveled by the user during this time period.

[0076] Meanwhile, calculate the time interval between these two time points. Divide the distance by the time interval to obtain the average moving speed of the user during this time period. To ensure the accuracy and timeliness of speed calculation, the execution entity will continuously update the time points and location data, and continuously calculate the moving speed, so as to achieve dynamic monitoring of the user's moving speed.

[0077] Step S302: Calculate the historical delay standard deviation based on the network delay data.

[0078] In the embodiment of the present application, obtain multiple network delay data of the user within a preset past time period. These data can be obtained from network monitoring systems, device logs and other channels. Organize the collected network delay data into a data set, and then calculate the average value of the data set, that is, the sum of all delay data divided by the number of data.

[0079] After that, for each delay data in the data set, calculate the difference between it and the average value, and square the difference. Add up all the squared differences and divide by the number of data to obtain the variance.

[0080] Finally, take the square root of the variance to obtain the historical delay standard deviation. This standard deviation can reflect the degree of dispersion of network delay data within the preset time period and is an important indicator to measure the stability of network quality.

[0081] Step S303: In response to the user's moving speed or historical delay standard deviation being greater than the corresponding preset threshold, generate a regional network delay heat map result based on the user location information, network status data and a pre-trained graph neural network model.

[0082] In the embodiment of the present application, compare the user's moving speed and historical delay standard deviation calculated in real time with the corresponding preset thresholds respectively. If the user's moving speed is greater than a preset speed threshold such as 5 m / s, or the historical delay standard deviation is greater than a preset standard deviation threshold such as 20 ms, it means that the fluctuations in the user's movement or network delay may cause relatively large delays due to network changes. At this time, the execution entity will integrate the user location information and network status data and use them as input data to be passed into a pre-trained graph neural network model. The model has been trained with a large amount of historical data and can learn the complex relationship between network features and network delays.

[0083] The model will analyze and process the input data and output the network delay prediction results for each region. Visualize these prediction results and present them in the form of a regional network delay heat map, where different colors represent different degrees of network delay, so as to intuitively display the distribution of regional network delays.

[0084] Step S304: Generate a delay prediction result within a preset time period based on the regional network delay heat map result.

[0085] In the embodiments of the present application, the colors and values in different regions of the heat map represent the current network delay situation. Combining historical network delay data and the dynamic change rules of the network, such as network usage peaks and troughs at different time periods, maintenance plans of network devices, and other factors. Using methods such as time series analysis and machine learning to predict the change trend of regional network delay within a preset time period.

[0086] By establishing a prediction model, taking the current heat map result and relevant influencing factors as inputs, the model will output the delay prediction values for each region within a preset time period. Finally, organize these prediction values into a clear delay prediction result to provide a basis for decisions such as network optimization and resource allocation.

[0087] Through the above process, it is possible to make predictions in advance based on the real-time location change of the user or the change of network delay, anticipate the future network delay situation in advance, detect such a situation before the network delay fluctuates greatly, and take corresponding measures in advance to ensure the stability of the network environment or service provided to the user.

[0088] Furthermore, the embodiments of the present application can also automatically adjust the thresholds of the user's moving speed and the historical delay standard deviation according to different time periods and different network scenarios (such as weekdays / weekends, indoor / outdoor).

[0089] Specifically, collect and analyze data for different time periods (such as day, night, working hours and rest hours on weekdays, etc.) and different network scenarios (distinguish indoor scenarios, such as home, office, and outdoor scenarios, such as streets, squares, etc.). By long-term monitoring and recording the moving speed data of users under these different conditions and the historical delay standard deviation data of the network, establish a corresponding database. Then, according to the distribution and characteristics of these data, use statistical methods or machine learning algorithms to calculate reasonable thresholds for the user's moving speed and the historical delay standard deviation in each time period and network scenario.

[0090] During the actual operation process, obtain the current time information and the network scenario information where the user is located in real time. When it is detected that the time or scenario changes, automatically retrieve the corresponding threshold from the database and apply it to subsequent judgments. For example, during the working hours on weekdays and in the office (indoor scenario), use a relatively low set of thresholds for the user's moving speed and the historical delay standard deviation; while in the outdoor scenario on weekends, switch to another set of thresholds more suitable for this scenario.

[0091] At the same time, set up user feedback channels, such as providing feedback portals in the application, so that users can evaluate and provide feedback on the quality of network services, including whether there are problems such as freezes and excessive delays. The system collects this feedback information and evaluates the current thresholds in combination with comprehensive evaluation indicators of service quality (such as the smoothness of video playback, application response time, etc.). If it is found that there are many user feedbacks or the service quality is poor under the current threshold, the system will analyze the relevant data again, further optimize the threshold setting, and adjust the size or range of the threshold to make it more in line with the actual network usage needs, thereby improving the user experience and the stability of network services.

[0092] Furthermore, the above can also further update the dynamic weight load model corresponding to each candidate cloud phone instance based on the delay prediction result obtained in the above process, combined with the packet loss rate data, user location information, delay weight coefficient, packet loss rate weight coefficient and geographical distance weight coefficient, and recalculate the weight scores of multiple candidate cloud phone instances based on the updated dynamic weight load model, and determine the optimal cloud phone instance again based on the weight score. Through this process, it is possible to ensure that the optimal cloud phone instance is updated in real time, dynamically and timely, so as to provide it to users and provide cloud phone services to users.

[0093] In some optional implementations of this embodiment, since the network environment or the user's location may change at any time, the selected optimal cloud phone instance may be affected by changes in the network environment or changes in the user's location. Therefore, in this embodiment, the migration of cloud phone instances can be achieved based on this situation. The migration of mobile phone instances can be achieved through edge computing nodes, which are mainly used to deploy cloud phone instances and support dynamic migration and operation of instances. The load-balancing-based cloud phone instance selection method also includes: in response to changes in the user's network status, cloud phone instance migration based on changes in network status through edge computing nodes; or, in response to network delay data or delay prediction results exceeding a preset delay threshold, cloud phone instance migration based on network delay data or delay prediction results through edge computing nodes; or, in response to packet loss rate data exceeding a preset packet loss rate threshold, cloud phone instance migration based on packet loss rate data through edge computing nodes.

[0094] In the above process, various situations that may cause significant changes in the network environment or network latency are considered, that is, when the user's network status changes (for example, the user switches from the mobile network of the mobile phone to the Wi-Fi network, or from the Wi-Fi network to the mobile network of the mobile phone, etc.), the network latency data or the latency prediction result may exceed the preset latency threshold (for example, greater than 150 ms), and the network packet loss rate data exceeds the preset packet loss rate threshold (for example, greater than 5%), the migration of the cloud mobile phone instance can be realized through the edge computing node, so as to ensure that the user can continue to use the cloud mobile phone instance in a relatively stable network environment and improve the stability of the cloud mobile phone instance provided for the user.

[0095] Furthermore, the above cloud mobile phone instance migration event can be recorded through the blockchain.

[0096] In this embodiment, the chain structure of the blockchain adopts the improved PBFT (Practical Byzantine Fault Tolerance) consensus (the number of nodes ≤ 10, the consensus time ≤ 2 seconds), and the hash value of the migration event (SHA-3 algorithm) is recorded through the smart contract to ensure that the operation cannot be tampered with. The trigger condition of the smart contract is the instance migration event (threshold: latency > 150 ms or packet loss rate > 5%). The contract records the hash value of the migration event to ensure the integrity and non-tamperability of the data. In the blockchain record, the sensitive information of the user can be encrypted to ensure the privacy of the user.

[0097] Please refer to Figure 4 , Figure 4 which is the flowchart of a method for selecting a cloud mobile phone instance based on load balancing provided by an embodiment of the present application, that is, a specific implementation manner is provided for step 303 in process 300 shown in Figure 3 . Other steps in process 300 are not adjusted, and a new complete embodiment is obtained by replacing step 303 with the specific implementation manner provided in this embodiment. The process 400 includes the following steps:

[0098] Step S401, using the user location information as the node feature and the network topology structure obtained based on the network status data as the edge feature.

[0099] In this embodiment, the user location information is collected, and accurate location data is obtained by means of the GPS positioning function of the user device, IP address positioning, or base station positioning, etc. At the same time, the network status data is collected from multiple data sources, such as the operation logs of network devices, the data of traffic monitoring systems, etc. The collected network status data is deeply analyzed, and the professional network topology discovery algorithm is used to identify the connection relationship and communication path between each node in the network, so as to construct the network topology structure.

[0100] Subsequently, the obtained user location information is used as node features, which means that each node corresponding to a user in the graph data has a feature attribute representing its location. And the constructed network topology structure is used as edge features, that is, the edges between nodes in the graph have attributes related to network connections. In this way, graph data can be successfully constructed based on user location information and network topology structure.

[0101] Step S402: Construct a graph structure based on node features and edge features. The features of the graph structure include: network latency data, packet loss rate data, and bandwidth of node features.

[0102] In the embodiment of the present application, first, each node in the graph is determined according to user location information, and each node represents a user. Then, edges are added between nodes according to the network topology structure, and these edges reflect the network connection relationship between users.

[0103] At the same time, the network latency data, packet loss rate data, and bandwidth data previously analyzed from network status data are assigned to the corresponding nodes and edges as features of the graph structure. For example, the network latency, packet loss rate, and bandwidth data of the network where the user corresponding to a certain node is located are added to the attributes of the node, and the relevant data of the network connection represented by the edge connecting two nodes are added to the attributes of the edge. The graph structure constructed in this way comprehensively includes the network latency data, packet loss rate data, and bandwidth of node features, providing rich and accurate data information for subsequent prediction.

[0104] Step S403: Predict the node features of the graph structure through a pre-trained graph neural network model to obtain the regional network latency heatmap result.

[0105] In the embodiment of the present application, the constructed graph structure is input into a pre-trained graph neural network (GNN) model. This pre-trained model is trained on a large amount of historical network data and has learned the complex relationship between network features and network latency. The graph neural network model performs layer-by-layer convolution and propagation operations on the node features in the graph structure, and through the neurons and weight parameters inside the model, performs non-linear transformation and feature extraction on the input node features.

[0106] In this process, the model comprehensively considers the features of the node itself and the relationship with adjacent nodes, so as to predict the network latency corresponding to each node. After the prediction is completed, the prediction results of each node are integrated and visualized, and presented in the form of a heatmap to obtain the regional network latency heatmap result. This heatmap can intuitively display the network latency conditions in different regions, helping network administrators quickly locate the regions with high network latency so as to take corresponding optimization measures.

[0107] To deepen the understanding and clarify the advantages and effects of the adaptive encoding and decoding optimization method provided in this embodiment, the present disclosure also provides an illustration in combination with specific application examples.

[0108] As an example: The network switching scenario of User A

[0109] User A uses a cloud mobile phone under a mobile network. It is detected in real time that the bandwidth is low and the latency is high. The geographical location information module determines that User A is located in a certain city, and the nearest edge node is Node X. The cloud mobile phone instance is allocated to Node X. When User A switches to a Wi-Fi network, the edge computing node senses the change in the network environment and automatically adjusts the network configuration of the instance to ensure seamless switching. The instance migration path is recorded by the blockchain recording module to ensure data security.

[0110] As an example: Instance optimization for cross-regional User B

[0111] User B accesses a cloud mobile phone instance in a certain country and detects that the current latency is high. Combining historical data, it is predicted that the future latency will further increase. The instance is migrated from the remote cloud server to an edge node near the user. The instance migration path is recorded by the blockchain to ensure that the data has not been tampered with.

[0112] Further referring to Figure 5 , as an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of a cloud mobile phone instance selection device based on load balancing. This device embodiment corresponds to Figure 2 the method embodiment shown, and this device can be specifically applied to various electronic devices.

[0113] As Figure 5 shown, the load balancing-based cloud mobile phone instance selection device 500 in this embodiment may include: a data acquisition module 501, a dynamic coefficient generation module 502, a dynamic weight load model generation module 503, a weight score calculation module 504, and a cloud mobile phone instance determination module 505. Among them, the data acquisition module 501 is configured to collect network status data and user location information; the dynamic coefficient generation module 502 is configured to generate dynamic coefficients of multiple candidate cloud mobile phone instances respectively based on the network status data through a pre-trained long short-term memory network model; the dynamic weight load model generation module 503 is configured to generate dynamic weight load models corresponding to each candidate cloud mobile phone instance respectively based on the dynamic coefficients, network status data, and user location information; the weight score calculation module 504 is configured to calculate the weight scores of multiple candidate cloud mobile phone instances respectively through the dynamic weight load models; the cloud mobile phone instance determination module 505 is configured to determine the optimal cloud mobile phone instance based on the weight scores.

[0114] In this embodiment, in the cloud mobile phone instance selection device 500 based on load balancing, the specific processing of the data acquisition module 501, the dynamic coefficient generation module 502, the dynamic weight load model generation module 503, the weight score calculation module 504, and the cloud mobile phone instance determination module 505 and the technical effects brought by them can be respectively referred to Figure 2 the relevant descriptions of steps 201-203 in the corresponding embodiment, which will not be elaborated here.

[0115] This embodiment exists as a device embodiment corresponding to the above method embodiment. The cloud mobile phone instance selection device based on load balancing provided in this embodiment can, based on the changes in the network environment where the cloud mobile phone instance is located and the location of the user, evaluate in real time and dynamically the impact of the network status data of each candidate cloud mobile phone instance on load balancing based on load balancing, so as to be able to more comprehensively, efficiently, and accurately evaluate the advantages and disadvantages of the candidate cloud mobile phone instances, allocate better cloud mobile phone instances for users to provide better services, and thus optimize the resource allocation in the cloud environment.

[0116] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor can implement the method for selecting a cloud mobile phone instance based on load balancing described in any of the above embodiments.

[0117] According to an embodiment of the present disclosure, the present disclosure also provides a readable storage medium, which stores computer instructions for enabling a computer to implement the method for selecting a cloud mobile phone instance based on load balancing described in any of the above embodiments when executed.

[0118] According to an embodiment of the present disclosure, the present disclosure also provides a computer program product, which can implement the method for selecting a cloud mobile phone instance based on load balancing described in any of the above embodiments when executed by a processor.

[0119] Please refer to Figure 6 , Figure 6 which is a schematic structural diagram of a computer device provided by an alternative embodiment of the present invention. As Figure 6As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting the components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some alternative embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (such as an array of servers, a set of blade servers, or a multi-processor system).

[0120] The processor 10 can be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 can further include a hardware chip. The above hardware chip can be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above programmable logic device can be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.

[0121] Among them, the memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiments.

[0122] The memory 20 can include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the computer device presented by a kind of landing page of a small program, etc. In addition, the memory 20 can include a high-speed random access memory and can also include a non-transitory memory, such as at least one disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 can optionally include a memory remotely set relative to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0123] The memory 20 can include a volatile memory, such as a random access memory; the memory can also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 can also include a combination of the above types of memories.

[0124] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or communication networks.

[0125] Embodiments of the present invention also provide a computer-readable storage medium. The methods according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code that is originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the methods described herein can be stored as such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the methods shown in the above embodiments are implemented.

[0126] Although embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for selecting cloud phone instances based on load balancing, characterized in that Including: Collecting network status data and user location information; Generating dynamic coefficients of multiple candidate cloud phone instances respectively based on the network status data through a pre-trained long short-term memory network model; Generating a dynamic weight load model corresponding to each of the candidate cloud phone instances respectively based on the dynamic coefficients, network status data, and user location information; Calculating weight scores of the multiple candidate cloud phone instances respectively through the dynamic weight load model; Determining an optimal cloud phone instance based on the weight scores.

2. The method according to claim 1, characterized in that, The dynamic coefficients include: a delay weight coefficient, a packet loss rate weight coefficient, and a geographical distance weight coefficient. The generating a dynamic weight load model corresponding to each of the candidate cloud phone instances respectively based on the dynamic coefficients, network status data, and user location information includes: Determining network delay data and packet loss rate data based on the network status data; Constructing the dynamic weight load model corresponding to each of the candidate cloud phone instances respectively based on the network delay data, packet loss rate data, user location information, and the delay weight coefficient, packet loss rate weight coefficient, and geographical distance weight coefficient.

3. The method according to claim 2, characterized in that, It further includes: Calculating the user's moving speed based on the user location information; Calculating the historical delay standard deviation based on the network delay data; In response to the user's moving speed or the historical delay standard deviation being greater than a corresponding preset threshold, generating a regional network delay heat map result based on the user location information, network status data, and a pre-trained graph neural network model; Generating a delay prediction result within a preset time period based on the regional network delay heat map result.

4. The method according to claim 3, wherein It further includes: In response to a change in the user's network status, migrating the cloud phone instance through an edge computing node based on the change in the network status; Or, In response to the network delay data or the delay prediction result exceeding a preset delay threshold, migrating the cloud phone instance through the edge computing node based on the network delay data or the delay prediction result; or, In response to the packet loss rate data exceeding a preset packet loss rate threshold, migrating the cloud phone instance through the edge computing node based on the packet loss rate data.

5. The method according to claim 4, characterized in that, It further includes: Recording the cloud phone instance migration event through a blockchain.

6. The method according to claim 3, characterized in that, The generating a dynamic weight load model corresponding to each of the candidate cloud phone instances respectively based on the dynamic coefficients, network status data, and user location information further includes: Updating the dynamic weight load model corresponding to each of the candidate cloud phone instances respectively based on the delay prediction result, packet loss rate data, user location information, and the delay weight coefficient, packet loss rate weight coefficient, and geographical distance weight coefficient.

7. The method according to claim 3, characterized in that, The generating a regional network delay heat map result based on the user location information, network status data, and a pre-trained graph neural network model includes: Using the user location information as node features and using the network topology structure obtained based on the network status data as edge features; Constructing a graph structure based on the node features and edge features, and the features of the graph structure include: network delay data, packet loss rate data, and bandwidth of the node features. Predict the node features of the graph structure through the pre-trained graph neural network model to obtain the regional network delay heat map result.

8. A cloud phone instance selection device based on load balancing, characterized in that, It includes: A data acquisition module configured to collect network status data and user location information; A dynamic coefficient generation module configured to generate dynamic coefficients of multiple candidate cloud mobile phone instances respectively based on the network status data through a pre-trained long short-term memory network model; A dynamic weight load model generation module configured to generate a dynamic weight load model corresponding to each of the candidate cloud mobile phone instances respectively based on the dynamic coefficients, network status data and user location information; A weight score calculation module configured to calculate the weight scores of the multiple candidate cloud mobile phone instances respectively through the dynamic weight load model; A cloud mobile phone instance determination module configured to determine an optimal cloud mobile phone instance based on the weight scores.

9. A computer device, characterized in that, It includes: A memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 7.

Images