Networking method, network management system, server and computer readable storage medium

Abstract

Embodiments of the present application relate to the technical field of communication, in particular to a networking method, a network management system, a server and a computer readable storage medium. The networking method comprises: collecting hardware operation diagnosis data of each network element in real time; predicting a current health condition of each network element according to the collected hardware operation diagnosis data of each network element and a pre-trained network element health prediction model; obtaining a network element whose current health condition matches a networking requirement according to the networking requirement and a preset matching relationship between the networking requirement and the network element health condition; and performing networking by using the network element whose current health condition matches the networking requirement. The networking method provided by the embodiments of the present application aims to automatically select a network element in a good health state to perform network slice networking, improve the performance of a slice network, meet actual networking requirements and business requirements of a user, and thus improve the user experience.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a networking method, a network management system, a server, and a computer-readable storage medium. Background Technology

[0002] With the rapid development of communication technology, 5G (5th Generation Mobile Communication Technology) has gradually begun to be commercialized and has reached a certain scale. 5G technology can use network slicing technology for networking. Network slicing technology is an on-demand networking method. Network slicing technology can slice the network for different application scenarios and services, which isolates the original physical network and decomposes it into several logical networks. Each slice is configured differently, so that the network used by users is smoother and provides users with high-quality services.

[0003] However, the creation of network slices requires the support of base station equipment (i.e., network elements). There are more than 10 million base station devices operating globally, distributed in tens of thousands of equipment rooms or signal towers in various cities and villages. Base station devices inevitably experience various failures during operation. If network slicing is based on network elements that have failed or have poor performance, the performance of the network will be greatly reduced, failing to meet the actual networking and service needs of users, and bringing a poor user experience. Summary of the Invention

[0004] The main objective of this application is to propose a networking method, network management system, server, and computer-readable storage medium, which aims to automatically select network elements in good health for network slicing, improve the performance of the sliced ​​network, meet the actual networking and business needs of users, and thus enhance the user experience.

[0005] To achieve the above objectives, embodiments of this application provide a networking method, the method comprising: real-time collection of hardware operation diagnostic data of each network element; predicting the current health status of each network element based on the collected hardware operation diagnostic data of each network element and a pre-trained network element health prediction model; obtaining network elements whose current health status matches the networking requirements based on networking requirements and a preset matching relationship between networking requirements and network element health status; and using the network elements whose current health status matches the networking requirements to form a network.

[0006] To achieve the above objectives, this application also provides a network management system, including a data acquisition module, a data storage module, an operation and maintenance analysis module, and a slice management module. The data acquisition module is used to collect hardware operation diagnostic data of each network element in real time and store the collected hardware operation diagnostic data of each network element in the data storage module. The operation and maintenance analysis module is used to obtain the collected hardware operation diagnostic data of each network element and a pre-trained network element health prediction model from the data storage module, and predict the current health status of each network element based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model. The slice management module is used to obtain network elements whose current health status matches the network requirements based on a preset matching relationship between network requirements and network element health status, and to use the network elements whose current health status matches the network requirements for network construction. The data storage module is used to store the hardware operation diagnostic data of each network element collected in real time by the data acquisition module, the pre-trained network element health prediction model, and the preset matching relationship between network requirements and network element health status.

[0007] To achieve the above objectives, embodiments of this application also provide a server, including: 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, the instructions being executed by the at least one processor to enable the at least one processor to perform the above-described networking method.

[0008] To achieve the above objectives, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described networking method.

[0009] The networking method, network management system, server, and computer-readable storage medium proposed in this application collect hardware operation diagnostic data of each network element in real time. Based on the collected hardware operation diagnostic data of each network element and a pre-trained network element health prediction model, the current health status of each network element is predicted. Then, based on the networking requirements and the preset matching relationship between networking requirements and network element health status, network elements whose current health status matches the networking requirements are obtained. Finally, network elements whose current health status matches the networking requirements are used for networking. Considering that various faults or problems are inevitable in operation, if network slicing is performed based on faulty or poor-performing network elements, the performance of the resulting sliced ​​network will be very poor and cannot meet the actual networking and service requirements of users. However, the embodiments of this application can determine the current health status of each network element before performing network slicing based on network elements, and automatically select network elements with good current health status that meet the networking requirements for network slicing, thereby improving the performance of each sliced ​​network, meeting the actual networking and service requirements of users, and thus improving the user experience. In addition, using a pre-trained network element health prediction model to predict the current health status of each network element can make the prediction of network element health status more convenient, accurate and efficient. Attached Figure Description

[0010] Figure 1 This is a flowchart of a networking method according to an embodiment of this application. Figure 1 ;

[0011] Figure 2 This is a flowchart of a training network element health prediction model provided in one embodiment of this application;

[0012] Figure 3 This is a flowchart of training a network element health prediction model based on hardware repair data obtained within a preset time period and hardware operation diagnostic data belonging to the same hardware as the hardware repair data, provided in one embodiment of this application.

[0013] Figure 4 This is a flowchart illustrating the output of the current health status of each network element to the operation and maintenance client, according to one embodiment of this application.

[0014] Figure 5 This is a schematic diagram of the structure of a network management system according to another embodiment of this application;

[0015] Figure 6 This is a schematic diagram of the structure of an electronic device according to another embodiment of this application. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the various embodiments of this application to help readers better understand this application. However, the technical solutions claimed in this application can be implemented even without these technical details and various changes and modifications based on the following embodiments. The division of the various embodiments below is for the convenience of description and should not constitute any limitation on the specific implementation of this application. The various embodiments can be combined with and referenced by each other without contradiction.

[0017] One embodiment of this application relates to a networking method applied to a network management system server. For ease of description, this embodiment and the following other embodiments will all be described using the server as an example. The implementation details of the networking method of this embodiment are described in detail below. The following content is only for the convenience of understanding the implementation details provided and is not necessary for implementing this solution.

[0018] The specific process of the networking method in this embodiment can be described as follows: Figure 1 As shown, it includes:

[0019] Step 101: Collect hardware operation diagnostic data of each network element in real time.

[0020] Specifically, the server can perform hardware operation diagnostics on the hardware of each network element in real time and collect hardware operation diagnostic data of each network element in real time.

[0021] In practical implementation, the hardware operation diagnostic data collected by the server for each network element includes diagnostic data of the network element's hardware during operation and environmental data that affects the hardware operation. Considering that the geographical locations of each network element vary greatly, the hardware of each network element may be affected by the external environment during operation, thereby affecting the performance of the network element's hardware and even causing the network element to malfunction. The hardware operation diagnostic data collected by the embodiments of this application not only includes diagnostic data of the network element's hardware during operation, but also includes environmental data that affects the hardware operation, which makes the collected hardware operation diagnostic data of each network element more scientific, accurate, and realistic, and in line with the actual working conditions of the network element.

[0022] In one example, the hardware operation diagnostic data of a network element includes more than 200 types of hardware operation diagnostic data, such as clock status data, resource utilization information, and light ranging. The server can collect tens of thousands of network element hardware operation diagnostic data every day and store them in the database.

[0023] In one example, the hardware of a network element can be a single board of the network element. The diagnostic data of the single board during operation can include, but is not limited to: the link status of the single board, the bit error rate of the single board, the power of the single board, the CPU utilization of the single board, and the temperature of the single board.

[0024] In one example, the hardware of a network element can be a single board of the network element. Environmental data that affects the operation of the single board of the network element may include, but is not limited to: the input voltage of the single board, the inlet and outlet temperatures of the single board, and the fan speed of the single board.

[0025] In one example, the network element's boards include: Baseband Unit (BBU) boards, Remote Radio Unit (RRU) boards, Centralized Unit (CU) boards, Distributed Unit (DU) boards, and Active Antenna Unit (AAU) boards, etc.

[0026] In one example, the hardware operation diagnostic data of each network element collected by the server can be shown in Table 1:

[0027] Table 1: Hardware Operation Diagnostic Data Table

[0028]

[0029] Where 1 indicates that the hardware operation diagnostic data is abnormal, and 0 indicates that the hardware operation diagnostic data is normal.

[0030] Step 102: Based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model, predict the current health status of each network element.

[0031] Specifically, after collecting hardware operation diagnostic data from each network element, the server can predict the current health status of each network element based on the collected hardware operation diagnostic data and a pre-trained network element health prediction model. The pre-trained network element health prediction model can be stored in the server's internal memory. This model can be trained by those skilled in the art according to actual needs, or it can be an open-source model obtained directly from the internet; this embodiment does not impose any specific limitations on it.

[0032] In practical implementation, the server can use the collected hardware operation diagnostic data of each network element as input data to input into the pre-trained network element health prediction model, obtain the fault probability of each network element output by the pre-trained network element health prediction model, and obtain the current health status of each network element based on the fault probability.

[0033] In one example, the input and output data of the pre-trained network element health prediction model can be shown in Table 2:

[0034] Table 2: Input and Output Data of the Pre-trained Network Element Health Prediction Model

[0035]

[0036] Where 1 indicates that the hardware operation diagnostic data is abnormal, and 0 indicates that the hardware operation diagnostic data is normal.

[0037] In one example, the pre-trained network element health prediction model can output the current health status of each network element as: very healthy, relatively healthy, moderate, and unhealthy.

[0038] Step 103: Based on the networking requirements and the preset matching relationship between networking requirements and network element health status, obtain the network elements whose current health status matches the networking requirements.

[0039] Specifically, after obtaining the predicted current health status of each network element, the server can obtain the network elements whose current health status matches the network requirements based on the network requirements and the preset matching relationship between the network requirements and the network element health status.

[0040] In one example, the server can determine the networking requirements before collecting hardware operation diagnostic data of each network element. That is, after obtaining the networking requirements, the server can collect hardware operation diagnostic data of each network element in real time.

[0041] In one example, the server can obtain the predicted current health status of each network element and then obtain the network configuration request to perform network configuration.

[0042] In one example, network requirements can be network speed requirements. The preset matching relationship between network requirements and network element health includes the matching relationship between network speed requirements and network element health. In this case, network speed requirements are directly proportional to the health of network elements, that is, the higher the network speed requirements, the better the health of the required network elements.

[0043] For example, if the server obtains the current health status of network elements as very healthy, relatively healthy, moderate, or unhealthy, the server will classify network speed requirements into three levels: high speed, medium speed, and low speed. If the network speed requirement is high speed, then a network element with a very healthy current health status will be matched; if the network speed requirement is medium speed, then a network element with a relatively healthy current health status will be matched; if the network speed requirement is low speed, then a network element with a moderate current health status will be matched. Network elements with an unhealthy current health status may be faulty, and the server will set network elements with an unhealthy health status not to participate in the networking process.

[0044] Step 104: Utilize network elements that match the current health status and networking requirements to form a network.

[0045] Specifically, after the server obtains the network elements whose current health status matches the network requirements based on the network requirements and the preset matching relationship between network requirements and network element health status, it can use the network elements whose current health status matches the network requirements to form a network.

[0046] In one example, the networking requirement can be a slicing networking requirement. The server can perform network slicing based on the network elements that match the current health status and networking requirements, resulting in several sliced ​​networks.

[0047] In this embodiment, the server collects hardware operation diagnostic data of each network element in real time. Based on the collected hardware operation diagnostic data and a pre-trained network element health prediction model, it predicts the current health status of each network element. Then, based on the networking requirements and the preset matching relationship between networking requirements and network element health status, it obtains network elements whose current health status matches the networking requirements. Finally, it uses network elements whose current health status matches the networking requirements to form a network. Considering that various faults or problems inevitably occur during operation, if network slicing is performed based on faulty or poor-performing network elements, the performance of the resulting sliced ​​network will be very poor and cannot meet the actual networking and service requirements of users. However, the embodiment of this application can determine the current health status of each network element before performing network slicing based on network elements, and automatically select network elements with good current health status that meet the networking requirements for network slicing, thereby improving the performance of each sliced ​​network, meeting the actual networking and service requirements of users, and thus improving the user experience. In addition, predicting the current health status of each network element based on a pre-trained network element health prediction model makes the prediction of network element health status more convenient, accurate, and efficient.

[0048] In one embodiment, the pre-trained network element health prediction model can be used as follows: Figure 2 The training process is carried out in the steps shown, specifically including:

[0049] Step 201: Obtain hardware repair data within a preset time period and hardware operation diagnostic data belonging to the same hardware as the hardware repair data.

[0050] In practical implementation, network element hardware inevitably encounters various faults or problems during operation. Telecommunication operators and hardware manufacturers often receive a large number of hardware that needs to be returned for repair. In the embodiments of this application, the server obtains hardware return data within a preset time period and obtains hardware operation diagnostic data belonging to the same hardware as the hardware return data from the work log or operation and maintenance database. The hardware return data can characterize whether the hardware is faulty.

[0051] In one example, the hardware repair data obtained by the server and the hardware operation diagnostic data belonging to the same hardware as the hardware repair data can be shown in Table 3:

[0052] Table 3: Hardware repair data and hardware operation diagnostic data belonging to the same hardware as the hardware repair data.

[0053]

[0054] In this context, 1 indicates that the hardware operation diagnostic data is abnormal, and 0 indicates that the hardware operation diagnostic data is normal. For the repair data column, 1 indicates that the hardware has a fault, and 0 indicates that the hardware does not have a fault.

[0055] Step 202: Train the network element health prediction model based on the hardware repair data obtained within a preset time period and the hardware operation diagnostic data belonging to the same hardware as the hardware repair data.

[0056] Specifically, after obtaining hardware repair data within a preset time period and hardware operation diagnostic data belonging to the same hardware as the hardware repair data, the server can train a network element health prediction model based on the hardware repair data within the preset time period and the hardware operation diagnostic data belonging to the same hardware as the hardware repair data.

[0057] In one example, the training steps of the network element health prediction model can be executed in a preset first cycle. The server obtains hardware repair data and hardware operation diagnostic data belonging to the same hardware as the hardware repair data within a preset time period. This preset time period can refer to the preset duration before the execution of the training steps of the network element health prediction model based on the hardware repair data and hardware operation diagnostic data belonging to the same hardware as the hardware repair data obtained within the preset time period. The network element health prediction model is updated periodically, which can make the prediction effect of the network element health prediction model more accurate and stable.

[0058] In this embodiment, the pre-trained network element health prediction model is trained through the following steps: acquiring hardware repair data and hardware operation diagnostic data belonging to the same hardware as the hardware repair data within a preset time period; training the network element health prediction model based on the hardware repair data and hardware operation diagnostic data belonging to the same hardware as the hardware repair data within the preset time period. Considering that the hardware repair data comes from front-line work and is real and reliable data, training the network element health prediction model based on the hardware repair data and hardware operation diagnostic data belonging to the same hardware as the hardware repair data can obtain a scientific, stable, and accurate network element health prediction model.

[0059] In one embodiment, the hardware repair data acquired within a preset time period includes repair data for multiple hardware components. The hardware operation diagnostic data belonging to the same hardware as the hardware repair data includes operation diagnostic data for multiple hardware components. The operation diagnostic data of each of the multiple hardware components, along with the repair data, serves as a training sample. The server trains a network element health prediction model based on the hardware repair data acquired within the preset time period and the hardware operation diagnostic data belonging to the same hardware as the hardware repair data. This model can be used as follows: Figure 3 The steps shown are implemented as follows:

[0060] Step 301: Input the operational diagnostic data from the training samples into the network element health prediction model, and obtain the rework data output from the intermediate layer of the network element health prediction model.

[0061] Specifically, the network element health prediction model includes an input layer, an intermediate layer, and an output layer. When training the network element health prediction model, the final output layer can be skipped first. The server inputs the running diagnostic data from the training samples into the input layer of the network element health prediction model and obtains the repair data output from the intermediate layer of the network element health prediction model.

[0062] In the specific implementation, each layer of the network element health prediction model is set with weights and biases. The server obtains the repair data output by the intermediate layer based on the operational diagnostic data input in the training samples and the weights and biases of each layer.

[0063] Step 302: Use the rework data in the training samples to verify the rework data output by the intermediate layer and obtain the verification value.

[0064] In the specific implementation, after the server obtains the rework data output by the intermediate layer of the network element health prediction model, it can use the rework data in the same training sample to verify the rework data output by the intermediate layer and obtain the verification value.

[0065] In one example, the server can call a preset cost function to verify the rework data output by the intermediate layer and obtain the verification value. The preset cost function can be set by those skilled in the art according to actual needs, and this embodiment does not specifically limit it.

[0066] Step 303: Adjust the model parameters of the network element health prediction model according to the verification value until the verification value indicates that the verification has passed.

[0067] In the specific implementation, after obtaining the verification value, the server can determine whether the verification value meets the preset verification standard. If the verification value meets the preset verification standard, the verification is confirmed to be successful. If the verification value does not meet the preset verification standard, the server can adjust the model parameters of the network element health prediction model and perform iterative training until the verification value indicates that the verification is successful.

[0068] In one example, the server can call a preset backpropagation algorithm to adjust the model parameters of the network element health prediction model, such as weights and biases, and perform iterative training based on the adjusted parameters. The preset backpropagation algorithm can be set by those skilled in the art according to actual needs, and this embodiment does not specifically limit it.

[0069] In this embodiment, the hardware repair data acquired within the preset time period includes repair data of multiple hardware components. The hardware operation diagnostic data belonging to the same hardware as the hardware repair data includes the operation diagnostic data of the multiple hardware components, and the operation diagnostic data of each of the multiple hardware components and the repair data serve as a training sample. The step of training the network element health prediction model based on the hardware repair data acquired within the preset time period and the hardware operation diagnostic data belonging to the same hardware as the hardware repair data includes: inputting the operation diagnostic data from the training sample into the network element health prediction model and acquiring the repair data output by the intermediate layer of the network element health prediction model; verifying the repair data output by the intermediate layer using the repair data from the training sample to obtain a verification value; adjusting the model parameters of the network element health prediction model according to the verification value until the verification value indicates that the verification is successful. The embodiments of this application can iteratively train the network element health prediction model based on a massive amount of training samples, continuously optimizing the model parameters and model structure, which can further improve the accuracy and stability of the network element health prediction model.

[0070] In one embodiment, the networking requirement is the networking requirement of 5G network slicing. The types of 5G network slicing networking requirements include enhanced mobile broadband (eMBB), ultra-reliable and low-latency communications (URLLC), and massive machine-type communications (mMTC). The health status of network elements includes a health score, which is used to represent the health status of network elements. The higher the health score, the better the network element performance. The preset matching relationship between networking requirements and network element health status may include: eMBB is matched with network elements whose health score is greater than or equal to a first threshold; uRLLC is matched with network elements whose health score is less than the first threshold but greater than or equal to a second threshold; and mMTC is matched with network elements whose health score is less than the second threshold but greater than or equal to a third threshold. The first threshold is greater than the second threshold, and the second threshold is greater than the third threshold.

[0071] In one example, the first threshold is 90 points, the second threshold is 80 points, and the third threshold is 70 points. If the network requirement is eMBB, the server can use network elements with a health score greater than or equal to 90 points for eMBB networking; if the network requirement is URLLC, the server can use network elements with a health score greater than or equal to 80 points and less than 90 points for URLLC networking; if the network requirement is mMTC, the server can use network elements with a health score greater than or equal to 70 points and less than 80 points for mMTC networking. Prioritizing network elements with higher scores for creating wireless slices can ensure a higher success rate for network deployment.

[0072] In another example, if the networking requirement is URLLC, the server can use network elements with a health score of 80 or higher for URLLC networking; if the networking requirement is mMTC, the server can use network elements with a health score of 70 or higher for mMTC networking. For some slicing scenarios where the health requirements of network elements are not high, network elements with slightly lower health scores can also be selected to try to meet the user's networking requirements.

[0073] In one embodiment, the step of the server predicting the current health status of each network element based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model can be executed in a preset second cycle.

[0074] In one embodiment, the step of the server predicting the current health status of each network element based on the collected hardware operation diagnostic data and a pre-trained network element health prediction model can be executed when a networking requirement is detected. Predicting the health status of network elements before networking is required allows for obtaining the current health status of the network elements, making the selection of network elements during networking more accurate, and thus ensuring that the actual performance of the network slice obtained after networking better matches the theoretical performance.

[0075] In one embodiment, after the server predicts the current health status of each network element based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model, it can also output the current health status of each network element to the operation and maintenance client. The embodiments of this application support the display of the health status of each network element, which makes it easier for operation and maintenance personnel to understand the current health status of each network element.

[0076] In one example, the server uses a health score to characterize the health status of network elements. The server can output information such as the current health score and fault location of each network element to the operation and maintenance client. For example, after the server detects network element A in City A, it determines that the health score of network element A in City A is 65 points. The server outputs the health score of network element A in City A as 65 points to the operation and maintenance client, and outputs the fault location information as: excessively high data center temperature, clock board reset, cabling board failure, community outage, and abnormal communication.

[0077] In one embodiment, the server outputs the current health status of each network element to the operation and maintenance client, which can be achieved by, for example... Figure 4 The steps shown are implemented as follows:

[0078] Step 401: Sort the health status of each network element according to its current health status.

[0079] In the specific implementation, after the server predicts the current health status of each network element based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model, it can sort the network elements according to their current health status. The server can sort them in order from worst to best health status or in order from best to worst health status.

[0080] In one example, the server uses health scores to characterize the health status of network elements. Network element A has a health score of 85, network element B has a health score of 23, network element C has a health score of 98, network element D has a health score of 67, and network element E has a health score of 76. The server sorts the network elements according to their health status from lowest to highest score as follows: network element B, network element D, network element E, network element A, and network element C.

[0081] Step 402: Present the ranking results of the health status of each network element to the operation and maintenance client in a visual manner.

[0082] In practice, the server can present the ranking results of the health status of each network element to the operation and maintenance client in a visual manner.

[0083] In one example, the server can present the health status ranking results of each network element in a table format on the operation and maintenance client interface.

[0084] In this embodiment, outputting the current health status of each network element to the operation and maintenance client includes: sorting the network elements according to their current health status; and presenting the sorting results of the network elements to the operation and maintenance client in a visual manner, which can more intuitively present the health status of each network element and provide a reference for operation and maintenance personnel to formulate network element health improvement work plans.

[0085] In one embodiment, after the server predicts the current health status of each network element based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model, it can perform operation and maintenance work on the network elements with poor current health status to improve the health status of the network elements and thus better meet the networking requirements.

[0086] In one example, the server represents the current health status of each network element using network element health level. After obtaining the network element health level of each network element, the server can perform maintenance operations on network elements whose health level is lower than a preset threshold according to the pre-stored operation and maintenance experience rules, thereby improving the health level of the network element. The preset threshold and the pre-stored operation and maintenance experience rules can be set by those skilled in the art according to actual needs, and the embodiments of this application do not specifically limit them.

[0087] In another example, the server represents the current health status of each network element using a network element health score. After obtaining the network element health score of each network element, the server can obtain the operation and maintenance operation instructions input by the operation and maintenance personnel. Based on the operation and maintenance operation instructions, the server can perform operation and maintenance operations on network elements whose health scores are lower than a preset threshold to improve the health score of the network element. The operation and maintenance operation instructions are formulated by the operation and maintenance personnel based on the analysis of the network element health scores and the hardware operation diagnostic data of each network element.

[0088] Another embodiment of this application relates to a network management system. The details of the network management system in this embodiment are described below. The following content is merely for ease of understanding and is not essential for implementing this example. Figure 5 This is a schematic diagram of the network management system described in this embodiment, including: a data acquisition module 501, a data storage module 502, an operation and maintenance analysis module 503, and a slice management module 504.

[0089] The data acquisition module 501 is connected to the data storage module 502. The data storage module 502 is connected to the operation and maintenance analysis module 503 and the slice management module 504 respectively. The operation and maintenance analysis module 503 is also connected to the slice management module 504.

[0090] The data acquisition module 501 is used to collect hardware operation diagnostic data of each network element in real time and store the collected hardware operation diagnostic data of each network element in the data storage module 502.

[0091] The operation and maintenance analysis module 503 is used to obtain the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model from the data storage module 502, and predict the current health status of each network element based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model.

[0092] The slice management module 504 is used to obtain network elements whose current health status matches the network requirements based on the preset matching relationship between network requirements and network element health status, and to use the network elements whose current health status matches the network requirements for networking.

[0093] The data storage module 502 is used to store the hardware operation diagnostic data of each network element collected in real time by the data acquisition module 501, the pre-trained network element health prediction model, and the preset matching relationship between network requirements and network element health status.

[0094] It is not difficult to see that this embodiment is a system embodiment corresponding to the above method embodiments, and this embodiment can be implemented in conjunction with the above method embodiments. The relevant technical details and technical effects mentioned in the above embodiments are still valid in this embodiment, and will not be repeated here to reduce repetition. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the above embodiments.

[0095] It is worth mentioning that all modules involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this invention, this embodiment does not introduce units that are not closely related to solving the technical problem proposed by this invention; however, this does not mean that other units are absent from this embodiment.

[0096] Another embodiment of this application relates to an electronic device, such as... Figure 6 As shown, it includes: at least one processor 601; and a memory 602 communicatively connected to the at least one processor 601; wherein the memory 602 stores instructions executable by the at least one processor 601, the instructions being executed by the at least one processor 601 to enable the at least one processor 601 to execute the networking methods in the above embodiments.

[0097] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0098] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0099] Another embodiment of this application relates to a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the method embodiments described above.

[0100] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0101] Those skilled in the art will understand that the above embodiments are specific examples of implementing the present invention, and in practical applications, various changes in form and detail may be made without departing from the spirit and scope of the present invention.

Claims

1. A networking method, characterized in that, include: Perform hardware operation diagnostics on the hardware of each network element and collect hardware operation diagnostic data of each network element in real time. Based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model, the current health status of each network element is predicted. Based on the networking requirements and the preset matching relationship between networking requirements and network element health status, obtain the network elements whose current health status matches the networking requirements; Networking is performed using network elements whose current health status matches the network requirements.

2. The networking method according to claim 1, characterized in that, The pre-trained network element health prediction model is trained through the following steps: Acquire hardware repair data within a preset time period and hardware operation diagnostic data belonging to the same hardware as the hardware repair data; A network element health prediction model is trained based on the hardware repair data obtained within a preset time period and the hardware operation diagnostic data belonging to the same hardware as the hardware repair data.

3. The networking method according to claim 2, characterized in that, The hardware repair data acquired within the preset time period includes repair data of multiple hardware components. The hardware operation diagnostic data belonging to the same hardware as the hardware repair data includes the operation diagnostic data of the multiple hardware components. The operation diagnostic data of each hardware component and the repair data are used as a training sample. The step of training a network element health prediction model based on hardware repair data acquired within a preset time period and hardware operation diagnostic data belonging to the same hardware as the hardware repair data includes: The operational diagnostic data from the training samples is input into the network element health prediction model, and the rework data output from the intermediate layer of the network element health prediction model is obtained. The rework data output by the intermediate layer is verified using the rework data in the training samples to obtain the verification value; The model parameters of the network element health prediction model are adjusted according to the verification value until the verification value indicates that the verification has passed.

4. The networking method according to claim 1, characterized in that, After predicting the current health status of each network element based on the collected hardware operation diagnostic data and the pre-trained network element health prediction model, the method further includes: Output the current health status of each network element to the operation and maintenance client.

5. The networking method according to claim 4, characterized in that, The step of outputting the current health status of each network element to the operation and maintenance client includes: Based on the current health status of each network element, the network elements are sorted according to their health status. The health status ranking results of each network element are presented to the operation and maintenance client in a visual manner.

6. The networking method according to claim 2, characterized in that, The training steps of the network element health prediction model are executed in a preset first cycle; The preset time period refers to the preset time period before the step of training the network element health prediction model based on the hardware repair data and hardware operation diagnostic data belonging to the same hardware as the hardware repair data obtained within the preset time period is executed.

7. The networking method according to claim 1, characterized in that, The step of predicting the current health status of each network element based on the collected hardware operation diagnostic data and the pre-trained network element health prediction model is executed in a preset second cycle; or, The step of predicting the current health status of each network element based on the collected hardware operation diagnostic data of each network element and the pre-trained network element health prediction model is executed when a networking requirement is detected.

8. The networking method according to claim 1, characterized in that, The networking requirements are those for 5G network slicing, and include the following requirement types: enhanced mobile broadband (eMBB), ultra-reliable low-latency communication (uRLLC), and massive machine-type communication (mMTC); the health status includes a health score. The matching relationships include: the eMBB is matched with network elements whose health score is greater than or equal to a first threshold; the uRLLC is matched with network elements whose health score is less than the first threshold but greater than or equal to a second threshold; and the mMTC is matched with network elements whose health score is less than the second threshold but greater than or equal to a third threshold; the first threshold is greater than the second threshold, and the second threshold is greater than the third threshold; wherein, the higher the health score, the better the network element performance.

9. The networking method according to claim 1, characterized in that, The hardware operation diagnostic data includes diagnostic data of the network element's hardware during operation and environmental data that affects the operation of the hardware.

10. The networking method according to claim 9, characterized in that, The hardware is the single board of the network element; The diagnostic data of the hardware during operation includes: the link status of the single board, the bit error rate of the single board, the power of the single board, the CPU utilization of the single board, and the temperature of the single board; The environmental data that affects the operation of the hardware includes: the input voltage of the board, the inlet and outlet temperatures of the board, and the fan speed of the board.

11. A network management system, characterized in that, It includes a data acquisition module, a data storage module, an operation and maintenance analysis module, and a slice management module: The data acquisition module is used to perform hardware operation diagnosis on the hardware of each network element, collect hardware operation diagnosis data of each network element in real time, and store the collected hardware operation diagnosis data of each network element into the data storage module. The operation and maintenance analysis module is used to obtain the collected hardware operation diagnostic data and pre-trained network element health prediction model of each network element from the data storage module, and predict the current health status of each network element based on the collected hardware operation diagnostic data and the pre-trained network element health prediction model. The slice management module is used to obtain network elements whose current health status matches the network requirements based on the preset matching relationship between network requirements and network element health status, and to use the network elements whose current health status matches the network requirements to form a network. The data storage module is used to store the hardware operation diagnostic data of each network element collected in real time by the data acquisition module, the pre-trained network element health prediction model, and the preset matching relationship between network requirements and network element health status.

12. A server, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the networking method as described in any one of claims 1 to 10.

13. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the networking method according to any one of claims 1 to 10.

Images