5G Slice Instance Backup Method, Apparatus, Computing Device, and Storage Medium

Abstract

The present invention discloses a backup method, apparatus, computing device and storage medium for a 5G slice instance. The method comprises: receiving a backup method formulation request triggered by a user's request to create a new slice instance; obtaining text data of the slice instance's service level agreement requirements, application scenario description and backup requirements from the backup method formulation request; inputting the text data into a trained backup method formulation device, and outputting a backup method combination for the slice instance within a preset period; and performing backup processing on the data of the slice instance according to the backup method combination within the preset period. It can be seen that the scheme of the present invention can efficiently and accurately generate a backup method combination that meets both SLA requirements and backup requirements and is suitable for its specific application scenario through a backup method formulation device, thereby facilitating backup processing.

Description

Technical Field

[0001] The present invention relates to the field of 5G slicing technology, and in particular to a backup device, computing equipment, and computer storage medium for a 5G slicing instance. Background Art

[0002] A network slice is an end-to-end collection of logical functions and the physical or virtual resources they require, encompassing access, transport, and core networks. It can be considered a virtualized "private network" within a 5G network. Built on a unified NFV infrastructure, network slicing enables cost-effective and efficient operations. Network slicing technology logically isolates communication networks, allowing network components and functions to be configured and reused within each network slice to meet specific industry application requirements.

[0003] A Network Slice Instance (NSI) is a real, running logical network that meets certain network characteristics or service requirements. A network slice instance may provide one or more services. Network slice instances can be created by a network management system (NMS). A single NMS may create multiple network slice instances and manage them simultaneously, including performance monitoring and fault management during the operation of the network slice instances. When multiple network slice instances coexist, they may share some network resources and network functions.

[0004] Currently, the backup method for 5G slice instances is mainly manually formulated based on expert experience. However, due to the diverse types of slice applications and the different needs of slice users, manually formulating the backup method for each slice instance is time-consuming, labor-intensive, and inefficient. It also easily leads to inaccurate backup methods and is unable to meet the personalized backup needs of slice users. Summary of the Invention

[0005] In view of the above problems, embodiments of the present invention are proposed to provide a backup method, apparatus, computing device and storage medium for a 5G slice instance that overcomes the above problems or at least partially solves the above problems.

[0006] According to one aspect of an embodiment of the present invention, a method for backing up a 5G slice instance is provided, including:

[0007] Receive a backup mode creation request triggered by a user's request to create a new slice instance;

[0008] Obtaining text data of the service level agreement requirements, application scenario description, and backup requirements of the slice instance from the backup method formulation request;

[0009] Input the text data into the trained backup mode setter, and output the backup mode combination of the slice instance within a preset period;

[0010] Within a preset period, the data of the slice instance is backed up according to the backup method combination.

[0011] According to another aspect of an embodiment of the present invention, a backup device for a 5G slice instance is provided, including:

[0012] A receiving module, adapted to receive a backup mode formulation request triggered by a user's request to create a new slice instance;

[0013] An extraction module adapted to obtain text data of a service level agreement requirement, an application scenario description, and a backup requirement of a slice instance from the backup mode formulation request;

[0014] A formulation module, adapted to input the text data into a trained backup mode formulater, and output a backup mode combination for the slice instance within a preset period;

[0015] The backup module is adapted to perform backup processing on the data of the slice instance according to the backup method combination within a preset period.

[0016] According to another aspect of an embodiment of the present invention, there is provided a computing device, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus;

[0017] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the backup method of the above-mentioned 5G slice instance.

[0018] According to another aspect of an embodiment of the present invention, a computer storage medium is provided, wherein the storage medium stores at least one executable instruction, and the executable instruction enables a processor to perform operations corresponding to the backup method of the above-mentioned 5G slicing instance.

[0019] According to the backup method, apparatus, computing device, and storage medium of the 5G slice instance according to the embodiment of the present invention, upon receiving a new creation request for a slice instance, the backup method is triggered to be specified, and the text data of the SLA requirements, backup requirements, and application scenario description of the slice instance are obtained, and then input into the backup method maker for backup method prediction. Then, a backup method combination within a preset period can be output, and the backup method combination can subsequently be used to complete the backup that meets personalized needs. It can be seen that the scheme of the present invention can use the backup method maker to formulate a backup method combination within a period T that meets both the SLA requirements and backup requirements of the slice instance and is suitable for its specific application scenario, thereby realizing on-demand personalized and differentiated backup processing of slice instance data.

[0020] The above description is only an overview of the technical solutions of the embodiments of the present invention. In order to more clearly understand the technical means of the embodiments of the present invention, they can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiments of the present invention more obvious and easy to understand, the specific implementation methods of the embodiments of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the embodiments of the present invention. The same reference numerals are used throughout the accompanying drawings to denote the same components. In the accompanying drawings:

[0022] Figure 1 A flowchart of a method for backing up a 5G slice instance provided by an embodiment of the present invention is shown;

[0023] Figure 2 A flowchart of a method for backing up a 5G slice instance provided by another embodiment of the present invention is shown;

[0024] Figure 3 A structural diagram of a backup mode setter constructed in a specific embodiment is shown;

[0025] Figure 4 The following shows the process of generating a personalized backup method in a specific example of the present invention;

[0026] Figure 5 A schematic structural diagram of a backup device for a 5G slicing instance provided in an embodiment of the present invention is shown;

[0027] Figure 6 A schematic structural diagram of a computing device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0028] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0029] Before implementing the embodiments of the present invention, several terms involved in this document are first explained as follows:

[0030] 1) Slice management architecture: mainly composed of CSMF, NSMF and NSSMF.

[0031] Among them, CSMF (Communication Service Management Function) completes the ordering and processing of user business communication service needs, is responsible for converting the communication service needs of operators / third-party customers into demand for network slicing, and sends the demand for network slicing to NSMF through the interface between NSMF (such as creation, termination, modification of network slice instance requests, etc.), and obtains network slice management data (such as performance, fault data, etc.) from NSMF.

[0032] Among them, NSMF (Network Slice Management Function) is responsible for receiving network slicing requirements sent by CSMF, managing the life cycle, performance, and faults of network slice instances, orchestrating the composition of network slice instances, decomposing the requirements of network slice instances into the requirements of each network slice subnet instance or network function, and sending network slice subnet instance management requests to each NSSMF.

[0033] In addition, NSSMF (Network Slice Subnet Management Function) receives network slice subnet deployment requirements issued by NSMF, manages network slice subnet instances, orchestrates the composition of network slice subnet instances, maps the SLA requirements of network slice subnets to the QoS requirements of network services, and issues network service deployment requests to the NFVO system in the ETSI NFV domain.

[0034] 2) Encoder-Decoder Neural Network: This is a method of organizing recurrent neural networks, primarily used to solve sequence prediction problems with multiple inputs or multiple outputs. It consists of an encoder and a decoder. The encoder encodes the input sequence word by word into a fixed-length vector, the context vector. The decoder reads the encoder's output context vector and generates the output sequence.

[0035] 3) Attention mechanism: It solves the limitations of the encoder-decoder structure. First, it provides the decoder with richer context obtained from the encoder. The encoder passes more data to the decoder. Compared with the traditional model, the encoder only passes the last hidden state of the encoding stage, while the encoder passes all hidden states to the decoder in the attention mechanism model. At the same time, attention provides a learning mechanism. When predicting the output sequence at each time step, the decoder can learn where to focus in the richer context. The attention network assigns an attention weight to each input. The more relevant the input is to the current operation, the closer the attention weight is to 1, and vice versa. These attention weights are recalculated at each output step.

[0036] 4) Long Short-Term Memory (LSTM) neurons: Long short-term memory (LSTM) is a special type of recurrent neural network, where the same neural network is reused. LSTM can learn long-term dependencies and retain information over time by controlling how long cached values ​​are retained, making it suitable for learning long sequences. Each neuron has four inputs and one output. Each neuron contains a cell that stores the memory value. Each LSTM neuron contains three gates: a forget gate, an input gate, and an output gate. LSTM neural networks are highly effective at learning long sequences.

[0037] 5) Slice instance backup methods: usually include full backup, incremental backup, and differential backup.

[0038] A full backup is a complete copy of all data or applications at a specific point in time. Performing a full backup of the system at regular intervals allows you to restore to the state at the time of the last backup using the previous backup if a system failure occurs during the backup interval and data is lost. For example, back up the entire system using one tape on Monday, another on Tuesday, and so on. This backup strategy offers the advantage of providing the most comprehensive and complete backup data. In the event of a data loss disaster, all data can be restored using a single tape (the backup tape from the day before the disaster).

[0039] Incremental backups involve performing a full backup first, followed by backups at shorter intervals, backing up only the data that changed during that time. This allows recovery to the previous full backup in the event of data loss. Restoring each daily backup, one at a time, allows recovery to the previous day's status. For example, a full backup can be performed on Sunday, followed by backups of only the new or modified data for the following six days. This backup strategy offers the advantages of fast backup speeds, no duplicate backup data, conserved tape space, and reduced backup time.

[0040] And, differential backup: refers to the backup of files that are added or modified between the time of a full backup and the time of a differential backup. When performing a recovery, we only need to restore the first full backup and the last differential backup. Differential backup avoids the shortcomings of the other two backup strategies while also possessing their respective advantages. First, it has the advantages of incremental backup in a short time and saving disk space; second, it has the characteristics of full backup recovery requiring fewer tapes and a shorter recovery time. The system administrator only needs two tapes, namely the full backup tape and the differential backup tape from the day before the disaster, to restore the system.

[0041] The specific embodiment of the present invention will be described in detail below:

[0042] Figure 1 FIG1 shows a flow chart of a backup method for a 5G slice instance provided by an embodiment of the present invention. Figure 1 As shown, the method includes the following steps:

[0043] Step S110: Receive a backup mode creation request triggered by a user's request to create a new slice instance.

[0044] The user initiates a request to create a new slice instance. When the new request is received, it can trigger the formulation of a backup method so that in the subsequent use of the slice instance, backup can be performed according to the formulated backup method.

[0045] Step S120: Obtain text data of the service level agreement requirements, application scenario description, and backup requirements of the slice instance from the backup method formulation request.

[0046] Among them, the service level agreement requirement is the SLA (Service-Level Agreement) requirement, the application scenario description refers to the description of the scenario where the slice instance is applicable, and the backup requirement reflects the demand for timeliness of data recovery, for example, how long it takes to complete data recovery.

[0047] In addition, the backup method specifies that the request carries text data of the slice instance's SLA requirements, application scenario description, and backup requirements.

[0048] Step S130: Input the text data into the trained backup method setter, and output the backup method combination of the slice instance within a preset period.

[0049] The backup method combination consists of backup methods for various time periods within a preset cycle. By inputting text data into the backup method generator, the backup method generator can predict a backup method combination within cycle T that meets both the SLA requirements and backup requirements of the slice instance and is suitable for its specific application scenario, so as to subsequently perform backup processing for the slice instance.

[0050] Step S140: within a preset period, back up the data of the slice instance according to the backup method combination.

[0051] Back up data using the backup methods listed in the backup method combination. For example, if the backup method combination is {Full Backup, Incremental Backup, Incremental Backup, Incremental Backup}, it means that except for the first backup, a full backup will be used, and all subsequent backups will be incremental backups.

[0052] According to the backup method for 5G slice instances provided in this embodiment, upon receiving a new request for a slice instance, the backup method is triggered to be specified, and the text data describing the SLA requirements, backup requirements, and application scenarios of the slice instance are obtained. This text data is then input into the backup method maker for backup method prediction, and a backup method combination within a preset period can be output. Subsequently, this backup method combination can be used to complete a backup that meets personalized requirements. It can be seen that in this embodiment, a backup method combination within a period T that simultaneously meets the SLA requirements and backup requirements of the slice instance and is suitable for its specific application scenario can be formulated by the backup method maker, thereby achieving on-demand personalized and differentiated backup processing of slice instance data.

[0053] Figure 2 FIG1 shows a flowchart of a backup method for a 5G slice instance provided by another embodiment of the present invention. Figure 2 As shown, the method includes the following steps:

[0054] Step S210: training the backup mode setter.

[0055] The backup mode generator is a codec neural network structure, including an encoder and a decoder. The specific training steps are as follows:

[0056] First, obtain the text data samples of the service level agreement requirements, application scenario descriptions, and backup requirements of each historical slice instance, and annotate the backup methods of each historical slice instance to obtain the annotation values ​​of the backup method combinations of the corresponding text data samples. In practice, the training data set composed of the text data samples and annotation values ​​needs to be cleaned and serialized, retaining all punctuation marks. If the text is in Chinese, the text is segmented; if the text is in English, the letters are unified into lowercase. At the same time, each word is indexed (tokenized) so that each text segment is converted into an index number, and zeros are added to the sequences that do not reach the maximum text length. Among them, the longest length L of the above three types of text data is taken as its index sequence length, and its dictionary size is taken as input_vocab_size. The longest length M of the corresponding slice backup strategy set is taken as its index sequence length, and its dictionary size is taken as backup_vocab_size.

[0057] Then, using a text data sample as training input data, the text data sample is input into the encoder, which performs feature extraction on the text data sample and merges it to obtain a context vector; the context vector is input into the decoder, which performs attention aggregation on the context vector to generate a prediction value of the backup method combination of each historical slice instance within a preset period. Furthermore, the encoder extracts features from the sequences corresponding to the three text data of the SLA requirements, application scenario description, and slice backup requirements of each historical slice instance, and encodes them separately into three fixed-length context vectors. The encoder merges the three context vectors to obtain a context vector; the merged context vector is then input into the decoder, and the attention decoder performs attention aggregation on the learned features to generate a prediction value of a backup method combination (a combination of the three backup methods of full backup, incremental backup, and differential backup) within a period T that meets both the SLA requirements and backup requirements of the slice instance and is suitable for its specific application scenario.

[0058] In addition, the labeled value of the backup mode combination is used as the training output data, the error between the predicted value and the labeled value is calculated, and the parameters of the encoder and decoder are adjusted according to the error until the error condition is met, then a backup mode maker is trained to obtain the backup mode maker, wherein the backup mode maker is a model composed of the parameters of the encoder and decoder when the error condition is met.

[0059] Figure 3 The following diagram shows the structure of a backup mode generator constructed in a specific embodiment. In this structure, a codec neural network based on long-short-term memory neurons is constructed, which also includes an encoder and a decoder. The structure of this structure and the training process using this structure are as follows:

[0060] 1) Encoder: Extract features of the SLA requirements, application scenario description, and backup requirement sequence of each slice instance, encode them separately into three fixed-length context vectors, and merge the three context vectors into one context vector h through the merging layer.

[0061] The first layer is the input layer: it inputs the SLA requirements, application scenario descriptions, and slice backup requirement sequences of the indexed slice instances. The length of each information sequence is L, so the shape of the output data of this layer is (None, L).

[0062] The second layer is the embedding layer: word embedding is used to convert each word into a vector. The input data dimension is input_vocab_size, and the output is set to the space vector that needs to be converted to 128 dimensions. The input sequence length is L, so the shape of the output data of this layer is (None, L, 128). The function of this layer is to perform vector mapping on the input words, converting the index of each word into a 128-dimensional fixed-shape vector. Among them, the SLA requirement sequence of the i-th slice application can be expressed as {S1i, S2i, S3i, ..., SLi}, the application scenario description sequence can be expressed as {V1i, V2i, V3i, ..., VLi}, and the backup requirement sequence can be expressed as {B1i, B2i, B3i, ..., BLi};

[0063] The third layer is the LSTM encoding layer: it contains three parallel LSTM layers, each with 128 LSTM neurons, and the activation function is set to "relu". The shape of the output data of this layer is (None, L, 128), which is encoded into three fixed-length context vectors;

[0064] The fourth layer is the concatenation layer: the three fixed-length context vectors are concatenated according to the column dimension to form a fixed-length context vector h;

[0065] 2) Decoder: The attention decoder aggregates the learned features to generate a backup strategy combination within a period T that meets both the SLA requirements and backup requirements of the slice instance and is suitable for its specific application scenario.

[0066] The fifth layer is the attention LSTM decoding layer, which contains 256 LSTM neurons and uses the relu activation function. The output data of this layer has a shape of (None, L, 256).

[0067] The sixth fully connected (Dense) layer (output layer) contains backup_vocab_size Dense fully connected neurons, with the activation function set to "softmax." The softmax output is fed into a multi-class cross-entropy loss function. The output data of this layer has a shape of (None, backup_vocab_size). This layer generates a backup method combination (a combination of full backup, incremental backup, and differential backup) within a period T that meets both the SLA requirements and backup requirements of the slice instance and is suitable for its specific application scenario. The output backup method combination is represented as {R1i, R2i, R3i, ..., RMi}.

[0068] During model training, the number of training rounds was set to 1000 (epochs = 1000), the batch size was set to 100 (batch_size = 100), categorical crossentropy was selected as the loss function (i.e., objective function) (loss = 'categorical_crossentropy'), and the Adam optimizer was selected for the gradient descent optimization algorithm to improve the learning speed of traditional gradient descent (optimizer = 'adam'). The objective function was calculated by comparing it with the results of the correct slicing backup strategy, and gradient descent was used to gradually find the weight value that minimized the objective function. The model weight after training convergence was used as the slice backup method setter.

[0069] Figure 4 The following figure shows the generation process of the personalized backup method in a specific example of the present invention. Figure 4 The example shown is used for illustration.

[0070] Step S220: Receive a backup mode creation request triggered by a new request for a slice instance from the user.

[0071] correspond Figure 4 In steps 1 and 2, the slice user sends a new slice instance creation request to CSMF (communication service management function module), and CSMF initiates a backup mode creation request for the slice instance to the backup mode creator. In other words, the backup mode creator receives the backup mode creation request triggered by CSMF based on the user's new slice instance creation request.

[0072] Step S230: Obtain text data of the service level agreement requirements, application scenario description, and backup requirements of the slice instance from the backup method formulation request.

[0073] The service level agreement requirements include, but are not limited to, one or more of the following requirements: security (or privacy, e.g., strong, medium, weak), visibility (or manageability), reliability (or availability, e.g., 99.999%), service type, air interface, customized network function (the aforementioned three items are specific service characteristics), latency (e.g., less than 5ms), throughput, packet loss rate, call drop rate, service scope, user scale, isolation (e.g., strong, medium, weak), access method, and max TP / site (e.g., 5Gbps);

[0074] In addition, the application scenario description includes but is not limited to at least one of the following scenarios: Internet of Vehicles, industrial control, intelligent manufacturing, intelligent transportation and logistics, environmental monitoring, intelligent meter reading, smart agriculture, video live broadcast and sharing, virtual reality, cloud access, and high-speed mobile Internet access.

[0075] In addition, the backup requirements include Recovery Time Objective (RTO) and Recovery Point Object (RPO). RTO refers to how long an application can be interrupted or shut down without causing significant damage to the business. Some slice applications may be down for several days without serious consequences, while some high-priority applications can only be stopped for a few seconds, otherwise it will be difficult for the enterprise and customers to cope and cause business loss. Based on this, the recovery time objective can be set according to the priority of the slice application carried by the slice instance; RPO refers to the enterprise's loss tolerance, the amount of data that may be lost before causing significant damage to the business. This target is expressed as a time measurement from the loss event to the most recent previous backup.

[0076] Step S240: Input the text data into the trained backup method setter, and output the backup method combination of the slice instance within a preset period.

[0077] The backup mode combination is a combination of full backup, incremental backup and / or differential backup.

[0078] correspond Figure 4 3 to 5, Figure 4 The data preprocessing module (equivalent to the word embedding layer) converts the three types of text data into integer sequences, which are then input into the backup method customization module, and the backup method combination is output. The backup method combination is then output to the NSMF (network slice management function module) of the slice management architecture.

[0079] Step S250: within a preset period, back up the data of the slice instance according to the backup method combination.

[0080] According to the backup method of the 5G slice instance provided in this embodiment, the attention encoding and decoding neural network can focus on the relevant parts of the input sequence as needed. The encoder extracts features of the SLA requirements, application scenario description, and backup requirement sequence of the slice instance, and merges them to obtain a context vector and input it into the decoder. The attention decoder aggregates the learned features to generate a backup method combination (a combination of the three backup methods of full backup, incremental backup, and differential backup) within a period T that can simultaneously meet the SLA requirements and backup requirements of the slice instance and is suitable for its specific application scenario. The backup policy of the slice instance is imported into the NSMF, and the NSMF implements backup for the slice according to the formulated slice instance backup strategy plan, thereby realizing on-demand personalization of the slice instance, customization of differentiated backup methods, and data backup.

[0081] Figure 5 FIG1 shows a schematic diagram of the structure of the backup device of the 5G slice instance provided by the embodiment of the present invention. Figure 5 As shown, the device includes:

[0082] A receiving module 510 is adapted to receive a backup mode formulation request triggered by a user's request to create a new slice instance;

[0083] An extraction module 520 is adapted to obtain text data of a service level agreement requirement, an application scenario description, and a backup requirement of a slice instance from the backup mode formulation request;

[0084] A formulation module 530 is adapted to input the text data into a trained backup mode formulater and output a backup mode combination for the slice instance within a preset period;

[0085] The backup module 540 is adapted to perform backup processing on the data of the slice instance according to the backup method combination within a preset period.

[0086] In an optional manner, the service level agreement requirements include one or more of the following requirements: security, visibility, reliability, service type, air interface, customized network function, latency, throughput, packet loss rate, call drop rate, service scope, user scale, isolation, and access method;

[0087] In addition, the application scenario description includes at least one of the following scenarios: Internet of Vehicles, industrial control, intelligent manufacturing, intelligent transportation and logistics, environmental monitoring, intelligent meter reading, smart agriculture, video live broadcast and sharing, virtual reality, cloud access, and high-speed mobile Internet access.

[0088] In an optional manner, the backup requirement includes a recovery time objective and a recovery point objective.

[0089] In an optional manner, the backup mode combination is a combination of full backup, incremental backup and / or differential backup.

[0090] In an optional manner, the backup mode formulater is a codec neural network structure, including an encoder and a decoder;

[0091] The device further comprises a training module adapted to:

[0092] Obtain text data samples of the service level agreement requirements, application scenario descriptions, and backup requirements of each historical slice instance, and annotate the backup method of each historical slice instance to obtain an annotation value of the backup method combination corresponding to the text data sample;

[0093] Inputting the text data sample into an encoder, the encoder extracts features from the text data sample and merges them to obtain a context vector;

[0094] The context vector is input to a decoder, and the decoder performs attention aggregation on the context vector to generate a prediction value of the backup mode combination of each historical slice instance within a preset period;

[0095] The error between the predicted value and the marked value is calculated, and the parameters of the encoder and decoder are adjusted according to the error until the error condition is met, and then the backup mode setter is trained.

[0096] In an optional manner, the training module is further adapted to:

[0097] The encoder extracts features from the sequences corresponding to the three text data and encodes them into three fixed-length context vectors.

[0098] The encoder merges the three context vectors to obtain a context vector.

[0099] In an optional manner, the receiving module is further adapted to: receive a backup mode formulation request triggered by the communication service management function module of the slice management structure according to a new creation request of a slice instance of the user;

[0100] The designated module is further adapted to: output the backup mode combination to the network slice management function module of the slice management architecture.

[0101] An embodiment of the present invention provides a non-volatile computer storage medium, which stores at least one executable instruction, and the computer executable instruction can execute the backup method of the 5G slice instance in any of the above method embodiments.

[0102] Figure 6The schematic diagram of the structure of the computing device provided by the embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the computing device.

[0103] like Figure 6 As shown, the computing device may include: a processor (processor) 602 , a communications interface (Communications Interface) 604 , a memory (memory) 606 , and a communication bus 608 .

[0104] Processor 602, communication interface 604, and memory 606 communicate with each other via communication bus 608. Communication interface 604 is used to communicate with other devices, such as clients or other server network elements. Processor 602 is used to execute program 610, which may specifically perform the steps of the embodiment of the method for backing up a 5G slice instance for a computing device.

[0105] Specifically, the program 610 may include program codes, which include computer operation instructions.

[0106] Processor 602 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in a computing device may be processors of the same type, such as one or more CPUs, or processors of different types, such as one or more CPUs and one or more ASICs.

[0107] The memory 606 is used to store the program 610. The memory 606 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0108] The program 610 may be specifically configured to enable the processor 602 to perform the following operations:

[0109] Receive a backup mode creation request triggered by a user's request to create a new slice instance;

[0110] Obtaining text data of the service level agreement requirements, application scenario description, and backup requirements of the slice instance from the backup method formulation request;

[0111] Input the text data into the trained backup mode setter, and output the backup mode combination of the slice instance within a preset period;

[0112] Within a preset period, the data of the slice instance is backed up according to the backup method combination.

[0113] In an optional manner, the service level agreement requirements include one or more of the following requirements: security, visibility, reliability, service type, air interface, customized network function, latency, throughput, packet loss rate, call drop rate, service scope, user scale, isolation, and access method;

[0114] In addition, the application scenario description includes at least one of the following scenarios: Internet of Vehicles, industrial control, intelligent manufacturing, intelligent transportation and logistics, environmental monitoring, intelligent meter reading, smart agriculture, video live broadcast and sharing, virtual reality, cloud access, and high-speed mobile Internet access.

[0115] In an optional manner, the backup requirement includes a recovery time objective and a recovery point objective.

[0116] In an optional manner, the backup mode combination is a combination of full backup, incremental backup and / or differential backup.

[0117] In an optional manner, the backup mode formulater is a codec neural network structure, including an encoder and a decoder;

[0118] The program 610 further causes the processor 602 to perform the following operations:

[0119] Obtain text data samples of the service level agreement requirements, application scenario descriptions, and backup requirements of each historical slice instance, and annotate the backup method of each historical slice instance to obtain an annotation value of the backup method combination corresponding to the text data sample;

[0120] Inputting the text data sample into an encoder, the encoder extracts features from the text data sample and merges them to obtain a context vector;

[0121] The context vector is input to a decoder, and the decoder performs attention aggregation on the context vector to generate a prediction value of the backup mode combination of each historical slice instance within a preset period;

[0122] The error between the predicted value and the marked value is calculated, and the parameters of the encoder and decoder are adjusted according to the error until the error condition is met, and then the backup mode setter is trained.

[0123] In an optional manner, the program 610 further causes the processor 602 to perform the following operations:

[0124] The encoder extracts features from the sequences corresponding to the three text data and encodes them into three fixed-length context vectors.

[0125] The encoder merges the three context vectors to obtain a context vector.

[0126] In an optional manner, the program 610 further causes the processor 602 to perform the following operations:

[0127] The communication service management function module of the receiving slice management structure generates a backup mode formulation request triggered by a new slice instance creation request of the user;

[0128] The backup method combination is output to the network slice management function module of the slice management architecture.

[0129] The algorithm or display provided herein is not inherently related to any particular computer, virtual system or other device. Various general-purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the embodiment of the present invention is not directed to any specific programming language. It should be understood that various programming languages ​​can be utilized to implement the content of the embodiment of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the embodiment of the present invention.

[0130] In the description provided herein, numerous specific details are described. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques are not shown in detail so as not to obscure the understanding of this description.

[0131] Similarly, it should be understood that in order to streamline the embodiments of the invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the invention, various features of the embodiments of the invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed approach should not be interpreted as reflecting an intention that the claimed embodiments of the invention require more features than are expressly recited in each claim. Rather, as reflected in the claims below, inventive aspects lie in less than all of the features of the individual embodiments disclosed above. Accordingly, the claims that follow the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of the invention.

[0132] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition may be divided into multiple submodules or subunits or subcomponents. All features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed herein may be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0133] Furthermore, those skilled in the art will appreciate that although some embodiments herein include certain features included in other embodiments but not other features, combinations of features from different embodiments are intended to be within the scope of the present invention and to form different embodiments. For example, in the claims below, any of the claimed embodiments may be used in any combination.

[0134] The various component embodiments of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components according to the embodiments of the present invention. The embodiments of the present invention can also be implemented as a device or apparatus program (e.g., a computer program and a computer program product) for executing part or all of the methods described herein. Such a program implementing an embodiment of the present invention can be stored on a computer-readable medium, or can have the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0135] It should be noted that the above embodiments illustrate rather than limit the embodiments of the invention, and that a person skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between brackets should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The embodiments of the invention may be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names. The steps in the above embodiments should not be understood as limiting the order of execution unless otherwise specified.

Claims

1. A method for backing up a 5G slice instance, comprising: Receive a backup mode creation request triggered by a user's request to create a new slice instance; Obtaining text data of the service level agreement requirements, application scenario description, and backup requirements of the slice instance from the backup method formulation request; Input the text data into a trained backup mode setter, and output a backup mode combination for the slice instance within a preset period; wherein the backup mode combination is a combination of at least two of full backup, incremental backup, and differential backup; The backup mode setter is a codec neural network structure, including an encoder and a decoder; the backup mode setter is trained by the following steps: Obtain text data samples of the service level agreement requirements, application scenario descriptions, and backup requirements of each historical slice instance, and annotate the backup method of each historical slice instance to obtain an annotation value of the backup method combination corresponding to the text data sample; Inputting the text data sample into an encoder, the encoder extracts features from the text data sample and merges them to obtain a context vector; The context vector is input to a decoder, and the decoder performs attention aggregation on the context vector to generate a prediction value of the backup mode combination of each historical slice instance within a preset period; Calculating the error between the predicted value and the marked value, and adjusting the parameters of the encoder and decoder according to the error until the error condition is met; Within a preset period, the data of the slice instance is backed up according to the backup method combination.

2. The method according to claim 1, wherein The service level agreement requirements include one or more of the following: security, visibility, reliability, service type, air interface, customized network function, latency, throughput, packet loss rate, call drop rate, service scope, user scale, isolation, and access method; In addition, the application scenario description includes at least one of the following scenarios: Internet of Vehicles, industrial control, intelligent manufacturing, intelligent transportation and logistics, environmental monitoring, intelligent meter reading, smart agriculture, video live broadcast and sharing, virtual reality, cloud access, and high-speed mobile Internet access.

3. The method according to claim 1, wherein The backup requirements include recovery time objective and recovery point objective.

4. The method according to claim 1, wherein The encoder extracts features from the text data sample and combines them to obtain a context vector, further comprising: The encoder extracts features from the sequences corresponding to the three text data and encodes them into three fixed-length context vectors. The encoder merges the three context vectors to obtain one context vector.

5. The method according to claim 1, wherein The receiving of the backup mode formulation request triggered by the user's request to create a new slice instance is specifically as follows: The communication service management function module of the receiving slice management structure generates a backup mode formulation request triggered by a new slice instance creation request of the user; The output of obtaining the backup mode combination of the slice instance within a preset period is specifically: outputting the backup mode combination to the network slice management function module of the slice management architecture.

6. A backup device for a 5G slice instance, comprising: A receiving module, adapted to receive a backup mode formulation request triggered by a user's request to create a new slice instance; An extraction module adapted to obtain text data of a service level agreement requirement, an application scenario description, and a backup requirement of a slice instance from the backup mode formulation request; A formulation module is adapted to input the text data into a trained backup mode formulater and output a backup mode combination for the slice instance within a preset period; wherein the backup mode formulater is a codec neural network structure including an encoder and a decoder; and the backup mode combination is a combination of at least two of full backup, incremental backup, and differential backup; A training module is adapted to obtain text data samples of the service level agreement requirements, application scenario descriptions, and backup requirements of each historical slice instance, and to annotate the backup methods of each historical slice instance to obtain an annotation value of the backup method combination of the corresponding text data sample; input the text data sample into an encoder, which performs feature extraction on the text data sample and merges the features to obtain a context vector; input the context vector into a decoder, which performs attention aggregation on the context vector to generate a predicted value of the backup method combination of each historical slice instance within a preset period; calculate the error between the predicted value and the annotated value, and adjust the parameters of the encoder and decoder according to the error until the error condition is met; The backup module is adapted to perform backup processing on the data of the slice instance according to the backup method combination within a preset period.

7. A computing device comprising: A processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform an operation corresponding to the backup method of the 5G slice instance as described in any one of claims 1-5.

8. A computer storage medium, wherein at least one executable instruction is stored in the storage medium, and the executable instruction enables a processor to perform operations corresponding to the backup method of the 5G slice instance as described in any one of claims 1-5.

Images