A signaling auto-completion method, apparatus, computing device, and storage medium

Abstract

The application discloses a signaling automatic completion method and device, a computing device and a storage medium. According to the technical scheme provided by the application, an original signaling collection result corresponding to a signaling query request is acquired; key information in each piece of signaling is extracted from the original signaling collection result, the key information is combined to obtain an original signaling text, and the original signaling text is preprocessed; the preprocessed original signaling text is input into a trained convolutional denoising auto-encoding neural network, signaling missing content is restored through the convolutional denoising auto-encoding neural network, and a completed signaling text is obtained. Through the application, the trained convolutional denoising auto-encoding neural network can be used as a signaling missing automatic completer to automatically restore the missing signaling, and the efficiency of an operation and maintenance personnel in locating faults and complaints according to the signaling is improved.

Description

Technical Field

[0001] This invention relates to the fields of mobile communication core networks and deep learning, specifically to a signaling automatic completion method, apparatus, computing device, and computer storage medium. Background Technology

[0002] With the development of mobile communication technology, the number of interactive signaling messages related to calls or value-added services in the mobile communication core network has increased significantly. Core network operation and maintenance personnel use signaling monitoring tools to obtain these interactive signaling messages and analyze service failures or user complaints based on them.

[0003] Due to the significant increase in the number of interactive signaling messages and the limitations of existing signaling monitoring tools, some signaling content loss, omissions, and missed collections inevitably occur during transmission or collection using these tools. These issues further reduce the efficiency of maintenance personnel in locating faults and handling complaints through interactive signaling. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed to provide a signaling automatic completion method and corresponding signaling automatic completion device, computing device and computer storage medium to overcome the above problems or at least partially solve the above problems.

[0005] According to one aspect of the present invention, a signaling auto-completion method is provided, the method comprising:

[0006] Obtain the original signaling collection results corresponding to the signaling query request;

[0007] Key information is extracted from each signaling message from the original signaling acquisition results, the key information is merged to obtain the original signaling text, and the original signaling text is preprocessed.

[0008] The preprocessed original signaling text is input into a trained convolutional denoising autoencoder neural network. The missing signaling content is restored through the convolutional denoising autoencoder neural network, resulting in the completed signaling text.

[0009] In the above scheme, the key information in the signaling includes: message type, source network element, and destination network element;

[0010] The step of extracting key information from each signaling message from the original signaling acquisition results further includes:

[0011] For each signaling message in the original signaling acquisition results, the message type is extracted from the message header, the source network element is extracted from the VIA header field, and the destination network element is extracted from the route header field.

[0012] In the above scheme, the preprocessing of the original signaling text further includes:

[0013] Remove punctuation marks from the original signaling text and perform integer serialization on the original signaling text.

[0014] In the above scheme, before inputting the preprocessed original signaling text into the trained convolutional denoising autoencoder neural network, the method further includes:

[0015] Obtain historical raw signaling acquisition results, and construct a noisy signaling sample set and a complete signaling sample set corresponding to the noisy signaling sample set based on the historical raw signaling acquisition results;

[0016] A convolutional denoising autoencoder neural network is trained based on the noisy signaling sample set and the complete signaling sample set.

[0017] In the above scheme, the step of constructing a noisy signaling sample set and a complete signaling sample set corresponding to the noisy signaling sample set based on the historical original signaling acquisition results further includes:

[0018] Extract key information from each signaling message from the historical original signaling acquisition results, and merge the key information to obtain the historical original signaling text;

[0019] Noise is added to the original historical signaling text, and the original historical signaling text is preprocessed to construct a noisy signaling sample set;

[0020] Based on the historical original signaling text, construct a complete signaling sample set corresponding to the noisy signaling sample set.

[0021] In the above scheme, training a convolutional denoising autoencoder neural network based on the noisy signaling sample set and the padded signaling sample set further includes:

[0022] Noise-added signaling samples are extracted from the noisy signaling sample set, and complete signaling samples corresponding to the noisy signaling samples are extracted from the complete signaling sample set.

[0023] The noisy signaling sample is input into an initial convolutional denoising autoencoder neural network for training to obtain an initial completion result corresponding to the noisy signaling sample;

[0024] Based on the loss between the initial completion result corresponding to the noisy signaling sample and the completed signaling sample, a loss function is obtained, and the weight parameters of the initial convolutional denoising autoencoder neural network are updated using the loss function.

[0025] The above steps are executed iteratively until the preset convergence condition is met, resulting in a trained convolutional denoising autoencoder neural network.

[0026] In the above scheme, obtaining the original signaling acquisition result corresponding to the signaling query request further includes:

[0027] Receive signaling query requests and use signaling monitoring tools to query the original signaling collection results corresponding to the signaling query requests.

[0028] According to another aspect of the present invention, an automatic signaling completion device is provided, comprising: an acquisition module, a preprocessing module, and a restoration module; wherein,

[0029] The acquisition module is used to: acquire the original signaling acquisition result corresponding to the signaling query request;

[0030] The preprocessing module is used to: extract key information from each signaling message from the original signaling acquisition results, merge the key information to obtain the original signaling text, and preprocess the original signaling text;

[0031] The restoration module is used to: input the preprocessed original signaling text into a trained convolutional denoising autoencoder neural network, restore the missing signaling content through the convolutional denoising autoencoder neural network, and obtain the completed signaling text.

[0032] According to another aspect of the present invention, a computing device is provided, 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;

[0033] The memory is used to store at least one executable instruction that causes the processor to perform an operation corresponding to the signaling auto-completion method described above.

[0034] According to another aspect of the present invention, a computer storage medium is provided, wherein at least one executable instruction is stored therein, the executable instruction causing a processor to perform an operation corresponding to the signaling auto-completion method described above.

[0035] According to the technical solution provided by this invention, the original signaling acquisition results corresponding to the signaling query request are obtained; key information in each signaling message is extracted from the original signaling acquisition results, the key information is merged to obtain the original signaling text, and the original signaling text is preprocessed; the preprocessed original signaling text is input into a trained convolutional denoising autoencoder neural network, and the missing signaling content is restored through the convolutional denoising autoencoder neural network to obtain the completed signaling text. This solves the problems of signaling content loss, omission, and missed acquisition that occur when signaling monitoring tools collect signaling, and consequently reduces the efficiency of maintenance personnel in locating faults and complaints. By using a trained convolutional denoising autoencoder neural network to automatically complete the missing parts of the signaling, the impact of signaling content loss and other problems is greatly reduced, and the efficiency of maintenance personnel in handling signaling, locating faults, and complaints is improved.

[0036] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0037] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0038] Figure 1 A flowchart illustrating a signaling auto-completion method according to an embodiment of the present invention is shown;

[0039] Figure 2 A flowchart illustrating a training method for a convolutional denoising autoencoder neural network according to another embodiment of the present invention is shown.

[0040] Figure 3 A schematic diagram of the structure of a convolutional denoising autoencoder neural network according to an embodiment of the present invention is shown;

[0041] Figure 4 A structural block diagram of a signaling automatic completion device according to an embodiment of the present invention is shown;

[0042] Figure 5 A schematic diagram of the structure of a computing device according to an embodiment of the present invention is shown. Detailed Implementation

[0043] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0044] Figure 1 A flowchart illustrating a signaling auto-completion method according to an embodiment of the present invention is shown, as follows: Figure 1 As shown, the method includes the following steps:

[0045] Step S101: Obtain the original signaling collection result corresponding to the signaling query request.

[0046] The original signaling collection result may be the interactive signaling related to the call or value-added service requested by the signaling query request.

[0047] Specifically, obtaining the original signaling collection results corresponding to the signaling query request further includes:

[0048] Receive signaling query requests and use signaling monitoring tools to query the original signaling collection results corresponding to the signaling query requests.

[0049] Step S102: Extract key information from each signaling message from the original signaling acquisition results, and merge the key information to obtain the original signaling text.

[0050] Specifically, the key information in the signaling includes: message type, source network element, and destination network element;

[0051] The signaling example is as follows:

[0052]

[0053]

[0054] In the above scheme, the step of extracting key information from each signaling message from the original signaling acquisition results further includes:

[0055] For each signaling message in the original signaling acquisition results, the message type is extracted from the message header, the source network element is extracted from the VIA header field, and the destination network element is extracted from the route header field.

[0056] After extracting the key information of the signaling, namely message type, source network element, and destination network element, the key information is merged in chronological order to generate the original signaling text.

[0057] Step S103: Preprocess the original signaling text.

[0058] The preprocessing of the original signaling text further includes:

[0059] Remove punctuation marks from the original signaling text and perform integer serialization on the original signaling text.

[0060] Specifically, after removing all punctuation marks, the integer serialization process of the original signaling text further includes: if the original signaling text contains Chinese characters, then the text is segmented into words; if the original signaling text contains English characters, then the letters in the text are uniformly converted to lowercase, and each word is tokenized, converting each word in the data into an integer sequence.

[0061] For example: ["qci": 40, "imsi": 105, "info": 8, "update": 278, "on": 89, "agent": 164, "modify": 59, "the": 21, "interrupted": 303, "command": 231,...].

[0062] Each segment of the original signaling text is converted into an index number, and zeros are padded to sequences that do not reach the maximum text length.

[0063] Step S104: Input the preprocessed original signaling text into the trained convolutional denoising autoencoder neural network, and restore the missing signaling content through the convolutional denoising autoencoder neural network to obtain the completed signaling text.

[0064] Specifically, before inputting the preprocessed original signaling text into the trained convolutional denoising autoencoder neural network, historical original signaling acquisition results are obtained. Based on the historical original signaling acquisition results, a noisy signaling sample set and a complete signaling sample set corresponding to the noisy signaling sample set are constructed.

[0065] The step of constructing a noisy signaling sample set and a complete signaling sample set corresponding to the noisy signaling sample set based on the historical original signaling acquisition results further includes:

[0066] Key information is extracted from each signaling message from the historical original signaling acquisition results, and the key information is merged to obtain the historical original signaling text. Noise is added to the historical original signaling text, and the historical original signaling text is preprocessed to construct a noisy signaling sample set. Based on the historical original signaling text, a complete signaling sample set corresponding to the noisy signaling sample set is constructed.

[0067] A convolutional denoising autoencoder neural network is trained based on the noisy signaling sample set and the complete signaling sample set.

[0068] Finally, the preprocessed original signaling text is input into a trained convolutional denoising autoencoder neural network. The missing signaling content is restored through the convolutional denoising autoencoder neural network, resulting in the completed signaling text.

[0069] The autoencoder neural network is a deep learning method that consists of an encoder and a decoder. By compressing and decompressing the original data, and comparing the original and decompressed data, the reconstruction error is calculated and backpropagated to gradually improve the accuracy of the autoencoder.

[0070] That is, f W，b (x)≈x.

[0071] A denoising autoencoder (DAE) builds upon an autoencoder by introducing noise into the training data to train the entire neural network. Since noise is unavoidable in real-world test data, training the neural network with noisy data allows it to learn the features of the input without noise, the main features of the noise, and more robust representations of the input data. This enhances its ability to remove noise and obtain noise-free data, as well as its generalization ability. In this way, the neural network reconstructs the true original input from data containing random noise by minimizing the denoising reconstruction error.

[0072] According to the signaling auto-completion method provided in this embodiment, the original signaling acquisition results corresponding to the signaling query request are obtained; key information in each signaling message is extracted from the original signaling acquisition results, the key information is merged to obtain the original signaling text, and the original signaling text is preprocessed; the preprocessed original signaling text is input into a trained convolutional denoising autoencoder neural network, and the missing signaling content is restored through the convolutional denoising autoencoder neural network to obtain the completed signaling text. Using the technical solution provided by this invention, interactive signaling obtained by signaling monitoring tools that has lost, missing, or omitted data can be automatically completed based on the trained denoising autoencoder neural network. The trained denoising autoencoder neural network can reconstruct the true original input from data containing random noise, restoring the missing parts of the signaling, effectively improving the accuracy of mobile communication core network operation and maintenance personnel obtaining interactive signaling related to calls or value-added services through signaling monitoring tools, as well as the efficiency of operation and maintenance personnel in locating faults and complaints and processing the interactive signaling. This solution effectively avoids the problems of loss, missing, and missed collection that occur in the process of collecting interactive signaling in existing signaling monitoring tools, as well as the resulting problems of locating faults, complaints, and low efficiency in processing interactive signaling.

[0073] According to the above method, Figure 2 A flowchart illustrating a training method for a convolutional denoising autoencoder neural network according to another embodiment of the present invention is shown, as follows: Figure 2 As shown, the method includes the following steps:

[0074] Step S201: Obtain historical original signaling acquisition results, extract key information from each signaling message from the historical original signaling acquisition results, and merge the key information to obtain the historical original signaling text.

[0075] Specifically, the key information in the signaling includes: message type, source network element, and destination network element; after extracting the key information of the signaling, namely message type, source network element, and destination network element, the key information is merged in chronological order to generate historical original signaling text.

[0076] The methods for extracting key information from signaling are as described in the above scheme and will not be repeated here.

[0077] Step S202: Add noise to the historical original signaling text and preprocess the historical original signaling text to construct a noisy signaling sample set; construct a complete signaling sample set corresponding to the noisy signaling sample set based on the historical original signaling text.

[0078] Specifically, the preprocessing of the historical original signaling text further includes:

[0079] After removing all punctuation marks, for historical original signaling texts containing Chinese characters, the text is segmented into words; for historical original signaling texts containing English characters, the letters in the text are uniformly converted to lowercase, and each word is indexed. Each word in the data is converted into an integer sequence, and each segment of the historical original signaling text is converted into an index number. Sequences that do not reach the maximum text length are padded with zeros.

[0080] In the set of historical original signaling text sequences, the i-th historical original signaling text can be represented as:

[0081] The noisy, i-th historical original signaling text can be represented as

[0082] The completed signaling text sequence corresponding to the i-th historical original signaling text can be represented as follows:

[0083] The longest sequence L in the noisy signaling sample set is taken as its index sequence length, and its dictionary size is taken as input_vocab_size. The longest sequence M in the completed signaling sample set is taken as its index sequence length, and its dictionary size is taken as output_vocab_size.

[0084] Preferably, in this embodiment of the invention, the obtained total dataset can be divided into a training set and a test set, with 80% of the total dataset designated as the training set and 20% as the test set. The training set is used to train the model in the neural network, and the test set is used to test the model in the neural network.

[0085] Step S203: Extract the noisy signaling samples from the noisy signaling sample set, and extract the completed signaling samples corresponding to the noisy signaling samples from the completed signaling sample set.

[0086] Specifically, the completed signaling sample is the complete and accurate signaling corresponding to the noisy signaling sample, which is used to compare with the completion result obtained in the neural network and to train the neural network.

[0087] Step S204: Input the noisy signaling sample into the initial convolutional denoising autoencoder neural network for training to obtain the initial completion result corresponding to the noisy signaling sample.

[0088] Step S205: Based on the loss between the initial completion result corresponding to the noisy signaling sample and the completed signaling sample, obtain the loss function, and use the loss function to update the weight parameters of the initial convolutional denoising autoencoder neural network.

[0089] Specifically, the initial completion result corresponding to the noisy signaling sample is compared with the completed signaling sample, and the loss function is derived based on the loss between the two.

[0090] The loss function is used to update the weight parameters of the initial convolutional denoising autoencoder neural network to enable the convolutional denoising autoencoder neural network to perform deep learning and further train the model in the convolutional denoising autoencoder neural network.

[0091] Step S206: Determine whether the preset convergence condition is met.

[0092] If so, then proceed to step S207 to finally obtain the trained convolutional denoising autoencoder neural network;

[0093] If not, then execute steps S203 to S205 again to obtain the noisy signaling sample, input it into the neural network for training, compare the initial completion result corresponding to the obtained noisy signaling sample with the completed signaling sample, and use the obtained loss function to update the weight parameters of the convolutional denoising autoencoder neural network again until the model in the convolutional denoising autoencoder neural network meets the preset convergence condition.

[0094] Step S207 yields a trained convolutional denoising autoencoder neural network.

[0095] Figure 3 A schematic diagram of the structure of a convolutional denoising autoencoder neural network according to an embodiment of the present invention is shown, as follows: Figure 3 As shown, the convolutional denoising autoencoder neural network includes an encoder, a decoder, and a max pooling layer. The original signaling text is input into the convolutional denoising autoencoder neural network, and the completed signaling text is output.

[0096] The encoder includes: an embedding layer, a convolutional layer, a max pooling layer, a convolutional layer, a max pooling layer, and a convolutional layer; the decoder includes: an upsampling layer, a convolutional layer, an upsampling layer, a convolutional layer, an upsampling layer, a convolutional layer, and a convolutional layer.

[0097] Specifically, the encoder is used to compress and reduce the dimensionality of the original signaling text, extracting feature vectors representing the input data. The encoder includes:

[0098] The first layer is the embedding layer: the input data dimension is set to the size L of the original signaling dataset text dictionary; the embedding layer is used to perform vector mapping (word embeddings) on each word in the input text. Optionally, the output is set to the size of the vector space to be converted into 128 dimensions, that is, to convert the integer sequence of each word in the text into a fixed-shape 128-dimensional vector.

[0099] The second layer is a convolutional layer (Conv1D): the convolutional layer is used to extract text features; optionally, the number of convolutional kernels is 128 (i.e., the output dimension), the spatial window length of the convolutional kernel is set to 2 (i.e., the convolutional kernel reads 2 words continuously each time), and the activation function is set to "relu";

[0100] The third layer is the MaxPooling1D layer: the MaxPooling1D layer is used to retain the maximum value among the feature values ​​extracted by the convolution kernel, and discard all other feature values; optional, the pooling window size is set to 2;

[0101] The fourth layer is a convolutional layer (Conv1D): optional, with 64 convolutional kernels, a spatial window length of 2 for each kernel, and the activation function set to "relu";

[0102] The fifth layer is the MaxPooling1D layer: the MaxPooling1D layer is used to retain the maximum value among the feature values ​​extracted by the convolution kernel, and discard all other feature values; optionally, the pooling window size is set to 2;

[0103] The sixth layer is a convolutional layer (Conv1D): optional, with 32 convolutional kernels, a spatial window length of 2 for each kernel, and the activation function set to "relu".

[0104] The seventh layer is the MaxPooling1D layer: the MaxPooling1D layer is used to retain the maximum value among the feature values ​​extracted by the convolution kernel, and discard all other feature values; optionally, the pooling window size is set to 2.

[0105] 2) Decoder: Used for restoring and reconstructing compressed data, it reconstructs the original input signaling text data based on the extracted feature vectors. The decoder includes:

[0106] The eighth layer is the upsampling layer (Upsampling1D): optional, with the upsampling factor set to 2, meaning that each time step is repeated twice on the time axis;

[0107] The ninth layer is a convolutional layer (Conv1D): optional, with 32 convolutional kernels, a spatial window length of 2 for each kernel, and the activation function set to "relu";

[0108] The tenth layer is the upsampling layer (Upsampling1D): optional, with the upsampling factor set to 2, meaning that each time step is repeated twice on the time axis;

[0109] The eleventh layer is a convolutional layer (Conv1D): optional, with 64 convolutional kernels, a spatial window length of 2 for each kernel, and the activation function set to "relu";

[0110] The twelfth layer is the upsampling layer (Upsampling1D): optional, the upsampling factor is set to 2, that is, each time step is repeated twice on the time axis;

[0111] The thirteenth layer is a convolutional layer (Conv1D): optional, with 128 convolutional kernels, a spatial window length of 2 for each kernel, and the activation function set to "relu";

[0112] The fourteenth layer is the output layer (also a convolutional layer Conv1D): optional, with 1 convolutional kernel, a spatial window length of 2 for the kernel, and the activation function set to "sigmoid".

[0113] Preferably, the number of training epochs for the neural network can be set to 1000, the batch size to 20, and categorical cross-entropy as the loss function (loss = 'categorical_crossentropy'). The Adam optimizer is selected to improve the learning speed of traditional gradient descent (optimizer = 'adam'). The neural network uses gradient descent to find the optimal weight values ​​that minimize the objective function, and learns these weights autonomously through training. Training is performed using a training set to minimize the objective function, and the model is evaluated and validated using a test set after each training epoch. Once the model converges, its weights are derived. After training convergence, a convolutional autoencoder neural network for automatic signaling missing completion is obtained.

[0114] Figure 4 A structural block diagram of a signaling automatic completion device according to an embodiment of the present invention is shown, as follows: Figure 4 As shown, the device includes: an acquisition module 401, a preprocessing module 402, and a restoration module 403; wherein,

[0115] The acquisition module 401 is used to acquire the original signaling acquisition result corresponding to the signaling query request.

[0116] Specifically, the acquisition module 401 is further configured to: receive a signaling query request and use a signaling monitoring tool to query the original signaling acquisition results corresponding to the signaling query request.

[0117] The acquisition module 402 is used to extract key information from each signaling message from the original signaling acquisition results, merge the key information to obtain the original signaling text, and preprocess the original signaling text.

[0118] Specifically, the key information in the signaling includes: message type, source network element, and destination network element;

[0119] The preprocessing module 402 is further configured to: extract the message type from the message header of each signaling in the original signaling acquisition result, extract the source network element from the via header field of the signaling, and extract the destination network element from the route header field of the signaling;

[0120] The preprocessing of the original signaling text further includes: removing punctuation marks from the original signaling text and performing integer serialization on the original signaling text.

[0121] The restoration module 403 is used to input the preprocessed original signaling text into a trained convolutional denoising autoencoder neural network, and restore the missing signaling content through the convolutional denoising autoencoder neural network to obtain the completed signaling text.

[0122] Specifically, the device further includes: a network training module 404, used to acquire historical original signaling acquisition results, construct a noisy signaling sample set and a complete signaling sample set corresponding to the noisy signaling sample set based on the historical original signaling acquisition results; and train a convolutional denoising autoencoder neural network based on the noisy signaling sample set and the complete signaling sample set.

[0123] The network training module 404 is further configured to: extract key information from each signaling message from the historical original signaling acquisition results, merge the key information to obtain historical original signaling text; add noise to the historical original signaling text, and preprocess the historical original signaling text to construct a noisy signaling sample set; and construct a complete signaling sample set corresponding to the noisy signaling sample set based on the historical original signaling text.

[0124] The network training module 404 is further configured to: extract noisy signaling samples from the noisy signaling sample set; extract completed signaling samples corresponding to the noisy signaling samples from the completed signaling sample set; input the noisy signaling samples into an initial convolutional denoising autoencoder neural network for training to obtain an initial completion result corresponding to the noisy signaling samples; obtain a loss function based on the loss between the initial completion result corresponding to the noisy signaling samples and the completed signaling samples; update the weight parameters of the initial convolutional denoising autoencoder neural network using the loss function; iteratively execute the above steps until a preset convergence condition is met to obtain a trained convolutional denoising autoencoder neural network.

[0125] According to the signaling auto-completion device provided in this embodiment, the original signaling acquisition results corresponding to the signaling query request are obtained; key information in each signaling message is extracted from the original signaling acquisition results, the key information is merged to obtain the original signaling text, and the original signaling text is preprocessed; the preprocessed original signaling text is input into a trained convolutional denoising autoencoder neural network, and the missing signaling content is restored through the convolutional denoising autoencoder neural network to obtain the completed signaling text. Using the technical solution provided by this invention, interactive signaling obtained by signaling monitoring tools that has lost, missing, or omitted data can be automatically completed based on the trained denoising autoencoder neural network. The trained denoising autoencoder neural network can reconstruct the true original input from data containing random noise, restoring the missing parts of the signaling, effectively improving the accuracy of mobile communication core network operation and maintenance personnel obtaining interactive signaling related to calls or value-added services through signaling monitoring tools, as well as the efficiency of operation and maintenance personnel in locating faults and complaints and processing the interactive signaling. This solution effectively avoids the problems of loss, missing, and missed collection that occur in the process of collecting interactive signaling in existing signaling monitoring tools, as well as the resulting problems of locating faults, complaints, and low efficiency in processing interactive signaling.

[0126] The present invention also provides a non-volatile computer storage medium storing at least one executable instruction that can execute the signaling auto-completion method in any of the above method embodiments.

[0127] Figure 5 The diagram illustrates the structure of a computing device according to an embodiment of the present invention. The specific embodiments of the present invention do not limit the specific implementation of the computing device.

[0128] like Figure 5 As shown, the computing device may include: a processor 502, a communications interface 504, a memory 506, and a communications bus 508.

[0129] in:

[0130] The processor 502, communication interface 504, and memory 506 communicate with each other via communication bus 508.

[0131] Communication interface 504 is used to communicate with other network elements such as clients or other servers.

[0132] The processor 502 is used to execute program 510, specifically to execute the relevant steps in the above-described signaling auto-completion method embodiment.

[0133] Specifically, program 510 may include program code that includes computer operation instructions.

[0134] Processor 502 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 computing device includes one or more processors, which 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.

[0135] Memory 506 is used to store program 510. Memory 506 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0136] Specifically, program 510 can be used to cause processor 502 to execute the signaling auto-completion method in any of the above method embodiments. The specific implementation of each step in program 510 can be found in the corresponding descriptions of the steps and units in the above-described signaling auto-completion method embodiments, and will not be repeated here. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the devices and modules described above can be referred to the corresponding process descriptions in the foregoing method embodiments, and will not be repeated here.

[0137] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0138] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0139] Similarly, it should be understood that, in order to streamline this disclosure 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 invention are sometimes grouped together in a single embodiment, figure, or description thereof. However, this method of disclosure should not be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the claims, inventive aspects lie in fewer than all features of a single foregoing disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of the invention.

[0140] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0141] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features but not others included in other embodiments, combinations of features from different embodiments are intended to be within the scope of the invention and form different embodiments. For example, in the claims, any of the claimed embodiments can be used in any combination.

[0142] The various component embodiments of the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) 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 present invention can also be implemented as a device or apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0143] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses 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 invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims 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 can be interpreted as names.

Claims

1. A method for automatic signaling completion, characterized in that, The method includes: Obtain the original signaling collection results corresponding to the signaling query request; The key information in each signaling message is extracted from the original signaling acquisition results, and the key information is merged to obtain the original signaling text. The original signaling text is then preprocessed. The key information in the signaling message includes: message type, source network element, and destination network element. Specifically, for each signaling message in the original signaling acquisition results, the message type is extracted from the message header, the source network element is extracted from the VIA header field, and the destination network element is extracted from the route header field. Obtain historical raw signaling acquisition results; construct a noisy signaling sample set and a complete signaling sample set corresponding to the noisy signaling sample set based on the historical raw signaling acquisition results; train a convolutional denoising autoencoder neural network based on the noisy signaling sample set and the complete signaling sample set. The preprocessed original signaling text is input into a trained convolutional denoising autoencoder neural network. The missing signaling content is restored through the convolutional denoising autoencoder neural network, resulting in the completed signaling text.

2. The method according to claim 1, characterized in that, The preprocessing of the original signaling text further includes: Remove punctuation marks from the original signaling text and perform integer serialization on the original signaling text.

3. The method according to claim 1, characterized in that, The step of constructing a noisy signaling sample set and a completed signaling sample set corresponding to the noisy signaling sample set based on the historical original signaling acquisition results further includes: Key information is extracted from each signaling message from the historical original signaling acquisition results, and the key information is merged to obtain the historical original signaling text; Noise is added to the original historical signaling text, and the original historical signaling text is preprocessed to construct a noisy signaling sample set; Based on the historical original signaling text, construct a complete signaling sample set corresponding to the noisy signaling sample set.

4. The method according to claim 1, characterized in that, The step of training a convolutional denoising autoencoder neural network based on the noisy signaling sample set and the padded signaling sample set further includes: Noise-added signaling samples are extracted from the noisy signaling sample set, and complete signaling samples corresponding to the noisy signaling samples are extracted from the complete signaling sample set. The noisy signaling sample is input into an initial convolutional denoising autoencoder neural network for training to obtain an initial completion result corresponding to the noisy signaling sample; Based on the loss between the initial completion result corresponding to the noisy signaling sample and the completed signaling sample, a loss function is obtained, and the weight parameters of the initial convolutional denoising autoencoder neural network are updated using the loss function. The above steps are executed iteratively until the preset convergence condition is met, resulting in a trained convolutional denoising autoencoder neural network.

5. The method according to claims 1-4, characterized in that, The step of obtaining the original signaling acquisition result corresponding to the signaling query request further includes: Receive signaling query requests and use signaling monitoring tools to query the original signaling collection results corresponding to the signaling query requests.

6. A signaling automatic completion device, comprising: The module consists of an acquisition module, a preprocessing module, and a restoration module; among which, The acquisition module is used to: acquire the original signaling acquisition result corresponding to the signaling query request; The preprocessing module is used to: extract key information from each signaling message in the original signaling acquisition results, merge the key information to obtain the original signaling text, and preprocess the original signaling text; the key information in the signaling message includes: message type, source network element, and destination network element; wherein, for each signaling message in the original signaling acquisition results, the message type is extracted from the message header of the signaling message, the source network element is extracted from the VIA header field of the signaling message, and the destination network element is extracted from the route header field of the signaling message; The restoration module is used to: acquire historical original signaling acquisition results; construct a noisy signaling sample set and a completed signaling sample set corresponding to the noisy signaling sample set based on the historical original signaling acquisition results; train a convolutional denoising autoencoder neural network based on the noisy signaling sample set and the completed signaling sample set; input the preprocessed original signaling text into the trained convolutional denoising autoencoder neural network, and restore the missing signaling content through the convolutional denoising autoencoder neural network to obtain the completed signaling text.

7. A computing device, comprising: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the bus. The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the signaling auto-completion method as described in any one of claims 1-5.

8. A computer storage medium storing at least one executable instruction that causes a processor to perform an operation corresponding to the signaling auto-completion method as described in any one of claims 1-5.

Images