Rru device detection method, apparatus, device, storage medium and program product

Abstract

The application provides a kind of RRU equipment detection method, device, equipment, storage medium and program product, wherein the method comprises: using convolutional variational autoencoder, the user measurement data of the RRU equipment to be detected is reduced dimension and reconstruction processing, and the reconstructed user measurement data is obtained;Determine the network statistical index of RRU equipment, using the Wasserstein generative adversarial network WGAN to the network statistical index is enhanced processing, and the enhanced network statistical index is obtained;Based on the ridge regression algorithm, the reconstructed user measurement data is associated with the enhanced network statistical index, and the time series data of the associated network statistical index is obtained;Time series data is input into the TCN prediction model constructed in advance, and the prediction time series data corresponding to the associated network statistical index is obtained, and according to time series data and prediction time series data, the abnormal detection result of RRU equipment is obtained.The application improves the efficiency and accuracy of RRU equipment detection.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, and in particular to a RRU device detection method, device, equipment, storage medium and program product. BACKGROUND

[0002] With the gradual popularization of 5G network, multiple cells will be set under a base station, and multiple remote radio units (RRU) devices will be set under the indoor cells to meet the growing demand for terminals. In the operation and maintenance process of the 5G network, it is usually necessary to detect the abnormality of the RRU device. At present, the artificial judgment method is used to detect the abnormality of the RRU device, which is low in efficiency and difficult to guarantee the accuracy. SUMMARY

[0003] The present application provides a RRU device detection method, device, equipment, storage medium and program product to solve the defects that the artificial judgment method is used to detect the abnormality of the RRU device in the prior art, which is low in efficiency and difficult to guarantee the accuracy.

[0004] In a first aspect, the present application provides a RRU device detection method, comprising:

[0005] The user measurement data of the remote radio unit (RRU) device to be detected is processed by dimension reduction and reconstruction using a convolutional variational autoencoder to obtain reconstructed user measurement data;

[0006] The network statistical indicators of the RRU device are determined, and the network statistical indicators are enhanced using a Wasserstein generative adversarial network (WGAN) to obtain enhanced network statistical indicators;

[0007] The reconstructed user measurement data and the enhanced network statistical indicators are associated based on a ridge regression algorithm to obtain time series data of the associated network statistical indicators;

[0008] The time series data is input into a pre-constructed time convolution network (TCN) prediction model to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and an abnormality detection result of the RRU device is obtained according to the time series data and the predicted time series data.

[0009] The TCN prediction model is trained based on the time series data of the sample network statistical indicators associated with the reconstructed sample user measurement data.

[0010] In some embodiments, the reconstructed user measurement data and the enhanced network statistical indicators are associated based on the ridge regression algorithm to obtain time series data of the associated network statistical indicators, comprising:

[0011] determine a regression function of the reconstructed user measurement data and the enhanced network statistic indicators based on a ridge regression algorithm;

[0012] calculate a correlation coefficient of the reconstructed user measurement data and the enhanced network statistic indicators according to the regression function;

[0013] determine a correlated network statistic indicator according to the correlation coefficient, to obtain time series data of the correlated network statistic indicator.

[0014] In some embodiments, the abnormality detection result of the RRU device is obtained according to the time series data and the predicted time series data, including:

[0015] determine the difference between the predicted time series data and the time series data to obtain a corresponding abnormality score;

[0016] determine an abnormality score threshold, compare the abnormality score with the abnormality score threshold, and determine whether the RRU device is abnormal, and if so, determine the abnormal time point of the RRU device.

[0017] In some embodiments, the network statistic indicators are enhanced by using a Wasserstein generative adversarial network (WGAN) to obtain enhanced network statistic indicators, including:

[0018] input the network statistic indicators into a pre-constructed WGAN model to obtain new network statistic indicators output by the WGAN model, the new network statistic indicators being similar to the network statistic indicators;

[0019] fuse the new network statistic indicators with the network statistic indicators to obtain the enhanced network statistic indicators;

[0020] The WGAN model includes a generator and a discriminator, and the WGAN model is obtained by cross-iterative training of the generator and the discriminator based on sample network statistic indicators.

[0021] In some embodiments, the user measurement data of the remote radio unit (RRU) device to be detected is processed by using a convolutional variational autoencoder for dimension reduction and reconstruction to obtain reconstructed user measurement data, including:

[0022] obtain user measurement data of a remote radio unit (RRU) device to be detected;

[0023] encode the user measurement data by using the convolutional variational autoencoder to obtain hidden features of the user measurement data, and decode the hidden features to obtain reconstructed user measurement data.

[0024] In some embodiments, the determination process of the TCN prediction model comprises:

[0025] obtaining sample user measurement data and sample network statistical indicators corresponding to a sample RRU device;

[0026] performing dimension reduction and reconstruction processing on the sample user measurement data by using a convolutional variational autoencoder to obtain reconstructed sample user measurement data;

[0027] performing enhancement processing on the sample network statistical indicators by using a Wasserstein generative adversarial network (WGAN) to obtain enhanced sample network statistical indicators;

[0028] associating the reconstructed sample user measurement data with the enhanced sample network statistical indicators based on a ridge regression algorithm to obtain time series data of a t time period and time series data of a t+1 time period of the associated network statistical indicators;

[0029] inputting the time series data of the t time period into a pre-constructed initial TCN prediction model to obtain predicted time series data of the t+1 time period output by the initial TCN prediction model;

[0030] calculating a loss function value based on the predicted time series data of the t+1 time period and the time series data of the t+1 time period;

[0031] performing iterative optimization on parameters of the initial TCN prediction model according to the loss function value to obtain the TCN prediction model.

[0032] In a second aspect, the present application further provides an RRU device detection apparatus, comprising:

[0033] a reconstruction unit configured to perform dimension reduction and reconstruction processing on user measurement data of a remote radio unit (RRU) device to be detected by using a convolutional variational autoencoder to obtain reconstructed user measurement data;

[0034] an enhancement unit configured to determine network statistical indicators of the RRU device, and perform enhancement processing on the network statistical indicators by using a Wasserstein generative adversarial network (WGAN) to obtain enhanced network statistical indicators;

[0035] an association unit configured to associate the reconstructed user measurement data with the enhanced network statistical indicators based on a ridge regression algorithm to obtain time series data of the network statistical indicators;

[0036] a prediction unit configured to input the time series data into a pre-constructed time convolution network (TCN) prediction model to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and obtain an anomaly detection result of the RRU device according to the time series data and the predicted time series data.

[0037] The TCN prediction model is trained based on the time series data of the sample network statistical indicators associated with the reconstructed sample user measurement data.

[0038] In a third aspect, the present application also provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the RRU device detection method according to any of the above aspects when executing the program.

[0039] In a fourth aspect, the present application also provides a non-transitory computer-readable storage medium, which stores a computer program executable by a processor to implement the RRU device detection method according to any of the above aspects.

[0040] In a fifth aspect, the present application also provides a computer program product, which includes a computer program executable by a processor to implement the RRU device detection method according to any of the above aspects.

[0041] The RRU device detection method, device, electronic device, storage medium, and program product provided by the present application can obtain reconstructed user measurement data by performing dimension reduction and reconstruction on the user measurement data of the RRU device to be detected, determine network statistical indicators of the RRU device, perform enhancement processing on the network statistical indicators to obtain enhanced network statistical indicators, provide a comprehensive data basis for anomaly detection, associate the reconstructed user measurement data with the enhanced network statistical indicators based on a ridge regression algorithm to obtain time series data of the associated network statistical indicators, input the time series data into a pre-constructed TCN prediction model to obtain predicted time series data corresponding to the associated network statistical indicators, determine differences according to the time series data and the predicted time series data, and obtain an anomaly detection result of the RRU device, thereby improving the efficiency and accuracy of RRU device detection. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0043] Figure 1is a flowchart of a RRU device detection method provided by an embodiment of the present application;

[0044] Figure 2 is a flowchart of a dimensionality reduction and reconstruction processing method for user measurement data using a convolutional variational autoencoder provided by an embodiment of the present application;

[0045] Figure 3 is a structural diagram of a RRU device detection apparatus provided by an embodiment of the present application;

[0046] Figure 4 is a structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0047] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described below in detail with reference to the accompanying drawings of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0048] At present, the abnormality of RRU devices is determined based on a single index or an index of single network side statistics, which is often not comprehensive enough. There are great differences in the service effects of different data such as registration, voice and video, and the measurement data. In addition to the statistical index of device granularity, the user dimension measurement perception data is also an important index for abnormality detection. How to combine the two types of structured data, i.e., the perception data measured by different users at different times and the statistical index of devices in different coverage areas, to detect RRU devices is the key point of RRU device detection.

[0049] At present, due to the small amount of abnormal data samples of RRU devices, it is difficult to use machine learning methods to realize reasonable evaluation of the abnormality. Although a simple autoencoder (AE) can effectively obtain reconstructed sample data through dimensionality reduction, the overfitting problem is difficult to solve, and the accuracy of the abnormal samples generated by the adversarial network is poor and the error is high.

[0050] To this end, the embodiment of the present application provides a RRU device detection method, device, equipment, storage medium and program product, the user measurement data of the RRU device to be detected is processed by dimension reduction and reconstruction, the reconstructed user measurement data is obtained, the network statistical index of the RRU device is determined, the network statistical index is enhanced, the enhanced network statistical index is obtained, a comprehensive data basis is provided for anomaly detection, the reconstructed user measurement data and the enhanced network statistical index are associated based on the ridge regression algorithm, the time series data of the associated network statistical index is obtained, the time series data is input into the TCN prediction model constructed in advance, the prediction time series data corresponding to the associated network statistical index is obtained, the difference is determined according to the time series data and the prediction time series data, and the anomaly detection result of the RRU device is obtained, thereby improving the efficiency and accuracy of RRU device detection.

[0051] Figure 1 The flowchart of the RRU device detection method provided by the embodiment of the present application is shown in Figure 1. Figure 1 As shown in Figure 1, a RRU device detection method is provided, which includes the following steps: step 110, step 120, step 130 and step 140.

[0052] In step 110, the convolutional variational autoencoder is used to process the user measurement data of the remote radio unit (RRU) device to be detected by dimension reduction and reconstruction, and the reconstructed user measurement data is obtained.

[0053] The remote radio unit is a key device in a wireless communication network, usually located at the top of a base station antenna tower or nearby, and connected with a base station control unit through an optical fiber or other transmission medium, and the RRU device is used to perform modulation, amplification and transmission of radio frequency signals.

[0054] Optionally, the user measurement data includes but is not limited to: user registration cell ID, user equipment received reference signal power, user equipment data volume to be transmitted, user equipment occupied network resources, user equipment used modulation and coding scheme in communication, user demodulation A / N result feedback, user rate and user time delay.

[0055] The user demodulation A / N result feedback indicates the confirmation or negation of the user equipment to the received data frame, which is the key information for judging whether the data transmission is successful or not; the user rate refers to the rate of data transmission or reception of the user equipment within a certain time; and the user time delay refers to the time delay of data transmission and / or reception of the user equipment.

[0056] The convolutional variational autoencoder (C-V-AE) is a deep learning model combining a convolutional neural network (CNN) and a variational autoencoder (VAE), which can map high-dimensional data to a latent space and reconstruct the original data from the latent space.

[0057] Figure 2 A flowchart of a process for dimensionality reduction and reconstruction of user measurement data using a convolutional variational autoencoder is provided for embodiments of the present application. Figure 2 As shown in some embodiments, the user measurement data of the remote radio unit (RRU) device to be detected is processed for dimensionality reduction and reconstruction using a convolutional variational autoencoder to obtain reconstructed user measurement data, including:

[0058] Obtaining user measurement data of the remote radio unit (RRU) device to be detected.

[0059] Encoding the user measurement data using a convolutional variational autoencoder to obtain hidden features of the user measurement data, and decoding the hidden features to obtain reconstructed user measurement data.

[0060] It should be noted that, as shown in Figure 2 The convolutional variational autoencoder includes an input layer (input), an encoder (Encoder), a hidden layer (latent), a decoder (Decoder), and an output layer (output), x represents the user measurement data, q(z|x) represents the probability of generating z from x, z represents the hidden feature, p(x|z) represents the probability of generating x from z, and x' represents the reconstructed user measurement data.

[0061] Optionally, the encoder maps the user measurement data to a latent space, and the decoder converts the vector of the latent space into reconstructed data similar to the user measurement data; the encoder encodes the user measurement data into a probability distribution for describing the position of each user measurement data in the latent space.

[0062] Optionally, an initial V-C-AE model is constructed to minimize the reconstruction error and the regularization error of the latent space, the initial V-C-AE model is trained, the encoder parameter and the decoder parameter are optimized to obtain the V-C-AE model.

[0063] The reconstruction error is used to describe the similarity between the reconstructed data and the original data, and the regularization error is used to constrain the distribution of the reconstructed data in the latent space, so that it is closer to the Gaussian distribution.

[0064] Optionally, the user measurement data x is input into an encoder, the mean m and the variance v are calculated, x is sampled according to a normal distribution N(μ(x),σ(x)), such as N(0,1), to obtain sampling data s; the hidden feature z=m+vs is calculated and input into a decoder to obtain a reconstructed sample x', and the reconstruction error and the regularization error are calculated.

[0065] It can be understood that the VCAE uses a variational encoder to compress the user measurement data into a hidden variable, and solves the problem that the gradient cannot be derived by using reparameterization.

[0066] In step 120, the network statistical indicators of the RRU device are determined, the network statistical indicators are enhanced by using a Wasserstein generative adversarial network WGAN, and enhanced network statistical indicators are obtained.

[0067] The network statistical indicators are used to evaluate the network performance of the RRU device.

[0068] Optionally, the network statistical indicators include but are not limited to: a radio resource control (RRC) connection number, a physical resource block (PRB) occupation number, and traffic.

[0069] The Wasserstein generative adversarial network (WGAN) introduces a Wasserstein distance as an index for measuring the difference between the generator and the discriminator, to solve the problems of unstable training and mode collapse in the traditional GAN.

[0070] The Wasserstein distance is a more stable and optimized distance measure, and the Wasserstein distance can more accurately measure the distance between two probability distributions, especially in high-dimensional space. By optimizing the Wasserstein distance, the WGAN can generate higher-quality and more realistic data samples.

[0071] In some embodiments, the network statistical indicators are enhanced by using the Wasserstein generative adversarial network WGAN, and the enhanced network statistical indicators are obtained, including:

[0072] The network statistical indicators are input into a pre-constructed WGAN model to obtain new network statistical indicators output by the WGAN model, and the new network statistical indicators are similar to the network statistical indicators.

[0073] The new network statistical indicator is fused with the network statistical indicator to obtain an enhanced network statistical indicator.

[0074] The WGAN model includes a generator and a discriminator, and the WGAN model is obtained by cross-iterative training of the generator and the discriminator based on the sample network statistical indicator.

[0075] It should be noted that the generator generates false data according to random noise by learning the internal distribution of real data, and the discriminator is used to judge whether the input data (real data and false data) is false data; the purpose of training the generator is to make the generated data more real, so that the discriminator cannot predict whether the input data is false data, and the purpose of training the discriminator is to make the judgment more accurate.

[0076] In step 130, the reconstructed user measurement data is associated with the enhanced network statistical indicator based on a ridge regression algorithm to obtain time series data of the associated network statistical indicator.

[0077] It can be understood that by associating the reconstructed user measurement data with the enhanced network statistical indicator, time series data of the associated network statistical indicator can be obtained, key data can be extracted, and redundant data processing can be avoided, thereby improving the efficiency and accuracy of RRU device detection.

[0078] In some embodiments, the reconstructed user measurement data is associated with the enhanced network statistical indicator based on a ridge regression algorithm to obtain time series data of the associated network statistical indicator, including:

[0079] Based on the ridge regression algorithm, a regression function of the reconstructed user measurement data and the enhanced network statistical indicator is determined.

[0080] According to the regression function, a correlation coefficient of the reconstructed user measurement data and the enhanced network statistical indicator is calculated.

[0081] According to the correlation coefficient, the associated network statistical indicator is determined to obtain time series data of the associated network statistical indicator.

[0082] The ridge regression algorithm introduces a regularization parameter (i.e. a ridge regression parameter) to solve the problem of unstable parameter estimation of ordinary linear regression in multiple collinearity.

[0083] Optionally, based on the ridge regression algorithm, a regression function of the reconstructed user measurement data and the enhanced network statistical indicator is determined, and an optimal regression parameter is determined.

[0084] Optionally, the calculation formula of the regression function is as follows:

[0085] ;

[0086] wherein, represents a regression function, represents reconstructed user measurement data, represents enhanced network statistical indicators, represents a correlation coefficient of the reconstructed user measurement data and the enhanced network statistical indicators, represents a regression parameter.

[0087] Step 140, inputting the time series data into a pre-constructed time convolution network (TCN) prediction model to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and obtaining an abnormality detection result of the RRU device according to the time series data and the predicted time series data.

[0088] wherein, the TCN prediction model is trained based on the time series data of the sample network statistical indicators associated with the reconstructed sample user measurement data.

[0089] It can be understood that by using the TCN prediction model for RRU device abnormality detection, the prediction accuracy, real-time performance and efficiency can be improved, and automatic and intelligent device state monitoring can be realized, thereby significantly improving network operation and maintenance effect and device reliability.

[0090] In some embodiments, according to the time series data and the predicted time series data, the abnormality detection result of the RRU device is obtained, including:

[0091] determining the difference between the predicted time series data and the time series data to obtain a corresponding abnormality score;

[0092] determining an abnormality score threshold, comparing the abnormality score with the abnormality score threshold, and determining whether the RRU device is abnormal, and if so, determining the abnormal time point of the RRU device.

[0093] It can be understood that by analyzing the difference between the time series data and the predicted time series data and combining the abnormality score threshold for judgment, the abnormality detection capability and operation efficiency of the RRU device can be significantly improved, and the stable operation and service quality of the network device can be effectively guaranteed.

[0094] In the embodiment of the present application, by dimension reduction and reconstruction processing on the user measurement data of the RRU device to be detected, the reconstructed user measurement data is obtained, the network statistical indicators of the RRU device are determined, the network statistical indicators are enhanced, the enhanced network statistical indicators are obtained, a comprehensive data basis is provided for anomaly detection, the reconstructed user measurement data and the enhanced network statistical indicators are associated based on the ridge regression algorithm, the time series data of the associated network statistical indicators is obtained, the time series data is input into the pre-constructed TCN prediction model, the prediction time series data corresponding to the associated network statistical indicators is obtained, the difference is determined according to the time series data and the prediction time series data, and the anomaly detection result of the RRU device is obtained, thereby improving the efficiency and accuracy of RRU device detection.

[0095] In some embodiments, the determination process of the TCN prediction model comprises:

[0096] Obtaining sample user measurement data and sample network statistical indicators corresponding to a sample RRU device;

[0097] Using a convolutional variational autoencoder to perform dimension reduction and reconstruction processing on the sample user measurement data, and obtaining reconstructed sample user measurement data;

[0098] Using a Wasserstein generative adversarial network (WGAN) to perform enhancement processing on the sample network statistical indicators, and obtaining enhanced sample network statistical indicators;

[0099] Associating the reconstructed sample user measurement data and the enhanced sample network statistical indicators based on a ridge regression algorithm, and obtaining time series data of the sample network statistical indicators in a t time period and time series data in a t+1 time period;

[0100] Inputting the time series data in the t time period into a pre-constructed initial TCN prediction model, and obtaining prediction time series data in the t+1 time period output by the initial TCN prediction model;

[0101] Based on the prediction time series data in the t+1 time period and the time series data in the t+1 time period, a loss function value is calculated;

[0102] According to the loss function value, the parameters of the initial TCN prediction model are iteratively optimized to obtain the TCN prediction model.

[0103] Optionally, a dynamic time warping (DTW) algorithm is used to calculate the distance between the prediction time series data in the t+1 time period and the time series data in the t+1 time period, and the calculation process is as follows:

[0104] 1) Initialization: Define two sequences A (predicted time series data of t+1 time period) and B (time series data of t+1 time period) to be matched, and initialize two pointers pointing to the starting positions of sequences A and B respectively;

[0105] 2) Calculate distance: Calculate the Euclidean distance between the elements pointed to by the two pointers;

[0106] 3) Move pointer: According to the calculated distance result, move the pointer, so that the elements of the two sequences are gradually aligned;

[0107] 4) Repeat calculation: Repeat steps 2) and 3) until the two pointers reach the end of the sequence;

[0108] 5) Calculate total distance: Calculate the total distance of the two sequences, which is the loss function value.

[0109] The RRU device detection apparatus provided by the embodiment of the application is described below, and the apparatus described below can be referred to in correspondence with the RRU device detection method described above.

[0110] Figure 3 The structure diagram of the RRU device detection apparatus provided by the embodiment of the application is shown in Figure 3 The RRU device detection apparatus 300 includes:

[0111] The reconstruction unit 310 is configured to use a convolutional variational autoencoder to perform dimensionality reduction and reconstruction processing on user measurement data of a remote radio unit (RRU) device to be detected, and obtain reconstructed user measurement data.

[0112] The enhancement unit 320 is configured to determine network statistical indicators of the RRU device, and perform enhancement processing on the network statistical indicators using a Wasserstein generative adversarial network (WGAN) to obtain enhanced network statistical indicators.

[0113] The association unit 330 is configured to associate the reconstructed user measurement data and the enhanced network statistical indicators based on a ridge regression algorithm to obtain time series data of the associated network statistical indicators.

[0114] The prediction unit 340 is configured to input the time series data into a pre-constructed time convolution network (TCN) prediction model to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and obtain an anomaly detection result of the RRU device according to the time series data and the predicted time series data.

[0115] The TCN prediction model is trained based on time series data of sample network statistical indicators associated with reconstructed sample user measurement data.

[0116] Optionally, the reconstructed user measurement data is associated with the enhanced network statistical indicators based on a ridge regression algorithm to obtain time series data of the associated network statistical indicators, including:

[0117] Based on the ridge regression algorithm, a regression function of the reconstructed user measurement data and the enhanced network statistical indicators is determined.

[0118] According to the regression function, the correlation coefficient of the reconstructed user measurement data and the enhanced network statistical indicators is calculated.

[0119] According to the correlation coefficient, the associated network statistical indicators are determined to obtain the time series data of the associated network statistical indicators.

[0120] Optionally, according to the time series data and the predicted time series data, an abnormal detection result of the RRU device is obtained, including:

[0121] Determine the difference between the predicted time series data and the time series data to obtain a corresponding anomaly score.

[0122] Determine an anomaly score threshold, compare the anomaly score with the anomaly score threshold, and determine whether the RRU device is abnormal, if so, determine the abnormal time point of the RRU device.

[0123] Optionally, the network statistical indicators are enhanced by using a Wasserstein generative adversarial network (WGAN) to obtain enhanced network statistical indicators, including:

[0124] The network statistical indicators are input into a pre-constructed WGAN model to obtain new network statistical indicators output by the WGAN model, and the new network statistical indicators are similar to the network statistical indicators.

[0125] The new network statistical indicators are fused with the network statistical indicators to obtain the enhanced network statistical indicators.

[0126] The WGAN model includes a generator and a discriminator, and the WGAN model is obtained by cross-iterative training of the generator and the discriminator based on sample network statistical indicators.

[0127] Optionally, a convolutional variational autoencoder is used to perform dimension reduction and reconstruction processing on user measurement data of a remote radio unit (RRU) device to be detected to obtain reconstructed user measurement data, including:

[0128] Obtain user measurement data of a remote radio unit (RRU) device to be detected.

[0129] The convolutional variational autoencoder is used to encode the user measurement data to obtain hidden features of the user measurement data, and the hidden features are decoded to obtain the reconstructed user measurement data.

[0130] Optionally, the determining process of the TCN prediction model comprises:

[0131] Obtaining sample user measurement data and sample network statistical indicators corresponding to the sample RRU device;

[0132] Performing dimension reduction and reconstruction processing on the sample user measurement data by using a convolutional variational autoencoder to obtain reconstructed sample user measurement data;

[0133] Performing enhancement processing on the sample network statistical indicators by using a Wasserstein generative adversarial network (WGAN) to obtain enhanced sample network statistical indicators;

[0134] Associating the reconstructed sample user measurement data with the enhanced sample network statistical indicators based on a ridge regression algorithm to obtain time series data of the t time period and time series data of the t+1 time period of the associated sample network statistical indicators;

[0135] Inputting the time series data of the t time period into a pre-constructed initial TCN prediction model to obtain predicted time series data of the t+1 time period output by the initial TCN prediction model;

[0136] Calculating a loss function value based on the predicted time series data of the t+1 time period and the time series data of the t+1 time period;

[0137] Iteratively optimizing parameters of the initial TCN prediction model according to the loss function value to obtain the TCN prediction model.

[0138] It should be noted that the RRU device detection apparatus provided by the embodiment of the present application can realize all the method steps realized by the RRU device detection method embodiment and achieve the same technical effects, and thus the same parts and beneficial effects of the method embodiment will not be described in detail.

[0139] Figure 4 The structural schematic diagram of the electronic device provided by the embodiment of the present application is as follows: Figure 4As shown, the electronic device can include a processor 410, a communications interface 420, a memory 430, and a communications bus 440, wherein the processor 410, the communications interface 420, and the memory 430 complete mutual communication through the communications bus 440. The processor 410 can invoke a logic instruction in the memory 430 to execute an RRU device detection method, which includes: using a convolutional variational autoencoder to perform dimensionality reduction and reconstruction processing on user measurement data of a remote radio unit (RRU) device to be detected, to obtain reconstructed user measurement data; determining network statistical indicators of the RRU device, and using a Wasserstein generative adversarial network (WGAN) to perform enhancement processing on the network statistical indicators, to obtain enhanced network statistical indicators; associating the reconstructed user measurement data with the enhanced network statistical indicators based on a ridge regression algorithm, to obtain time series data of the associated network statistical indicators; inputting the time series data into a pre-constructed time convolution network (TCN) prediction model, to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and obtaining an anomaly detection result of the RRU device according to the time series data and the predicted time series data; wherein the TCN prediction model is trained based on time series data of sample network statistical indicators associated with reconstructed sample user measurement data.

[0140] In addition, the logic instruction in the memory 430 described above can be implemented in the form of a software function unit and sold or used as an independent product, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0141] In another aspect, the present application also provides a computer program product comprising a computer program, which can be stored on a non-transitory computer-readable storage medium, and the computer program can be executed by a processor to enable a computer to perform the RRU device detection method provided by the above method, which comprises: using a convolutional variational autoencoder to perform dimension reduction and reconstruction processing on user measurement data of a remote radio unit (RRU) device to be detected, to obtain reconstructed user measurement data; determining network statistical indicators of the RRU device, and using a Wasserstein generative adversarial network (WGAN) to perform enhancement processing on the network statistical indicators, to obtain enhanced network statistical indicators; associating the reconstructed user measurement data with the enhanced network statistical indicators based on a ridge regression algorithm, to obtain time series data of the associated network statistical indicators; inputting the time series data into a pre-constructed time convolution network (TCN) prediction model, to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and obtaining an anomaly detection result of the RRU device according to the time series data and the predicted time series data; wherein the TCN prediction model is trained based on time series data of sample network statistical indicators associated with reconstructed sample user measurement data.

[0142] In another aspect, the present application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and the computer program can be executed by a processor to implement the RRU device detection method provided by the above method, which comprises: using a convolutional variational autoencoder to perform dimension reduction and reconstruction processing on user measurement data of a remote radio unit (RRU) device to be detected, to obtain reconstructed user measurement data; determining network statistical indicators of the RRU device, and using a Wasserstein generative adversarial network (WGAN) to perform enhancement processing on the network statistical indicators, to obtain enhanced network statistical indicators; associating the reconstructed user measurement data with the enhanced network statistical indicators based on a ridge regression algorithm, to obtain time series data of the associated network statistical indicators; inputting the time series data into a pre-constructed time convolution network (TCN) prediction model, to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and obtaining an anomaly detection result of the RRU device according to the time series data and the predicted time series data; wherein the TCN prediction model is trained based on time series data of sample network statistical indicators associated with reconstructed sample user measurement data.

[0143] The device embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purposes of the embodiments according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0144] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0145] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting RRU equipment, characterized in that, The method comprises the following steps: dimension reduction and reconstruction processing of user measurement data of a remote radio unit (RRU) device to be detected are performed by using a convolutional variational autoencoder to obtain reconstructed user measurement data; network statistical indicators of the RRU device are determined, and the network statistical indicators are enhanced by using a Wasserstein generative adversarial network (WGAN) to obtain enhanced network statistical indicators; the reconstructed user measurement data and the enhanced network statistical indicators are associated based on a ridge regression algorithm to obtain time series data of the associated network statistical indicators; the time series data are input into a pre-constructed time convolution network (TCN) prediction model to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and an abnormality detection result of the RRU device is obtained according to the time series data and the predicted time series data; wherein the TCN prediction model is trained based on sample network statistical indicators time series data associated with reconstructed sample user measurement data.

2. The method of claim 1, wherein the RRU device detection method is characterized by, The method comprises the following steps: a regression function of the reconstructed user measurement data and the enhanced network statistical indicators is determined based on a ridge regression algorithm; a correlation coefficient of the reconstructed user measurement data and the enhanced network statistical indicators is calculated according to the regression function; associated network statistical indicators are determined according to the correlation coefficient to obtain time series data of the associated network statistical indicators.

3. The method of claim 1, wherein the RRU device detection is performed by a base station. The method comprises the following steps: a difference between the predicted time series data and the time series data is determined to obtain a corresponding abnormality score; an abnormality score threshold is determined, and the abnormality score is compared with the abnormality score threshold to determine whether the RRU device is abnormal, and if so, an abnormal time point of the RRU device is determined.

4. The method of claim 1, wherein the RRU device detection is performed by a base station. The method comprises the following steps: the network statistical indicators are input into a pre-constructed WGAN model to obtain new network statistical indicators output by the WGAN model, the new network statistical indicators being similar to the network statistical indicators; the new network statistical indicators and the network statistical indicators are fused to obtain the enhanced network statistical indicators; wherein the WGAN model comprises a generator and a discriminator, and the WGAN model is obtained by cross-iterative training of the generator and the discriminator based on sample network statistical indicators.

5. The method of claim 1, wherein the RRU device detection is performed by a base station. The method comprises the following steps: user measurement data of a remote radio unit (RRU) device to be detected are obtained; The convolutional variational autoencoder is used to encode the user measurement data to obtain hidden features of the user measurement data, and the hidden features are decoded to obtain reconstructed user measurement data.

6. The method of claim 1, wherein the RRU device detection is performed by a base station. The determination process of the TCN prediction model includes: Obtain sample user measurement data and sample network statistical indicators corresponding to a sample RRU device; The convolutional variational autoencoder is used to reduce dimension and reconstruct the sample user measurement data to obtain reconstructed sample user measurement data; The sample network statistical indicators are enhanced by using a Wasserstein generative adversarial network (WGAN) to obtain enhanced sample network statistical indicators; The reconstructed sample user measurement data and the enhanced sample network statistical indicators are associated based on a ridge regression algorithm to obtain time series data of the associated sample network statistical indicators in a t time period and time series data in a t+1 time period; The t time period time series data is input into a pre-constructed initial TCN prediction model to obtain predicted time series data in the t+1 time period output by the initial TCN prediction model; Based on the predicted time series data in the t+1 time period and the time series data in the t+1 time period, a loss function value is calculated; According to the loss function value, the parameters of the initial TCN prediction model are iteratively optimized to obtain the TCN prediction model.

7. An RRU device detection apparatus, comprising: It includes: A reconstruction unit is configured to use a convolutional variational autoencoder to reduce dimension and reconstruct user measurement data of a remote radio unit (RRU) device to be detected to obtain reconstructed user measurement data; An enhancement unit is configured to determine network statistical indicators of the RRU device, and use a Wasserstein generative adversarial network (WGAN) to enhance the network statistical indicators to obtain enhanced network statistical indicators; An association unit is configured to associate the reconstructed user measurement data and the enhanced network statistical indicators based on a ridge regression algorithm to obtain time series data of the associated network statistical indicators; A prediction unit is configured to input the time series data into a pre-constructed time convolution network (TCN) prediction model to obtain predicted time series data corresponding to the associated network statistical indicators output by the TCN prediction model, and obtain an anomaly detection result of the RRU device according to the time series data and the predicted time series data. The TCN prediction model is trained based on the time series data of the associated sample network statistical indicators of the reconstructed sample user measurement data.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the RRU device detection method of any one of claims 1 to 6. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the RRU device detection method of any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the RRU device detection method of any one of claims 1 to 6.

Images