Base station energy saving index data anomaly identification method and device, computer device and storage medium

Abstract

The application relates to a base station energy-saving index data abnormality identification method and device, computer equipment and a storage medium. The method comprises the following steps: determining fluctuation representation coefficients and autocorrelation coefficients corresponding to each base station energy-saving index item according to base station energy-saving index data sequences corresponding to the base station energy-saving index items; identifying static base station energy-saving index items from the base station energy-saving index items according to the fluctuation representation coefficients; identifying periodic base station energy-saving index items from the base station energy-saving index items according to the autocorrelation coefficients; determining unstable base station energy-saving index items according to the base station energy-saving index items other than the periodic base station energy-saving index items and the static base station energy-saving index items; and performing abnormality identification on base station energy-saving index data sequences corresponding to the static base station energy-saving index items, the periodic base station energy-saving index items and the unstable base station energy-saving index items respectively according to corresponding abnormality identification modes. The method can improve the abnormality identification accuracy.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a method, apparatus, computer equipment, and storage medium for identifying anomalies in base station energy-saving index data. Background Technology

[0002] Base station energy-saving systems involve multiple layers, large data volumes, and significant maintenance costs. Due to the system's sensitivity to anomalies and high complexity, any abnormalities can lead to adverse consequences. Therefore, identifying anomalies in base station energy-saving indicators is crucial.

[0003] Currently, base station energy-saving indicators are selected and judged based on human experience, which is highly subjective and leads to a high rate of missed or false positives in identifying anomalies in base station energy-saving indicators. Therefore, there is an urgent need to propose a solution that can improve the accuracy of anomaly identification. Summary of the Invention

[0004] Based on this, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for identifying anomalies in base station energy-saving index data with improved accuracy, in order to address the aforementioned technical problems.

[0005] Firstly, this application provides a method for identifying anomalies in base station energy-saving indicator data. The method includes:

[0006] Obtain the base station energy-saving index data sequence corresponding to each base station energy-saving index item; the base station energy-saving index data sequence includes the base station energy-saving index data of the base station energy-saving index item at different time points;

[0007] Based on the data sequence of each base station energy-saving index, determine the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each base station energy-saving index item;

[0008] Based on the fluctuation characterization coefficient, identify static base station energy-saving indicators from each base station energy-saving indicator item;

[0009] Based on the autocorrelation coefficient, periodic detection is performed on the data sequence of each base station energy-saving index to identify periodic base station energy-saving index items from each base station energy-saving index item;

[0010] Based on the base station energy-saving indicators other than the periodic base station energy-saving indicators and the static base station energy-saving indicators, determine the unstable base station energy-saving indicators;

[0011] The base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items are respectively identified for anomalies according to the corresponding anomaly identification methods.

[0012] In one embodiment, the fluctuation characterization coefficient is the coefficient of variation; the step of determining the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each base station energy-saving index item based on the data sequence of each base station energy-saving index includes:

[0013] For each base station energy-saving indicator item, determine the standard deviation of the base station energy-saving indicator data sequence corresponding to the base station energy-saving indicator item;

[0014] Determine the mean value of the base station energy-saving index data in the base station energy-saving index data sequence;

[0015] Based on the standard deviation and mean, determine the coefficient of variation corresponding to the base station energy-saving index item;

[0016] The step of identifying static base station energy-saving indicators from each base station energy-saving indicator item based on the fluctuation characterization coefficient includes:

[0017] Base station energy-saving indicators with a coefficient of variation less than or equal to the first preset threshold are defined as static base station energy-saving indicators.

[0018] In one embodiment, the static base station energy-saving index item corresponds to the first base station energy-saving index data sequence;

[0019] The step of identifying anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items according to the corresponding anomaly identification methods includes:

[0020] The data is slid across the energy-saving index data sequence of the first base station using a first moving average time window.

[0021] The average value of static base station energy-saving index data within the first moving average time window is calculated to obtain the first average value.

[0022] The first mean value is compared with the first dynamic threshold, and abnormal data in the first base station energy-saving index data sequence is identified based on the comparison result.

[0023] In one embodiment, the unstable base station energy-saving index item corresponds to the second base station energy-saving index data sequence;

[0024] The step of identifying anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items according to the corresponding anomaly identification methods includes:

[0025] When there are multiple energy-saving indicators for unstable base stations, the energy-saving indicators for unstable base stations with abnormal labels are determined from the multiple energy-saving indicators for unstable base stations to obtain the abnormal label indicators.

[0026] For the abnormal label indicator item, obtain the historical unstable base station energy saving indicator data of the abnormal label indicator item, and determine the second dynamic threshold.

[0027] By sliding the second moving average time window over the second base station energy-saving index data sequence, the average value of the unstable base station energy-saving index data located within the second moving average time window is calculated to obtain the second average value.

[0028] Based on the comparison between the second mean and the second dynamic threshold, abnormal data in the second base station energy-saving index data sequence are identified.

[0029] In one embodiment, the method further includes:

[0030] For the energy-saving indicators without abnormal labels among the multiple unstable base station energy-saving indicators, the energy-saving indicator data of unstable base stations without abnormal labels is initially identified through an unsupervised anomaly identification model to obtain the energy-saving indicator data of unstable base stations with preliminary anomalies.

[0031] Based on the second preset threshold, the energy-saving index data of unstable base stations that were initially identified as abnormal are further identified as abnormal, and the energy-saving index data of unstable base stations that are finally identified as abnormal are obtained.

[0032] In one embodiment, the periodic base station energy-saving index item corresponds to a third base station energy-saving index data sequence;

[0033] The step of identifying anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items according to the corresponding anomaly identification methods includes:

[0034] The energy-saving index data sequence of the third base station is divided according to the period to obtain the periodic base station energy-saving index data for each period.

[0035] By using a pre-trained periodic anomaly identification model, the periodic base station energy-saving index data for the current period is predicted based on the periodic base station energy-saving index data of the previous period, and the difference between the predicted periodic base station energy-saving index data for the current period and the periodic base station energy-saving index data for the current period in the third base station energy-saving index data sequence is determined.

[0036] If the difference is greater than a preset difference threshold, it is determined that the periodic base station energy-saving index data for the current period is abnormal.

[0037] In one embodiment, each base station energy-saving indicator is a key base station energy-saving indicator; obtaining the base station energy-saving indicator data sequence corresponding to each base station energy-saving indicator includes:

[0038] Obtain the base station energy-saving index data sequence corresponding to each candidate base station energy-saving index item;

[0039] For each candidate base station energy saving indicator item, the base station energy saving indicator data sequence corresponding to the candidate base station energy saving indicator item is divided into two sub-sequences;

[0040] Hypothesis testing was performed on the two sets of subsequences to obtain the hypothesis test results;

[0041] Based on the hypothesis test results, key base station energy-saving indicators that are sensitive to anomalies are identified from the candidate base station energy-saving indicators.

[0042] Secondly, this application also provides an anomaly identification device for base station energy-saving index data, the device comprising:

[0043] The indicator acquisition module is used to acquire the base station energy-saving indicator data sequence corresponding to each base station energy-saving indicator item; the base station energy-saving indicator data sequence includes the base station energy-saving indicator data of the base station energy-saving indicator item at different time points.

[0044] The coefficient determination module is used to determine the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each energy-saving index item of each base station based on the data sequence of each base station energy-saving index.

[0045] The indicator identification module is used to identify static base station energy-saving indicators from each base station energy-saving indicator item based on the fluctuation characterization coefficient; to perform periodic detection on each base station energy-saving indicator data sequence based on the autocorrelation coefficient to identify periodic base station energy-saving indicators from each base station energy-saving indicator item; and to determine unstable base station energy-saving indicators based on base station energy-saving indicators other than the periodic base station energy-saving indicators and the static base station energy-saving indicators.

[0046] The anomaly identification module is used to identify anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items, respectively, according to the corresponding anomaly identification methods.

[0047] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps described in the embodiments of this application.

[0048] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the steps described in the embodiments of this application.

[0049] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the steps described in the embodiments of this application.

[0050] The aforementioned method, apparatus, computer equipment, storage medium, and computer program product for anomaly identification of base station energy-saving index data determine the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each base station energy-saving index item based on the base station energy-saving index data sequence corresponding to each base station energy-saving index item. This allows for accurate and convenient identification of static and periodic base station energy-saving index items from each base station energy-saving index item, and further identification of unstable base station energy-saving index items. Then, the base station energy-saving index data sequences corresponding to the static, periodic, and unstable base station energy-saving index items are respectively subjected to anomaly identification according to the corresponding anomaly identification methods, achieving adaptive anomaly identification processing and greatly improving the accuracy of anomaly identification. Attached Figure Description

[0051] Figure 1 This is an application scenario diagram of an anomaly identification method for base station energy-saving index data applicable to this application in one embodiment;

[0052] Figure 2 This is a schematic diagram illustrating the principle of classifying base station energy-saving indicators in one embodiment;

[0053] Figure 3 This is a schematic diagram illustrating the principle of anomaly identification in one embodiment;

[0054] Figure 4 This is a schematic diagram illustrating the selection of key base station energy-saving indicators in one embodiment;

[0055] Figure 5 This is a structural block diagram of an anomaly detection device for base station energy-saving index data in one embodiment;

[0056] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0058] In one embodiment, such as Figure 1 As shown, a method for anomaly identification of base station energy-saving index data is provided. Taking the application of this method to a computer device as an example, the computer device can be a terminal or a server. The method can be executed independently by the terminal or server, or it can be implemented through interaction between the terminal and the server. Specifically, it includes the following steps:

[0059] Step 102: Obtain the base station energy saving index data sequence corresponding to each base station energy saving index item.

[0060] The base station energy-saving index data sequence includes base station energy-saving index data at different time points. Each base station energy-saving index is used to characterize a specific base station energy-saving indicator. Each base station energy-saving index has different index data at different times; that is, the base station energy-saving index data, arranged sequentially at multiple time points, constitutes the base station energy-saving index data sequence.

[0061] It is understandable that the energy-saving indicators of each base station can be the initial base station energy-saving indicators, or they can be the key base station energy-saving indicators selected from the initial base station energy-saving indicators and to be analyzed for anomalies.

[0062] In some embodiments, base station energy-saving indicators may include at least one of the following: number of energy-saving deployed cells, energy saving, or number of energy-saving recommended cells.

[0063] Step 104: Based on the energy-saving index data sequence of each base station, determine the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each energy-saving index item of each base station.

[0064] Among them, the volatility characterization coefficient is used to characterize the degree of volatility of the base station energy-saving index data sequence. The autocorrelation coefficient is used to characterize the correlation between the base station energy-saving index data at different time points for the same base station energy-saving index item.

[0065] Specifically, for each base station energy-saving indicator, the computer equipment can perform fluctuation analysis based on the corresponding base station energy-saving indicator data sequence to calculate the fluctuation characterization coefficient. The computer equipment can also obtain the autocorrelation coefficient based on the correlation between the base station energy-saving indicator data at different time points in the data sequence.

[0066] In some embodiments, the computer device can calculate the autocorrelation coefficient ACF(k) using the following formula:

[0067]

[0068] Where ACF(k) represents the autocorrelation coefficient with a lag of K periods, Z tLet n be the base station energy-saving index data sequence at any time t, and n be the length of the base station energy-saving index data sequence. This is the mean.

[0069] Step 106: Identify static base station energy-saving indicators from each base station energy-saving indicator item based on the fluctuation characterization coefficient.

[0070] Specifically, computer equipment can select static base station energy-saving indicators from the base station energy-saving indicators based on fluctuation characterization coefficients. It can be understood that the fluctuation characterization coefficients of the selected static base station energy-saving indicators are smaller than those of the other remaining base station energy-saving indicators.

[0071] In some embodiments, the fluctuation characterization coefficient may be the coefficient of variation. The computer equipment may determine base station energy-saving indicators with a coefficient of variation less than or equal to a first preset threshold as static base station energy-saving indicators.

[0072] In some embodiments, the step of calculating the coefficient of variation includes: for each base station energy saving index item, determining the standard deviation of the base station energy saving index data sequence corresponding to the base station energy saving index item; determining the mean of the base station energy saving index data in the base station energy saving index data sequence; and determining the coefficient of variation corresponding to the base station energy saving index item based on the standard deviation and the mean.

[0073] In some embodiments, the coefficient of variation C v This can be achieved using the following formula:

[0074]

[0075] Wherein, σ i For the base station energy saving index data x in the base station energy saving index data sequence i Standard deviation, μ i For the base station energy saving index data x in the base station energy saving index data sequence i The mean.

[0076] Step 108: Based on the autocorrelation coefficient, perform periodic detection on the energy-saving index data sequence of each base station to identify periodic base station energy-saving index items from each base station energy-saving index item.

[0077] Specifically, the computer equipment can iterate through the phase differences of the base station energy-saving index data sequence. When the autocorrelation coefficient of a certain phase difference is greater than the preset coefficient, it is considered that the base station energy-saving index item has periodicity, that is, the base station energy-saving index item is determined to be a periodic base station energy-saving index item.

[0078] Step 110: Determine the energy-saving indicators for unstable base stations based on the base station energy-saving indicators other than the periodic base station energy-saving indicators and the static base station energy-saving indicators.

[0079] In some embodiments, the computer device can determine the remaining base station energy-saving indicators, excluding periodic and static base station energy-saving indicators, and identify these remaining indicators as unstable base station energy-saving indicators. It is understood that the computer device can further screen and analyze the remaining base station energy-saving indicators to identify unstable base station energy-saving indicators.

[0080] In some embodiments, the computer device may first determine the base station energy-saving index items with a coefficient of variation greater than a first preset threshold, and then remove the periodic base station energy-saving index items from the determined base station energy-saving index items (i.e., base station energy-saving index items with a coefficient of variation greater than the first preset threshold), and the remaining base station energy-saving index items are the unstable base station energy-saving index items.

[0081] In some embodiments, a base station energy-saving indicator is determined to be an unstable base station energy-saving indicator if it meets the following conditions:

[0082] c v >c, and does not belong to the periodic base station energy saving indicators;

[0083] Where c is the first preset threshold, c v >c means that the coefficient of variation is greater than the first preset threshold.

[0084] Step 112: The base station energy saving index data sequences corresponding to the static base station energy saving index items, periodic base station energy saving index items, and unstable base station energy saving index items are respectively identified for anomalies according to the corresponding anomaly identification methods.

[0085] It can be understood that static base station energy-saving indicators, periodic base station energy-saving indicators, and unstable base station energy-saving indicators belong to three different categories of base station energy-saving indicators. Therefore, the base station energy-saving indicator data sequences corresponding to these three different categories can be anomaly identified according to their respective anomaly identification methods. That is, different categories of base station energy-saving indicators can adaptively adopt different anomaly identification methods for anomaly identification.

[0086] To facilitate understanding of the indicator category classification, we will now combine... Figure 2 A brief explanation is provided. Figure 2 This is a schematic diagram illustrating the principle of classifying base station energy-saving indicators in one embodiment. See also... Figure 2Based on the coefficient of variation and autocorrelation coefficient, energy-saving indicators for each base station can be divided into static base station energy-saving indicators, periodic base station energy-saving indicators, and unstable base station energy-saving indicators. Static base station energy-saving indicators are selected by having a coefficient of variation less than a first preset threshold. Periodic base station energy-saving indicators are determined by iterating through phase differences; if the autocorrelation coefficient of a certain phase difference is greater than a preset coefficient, it is considered periodic. Unstable base station energy-saving indicators are those with a coefficient of variation greater than the first preset threshold that do not belong to the periodic base station energy-saving indicator category.

[0087] The above-mentioned method for identifying anomalies in base station energy-saving index data determines the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each base station energy-saving index item based on the base station energy-saving index data sequence corresponding to each base station energy-saving index item. This allows for accurate and convenient identification of static and periodic base station energy-saving index items from each base station energy-saving index item, and further identification of unstable base station energy-saving index items. Then, the base station energy-saving index data sequences corresponding to the static, periodic, and unstable base station energy-saving index items are respectively subjected to anomaly identification according to the corresponding anomaly identification methods. This achieves adaptive anomaly identification processing and greatly improves the accuracy of anomaly identification.

[0088] In some embodiments, the static base station energy-saving index item corresponds to the first base station energy-saving index data sequence. In this embodiment, the base station energy-saving index data sequences corresponding to the static base station energy-saving index item, the periodic base station energy-saving index item, and the unstable base station energy-saving index item are respectively anomaly identified according to the corresponding anomaly identification methods, including: sliding a first moving average time window on the first base station energy-saving index data sequence; calculating the mean of the static base station energy-saving index data located within the first moving average time window to obtain a first mean; comparing the first mean with a first dynamic threshold, and identifying abnormal data in the first base station energy-saving index data sequence based on the comparison result.

[0089] It is understandable that the first base station energy-saving index data sequence includes static base station energy-saving index data at different time points. The first dynamic threshold is a threshold that changes dynamically and adaptively over time.

[0090] Specifically, the computer equipment can slide across the first base station energy-saving index data sequence using a first moving average time window. This will continuously generate static base station energy-saving index data that fall within the first moving average time window. The computer equipment can calculate the average of the static base station energy-saving index data within the first moving average time window to obtain a first average value. The computer equipment compares the first average value with a first dynamic threshold. If the first average value is lower than the first dynamic threshold, it is determined that there is abnormal data in the static base station energy-saving index data within the first moving average time window.

[0091] In some embodiments, the first dynamic threshold may be determined based on a first mean and a threshold coefficient α. That is, the first dynamic threshold may be determined based on the product of the first mean and the threshold coefficient α.

[0092] The above embodiments can conveniently identify anomalies and more accurately identify abnormal data based on dynamic thresholds.

[0093] In some embodiments, the periodic base station energy-saving index item corresponds to a third base station energy-saving index data sequence. In this embodiment, the step of identifying anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index item, the periodic base station energy-saving index item, and the unstable base station energy-saving index item, respectively, according to corresponding anomaly identification methods, includes: dividing the third base station energy-saving index data sequence according to a period to obtain periodic base station energy-saving index data for each period; predicting the periodic base station energy-saving index data for the current period based on the periodic base station energy-saving index data of the previous period using a pre-trained periodic anomaly identification model; determining the difference between the predicted periodic base station energy-saving index data for the current period and the periodic base station energy-saving index data for the current period in the third base station energy-saving index data sequence; if the difference is greater than a preset difference threshold, then determining that the periodic base station energy-saving index data for the current period is abnormal.

[0094] Specifically, for periodic base station energy-saving indicators, the computer equipment can pre-train a model using historical periodic base station energy-saving indicator data sequences under the periodic base station energy-saving indicator item to obtain a periodic anomaly identification model. The computer equipment can divide the current periodic base station energy-saving indicator data sequence corresponding to the periodic base station energy-saving indicator item into periods, obtaining periodic base station energy-saving indicator data for each period. For each current period requiring anomaly identification, the periodic base station energy-saving indicator data for the current period is predicted based on the periodic base station energy-saving indicator data of the previous period using the periodic anomaly identification model, and the difference between the predicted periodic base station energy-saving indicator data for the current period and the actual periodic base station energy-saving indicator data for the current period in the third base station energy-saving indicator data sequence is determined. The computer equipment can compare this difference with a preset difference threshold; if it exceeds the preset difference threshold, it is determined that there is an anomaly in the periodic base station energy-saving indicator data of these two adjacent periods.

[0095] The above embodiments are capable of identifying anomalies based on the characteristics of periodic base station energy-saving index data, and can more accurately identify abnormal data.

[0096] In some embodiments, the energy-saving index item of an unstable base station corresponds to a second base station energy-saving index data sequence. In this embodiment, the base station energy-saving index data sequences corresponding to the static base station energy-saving index item, the periodic base station energy-saving index item, and the unstable base station energy-saving index item are respectively anomaly identified according to the corresponding anomaly identification method, including: when there are multiple unstable base station energy-saving index items, determining the unstable base station energy-saving index item with anomaly label from the multiple unstable base station energy-saving index items to obtain the anomaly label index item; for the anomaly label index item, obtaining the historical unstable base station energy-saving index data of the anomaly label index item to determine a second dynamic threshold; sliding a second moving average time window on the second base station energy-saving index data sequence to calculate the mean of the unstable base station energy-saving index data located within the second moving average time window to obtain a second mean; and identifying abnormal data in the second base station energy-saving index data sequence based on the comparison result between the second mean and the second dynamic threshold.

[0097] It is understandable that when there are multiple energy-saving indicators for unstable base stations, computer equipment can further classify these indicators based on the presence or absence of abnormal tags, dividing them into indicators with abnormal tags and those without. Different processing methods can be applied to indicators with abnormal tags and those without. The second base station energy-saving indicator data sequence includes energy-saving indicator data for unstable base stations at different time points.

[0098] Specifically, the computer equipment can identify unstable base station energy-saving indicator items with abnormal labels from multiple unstable base station energy-saving indicator items, thus obtaining abnormally labeled indicator items. It can be understood that abnormally labeled indicator items are unstable base station energy-saving indicator items that have been identified as having abnormal data in previous processing, and therefore, abnormal labels are added to them. For abnormally labeled indicator items, the computer equipment can obtain historical unstable base station energy-saving indicator data for the abnormally labeled indicator items to determine a second dynamic threshold. It can be understood that the second dynamic threshold is adaptively determined based on the historical unstable base station energy-saving indicator data of the abnormally labeled indicator items, rather than being a fixed threshold, making it more accurate.

[0099] The computer equipment can slide across the second base station energy-saving index data sequence using a second moving average time window to calculate the average of unstable base station energy-saving index data within the second moving average time window, thus obtaining a second average value. The computer equipment can then identify abnormal data in the second base station energy-saving index data sequence based on the comparison between the second average value and a second dynamic threshold. Specifically, if the second average value is lower than the second dynamic threshold, it is determined that there is abnormal data in the unstable base station energy-saving index data within the second moving average time window.

[0100] In other embodiments, the computer device may also perform supervised training in advance based on historical unstable base station energy-saving index data under the abnormal label index item to train an anomaly identification model. The computer device can input the second base station energy-saving index data sequence to be identified into the anomaly identification model to perform anomaly identification and identify abnormal data in the second base station energy-saving index data sequence.

[0101] In some embodiments, the method further includes: for the energy-saving index items without abnormal labels among multiple unstable base station energy-saving index items, performing preliminary anomaly identification on the unstable base station energy-saving index data without abnormal labels through an unsupervised anomaly identification model to obtain the unstable base station energy-saving index data that is initially identified as abnormal; and performing advanced anomaly identification on the unstable base station energy-saving index data that is initially identified as abnormal according to a second preset threshold to obtain the unstable base station energy-saving index data that is finally identified as abnormal.

[0102] It is understandable that, for multiple unstable base station energy-saving indicators without abnormal labels, computer equipment can identify abnormal data through unsupervised training combined with a second preset threshold.

[0103] Specifically, the computer equipment can perform preliminary anomaly identification on the energy-saving index data of unstable base stations under no-abnormal label indicators using an unsupervised anomaly identification model to obtain preliminarily identified abnormal energy-saving index data of unstable base stations. Then, the computer equipment can perform advanced anomaly identification on the preliminarily identified abnormal energy-saving index data of unstable base stations according to a second preset threshold to obtain the final identified abnormal energy-saving index data of unstable base stations.

[0104] In some embodiments, for the sequence of unstable base station energy-saving index data that has been initially identified as abnormal, i.e., the initial anomaly identification sequence, a third moving average time window can be used to slide on the initial anomaly identification sequence to calculate the average value of the unstable base station energy-saving index data within the third moving average time window, and a third average value is obtained. If the third average value is lower than the second preset threshold, it is determined that there is abnormal data in the unstable base station energy-saving index data within the third moving average time window.

[0105] Figure 3 This is a schematic diagram illustrating the principle of anomaly identification in one embodiment. See also... Figure 3 For static base station energy-saving indicators, a first dynamic threshold can be determined. If the average value of the static base station energy-saving indicator data within a time window is lower than the first dynamic threshold, it is considered an anomaly. For periodic base station energy-saving indicators, a periodic anomaly identification model can be pre-trained for anomaly identification. For anomaly identification of unstable base station energy-saving indicators, it is divided into anomaly-labeled indicators and unlabeled indicators. For anomaly-labeled indicators, the threshold can be updated based on historical data. For unlabeled indicators, the threshold can be combined with an unsupervised anomaly identification method for anomaly identification.

[0106] The above embodiments can perform targeted anomaly identification processing based on the presence or absence of labels, and can more accurately identify abnormal data.

[0107] In some embodiments, each base station energy-saving indicator is a key base station energy-saving indicator. The step of obtaining the base station energy-saving indicator data sequence corresponding to each base station energy-saving indicator includes: obtaining the base station energy-saving indicator data sequence corresponding to each candidate base station energy-saving indicator; for each candidate base station energy-saving indicator, dividing the base station energy-saving indicator data sequence corresponding to the candidate base station energy-saving indicator into two sub-sequences; performing hypothesis testing on the two sub-sequences to obtain hypothesis testing results; and identifying key base station energy-saving indicators that are sensitive to anomalies from the candidate base station energy-saving indicators based on the hypothesis testing results.

[0108] Specifically, computer equipment can divide the base station energy-saving indicator data sequence corresponding to the candidate base station energy-saving indicator items into two sub-sequences based on the target division criteria. The target division criteria can be the energy saving in the base station energy-saving system. That is, based on the two cases where the energy saving increment at multiple time points is greater than 0 or equal to 0, the base station energy-saving indicator data sequence corresponding to each candidate base station energy-saving indicator item can be divided into two sub-sequences. For example, based on whether the energy saving increment at m time points is greater than 0 or equal to 0, the m base station energy-saving indicator data in the base station energy-saving indicator data sequence of each candidate base station energy-saving indicator item can be divided into two groups, resulting in two sub-sequences.

[0109] In hypothesis testing, the null hypothesis is that the two subsequences are indistinguishable, while the alternative hypothesis is that they are different. The goal is to identify base station energy-saving indicators that differ between the two subsequences. Significant differences indicate that the energy-saving indicator is sensitive to anomalies. Therefore, the computer can perform hypothesis testing on the two subsequences to obtain the results. Based on these results, key energy-saving indicators sensitive to anomalies are identified from the candidate energy-saving indicators. Specifically, the computer compares the hypothesis test probability P-value with a preset probability threshold K. If P-value is less than K, the null hypothesis is rejected, and the alternative hypothesis is accepted. This indicates that the energy-saving indicator is highly sensitive to system anomalies and is a key indicator for subsequent anomaly identification. By screening for highly sensitive and significant energy-saving indicators, the health status of the base station energy-saving system can be more accurately assessed.

[0110] Figure 4 This is a schematic diagram illustrating the selection of key base station energy-saving indicators in one embodiment. From Figure 4 It can be seen that, firstly, the base station energy-saving index data sequences under the candidate base station energy-saving index items of each layer of the system are used as input. Then, the base station energy-saving index data sequences are grouped according to the gold standard (i.e., the target division basis) in the base station energy-saving system, and finally, hypothesis testing is performed. Specifically, each candidate base station energy-saving index item's base station energy-saving index data sequence contains base station energy-saving index data at m time points, denoted as {k1,k2,…,k...}. m Based on the gold standard, they are divided into two groups, for example, {k1,k4,…,k}. m-1} and {k2,k3,…,k m These two groups were selected. Base station energy-saving indicators with a hypothesis test probability P-value less than a probability threshold were chosen as the key base station energy-saving indicators.

[0111] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0112] Based on the same inventive concept, this application also provides an anomaly identification device for base station energy-saving index data, used to implement the above-mentioned anomaly identification method for base station energy-saving index data. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations in one or more embodiments of the base station energy-saving index data anomaly identification device provided below can be found in the limitations of the base station energy-saving index data anomaly identification method described above, and will not be repeated here.

[0113] In one embodiment, such as Figure 5 As shown, an anomaly identification device for base station energy-saving index data is provided, comprising: an index acquisition module 502, a coefficient determination module 504, an index identification module 506, and an anomaly identification module 508, wherein:

[0114] The indicator acquisition module 502 is used to acquire the base station energy-saving indicator data sequence corresponding to each base station energy-saving indicator item; the base station energy-saving indicator data sequence includes the base station energy-saving indicator data of the base station energy-saving indicator item at different time points.

[0115] The coefficient determination module 504 is used to determine the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each energy-saving index item of each base station based on the data sequence of each base station energy-saving index.

[0116] The indicator identification module 506 is used to identify static base station energy-saving indicators from each base station energy-saving indicator item according to the fluctuation characterization coefficient; to perform periodic detection on each base station energy-saving indicator data sequence according to the autocorrelation coefficient to identify periodic base station energy-saving indicators from each base station energy-saving indicator item; and to determine unstable base station energy-saving indicators based on base station energy-saving indicators other than the periodic base station energy-saving indicators and the static base station energy-saving indicators.

[0117] The anomaly identification module 508 is used to identify anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items, respectively, according to the corresponding anomaly identification methods.

[0118] In some embodiments, the coefficient determination module 504 is further configured to, for each base station energy-saving index item, determine the standard deviation of the base station energy-saving index data sequence corresponding to the base station energy-saving index item; determine the mean of the base station energy-saving index data in the base station energy-saving index data sequence; and determine the coefficient of variation corresponding to the base station energy-saving index item based on the standard deviation and the mean.

[0119] The indicator identification module 506 is also used to identify base station energy-saving indicator items with a coefficient of variation less than or equal to a first preset threshold as static base station energy-saving indicator items.

[0120] In some embodiments, the static base station energy-saving index item corresponds to the first base station energy-saving index data sequence; the anomaly identification module 508 is further configured to slide on the first base station energy-saving index data sequence through a first moving average time window; calculate the mean of the static base station energy-saving index data located within the first moving average time window to obtain a first mean; compare the first mean with a first dynamic threshold, and identify abnormal data in the first base station energy-saving index data sequence based on the comparison result.

[0121] In some embodiments, the unstable base station energy-saving indicator item corresponds to a second base station energy-saving indicator data sequence; the anomaly identification module 508 is further configured to, when there are multiple unstable base station energy-saving indicator items, determine the unstable base station energy-saving indicator item with an anomaly label from the multiple unstable base station energy-saving indicator items, and obtain the anomaly label indicator item; for the anomaly label indicator item, obtain the historical unstable base station energy-saving indicator data of the anomaly label indicator item, and determine a second dynamic threshold; slide the second moving average time window on the second base station energy-saving indicator data sequence, calculate the mean of the unstable base station energy-saving indicator data located within the second moving average time window, and obtain a second mean; and identify abnormal data in the second base station energy-saving indicator data sequence based on the comparison result between the second mean and the second dynamic threshold.

[0122] In some embodiments, the anomaly identification module 508 is further configured to, for the energy-saving index items without anomaly labels among the plurality of unstable base station energy-saving index items, perform preliminary anomaly identification on the unstable base station energy-saving index data under the unsupervised anomaly identification model to obtain the unstable base station energy-saving index data that is initially identified as abnormal; and perform advanced anomaly identification on the unstable base station energy-saving index data that is initially identified as abnormal according to a second preset threshold to obtain the unstable base station energy-saving index data that is finally identified as abnormal.

[0123] In some embodiments, the periodic base station energy-saving index item corresponds to a third base station energy-saving index data sequence; the anomaly identification module 508 is further configured to divide the third base station energy-saving index data sequence according to a period to obtain periodic base station energy-saving index data for each period; through a pre-trained periodic anomaly identification model, predict the periodic base station energy-saving index data for the current period based on the periodic base station energy-saving index data of the previous period, and determine the difference between the predicted periodic base station energy-saving index data for the current period and the periodic base station energy-saving index data for the current period in the third base station energy-saving index data sequence; if the difference is greater than a preset difference threshold, it is determined that the periodic base station energy-saving index data for the current period is abnormal.

[0124] In some embodiments, the indicator acquisition module 502 is further configured to acquire base station energy-saving indicator data sequences corresponding to candidate base station energy-saving indicator items respectively; for each candidate base station energy-saving indicator item, divide the base station energy-saving indicator data sequence corresponding to the candidate base station energy-saving indicator item into two sub-sequences; perform hypothesis testing on the two sub-sequences to obtain hypothesis testing results; and identify key base station energy-saving indicator items that are sensitive to anomalies from the candidate base station energy-saving indicator items based on the hypothesis testing results.

[0125] The modules in the aforementioned anomaly detection device for base station energy-saving indicators can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0126] In one embodiment, a computer device is provided, which may be any one of a main base station or other network element device, and its internal structure diagram may be as follows: Figure 6 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for identifying anomalies in base station energy-saving indicator data.

[0127] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0128] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above embodiments.

[0129] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above embodiments.

[0130] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above embodiments.

[0131] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0132] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0133] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0134] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for identifying anomalies in base station energy-saving index data, characterized in that, The method includes: Obtain the base station energy-saving index data sequence corresponding to each base station energy-saving index item; the base station energy-saving index data sequence includes the base station energy-saving index data of the base station energy-saving index item at different time points; Based on the data sequence of each base station energy-saving index, determine the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each base station energy-saving index item; Based on the fluctuation characterization coefficient, identify static base station energy-saving indicators from each base station energy-saving indicator item; Based on the autocorrelation coefficient, periodic detection is performed on the data sequence of each base station energy-saving index to identify periodic base station energy-saving index items from each base station energy-saving index item; Based on the base station energy-saving indicators other than the periodic base station energy-saving indicators and the static base station energy-saving indicators, determine the unstable base station energy-saving indicators; The base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items are respectively identified as anomalies according to the corresponding anomaly identification methods. The static base station energy-saving index is selected by having a coefficient of variation less than a first preset threshold. The periodic base station energy-saving index is a base station energy-saving index with an autocorrelation coefficient of phase difference greater than a preset coefficient. The unstable base station energy-saving index is a base station energy-saving index with a coefficient of variation greater than the first preset threshold that does not belong to the periodic base station energy-saving index.

2. The method according to claim 1, characterized in that, The fluctuation characterization coefficient is the coefficient of variation; The step of determining the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each energy-saving index item of each base station based on the data sequence of each base station energy-saving index includes: For each base station energy-saving indicator item, determine the standard deviation of the base station energy-saving indicator data sequence corresponding to the base station energy-saving indicator item; Determine the mean value of the base station energy-saving index data in the base station energy-saving index data sequence; Based on the standard deviation and mean, determine the coefficient of variation corresponding to the base station energy-saving index item; The step of identifying static base station energy-saving indicators from each base station energy-saving indicator item based on the fluctuation characterization coefficient includes: Base station energy-saving indicators with a coefficient of variation less than or equal to the first preset threshold are defined as static base station energy-saving indicators.

3. The method according to claim 1, characterized in that, The static base station energy-saving index item corresponds to the first base station energy-saving index data sequence; The step of identifying anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items according to the corresponding anomaly identification methods includes: The data is slid across the energy-saving index data sequence of the first base station using a first moving average time window. The average value of static base station energy-saving index data within the first moving average time window is calculated to obtain the first average value. The first mean value is compared with the first dynamic threshold, and abnormal data in the first base station energy-saving index data sequence is identified based on the comparison result.

4. The method according to claim 1, characterized in that, The unstable base station energy-saving index item corresponds to the second base station energy-saving index data sequence; The step of identifying anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items according to the corresponding anomaly identification methods includes: When there are multiple energy-saving indicators for unstable base stations, the energy-saving indicators for unstable base stations with abnormal labels are determined from the multiple energy-saving indicators for unstable base stations to obtain the abnormal label indicators. For the abnormal label indicator item, obtain the historical unstable base station energy saving indicator data of the abnormal label indicator item, and determine the second dynamic threshold. By sliding the second moving average time window over the second base station energy-saving index data sequence, the average value of the unstable base station energy-saving index data located within the second moving average time window is calculated to obtain the second average value. Based on the comparison between the second mean and the second dynamic threshold, abnormal data in the second base station energy-saving index data sequence are identified.

5. The method according to claim 4, characterized in that, The method further includes: For the energy-saving indicators without abnormal labels among the multiple unstable base station energy-saving indicators, the energy-saving indicator data of unstable base stations without abnormal labels is initially identified through an unsupervised anomaly identification model to obtain the energy-saving indicator data of unstable base stations with preliminary anomalies. Based on the second preset threshold, the energy-saving index data of unstable base stations that were initially identified as abnormal are further identified as abnormal, and the energy-saving index data of unstable base stations that are finally identified as abnormal are obtained.

6. The method according to claim 1, characterized in that, The periodic base station energy-saving index item corresponds to the third base station energy-saving index data sequence; The step of identifying anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items according to the corresponding anomaly identification methods includes: The energy-saving index data sequence of the third base station is divided according to the period to obtain the periodic base station energy-saving index data for each period. By using a pre-trained periodic anomaly identification model, the periodic base station energy-saving index data for the current period is predicted based on the periodic base station energy-saving index data of the previous period, and the difference between the predicted periodic base station energy-saving index data for the current period and the periodic base station energy-saving index data for the current period in the third base station energy-saving index data sequence is determined. If the difference is greater than a preset difference threshold, it is determined that the periodic base station energy-saving index data for the current period is abnormal.

7. The method according to any one of claims 1 to 6, characterized in that, Each base station energy-saving indicator is a key base station energy-saving indicator; obtaining the base station energy-saving indicator data sequence corresponding to each base station energy-saving indicator includes: Obtain the base station energy-saving index data sequence corresponding to each candidate base station energy-saving index item; For each candidate base station energy saving indicator item, the base station energy saving indicator data sequence corresponding to the candidate base station energy saving indicator item is divided into two sub-sequences; Hypothesis testing was performed on the two sets of subsequences to obtain the hypothesis test results; Based on the hypothesis test results, key base station energy-saving indicators that are sensitive to anomalies are identified from the candidate base station energy-saving indicators.

8. A device for identifying anomalies in base station energy-saving index data, characterized in that, The device includes: The indicator acquisition module is used to acquire the base station energy-saving indicator data sequence corresponding to each base station energy-saving indicator item; the base station energy-saving indicator data sequence includes the base station energy-saving indicator data of the base station energy-saving indicator item at different time points. The coefficient determination module is used to determine the fluctuation characterization coefficient and autocorrelation coefficient corresponding to each energy-saving index item of each base station based on the data sequence of each base station energy-saving index. The indicator identification module is used to identify static base station energy-saving indicators from each base station energy-saving indicator item based on the fluctuation characterization coefficient; to perform periodic detection on each base station energy-saving indicator data sequence based on the autocorrelation coefficient to identify periodic base station energy-saving indicators from each base station energy-saving indicator item; and to determine unstable base station energy-saving indicators based on base station energy-saving indicators other than the periodic base station energy-saving indicators and the static base station energy-saving indicators. The anomaly identification module is used to identify anomalies in the base station energy-saving index data sequences corresponding to the static base station energy-saving index items, periodic base station energy-saving index items, and unstable base station energy-saving index items, respectively, according to the corresponding anomaly identification methods. The static base station energy-saving index is selected by having a coefficient of variation less than a first preset threshold. The periodic base station energy-saving index is a base station energy-saving index with an autocorrelation coefficient of phase difference greater than a preset coefficient. The unstable base station energy-saving index is a base station energy-saving index with a coefficient of variation greater than the first preset threshold that does not belong to the periodic base station energy-saving index.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images