Passive indoor distribution system monitoring model training method and device and computer equipment

Abstract

The invention discloses a passive indoor distribution system monitoring model training method and device and computer equipment. The method comprises the following steps: collecting standing-wave ratio data of a plurality of passive indoor distribution systems; determining the fluctuation condition of the standing-wave ratio of the plurality of passive indoor distribution systems according to the standing-wave ratio data of the plurality of passive indoor distribution systems, and determining a first abnormal passive indoor distribution system from the standing-wave ratio data of the plurality of passive indoor distribution systems according to the fluctuation condition of the standing-wave ratio of the plurality of passive indoor distribution systems; determining a training data set according to the user terminal data under the coverage of the first abnormal passive indoor distribution system; and training based on the training data set to obtain an identification model, the identification model being used for identifying the user terminal data under the coverage of the to-be-monitored indoor distribution system to determine whether the to-be-monitored indoor distribution system is abnormal. The technical problem of low fault prediction accuracy of the passive indoor distribution system in related technologies is at least solved.

Description

Technical Field

[0001] The present application relates to the field of wireless communication technologies, and in particular, to a method, apparatus, and computer device for training a monitoring model of a passive indoor distribution system. Background Art

[0002] The indoor distribution system is the main means for operators to solve the problem of insufficient indoor deep coverage. There are a large number of passive devices in the system. If the connections between devices are loose or dropped, the network quality will rapidly decline. As typical deployment scenarios of the indoor distribution system, hospitals, large shopping malls, office buildings, and universities are prone to damage system devices during irregular renovations, and the user perception deteriorates significantly. Usually, problems can only be discovered after user complaints. Currently, the commonly used indoor distribution monitoring methods in the industry are to monitor performance indicators such as the excellent rate of RSRP (Reference Signal Received Power Excellence Rate), the utilization rate of PRB (Physical Resource Block Utilization), and the number of users. However, in the indoor distribution scenario, the user tidal effect is serious, and the performance indicators fluctuate greatly. It is not easy to judge whether the indoor distribution system is abnormal based on the changes in user performance indicators, and the monitoring accuracy is relatively low. In addition, using this monitoring method is prone to misjudgment caused by the deterioration of the performance of resident users due to coverage blind spots in the indoor distribution system. For example, there is no abnormality in the indoor distribution system of a certain building, but the meeting room is not covered. When resident users enter the meeting room to have a meeting, the relevant indicators will deteriorate sharply, causing algorithm misjudgment.

[0003] Facing many problems such as no alarms for passive devices and a serious user tidal effect, traditional monitoring methods are difficult to accurately predict and locate hidden obstacles in the indoor distribution before they occur. Summary of the Invention

[0004] Embodiments of the present application provide a method, apparatus, and computer device for training a monitoring model of a passive indoor distribution system, so as to at least solve the technical problem of relatively low accuracy in predicting faults of passive indoor distribution systems in related technologies.

[0005] According to an aspect of an embodiment of the present application, a method for training a monitoring model of a passive indoor distribution system is provided, including: collecting standing wave ratio data of multiple passive indoor distribution systems; determining the fluctuation of the standing wave ratio of the multiple passive indoor distribution systems according to the standing wave ratio data of the multiple passive indoor distribution systems, and determining a first abnormal passive indoor distribution system from the standing wave ratio data of the multiple passive indoor distribution systems according to the fluctuation of the standing wave ratio of the multiple passive indoor distribution systems; determining a training data set according to the user terminal data covered by the first abnormal passive indoor distribution system; and training an identification model based on the training data set, where the identification model is used to identify the user terminal data covered by the to-be-monitored indoor distribution system to determine whether the to-be-monitored indoor distribution system is abnormal.

[0006] Optionally, determining the fluctuation condition of the standing wave ratios of the plurality of passive distributed antenna systems according to the standing wave ratio data of the plurality of passive distributed antenna systems includes: determining the moving average of the standing wave ratios of the plurality of passive distributed antenna systems according to the standing wave ratio data of the plurality of passive distributed antenna systems; obtaining the standard deviation of the moving average, and determining the weighted standard deviation according to a preset weight and the standard deviation; determining the upper limit of standing wave ratio fluctuation and the lower limit of standing wave ratio fluctuation according to the moving average and the weighted standard deviation respectively; and determining the fluctuation condition of the standing wave ratios of the plurality of passive distributed antenna systems according to the comparison results between the standing wave ratios of the plurality of passive distributed antenna systems and the upper limit of standing wave ratio fluctuation and the lower limit of standing wave ratio fluctuation.

[0007] Optionally, determining a training data set according to the user terminal data covered by the first abnormal passive distributed antenna system includes: obtaining the user terminal data covered by the first abnormal passive distributed antenna system, where the user terminal data includes at least one of the following: the received power of the reference signal received by the user terminal, the physical resource block utilization rate, and the number of user terminals; associating the user terminal data covered by the first abnormal passive distributed antenna system with the standing wave ratio data of the first abnormal passive distributed antenna system to obtain processed data; and converting the processed data into a multi-dimensional tensor format to obtain the training data set, where the multi-dimensional tensor includes at least: the number of samples in the training data set, the number of standing wave ratios in the samples, and the labels corresponding to the samples.

[0008] Optionally, training an identification model based on the training data set includes: obtaining an initial model, where the initial model is a long short-term memory model; dividing the training data set into a training set, a validation set, and a test set; training the initial model with the training set, and determining that the initial model converges to obtain the identification model when the loss function values of the initial model on the training set and the validation set reach a first threshold and the accuracy rate of the initial model on the training set and the validation set reaches a second threshold.

[0009] Optionally, the method further includes: respectively obtaining the numbers of true positives, false positives, true negatives, and false negatives output by the initial model; and determining the accuracy rate of the initial model according to the numbers of true positives, false positives, true negatives, and false negatives output by the initial model.

[0010] Optionally, the method further includes: obtaining the alarm data of the plurality of passive distributed antenna systems, determining the second abnormal passive distributed antenna system among the plurality of passive distributed antenna systems according to the alarm data; and adding the user terminal data covered by the second abnormal passive distributed antenna system to the training data set.

[0011] Optionally, identify the user terminal data covered by the in-building distribution system to be monitored, including: obtaining the user terminal data covered by the in-building distribution system to be monitored at a preset period; using the identification model to analyze the obtained user terminal data covered by the in-building distribution system to be monitored to determine whether the in-building distribution system to be monitored is abnormal.

[0012] According to another aspect of the embodiments of the present application, there is also provided a passive in-building distribution system monitoring model training device, including: an acquisition module for acquiring original data, where the original data includes: historical transaction records of data commodities, data commodities to be traded, and information of users to be traded; an extraction module for extracting features from the original data to obtain features of data commodities to be traded and features of users to be traded, where the features of data commodities to be traded and the features of users to be traded both include: category features, numerical features, and sequence features; a vector module for respectively inputting the features of data commodities to be traded and the features of users to be traded into a user tower and a commodity tower in a data commodity recommendation model to obtain a user tower embedding vector and a commodity tower embedding vector; a recommendation module for determining recommended data commodities corresponding to the users to be traded based on the user tower embedding vector and the commodity tower embedding vector.

[0013] According to yet another aspect of the embodiments of the present application, there is also provided a computer device, including: a memory and a processor, where the memory is used to store program instructions; the processor is connected to the memory and is used to execute the above-mentioned passive in-building distribution system monitoring model training method.

[0014] According to still another aspect of the embodiments of the present application, there is also provided a non-volatile storage medium, which includes a stored computer program, where the device where the non-volatile storage medium is located executes the above-mentioned passive in-building distribution system monitoring model training method by running the computer program.

[0015] According to still another aspect of the embodiments of the present application, there is also provided a computer program product, including computer instructions, where the computer instructions, when executed by a processor, implement the above-mentioned passive in-building distribution system monitoring model training method.

[0016] In the embodiments of the present application, the standing wave ratio data of multiple passive distributed antenna systems is collected; the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems is determined according to the standing wave ratio data of the multiple passive distributed antenna systems, and a first abnormal passive distributed antenna system is determined from the standing wave ratio data of the multiple passive distributed antenna systems according to the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems; a training data set is determined according to the user terminal data covered by the first abnormal passive distributed antenna system; an identification model is trained based on the training data set, and the identification model is used to identify the user terminal data covered by the to-be-monitored distributed antenna system to determine whether the to-be-monitored distributed antenna system is abnormal, thereby achieving the purpose of determining the abnormal distributed antenna system data according to the fluctuation condition of the standing wave ratio and training an abnormal distributed antenna system identification model by using the determined abnormal distributed antenna system data, thereby realizing the technical effect of improving the fault prediction accuracy of the passive distributed antenna system, and further solving the technical problem of low fault prediction accuracy of the passive distributed antenna system in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The illustrative embodiments and descriptions thereof of the present application are used to explain the present application, and do not constitute an improper limitation of the present application. In the drawings:

[0018] Figure 1 is a hardware structure block diagram of a computer terminal for implementing a method for training a monitoring model of a passive distributed antenna system according to an embodiment of the present application;

[0019] Figure 2 is a flowchart of a method for training a monitoring model of a passive distributed antenna system according to an embodiment of the present application;

[0020] Figure 3 is a schematic structural diagram of a monitoring model of a passive distributed antenna system according to an embodiment of the present application;

[0021] Figure 4 is a structural diagram of a device for training a monitoring model of a passive distributed antenna system according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0023] It should be noted that the terms "first", "second", etc. in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0024] The information collected in the embodiments of the present application is information and data authorized by the user or fully authorized by all parties. And for the processing of relevant data such as collection, storage, use, processing, transmission, provision, disclosure and application, etc., all comply with the relevant laws, regulations and standards of the relevant regions, take necessary confidentiality measures, do not violate public order and good customs, and provide corresponding operation entrances for users to choose to authorize or reject the results of automated decision-making; if the user chooses to reject, the expert decision-making process will be entered.

[0025] To solve the problems existing in the related art, the embodiments of the present application provide a method for training a monitoring model of a passive indoor distribution system, and this method can run on Figure 1 the computer terminal shown below. The following is an explanatory description of this computer terminal.

[0026] The embodiments of the method for training a monitoring model of a passive indoor distribution system provided by the embodiments of the present application can be executed on a mobile terminal, a computer terminal or a similar computing device. Figure 1 The following shows a hardware structure block diagram of a computer terminal for implementing the method for training a monitoring model of a passive indoor distribution system. As Figure 1 shown, the computer terminal 10 may include one or more (shown as 102a, 102b,..., 102n in the figure) processors (the processor may include, but is not limited to, processing devices such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission module 106 for communication functions connected by wired and / or wireless networks. In addition, it may further include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, and a BUS bus. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 may further include more Figure 1more or fewer components as shown, or having a configuration different from that shown in Figure 1 that shown.

[0027] It should be noted that one or more of the above-mentioned processors and / or other data processing circuits can generally be referred to as "data processing circuits" herein. The data processing circuit can be embodied in software, hardware, firmware, or any combination thereof, in whole or in part. In addition, the data processing circuit can be a single independent processing module, or be incorporated in whole or in part into any one of the other elements in the computer terminal 10. As involved in the embodiments of the present application, the data processing circuit is a processor control (such as the selection of a variable resistor terminal path connected to an interface).

[0028] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the passive indoor distribution system monitoring model training method in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the above-mentioned passive indoor distribution system monitoring model training method. The memory 104 can include high-speed random access memory, and can also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memories. In some instances, the memory 104 can further include memories remotely located relative to the processor, and these remote memories can be connected to the computer terminal 10 through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, enterprise intranets, local area networks, mobile communication networks, and combinations thereof.

[0029] The transmission module 106 is used to receive or send data via a network. Specific examples of the above-mentioned network can include the wireless network provided by the communication provider of the computer terminal 10. In one instance, the transmission module 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one instance, the transmission module 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0030] The display can be, for example, a touch-screen liquid crystal display (LCD), which enables the user to interact with the user interface of the computer terminal 10.

[0031] It should be noted here that in some alternative embodiments, the above Figure 1 shown computer terminal can include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware elements and software elements. It should be pointed out thatFigure 1 This is just an example of a specific concrete instance and is intended to illustrate the types of components that may exist in the above computer terminal.

[0032] Under the above operating environment, an embodiment of a method for training a monitoring model of a passive indoor distribution system is provided in an embodiment of the present application. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0033] Figure 2 is a flowchart of a method for training a monitoring model of a passive indoor distribution system according to an embodiment of the present application, as Figure 2 shown, the method includes the following steps:

[0034] Step S202, collect the standing wave ratio data of multiple passive indoor distribution systems;

[0035] In step S202, the transmission quality of the physical link signal is often measured by the standing wave ratio. When the standing wave ratio is 1, the impedance of the feeder and the antenna is completely matched, and all the energy is radiated by the antenna without reflection; when the standing wave ratio is greater than 1, part of the energy is reflected and becomes heat energy, and the feeder heats up. The larger the standing wave ratio, the worse the impedance matching, the more energy is reflected, and the smaller the antenna signal strength. When the indoor distribution system is damaged or the connection of passive devices is loose, the standing wave ratio increases. Therefore, the standing wave ratio can be used as an evaluation index for monitoring whether the indoor distribution system is abnormal. However, in actual application scenarios, the common situation is that part of the indoor distribution system is damaged, and at this time, no standing wave alarm is generated. Since the installation positions and environments of each signal source are different, there is no unified standard for the standing wave ratio index of the indoor distribution system. For example, the standing wave ratio of a normal indoor distribution system may be the same as that of a faulty indoor distribution system. Therefore, in addition to the standing wave alarm, it is impossible to directly judge whether the indoor distribution system is abnormal only based on a single standing wave ratio value.

[0036] Step S204, determine the fluctuation situation of the standing wave ratios of the multiple passive indoor distribution systems according to the standing wave ratio data of the multiple passive indoor distribution systems, and determine the first abnormal passive indoor distribution system from the standing wave ratio data of the multiple passive indoor distribution systems according to the fluctuation situation of the standing wave ratios of the multiple passive indoor distribution systems;

[0037] In step S204, the first abnormal passive distributed antenna system includes: determined according to the fluctuation of the standing wave ratio, which is a hidden abnormality and needs to be explained. A hidden abnormality refers to a concealed problem in the distributed antenna system where there is no direct hardware alarm, but the signal transmission quality gradually deteriorates. Such abnormalities may be caused by imperceptible reasons such as loose connections, fine-tuning of antenna positions, and signal attenuation caused by environmental changes. These reasons may not immediately cause significant changes in performance indicators or hardware alarms, but will gradually affect the signal transmission quality over a period of time. The characteristic of a hidden abnormality is that it may have a relatively small impact on the user experience in the initial stage and is not easily detected by conventional monitoring methods. However, if not dealt with in a timely manner, it will gradually evolve into an obvious abnormality, seriously affecting user communication. An obvious abnormality refers to an obvious fault in the distributed antenna system that can be directly detected through an alarm system or intuitive detection means. Such abnormalities are usually accompanied by clear hardware alarms, such as antenna damage, feeder disconnection, power splitter failure, etc. These faults will cause physical link signal transmission quality indicators such as the standing wave ratio to exceed the normal range, thus triggering an alarm. An obvious abnormality can directly affect the signal transmission quality, resulting in a significant decline in user perception and usually requires immediate on-site maintenance and repair.

[0038] Step S206, determining a training data set according to the user terminal data covered by the first abnormal passive distributed antenna system;

[0039] Step S208, training an identification model based on the training data set, where the identification model is used to identify the user terminal data covered by the to-be-monitored distributed antenna system to determine whether the to-be-monitored distributed antenna system is abnormal.

[0040] In step S208, the identification model is used to identify according to the real-time data of the user terminal to determine the abnormal distributed antenna system.

[0041] Through the above steps S202 to S208, the standing wave ratio data of multiple passive distributed antenna systems is collected; the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems is determined according to the standing wave ratio data of the multiple passive distributed antenna systems, and a first abnormal passive distributed antenna system is determined from the standing wave ratio data of the multiple passive distributed antenna systems according to the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems; a training data set is determined according to the user terminal data covered by the first abnormal passive distributed antenna system; an identification model is trained based on the training data set, and the identification model is used to identify the user terminal data covered by the to-be-monitored distributed antenna system to determine whether the to-be-monitored distributed antenna system is abnormal, thereby achieving the purpose of determining abnormal distributed antenna system data according to the fluctuation condition of the standing wave ratio and training an abnormal distributed antenna system identification model by using the determined abnormal distributed antenna system data, thereby realizing the technical effect of improving the fault prediction accuracy of the passive distributed antenna system, and further solving the technical problem of low fault prediction accuracy of the passive distributed antenna system in the related art. The following is a detailed description.

[0042] In some embodiments of the present application, the specific steps for determining the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems according to the standing wave ratio data of the multiple passive distributed antenna systems are as follows: determining the moving average of the standing wave ratio of the multiple passive distributed antenna systems according to the standing wave ratio data of the multiple passive distributed antenna systems; obtaining the standard deviation of the moving average, and determining the weighted standard deviation according to a preset weight and the standard deviation; determining the standing wave ratio fluctuation upper limit and the standing wave ratio fluctuation lower limit according to the moving average and the weighted standard deviation respectively; determining the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems according to the comparison result between the standing wave ratio of the multiple passive distributed antenna systems and the standing wave ratio fluctuation upper limit and the standing wave ratio fluctuation lower limit.

[0043] Specifically, when initially screening abnormal distributed antenna data from a large amount of data as a training set, the long-term change characteristics of the standing wave ratio can be calculated by using the statistical characteristics of statistics. According to the change rule, the distributed antenna systems with hidden obstacles can be initially screened out. If the standing wave ratio fluctuates within the normal range, no intervention is required; if the standing wave ratio is abnormally beyond the normal range, even if there is no standing wave alarm for the device at this time, the distributed antenna system still has hidden abnormalities and needs on-site troubleshooting. In an optional manner, the distributed antenna systems with hidden obstacles can be screened out by the change trend of the moving average and the trend weighting of the standard deviation. The specific calculation is as shown in the following formula:

[0044] VSWR UP =VSWR MID +m……std(VSWR MID ) Equation (1-1)

[0045] VSWR LOW =VSWR MID-m * std(VSWR MID ) formula (1-2)

[0046] Wherein, VSWR UP represents the upper limit of the standing wave ratio fluctuation, VSWR LOW represents the lower limit of the standing wave ratio fluctuation, m represents a preset weight, std(·) represents the standard deviation, and VSWR MID represents the moving average of the standing wave ratio.

[0047] It should be noted that m is used to adjust the degree of fluctuation.

[0048] In an optional manner, when the standing wave ratio is greater than VSWR UP , it is determined that the indoor distribution system is abnormal; when the maintenance personnel finish troubleshooting on-site, if the standing wave ratio is less than VSWR LOW , it is determined that the troubleshooting is completed. The determination of the m value is crucial for accurately screening out abnormal indoor distribution. A larger m value results in a wider normal fluctuation range, and a smaller m value results in a narrower normal fluctuation range. Since the indoor distribution system is usually in a relatively stable state, m can be taken as 1.3.

[0049] Specifically, the specific steps for determining the training data set based on the user terminal data covered by the first abnormal passive indoor distribution system are as follows: Obtain the user terminal data covered by the first abnormal passive indoor distribution system, and the user terminal data includes at least one of the following: the received power of the reference signal received by the user terminal, the physical resource block utilization rate, and the number of user terminals; Associate the user terminal data covered by the first abnormal passive indoor distribution system with the standing wave ratio data of the first abnormal passive indoor distribution system to obtain processed data; Convert the processed data into a multi-dimensional tensor format to obtain the training data set, where the multi-dimensional tensor includes at least: the number of samples in the training data set, the number of standing wave ratios in the sample, and the label corresponding to the sample.

[0050] Specifically, the multi-dimensional tensor can be expressed as [batch_size, sequence_length, feature], where batch_size represents the number of samples in the training data set, sequence_length represents the number of standing wave ratios in the sample, and feature represents the label corresponding to the sample.

[0051] Taking the standing wave ratio as an example, the data structure of this system is as follows:

[0052] [3.32, 3.23, ……, 3.24, 3.80][3];

[0053] [1.49, 1.44, ……, 1.41, 1.52][1];

[0054] [1.50,1.51,……,1.53,1.54][2].

[0055] It can be understood that [3.32, 3.23, ..., 3.24, 3.80] represent the standing wave ratio values ​​at consecutive sampling points in a specific time period, and [1], [2], and [3] represent the distributed indoor system in different states, for example: Normal state: indicates that the operating state of the distributed indoor system in this time series is considered normal, and no obvious faults or obstacles are detected. Explicit abnormal state: indicates that there are obvious faults in the distributed indoor system, which may have triggered alarms or can be easily identified through conventional monitoring methods. Hidden abnormal state: indicates that although there are no obvious fault alarms in the distributed indoor system, the fluctuation or decline in signal transmission quality may indicate potential obstacles or problems in the system. This state often requires more advanced analysis methods to identify. Among them, label "1" represents the normal state, label "3" represents the explicit abnormal state, and label "2" represents the hidden abnormal state.

[0056] In some embodiments of the present application, the specific process of training a recognition model based on the training data set is as follows: obtaining an initial model, which is a long short-term memory model; dividing the training data set into a training set, a validation set, and a test set; using the training set to train the initial model, and when the loss function value of the initial model on the training set and the validation set reaches a first threshold and the accuracy of the initial model on the training set and the validation set reaches a second threshold, determining that the initial model converges to obtain the recognition model.

[0057] like Figure 3 As shown, a long short-term memory network (LSTM) model is constructed and its network architecture is designed, which includes the input layer, LSTM layer, and output layer in sequence, and may also include a fully connected layer. Define the number of neurons, activation function, loss function, and optimizer (such as using cross entropy loss and Adam optimizer) in each layer of the model. In the LSTM layer, the data is calculated in the order of equations (1-3) to (1-8):

[0058] f t =σ(W f ·[h t-1 ,x t ]+b f ) Formula (1-3)

[0059] i t =σ(W i ·[h t-1 ,x t ]+b i ) Formula (1-4)

[0060]

[0061] o t = σ(W o ·[h t-1 , x t + b o ) Equation (1-7)

[0062] h t = o t * tanh(C t ) Equation (1-8)

[0063] Among them, Equation (1-3) is the calculation process of the forget gate. h t-1 is the hidden layer state at the previous sequence moment, x t is the data at the current input sequence moment, f t is the output value, σ is the sigmoid activation function, W f is the weight matrix, b f is the bias vector. The forget gate determines how much of the cell state at the previous moment needs to be retained; Equations (1-4) to (1-6) are the calculation process of the input gate, where i t is the output of the memory gate, is the temporary cell state, C t is the cell state at the current moment, tanh is the activation function, and the input gate determines which new information needs to be added to the memory cell; Equations (1-7) to (1-8) are the calculation process of the output gate, where o t is the value of the output gate, h t is the output value, and the output gate determines which information needs to be output.

[0064] Specifically, the training set data is used to train the model, and the model parameters are adjusted through the backpropagation algorithm to minimize the loss function. During the training process, the validation set data is used to monitor the model performance, and hyperparameters such as the learning rate are adjusted to avoid overfitting. Monitor the changes in the loss value and accuracy during the training process to ensure that the model converges and reaches the expected performance. Use the test set data to finally evaluate the trained model and calculate its prediction accuracy. Evaluate the performance of the model on unseen data to ensure that it can be generalized to the monitoring of in-building distribution systems in the real world. The loss function J is shown as follows:

[0065]

[0066] In the formula, N is the number of samples, K is the number of label categories, y ic is the target value of the sample, h θ (x i ) c is the observed sample x iThe predicted probability value belonging to category c, where i represents the sample number.

[0067] The accuracy is determined as follows: respectively obtain the numbers of true positives, false positives, true negatives, and false negatives output by the initial model; determine the accuracy of the initial model according to the numbers of true positives, false positives, true negatives, and false negatives output by the initial model.

[0068] Specifically, it is shown as the following formula:

[0069] Accuracy = (TP + TN) / (TP + TN + FP + FN) Formula (1 - 10);

[0070] In the formula, TP represents true positives, FP represents false positives, TN represents true negatives, FN represents false negatives, and Accuracy represents the accuracy.

[0071] In order to further improve the performance of the recognition model, the user terminal data under the coverage of the passive distributed antenna system with obvious anomalies can be added to the training data set so that the model can recognize more types of abnormal states. The specific steps are as follows: obtain the alarm data of multiple passive distributed antenna systems, determine the second abnormal passive distributed antenna system among the multiple passive distributed antenna systems according to the alarm data; add the user terminal data under the coverage of the second abnormal passive distributed antenna system to the training data set.

[0072] It can be understood that the second abnormal passive distributed antenna system is an obvious fault situation that can be directly queried according to the alarm information.

[0073] After obtaining the recognition model, the user terminal data under the coverage of the distributed antenna system to be monitored can be recognized by using the recognition model. The specific steps are as follows: obtain the user terminal data under the coverage of the distributed antenna system to be monitored according to a preset period; analyze the user terminal data under the coverage of the distributed antenna system to be monitored obtained by using the recognition model to determine whether the distributed antenna system to be monitored is abnormal.

[0074] Figure 4 A training device for a passive distributed antenna system monitoring model according to an embodiment of the present application, the device includes:

[0075] An acquisition module 40, configured to acquire the standing wave ratio data of multiple passive distributed antenna systems;

[0076] A selection module 42, configured to determine the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems according to the standing wave ratio data of the multiple passive distributed antenna systems, and determine the first abnormal passive distributed antenna system from the standing wave ratio data of the multiple passive distributed antenna systems according to the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems;

[0077] A determination module 44, configured to determine a training data set according to user terminal data under the coverage of the first abnormal passive distributed antenna system;

[0078] An identification module 46, configured to train an identification model based on the training data set, identify user terminal data under the coverage of the distributed antenna system to be monitored, and determine whether the distributed antenna system to be monitored is abnormal.

[0079] Through the above passive distributed antenna system monitoring model training device, the standing wave ratio data of multiple passive distributed antenna systems is collected; the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems is determined according to the standing wave ratio data of the multiple passive distributed antenna systems, and a first abnormal passive distributed antenna system is determined from the standing wave ratio data of the multiple passive distributed antenna systems according to the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems; a training data set is determined according to user terminal data under the coverage of the first abnormal passive distributed antenna system; an identification model is trained based on the training data set, and the identification model is used to identify user terminal data under the coverage of the distributed antenna system to be monitored to determine whether the distributed antenna system to be monitored is abnormal, thereby achieving the purpose of determining abnormal distributed antenna system data according to the fluctuation condition of the standing wave ratio, and training an abnormal distributed antenna system identification model by using the determined abnormal distributed antenna system data, thereby realizing the technical effect of improving the fault prediction accuracy of the passive distributed antenna system, and further solving the technical problem of low fault prediction accuracy of the passive distributed antenna system in the related art.

[0080] The selection module 42 includes: a determination sub-module, configured to determine the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems according to the standing wave ratio data of the multiple passive distributed antenna systems, including: determining the moving average of the standing wave ratio of the multiple passive distributed antenna systems according to the standing wave ratio data of the multiple passive distributed antenna systems; obtaining the standard deviation of the moving average, and determining the weighted standard deviation according to a preset weight and the standard deviation; determining a standing wave ratio fluctuation upper limit and a standing wave ratio fluctuation lower limit according to the moving average and the weighted standard deviation respectively; determining the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems according to the comparison result between the standing wave ratio of the multiple passive distributed antenna systems and the standing wave ratio fluctuation upper limit and the standing wave ratio fluctuation lower limit.

[0081] The determination sub-module includes a determination unit for determining a training data set according to the user terminal data under the coverage of the first abnormal passive distributed antenna system, including: obtaining the user terminal data under the coverage of the first abnormal passive distributed antenna system, where the user terminal data includes at least one of the following: the received power of the reference signal received by the user terminal, the physical resource block utilization rate, and the number of user terminals; associating the user terminal data under the coverage of the first abnormal passive distributed antenna system with the standing wave ratio data of the first abnormal passive distributed antenna system to obtain processed data; converting the processed data into a multi-dimensional tensor format to obtain the training data set, where the multi-dimensional tensor includes at least: the number of samples in the training data set, the number of standing wave ratios in the samples, and the labels corresponding to the samples.

[0082] The recognition module 46 includes a training sub-module for training an identification model based on the training data set, including: obtaining an initial model, where the initial model is a long short-term memory model; dividing the training data set into a training set, a validation set, and a test set; training the initial model using the training set, and determining that the initial model converges to obtain the recognition model when the loss function values of the initial model on the training set and the validation set reach a first threshold and the accuracy rate of the initial model on the training set and the validation set reaches a second threshold.

[0083] The training sub-module includes a verification unit for respectively obtaining the numbers of true positives, false positives, true negatives, and false negatives output by the initial model; determining the accuracy rate of the initial model according to the numbers of true positives, false positives, true negatives, and false negatives output by the initial model.

[0084] The recognition module 46 further includes a data set sub-module for obtaining the alarm data of multiple passive distributed antenna systems, determining the second abnormal passive distributed antenna system among the multiple passive distributed antenna systems according to the alarm data; and adding the user terminal data under the coverage of the second abnormal passive distributed antenna system to the training data set.

[0085] The recognition module 46 further includes an identification sub-module for identifying the user terminal data under the coverage of the to-be-monitored distributed antenna system, including: obtaining the user terminal data under the coverage of the to-be-monitored distributed antenna system at a preset period; analyzing the user terminal data under the coverage of the to-be-monitored distributed antenna system obtained by using the recognition model to determine whether the to-be-monitored distributed antenna system is abnormal.

[0086] It should be noted that Figure 4 The shown training device for the passive distributed antenna system monitoring model is used to execute Figure 2The method for training the monitoring model of the passive in-building distribution system shown, so the relevant explanations in the above method for training the monitoring model of the passive in-building distribution system also apply to this device for training the monitoring model of the passive in-building distribution system, and will not be repeated here.

[0087] An embodiment of the present application further provides a computer device, including: a memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the above method for training the monitoring model of the passive in-building distribution system.

[0088] The method executed by the above computer device includes collecting the standing wave ratio data of multiple passive in-building distribution systems; determining the fluctuation conditions of the standing wave ratios of the multiple passive in-building distribution systems according to the standing wave ratio data of the multiple passive in-building distribution systems, and determining the first abnormal passive in-building distribution system from the standing wave ratio data of the multiple passive in-building distribution systems according to the fluctuation conditions of the standing wave ratios of the multiple passive in-building distribution systems; determining a training data set according to the user terminal data covered by the first abnormal passive in-building distribution system; training an identification model based on the training data set, and the identification model is used to identify the user terminal data covered by the in-building distribution system to be monitored to determine whether the in-building distribution system to be monitored is abnormal, so as to achieve the purpose of determining the abnormal in-building distribution system data according to the fluctuation conditions of the standing wave ratio and training an abnormal in-building distribution system identification model by using the determined abnormal in-building distribution system data, thus realizing the technical effect of improving the fault prediction accuracy of the passive in-building distribution system, and further solving the technical problem of low fault prediction accuracy of the passive in-building distribution system in the related art.

[0089] An embodiment of the present application further provides a non-volatile storage medium, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the above method for training the monitoring model of the passive in-building distribution system by running the computer program.

[0090] The method stored in the above non-volatile storage medium adopts collecting the standing wave ratio data of multiple passive distributed antenna systems; determining the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems according to the standing wave ratio data of the multiple passive distributed antenna systems, and determining a first abnormal passive distributed antenna system from the standing wave ratio data of the multiple passive distributed antenna systems according to the fluctuation condition of the standing wave ratio of the multiple passive distributed antenna systems; determining a training data set according to the user terminal data covered by the first abnormal passive distributed antenna system; training an identification model based on the training data set, where the identification model is used to identify the user terminal data covered by the to-be-monitored distributed antenna system to determine whether the to-be-monitored distributed antenna system is abnormal, so as to achieve the purpose of determining the abnormal distributed antenna system data according to the fluctuation condition of the standing wave ratio and training an abnormal distributed antenna system identification model by using the determined abnormal distributed antenna system data, thereby realizing the technical effect of improving the fault prediction accuracy of the passive distributed antenna system, and further solving the technical problem of the low fault prediction accuracy of the passive distributed antenna system in the related art.

[0091] The embodiment of the present application also provides a computer program product, including computer instructions, which implement the steps of the method for training the monitoring model of the passive distributed antenna system in the present application when executed by a processor.

[0092] The serial numbers of the above embodiments of the present application are only for description and do not represent the advantages or disadvantages of the embodiments.

[0093] In the above embodiments of the present application, the descriptions of each embodiment have their own emphases. For the parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0094] In several embodiments provided by the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units can be a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces, and the indirect coupling or communication connection of units or modules can be in an electrical or other form.

[0095] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0096] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0097] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0098] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. A passive indoor distributed system monitoring model training method, characterized in that: include: Collect standing wave ratio data of multiple passive indoor distributed systems; Determining fluctuations of the standing wave ratios of the plurality of passive indoor distributed systems according to the standing wave ratio data of the plurality of passive indoor distributed systems, and determining a first abnormal passive indoor distributed system from the standing wave ratio data of the plurality of passive indoor distributed systems according to the fluctuations of the standing wave ratios of the plurality of passive indoor distributed systems; Determine a training data set based on user terminal data covered by the first abnormal passive indoor distributed system; A recognition model is obtained by training based on the training data set, and the recognition model is used to recognize user terminal data covered by the indoor distributed system to be monitored to determine whether the indoor distributed system to be monitored is abnormal.

2. The method according to claim 1, characterized in that Determining fluctuations of the standing wave ratios of the plurality of passive indoor distributed systems according to the standing wave ratio data of the plurality of passive indoor distributed systems comprises: Determine a moving average of the standing wave ratios of the plurality of passive indoor distributed systems according to the standing wave ratio data of the plurality of passive indoor distributed systems; Obtaining the standard deviation of the moving average, and determining a weighted standard deviation according to a preset weight and the standard deviation; Determine the standing wave ratio fluctuation upper limit and the standing wave ratio fluctuation lower limit respectively according to the moving average and the weighted standard deviation; The fluctuation of the standing wave ratio of the multiple passive indoor distributed systems is determined according to the comparison result of the standing wave ratio of the multiple passive indoor distributed systems with the standing wave ratio fluctuation upper limit and the standing wave ratio fluctuation lower limit.

3. The method according to claim 2, characterized in that Determining a training data set according to user terminal data covered by the first abnormal passive indoor distributed system includes: Acquire user terminal data under the coverage of the first abnormal passive indoor distributed system, wherein the user terminal data includes at least one of the following: received power of a reference signal received by the user terminal, utilization rate of a physical resource block, and number of user terminals; Associating the user terminal data covered by the first abnormal passive indoor distributed system with the standing wave ratio data of the first abnormal passive indoor distributed system to obtain processed data; The processed data is converted into a multidimensional tensor format to obtain the training data set, wherein the multidimensional tensor includes at least: the number of samples in the training data set, the number of standing wave ratios in the samples, and labels corresponding to the samples.

4. The method according to claim 1, characterized in that: The recognition model is obtained by training based on the training data set, including: Acquire an initial model, wherein the initial model is a long short-term memory model; Dividing the training data set into a training set, a validation set and a test set; The initial model is trained using the training set. When the loss function value of the initial model on the training set and the validation set reaches a first threshold and the accuracy of the initial model on the training set and the validation set reaches a second threshold, it is determined that the initial model converges to obtain the recognition model.

5. The method according to claim 4, characterized in that The method further comprises: Respectively obtain the number of true positive examples, false positive examples, true negative examples, and false negative examples output by the initial model; The accuracy of the initial model is determined according to the number of true positive examples, false positive examples, true negative examples, and false negative examples output by the initial model.

6. The method according to claim 1, characterized in that The method further comprises: Acquire alarm data of a plurality of passive indoor distributed systems, and determine a second abnormal passive indoor distributed system among the plurality of passive indoor distributed systems according to the alarm data; The user terminal data under the coverage of the second abnormal passive indoor distributed system is added to the training data set.

7. The method according to claim 1, characterized in that Identify the user terminal data under the coverage of the indoor distributed system to be monitored, including: Acquire user terminal data under the coverage of the indoor distributed system to be monitored according to a preset period; The recognition model is used to analyze the acquired user terminal data under the coverage of the indoor distributed system to be monitored to determine whether the indoor distributed system to be monitored is abnormal.

8. A passive room distribution system monitoring model training device, characterized in that: include: An acquisition module is used to collect standing wave ratio data of multiple passive indoor distributed systems; A selection module, used for determining the fluctuation of the standing wave ratio of the plurality of passive indoor distributed systems according to the standing wave ratio data of the plurality of passive indoor distributed systems, and determining a first abnormal passive indoor distributed system from the standing wave ratio data of the plurality of passive indoor distributed systems according to the fluctuation of the standing wave ratio of the plurality of passive indoor distributed systems; A determination module, used to determine a training data set according to user terminal data covered by the first abnormal passive indoor distributed system; The identification module is used to obtain an identification model based on the training data set, identify the user terminal data covered by the indoor distributed system to be monitored, and determine whether the indoor distributed system to be monitored is abnormal.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the passive indoor distributed system monitoring model training method described in any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by the processor, the passive indoor distributed system monitoring model training method described in any one of claims 1 to 7 is implemented.

Images