CN112836843A - Method and device for predicting base station out-of-service alarm - Google Patents
Method and device for predicting base station out-of-service alarm Download PDFInfo
- Publication number
- CN112836843A CN112836843A CN201911168434.1A CN201911168434A CN112836843A CN 112836843 A CN112836843 A CN 112836843A CN 201911168434 A CN201911168434 A CN 201911168434A CN 112836843 A CN112836843 A CN 112836843A
- Authority
- CN
- China
- Prior art keywords
- base station
- data
- weather information
- service
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000013528 artificial neural network Methods 0.000 claims abstract description 50
- 230000000306 recurrent effect Effects 0.000 claims abstract description 45
- 238000012549 training Methods 0.000 claims description 40
- 238000012544 monitoring process Methods 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims description 9
- 230000015654 memory Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 5
- 238000012423 maintenance Methods 0.000 abstract description 17
- 238000007726 management method Methods 0.000 description 10
- 239000011159 matrix material Substances 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 238000004422 calculation algorithm Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 125000004122 cyclic group Chemical group 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000008034 disappearance Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 238000002360 preparation method Methods 0.000 description 3
- 238000013024 troubleshooting Methods 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011056 performance test Methods 0.000 description 2
- 230000002441 reversible effect Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000002759 z-score normalization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The embodiment of the invention provides a method and a device for predicting a base station out-of-service alarm, wherein the method comprises the following steps: acquiring characteristic data of a base station and weather information data of an area where the base station is located; inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network. The device performs the above method. The method and the device for predicting the service quitting alarm of the base station, provided by the embodiment of the invention, can be used for predicting the service quitting alarm of the base station in advance, accurately and efficiently, so that the efficiency and the quality of network operation and maintenance are improved.
Description
Technical Field
The invention relates to the technical field of network operation and maintenance, in particular to a method and a device for predicting base station out-of-service alarm.
Background
The network operation and maintenance troubleshooting is the main work content of the network operation and maintenance, wherein the most core work is to troubleshoot the base station out-of-service alarm.
The network operation and maintenance troubleshooting technical means adopted at present mainly comprises a passive coping scheme adopted after a base station gives an alarm of quitting service, and root cause analysis and processing are further carried out by means of manual experience of operation and maintenance personnel and an equipment manual. Due to the technical means, after the base station gives an alarm for quitting service and stops service, the user cannot normally access the network within a period of time. Therefore, it is necessary to predict the base station out-of-service alarm in advance to prevent the network from getting ill in the future and improve the efficiency and quality of network operation and maintenance.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides a method and a device for predicting a base station out-of-service alarm.
The embodiment of the invention provides a method for predicting a base station out-of-service alarm, which comprises the following steps:
acquiring characteristic data of a base station and weather information data of an area where the base station is located;
inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
Wherein, the training of the recurrent neural network by adopting the sample data for extracting the characteristic data of the base station and the weather information data comprises the following steps:
training a recurrent neural network by adopting sample data of extracted site parameter resource data and the weather information data; the station parameter resource data is parameter configuration state data and cell identification data which can represent the intrinsic property of the base station.
Wherein, the training of the recurrent neural network by adopting the sample data for extracting the characteristic data of the base station and the weather information data comprises the following steps:
training a recurrent neural network by adopting sample data for extracting network management performance statistical data and the weather information data; the network management performance statistical data is key performance index data of the base station for wireless communication service.
Wherein, the training of the recurrent neural network by adopting the sample data for extracting the characteristic data of the base station and the weather information data comprises the following steps:
training a recurrent neural network by adopting sample data extracted from the dynamic environment monitoring data and the weather information data; the power environment monitoring data are power monitoring data and environment monitoring data in a base station machine room.
Wherein, the training of the recurrent neural network by adopting the sample data for extracting the characteristic data of the base station and the weather information data comprises the following steps:
training a recurrent neural network by adopting sample data for extracting historical alarm data and the weather information data; the historical alarm data is the type and frequency of the out-of-service alarm of the base station in a preset historical time period.
The key performance index data is at least one of a wireless connection rate, a wireless disconnection rate, a connection time delay and a block error rate.
Wherein the recurrent neural network is LSTM.
The embodiment of the invention provides a device for predicting a base station out-of-service alarm, which comprises:
the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring characteristic data of a base station and weather information data of an area where the base station is located;
the prediction unit is used for inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
An embodiment of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein,
the processor, when executing the computer program, implements the method steps of:
acquiring characteristic data of a base station and weather information data of an area where the base station is located;
inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
An embodiment of the invention provides a non-transitory computer readable storage medium having a computer program stored thereon, which when executed by a processor implements the following method steps:
acquiring characteristic data of a base station and weather information data of an area where the base station is located;
inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
According to the method and the device for predicting the service quitting alarm of the base station, provided by the embodiment of the invention, the characteristic data of the base station and the weather information data of the area where the base station is located are input to the preset base station service quitting alarm prediction model, and the output result of the model is used as the prediction result of the service quitting alarm of the base station, so that the service quitting alarm of the base station can be predicted accurately and efficiently in advance, and the efficiency and the quality of network operation and maintenance are further improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a flowchart of a method for predicting a station fallback alarm according to an embodiment of the present invention;
FIG. 2 is a flowchart of a method for predicting a station fallback alarm according to another embodiment of the present invention;
FIG. 3 is a diagram of a single sample of a training set according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of the basic elements of an LSTM according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of an embodiment of a base station fallback warning prediction apparatus according to the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a flowchart of an embodiment of a method for predicting a base station fallback alarm, and as shown in fig. 1, the method for predicting the base station fallback alarm provided by the embodiment of the present invention includes the following steps:
s101: and acquiring the characteristic data of the base station and the weather information data of the area where the base station is located.
Specifically, the characteristic data of the base station and the weather information data of the area where the base station is located are obtained. The method steps can be executed by computer equipment, and can be embodied as a network management platform. Because the preset base station out-of-service alarm prediction model in the embodiment of the invention is based on the recurrent neural network, and the recurrent neural network has the capabilities of data dimension reduction and feature learning in the training process, in the feature extraction stage, feature data corresponding to all reasons causing the base station out-of-service or abnormal feature data generated before the out-of-service occurs should be considered as much as possible. The method can specifically comprise the following steps: site engineering parameter resource data, network management performance statistical data, dynamic environment monitoring data, historical alarm data and weather information data. Are specifically described as follows:
1) site engineering parameter resource data
The data is parameter configuration state data and cell identification data which can represent the intrinsic property of the base station. For example: the method comprises the following steps of opening time, geographic position, manufacturer, frequency point number, Physical Cell Identifier (PCI for short), coverage scene, antenna hanging height, direction angle downward inclination angle and the like. These data characterize the intrinsic properties of the base station, such as its geographical location, coverage scenario, and the likelihood that the base station will be subjected to extreme weather.
2) Network management performance statistics
The data is Key Performance index data of a base station for wireless communication service, and may include Key Performance Indicator (KPI) and Performance test (PM) data of a 15-minute granularity wireless network, which are extracted from a network management platform. Before a major failure occurs, the performance of the base station may fluctuate or deteriorate due to influences such as aged deterioration of equipment, bad weather, and the like. By capturing changes in these performance data, it can be used to estimate the occurrence of a fault. The key performance indicator data may further include: and the performance indexes such as wireless connection rate, wireless disconnection rate, connection delay, block error rate and the like.
3) Dynamic environment monitoring data
The data are power monitoring data and environment monitoring data in a base station machine room. Part of the base station failure is due to the base station powering off and the battery draining, or due to extreme weather causing equipment failure. The power monitoring is to monitor the voltage and the current of a power supply and the residual electric quantity of a storage battery in real time and is directly related to whether a base station has power failure risk or not; the environmental monitoring detects the temperature, humidity and other information in the machine room, and has a pilot prediction effect on predicting the short circuit and burning of the plate circuit.
4) Historical alarm data
The data is the type and frequency of the out-of-service alarm of the base station in a preset historical time period, and the specific numerical value of the preset historical time period can be set independently according to the actual situation. And counting the category and frequency of other secondary alarms accompanied when the out-of-service alarm causing the fault occurs. Based on the standard alarm dictionary table, the categories of the statistical alarms may include: the system has no License operation alarm, the base station S1 control plane transmission interruption alarm, the monitoring equipment maintenance link abnormal alarm, etc.
5) Weather information data
This type of data describes the future weather conditions in the area of the base station. The temperature, humidity, rain and snow weather probability and some extreme weather early warning information (such as typhoon, slush and high temperature) of the city/county where the base station is located can be crawled by using the crawler for assisting fault prediction.
S102: inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
Specifically, the base station characteristic data and the weather information data are input to a preset base station out-of-service alarm prediction model, and an output result of the preset base station out-of-service alarm prediction model is used as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network. The data modeling using the deep neural network may use various models, such as a deep boltzmann network DBN, a convolutional neural network CNN, a cyclic neural network RNN, and the like.
The embodiment of the invention provides a prediction method aiming at the occurrence probability of the fallback alarm prediction of a base station. The prediction method is based on the input of a model by using multi-dimensional base station characteristic data (and combining the data) in a period of time of a base station, is used for predicting the probability of the occurrence of out-of-service alarm of the base station in a period of time in the future, and is a prediction method based on a time sequence.
The RNN is very suitable for predicting time series related data due to the memorability, is commonly used in the data prediction fields of temperature prediction, stock prediction and the like, and meets the requirements of the invention. Meanwhile, the naive RNN algorithm has the problem of gradient disappearance when the model scale is too large due to the long-term dependence problem. The long-short term memory network LSTM realizes long-range correlation of information by introducing a leakage unit and a gate control circulation unit, and avoids gradient disappearance.
Furthermore, the embodiment of the invention can select a variety of the recurrent neural network RNN, namely a long short-term memory network LSTM, to realize the prediction of the fallback alarm of the base station.
Fig. 2 is a flowchart of a method for predicting a fallback alarm of a base station according to another embodiment of the present invention, as shown in fig. 2, including a data preparation phase, a model training phase, and a model prediction phase. These three phases are specified below:
step 1) data preparation phase
In the stage, preparation data is generated for a preset base station out-of-service alarm prediction model, and a training sample set is constructed. The training sample set is formed by historical data of a plurality of base stations by taking the base stations as basic units. Fig. 3 is a schematic diagram of a single sample of a training set according to an embodiment of the present invention, and as shown in fig. 3, a base station continuously forms a sample with data for a period of time (assuming that input data includes a characteristic of T time, output data needs to include data of T +1 time, and it is guaranteed that a target to be predicted is in a future time of the input characteristic). Each sample includes both the feature data for the input and the actual values of the desired output, i.e., the training targets. The input data comprises site engineering parameter resource data, network management performance statistical data, power environment monitoring data, alarm data and weather information data of the base station, and the training target data is whether a service quitting alarm occurs in a period of time (T to T +1) in the future of the base station, wherein the occurrence is 1, and the non-occurrence is 0. And inputting characteristic data into the model to obtain the probability of giving out alarm in the future period of time, wherein the interval of the probability value is [0,1 ].
By mapping the input feature data with the output target data, a training sample pattern is composed as shown in fig. 3. Matrix X is the input matrix and matrix Y is the target matrix.
The X matrix comprises n rows which are respectively constructed by various data sourcesEstablishing n-dimensional characteristics; line T + 1, time sequence 1-time sequence T +1, where line T +1 may be empty, and fs,lThe value of the l-th characteristic at the s moment is obtained.
The Y matrix comprises m rows which are respectively target types to be predicted (in the embodiment of the invention, the type of the possible out-of-service alarm is mainly referred to); t +1 rows, time sequence 1 to time sequence T +1, respectively, since the embodiment of the present invention aims to predict the target at a future time based on the current information, the Y matrix needs to retain data of 2 to T +1 rows, i.e. delay 1 time unit, K, from the X matrixs,lIf the first out-of-service alarm occurs at the time s, the occurrence is 1, and the non-occurrence is 0.
Two operations are required after the training set is prepared. The first step is normalization, which is normalized by columns by using a Z-Score normalization method to ensure the stability of the data set distribution:
X:,j＝(f:,j-fj,mean)/fj,std
y:,j＝(k:,j-kj,mean)/kj,std
wherein f is:,jFor all elements of the j-th column before normalization, x:,jFor all elements of column j after normalization, fj,meanIs the mean, f, of all the element values of column j of the entire sample setj,stdThe standard deviation of all elements in column j for the entire sample set. Normalized by Z-Score such that the entire sample set is normalized to a standard normal distribution with a mean of 0 and a variance of 1. Similar operations on the Y matrix also achieve normalization of the entire sample set.
The second step is to fill in the defect data. For cells without data, filling the cells by using a standard normal distribution random generation numerical value, and the statistical property of each column of data cannot be changed.
Step 2) model training phase
Training LSTM network weight by using the data set constructed in the step 1) to realize mapping from cell fault alarm related characteristic index input to out-of-service alarm probability output. The LSTM model is formed by cascading a plurality of LSTM units, is a prediction model based on time steps, and can input corresponding characteristics at each time step. The data set established based on the service logic implies the symptoms of the occurrence of faults such as the out-of-service alarm, and the like, and the characteristic information of each time step has different degrees of influence on the probability of the occurrence of the out-of-service alarm, and the influence degree is complex and cannot be determined manually. The LSTM algorithm can automatically learn the influence degree information implied in the data by changing the weight transmitted by the unit, thereby realizing the modeling of the data. Fig. 4 is a schematic diagram of basic units of an LSTM according to an embodiment of the present invention, as shown in fig. 4:
the LSTM network is formed by a plurality of identical basic unit tail-ends, each unit corresponding to a time step. Wherein, XtFeature vectors, h, input for t time stepstFor output of t time steps, CtIs the state at t time step. f. oft,it,otRespectively corresponding to the forgetting gate, the input gate and the output gate. σ (-) is a Sigmoid activation function.
The training of the model is divided into two stages of forward propagation and reverse training. And in the forward propagation stage, the characteristic vectors are input according to the time step sequence, and the unit state and the output of each unit step are calculated according to the following formula, wherein W corresponds to the network weight, and b corresponds to the network bias. Firstly, to the forgetting door ftAnd an input gate itAnd (6) performing calculation. The forgetting gate determines the influence of the last time step state on the next time step state; the input gate determines the effect of the input on the current time step on the current state. The values of the two are 0-1, the 0 mark is completely abandoned, and the 1 mark is completely reserved. The formula is as follows:
ft＝σ(Wf·[ht-1,xt]+bf)
it＝σ(Wi·[ht-1,xt]+bi)
second, for the content of the input gate input, an alternative input vector needs to be constructed
And use
ft、it、Ct-1To construct the current time step state CtThe formula is as follows:
then, the output gate output is calculated per unit step. Output gate otAnd determining the quantity of the output information of the current time step, wherein the value is 0-1, the 0 mark is completely abandoned, and the 1 mark represents the complete reservation. The calculation formula is as follows:
Ot＝σ(Wo·[ht-1,xt]+bo)
ht＝ot*tanh(Ct)
finally, an output prediction value of the current time sequence T is calculated, namely the probability of the occurrence of the out-of-service alarm at the moment of T +1 is estimated in the embodiment of the invention, and the calculation formula is as follows:
in the forward propagation process, base station characteristics based on time step are mapped into alarm occurrence probability through an LSTM network; and (3) carrying out a reverse training process, changing the weights W, b and the like of the network through training, and enabling the predicted probability of the network to be closer to the true probability. The training process is implemented by a time-based back propagation algorithm (BPTT).
The objective of the BPTT algorithm is to minimize the loss function of the model. The loss function uses a cross-entropy function to represent the difference between the predicted and true values, as follows:
the gradient of the loss function with respect to each weight parameter can be calculated using the partial derivative chain rule, taking Wf as an example:
for other weights, bias parameters Wi, Wo, bf, bi, bo, V, c, etc., the gradients may be similarly solved. All the weights and the bias parameters can be trained and updated by using a random gradient descent method.
Step 3) model prediction phase
After multiple times of iterative training, the model is converged and can be used for out-of-service alarm prediction of base station faults. During prediction, performance data, historical alarm data, work parameter data, dynamic ring data and local weather data of a current base station in a past period (less than T) are input, input characteristics of each time step are formed according to a time sequence, and a model is input, so that a failure probability prediction value at the moment of T +1 can be obtained.
The embodiment of the invention obtains the service requirement for predicting the out-of-service alarm probability based on the deep research on the operation and maintenance troubleshooting of the wireless network. According to the characteristics of the demand, a deep learning algorithm LSTM based on time series prediction is selected as a business modeling algorithm, and a fallback alarm probability prediction model is constructed. The embodiment of the invention considers all the details involved in scenes such as internal and external reasons causing the wireless base station to generate the out-of-service alarm, performance degradation condition representing out-of-service occurrence and the like as much as possible, and realizes accurate mapping from input characteristic data to output probability through training of model weight. Compared with the traditional method for regional quit alarm prediction by experience, the embodiment of the invention realizes the refined quit probability prediction of the cell level. The method has guiding significance for releasing the operation and maintenance resources of the wireless base station and selecting the routing inspection base station. Meanwhile, by preventing the trouble in the bud and utilizing a prediction mechanism, the system is overhauled in advance before the fault occurs, so that the frequency of actual quit of service can be reduced, and the perception of the client is smoother.
According to the method for predicting the service quitting alarm of the base station, provided by the embodiment of the invention, the characteristic data of the base station and the weather information data of the area where the base station is located are input to the preset base station service quitting alarm prediction model, and the output result of the model is used as the prediction result of the service quitting alarm of the base station, so that the service quitting alarm of the base station can be predicted accurately and efficiently in advance, and the efficiency and the quality of network operation and maintenance are further improved.
On the basis of the above embodiment, the training of the recurrent neural network by using the sample data obtained by extracting the feature data of the base station and the weather information data includes:
specifically, a cyclic neural network is trained by adopting sample data of extracted site work parameter resource data and the weather information data; the station parameter resource data is parameter configuration state data and cell identification data which can represent the intrinsic property of the base station. Reference is made to the above description and no further description is made.
According to the method for predicting the base station out-of-service alarm provided by the embodiment of the invention, the recurrent neural network is trained by adopting the sample data of the extracted site engineering parameter resource data and the extracted weather information data, so that the base station out-of-service alarm can be predicted in advance, accurately and efficiently, and the efficiency and quality of network operation and maintenance are further improved.
On the basis of the above embodiment, the training of the recurrent neural network by using the sample data obtained by extracting the feature data of the base station and the weather information data includes:
specifically, a recurrent neural network is trained by adopting sample data for extracting network management performance statistical data and the weather information data; the network management performance statistical data is key performance index data of the base station for wireless communication service. Reference is made to the above description and no further description is made.
According to the method for predicting the base station out-of-service alarm provided by the embodiment of the invention, the cyclic neural network is trained by adopting the sample data extracting the network management performance statistical data and the weather information data, so that the prediction of the base station out-of-service alarm can be further performed in advance, accurately and efficiently, and the efficiency and quality of network operation and maintenance are further improved.
On the basis of the above embodiment, the training of the recurrent neural network by using the sample data obtained by extracting the feature data of the base station and the weather information data includes:
specifically, a recurrent neural network is trained by adopting sample data of extracted dynamic environment monitoring data and weather information data; the power environment monitoring data are power monitoring data and environment monitoring data in a base station machine room. Reference is made to the above description and no further description is made.
According to the method for predicting the base station out-of-service alarm provided by the embodiment of the invention, the recurrent neural network is trained by adopting the sample data for extracting the dynamic environment monitoring data and the weather information data, so that the base station out-of-service alarm can be predicted in advance, accurately and efficiently, and the efficiency and quality of network operation and maintenance are further improved.
On the basis of the above embodiment, the training of the recurrent neural network by using the sample data obtained by extracting the feature data of the base station and the weather information data includes:
specifically, a recurrent neural network is trained by adopting sample data for extracting historical alarm data and the weather information data; the historical alarm data is the type and frequency of the out-of-service alarm of the base station in a preset historical time period. Reference is made to the above description and no further description is made.
According to the method for predicting the service quitting alarm of the base station, provided by the embodiment of the invention, the cyclic neural network is trained by adopting the sample data for extracting the historical alarm data and the weather information data, so that the service quitting alarm of the base station can be predicted accurately and efficiently in advance, and the efficiency and quality of network operation and maintenance are further improved.
On the basis of the foregoing embodiment, specifically, the key performance indicator data is at least one of a wireless connection rate, a wireless disconnection rate, a connection delay, and a block error rate. Reference is made to the above description and no further description is made.
According to the method for predicting the base station out-of-service alarm provided by the embodiment of the invention, the key performance index data is determined to be at least one of the wireless connection rate, the wireless disconnection rate, the connection delay and the block error rate, so that the key performance index can be comprehensively obtained, the characteristics influencing the base station out-of-service alarm are further comprehensively considered, and the prediction of the base station out-of-service alarm is optimized.
On the basis of the above embodiment, specifically, the recurrent neural network is LSTM. Reference is made to the above description and no further description is made.
The base station out-of-service alarm prediction method provided by the embodiment of the invention can avoid gradient disappearance and optimize the training process of the model.
Fig. 5 is a schematic structural diagram of an embodiment of the device for predicting a base station fallback alarm, and as shown in fig. 5, an embodiment of the present invention provides a device for predicting a base station fallback alarm, which includes an obtaining unit 501 and a predicting unit 502, where:
the obtaining unit 501 is configured to obtain characteristic data of a base station and weather information data of an area where the base station is located; the prediction unit 502 is configured to input the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and use an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
Specifically, the obtaining unit 501 is configured to obtain characteristic data of a base station and weather information data of an area where the base station is located; the prediction unit 502 is configured to input the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and use an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
According to the base station out-of-service alarm prediction device provided by the embodiment of the invention, the base station out-of-service alarm prediction model is input by the base station characteristic data and the weather information data of the area where the base station is located, and the output result of the model is used as the base station out-of-service alarm prediction result, so that the base station out-of-service alarm prediction can be carried out in advance, accurately and efficiently, and the efficiency and quality of network operation and maintenance are further improved.
The base station fallback alarm prediction device provided by the embodiment of the present invention may be specifically configured to execute the processing flows of the above method embodiments, and the functions of the device are not described herein again, and refer to the detailed description of the above method embodiments.
Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 6, the electronic device includes: a processor (processor)601, a memory (memory)602, and a bus 603;
the processor 601 and the memory 602 complete mutual communication through a bus 603;
the processor 601 is configured to call program instructions in the memory 602 to perform the methods provided by the above-mentioned method embodiments, for example, including: acquiring characteristic data of a base station and weather information data of an area where the base station is located; inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
The present embodiment discloses a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the method provided by the above-mentioned method embodiments, for example, comprising: acquiring characteristic data of a base station and weather information data of an area where the base station is located; inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
The present embodiments provide a non-transitory computer-readable storage medium storing computer instructions that cause the computer to perform the methods provided by the above method embodiments, for example, including: acquiring characteristic data of a base station and weather information data of an area where the base station is located; inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for predicting a base station out-of-service alarm is characterized by comprising the following steps:
acquiring characteristic data of a base station and weather information data of an area where the base station is located;
inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model, and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
2. The method according to claim 1, wherein the training of the recurrent neural network with the sample data for extracting the base station characteristic data and the weather information data comprises:
training a recurrent neural network by adopting sample data of extracted site parameter resource data and the weather information data; the station parameter resource data is parameter configuration state data and cell identification data which can represent the intrinsic property of the base station.
3. The method according to claim 1, wherein the training of the recurrent neural network with the sample data for extracting the base station characteristic data and the weather information data comprises:
training a recurrent neural network by adopting sample data for extracting network management performance statistical data and the weather information data; the network management performance statistical data is key performance index data of the base station for wireless communication service.
4. The method according to claim 1, wherein the training of the recurrent neural network with the sample data for extracting the base station characteristic data and the weather information data comprises:
training a recurrent neural network by adopting sample data extracted from the dynamic environment monitoring data and the weather information data; the power environment monitoring data are power monitoring data and environment monitoring data in a base station machine room.
5. The method according to claim 1, wherein the training of the recurrent neural network with the sample data for extracting the base station characteristic data and the weather information data comprises:
training a recurrent neural network by adopting sample data for extracting historical alarm data and the weather information data; the historical alarm data is the type and frequency of the out-of-service alarm of the base station in a preset historical time period.
6. The method according to claim 3, wherein the key performance indicator data is at least one of a radio connection rate, a radio disconnection rate, a connection delay, and a block error rate.
7. The method of predicting base station fallback alarms according to any one of claims 1 to 6, wherein the recurrent neural network is an LSTM.
8. A base station fallback alarm prediction device, comprising:
the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring characteristic data of a base station and weather information data of an area where the base station is located;
the prediction unit is used for inputting the base station characteristic data and the weather information data to a preset base station out-of-service alarm prediction model and taking an output result of the preset base station out-of-service alarm prediction model as a base station out-of-service alarm prediction result; and the preset base station out-of-service alarm prediction model is obtained by adopting sample data extracted from the base station characteristic data and the weather information data to train a recurrent neural network.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method according to any of claims 1 to 7 are implemented when the computer program is executed by the processor.
10. A non-transitory computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911168434.1A CN112836843B (en) | 2019-11-25 | 2019-11-25 | Base station out-of-service alarm prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911168434.1A CN112836843B (en) | 2019-11-25 | 2019-11-25 | Base station out-of-service alarm prediction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112836843A true CN112836843A (en) | 2021-05-25 |
| CN112836843B CN112836843B (en) | 2024-04-09 |
Family
ID=75922332
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911168434.1A Active CN112836843B (en) | 2019-11-25 | 2019-11-25 | Base station out-of-service alarm prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112836843B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113365297A (en) * | 2021-08-10 | 2021-09-07 | 广东南方电信规划咨询设计院有限公司 | Base station antenna performance monitoring method and device, electronic equipment and storage medium |
| CN114375035A (en) * | 2021-12-29 | 2022-04-19 | 中国电信股份有限公司 | Base station parameter adjusting method and device, electronic equipment and readable storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20030046905A (en) * | 2001-12-07 | 2003-06-18 | 에스케이 텔레콤주식회사 | Method for Call Fail Cause Analysis in Mobile Communication System |
| CN101242549A (en) * | 2007-11-22 | 2008-08-13 | 中国移动通信集团山东有限公司 | Neutral network construction method for communication network alarm association |
| CN101742541A (en) * | 2008-11-13 | 2010-06-16 | 中国移动通信集团河北有限公司 | Early warning method and system of base station service quit |
| CN104270779A (en) * | 2014-10-17 | 2015-01-07 | 浪潮通信信息系统有限公司 | Wireless network fault reason preprocessing method and wireless network fault work order distribution system |
| CN107295553A (en) * | 2017-08-21 | 2017-10-24 | 中国铁塔股份有限公司 | A kind of base station maintenance method and device |
| CN108063676A (en) * | 2016-11-08 | 2018-05-22 | 中国移动通信集团山西有限公司 | Communication network failure method for early warning and device |
-
2019
- 2019-11-25 CN CN201911168434.1A patent/CN112836843B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20030046905A (en) * | 2001-12-07 | 2003-06-18 | 에스케이 텔레콤주식회사 | Method for Call Fail Cause Analysis in Mobile Communication System |
| CN101242549A (en) * | 2007-11-22 | 2008-08-13 | 中国移动通信集团山东有限公司 | Neutral network construction method for communication network alarm association |
| CN101742541A (en) * | 2008-11-13 | 2010-06-16 | 中国移动通信集团河北有限公司 | Early warning method and system of base station service quit |
| CN104270779A (en) * | 2014-10-17 | 2015-01-07 | 浪潮通信信息系统有限公司 | Wireless network fault reason preprocessing method and wireless network fault work order distribution system |
| CN108063676A (en) * | 2016-11-08 | 2018-05-22 | 中国移动通信集团山西有限公司 | Communication network failure method for early warning and device |
| CN107295553A (en) * | 2017-08-21 | 2017-10-24 | 中国铁塔股份有限公司 | A kind of base station maintenance method and device |
Non-Patent Citations (5)
| Title |
|---|
| YIN-HSIN LIU, 等: "Predicting malfunction of mobile network base station using machine learning approach", 《2019 20TH ASIA-PACIFIC NETWORK OPERATIONS AND MANAGEMENT SYMPOSIUM》, pages 1 - 4 * |
| 刘洪波,等: "基于神经网络的通信网络告警关联分析及应用", 《电信技术》, no. 5, pages 32 - 35 * |
| 王洋, 等: "基于集中监控数据资源的4G基站退服故障预警模型", 《电信科学》, vol. 32, no. 7, pages 188 - 196 * |
| 石少民: "数据挖掘在移动基站退服故障预测中的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》, no. 10, pages 136 - 82 * |
| 魏鹏涛,等: "基于大数据的 5G 基站退服成本估算", 《电子技术应用》, vol. 45, no. 10, pages 15 - 16 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113365297A (en) * | 2021-08-10 | 2021-09-07 | 广东南方电信规划咨询设计院有限公司 | Base station antenna performance monitoring method and device, electronic equipment and storage medium |
| CN114375035A (en) * | 2021-12-29 | 2022-04-19 | 中国电信股份有限公司 | Base station parameter adjusting method and device, electronic equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112836843B (en) | 2024-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108513251B (en) | Positioning method and system based on MR data | |
| CN109146093A (en) | A kind of electric power equipment on-site exploration method based on study | |
| CN113486584B (en) | Method and device for predicting equipment failure, computer equipment and computer readable storage medium | |
| CN107077132A (en) | The method of the subsystem managed a processing factory using dcs | |
| CN112836843A (en) | Method and device for predicting base station out-of-service alarm | |
| CN105471647A (en) | Power communication network fault positioning method | |
| CN108022123B (en) | Automatic adjustment method and device for business model | |
| CN112270429A (en) | Cloud edge cooperation-based power battery pole piece manufacturing equipment maintenance method and system | |
| Eltyshev et al. | Intelligent decision support in the electrical equipment diagnostics | |
| CN115392627A (en) | Online risk assessment method and system for power distribution network | |
| CN103942251A (en) | Method and system for inputting high altitude meteorological data into database based on multiple quality control methods | |
| AU2019100631A4 (en) | Self-correcting multi-model numerical rainfall ensemble forecasting method | |
| CN115033450A (en) | Bayesian cluster monitoring early warning analysis method based on distribution | |
| CN114553671A (en) | Diagnosis method for power communication network fault alarm | |
| CN113825165B (en) | 5G slice network congestion early warning method and device based on time diagram network | |
| CN109978299B (en) | Data analysis method and device for offshore wind power business and storage medium | |
| CN115409100A (en) | Method, device, medium and equipment for generating power accident emergency scheme | |
| CN115689320A (en) | Health management method and device for base station operation and computer readable storage medium | |
| CN113255994A (en) | Optimal configuration method, device, equipment and medium for power grid CPS | |
| CN113487087A (en) | Method and device for predicting service life of equipment, computer equipment and computer-readable storage medium | |
| CN112104467B (en) | Cutover operation risk rating method and device and computing equipment | |
| CN117131947B (en) | Overhead transmission line fault prediction method, device, equipment and storage medium | |
| CN112801815B (en) | Power communication network fault early warning method based on federal learning | |
| CN117217546B (en) | Power transmission line lightning trip prediction model, method, system and storage medium | |
| CN117113267B (en) | Prediction model training method based on big data and photovoltaic power generation performance detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |