Wireless communication network interference signal automatic alarm device based on big data

Abstract

According to the wireless communication network interference signal automatic alarm device based on big data, historical data of wireless communication network interference signals are acquired through the data acquisition module, the interference signal identification model is established, and real-time detection and alarm are performed on the wireless communication network through the alarm module; the device optimizes an interference signal alarm program of the wireless communication network, enlarges the capture and identification range of the wireless communication network to the interference signal, improves the alarm efficiency of the interference signal of the wireless communication network, and ensures the stable operation of the wireless communication network.

Description

Technical Field

[0001] The present application belongs to the field of wireless communication network interference signal detection, and specifically relates to a wireless communication network interference signal automatic alarm device based on big data. Background Art

[0002] A wireless communication network refers to a network that can interconnect various communication devices without the need for wiring. A wireless communication network completes interaction, data transmission and information sharing through signals. A wireless communication network is an important part of the information construction of modern society, and meets people's needs for communication speed, quality and security. With the development of science and technology, people's demand for wireless communication networks is getting higher and higher, which continuously promotes the advancement of wireless communication technology. An interference signal refers to a signal that damages the reception of useful signals. An interference signal will affect the normal operation of a wireless communication network to a certain extent, and bring adverse effects to the signal transmission of the wireless communication network. In order to ensure the normal operation of the wireless communication network and the normal transmission of wireless signals, it is necessary to capture and alarm the interference signal. The overall alarm efficiency of the automatic alarm method for network interference signals in the existing technology is low, and it is impossible to capture the interference signal in time and issue an alarm. Summary of the invention

[0003] In order to solve the problem that the existing technology cannot capture interference signals in time and issue an alarm, a wireless communication network interference signal automatic alarm device based on big data is proposed. The device collects historical data of wireless communication network interference signals based on the fusion of big data, establishes an interference signal recognition model, and realizes the recognition of wireless communication network interference signals through the interference signal recognition model;

[0004] A wireless communication network interference signal automatic alarm device based on big data, comprising: a data acquisition module, a model building module, a model training module and an alarm module;

[0005] A data acquisition module, used to obtain historical data of interference signals of the wireless communication network, and send the historical data of interference signals of the wireless communication network to the model building module;

[0006] A model building module, used to build an interference signal recognition model based on the historical data of interference signals of the wireless communication network; the historical data of interference signals of the wireless communication network includes a waveform image of the interference signal;

[0007] A model training module, used to use the interference signal historical data as an input parameter of the interference signal recognition model, and use the interference signal type corresponding to the interference signal historical data as an output parameter of the interference signal recognition model to train the interference signal recognition model;

[0008] The alarm module is used to obtain real-time data of the wireless communication network, and input the real-time data of the wireless communication network into the trained interference signal recognition model to obtain the recognition result; and to issue an alarm based on the recognition result; the real-time data of the wireless communication network includes the waveform image of the wireless communication network signal at the current moment.

[0009] Preferably, the interference signal recognition model is a YOLOv2 algorithm network.

[0010] Preferably, the interference signal recognition model includes a feature extraction module, a fusion module and a recognition output module;

[0011] The feature extraction module is used to make the input interference signal waveform image pass through the No. 1 vector convolution layer conv, the No. 1 maximum pooling layer Maxpool, the No. 2 vector convolution layer conv, the No. 2 maximum pooling layer Maxpool, the No. 1 convolution module, the No. 3 maximum pooling layer Maxpool, the No. 2 convolution module, the No. 4 maximum pooling layer Maxpool, and the No. 3 convolution module in sequence to obtain the first interference signal waveform feature map; make the input interference signal waveform image pass through the No. 1 vector convolution layer conv, the No. 1 maximum pooling layer Maxpool, the No. 2 vector convolution layer conv, the No. 2 vector convolution layer conv, the No. 3 maximum pooling layer Maxpool, and the No. 3 convolution module in sequence to obtain the first interference signal waveform feature map. Layer conv, No. 2 maximum pooling layer Maxpool, No. 1 convolution module, No. 3 maximum pooling layer Maxpool, No. 2 convolution module, No. 4 maximum pooling layer Maxpool, No. 3 convolution module, No. 5 maximum pooling layer Maxpool, No. 4 convolution module to obtain the second interference signal waveform feature map; the obtained first interference signal waveform feature map and the second interference signal waveform feature map are sent to the fusion module; No. 1 convolution module and No. 2 convolution module are implemented by three vector convolution layers, and No. 3 convolution module and No. 4 convolution module are implemented by five vector convolution layers;

[0012] The fusion module convolves the received first interference signal waveform feature map through the two-dimensional matrix convolution layer conv2d, and convolves the received second interference signal waveform feature map through the two-dimensional matrix convolution module, and fuses the convolved first interference signal waveform feature map and the second interference signal waveform feature map through the fusion layer Concate to obtain an interference signal waveform fusion feature map, and inputs the interference signal waveform fusion feature map into the recognition output module; the two-dimensional matrix convolution module is implemented by two two-dimensional matrix convolution layers;

[0013] The recognition output module is used to make the input interference signal waveform fusion feature map pass through the two-dimensional matrix convolution layer conv2d, the average pooling layer Avgpool and the classification layer Softmax in sequence, and finally obtain the interference signal type.

[0014] Beneficial Effects

[0015] The present application discloses an automatic alarm device for interference signals of wireless communication network based on big data, and the device comprises a data acquisition module, a model building module, a model training module and an alarm module; the device obtains historical data of interference signals of wireless communication network through the data acquisition module, establishes an interference signal recognition model through the model building module, and performs real-time detection and alarm of wireless communication network through the alarm module; the device optimizes the alarm procedure of interference signals of wireless communication network, expands the capture and recognition range of interference signals of wireless communication network, improves the alarm efficiency of interference signals of wireless communication network, and ensures the stable operation of wireless communication network. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a structural diagram of the wireless communication network interference signal automatic alarm device based on big data in this embodiment;

[0017] Figure 2 This is a flow chart of the automatic alarm method for wireless communication network interference signals based on big data according to this embodiment;

[0018] Figure 3 This is a structural diagram of the interference signal identification model of this embodiment. DETAILED DESCRIPTION

[0019] Specific implementation method 1: The following will be combined with the attached embodiment of the present invention Figure 1 To Attachment Figure 3 , illustrate this implementation mode, and clearly and completely describe the technical solutions in the embodiments of the present invention:

[0020] The present application discloses a wireless communication network interference signal automatic alarm device based on big data, comprising: a data acquisition module, a model building module, a model training module and an alarm module;

[0021] A data acquisition module, used to obtain historical data of interference signals of the wireless communication network, and send the historical data of interference signals of the wireless communication network to the model building module;

[0022] A model building module, used to build an interference signal recognition model based on the historical data of interference signals of the wireless communication network; the historical data of interference signals of the wireless communication network includes a waveform image of the interference signal;

[0023] A model training module, used to use the interference signal historical data as an input parameter of the interference signal recognition model, and use the interference signal type corresponding to the interference signal historical data as an output parameter of the interference signal recognition model to train the interference signal recognition model;

[0024] The alarm module is used to obtain real-time data of the wireless communication network, and input the real-time data of the wireless communication network into the trained interference signal recognition model to obtain the recognition result; and to issue an alarm based on the recognition result; the real-time data of the wireless communication network includes the waveform image of the wireless communication network signal at the current moment.

[0025] Specifically, wireless communication network signal parameters are set according to communication requirements, a wireless signal receiving device collects wireless communication network interference signals of different time periods and different types to obtain a wireless communication network interference signal waveform image, and performs noise reduction processing on the collected wireless communication network interference signal waveform image; a wireless communication network interference signal waveform image data set is constructed; and frequency, power, bandwidth, occupancy and other information of the corresponding wireless communication network interference signal can be obtained according to the wireless communication network interference signal waveform image;

[0026] The wireless communication network interference signal waveform image data set is divided into a training set and a test set in proportion, and the division ratio can be 80% for training and 20% for testing; the specific division method is: all wireless communication network interference signal waveform images are shuffled in order, and according to the set ratio, the first 80% of the image files are used as the training set, and the remaining 20% ​​of the image files are used as the test set;

[0027] Furthermore, the data acquisition module is a radio frequency wireless receiver.

[0028] Furthermore, the interference signal recognition model is a YOLOv2 algorithm network.

[0029] Furthermore, combined with Figure 3 Interference signal recognition model structure diagram, the interference signal recognition model includes a feature extraction module, a fusion module and a recognition output module;

[0030] The feature extraction module is used to make the input interference signal waveform image pass through the No. 1 vector convolution layer conv, the No. 1 maximum pooling layer Maxpool, the No. 2 vector convolution layer conv, the No. 2 maximum pooling layer Maxpool, the No. 1 convolution module, the No. 3 maximum pooling layer Maxpool, the No. 2 convolution module, the No. 4 maximum pooling layer Maxpool, and the No. 3 convolution module in sequence to obtain the first interference signal waveform feature map; make the input interference signal waveform image pass through the No. 1 vector convolution layer conv, the No. 1 maximum pooling layer Maxpool, the No. 2 vector convolution layer conv, the No. 2 vector convolution layer conv, the No. 3 maximum pooling layer Maxpool, and the No. 3 convolution module in sequence to obtain the first interference signal waveform feature map. Layer conv, No. 2 maximum pooling layer Maxpool, No. 1 convolution module, No. 3 maximum pooling layer Maxpool, No. 2 convolution module, No. 4 maximum pooling layer Maxpool, No. 3 convolution module, No. 5 maximum pooling layer Maxpool, No. 4 convolution module to obtain the second interference signal waveform feature map; the obtained first interference signal waveform feature map and the second interference signal waveform feature map are sent to the fusion module; No. 1 convolution module and No. 2 convolution module are implemented by three vector convolution layers, and No. 3 convolution module and No. 4 convolution module are implemented by five vector convolution layers;

[0031] The fusion module convolves the received first interference signal waveform feature map through the two-dimensional matrix convolution layer conv2d, and convolves the received second interference signal waveform feature map through the two-dimensional matrix convolution module, and fuses the convolved first interference signal waveform feature map and the second interference signal waveform feature map through the fusion layer Concate to obtain an interference signal waveform fusion feature map, and inputs the interference signal waveform fusion feature map into the recognition output module; the two-dimensional matrix convolution module is implemented by two two-dimensional matrix convolution layers;

[0032] The recognition output module is used to make the input interference signal waveform fusion feature map pass through the two-dimensional matrix convolution layer conv2d, the average pooling layer Avgpool and the classification layer Softmax in sequence, and finally obtain the interference signal type.

[0033] Furthermore, the softmax classification formula is,

[0034] Among them, z i is the logical value of the i-th category, N is the total number of logical values, P i is the probability distribution of the ith category, where

[0035] In the process of training the interference signal recognition model in this embodiment,

[0036] An interference signal recognition model is established based on the wireless communication network interference signal waveform image dataset, and the interference signal recognition model is trained based on the wireless communication network interference signal waveform image training set. YOLOv2 uses the deep learning framework darknet. After multiple trainings and parameter adjustments, the initial learning rate is finally set to 0.001, the number of unit rounds is 8, and the maximum number of iteration rounds is 19000. The features of the wireless communication network interference signal waveform image are extracted and the corresponding weights are generated. Finally, the YOLOv2 training process is visualized. When the value of the average loss function is smaller, it means that the actual output is closer to the expected output, and the training effect is relatively good.

[0037] The input of the interference signal recognition model is the interference signal waveform image of the wireless communication network, and the output is the corresponding interference signal type; in the training process of different data sets, YOLOv2 generalizes the interference signal waveform image feature detection through a large amount of learning and parameter optimization, which can not only verify and identify the collected data sets, but also be suitable for detecting other untrained scenarios. Finally, weights with interference signal waveform image features are generated, and they are called in the fusion algorithm and the interference signal waveform image features are screened for the second time; the wireless communication network interference signal waveform image test set is used to test the interference signal recognition model to verify whether the interference signal recognition model obtained after training has overfitting phenomenon.

[0038] Furthermore, the alarm module issues alarms of different levels according to the recognition results.

[0039] Specifically, the interference signal recognition model obtains recognition results based on real-time data of the wireless communication network. Since different types of interference signals have different impacts on the wireless communication network, the alarm module issues different levels of alarms based on the recognition results, which is conducive to timely processing by relevant technical personnel.

[0040] Furthermore, the wireless communication network interference signal automatic alarm device based on big data also includes a mobile terminal, and the alarm module sends the identification result to the mobile terminal. Specific implementation method 2:

[0042] The following will be combined Figure 2 , explain this implementation mode, and clearly and completely describe the technical solutions in the embodiments of the present invention;

[0043] A method for automatic warning of interference signals in a wireless communication network based on big data, comprising:

[0044] Obtain historical data of interference signals of the wireless communication network, and establish an interference signal recognition model according to the historical data of the interference signals; the historical data of the interference signals include waveform images of the interference signals; use the historical data of the interference signals as input parameters of the interference signal recognition model, and use the interference signal types corresponding to the historical data of the interference signals as output parameters of the interference signal recognition model to train the interference signal recognition model; obtain real-time data of the wireless communication network, and input the real-time data of the wireless communication network into the trained interference signal recognition model to obtain recognition results, and issue an alarm according to the recognition results; the real-time data of the wireless communication network include waveform images of the wireless communication network signals at the current moment.

[0045] Specifically, the wireless communication network interference signal automatic alarm method based on big data establishes an interference signal recognition model, identifies the interference signal through the interference signal recognition model and issues an alarm based on the recognition result, thereby realizing the monitoring of the wireless communication network interference signal; improving the alarm efficiency of the wireless communication network interference signal and ensuring the stable operation of the wireless communication network.

[0046] Furthermore, the interference signal recognition model is a YOLOv2 algorithm network.

[0047] Specifically, the interference signal recognition model based on the YOLOv2 algorithm network can effectively realize the recognition of interference signals and improve the efficiency of interference signal recognition.

[0048] Furthermore, the method issues different levels of warnings based on the recognition results.

[0049] Specifically, the interference signal recognition model obtains recognition results based on real-time data of the wireless communication network. Since different types of interference signals have different impacts on the wireless communication network, the alarm module issues different levels of alarms based on the recognition results, which is conducive to timely processing by relevant technical personnel.

[0050] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in a manner different from that described in the original claims. It should also be understood that features described in conjunction with individual embodiments may be used in other described embodiments.

Claims

1. A wireless communication network interference signal automatic alarm device based on big data, characterized in that: include: Data collection module, model building module, model training module and alarm module; A data acquisition module, used to obtain historical data of interference signals of the wireless communication network, and send the historical data of interference signals of the wireless communication network to the model building module; A model building module, used to build an interference signal recognition model based on the historical data of interference signals of the wireless communication network; the historical data of interference signals of the wireless communication network includes a waveform image of the interference signal; A model training module, used to use the interference signal historical data as an input parameter of the interference signal recognition model, and use the interference signal type corresponding to the interference signal historical data as an output parameter of the interference signal recognition model to train the interference signal recognition model; The alarm module is used to obtain real-time data of the wireless communication network, and input the real-time data of the wireless communication network into the trained interference signal recognition model to obtain the recognition result; and to issue an alarm based on the recognition result; the real-time data of the wireless communication network includes the waveform image of the wireless communication network signal at the current moment.

2. According to the big data-based automatic alarm device for wireless communication network interference signals according to claim 1, it is characterized in that: The data acquisition module is a radio frequency wireless receiver.

3. According to the big data-based wireless communication network interference signal automatic alarm device of claim 1, it is characterized by: The interference signal recognition model is the YOLOv2 algorithm network.

4. According to claim 3, a wireless communication network interference signal automatic alarm device based on big data is characterized in that: The interference signal recognition model includes a feature extraction module, a fusion module and a recognition output module; The feature extraction module is used to make the input interference signal waveform image pass through the No. 1 vector convolution layer conv, the No. 1 maximum pooling layer Maxpool, the No. 2 vector convolution layer conv, the No. 2 maximum pooling layer Maxpool, the No. 1 convolution module, the No. 3 maximum pooling layer Maxpool, the No. 2 convolution module, the No. 4 maximum pooling layer Maxpool, and the No. 3 convolution module in sequence to obtain the first interference signal waveform feature map; make the input interference signal waveform image pass through the No. 1 vector convolution layer conv, the No. 1 maximum pooling layer Maxpool, the No. 2 vector convolution layer conv, the No. 2 vector convolution layer conv, the No. 3 maximum pooling layer Maxpool, and the No. 3 convolution module in sequence to obtain the first interference signal waveform feature map. Layer conv, No. 2 maximum pooling layer Maxpool, No. 1 convolution module, No. 3 maximum pooling layer Maxpool, No. 2 convolution module, No. 4 maximum pooling layer Maxpool, No. 3 convolution module, No. 5 maximum pooling layer Maxpool, No. 4 convolution module to obtain the second interference signal waveform feature map; the obtained first interference signal waveform feature map and the second interference signal waveform feature map are sent to the fusion module; No. 1 convolution module and No. 2 convolution module are implemented by three vector convolution layers, and No. 3 convolution module and No. 4 convolution module are implemented by five vector convolution layers; The fusion module convolves the received first interference signal waveform feature map through the two-dimensional matrix convolution layer conv2d, and convolves the received second interference signal waveform feature map through the two-dimensional matrix convolution module, and fuses the convolved first interference signal waveform feature map and the second interference signal waveform feature map through the fusion layer Concate to obtain an interference signal waveform fusion feature map, and inputs the interference signal waveform fusion feature map into the recognition output module; the two-dimensional matrix convolution module is implemented by two two-dimensional matrix convolution layers; The recognition output module is used to make the input interference signal waveform fusion feature map pass through the two-dimensional matrix convolution layer conv2d, the average pooling layer Avgpool and the classification layer Softmax in sequence, and finally obtain the interference signal type.

5. According to claim 4, a wireless communication network interference signal automatic alarm device based on big data is characterized in that: The Softmax classification formula is: Among them, z i is the logical value of the i-th category, N is the total number of logical values, P i is the probability distribution of the ith category, 6. According to the big data-based wireless communication network interference signal automatic alarm device of claim 1, it is characterized by: The alarm module issues different levels of alarms based on the recognition results.

7. According to claim 1, a wireless communication network interference signal automatic alarm device based on big data is characterized in that: It also includes a mobile terminal, and the alarm module sends the identification result to the mobile terminal.

8. A method for automatically warning of interference signals in a wireless communication network based on big data according to claim 1, characterized in that: Obtain historical data of interference signals of the wireless communication network, and establish an interference signal recognition model based on the historical data of the interference signals; the historical data of the interference signals include waveform images of the interference signals; use the historical data of the interference signals as input parameters of the interference signal recognition model, and use the interference signal types corresponding to the historical data of the interference signals as output parameters of the interference signal recognition model to train the interference signal recognition model; obtain real-time data of the wireless communication network, and input the real-time data of the wireless communication network into the trained interference signal recognition model to obtain recognition results, and issue an alarm based on the recognition results; the real-time data of the wireless communication network include waveform images of the wireless communication network signals at the current moment.

9. The method for automatic warning of interference signals in wireless communication networks based on big data according to claim 6 is characterized in that: The interference signal recognition model is a YOLOv2 algorithm network.

10. The method for automatic warning of interference signals in wireless communication networks based on big data according to claim 7, characterized in that: This method issues different levels of warnings based on the recognition results.

Images