Operation monitoring protection method of intelligent module based on Internet of Things technology

Abstract

The invention relates to the technical field of communication monitoring and early warning, and discloses an intelligent module operation monitoring protection method based on the Internet of Things technology, and the method comprises the following steps: S101, collecting time domain signal data of a communication module; step S102, preprocessing the time domain signal data to obtain a second feature sequence; s103, constructing a communication state recognition model according to the second feature sequence to output probability values of different communication states; and S104, generating a solution according to the probability values of different communication states output by the communication state recognition model. According to the method, the communication state recognition model is constructed to set the probability value of each communication state, and a corresponding solution is adopted, so that intelligent and automatic management of the communication link can be realized, and the reliability and the safety of the communication module are effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of communication monitoring and early warning, and in particular relates to an operation monitoring and protection method of an intelligent module based on the Internet of Things technology. Background Art

[0002] In the IoT system, the communication module is the core component for the device to interact with the outside world. The communication module not only undertakes the key task of information transmission, but also needs to ensure the stability and reliability of data transmission. Whether it is a smart home, industrial automation, or telemedicine, the efficient operation of the communication module is the basis for the normal operation of the system. For the communication module, the source and type of interference are very important, because different sources of interference may require different treatment methods. For example, electromagnetic interference may require shielding technology, while frequency interference may require frequency adjustment or the use of more anti-interference protocols. Therefore, being able to identify the specific source and type of interference is crucial to optimizing the communication link and ensuring the normal operation of the device. Common types of interference include: electromagnetic interference, wireless frequency interference, physical obstruction, and bandwidth interference.

[0003] Current communication module monitoring technologies include: signal strength monitoring, which monitors the signal strength received by the device; and spectrum analysis, which monitors the distribution of the signal in the frequency domain to obtain the source of interference. However, signal strength monitoring can only reflect the strength of the signal, but cannot identify the specific source of interference. Spectrum analysis can only identify wireless frequency interference of the same frequency, and is costly and requires manual intervention.

[0004] Therefore, there is an urgent need for an operation monitoring and protection method of an intelligent module based on Internet of Things technology to solve the above problems. Summary of the invention

[0005] The present invention provides an operation monitoring and protection method for an intelligent module based on the Internet of Things technology, which solves the technical problems in the related art that the specific source of communication interference cannot be identified and the labor cost is high.

[0006] The present invention provides an operation monitoring and protection method of an intelligent module based on Internet of Things technology, comprising the following steps:

[0007] Step S101, within a preset time period T, collecting time domain signal data of the communication module at a preset time interval t, and representing it by a first characteristic sequence;

[0008] The first characteristic sequence includes N sequence units, and the nth sequence unit represents the time domain signal data at the nth time point;

[0009] The time domain signal data is represented by a time domain waveform, the horizontal axis represents the time point, and the vertical axis represents the signal amplitude;

[0010] Step S102, preprocessing the first feature sequence to obtain a second feature sequence;

[0011] Step S103, constructing a communication state recognition model according to the second feature sequence, the input of the communication state recognition model is the second feature sequence, and the output is the probability value of different communication states;

[0012] Communication status includes: normal, electromagnetic interference, wireless frequency interference, physical blocking and bandwidth interference;

[0013] Step S104: generating a solution according to the probability values ​​of different communication states output by the communication state recognition model.

[0014] Furthermore, N, n, T and t are all custom parameters, 1≤n≤N, N=T / t.

[0015] Furthermore, the step of preprocessing the first feature sequence includes:

[0016] Step S201, extracting amplitude features from the time domain signal data;

[0017] Step S202, extracting a signal strength value feature and a signal strength mean feature from the time domain signal data;

[0018] Step S203, using fast Fourier transform to convert the time domain waveform representation into a frequency domain waveform representation, the horizontal axis of the frequency domain waveform represents frequency, and the vertical axis represents amplitude, extracting the frequency point with the largest amplitude in the frequency domain waveform to obtain a frequency peak feature;

[0019] Step S204, extracting center frequency features according to the frequency domain waveform;

[0020] Step S205, obtaining a frequency band width characteristic according to the span of the frequency component in the frequency domain waveform;

[0021] Step S206, calculating the signal-to-noise ratio according to the time domain waveform to obtain a signal-to-noise ratio feature;

[0022] Step S207, concatenating the above feature vectors into a communication feature vector.

[0023] Further, the communication state recognition model includes a first hidden layer, a second hidden layer, a generator and a classifier;

[0024] The first hidden layer includes N first units, the nth first unit inputs the nth sequence unit of the second feature sequence, and outputs a hidden state feature, and the N hidden state features are represented by a hidden state sequence;

[0025] The second hidden layer includes N second units, the nth second unit inputs the nth sequence unit of the hidden state sequence, and outputs a weight feature;

[0026] The generator is used to input N hidden state features and N weight features, and output the final hidden state features;

[0027] The final hidden state features are input into the classifier, and the classification space of the classifier represents the probability values ​​of various communication states.

[0028] Furthermore, the formula of the nth first unit includes:

[0029]

[0030]

[0031]

[0032]

[0033] where x n represents the nth sequence unit of the second feature sequence of the nth first unit input, h n and h n-1 denote the second candidate hidden features of the nth and n-1th first unit outputs, respectively, h 0 Assign a value of 0, r n 、z n and They represent the reset gate, update gate and first candidate hidden feature of the nth first unit of the first hidden layer, respectively. and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the reset gate of the nth first unit of the first hidden layer, and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the reset gate of the nth first unit of the first hidden layer, and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the first candidate hidden feature of the nth first unit of the first hidden layer, ⊙ represents point-by-point multiplication, sigmoid represents the sigmoid activation function, and tanh represents the tanh activation function.

[0034] Furthermore, the formulas for the second hidden layer and the generator include:

[0035]

[0036]

[0037]

[0038]

[0039]

[0040] in represents the final hidden state features of the second hidden layer, represents the Nth intermediate hidden feature of the second hidden layer, C N represents the context vector feature of the Nth sequence unit in the second hidden layer, z N and r N represents the update gate and reset gate of the Nth first unit of the first hidden layer, h N and h N-1 represents the second candidate hidden features output by the Nth and N-1th first units of the first hidden layer, and represents the first weight parameter, the second weight parameter and the bias parameter of the Nth first unit of the first hidden layer, represents the first weight parameter corresponding to the reset gate of the Nth first unit of the first hidden layer, x n represents the Nth sequence unit of the second feature sequence input by the Nth first unit of the first hidden layer, α n represents the attention weight of the nth second unit in the second hidden layer, e N,n represents the attention score of the Nth second unit in the second hidden layer, e N,k represents the attention score outside the Nth second unit of the second hidden layer, and represents the first weight parameter and bias parameter of the attention score of the Nth second unit of the second hidden layer, (h N ;h n ) means h N and h n Vertical splicing.

[0041] Furthermore, time domain signal data is generated by using a generative adversarial network model as training samples for the communication state recognition model.

[0042] Furthermore, the loss function of the communication state recognition model is a mean square error loss function.

[0043] The present invention provides an electronic device, comprising a processor and a memory storing a computer program, wherein when the processor executes the computer program, the operation monitoring and protection method of an intelligent module based on the Internet of Things technology provided above is implemented.

[0044] The present invention provides a computer-readable storage medium storing instructions, which, when executed on a computer, enables the computer to execute the above-mentioned operation monitoring and protection method of an intelligent module based on Internet of Things technology.

[0045] The beneficial effect of the present invention is that the present invention can realize intelligent and automatic management of communication links and effectively improve the reliability and security of communication modules by constructing a communication state recognition model and setting the probability value of each communication state to adopt corresponding solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flow chart of an operation monitoring and protection method of an intelligent module based on Internet of Things technology of the present invention;

[0047] Figure 2 It is a flow chart of preprocessing time domain signal data according to the present invention. DETAILED DESCRIPTION

[0048] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0049] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in one or more embodiments of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0050] like Figure 1-Figure 2 As shown, a method for monitoring and protecting the operation of an intelligent module based on the Internet of Things technology comprises the following steps:

[0051] Step S101, within a preset time period T, collecting time domain signal data of the communication module at a preset time interval t, and representing it by a first characteristic sequence;

[0052] The first characteristic sequence includes N sequence units, and the nth sequence unit represents the time domain signal data at the nth time point;

[0053] The time domain signal data is represented by a time domain waveform, the horizontal axis represents the time point, and the vertical axis represents the signal amplitude;

[0054] Step S102, preprocessing the first feature sequence to obtain a second feature sequence;

[0055] Step S103, constructing a communication state recognition model according to the second feature sequence, the input of the communication state recognition model is the second feature sequence, and the output is the probability value of different communication states;

[0056] Communication status includes: normal, electromagnetic interference, wireless frequency interference, physical blocking and bandwidth interference;

[0057] Step S104: generating a solution according to the probability values ​​of different communication states output by the communication state recognition model.

[0058] In one embodiment of the present invention, N, n, T and t are all custom parameters, 1≤n≤N, N=T / t.

[0059] In one embodiment of the present invention, a data acquisition card is used to automatically and real-timely acquire time domain signal data of the communication module according to the signal bandwidth of the communication module, and the sampling frequency is set to 5 times of the signal bandwidth.

[0060] In one embodiment of the present invention, the step of preprocessing the first feature sequence includes:

[0061] Step S201, extracting amplitude features from the time domain signal data, where the amplitude is represented by the vertical axis of the time domain waveform;

[0062] Step S202: extracting a signal strength value feature and a signal strength mean feature from the time domain signal data. The signal strength feature is represented by the square of the signal amplitude. The calculation formula of the signal strength feature is: SSV = |x n | 2 , where x n The signal amplitude at the nth time point is represented by the signal strength mean feature obtained using the sliding time window method, specifically: at each time point, the amplitudes of the K adjacent time points before and after the point are taken, and the average of these amplitudes is calculated, and the amplitude of the current time point is replaced by the average value. Preferably, K is set to 3;

[0063] Step S203, using fast Fourier transform to convert the time domain waveform representation into a frequency domain waveform representation, the horizontal axis of the frequency domain waveform represents frequency, and the vertical axis represents amplitude, extracting the frequency point with the largest amplitude in the frequency domain waveform to obtain a frequency peak feature;

[0064] Step S204, extracting the center frequency feature according to the frequency domain waveform, the calculation formula is: where f n Represents the frequency component at a time point, A(f n ) represents the amplitude of the frequency component;

[0065] Step S205, obtaining a frequency band width characteristic according to the span of the frequency component in the frequency domain waveform;

[0066] Step S206, calculating the signal-to-noise ratio according to the time domain waveform to obtain a signal-to-noise ratio feature;

[0067] Step S207, concatenating the above feature vectors into a communication feature vector.

[0068] In one embodiment of the present invention, step S206 specifically includes:

[0069] Step S2061, calculating the power of the signal according to the time domain waveform, the calculation formula is: SP represents the signal power, x n Represents the amplitude of the signal at the nth time point;

[0070] Step S2062, the noise power is calculated based on the difference between the total signal power and the main component power of the signal, where the total signal power represents the total power of the entire signal in the entire time period, and the calculation formula of the total signal power is: STP = ∑ f A(f) 2 , where STP represents the total signal power, A(f) represents the amplitude of the frequency classification; the principal component power represents the power within the first frequency band of the maximum frequency in the frequency domain waveform. Preferably, the first frequency band is plus or minus 5 Hz of the maximum frequency. The calculation formula of the principal component power is: PCP stands for principal component power;

[0071] Step S2063: Calculate the signal-to-noise ratio according to the noise power and the main component power. The calculation formula of the signal-to-noise ratio is: Where SNR stands for signal-to-noise ratio.

[0072] In one embodiment of the present invention, the communication state recognition model includes a first hidden layer, a second hidden layer, a generator and a classifier;

[0073] The first hidden layer includes N first units, the nth first unit inputs the nth sequence unit of the second feature sequence, and outputs a hidden state feature, and the N hidden state features are represented by a hidden state sequence;

[0074] The second hidden layer includes N second units, the nth second unit inputs the nth sequence unit of the hidden state sequence, and outputs a weight feature;

[0075] The generator is used to input N hidden state features and N weight features, and output the final hidden state features;

[0076] The final hidden state features are input into the classifier, and the classification space of the classifier represents the probability values ​​of various communication states.

[0077] In one embodiment of the present invention, the formula of the nth first unit includes:

[0078]

[0079]

[0080]

[0081]

[0082] where x n represents the nth sequence unit of the second feature sequence of the nth first unit input, h n and h n-1 denote the second candidate hidden features of the nth and n-1th first unit outputs, respectively, h 0 Assign a value of 0, r n 、z n and They represent the reset gate, update gate and first candidate hidden feature of the nth first unit of the first hidden layer, respectively. and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the reset gate of the nth first unit of the first hidden layer, and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the reset gate of the nth first unit of the first hidden layer, and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the first candidate hidden feature of the nth first unit of the first hidden layer, ⊙ represents point-by-point multiplication, sigmoid represents the sigmoid activation function, and tanh represents the tanh activation function.

[0083] In one embodiment of the present invention, the formulas for the second hidden layer and the generator include:

[0084]

[0085]

[0086]

[0087]

[0088]

[0089] in represents the final hidden state features of the second hidden layer, represents the Nth intermediate hidden feature of the second hidden layer, C N represents the context vector feature of the Nth sequence unit in the second hidden layer, z N and r N represents the update gate and reset gate of the Nth first unit of the first hidden layer, h N and h N-1 represents the second candidate hidden features output by the Nth and N-1th first units of the first hidden layer, and represents the first weight parameter, the second weight parameter and the bias parameter of the Nth first unit of the first hidden layer, represents the first weight parameter corresponding to the reset gate of the Nth first unit of the first hidden layer, x n represents the Nth sequence unit of the second feature sequence input by the Nth first unit of the first hidden layer, α n represents the attention weight of the nth second unit in the second hidden layer, e N,n represents the attention score of the Nth second unit in the second hidden layer, e N,k represents the attention score outside the Nth second unit of the second hidden layer, and represents the first weight parameter and bias parameter of the attention score of the Nth second unit of the second hidden layer, (h N ;h n ) means h N and h n Vertical splicing.

[0090] In one embodiment of the present invention, the communication state recognition model calculates the weights of N time points to highlight the first candidate hidden features of important time points, thereby giving these time points greater weights, so that the model pays more attention to important time points when generating the final hidden state features.

[0091] In one embodiment of the present invention, a generative adversarial network model is used to generate time domain signal data as a training sample for a communication status recognition model. The generative adversarial network model includes a generator and a discriminator. The input of the generator is random noise sampled from a normal distribution, and the output is a communication status data sample. The input of the discriminator is real communication status data and communication status data samples, and the output input data is a probability value of real data. The steps of training the generative adversarial network model are: first fix the generator and train the discriminator, then fix the discriminator and train the generator, and use the cross entropy function as the loss function to update the generator and the discriminator. At the same time, the generated communication status data is annotated by manual annotation to obtain sample labels for the communication status recognition model.

[0092] In one embodiment of the present invention, different initial thresholds are set according to the communication state probability value output by the communication state identification model to adopt corresponding solutions to ensure the reliability of the communication link. The initial threshold of the probability of normal communication state is set to 0.9. When the probability of normal communication state is greater than 0.9, it indicates that there is no problem with communication and no measures need to be taken. The initial threshold of the probability of electromagnetic interference state is set to 0.2. When the probability of electromagnetic interference state exceeds 0.2, it is preliminarily judged as an electromagnetic interference problem, and anti-interference strategies should be initiated, such as adjusting the communication frequency band or increasing the transmission power. The initial threshold of the probability of wireless frequency interference state is set to 0.2. When the probability of wireless frequency interference state exceeds 0.2, it is preliminarily judged as co-frequency interference, and switching channels or changing communication frequencies should be considered. The initial threshold of the probability of physical obstruction state is set to 0.1. When the probability of physical obstruction state output by the model exceeds 0.1, the physical environment in the communication link should be checked, and attempts should be made to adjust the device position or add relay nodes. The initial threshold of the probability of bandwidth interference state is set to 0.1. When the probability of bandwidth interference state output by the model exceeds 0.1, there may be channel occupancy or network congestion, and optimization of communication protocols or traffic management should be considered.

[0093] In one embodiment of the present invention, the loss function of the communication state recognition model is a mean square error loss function.

[0094] An embodiment of the present invention provides an electronic device, including a processor and a memory storing a computer program, wherein when the processor executes the computer program, an operation monitoring and protection method of an intelligent module based on Internet of Things technology provided by the above embodiments is implemented.

[0095] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0096] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A method for monitoring and protecting the operation of an intelligent module based on Internet of Things technology, characterized in that: The following steps are involved: Step S101, within a preset time period T, collecting time domain signal data of the communication module at a preset time interval t, and representing it by a first characteristic sequence; The first characteristic sequence includes N sequence units, and the nth sequence unit represents the time domain signal data at the nth time point; The time domain signal data is represented by a time domain waveform, the horizontal axis represents the time point, and the vertical axis represents the signal amplitude; Step S102, preprocessing the first feature sequence to obtain a second feature sequence; Step S103, constructing a communication state recognition model according to the second feature sequence, the input of the communication state recognition model is the second feature sequence, and the output is the probability value of different communication states; Communication status includes: normal, electromagnetic interference, wireless frequency interference, physical blocking and bandwidth interference; Step S104: generating a solution according to the probability values ​​of different communication states output by the communication state recognition model.

2. The operation monitoring and protection method of an intelligent module based on Internet of Things technology according to claim 1 is characterized in that: N, n, T and t are all custom parameters, 1≤n≤N, N=T / t.

3. The operation monitoring and protection method of an intelligent module based on Internet of Things technology according to claim 1 is characterized in that: The step of preprocessing the first feature sequence comprises: Step S201, extracting amplitude features from the time domain signal data; Step S202, extracting a signal strength value feature and a signal strength mean feature from the time domain signal data; Step S203, using fast Fourier transform to convert the time domain waveform representation into a frequency domain waveform representation, the horizontal axis of the frequency domain waveform represents frequency, and the vertical axis represents amplitude, extracting the frequency point with the largest amplitude in the frequency domain waveform to obtain a frequency peak feature; Step S204, extracting center frequency features according to the frequency domain waveform; Step S205, obtaining a frequency band width characteristic according to the span of the frequency component in the frequency domain waveform; Step S206, calculating the signal-to-noise ratio according to the time domain waveform to obtain a signal-to-noise ratio feature; Step S207, concatenating the above feature vectors into a communication feature vector.

4. The operation monitoring and protection method of an intelligent module based on Internet of Things technology according to claim 1 is characterized in that: The communication state recognition model includes a first hidden layer, a second hidden layer, a generator and a classifier; The first hidden layer includes N first units, the nth first unit inputs the nth sequence unit of the second feature sequence, and outputs a hidden state feature, and the N hidden state features are represented by a hidden state sequence; The second hidden layer includes N second units, the nth second unit inputs the nth sequence unit of the hidden state sequence, and outputs a weight feature; The generator is used to input N hidden state features and N weight features, and output the final hidden state features; The final hidden state features are input into the classifier, and the classification space of the classifier represents the probability values ​​of various communication states.

5. The operation monitoring and protection method of an intelligent module based on Internet of Things technology according to claim 1 is characterized in that: The formulas for the nth first unit include: where x n represents the nth sequence unit of the second feature sequence of the nth first unit input, h n and h n-1 They represent the second candidate hidden features of the nth and n-1th first unit outputs, h0 is assigned to 0, r n 、z n and They represent the reset gate, update gate and first candidate hidden feature of the nth first unit of the first hidden layer, respectively. and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the reset gate of the nth first unit of the first hidden layer, and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the reset gate of the nth first unit of the first hidden layer, and They respectively represent the first weight parameter, the second weight parameter and the bias parameter corresponding to the first candidate hidden feature of the nth first unit of the first hidden layer, ⊙ represents point-by-point multiplication, sigmoid represents the sigmoid activation function, and tanh represents the tanh activation function.

6. The operation monitoring and protection method of an intelligent module based on Internet of Things technology according to claim 1 is characterized in that: The formulas for the second hidden layer and the generator include: in represents the final hidden state features of the second hidden layer, represents the Nth intermediate hidden feature of the second hidden layer, C N represents the context vector feature of the Nth sequence unit in the second hidden layer, z N and r N represents the update gate and reset gate of the Nth first unit of the first hidden layer, h N and h N-1 represents the second candidate hidden features output by the Nth and N-1th first units of the first hidden layer, and represents the first weight parameter, the second weight parameter and the bias parameter of the Nth first unit of the first hidden layer, represents the first weight parameter corresponding to the reset gate of the Nth first unit of the first hidden layer, x n represents the Nth sequence unit of the second feature sequence input by the Nth first unit of the first hidden layer, α n represents the attention weight of the nth second unit in the second hidden layer, e N,n represents the attention score of the Nth second unit in the second hidden layer, e N,k represents the attention score outside the Nth second unit of the second hidden layer, and represents the first weight parameter and bias parameter of the attention score of the Nth second unit of the second hidden layer, (h N ;h n ) means h N and h n Vertical splicing.

7. The operation monitoring and protection method of an intelligent module based on Internet of Things technology according to claim 1 is characterized in that: Time domain signal data is generated by using a generative adversarial network model as training samples for the communication state recognition model.

8. The operation monitoring and protection method of an intelligent module based on Internet of Things technology according to claim 1 is characterized in that: The loss function of the communication status recognition model is the mean square error loss function.

9. An electronic device comprising a processor and a memory storing a computer program, characterized in that: When the processor executes the computer program, the operation monitoring and protection method of the intelligent module based on the Internet of Things technology as described in any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements an operation monitoring and protection method for an intelligent module based on Internet of Things technology as described in any one of claims 1 to 8.

Images