CN114121305A - Sensor safety detection method and system based on frequency sweep technology - Google Patents
Sensor safety detection method and system based on frequency sweep technology Download PDFInfo
- Publication number
- CN114121305A CN114121305A CN202111394732.XA CN202111394732A CN114121305A CN 114121305 A CN114121305 A CN 114121305A CN 202111394732 A CN202111394732 A CN 202111394732A CN 114121305 A CN114121305 A CN 114121305A
- Authority
- CN
- China
- Prior art keywords
- frequency
- sensor
- data
- sweep
- band
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 91
- 238000005516 engineering process Methods 0.000 title claims abstract description 27
- 238000012360 testing method Methods 0.000 claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 26
- 230000005284 excitation Effects 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 20
- 238000007781 pre-processing Methods 0.000 claims abstract description 9
- 230000002159 abnormal effect Effects 0.000 claims description 16
- 238000003062 neural network model Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 11
- 230000000737 periodic effect Effects 0.000 claims description 9
- 238000010408 sweeping Methods 0.000 claims description 9
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 claims description 7
- 238000013480 data collection Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 5
- 230000005856 abnormality Effects 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 230000003287 optical effect Effects 0.000 claims description 2
- 239000000126 substance Substances 0.000 claims description 2
- 238000004422 calculation algorithm Methods 0.000 abstract description 5
- 238000010801 machine learning Methods 0.000 abstract description 5
- 238000001228 spectrum Methods 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 abstract description 2
- 230000001133 acceleration Effects 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008447 perception Effects 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000012806 monitoring device Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 240000007651 Rubus glaucus Species 0.000 description 1
- 235000011034 Rubus glaucus Nutrition 0.000 description 1
- 235000009122 Rubus idaeus Nutrition 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000005389 magnetism Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
- G06F2218/16—Classification; Matching by matching signal segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y30/00—IoT infrastructure
- G16Y30/10—Security thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Signal Processing (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Security & Cryptography (AREA)
- Medical Informatics (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a method and a system for detecting the vulnerability of a sensor based on a frequency sweep technology, which comprises the following steps: firstly, generating a sweep frequency excitation signal; generating an out-of-band signal; step three, testing out-of-band signals; step four, collecting sensor data; preprocessing sensor data; step six, a model training stage; seventhly, detecting out-of-band vulnerability; and step eight, generating a test report. The system can be used for detecting the out-of-band vulnerability of the sensor, has good generalization capability, can conveniently detect the out-of-band vulnerability of any type of sensor comprehensively, and can cover the sensor loopholes in the full spectrum range such as the equal spectrum range. The system provided by the invention adopts a digital frequency sweep technology, realizes automatic detection and analysis of the vulnerability of the sensor by combining a machine learning algorithm, and provides a good reference basis for the safety protection of a sensing system in the Internet of things.
Description
Technical Field
The invention belongs to the field of vulnerability detection, and particularly relates to a sensor safety detection method and system based on a frequency sweep technology.
Background
With the continuous development of the internet of things and the continuous intellectualization of the terminal equipment, the sensor is used as the terminal equipment to sense the 'eyes' of the physical world, is a link for the interactive connection between the internet of things and the physical world, and is widely used in the internet of things system. However, the security risk for the sensors is not paid enough attention, and the data acquired by all the sensors are generally credible by default at the upper layer, and the measurement data of the sensors are tampered once the sensors are attacked, so that the blind trust on the hardware causes a great threat to the secure operation of the whole internet of things. If the automatic driving system carries out driving decision based on the sensor perception information, when the perception signal of the sensor is wrong, wrong driving decision can be led out, and serious consequences are caused. Therefore, the testing accuracy of the sensor is guaranteed, and the safety operation of the Internet of things system is very important.
The traditional sensor tests the sensitivity, the precision and other performances, uses an in-band signal for testing, and does not consider that the sensor is possibly influenced by an out-of-band signal, so that the safety problem of the sensor occurs. The in-band (in-band) signal refers to a detection signal within the design range of the sensor, and is otherwise called an out-of-band (out-of-band) signal, for example, the speed detection of the accelerometer is the in-band signal detection, and the sound wave detection of the accelerometer is the out-of-band signal detection. In a real application scene of the internet of things based on sensor perception, most of emergencies or malicious attacks are caused by the out-of-band vulnerability of the sensor. In addition, except for hardware factory test, the existing multi-sensor fusion technology can reduce the risk caused by abnormal work of a single sensor to a certain extent, but the problem of out-of-band vulnerability of the sensor is still not solved from the source.
In conclusion, how to realize a sensor out-of-band vulnerability detection system with convenient operation, strong portability, accurate and reliable measurement has great significance for the safety research and protection of the Internet of things system based on sensor perception.
Disclosure of Invention
The invention aims to provide a sensor safety detection method and system based on a frequency sweeping technology, and aims to solve the problem that the existing sensor performance detection only detects in-band signals and limit bearing environments of a sensor, but lacks the detection of vulnerability of out-of-band signals, and has limitation in practical application scenes.
The invention adopts the following technical scheme for solving the problems:
a sensor safety detection method based on a frequency sweep technology comprises the following steps:
step one, an excitation signal generation stage: setting excitation signal types, frequency sweep starting frequency, frequency sweep cut-off frequency, the number of equally spaced frequency sweep frequency points, the amplitude of the excitation signal and the detection duration parameter of the same-frequency signal by adopting a digital frequency sweep technology, and outputting a periodic excitation electric signal by an arbitrary waveform generator;
step two, out-of-band signal generation stage: converting the periodic excitation electrical signal into an out-of-band signal to be tested by using a transducer;
step three, out-of-band signal testing stage: according to the contact type of the out-of-band signal to be tested and the sensor, the out-of-band signal to be tested acts on the sensor and equipment for installing the sensor;
step four, a test data collection stage: acquiring output data of a sensor and characteristic data of equipment provided with the sensor, and preprocessing the output data and the characteristic data to be used as normal sample data;
step five, adding interference in the step three, and repeating the step three and the step four to obtain abnormal sample data;
step six, constructing a neural network model, and training the model by using normal sample data and abnormal sample data;
and seventhly, performing safety detection on the sensor to be detected by using the trained model, wherein in the safety detection process, firstly, frequency sweeping is performed in a wide frequency band range, and an optimal vulnerability frequency point is searched by combining a frequency sweeping result with a dichotomy.
Further, the type of the out-of-band signal includes any one of acoustic, optical, electrical, magnetic, thermal, chemical signals.
Further, the characteristic data of the sensor-mounted device includes an operating temperature and a vibration amplitude of the device.
Further, the data preprocessing process comprises:
4.1) taking the output data of the sensor and the characteristic data of the equipment provided with the sensor as raw data, and filtering and normalizing the raw data;
and 4.2) dividing the normalized data into a plurality of samples according to the number of the frequency points of the sweep frequency at equal intervals.
Further, the length of the abnormal sample is the same as the length of the normal sample.
Further, when the model is trained in the step six, the probability value in the range of (0,1) is output by the model, and the smaller the probability value is, the higher the possibility of sensor abnormality is.
Further, the seventh step is specifically:
7.1) firstly setting the number of m equally-spaced frequency points in a broadband range [ L, R ], wherein L is the frequency sweep starting frequency, R is the frequency sweep cut-off frequency, and m is an even number; taking the output value of the sensor to be detected, the vibration amplitude of equipment on which the sensor to be detected is installed and the working temperature data obtained in the frequency band range as original data, and carrying out filtering and normalization pretreatment on the original data;
7.2) dividing the preprocessed data into m sections according to the number of the frequency points of the sweep frequency at equal intervals, wherein the length of a sample to be tested of each section is consistent with the length of a training sample during training of the neural network model, and obtaining m sample data to be tested;
7.3) taking m sample data to be tested as the input of the trained neural network model according to the segmented time sequence to obtain m output results;
7.4) taking the size of the output mean value of the normal sample in the training process as a standard value, respectively calculating the output mean value of the front half part and the output mean value of the rear half part of the model, comparing the calculated values with the standard value, if the mean value of the front half part or the rear half part is smaller than the standard value and the difference value of the two exceeds a detection threshold value, resetting the frequency band range in the front half part or the rear half part by using a bisection method, setting the frequency sweep range of the front half part as [ L (L + R)/2], or setting the frequency sweep range of the rear half part as [ (L + R)/2, R ], repeating the steps 7.1) to 7.4) until the frequency sweep range is lower than the frequency sweep threshold value, and outputting the corresponding fragile frequency point.
Step eight, generating a detection report according to the detection result; the detection report comprises a frequency sweep detection object, a frequency sweep range, test precision, a vulnerability frequency point and a sensor output value under the vulnerability frequency point.
A sensor safety detection system based on a frequency sweeping technology is used for realizing the sensor safety detection method.
The sensor out-of-band vulnerability detection system based on the frequency sweep technology has the advantages of reasonable design, simple structure, convenience in operation, strong generalization capability and reliable detection result. The main beneficial effects comprise:
(1) the invention aims at the out-of-band vulnerability detection system of the sensor, breaks through the defect that the traditional performance detection is only carried out on the in-band signal domain, perfects the performance detection system of the sensor and has important significance for ensuring the reliability of the sensing data in the Internet of things system.
(2) The invention adopts the digital frequency sweep technology to provide stable excitation signals for a transducer device, the type of the transducer in the prior art can cover the out-of-band leak detection of the sensors of 6 types, 104 types, including sound, light, electricity, magnetism, heat and chemistry, and the detection can be divided into contact detection and non-contact detection according to whether the detection signals directly contact the sensor to be detected, the contact detection is to directly apply the out-of-band signals in a contact mode to the sensor and equipment for installing the sensor, and the non-contact detection is to apply the out-of-band signals in a non-contact mode to the sensor and the equipment for installing the sensor, for example, through air propagation and the like. Therefore, the invention establishes a complete out-of-band vulnerability detection system independent of the sensor type, and has good generalization performance.
(3) In the system for detecting the out-of-band vulnerability of the sensor, a high-efficiency digital frequency sweeping strategy is adopted to provide a stable excitation signal, so that the efficiency of detecting the out-of-band signal of the sensor is effectively improved; an anomaly detection model based on a machine learning algorithm is introduced, detection tasks are converted into classification problems, and detection accuracy and detection speed are further improved.
Drawings
FIG. 1 is a schematic view of the overall flow chart of the sensor safety detection system based on the frequency sweep technology;
FIG. 2 is a schematic diagram of out-of-band signal generation based on frequency sweep technology according to the present invention;
FIG. 3 is a schematic diagram of the data acquisition and preprocessing process of the present invention;
FIG. 4 is a schematic diagram of vulnerability detection based on a machine learning algorithm of the present invention;
FIG. 5 is a flow chart of the present invention for finding weak frequency points using modified dichotomy;
fig. 6 is a diagram illustrating a content structure of a test report according to an embodiment of the present invention.
Detailed Description
The invention provides a sensor safety detection method and system based on a frequency sweep technology, which utilize the characteristic that a sensor outputs abnormal data at an out-of-band fragile point, and combine with a machine learning classification algorithm to realize automatic vulnerability mining and detection, can effectively detect the vulnerability of the sensor in an out-of-band physical signal, and overcomes the defect that the traditional sensor detection system only aims at in-band signal detection. The detection method and the detection system of the invention complement the defects of the sensor in out-of-band signal leak detection in the full spectrum range such as the sound wave resonance leak, the filter leak, the saturation leak, the mixing leak, the photoelectric coupling leak, the electromagnetic coupling leak, the nonlinear intermodulation distortion leak, the envelope extraction leak and the like.
The embodiments and technical solutions of the present invention are further described in detail below with reference to the accompanying drawings:
as shown in fig. 1, the following describes an embodiment of the sensor security detection based on the frequency sweep technology according to the present invention, taking the detection of out-of-band vulnerability of the acoustic wave by the triaxial acceleration sensor as an example. In this embodiment, the swept-frequency signal source module employs an arbitrary waveform generator (DG4012, supporting a maximum frequency of 100MHz and a sampling rate of 500MSa/s) and a high-voltage amplifier (a high-power amplifier (HAS 4051, supporting a maximum signal frequency of 500KHz) of NF corporation, a high-voltage radio-frequency amplifier (ZHL-100W-GAN +, supporting an amplification frequency of 20-500MHz)) of Coaxial corporation. The out-of-band signal emission module is a high-power loudspeaker, the sensor to be detected is a three-axis acceleration sensor, the sensor data acquisition platform and the microcontroller module are integrated on the raspberry pi 4B, and the display module adopts a high-definition liquid crystal display screen.
Step 1: an excitation signal is generated. An acoustic wave frequency sweep test system on the control host is turned on, and acoustic wave frequency sweep test parameters are set in a system interface, as shown in fig. 2, and include excitation signal types, frequency sweep starting frequencies, frequency sweep cut-off frequencies, the number of equally spaced frequency sweep frequency points, excitation signal amplitude values and same-frequency signal detection duration. And clicking a 'start' button, transmitting a control instruction to a swept-frequency signal source module, namely an arbitrary waveform generator, and enabling the arbitrary waveform generator to enter a swept-frequency mode to generate an original excitation signal with periodic change at equal intervals. The original excitation signal is amplified by a high-voltage amplifier, and the high-voltage amplifier transmits the amplified excitation signal to an out-of-band signal emission module, namely a high-power loudspeaker.
Step 2: out-of-band acoustic signal generation. The out-of-band acoustic signal generation stage mainly means that the signal output end of the high-voltage amplifier is connected with the input end of a loudspeaker, and the loudspeaker converts a periodic sweep frequency excitation signal into a periodic frequency-converted out-of-band acoustic signal.
And step 3: and (3) testing the out-of-band signal, wherein the sound wave signal is mainly propagated through air, so that a non-contact testing mode is adopted in the testing stage, the loudspeaker is close to a sensor module to be tested, namely an anti-shaking module of the camera, and the out-of-band sound wave signal is propagated to a receiving end of the sensor to be tested through air and acts on the MEMS vibration module of the three-axis accelerometer, so that the output abnormality of the sensor is caused, and the platform shakes and generates heat.
And 4, step 4: and collecting sensor data. As shown in fig. 3, the data collection stage includes accessing the three-axis acceleration sensor to be measured to the sensor data collection platform and fixing the sensor data collection platform on the mechanical device, and the data collection platform integrates an analog-to-digital conversion module supporting high input frequency and having high linearity and ultra-low power consumption. Measuring the vibration amplitude of the sensor by using a vibration monitoring device, and connecting the output of the vibration monitoring device with a data acquisition platform; and collecting the real-time working temperature of the sensor by adopting an infrared temperature measuring device, and connecting the output of the infrared temperature measuring device with a data acquisition platform.
After the sensor to be detected is electrified and works, the data acquisition platform adopts a high-speed analog-to-digital converter to convert analog output signals of three shafts of the sensor, analog output signals of the vibration monitoring device and analog output signals of the infrared temperature measuring device into stable digital signals in real time, and the stable digital signals are stored according to time sequence. The acquired data is divided into normal sample data and abnormal sample data, the normal sample data refers to data acquired under the condition that the sensor is in a stable and normal operating state without interference, and both the normal sample data and the abnormal sample data adopt a high-speed analog-to-digital converter to convert an analog output signal of the sensor into a stable digital signal in real time and store the stable digital signal; the abnormal sample data refers to data collected when the sensor output is abnormal due to artificial external interference, such as output abnormality of a short-time vibration accelerometer.
And 5: sensor data preprocessing, as shown in fig. 3, since the triaxial accelerometer is affected by ambient noise in a stationary state, a filtering operation is performed on raw data in a preprocessing stage. Considering that the numerical output of different sensors, the vibration monitoring and the dimension and magnitude of the output signal of the temperature monitoring platform are different, in order to improve the generalization of the detection algorithm, the original data x needs to be subjected to1,x2,x3,x4,x5Carrying out normalization processing to obtain a preprocessed data sample x'1,x′2,x′3,x′4,x′5In this example, the normalization method is used, as shown in the following formula.
And taking 50% of the preprocessed data as a training set and the rest as a test set.
Wherein x is1,x2,x3,x4,x5Respectively representing the x-axis output, the y-axis output, the z-axis output, the vibration amplitude and the working temperature of the acceleration sensor; x ' represents a normalized data set x ' ═ x '1,x′2,x′3,x′4,x′5]And x represents the original data set x ═ x1,x2,x3,x4,x5]And μ denotes a set of mean values μ ═ μ of the raw data1,μ2,μ3,μ4,μ5]And σ denotes a variance set σ of the original data [ σ ]1,σ2,σ3,σ4,σ5]。
Step 6: and in the model training stage, the sensor training set data preprocessed under a certain testing frequency is cut into a plurality of samples at equal intervals, the samples are input into the input end of the machine learning model, the weight is initialized randomly, and the model parameters are optimized by adopting a gradient descent algorithm. The method comprises the following specific steps:
step 6.1: as shown in fig. 4, a neural network model composed of an input layer, a hidden layer and an output layer is built, the vulnerability detection problem is converted into a classification problem, and because each group of data has 5 inputs, the input layer is composed of 5 neural units;
step 6.2: initializing parameters of the neural network model, wherein the weight w of each neuron is randomly generated to be initializednAnd an offset value bnInitialized to random numbers, where wnRepresenting the weight matrix at the nth iteration, bnRepresenting a neural network bias value matrix at the nth iteration; reset activation function step 6.3: pre-training a neural network model by using a training set, firstly, calculating a neural network activation value:
wherein n represents the number of iterations,
an output value representing the hidden layer of the sample at the nth iteration; o isnAnd representing an activation function at the nth iteration to enhance the nonlinearity of the model, and concentrating the hidden layer output value to a probability value in a range of (0,1) to be used as the final output of the neural network model.
Step 6.4: and calculating errors, reversely transmitting the errors of the neural network model according to the loss function values, and optimizing parameters and a learning rate by adopting a gradient descent method. Assuming that the model prediction output value of the sample data is P (where the model output value P of the normal sample data is 1 and the model output value P of the abnormal sample data is 0), the output layer error Err for the nth iteration is calculatednCan be expressed as:
through error reverse transmission, the update weight is:
wherein, beta represents the learning rate,
representing the neuron weights at the nth iteration. When the output result of the proposed neural network is higher than 0.5, the sample to be detected is considered to be normal, namely the sensor to be detected is positioned atThe method has the advantages that the vulnerability does not exist in the out-of-band signals of the current frequency points; otherwise, the sample is abnormal, namely the out-of-band signal of the current frequency point is the out-of-band fragile frequency point of the sensor. The training process is not stopped until the predicted error rate is lower than a certain threshold value, so that a trained network is obtained; bn+1Is the neural network bias value at the n +1 th iteration, bnIs the neural network bias value at the nth iteration.
And 7: and in the detection stage, the trained neural network model is adopted to convert the abnormal detection problem into a classification problem, and the dichotomy is combined to search the optimal vulnerability frequency point, so that the detection efficiency is effectively improved, and the specific process is shown in a flow chart 5.
Step 7.1: first in the wide frequency band [ L, R]Setting m equal interval sweep frequency point numbers, preprocessing the output value of the sensor to be detected obtained in the frequency band range, the vibration amplitude of the equipment installed with the sensor to be detected and the working temperature data according to the method in the step 5, dividing the preprocessed data into m sections at equal intervals according to the number of frequency points, wherein the length of the sample to be detected of each section is consistent with the length of a training sample during training a neural network model, sequentially inputting m sample data to be detected obtained under the m frequency points into the trained neural network model, and obtaining m model outputs O ═1,O2,...,Om-1,Om]And m is an even number.
Step 7.2: the improved binary search method searches weak frequency points. Because the vulnerability frequency point appears in a certain narrow bandwidth range and a plurality of vulnerabilities may exist, workload is increased if frequency sweep detection is carried out on each narrow band frequency point one by one in a full frequency band, and the improved binary search method can effectively reduce frequency sweep times, improve search speed, accurately position the vulnerability and enhance detection performance.
The specific implementation method comprises the following steps: calculating the average value output by the front half part of the model and the average value output by the rear half part of the model, comparing the average value output by the front half part of the model and the average value output by the rear half part of the model with the average value output by the normal sample model, if the average value output by the front half part (or the rear half part) is smaller than the average value output by the normal sample model and the difference value between the two exceeds a detection threshold value, redefining the sweep frequency range in the front half part (or the rear half part), for example, if the average value output by the front half part is smaller than the average value output by the normal sample model, setting the sweep frequency range as [ L (L + R)/2], setting the sweep frequency range as [ (L + R)/2, R ], repeating the steps 7.1 to 7.2 until the sweep frequency range is lower than the threshold value, outputting corresponding fragile frequency points, namely outputting the final sweep frequency range. Note that in the detection process, there may be a plurality of fragile frequency points, so if both the front and rear parts are smaller than the normal sample model output mean value, then the front and rear parts need to be subjected to dichotomy detection.
In this embodiment, through the out-of-band vulnerability detection in step 7, first, the sensor is swept from the frequency range of 1-100KHz, the number of frequency points for equal-interval frequency sweeping is set to be 10, and it is found through a binary search method that the difference value of the output mean value of the model output mean value in the first half frequency division rate range from the output mean value of the normal sample model exceeds the detection threshold value, and the difference value of the output mean value in the second half frequency division rate range from the output mean value of the normal sample model does not exceed the detection threshold value, so that the vulnerability can be determined to be in the range of 1-50 KHz. Similarly, frequency subdivision, frequency sweep and judgment are further carried out on [1KHz and 50KHz ], finally, the vulnerability frequency points of the acoustic wave signals of the accelerometer outside the band are detected to be about 0.8KHz and about 3KHz, the vulnerability about 0.8KHz is caused by resonance of the Y axis, the vulnerability about 3KHz is caused by resonance of the X axis, and therefore the vulnerability type is acoustic wave resonance vulnerability.
And 8: and generating a test report. And after the sweep frequency test is finished, generating a complete test report. The test report content comprises a swept frequency detection object (a three-axis acceleration sensor), a test frequency point range (1Hz-100KHz), a test precision (100Hz), vulnerability frequency points (0.8KHz and 3KHz), and a sensor output value under the vulnerability frequency points (when the frequency point is 0.8KHz, the Y-axis output value of the sensor is 0.19g/s, and when the frequency point is 3KHz, the X-axis output value of the sensor is 0.42 g/s). Through further professional analysis, the test report may further include the types of the bugs of the sensor to be tested, the application field, the test principle, the protection suggestion, and the like, and the specific contents are shown in fig. 6. In the case, aiming at the fact that the vulnerability of the out-of-band acoustic wave signal detected by the acceleration sensor is an acoustic wave resonance leak, the protection suggestion is to adopt an acoustic wave blocking method, namely, a layer of acoustic wave shielding material is added on the periphery of the sensor to shield or attenuate the malicious acoustic wave signal.
Corresponding to the foregoing embodiment of the method for detecting sensor security based on frequency sweep technology, the present application further provides an embodiment of a system for detecting sensor security based on frequency sweep technology, which includes:
and the sweep frequency signal source module generates a periodic excitation electric signal according to the excitation signal type, the sweep frequency starting frequency, the sweep frequency cut-off frequency, the number of equally spaced sweep frequency points, the amplitude of the excitation signal and the detection duration parameter of the same-frequency signal.
And the out-of-band signal transmitting module is used for converting the periodic excitation electric signal into an out-of-band signal to be tested, and applying the out-of-band signal to be tested to the sensor and equipment for mounting the sensor according to the contact type of the out-of-band signal to be tested and the sensor.
And the sensor data acquisition platform acquires output data of the sensor and characteristic data of equipment for mounting the sensor, and preprocesses the data.
And the neural network model training module is used for training the model by using the normal sample data and the abnormal sample data.
And the detection module performs safety detection by using the trained model, firstly performs frequency sweep in a wide frequency band range in the safety detection process, and searches an optimal vulnerability frequency point by combining a frequency sweep result with a dichotomy.
And the display module is used for displaying the detection result.
For the system embodiment, since it basically corresponds to the method embodiment, reference may be made to the partial description of the method embodiment for relevant points. The system embodiments described above are merely illustrative, and the detection modules described herein may or may not be physically separate. In addition, each functional module in the present invention may be integrated into one processing unit, or each module may exist alone physically, or two or more modules may be integrated into one unit. The integrated modules or units can be implemented in the form of hardware, or in the form of software functional units, so that part or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of the application.
The foregoing lists merely illustrate specific embodiments of the invention. It is obvious that the invention is not limited to the above embodiments, but that many variations are possible. All modifications which can be derived or suggested by a person skilled in the art from the disclosure of the present invention are to be considered within the scope of the invention.
Claims (9)
1. A sensor safety detection method based on a frequency sweeping technology is characterized by comprising the following steps:
step one, an excitation signal generation stage: setting excitation signal types, frequency sweep starting frequency, frequency sweep cut-off frequency, the number of equally spaced frequency sweep frequency points, the amplitude of the excitation signal and the detection duration parameter of the same-frequency signal by adopting a digital frequency sweep technology, and outputting a periodic excitation electric signal by an arbitrary waveform generator;
step two, out-of-band signal generation stage: converting the periodic excitation electrical signal into an out-of-band signal to be tested by using a transducer;
step three, out-of-band signal testing stage: according to the contact type of the out-of-band signal to be tested and the sensor, the out-of-band signal to be tested acts on the sensor and equipment for installing the sensor;
step four, a test data collection stage: acquiring output data of a sensor and characteristic data of equipment provided with the sensor, and preprocessing the output data and the characteristic data to be used as normal sample data;
step five, adding interference in the step three, and repeating the step three and the step four to obtain abnormal sample data;
step six, constructing a neural network model, and training the model by using normal sample data and abnormal sample data;
and seventhly, performing safety detection on the sensor to be detected by using the trained model, wherein in the safety detection process, firstly, frequency sweeping is performed in a wide frequency band range, and an optimal vulnerability frequency point is searched by combining a frequency sweeping result with a dichotomy.
2. A frequency sweep technology based sensor security detection method as claimed in claim 1, characterized in that the type of out-of-band signal comprises any of acoustic, optical, electrical, magnetic, thermal, chemical signals.
3. A method as claimed in claim 1, wherein the characteristic data of the sensor-mounted device includes operating temperature and vibration amplitude of the device.
4. A sweep frequency technology-based sensor security detection method as claimed in claim 1, characterized in that the preprocessing process of the data is:
4.1) taking the output data of the sensor and the characteristic data of the equipment provided with the sensor as raw data, and filtering and normalizing the raw data;
and 4.2) dividing the normalized data into a plurality of samples according to the number of the frequency points of the sweep frequency at equal intervals.
5. A swept frequency technology-based sensor safety detection method according to claim 4, wherein the length of the abnormal sample is the same as the length of the normal sample.
6. A swept frequency technology-based sensor security detection method as claimed in claim 1, wherein in the step of six pairs of model training, a probability value in a range of (0,1) is output by the model, and the smaller the probability value, the greater the possibility of sensor abnormality.
7. A swept frequency technology-based sensor safety detection method according to claim 6, wherein the seventh step specifically is:
7.1) firstly setting the number of m equally-spaced frequency points in a broadband range [ L, R ], wherein L is the frequency sweep starting frequency, R is the frequency sweep cut-off frequency, and m is an even number; taking the output value of the sensor to be detected, the vibration amplitude of equipment on which the sensor to be detected is installed and the working temperature data obtained in the frequency band range as original data, and carrying out filtering and normalization pretreatment on the original data;
7.2) dividing the preprocessed data into m sections according to the number of the frequency points of the sweep frequency at equal intervals, wherein the length of a sample to be tested of each section is consistent with the length of a training sample during training of the neural network model, and obtaining m sample data to be tested;
7.3) taking m sample data to be tested as the input of the trained neural network model according to the segmented time sequence to obtain m output results;
7.4) taking the size of the output mean value of the normal sample in the training process as a standard value, respectively calculating the output mean value of the front half part and the output mean value of the rear half part of the model, comparing the calculated values with the standard value, if the mean value of the front half part or the rear half part is smaller than the standard value and the difference value of the two exceeds a detection threshold value, resetting the frequency band range in the front half part or the rear half part by using a bisection method, setting the frequency sweep range of the front half part as [ L (L + R)/2], or setting the frequency sweep range of the rear half part as [ (L + R)/2, R ], repeating the steps 7.1) to 7.4) until the frequency sweep range is lower than the frequency sweep threshold value, and outputting the corresponding fragile frequency point.
8. A sweep frequency technology-based sensor security detection method as claimed in claim 1, characterized by further comprising step eight of generating a detection report according to the detection result; the detection report comprises a frequency sweep detection object, a frequency sweep range, test precision, a vulnerability frequency point and a sensor output value under the vulnerability frequency point.
9. A sensor safety detection system based on a frequency sweep technology is characterized by being used for realizing the sensor safety detection method of claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111394732.XA CN114121305B (en) | 2021-11-23 | 2021-11-23 | Sensor safety detection method and system based on sweep frequency technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111394732.XA CN114121305B (en) | 2021-11-23 | 2021-11-23 | Sensor safety detection method and system based on sweep frequency technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114121305A true CN114121305A (en) | 2022-03-01 |
| CN114121305B CN114121305B (en) | 2024-02-23 |
Family
ID=80440248
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111394732.XA Active CN114121305B (en) | 2021-11-23 | 2021-11-23 | Sensor safety detection method and system based on sweep frequency technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114121305B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115021971A (en) * | 2022-05-10 | 2022-09-06 | 电子科技大学 | Hardware security testing method for FPGA cloud platform |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104374468A (en) * | 2014-10-24 | 2015-02-25 | 哈尔滨工业大学深圳研究生院 | Intrinsic frequency measuring method for long and thin object |
| CN105136281A (en) * | 2015-10-14 | 2015-12-09 | 基康仪器股份有限公司 | Vibrating wire transducer frequency measurement device and method |
| US20170356739A1 (en) * | 2016-06-08 | 2017-12-14 | Zygo Corporation | Precision positioning system using a wavelength tunable laser |
| CN107525849A (en) * | 2017-08-25 | 2017-12-29 | 北京航空航天大学 | A kind of single-input single-output test modal analysis system and method based on fiber grating |
| US20180014217A1 (en) * | 2013-03-15 | 2018-01-11 | DGS Global Systems, Inc. | Systems, methods, and devices for automatic signal detection with temporal feature extraction within a spectrum |
| CN111157099A (en) * | 2020-01-02 | 2020-05-15 | 河海大学常州校区 | Distributed optical fiber sensor vibration signal classification method and identification classification system |
| CN111325095A (en) * | 2020-01-19 | 2020-06-23 | 西安科技大学 | Intelligent equipment health state detection method and system based on sound wave signals |
-
2021
- 2021-11-23 CN CN202111394732.XA patent/CN114121305B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180014217A1 (en) * | 2013-03-15 | 2018-01-11 | DGS Global Systems, Inc. | Systems, methods, and devices for automatic signal detection with temporal feature extraction within a spectrum |
| CN104374468A (en) * | 2014-10-24 | 2015-02-25 | 哈尔滨工业大学深圳研究生院 | Intrinsic frequency measuring method for long and thin object |
| CN105136281A (en) * | 2015-10-14 | 2015-12-09 | 基康仪器股份有限公司 | Vibrating wire transducer frequency measurement device and method |
| US20170356739A1 (en) * | 2016-06-08 | 2017-12-14 | Zygo Corporation | Precision positioning system using a wavelength tunable laser |
| CN107525849A (en) * | 2017-08-25 | 2017-12-29 | 北京航空航天大学 | A kind of single-input single-output test modal analysis system and method based on fiber grating |
| CN111157099A (en) * | 2020-01-02 | 2020-05-15 | 河海大学常州校区 | Distributed optical fiber sensor vibration signal classification method and identification classification system |
| CN111325095A (en) * | 2020-01-19 | 2020-06-23 | 西安科技大学 | Intelligent equipment health state detection method and system based on sound wave signals |
Non-Patent Citations (2)
| Title |
|---|
| ILIAS GIECHASKIEL等: "Taxonomy and Challenges of Out-of-Band Signal Injection Attacks and Defenses", 《IEEE COMMUNICATIONS SURVEYS & TUTORIALS》 * |
| 邰寒松: "正弦动态压力校准中扫频速率对校准结果的影响研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115021971A (en) * | 2022-05-10 | 2022-09-06 | 电子科技大学 | Hardware security testing method for FPGA cloud platform |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114121305B (en) | 2024-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | A high-efficiency multiple events discrimination method in optical fiber perimeter security system | |
| US20180357542A1 (en) | 1D-CNN-Based Distributed Optical Fiber Sensing Signal Feature Learning and Classification Method | |
| CN105261136B (en) | The method and device of weather interference is shielded in a kind of fiber-optic monitoring warning system | |
| CN103116957B (en) | A kind of method of optical fiber perimeter security and protection system shielding climate impact | |
| Chen et al. | A real-time distributed deep learning approach for intelligent event recognition in long distance pipeline monitoring with DOFS | |
| CN106404148A (en) | Portable micro vibration testing device and method | |
| CN110824550B (en) | Tunnel bad geologic body microseismic advanced detection system and method | |
| CN114121305B (en) | Sensor safety detection method and system based on sweep frequency technology | |
| CN102288882A (en) | Partial discharge monitoring system and method for multichannel electrical equipment | |
| CN115265750A (en) | Optical fiber distributed acoustic wave sensing system and method | |
| CN109827652A (en) | One kind being directed to Fibre Optical Sensor vibration signal recognition and system | |
| CN106767960A (en) | Fibre Optical Sensor Dynamic Signal demodulation method based on spectrum simulation Yu difference algorithm | |
| CN118004704B (en) | DAS-based belt conveyor carrier roller fault monitoring method in complex sound field | |
| WO2016182183A1 (en) | Aircraft noise analysis method based on natural cutoff frequency | |
| EP1188075A1 (en) | Transient monitoring system and method | |
| CN103985213B (en) | A kind of configurable continuous-flow type optical fiber disturbance alarm detection procedure and detector thereof | |
| CN207007993U (en) | A kind of measuring device with electricity of partial discharge of transformer | |
| CN105737965A (en) | Aerogenerator vibration detection device and analysis method | |
| CN116026922A (en) | Acoustic emission signal acquisition device and positioning method in rock mechanical test | |
| CN110412407A (en) | Failure cable essence distance measuring method based on sounding analytic approach | |
| CN113758508A (en) | Transmission line environment conflagration and external damage monitoring devices | |
| CN219038172U (en) | Integrated cable and fence detection system | |
| CN108344493A (en) | A kind of measuring apparatus and measurement method of phase generated carrier optical fiber sensing system signal demodulation factor | |
| Puranto et al. | Development of microcontroller-based acquisition and processing unit for fiber optic vibration sensor | |
| CN105527552B (en) | The detection method of taiwan area discharge fault detection device based on acoustooptic technique |
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 |