A touch screen vulnerability analysis and protection method and system

Abstract

The present invention discloses a touch screen vulnerability analysis and protection method and system, which relates to the field of electronic terminal technology. It includes: preliminarily judging whether the touch screen has vulnerability and determining the vulnerable frequency point; obtaining the touch screen drive signal, and designing the interference signal based on the drive signal to realize the false point injection; collecting the abnormal sample data of the false touch point generated by the interference signal and the normal sample data generated by the normal touch; pre-processing the sample data; establishing and training the reporting point enhancement model based on machine learning, collecting the contact point related data of the device to be tested in real time, and the touch screen does not respond to the contact points identified as abnormal, so as to realize the vulnerability protection. The present invention analyzes the vulnerability of the capacitive touch screen from the two aspects of conducted interference and radiated interference, and realizes the vulnerability defense in the software by using the reporting point enhancement model, and realizes the vulnerability defense in the hardware by combining the filtering module, which can effectively improve the anti-interference ability of the touch screen and reduce the false alarm rate of the touch screen.

Description

Technical Field

[0001] The invention belongs to the field of electronic terminals, and in particular relates to a touch screen vulnerability analysis and protection method and system. Background Art

[0002] With the continuous development of electronic technology, touch screens, as an important human-computer interaction interface, have been widely used in various smart devices, such as smart phones, tablets and laptops. Among them, mutual capacitive touch screens have become the mainstream touch screens on the market with their advantages of low power consumption, precise touch, and support for multi-touch. Mutual capacitive touch screens usually consist of two groups of cross-arranged signal lines to form a mutual capacitance array: drive lines and sensing lines. The change in mutual capacitance is detected to determine whether there is a touch point. The drive line is responsible for transmitting high-frequency drive signals, and the sensing line is used to detect changes in capacitance. When the conductor is close, a coupling capacitor is formed between the conductor and the touch screen, which destroys the original electrostatic balance and changes the capacitance value detected by the sensing line.

[0003] With the continuous advancement of touch technology, capacitive touch screens have also derived a variety of driving strategies, such as sequential driving and parallel driving according to the form of driving signals. While improving performance advantages, capacitive touch screens also bring many troubles and security risks to users. For example, when smart devices are connected to chargers, they will be interfered by common-mode signals and deviate from the actual touch point. Attackers can even remotely operate the victim's mobile phone by constructing sequential common-mode interference signals or electromagnetic interference to perform malicious operations such as transferring money and implanting viruses, posing a huge threat to the safe operation of mobile terminals. Although touch devices currently perform common-mode interference detection before leaving the factory, due to the single form of the detection signal and the small amplitude of the interference signal, it is impossible to fully explore the vulnerability of the touch screen, and the defects in the touch screen reporting algorithm also provide an attack entry for attackers to perform malicious operations.

[0004] In summary, building a systematic capacitive touch screen vulnerability analysis framework and formulating corresponding reporting point enhancement strategies are of great significance to the security research and protection of mobile electronic devices. Summary of the invention

[0005] The purpose of the present invention is to provide a vulnerability analysis method for mutual capacitive touch screens and propose corresponding protection strategies, aiming to solve the problem that existing touch screens have software and hardware vulnerabilities and lack systematic hardware interference detection and reporting point enhancement strategies, which poses a safety hazard in actual application scenarios.

[0006] The present invention adopts the following technical solutions to solve the above problems:

[0007] In a first aspect, the present invention proposes a touch screen vulnerability analysis and protection method, comprising the following steps:

[0008] Step 1: Construct a frequency sweep signal to perform interference detection on the test equipment, and preliminarily judge the vulnerability through the working status of the touch screen. If there is vulnerability, record the vulnerable frequency point and vulnerability type, and execute the subsequent steps, otherwise stop the detection;

[0009] Step 2, collecting the driving signal of the touch screen of the test device with the fragile frequency point, and determining its driving mode;

[0010] Step 3, constructing an interference signal according to the driving mode and fragile type of the touch screen of the test device, taking the contact point related data generated by the interference signal as abnormal sample data, and taking the contact point related data generated by normal touch as normal sample data;

[0011] Step 4: Build a machine learning model of multiple biometric features and train it using abnormal sample data and normal sample data;

[0012] Step 5: Place the trained multi-biometric machine learning model into the test device as a reporting point enhancement model to collect contact-related data of the device to be tested in real time. The touch screen does not respond to abnormal contacts to achieve vulnerability protection.

[0013] Furthermore, the fragile frequency points described in step 1 include a fragile frequency point 1 for conduction and a fragile frequency point 2 for radiation; when either the fragile frequency point 1 or the fragile frequency point 2 exists, it is preliminarily determined that the test equipment is vulnerable.

[0014] Furthermore, when the test equipment has a fragile frequency point 1 for conduction, a common-mode signal filtering module for the fragile frequency point 1 is installed in the test equipment; when the test equipment has a fragile frequency point 2 for radiation, a differential-mode signal filtering module for the fragile frequency point 2 is installed in the test equipment.

[0015] Furthermore, the method for detecting the fragile frequency point 1 is:

[0016] The arbitrary waveform generator is connected to a high-voltage amplifier or a radio frequency amplifier, and the amplified sweep signal is used as a vulnerability detection signal source. The sweep signal source based on the conduction vulnerability detection and the sweep signal source based on the radiation vulnerability detection are obtained by setting and adjusting the sweep signal parameters.

[0017] The positive pole of the output terminal of the sweep signal source based on the conduction vulnerability detection is connected to the ground terminal of the data line connected to the test equipment, and the negative pole of the output terminal is suspended, so as to inject the common mode interference signal into the test equipment; in the interference injection stage, the touch screen is touched to observe whether the contact track is abnormal to determine whether there is conduction vulnerability, and the frequency point with the strongest interference is defined as the fragile frequency point 1;

[0018] Connect the two ends of the dipole antenna to the positive and negative output terminals of the sweep signal source based on radiation vulnerability detection, and place the antenna directly above the touch screen of the test device to inject radiation interference signals into the test device; during the interference injection phase, perform touch operations on the touch screen to observe whether the contact point trajectory is abnormal to determine whether there is radiation vulnerability, and define the frequency point with the strongest interference as the second vulnerable frequency point;

[0019] The abnormality refers to a touch abnormality or a sliding track deviating from an actual position.

[0020] Furthermore, the driving mode of the touch screen of the testing device includes sequential driving and parallel driving.

[0021] Furthermore, the interference signal is constructed according to the driving mode of the touch screen of the test device, specifically:

[0022] If the driving mode of the touch screen of the test equipment is sequential driving, the driving signals on different driving electrodes are distributed sequentially over time, and the high-frequency signal inside each driving signal is the same. By designing the signal period and duration and adjusting the delay, a timing interference signal is constructed;

[0023] If the driving mode of the touch screen of the test equipment is parallel driving, the driving signals of the touch screen are generated simultaneously, and the driving signals on different electrodes are code-division multiplexed. Therefore, the signal replay method is used to construct a coded interference signal according to the coding form of the driving signal.

[0024] Furthermore, the abnormal sample data is obtained by injecting an interference signal into a designated driving electrode position and collecting data related to a false touch point as the abnormal sample data.

[0025] Furthermore, the contact-related data include the capacitance change at the contact position, the average capacitance change near the contact, the pressure data at the contact position, the duration of the capacitance change at the contact position, and the time interval between two false points or touch points.

[0026] Furthermore, the multi-biometric machine learning model is a neural network model consisting of three parts: an input layer, a hidden layer, and an output layer.

[0027] In a second aspect, the present invention proposes a touch screen vulnerability analysis and protection system for implementing the above touch screen vulnerability analysis and protection method.

[0028] The vulnerability analysis and protection method for touch screens proposed by the present invention is comprehensive, easy to operate, has strong generalization ability, and reliable detection results. The proposed reporting point enhancement strategy reduces the potential threats to the touch screen. The main beneficial effects include:

[0029] (1) The present invention analyzes the hardware-based anti-interference vulnerability of capacitive touch screens from the perspectives of conducted interference and radiated interference, and fully exploits the potential vulnerability of the touch screen's software-based reporting algorithm by designing interference detection signals. Compared with the simple common-mode interference test in the factory, the present invention provides a more systematic and comprehensive vulnerability analysis and anti-interference test signal construction guidance for touch screen security research, rather than directly injecting a single AC test signal, which is of great significance for ensuring the safety of human-computer interaction in electronic terminals.

[0030] (2) The present invention proposes a neural network-based point reporting enhancement strategy to explore the software vulnerability of capacitive touch screens. Compared with the traditional method of judging whether there is a touch point based only on the capacitance threshold, the present invention combines multiple touch operation features, especially adding live click features based on pressure, touch time, etc., to make the touch screen point reporting more accurate and reliable.

[0031] (3) In the hardware vulnerability mining process of the present invention, an efficient digital scanning strategy is adopted to provide a stable interference signal, which effectively improves the electromagnetic interference detection efficiency of the touch screen; a contact detection model based on a machine learning algorithm is introduced to transform the contact judgment task into a classification problem, which further improves the detection accuracy and speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a schematic diagram of a typical capacitive touch screen structure according to an exemplary embodiment;

[0033] Figure 2 is a schematic diagram of a driving signal based on sequential driving according to an exemplary embodiment;

[0034] Figure 3 is a schematic diagram of a driving signal based on parallel driving according to an exemplary embodiment;

[0035] Figure 4 is a schematic diagram of a workflow based on multi-feature reporting according to an exemplary embodiment;

[0036] Figure 5 is a schematic diagram of driving signal acquisition according to an exemplary embodiment;

[0037] Figure 6 is a schematic diagram of contact point detection based on a machine learning algorithm according to an exemplary embodiment;

[0038] Figure 7 It is a schematic diagram showing the composition of a system implementation module according to an exemplary embodiment. DETAILED DESCRIPTION

[0039] The present invention is further described below in conjunction with the accompanying drawings and embodiments. The accompanying drawings are only schematic diagrams of the present invention, and some of the block diagrams shown in the accompanying drawings are functional entities, which do not necessarily correspond to physically or logically independent entities, and these functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor systems and / or microcontroller systems.

[0040] The present invention proposes a vulnerability analysis and protection method for capacitive touch screens, which conducts non-invasive and systematic security analysis on capacitive touch screens without disassembling and modifying the hardware. It mines the vulnerability of the device's software and hardware from the two interference signal propagation modes of conduction and radiation, and uses a frequency scanning method to increase the mining rate. In addition, a vulnerability detection signal construction method based on the touch screen driving method is proposed, which overcomes the shortcomings of traditional detection that only uses a single form of continuous noise and lacks targeted interference signal construction, thereby failing to fully mine its vulnerabilities. Based on the vulnerability, the present invention proposes multi-biological touch features, combines machine learning classification algorithms, and constructs a touch reporting enhancement model based on multi-biological features to achieve automated touch detection and improve the anti-interference ability and reporting accuracy of the touch screen.

[0041] like Figure 1 As shown, taking a capacitive touch screen arranged in a grid as an example, a specific implementation scheme of the vulnerability analysis and protection method of the touch screen of the present invention is introduced, and the implementation flow chart is shown as follows Figure 4 As shown in Figure 2, the system structure consists of six modules: Figure 7 shown.

[0042] In this embodiment, the devices used include an arbitrary waveform generator (DG4012, supporting a maximum frequency of 100 MHz and a sampling rate of 500 MSa / s) from iPuyuan, a high-voltage amplifier (NF's high-power amplifier (HAS 4051, supporting a maximum signal frequency of 500 KHz), Coaxial's high-voltage radio frequency amplifier (ZHL-100W-GAN+, supporting an amplification frequency of 20-500 MHz) and a signal oscilloscope (RIGOL MSO8104). The transmission carrier of the common-mode interference signal is a standard data line, the transmission carrier of the radiation signal is a homemade antenna, and the touch screen device to be tested is a commercial mobile phone Google Nexus 5X. The driving signal is collected through copper sheets and wires, and data such as the touch path and capacitance are obtained through the developer mode of the mobile phone.

[0043] Combination Figure 4 The touch screen vulnerability analysis and protection method in this embodiment includes the following steps:

[0044] Step 1: Vulnerability detection.

[0045] Construct a frequency sweep signal, perform two interference tests, conducted and radiated, on the vulnerability detection object, respectively, observe the working state of the touch screen to determine whether the corresponding vulnerability exists, and determine the vulnerable frequency point 1 based on conducted interference and the vulnerable frequency point 2 based on radiated interference. Based on the above analysis, step 1 specifically includes the following steps.

[0046] Step 1.1, construction of frequency sweep signal source.

[0047] Turn on the arbitrary waveform generator and adjust the sweep signal parameters, including the interference signal type T, the sweep start frequency fs, the sweep cutoff frequency fe, the number of equally spaced sweep frequency points N, the interference signal amplitude I, and the same-frequency signal detection time t. Connect the arbitrary waveform generator to the high-voltage amplifier or the radio frequency amplifier, adjust the amplification factor A, and the amplified sweep signal is the vulnerability detection signal source. Before the test, the touch display function in the developer mode of the test equipment should be turned on first. For example, in this embodiment, the sweep signal source for the conduction vulnerability detection is set as follows: the signal type is a sine signal, fs = 0kHz, the sweep cutoff frequency fe = 500kHz, the number of equally spaced sweep frequency points N = 100, the interference signal amplitude I = 3V, the same-frequency signal detection time t = 2s, and the amplification factor A = 100. The settings for the sweep signal source for radiation vulnerability detection are: signal type is sinusoidal signal, fs=1MHz, sweep cutoff frequency fe=100MHz, number of equally spaced sweep frequency points N=1000, interference signal amplitude I=1V, and co-frequency signal detection time t=2s. Since the RF amplifier automatically amplifies the signal, there is no need to manually adjust the amplification factor.

[0048] Step 1.2, conduction-oriented vulnerability detection.

[0049] This step is used to detect whether the touch screen has conductive vulnerability. The positive output terminal of the sweep signal source based on conductive vulnerability detection is connected to the ground terminal of the data line connected to the test equipment, and the negative output terminal is left floating, so as to inject a common-mode interference signal into the test equipment. During the interference injection stage, the touch screen is touched to observe whether the contact track is abnormal to determine whether there is conductive vulnerability, and the frequency point with the strongest interference is defined as the fragile frequency point one. In this embodiment, it is found that the touch track of the click and slide operations of the Nexus 5X deviates near a certain frequency band, so it is determined that the device has conductive vulnerability based on common-mode interference. Through finer-grained sweep operations and comparison of the degree of deviation of the touch track, it is determined that the fragile frequency point one of the Nexus 5X is 285kHz.

[0050] Step 1.3, radiation-oriented vulnerability detection.

[0051] This step is used to detect whether the touch screen has conductive vulnerability. Since the capacitive touch screen is composed of an electrode array, there may be a risk of electromagnetic coupling, so that the interference signal radiated in the air is converted into an electrical signal detected by the device itself. Connect the two ends of the homemade dipole antenna to the positive and negative output terminals of the sweep signal source based on radiation vulnerability detection, and place the antenna directly above the touch screen to inject the radiation interference signal into the test device. In the interference injection stage, the touch screen is touched to observe whether the contact track is abnormal to determine whether there is radiation vulnerability, and the frequency point with the strongest interference is determined as the fragile frequency point 2. In this embodiment, the length of the homemade dipole antenna is 15 cm, and the antenna is placed 1 cm above the touch screen. It is found that the touch track of the click and slide operation of the Nexus 5X deviates near a certain frequency band, so it is determined that the device has radiation vulnerability based on electromagnetic interference. Through a more fine-grained sweep operation and comparing the degree of deviation of the touch track, it is determined that the fragile frequency point 2 of the Nexus 5X is 46MHz.

[0052] Step 2: Construct targeted interference signals.

[0053] Since the vulnerability detection in step 1 only roughly determines whether the test device has a vulnerability when clicked by a human hand, and does not construct a specific detection signal based on the characteristics of the touch screen, this universal interference will be shielded by the pocket anti-mistouch mode of the touch screen, and no contact will be generated when there is no human hand touching it. In addition, factories usually have a certain interference tolerance for equipment, but still release vulnerable devices. The present invention will conduct more in-depth vulnerability mining based on device characteristics, construct detection signals from the perspective of malicious injection of false points by attackers, avoid conventional interference detection methods, more fully reveal the software and hardware vulnerabilities of the device, and provide technical references for subsequent reporting strategies. Based on different touch screen drive signal forms, step 2 specifically includes the following steps.

[0054] Step 2.1, drive signal acquisition and drive mode determination.

[0055] like Figure 5 As shown, two square copper sheets with a side length of 1 cm are welded on two wires to form a receiving antenna array. The copper sheet end is placed on the screen of a mobile phone in working state, and the wire end is connected to different channels of the positive input end of the oscilloscope. The negative input end of the oscilloscope is connected to the ground end of the mobile phone. In this embodiment, the negative input end of the oscilloscope is directly connected to the metal head end of the data line connected to the mobile phone. In this embodiment, it can be judged from the form of the dual-channel driving signal that Nexus 5X adopts a sequential driving method, the scanning period is 120Hz, and the frequency of the internal driving signal is 285kHz.

[0056] Step 2.2, construction of interference signal for sequential driving device.

[0057] like Figure 2 As shown, the driving signals on different driving electrodes are distributed sequentially over time, and the high-frequency signal inside each driving signal is the same. Therefore, the detection signal timing (signal period, duration) in step 1 can be designed according to the form of the driving signal, and the timing interference signal can be constructed by adjusting the delay, so as to inject the false point into the specified driving electrode position. In this embodiment, the interference signal period is 8.354m, and the high-frequency signal duration is 3ms.

[0058] Step 2.3, construction of interference signals for parallel driving devices.

[0059] like Figure 3 As shown in FIG. 1 , in a parallel driving device, the driving signals of the touch screen are generated simultaneously, and the driving signals on different electrodes are usually multiplexed using code division. Therefore, using the idea of ​​signal replay, the detection signal in step 1 is encoded according to the encoding form of the driving signal to construct a coded interference signal, thereby achieving the injection of false points on the specified electrode.

[0060] Step 3: Sample data collection.

[0061] Collect the relevant data of false touch points generated by the interference signal in step 2 as abnormal sample data, collect the relevant data of human hand clicks as normal sample data, and use these data as original sample data, recorded as x1, x2, x3, x4, x5. The specific steps include:

[0062] Step 3.1, abnormal sample data collection.

[0063] Construct the interference signal in step 2 according to the form of the touch screen drive signal. For devices with conductive vulnerability, use wired injection to inject interference signals into the ground wire of the test device to generate false touch points. For devices with radiation vulnerability, use antennas to emit radiation interference to generate false touch points. And record the capacitance change at the false point position, the average capacitance change near the false point, the pressure data at the false point position, the duration of the capacitance change at the false point position, and the time interval between the two false points, and use them as abnormal samples.

[0064] Step 3.2, normal sample data collection.

[0065] Without injecting false points, perform normal click operations on the touch screen, and record the change in capacitance at the touch point, the average change in capacitance near the touch point, the pressure data at the touch point, the duration of the capacitance change at the touch point, and the time interval between the two touch points as normal samples.

[0066] Step 4: Sample data processing.

[0067] This step is to overcome the dimension and order of magnitude differences of the sample data and improve the generalization of the detection algorithm. The sample data is normalized and the processed data is used as the original data for subsequent model training. In this case, the original data x1, x2, x3, x4, x5 are normalized to obtain the preprocessed data samples x′1, x′2, x′3, x′4, x′5. The standardization method is used in this embodiment as shown in the following formula. 50% of the preprocessed data is used as the training set, and the rest is used as the test set.

[0068]

[0069] Among them, x′ represents the normalized data set x′=[x′1,x′2,x′3,x′4,x′5], x represents the original data set x=[x1,x2,x3,x4,x5], μ represents the mean set of the original data μ=[μ1,μ2,μ3,μ4,μ5], and σ represents the variance set of the original data σ=[σ1,σ2,σ3,σ4,σ5].

[0070] Step 5: Establish and train the reporting point enhancement model.

[0071] Build a machine learning model with multiple biometric features, divide the preprocessed sample data into several samples at equal intervals, input them into the input of the machine learning model, initialize the weights randomly, and use the gradient descent algorithm to optimize the model parameters. The specific steps are:

[0072] Step 5.1: Model construction, such as Figure 6 As shown in the figure, a neural network model consisting of an input layer, a hidden layer, and an output layer is built to transform the touch point detection problem into a classification problem. Since each set of data has 5 inputs, the input layer consists of 5 neural units.

[0073] Step 5.2: Parameter initialization. Initialize the parameters of the neural network model by randomly generating initial weights and setting the weight w of each neuron n and the bias value b n Initialized to a random number, where w n represents the weight matrix at the nth iteration, b n Represents the neural network bias value matrix at the nth iteration;

[0074] Step 5.3: Model pre-training. Use the training set to pre-train the neural network model. First, calculate the neural network activation value:

[0075]

[0076]

[0077] Where n represents the number of iterations, Represents the output value of the hidden layer of the sample at the nth iteration; O n Represents the activation function at the nth iteration to enhance the nonlinearity of the model, concentrate the hidden layer output value to a probability value in the range of (0,1), and use it as the final output of the neural network model.

[0078] Step 5.4: Error calculation. According to the loss function value, the error of the neural network model is reversely transmitted, and the gradient descent method is used to optimize the parameters and learning rate. Let the model prediction output value of the sample data be P (where the model output value of normal sample data is P = 1, and the model output value of abnormal sample data is P = 0). For the output layer error Err of the nth iteration n It can be expressed as:

[0079]

[0080] Through error back propagation, the updated weight is:

[0081]

[0082]

[0083] Among them, β represents the learning rate, represents the neuron weight at the nth iteration. When the output result of the proposed neural network is higher than 0.5, it is considered that there is a touch operation and the reporting information is transmitted to the processor; otherwise, there is no touch operation and no reporting is performed. The training process stops when the predicted error rate is lower than a certain threshold, thus obtaining a trained network; b n+1 is the bias value of the neural network at the n+1th iteration, b n is the bias value of the neural network at the nth iteration.

[0084] Step 6, report point enhancement strategy.

[0085] This step proposes a protection strategy based on detection algorithm and hardware anti-interference. In this embodiment, for Nexus 5X, due to the sequential drive mode adopted by this mobile phone, there are vulnerabilities in both conduction and radiation. For the detection algorithm, the protection strategy of the present invention is: combining multi-touch feature data, such as adding pressure data, the duration of a single touch, and the time interval between two clicks to resist false points caused by malicious electromagnetic interference, and using a machine learning algorithm for touch point detection, which can effectively improve the efficiency and accuracy of reporting points. For example, the trained machine learning model is placed in the test device as a reporting point enhancement model, and the contact point related data of the device to be tested is collected in real time. The touch screen does not respond to the abnormal contact point, that is, the false point is automatically shielded. In terms of hardware, when the test device has a fragile frequency point 1 for conduction, a common-mode signal filtering module for fragile frequency point 1 is installed in the test device; when the test device has a fragile frequency point 2 for radiation, a differential-mode signal filtering module for fragile frequency point 2 is installed in the test device. In this embodiment, the filter module is installed inside the mobile phone charging port to reduce the common-mode interference of wired conduction and the differential-mode interference of radiation.

[0086] In this embodiment, a vulnerability analysis and protection system for a touch screen is also provided, which is used to implement the above-mentioned embodiment, and the descriptions that have been made will not be repeated. The terms "module", "unit", etc. used below can be a combination of software and / or hardware that implements a predetermined function. Although the system described in the following embodiments is preferably implemented in software, it is also possible to implement hardware, or a combination of software and hardware.

[0087] The system comprises:

[0088] A vulnerability detection module is used to construct a frequency sweep signal, perform interference detection on the test equipment, and preliminarily judge the vulnerability through the working status of the touch screen. If there is vulnerability, the vulnerable frequency point and vulnerability type are recorded;

[0089] An interference signal construction module, which is used to determine the driving mode and construct an interference signal according to the driving mode of the touch screen of the test device;

[0090] A data collection module, which is used to collect contact point related data generated by interference signals as abnormal sample data, and collect contact point related data generated by normal touch as normal sample data;

[0091] A data processing module, which is used to normalize normal sample data and abnormal sample data;

[0092] A model building and training module, which is used to build a machine learning model of multiple biological features and train it using normalized abnormal sample data and normal sample data;

[0093] The reporting point enhancement strategy module is used to place the trained machine learning model into the test equipment as a reporting point enhancement model, collect the contact point related data of the equipment to be tested in real time, and identify the touch points that the touch screen does not respond to as abnormal.

[0094] For the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The system embodiment described above is only schematic, wherein the vulnerability detection module may or may not be physically separated. In addition, the functional modules in the present invention may be integrated into a processing unit, or each module may exist physically separately, or two or more modules may be integrated into one unit. The above-mentioned integrated modules or units can be implemented in the form of hardware or in the form of software functional units, so as to select some or all of the modules according to actual needs to achieve the purpose of the present application.

[0095] The above examples are only specific embodiments of the present invention. Obviously, the present invention is not limited to the above examples, and many variations are possible. All variations that can be directly derived or associated with the contents disclosed by a person skilled in the art should be considered as the protection scope of the present invention.

Claims

1. A touch screen vulnerability analysis and protection method, characterized in that: The following steps are involved: Step 1, construct a sweep signal, perform interference detection on the test equipment, preliminarily judge the vulnerability through the working state of the touch screen, if there is vulnerability, record the vulnerable frequency point and vulnerable type, and execute the subsequent steps, otherwise stop the detection; the vulnerable types include conducted interference and radiated interference; Step 2, collecting the driving signal of the touch screen of the test device with the fragile frequency point, and determining its driving mode; the driving mode includes sequential driving and parallel driving; Step 3, constructing an interference signal according to the driving mode and fragile type of the touch screen of the test device, taking the contact point related data generated by the interference signal as abnormal sample data, and taking the contact point related data generated by normal touch as normal sample data; Step 4: Build a machine learning model of multiple biometric features and train it using abnormal sample data and normal sample data; Step 5: Place the trained multi-biometric machine learning model into the test device as a reporting point enhancement model to collect contact-related data of the device to be tested in real time. The touch screen does not respond to abnormal contacts to achieve vulnerability protection.

2. A touch screen vulnerability analysis and protection method according to claim 1, characterized in that: The fragile frequency points described in step 1 include a first fragile frequency point for conduction and a second fragile frequency point for radiation; when either the first fragile frequency point or the second fragile frequency point exists, it is preliminarily determined that the test equipment is vulnerable.

3. A touch screen vulnerability analysis and protection method according to claim 2, characterized in that: When the test equipment has a fragile frequency point 1 for conduction, a common-mode signal filtering module for the fragile frequency point 1 is installed in the test equipment; when the test equipment has a fragile frequency point 2 for radiation, a differential-mode signal filtering module for the fragile frequency point 2 is installed in the test equipment.

4. A touch screen vulnerability analysis and protection method according to claim 2, characterized in that: The method for detecting the fragile frequency point 1 is: The arbitrary waveform generator is connected to a high-voltage amplifier or a radio frequency amplifier, and the amplified sweep signal is used as a vulnerability detection signal source. The sweep signal source based on the conduction vulnerability detection and the sweep signal source based on the radiation vulnerability detection are obtained by setting and adjusting the sweep signal parameters. The positive pole of the output terminal of the sweep signal source based on the conduction vulnerability detection is connected to the ground terminal of the data line connected to the test equipment, and the negative pole of the output terminal is suspended, so as to inject the common mode interference signal into the test equipment; in the interference injection stage, the touch screen is touched to observe whether the contact track is abnormal to determine whether there is conduction vulnerability, and the frequency point with the strongest interference is defined as the fragile frequency point 1; Connect the two ends of the dipole antenna to the positive and negative output terminals of the sweep signal source based on radiation vulnerability detection, and place the antenna directly above the touch screen of the test device to inject radiation interference signals into the test device; during the interference injection phase, perform touch operations on the touch screen to observe whether the contact point trajectory is abnormal to determine whether there is radiation vulnerability, and define the frequency point with the strongest interference as the second vulnerable frequency point; The abnormality refers to a touch abnormality or a sliding track deviating from an actual position.

5. A touch screen vulnerability analysis and protection method according to claim 4, characterized in that: The interference signal is constructed according to the driving mode of the touch screen of the test device, specifically: If the driving mode of the touch screen of the test equipment is sequential driving, the driving signals on different driving electrodes are distributed sequentially over time, and the high-frequency signal inside each driving signal is the same. By designing the signal period and duration and adjusting the delay, a timing interference signal is constructed; If the driving mode of the touch screen of the test equipment is parallel driving, the driving signals of the touch screen are generated simultaneously, and the driving signals on different electrodes are code-division multiplexed. Therefore, the signal replay method is used to construct a coded interference signal according to the coding form of the driving signal.

6. A touch screen vulnerability analysis and protection method according to claim 1 or 5, characterized in that: The abnormal sample data is obtained by injecting an interference signal into a designated driving electrode position and collecting data related to a false touch point as the abnormal sample data.

7. A touch screen vulnerability analysis and protection method according to claim 1, characterized in that: The contact-related data include the capacitance change at the contact position, the average capacitance change near the contact, the pressure data at the contact position, the duration of the capacitance change at the contact position, and the time interval between two false points or touch points.

8. A touch screen vulnerability analysis and protection method according to claim 1, characterized in that: The multi-biometric machine learning model is a neural network model consisting of three parts: an input layer, a hidden layer, and an output layer.

9. A touch screen vulnerability analysis and protection system, used to implement the touch screen vulnerability analysis and protection method according to claim 1, characterized in that: The system comprises: A vulnerability detection module is used to construct a frequency sweep signal, perform interference detection on the test equipment, and preliminarily judge the vulnerability through the working status of the touch screen. If there is vulnerability, the vulnerable frequency point and vulnerability type are recorded; An interference signal construction module, which is used to determine the driving mode and construct an interference signal according to the driving mode of the touch screen of the test device; A data collection module, which is used to collect contact point related data generated by interference signals as abnormal sample data, and collect contact point related data generated by normal touch as normal sample data; A data processing module, which is used to normalize normal sample data and abnormal sample data; A model building and training module, which is used to build a machine learning model of multiple biological features and train it using normalized abnormal sample data and normal sample data; The reporting point enhancement strategy module is used to place the trained machine learning model into the test equipment as a reporting point enhancement model, collect the contact point related data of the equipment to be tested in real time, and identify the touch points that the touch screen does not respond to as abnormal.

Images