Apparatus, method of interfering with a wearable photoplethysmography device

Abstract

The application relates to the field of mobile and wireless security technology, and discloses a device and a method for interfering with a wearable photoplethysmography device, the device comprising: an infrared light source configured to emit a modulated infrared light beam; and a control unit configured to generate a modulation waveform and control the infrared light source to emit the infrared light beam according to the modulation waveform so as to change heart rate and blood oxygen saturation readings of the wearable photoplethysmography device; wherein the control unit is configured to generate the modulation waveform using an adaptive genetic algorithm based on an upper layer structure and a lower layer structure, and the algorithm is used to optimize an index related to a manipulation effect so as to improve the manipulation effect on the readings. In this way, controllable interference data can be quickly and accurately generated in a medical health data security test process, so as to ensure the accuracy and reliability of the medical health data security test process. In addition, confusion and deception of data attack behaviors can be realized, and the security of medical health data is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of mobile and wireless security technology, for example to an apparatus, method of interfering with a wearable photoplethysmography device. BACKGROUND

[0002] The rapid development of the Internet of Things in the healthcare field, driven by convenience, cost-effectiveness and user-friendliness, exposes a large number of vulnerabilities due to the lack of effective regulation. Among them, the most critical is the eavesdropping of vital signs, which exploits communication vulnerabilities to compromise data privacy. And insecure mobile medical applications further facilitate unauthorized access and operation, mislead doctors, caregivers and family members, and endanger patients.

[0003] In the existing process of medical health data security testing, various interference means mainly focus on changing the vital sign readings of wearable devices by exploiting vulnerabilities in data transmission protocols and databases, but the existing security testing process ignores the risk of unreliable PPG readings, thus making the medical health data security testing exist the risk of unreliability.

[0004] Therefore, in the existing process of medical health data security testing, there is a problem of how to generate controllable interference data based on the detection principle of PPG to realize accurate and reliable medical health data security testing.

[0005] It should be noted that the information disclosed in the above background section is only used to strengthen the understanding of the background of the present application. SUMMARY

[0006] To have a basic understanding of some aspects of the disclosed embodiments, a brief overview is given below. The summary is not an overall description of the application, nor is it intended to identify key / important elements or delineate the scope of the embodiments, but to serve as a prelude to the detailed description below.

[0007] The apparatus and method for interfering with a wearable photoplethysmography device provided by the embodiments of the present disclosure can solve the problem of uncontrollable interference data and inaccurate medical health data security testing.

[0008] The present disclosure provides an apparatus for interfering with a wearable photoplethysmography device, the apparatus comprising:

[0009] an infrared light source configured to emit a modulated infrared light beam;

[0010] a control unit configured to generate a modulation waveform and control the infrared light source to emit the infrared light beam according to the modulation waveform to change the heart rate and blood oxygen saturation readings of the wearable photoplethysmography device;

[0011] The control unit is configured to generate the modulation waveform using an adaptive genetic algorithm based on the upper layer structure and the lower layer structure, and the algorithm improves the manipulation effect on the readings by optimizing an index related to the manipulation effect.

[0012] In some embodiments, the adaptive genetic algorithm adopts a double-layer structure, including an upper layer structure and a lower layer structure:

[0013] The upper layer structure is configured to filter short waveform segments with good performance and determine the initial population of the lower layer structure by iterating multiple generations, and each generation performs the following steps:

[0014] A Gaussian process surrogate model is constructed using a Bayesian optimization algorithm, which is based on the evaluated waveform segments and their jamming success rates, and is used to predict the expected improvement value of the candidate waveform segments;

[0015] According to the expected improvement value, waveform segments that have the potential to meet the preset conditions are selected from the candidate waveform segments as experimental samples;

[0016] The jamming success rate of the experimental samples is evaluated by experiments, and the evaluation results are used to update the Gaussian process surrogate model;

[0017] Based on the jamming success rate evaluated by experiments, waveform segments with a jamming success rate meeting a preset success threshold are selected as the initial population of the lower layer structure;

[0018] The lower layer structure is configured to generate its initial population by combining the short waveform segments determined by the upper layer structure, and optimize the combination of waveform segments by genetic algorithm to improve the ability of the waveform to resist jamming suppression, and the genetic algorithm includes:

[0019] A child sequence is generated by performing a single-point crossover operation at the substring boundary and a mutation operation by replacing the substring with another substring with a Hamming distance less than a preset threshold;

[0020] The sequence for the next iteration is selected according to the anti-jamming index, and the steps of crossover, mutation and selection are repeated until the anti-jamming index tends to be stable.

[0021] In some embodiments, the anti-jamming index is calculated as follows:

[0022]

[0023] Wherein:

[0024] AIM represents the anti-jamming index;

[0025] PSL is the autocorrelation lobe, defined as PSL = max(|R(k)| / R(0), k≠0), where k = 1, 2,..., N-1, and R(k) is the autocorrelation function;

[0026] SE is the spectral entropy, defined as p(f) is the normalized power spectrum;

[0027] w PSL and w SE are the weights of the autocorrelation sidelobe and the spectral entropy, respectively, and w PSL +w SE = 1;

[0028] PSL min and PSL max are the minimum and maximum of the autocorrelation sidelobe in the current population, respectively;

[0029] SE min and SE max are the minimum and maximum of the spectral entropy in the current population, respectively.

[0030] In some embodiments, the infrared light beam emitted by the infrared light source has a wavelength in a range from 500 nanometers to 1000 nanometers.

[0031] In some embodiments, the interference success rate is determined by:

[0032] For n participants wearing the wearable photoplethysmography device, m interference experiments are performed using the waveform segments;

[0033] The number of successful interference q is determined, where successful interference refers to the waveform segment causing the target physiological parameter of the participant to change by a preset amount;

[0034] The interference success rate is calculated as

[0035] In some embodiments, the target physiological parameter is selected from at least one of heart rate and blood oxygen saturation, and the preset change is an increase or decrease in the value of the target physiological parameter.

[0036] In some embodiments, the adaptive genetic algorithm described above generates the modulated waveform in the form of a binary string, and the control unit is configured to control the infrared light source to emit the modulated infrared light beam using at least one modulation method selected from on-off keying modulation and pulse width modulation based on the binary string.

[0037] In some embodiments, the Gaussian process surrogate model uses a Hamming distance kernel function, and the Gaussian process surrogate model includes a noise regularization term.

[0038] In some embodiments, the genetic algorithm of the lower layer structure uses the short waveform segments screened out by the upper layer structure as coding units for combination optimization, and both the crossover operation and the mutation operation are performed in units of short waveform segments.

[0039] The embodiment of the present disclosure also provides a method for interfering with the reading of a wearable photoplethysmography device, which comprises the following steps of:

[0040] A modulation waveform is generated by using the adaptive genetic algorithm based on the upper layer structure and the lower layer structure, and the modulation waveform is generated by optimizing and manipulating the indicators related to the manipulation effect, so as to improve the manipulation effect on the reading, wherein the modulation waveform generated by the adaptive genetic algorithm based on the upper layer structure and the lower layer structure is obtained according to the device;

[0041] The infrared light source is controlled to emit a modulated infrared light beam according to the modulation waveform;

[0042] The modulated infrared light beam is guided to the wearable photoplethysmography device, so as to change the heart rate and / or blood oxygen saturation reading of the device.

[0043] The device and method for interfering with the wearable photoplethysmography device provided by the embodiment of the present disclosure can achieve the following technical effects:

[0044] The device and method for interfering with the wearable photoplethysmography device provided by the embodiment of the present disclosure can achieve the following technical effects:

[0045] The foregoing general description and the following description are only exemplary and explanatory, and are not used to limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0046] One or more embodiments are exemplarily illustrated by corresponding drawings, which do not constitute limitations on the embodiments, elements with the same reference numerals in the drawings are shown as similar elements, the drawings do not constitute proportional limits, and wherein:

[0047] Figure 1 is a device structure schematic diagram of the wearable photoplethysmography device provided by the embodiment of the present disclosure;

[0048] Figure 2 is a whole architecture schematic diagram of an interference schematic diagram model of the wearable photoplethysmography device interfered by the infrared LED provided by the embodiment of the present disclosure;

[0049] Figure 3is a schematic diagram of the error between the false heart rate measurement value and the reference heart rate value of the wearable PPG device in the interference of reducing the heart rate reading provided by the embodiments of the present disclosure.

[0050] Figure 4 is a schematic diagram of the error between the false heart rate measurement value and the reference heart rate value of the wearable PPG device in the interference of reducing the heart rate reading provided by the embodiments of the present disclosure.

[0051] Figure 5 is a schematic diagram of the error between the false heart rate measurement value and the reference heart rate value of the wearable PPG device in the interference of reducing the heart rate reading provided by the embodiments of the present disclosure.

[0052] Figure 6 is a schematic diagram of the error between the false heart rate measurement value and the reference heart rate value of the wearable PPG device in the interference of reducing the heart rate reading provided by the embodiments of the present disclosure.

[0053] Figure 7 is a schematic diagram of a method for interfering with a wearable photoplethysmography device provided by the embodiments of the present disclosure.

[0054] Figure 8 is a schematic diagram of a device structure for interfering with a wearable photoplethysmography device provided by the embodiments of the present disclosure. DETAILED DESCRIPTION

[0055] In order to enable a more detailed understanding of the features and technical content of the embodiments of the present disclosure, the implementation of the embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings, which are for reference only and do not limit the embodiments of the present disclosure. In the following technical description, in order to facilitate explanation, a plurality of details are provided to provide a full understanding of the disclosed embodiments. However, one or more embodiments can still be implemented without these details. In other cases, well-known structures and devices can be simplified to facilitate the drawings.

[0056] The terms "first", "second", and the like in the embodiments of the present disclosure are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present disclosure described herein can be implemented. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion.

[0057] Unless otherwise specified, the term "a plurality of" means two or more.

[0058] In the embodiments of the present disclosure, the character " / " represents an "or" relationship between the objects before and after it. For example, A / B means A or B.

[0059] The term "and / or" is a descriptive representation of a relationship between objects, indicating that there can be three relationships. For example, A and / or B, indicating that there are three relationships: A or B, or A and B.

[0060] The term "corresponds to" can refer to a relationship or a binding relationship. A corresponds to B means that there is a relationship or a binding relationship between A and B.

[0061] To solve the above problems, the present disclosure provides a device and method for interfering with a wearable photoplethysmography device.

[0062] For the existing medical health data security test scene, the device and method for interfering with a wearable photoplethysmography device provided by the embodiments of the present disclosure will be described below with reference to the accompanying drawings. In addition, the device and method for interfering with a wearable photoplethysmography device can be used in the process of data security confrontation to realize the confusion and deception of attack behavior, thereby improving the security of medical health data.

[0063] Reference Figure 1 The device for interfering with a wearable photoplethysmography device proposed in the present application includes an infrared light source 101 and a control unit 102. Reference Figure 2 The device emits a modulated infrared light beam x(t) through the infrared light source 101 Figure 2 (schematically shown as an LED) to interfere with a target photoplethysmography (PPG) device (including a photoelectric detector (PD) and its own LED (Light-Emitting Diode)).

[0064] In an actual interference scene, the effective signal strength F(t) of the modulated infrared light beam emitted by the infrared light source 101 will be affected by various complex physical propagation factors in the process of reaching the target PPG device and affecting it. These factors include but are not limited to the distance d between the light source and the target, the light-emitting angle characteristics Φ of the light source and the half-power angle Φ 1 / 2 , the angle ψ at which the light beam is incident on the human skin, and the reflection, absorption and scattering characteristics of the human skin tissue to the specific wavelength infrared light. At the same time, the photoelectric detector PD of the target PPG device will also be disturbed by environmental noise when receiving its physiological signal I0. The combined effects of these physical factors make the accurate modeling from x(t) to F(t) very complex and variable in actual applications.

[0065] Due to these complex physical propagation characteristics and environmental factors, the present application proposes a method of directly optimizing and screening the interference modulation waveform through a double-layer adaptive genetic algorithm. The goal of this algorithm is not to explicitly model and invert each physical propagation parameter accurately, but rather to iteratively find and generate modulation waveforms that can most effectively change the readings of the target PPG device in a data-driven manner under real experimental conditions that include all these physical propagation effects. Therefore, through its experimental feedback mechanism, this optimization algorithm indirectly adapts to and overcomes these complex propagation effects to achieve the best interference effect.

[0066] In combination Figure 1 and Figure 2 , the device for interfering with a wearable photoplethysmography device can include:

[0067] an infrared light source 101 configured to emit a modulated infrared light beam;

[0068] a control unit 102 configured to generate the modulation waveform and control the infrared light source to emit the infrared light beam according to the modulation waveform to change the heart rate and blood oxygen saturation readings of the wearable photoplethysmography device;

[0069] In typical wearable photoplethysmography (PPG) signal collection, the built-in photodetector of the device does not strictly distinguish the wavelengths of incident light, and its effective sensing spectral range generally covers 500-1000 nanometers. Although the device will subsequently separate the received mixed light signal into different signal channels according to the preset wavelength through a time-division multiplexing synchronization mechanism with its own light-emitting diode (LED), in order to ensure that the external interference signal emitted by the infrared light source 101 can be effectively received by the photodetector of the target PPG device and produce an interference effect, the wavelength of the infrared light beam emitted by the infrared light source 101 should be set within the range of 500-1000 nanometers. Therefore, in this embodiment, the control unit 102 controls the infrared light source 101 (e.g., an infrared LED) to work within this wavelength range.

[0070] After determining the emission wavelength range of the infrared light source 101, the control unit 102 needs to select an appropriate modulation method to convert the digital sequence generated by the subsequent optimization algorithm into a physical signal that controls the output of the infrared light source 101. In the research and experiments of the present application, two basic digital modulation techniques are focused on to drive the infrared light source 101: On-Off Keying (OOK) modulation and Pulse Width Modulation (PWM). The reason for choosing these two modulation methods is that their implementation circuits are relatively simple, require fewer components, and the physical size can be controlled to the square millimeter level, which makes the interference device containing the infrared light source 101 and the corresponding modulation circuit have good concealability in actual deployment. The subsequent experimental results (such as the attached Figure 3 to the attached Figure 6 will show that by using OOK or PWM modulation methods to drive the infrared light source 101 through the control unit 102, the heart rate and blood oxygen saturation readings of the target PPG device will produce different degrees and characteristics of interference effects. The control unit 102 will generate a binary sequence according to the optimization algorithm, and combine it with the pre-set or dynamically selected OOK or PWM modulation scheme to accurately control the output of the infrared light source 101, so that it emits a specific modulated infrared light beam.

[0071] In a preferred embodiment of the present application, the control unit 102 adopts a modulation waveform generation method based on a double-layer adaptive genetic algorithm, which includes an upper layer structure and a lower layer structure, and the specific implementation is as follows:

[0072] The upper layer structure aims to filter out fragments with high interference success rate from short binary strings (16-bit 0 / 1 strings in this embodiment) through Bayesian optimization combined with genetic ideas.

[0073] The optimization filtering process starts with an initialization step. In this step, the control unit randomly generates 32 16-bit binary strings as initial candidates. In order to evaluate the actual interference effect of these initial strings, they will be subjected to interference experiments one by one, and the respective interference success rates SR(x) will be recorded. The interference success rate SR(x) is defined here as the number of times q that successfully causes the target PPG device reading to change by a preset amount, divided by the total number of interference experiments (i.e. the product of the number of interference times m for each participant and the number of participants n: SR(x) = q / (m*n)). These initially generated binary strings and their corresponding experimentally measured SR(x) values will jointly constitute the initial training data set for the subsequent Gaussian process surrogate model construction.

[0074] In order to be able to effectively predict the interference effect of the untested binary string based on the existing experimental data, and guide the subsequent optimization search direction, the control unit constructs and maintains a Gaussian process proxy model. The core of the Gaussian process model lies in the selected kernel function. In the embodiment, the Hamming distance kernel function is used to measure the similarity between binary strings, and its specific form is k(x, x') = exp(-aHD(x, x')), where HD(x, x') represents the Hamming distance between binary strings x and x', and in the experiment of the embodiment, the attenuation factor a is set to 0.05. Based on this Gaussian process model, for any new binary string x* that has not been experimentally evaluated, the control unit can predict the mean value μ(x*) and the uncertainty (in the form of variance σ 2 (x*) of the interference success rate it may produce. The calculation of these predicted values follows the standard Gaussian process regression formula:

[0075] μ(x * )=K(x * ,X)[K(X,X)+σ 2 n I] -1 y,

[0076] σ 2 (x * )=K(x * ,x * )-K(x * ,X)[K(X,X)+σ 2 n I] -1 K(X,x * ),

[0077] In the formula, X represents the set of binary strings that have been experimentally evaluated, y is the SR observation value corresponding to these strings, K is the corresponding kernel matrix calculated by the selected kernel function, and I is the unit matrix. In order to more accurately model the noise that may exist in actual observation, the observation noise variance σ 2 n is introduced in the model. In the embodiment, the standard deviation of the observation noise is set to 0.05.

[0078] After the Gaussian process model is constructed and initialized, the optimization screening process enters an iterative optimization loop. In each iteration of the generation, the control unit first generates a new set of candidate binary strings by selecting and performing genetic operations on the existing binary strings in the current population (the initial population is the aforementioned 32 strings). In the experiments of this embodiment, the size of the candidate population generated in each generation is about 50, and the specific composition is as follows: 25 sub-generations are generated by performing crossover operations on the parents extracted from the current population and selected based on the GP model, 15 sub-generations are generated by performing mutation operations, and 10 new individuals are additionally generated at random. The selection operation includes extracting 3 strings from the current population, using the Gaussian process model to select the individual with the optimal predicted SR as one of the parents, and repeating this process to obtain several parents. The genetic operation includes single-point random crossover, two-point random crossover, and bit mutation, and in the experiments of this embodiment, the mutation probability of each bit is set to 0.2. To avoid unnecessary repeated experiments, the system ensures that the newly generated candidate strings are not identical to all the evaluated strings, and if a repetition occurs, it is regenerated.

[0079] Next, in order to determine which newly generated candidate strings are most worthy of performing expensive actual interference experiments, the control unit uses the Expected Improvement (EI) as the acquisition function. For each candidate string, the current Gaussian process model is used to predict its μ(x*) and σ 2 (x*), and the EI value is calculated accordingly. The formula for calculating the EI value is:

[0080]

[0081] In this formula, f max represents the highest interference success rate SR value observed in all the binary strings that have been evaluated by the current experiments, and Φ(·) and are the cumulative distribution function (CDF) and probability density function (PDF) of the standard normal distribution, respectively. The control unit then selects the top 10 candidate strings with the highest EI values and performs actual interference experiments on them to obtain their true SR(x) values. It should be noted that the number of candidate strings with the highest EI values described above can be adjusted according to actual conditions and is not limited here.

[0082] After obtaining a new batch of experimental data, the control unit adds these newly evaluated binary strings and their corresponding SR values to the training data set of the Gaussian process model, and updates or re-trains the Gaussian process model using the updated complete data set to improve the accuracy of subsequent model predictions. At the same time, in order to maintain and evolve the population of binary strings with excellent performance, from all strings that have been experimentally evaluated, the real SR values are sorted, and the 32 strings with the best performance are selected to form the population of the next iteration, so that the population size is maintained at 32. It should be noted that the number of strings selected with the best performance can be adjusted according to actual conditions, and is not limited here.

[0083] This iterative optimization process will continue until the preset termination condition is met. In the experiment of the embodiment, the termination condition is set to reach the maximum number of iterations of 20 rounds, at which time it is observed that the average success rate of the selected substring waveform can reach more than 80%, indicating that the optimization has converged to an excellent degree. The optimization and selection process of the first stage ends at this time. The 32 16-bit binary strings finally retained in the population are considered to be high-quality short waveform fragments after optimization and selection, and they will be used as basic building blocks for the subsequent long waveform combination optimization in the second stage.

[0084] The lower structure of the application aims to use the 32 high-quality 16-bit short waveform fragments selected by the upper structure as basic building blocks. The combination of these short fragments is optimized by a genetic algorithm to construct a 128-bit long sequence composed of 8 short fragments. The design goal of these long sequences is to have more complex pattern characteristics, so that they are more difficult to be predicted and filtered out by the adaptive algorithm of the target PPG device, so as to achieve a more effective interference effect.

[0085] First, in the encoding and initialization step, the long interference waveform is encoded in this stage as a 128-bit sequence composed of 8 16-bit short fragments selected from the first stage results. The control unit constructs an initial population consisting of 100 such 128-bit long waveform sequences. The generation of the initial population includes randomly extracting and splicing long sequences from the 32 short fragment library, while also introducing heuristic rules, specifically, those substrings with higher "interference success rate" in the first stage experiment are preferentially arranged and combined together, in order to include some individuals with higher potential in the initial stage.

[0086] Next, to evaluate the potential jamming performance of each 128-bit long waveform sequence, the control unit adopts an Anti-jamming Integrated Measure (AIM). The AIM is designed to balance two key properties of the waveform: one is the aperiodicity, which is embodied by a lower Peak Sidelobe Level (PSL), and the other is the dispersion of spectral energy, which is embodied by a higher Spectral Entropy (SE). The PSL is calculated based on the autocorrelation function R xx (k) of the sequence. For a binary sequence s of length N (N = 128 in this embodiment), the autocorrelation function is defined as This function measures the similarity of the sequence to itself at different delays k. The PSL is accordingly defined as:

[0087]

[0088] The SE is calculated by first obtaining the power spectrum P(f) of the long waveform sequence, normalizing it to obtain the probability distribution P(f), and then defining the spectral entropy SE as The final AIM is obtained by weighted summation of the two normalized indicators, and the specific calculation formula is:

[0089]

[0090] The weight coefficients w PSL and w SE are both initially set to 0.5 in the experiments of this embodiment. PSL min and PSL max are the minimum and maximum values of the autocorrelation sidelobes in the current population, SE min and SE max are the minimum and maximum values of the spectral entropy in the current population.

[0091] Then, the optimization process enters the iterative loop of the genetic algorithm. In each iteration of a generation, first parental selection is performed: 10 sequences are randomly drawn from the current population of 100 sequences, and the individual with the best AIM indicator is selected as the parent. The crossover operation is performed in basic units of 16-bit short waveform segments, using a single-point crossover method, with the crossover point selected at the boundary between these segments. The mutation operation is the replacement of one or more 16-bit short segments in the long waveform; the new segment to be replaced is randomly selected from the 32-segment library screened in the first stage, and this replacement is only performed in segments with a Hamming distance of less than 3 from the current substring, in order to ensure that the new sequence, while being locally optimized, does not completely destroy the good properties that have been formed. Through crossover and mutation operations, in this embodiment, 100 offspring individuals are generated in each generation, of which 70 are generated by crossover and 30 by mutation. Subsequently, population updating is performed: the newly generated 100 offspring are merged with the original parent population (100). The AIM values of all individuals in the new population are calculated, and the individuals are sorted according to the AIM values, and the 100 individuals with the highest fitness are selected to enter the next generation, so that the population size remains unchanged.

[0092] The iterative process of this genetic algorithm will continue until the preset termination condition is met. In this embodiment, when it is observed that the average or optimal value of the AIM of the population no longer improves significantly in continuous multiple generations of iterations, i.e., the algorithm reaches a state of convergence, the optimization process of the second stage is terminated.

[0093] Finally, after the genetic algorithm iteration is completed, the control unit retains 100 candidate long waveforms from the final obtained population. In order to finally determine the best jamming waveform, actual jamming experiments will be performed on these candidate long waveforms one by one, and their jamming success rates SR(x) in real scenarios are evaluated, which are defined the same as in the first stage. Finally, the waveform with the highest SR(x) value is selected as the best jamming waveform output by the present application.

[0094] In this embodiment, at least two modulation methods are supported: On-Off Keying (OOK) modulation and Pulse Width Modulation (PWM).

[0095] For On-Off Keying (OOK) modulation:

[0096] The control unit directly maps each bit in the 128-bit binary sequence to the on-off state of the infrared light source. Specifically, when a bit in the sequence is '1', the infrared light source is turned on and emits infrared light in the corresponding time unit; when a bit in the sequence is '0', the infrared light source is turned off and does not emit infrared light in the corresponding time unit.

[0097] For pulse width modulation (PWM):

[0098] The control unit maps each bit (or combination of bits) in the 128-bit binary sequence to a duty cycle of a PWM signal period. In this embodiment, each bit in the sequence can directly determine the duty cycle of a PWM period. For example, when a bit in the sequence is '0', the control unit generates a PWM signal with a lower duty cycle (20%) to drive the infrared light source; when a bit in the sequence is '1', it generates a PWM signal with a higher duty cycle (80%).

[0099] The control unit converts the optimized binary sequence into a corresponding driving signal using one of the above OOK or PWM modulation methods according to a pre-configured or dynamically selected configuration. The driving signal is then applied to the infrared light source (such as an infrared LED), causing it to emit a modulated infrared beam according to the generated modulation waveform, which in turn is used to interfere with the target wearable PPG device.

[0100] In one specific embodiment, based on Figure 1 The device for interfering with a wearable photoplethysmography device was used to conduct interference experiments, and the results were recorded.

[0101] First, the "heart rate" interference was tested using OOK and PWM modulation techniques, with the following specific configurations:

[0102] OOK modulation method: Commercial infrared LED with a wavelength of 850 nm, flashing at an interval of 0.25 s at a power of 5 W;

[0103] PWM modulation method: PWM infrared LED with a symbol period (duration of each PWM signal modulation period) of 0.2-0.8 s, wavelength of 850 nm, and power of 5 W.

[0104] 23 participants (19 healthy people and 4 people with respiratory diseases) wore the interfered object, and we interfered with it at a distance of 100 cm and an incident angle of 30°. The best waveform was selected using the proposed infrared waveform optimization framework, and the heart rate value recorded during the non-interference period was used as the reference value for the wearable device. In addition, we used electrocardiography (ECG) to obtain the true heart rate to more comprehensively evaluate the interference effect.

[0105] 1. Heart rate increase experiment

[0106] In a total of 2000 "heart rate increase" interferences (each interference for 1 minute), Figure 3 The error distribution between the PPG device measurement and the reference heart rate is shown:

[0107] OOK: 83.35% success rate, 54.00% of the errors are greater than 4 bpm;

[0108] PWM: 75.00% success rate, 44.10% of the errors are greater than 4 bpm;

[0109] In addition, the Bland-Altman plot Figure 3 shows that, under OOK modulation, 95% of the errors are between -3.56 bpm and 11.66 bpm; under PWM modulation, 95% of the errors are between -3.85 bpm and 9.63 bpm. The upper limits of agreement are 11.66 bpm (OOK) and 9.63 bpm (PWM), respectively, indicating that there is a high probability that OOK and PWM will overestimate the heart rate of the PPG device within this range.

[0110] When compared with the true heart rate of ECG, 72.20% and 66.64% of the instances have errors greater than 4 bpm for OOK and PWM, respectively.

[0111] It can be seen that in this scenario, the interference effect of OOK modulation mode is slightly better than that of PWM.

[0112] 2. Heart rate reduction experiment

[0113] Similarly, we counted the deviation of the PPG device measurement value from the reference heart rate for 2000 times of "heart rate reduction" interference (1 minute per interference), as shown in Figure 4 .

[0114] OOK: 86.35% success rate, 58.95% of the errors are greater than 4 bpm;

[0115] PWM: 74.35% success rate, 47.35% of the errors are greater than 4 bpm;

[0116] In addition, the Bland-Altman plot Figure 4 shows that, under OOK modulation, 95% of the errors are between -12.48 bpm and 2.94 bpm; under PWM modulation, 95% of the errors are between -10.70 bpm and 4.48 bpm. The lower limits of agreement are -12.48 bpm (OOK) and -10.70 bpm (PWM), respectively, indicating that there is a high probability that OOK and PWM will underestimate the heart rate of the PPG device within this range.

[0117] When compared with the ECG value, 70.00% and 55.10% of the samples have errors greater than 4 bpm for OOK and PWM, respectively.

[0118] Overall, OOK modulation is slightly better than PWM in both "increasing" and "decreasing" heart rate readings.

[0119] II. Blood oxygen saturation interference experiment and results

[0120] The experimental setup of the blood oxygen saturation interference experiment is the same as that of the heart rate interference experiment.

[0121] Since the blood oxygen saturation of healthy people is usually close to 100%, there is limited room for further "increasing" its readings, so we only count cases where the reference blood oxygen saturation is strictly less than 100%. In addition to the reference readings of the PPG device during the non-interference period, we also record the true blood oxygen saturation using a shielded clinical blood oxygen device. The experiment also uses two infrared modulation methods, OOK and PWM.

[0122] 1. Blood oxygen saturation increase experiment

[0123] We conducted a total of 2000 "blood oxygen saturation increase" interferences (1 minute per interference), Figure 5 The error distribution of the false blood oxygen saturation and the reference blood oxygen saturation (the range of the subjects' blood oxygen saturation is about 89%~99%) is shown:

[0124] OOK: 61.70% of the interferences caused an increase in blood oxygen saturation, and 17.00% of the sample errors exceeded 4%;

[0125] PWM: the success rate of interference reached 71.70%, and 19.00% of the errors were greater than 4%;

[0126] In addition, the Bland-Altman plot ( Figure 5 ) shows that under OOK modulation, 95% of the error values are between -3.27% and 6.17%; under PWM modulation, 95% of the error values are between -3.73% and 6.13%. The upper limit of consistency reaches 6.17% (OOK) and 6.13% (PWM), respectively, indicating that within this range, both modulation methods can cause the PPG device to overestimate the blood oxygen saturation.

[0127] If compared with the true blood oxygen saturation, 17.89% and 20.22% of the instances have errors greater than 4% under OOK and PWM, respectively.

[0128] 2. Blood oxygen saturation decrease experiment

[0129] For 2000 "blood oxygen saturation decrease" interferences (1 minute per interference), Figure 6 The error distribution of the false blood oxygen saturation and the reference blood oxygen saturation is shown:

[0130] OOK: the success rate is 76.70%, and 28.35% of the sample errors are greater than 4%;

[0131] PWM: success rate 78.00%, 27.70% of errors are more than 4%;

[0132] In addition, the Bland-Altman plot Figure 6 shows that, under OOK modulation, 95% of the error values are between -7.18% and 2.62%; under PWM modulation, 95% of the error values are between -6.77% and 2.49%. Among them, the lower limit of consistency reaches -7.18% (OOK) and -6.77% (PWM), respectively, indicating that in this range, both modulation methods can cause PPG devices to overestimate blood oxygen saturation.

[0133] Compared with the true blood oxygen saturation, 33.61% and 33.05% of the instances of OOK and PWM, respectively, have errors greater than 4%.

[0134] It can be seen that in the blood oxygen saturation interference, the PWM modulation method is slightly better.

[0135] The device for interfering with a wearable photoplethysmography device proposed in the embodiments of the present disclosure realizes control of the interference of the heart rate and blood oxygen saturation reading of the interference object, and the selected optimization framework is simple and effective and has good convergence effect, and can quickly converge to a high-success-rate waveform. Therefore, in the existing medical health data security test process, controllable interference data can be generated to analyze and find the generation behavior of the interference data in the test process, and thus the accuracy and reliability of the medical health data security test process are ensured. In addition, the interference generation method can also be used in the process of data security confrontation to realize the confusion and deception of the attack behavior, and thus the security of the medical health data is improved.

[0136] Corresponding to the device for interfering with a wearable photoplethysmography device in Figure 1 , the present disclosure also provides a method for interfering with a wearable photoplethysmography device, as shown in Figure 7 , which can specifically include:

[0137] S701, generating a modulation waveform using the adaptive genetic algorithm based on the upper layer structure and the lower layer structure, and the algorithm improves the manipulation effect on the reading by optimizing and manipulating the indicators related to the manipulation effect, wherein the modulation waveform generated by the adaptive genetic algorithm based on the upper layer structure and the lower layer structure is obtained according to the device in Figure 1 ;

[0138] S702, controlling the infrared light source to emit a modulated infrared light beam according to the modulation waveform;

[0139] S703, direct the modulated infrared light beam to the wearable photoplethysmography device to change the heart rate and / or blood oxygen saturation reading of the device.

[0140] In addition, it also needs to be explained that the various parameters selected for use in the present disclosure can be adjusted according to the actual situation, which is not limited here.

[0141] In combination Figure 8 As shown in the figure, the embodiments of the present disclosure also provide a device 800 for interfering with a wearable photoplethysmography device, which includes a processor 804 and a memory 801. Optionally, the system can also include a communication interface 802 and a bus 803. Wherein the processor 804, the communication interface 802, the memory 801 can complete the communication between each other through the bus 803. The communication interface 802 can be used for information transmission. The processor 804 can call the logic instructions in the memory 801 to execute the method for interfering with the wearable photoplethysmography device in the above-mentioned embodiments.

[0142] In addition, the logic instructions in the memory 801 described above can be realized in the form of a software function unit and sold or used as an independent product when used, which can be stored in a computer readable storage medium.

[0143] The memory 801 as a kind of computer readable storage medium can be used to store software programs, computer executable programs, such as the program instructions / modules corresponding to the method in the embodiments of the present disclosure. The processor 804 executes the program instructions / modules stored in the memory 801, thereby executing function application and data processing, that is, realizing the method for interfering with the wearable photoplethysmography device in the above-mentioned embodiments.

[0144] The memory 801 can include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function; The data storage area can store data created according to the use of the terminal device and the like. In addition, the memory 801 can include a high-speed random access memory, and can also include a non-volatile memory.

[0145] The embodiments of the present disclosure provide a computer readable storage medium, which stores computer executable instructions, and the computer executable instructions are set as the method for interfering with the wearable photoplethysmography device.

[0146] The above-mentioned computer readable storage medium can be a transitory computer readable storage medium, or a non-transitory computer readable storage medium.

[0147] The technical solutions of the embodiments of the present disclosure can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes one or more instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method of the embodiments of the present disclosure. The aforementioned storage medium can be a non-transitory storage medium, including: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes, and can also be a transitory storage medium.

[0148] The above description and drawings sufficiently illustrate the embodiments of the present disclosure to enable one skilled in the art to practice them. Other embodiments can include structural, logical, electrical, process, and other changes. The embodiments represent only a few of the possible variations. Individual components and functions are optional unless explicitly required, and the order of operations can be changed. Parts and features of some embodiments can be included in or replace parts and features of other embodiments. As used in the description of the embodiments, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. Similarly, as used in this application, the term "and / or" refers to any and all possible combinations of one or more associated listed items. In addition, when used in this application, the term "comprise" and its variants "comprises" and / or "comprising" and the like mean the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. Without more limitations, the element defined by the phrase "comprising a" does not exclude the presence of additional identical elements in the process, method, or device including the element. In this document, each embodiment focuses on the differences from other embodiments, and the same or similar parts between various embodiments can be referred to each other. For the method, product, etc. disclosed by the embodiments, if it corresponds to the method part disclosed by the embodiments, the relevant part can be referred to the description of the method part.

[0149] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software manner can depend on specific application and design constraints of the technical solutions. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of the present disclosure. Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0150] In the embodiments disclosed herein, the disclosed methods, products (including but not limited to devices, equipment, etc.) can be implemented in other ways. For example, the above-described device embodiments are merely illustrative, for example, the division of units can be merely a logical function division, and actual implementation can have another division manner, for example, multiple units or components can be combined or integrated into another system, or some features can be omitted or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed units can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms. The units described as separate components can be or can not be physically separated, and the components displayed as units can be or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. Part or all of the units can be selected according to actual needs to implement the embodiments. In addition, the functional units in the embodiments of the present disclosure can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit.

[0151] The computer program instructions can also be loaded onto a computer or other programmable apparatus to cause a series of operations to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0152] The various embodiments of the systems and techniques described above can be implemented in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a load programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0153] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general or special purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, implements the functions / acts specified in the flowchart and / or block diagram block or blocks. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a standalone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0154] In the context of this disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0155] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0156] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0157] The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server can arise by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0158] It should be understood that the various forms of flow shown above can be reordered, added to, or have steps deleted, for example. The steps recited in the present disclosure can be performed in parallel, in series, or in a different order, as long as the desired results of the technical solutions of the present disclosure are achieved, which are not limited herein.

[0159] The specific implementation described above does not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A device for interfering with a wearable photoplethysmography device, characterized in that, The device includes: An infrared light source, configured to emit a modulated infrared beam; The control unit is configured to generate a modulated waveform and control the infrared light source to emit an infrared beam according to the modulated waveform, so as to change the heart rate and blood oxygen saturation readings of the wearable photoplethysmography device. The control unit is configured to generate the modulation waveform using an adaptive genetic algorithm based on an upper-layer structure and a lower-layer structure. The algorithm improves the manipulation effect on the reading by optimizing the indicators related to the manipulation effect. The adaptive genetic algorithm adopts a two-layer structure, including an upper layer and a lower layer: The upper-level structure is configured to iterate through multiple generations to select short waveform segments with good performance and determine the initial population of the lower-level structure. Each generation performs the following steps: A Gaussian process surrogate model is constructed using a Bayesian optimization algorithm. This Gaussian process surrogate model is based on the evaluated waveform segments and their interference success rates, and is used to predict the expected improvement value of candidate waveform segments. Based on the desired improvement value, select waveform segments with the potential to meet the preset conditions from the candidate waveform segments as experimental samples; The perturbation success rate of the experimental samples was evaluated through experiments, and the evaluation results were used to update the Gaussian process proxy model. Based on the interference success rate evaluated in experiments, waveform segments with interference success rates that meet the preset success threshold are selected as the initial population of the lower-level structure. The lower-level structure is configured to generate an initial population by combining short waveform segments determined by the upper-level structure, and to optimize the combination of the waveform segments using a genetic algorithm to improve the waveform's ability to resist interference suppression. The genetic algorithm includes: The offspring sequence is generated by performing a single-point crossover operation at the substring boundary and a mutation operation by replacing the substring with other substrings whose Hamming distance is less than a preset threshold. Sequences for the next generation of iterations are selected based on the anti-interference metric, and the crossover, mutation, and selection steps are repeated until the anti-interference metric tends to stabilize.

2. The apparatus according to claim 1, wherein, The calculation method for the anti-interference index is as follows: ， in: Indicates anti-interference index; It is an autocorrelation sidelobe, defined as ,in , It is an autocorrelation function; It is the spectral entropy, defined as , This is the normalized power spectrum; and These are the weights of the autocorrelation sidelobes and the spectral entropy, respectively. ; and These are the minimum and maximum values ​​of the autocorrelation sidelobe in the current population, respectively. and These are the minimum and maximum values ​​of the spectral entropy in the current population, respectively.

3. The apparatus according to claim 1, characterized in that, The infrared light emitted by the infrared light source has a wavelength in the range of 500 nanometers to 1000 nanometers.

4. The apparatus according to claim 1, wherein, The success rate of the interference is determined in the following way: right Participants wearing the wearable photoplethysmography device used the waveform segment to perform... Secondary interference experiment; Determine the number of successful interferences Successful interference refers to the waveform segment causing a preset change in the target physiological parameters of the participant; The interference success rate is calculated as follows: .

5. The apparatus according to claim 4, wherein, The target physiological parameter is selected from at least one of heart rate and blood oxygen saturation, and the preset change is an increase or decrease in the value of the target physiological parameter.

6. The apparatus according to any one of claims 1 to 5, wherein, The adaptive genetic algorithm generates the modulated waveform in the form of a binary string, and the control unit is configured to control the infrared light source to emit the modulated infrared beam based on the binary string using at least one modulation method selected from on-off keying modulation and pulse width modulation.

7. The apparatus according to claim 1, wherein, The Gaussian process surrogate model uses the Hamming distance kernel function and includes a noise regularization term.

8. The apparatus according to claim 1, wherein, The genetic algorithm of the lower layer uses the short waveform segments selected by the upper layer as encoding units for combination optimization, and the crossover operation and the mutation operation are both performed on the short waveform segments.

9. A method for interfering with a wearable photoplethysmography device, characterized in that, The method includes: Modulated waveforms are generated using an adaptive genetic algorithm based on an upper-level and lower-level structure. The algorithm improves the manipulation effect on the readings by optimizing indices related to the manipulation effect. The modulated waveforms generated by the adaptive genetic algorithm based on the upper-level and lower-level structure are obtained by the apparatus according to any one of claims 1-8. Control the infrared light source to emit a modulated infrared beam according to the modulated waveform; The modulated infrared beam is directed to the wearable photoplethysmography device to alter the device’s heart rate and / or blood oxygen saturation readings.

Images