A neural sequence data privacy protection circuit and method

Abstract

The present invention discloses a neural sequence data privacy protection circuit and method, belonging to the technical field of information security. The present invention performs high-entropy modulation on the original neural sequence data (low-entropy data) through a true random number circuit and random slot time slot insertion, and can randomly or manually specify to select some sensitive data according to the protection level to perform reversible or irreversible desensitization protection on the original neural sequence data; at the same time, for the reversibly desensitized protected neural data, the original neural data signal can also be restored through desensitization reverse processing. It is an effective circuit design for protecting the digital neural sequence data after neural electrical sensor sampling and digital quantization, and the hardware implementation of this circuit is simple and effective. It solves the problem of "lacking a privacy protection circuit for biological neural sequence data" existing in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of information security technology, and particularly to a neural sequence data privacy protection circuit and method. Background Art

[0002] The statements in this section merely mention the background art related to the present invention and do not necessarily constitute prior art.

[0003] In the fields of biometrics and artificial intelligence, more and more biological signals are used as a medium to connect the fields of artificial intelligence and biometric identification. In the increasingly open field of AI biometrics, the privacy encryption of biological signals and the security issues of transmission authentication are particularly prominent, and are at risk of being counterfeited and illegally exploited at any time.

[0004] Intracranial neural electricity or extracranial brain electrical signals, as the most important features that can reflect physiological activities, are closely related to the functional state of the brain nervous system. Nowadays, the objects of privacy encryption protection for the vast majority of biological signals mostly focus on the surface of organisms, such as biological signals like facial features, fingerprints, voiceprints, etc., while the privacy protection circuit for biological neural sequence data is rarely seen, seriously affecting the security during data transmission and use. Summary of the Invention

[0005] In order to solve the deficiencies of the prior art, the present invention provides a neural sequence data privacy protection circuit, method and electronic device, which realize the encryption protection of neural sequence data through a simple circuit design.

[0006] In a first aspect, the present invention provides a neural sequence data privacy protection circuit;

[0007] A neural sequence data privacy protection circuit includes:

[0008] A true random number circuit, which is used to obtain an enable signal, find the jitter interval of the clock frequency according to the enable signal, and generate non-deterministic data and deterministic data;

[0009] A first exclusive OR operation circuit, which is used to perform an exclusive OR operation on the non-deterministic data and the deterministic data, and collect the non-deterministic data as high-entropy modulation data;

[0010] A shift register, which is used to store the high-entropy modulation data;

[0011] A modulation circuit, which is used to randomly obtain the high-entropy modulation data in the shift register, and obtain the privacy desensitized neural sequence data according to the high-entropy modulation data and the biological sequence data; wherein, the biological sequence data is the data at the position of the slot time slot in the original neural sequence data.

[0012] Further, the modulation circuit is a second exclusive - OR operation circuit, and the second exclusive - OR operation circuit is used to randomly obtain high - entropy modulation data in the shift register, perform an exclusive - OR operation on the high - entropy modulation data and the biological sequence data, and obtain the neuralsequence data after privacy desensitization processing.

[0013] Preferably, the second exclusive - OR operation circuit is further used to perform an inverse exclusive - OR operation on the neuralsequence data after privacy desensitization processing and the biological sequence data to obtain the original neuralsequence data.

[0014] Preferably, the true random number circuit includes a high - frequency oscillation loop circuit and a low - frequency oscillation loop circuit;

[0015] The output end of the high - frequency oscillation loop circuit is connected to the input end of the low - frequency oscillation loop circuit, the output end of the low - frequency oscillation loop circuit is connected to the input end of the first exclusive - OR operation circuit, the output end of the first exclusive - OR operation circuit is connected to the input end of the shift register, and the output end of the shift register is connected to the input end of the second exclusive - OR operation circuit.

[0016] Further, the modulation circuit is a replacement unit, and the replacement unit is used to use high - entropy modulation data to replace the biological sequence data to obtain the neuralsequence data after privacy desensitization processing.

[0017] Further, the high - frequency oscillation loop circuit is used to generate a high - frequency clock signal, and the low - frequency oscillation loop circuit is used to perform a first - order sampling and a second - order sampling on the jitter space of the high - frequency clock signal according to a low - frequency clock signal to obtain non - deterministic data and deterministic data.

[0018] In a second aspect, the present invention provides a method for protecting the privacy of neuralsequence data;

[0019] A method for protecting the privacy of neuralsequence data, based on the above - mentioned neuralsequence data privacy protection circuit, includes the following steps:

[0020] Obtain the original neuralsequence data, and divide the original neuralsequence data into equal - length period segments; for each period segment, select positions to insert slot time slots according to the privacy protection degree;

[0021] Obtain high - entropy modulation data, and obtain the neuralsequence data after privacy desensitization processing according to the biological sequence data and the high - entropy modulation data at the positions where the slot time slots are located.

[0022] Further, obtaining the neuralsequence data after privacy desensitization processing according to the biological sequence data and the high - entropy modulation data at the positions where the slot time slots are located is:

[0023] Perform an exclusive OR operation on the biological sequence data and the high-entropy modulation data at the position of the slot time slot to obtain the neural sequence data after privacy desensitization processing;

[0024] Or,

[0025] Use the high-entropy modulation data to replace the biological sequence data at the position of the slot time slot to obtain the neural sequence data after privacy desensitization processing.

[0026] Furthermore, the privacy protection degree is the ratio of the width of the slot time slot to the width of the period.

[0027] In a third aspect, the present invention provides an electronic device;

[0028] An electronic device includes a memory, a processor, and a neural sequence data privacy protection circuit as described in any one of claims 1-7.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] The technical solution provided by the present invention is a circuit design for effectively protecting the digital neural sequence data after neural electro-sensor sampling and digital quantization, and the circuit hardware implementation is simple and effective; through a true random number circuit and random slot time slot insertion, high-entropy modulation is performed on the original neural sequence signal (low-entropy data), and partial sensitive data can be randomly or manually specified to perform reversible or irreversible desensitization protection on the original neural sequence data according to the protection degree; at the same time, for the reversibly desensitized protected neural data, the original neural data signal can also be restored through desensitization reverse processing.

[0031] The original neural information data is cut into equal-length period segments, and after passing through the slot time slot and the privacy protection degree "ε", user-defined desensitized privacy encryption protection or manually specified allocated sensitive privacy encryption protection can be realized. After high-entropy modulation of the data through a true random number circuit, the modulated data is obtained. Finally, for the reversible privacy protection operation based on the XOR operation, the original data signal is also restored through XOR desensitization reverse processing, thus realizing the design and implementation of the protection circuit for biological neural sequence signals. Description of the Drawings

[0032] The specification drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0033] Figure 1 It is a schematic diagram of the neural sequence data privacy protection circuit provided by the embodiment of the present invention;

[0034] Figure 2 Schematic diagram of reversible desensitization processing of original neural sequence data provided by an embodiment of the present invention;

[0035] Figure 3 Interpretation diagram of irreversible desensitization processing of original neural sequence data provided by an embodiment of the present invention. Detailed implementation manners

[0036] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0037] It should be noted that the terms used herein are only for describing specific implementation manners and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units need not be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0038] In the case of no conflict, the embodiments in the present invention and the features in the embodiments may be combined with each other.

[0039] Embodiment 1

[0040] In the prior art, the security issues of privacy encryption and transmission authentication of biological signals are particularly prominent, and are at risk of being counterfeited and illegally used at any time. As the neural sequence data can reflect the most important characteristics of the human body physiology, the security issues cannot be ignored; therefore, the present invention provides a neural sequence data privacy protection circuit.

[0041] Next, in combination with Figure 1 - Figure 2A neural sequence data privacy protection circuit disclosed in this embodiment will be described in detail. The neural sequence data privacy protection circuit includes a true random number circuit, a first exclusive OR operation circuit, a shift register, and a second exclusive OR operation circuit. The true random number circuit is used to obtain an enable signal, find the jitter interval of the clock frequency according to the enable signal, and generate non-deterministic data and deterministic data. The first exclusive OR operation circuit is used to perform an exclusive OR operation on the non-deterministic data and the deterministic data, and collect the non-deterministic data as high-entropy modulation data. The shift register is used to store the high-entropy modulation data as a random sequence pool of the high-entropy modulation data. The second exclusive OR operation circuit is used to randomly obtain the high-entropy modulation data in the shift register, perform an exclusive OR operation on the high-entropy modulation data and the biological sequence data, and obtain the privacy-desensitized neural sequence data.

[0042] In this embodiment, the high-entropy modulation data is a random number output by the true random number circuit, which is high-entropy. High-entropy means that the data has undergone randomization and scrambling processing; low-entropy refers to the original neural sequence data, which has relatively poor randomness compared to the random number because the original neural sequence data has not undergone randomization processing.

[0043] Among them, the biological sequence data is the data at the position of the slot time slot in the original neural sequence data. The position of the slot time slot is determined by dividing the original neural sequence data into multiple periodic segments and randomly or specifying the insertion of the slot time slot within the periodic segment.

[0044] The true random number circuit includes a high-frequency oscillation loop circuit and a low-frequency oscillation loop circuit; the output end of the high-frequency oscillation loop circuit is connected to the input end of the low-frequency oscillation loop circuit, the output end of the low-frequency oscillation loop circuit is connected to the input end of the first exclusive OR operation circuit, the output end of the first exclusive OR operation circuit is connected to the input end of the shift register, and the output end of the shift register is connected to the input end of the second exclusive OR operation circuit.

[0045] Specifically, based on the true random number circuit to generate random numbers, the true random number circuit uses the clock signal generated by the low-frequency oscillation loop circuit (the low-frequency clock signal generated by the low-frequency oscillation loop circuit has no clock jitter), and samples the jitter interval of the clock generated by the high-frequency oscillation loop circuit with a certain probability, hoping to collect the jitter interval to obtain the non-deterministic "X" bit. Then sample again to convert the non-deterministic "X" bit into a deterministic "1" or "0", that is, to achieve random bit collection; among them, the frequency of the clock signal generated by the low-frequency oscillation loop circuit can be adjusted to select the appropriate clock frequency of the low-frequency oscillation loop and improve the sampling hit probability of the jitter interval of the low-frequency sampling of the high-frequency; and ensure that the low-frequency oscillation loop frequency and the high-frequency oscillation loop frequency are incoherent, that is, there is no phase dependence relationship between the low-frequency and the high-frequency.

[0046] For multiple channels, random jitter can be collected. The low-frequency RO will hit the random jitter accumulated by multiple channels. That is, every time the low-frequency RO clock hits a high-frequency jitter interval, an indeterminate "x" state will be output. For high frequency, the fixed frequency does not require a specific frequency range. The frequency of the low-frequency RO can be adjusted so that after the collected 'x' state sequence is converted into a definite state '1' or '0', it has high randomness.

[0047] The true random number circuit used collects the acquisition results (indeterminate state "X" or definite state "1 / 0") of multiple high-frequency oscillator circuits, and is collected based on the subsequent first XOR operation circuit, and further stored in the subsequent shift register as a random sequence pool for high-entropy modulation data to provide high-entropy modulation data.

[0048] In this embodiment, through privacy desensitization reverse processing, the second XOR operation circuit that is still simple and easy to implement in hardware can be used to reversibly restore the original bio-neural information. By saving and transmitting the previous bio-sequence data as a Key, the other party receives the encrypted neural sequence data and the received key (i.e., the bio-sequence data), and can perform another reverse XOR "exclusive OR operation" to restore the original neural information (reversible restoration).

[0049] Furthermore, the true random number circuit used in this embodiment has multiple high-frequency oscillator circuits to collect more indeterminate states "X". And the first circuit gate of each high-frequency oscillator circuit is a NAND gate, so that a single high-frequency oscillator circuit can be enabled to pass through this NAND gate to selectively activate or prohibit this high-frequency oscillator circuit.

[0050] In this embodiment, the high-frequency RO (Ring Oscillator) oscillator circuit includes a NAND gate (note: one of the fan-ins of the NAND gate is the enable port Pin). The NAND gate is serially connected with an odd number of inverters behind it, and the output Pin of the last inverter is fed back and connected to the other fan-in Pin of the NAND gate. The total gate delay of this high-frequency RO oscillator loop (including: one NAND gate and the weighted total delay of an odd number of inverters) is half of the period of the high-frequency RO oscillator, that is, the period Period of the high-frequency RO oscillator is twice the total loop delay (the total loop delay is: one NAND gate delay plus the delay of an odd number of inverters). The low-frequency RO oscillator circuit structure is the same as the high-frequency RO oscillator circuit composition structure, except that the low-frequency RO oscillator circuit loop includes more odd number of inverters. The first XOR operation circuit is an XOR gate, and the second XOR operation circuit is an XOR gate.

[0051] Embodiment 2

[0052] Combined withFigure 3 , the difference between this embodiment and the first embodiment is that the modulation circuit in this embodiment is a replacement unit, and the replacement unit is used to use high-entropy modulation data to replace the biological sequence data to obtain the neuralsequence data after privacy desensitization processing, which is irreversible.

[0053] Specifically, directly use the modulation data (random high-entropy data) to replace the data that originally needed to be modulated (low-entropy data). From the perspective of circuit implementation, it is to directly assign the data in the register bank storing the random number sequence to the register bank storing the original neuralsequence data; that is, directly transfer and copy the data between the register banks.

[0054] According to the privacy protection degree defined by the user, use the high-entropy data generated by the random number generator to replace the original data after XOR processing, that is, this modulation can perform a certain degree of random scrambling on the specified sensitive information part or local neural data in the original neuralsequence data; it can also use the high-entropy data generated by the random number generator to replace the original local low-entropy neural data to achieve the purpose of protecting the original neuralsequence data.

[0055] Embodiment Three

[0056] This embodiment discloses a method for protecting the privacy of neuralsequence data. Based on the above-mentioned circuit for protecting the privacy of sequence data, it includes the following steps:

[0057] S1. Cut the data frame of the original neuralsequence data to be protected into equal-length periodic segments. Specifically, as Figure 1 shown, divide the data frame of the original neural data into different periods: Period 1, Period 2, Period 3... Period K.

[0058] The cutting granularity parameter of the data frame of the original neuralsequence data to be protected can be adjusted. For example, the original neural information frame can be cut into 100 segments, or 1000 segments, etc.; if the cutting granularity is finer, that is, the number of cut periodic segments is more, and then high-entropy insertion modulation processing is performed within the period, the overall desensitization processing effect obtained is better; similarly, the coarser the cutting granularity, that is, the fewer the number of cut periodic segments, the worse the overall desensitization processing effect, but it saves the modulation operation within the period.

[0059] S2. For each periodic segment, select a position to insert a slot time slot according to the privacy protection degree; exemplarily, for each cut periodic segment, randomly select a slot time slot position within its period, that is, for different periodic segments, the selected slot time slot positions are randomly distributed within a single period.

[0060] In addition, a privacy protection degree "ε" is predefined. Within a single cycle, the ratio of the width of the slot time slot to the width of the period is the privacy protection degree "ε".

[0061] According to the predefined privacy protection degree, what proportion of the original neural sequence data is subjected to privacy processing, that is, the randomization processing ratio. If the privacy protection degree is 0, the original neural data is retained; if the privacy protection degree is 100%, privacy processing is performed on all the original neural sequence data; if the privacy protection degree is >0 and <100%, the original neural data is evenly divided into multiple cycles, and then within the random time slot inside each cycle, randomization processing is performed, and the ratio of the width of the time slot to the width of a single cycle is equal to the privacy protection degree.

[0062] S3. Obtain high-entropy modulation data, and obtain the privacy-desensitized neural sequence data according to the biological sequence data and the high-entropy modulation data at the position where the slot time slot is located. Specifically, perform an exclusive OR operation on the biological sequence data and the high-entropy modulation data at the position where the slot time slot is located to obtain the privacy-desensitized neural sequence data. Thus, the original data (low-entropy data) is inserted with high-entropy data and modulated to obtain the modulated binary sequence data, and the original data is reversible (that is, the original data can be restored).

[0063] Furthermore, in some embodiments, the biological sequence data at the position where the slot time slot is located is replaced with high-entropy modulation data to obtain the privacy-desensitized neural sequence data.

[0064] Embodiment Four

[0065] Embodiment Four of the present invention provides an electronic device, including a memory, a processor, and the above-mentioned neural sequence data privacy protection circuit.

[0066] In the above embodiments, the descriptions of each embodiment have their own emphases. For parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0067] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A neural sequence data privacy protection circuit, characterized in that Including: A true random number circuit, which is used to obtain an enable signal, find the jitter interval of the clock frequency according to the enable signal, and generate non-deterministic data and deterministic data; A first exclusive OR operation circuit, which is used to perform an exclusive OR operation on the non-deterministic data and the deterministic data, and collect the non-deterministic data as high-entropy modulation data; A shift register, which is used to store the high-entropy modulation data; A modulation circuit, which is used to randomly obtain the high-entropy modulation data in the shift register, and obtain the privacy-desensitized neural sequence data according to the high-entropy modulation data and the biological sequence data; wherein, the biological sequence data is the data at the position of the slot time slot in the original neural sequence data.

2. The neural sequence data privacy protection circuit according to claim 1, characterized in that The modulation circuit is a second exclusive OR operation circuit, which is used to randomly obtain the high-entropy modulation data in the shift register, perform an exclusive OR operation on the high-entropy modulation data and the biological sequence data, and obtain the privacy-desensitized neural sequence data.

3. The neural sequence data privacy protection circuit according to claim 2, wherein The second exclusive OR operation circuit is further used to perform an inverse exclusive OR operation on the privacy-desensitized neural sequence data and the biological sequence data to obtain the original neural sequence data.

4. The neural sequence data privacy protection circuit according to claim 2, wherein The true random number circuit includes a high-frequency oscillation loop circuit and a low-frequency oscillation loop circuit; The output end of the high-frequency oscillation loop circuit is connected to the input end of the low-frequency oscillation loop circuit, the output end of the low-frequency oscillation loop circuit is connected to the input end of the first exclusive OR operation circuit, the output end of the first exclusive OR operation circuit is connected to the input end of the shift register, and the output end of the shift register is connected to the input end of the second exclusive OR operation circuit.

5. The neural sequence data privacy protection circuit according to claim 1, characterized in that The modulation circuit is a replacement unit, which is used to replace the biological sequence data with the high-entropy modulation data to obtain the privacy-desensitized neural sequence data.

6. The neural sequence data privacy protection circuit according to claim 4, wherein, The high-frequency oscillation loop circuit is used to generate a high-frequency clock signal, and the low-frequency oscillation loop circuit is used to perform a first sampling and a second sampling on the jitter space of the high-frequency clock signal according to the low-frequency clock signal to obtain non-deterministic data and deterministic data.

7. A method for protecting the privacy of neural sequence data, characterized in that, The neural sequence data privacy protection circuit according to any one of claims 1-6 includes the following steps: Obtain the original neural sequence data, and divide the original neural sequence data into equal-length periodic segments; for each periodic segment, select a position to insert a slot time slot according to the privacy protection degree; Obtain the high-entropy modulation data, and obtain the privacy-desensitized neural sequence data according to the biological sequence data at the position of the slot time slot and the high-entropy modulation data.

8. The neural sequence data privacy protection method according to claim 7, wherein, The obtaining of the privacy-desensitized neural sequence data according to the biological sequence data at the position of the slot time slot and the high-entropy modulation data is: Perform an exclusive OR operation on the biological sequence data at the position of the slot time slot and the high-entropy modulation data to obtain the privacy-desensitized neural sequence data; Or, Use the high-entropy modulation data to replace the biological sequence data at the position of the slot time slot to obtain the privacy-desensitized neural sequence data.

9. The method for protecting the privacy of neural sequence data according to claim 7, characterized in that The privacy protection degree is the ratio of the width of the slot time slot to the width of the period.

10. An electronic device, characterized in that, It includes a memory, a processor, and a neural sequence data privacy protection circuit as described in any one of claims 1-6.

Images