Method and device for realizing double-entropy-source physical unclonable function

Abstract

The invention discloses an implementation method of a double-entropy-source physical unclonable function, which comprises a first entropy source unit used for generating a first random response and applying an SET voltage to a memristor in the first entropy source unit, so that high and low resistance states of the memristor have randomness, and a second entropy source unit used for generating a second random response; the current of the column where the memristor in the first entropy source unit is located generates random fluctuation as a first random response; the second entropy source unit is used for generating a second random response and applying RESET voltage to the memristor in the second entropy source unit, so that the fracture degree of a conductive filament of the memristor has randomness and the current of the column where the memristor is located generates random fluctuation, and the second random response is obtained; and the PUF function output unit is used for calculating the first random response and the second random response and then taking the first random response and the second random response as output response. Through the double-entropy-source physical unclonable function implementation method disclosed by the invention, the problems that the entropy source of the existing single-entropy-source physical unclonable function is insufficient in reliability and is easily attacked by machine learning are solved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of data security, and more particularly relates to a method and device for implementing a dual-entropy-source physical unclonable function. BACKGROUND

[0002] A physical unclonable function (PUF) is a secure primitive that generates unique responses by exploiting the random physical variations inherent in the microfabrication process. Each PUF instance, even if it is designed and manufactured identically, will produce a unique response to the same challenge due to its inherent physical uniqueness. This characteristic makes PUFs very suitable for security fields such as device identity authentication, key generation, and intellectual property protection. Existing PUFs mainly rely on a single entropy source: PUFs based on static manufacturing variations (such as path delay, SRAM power-on state) have good stability, but their fixed characteristics are vulnerable to machine learning modeling attacks, and their security is insufficient; while PUFs based on dynamic transient noise (such as thermal noise, power supply noise) have strong anti-modeling ability, but their response stability is poor and they are easily affected by environmental interference, and their reliability is low. Therefore, traditional single-entropy-source PUFs face the core contradiction of "high stability and weak security, high security and poor stability", and it is difficult to balance security and reliability, which limits their application in high-end security scenarios. SUMMARY

[0003] In view of the above defects or improvement needs of the prior art, the present application provides a method and device for implementing a dual-entropy-source physical unclonable function to solve the technical problem of insufficient entropy source reliability of existing single-entropy-source physical unclonable functions, which are vulnerable to machine learning attacks.

[0004] In order to achieve the above purpose, the present application provides a method and device for implementing a dual-entropy-source physical unclonable function, which comprises: A first entropy source unit for generating a first random response, a SET voltage is applied to the memristor in the first entropy source unit, so that the high and low resistance states of the memristor have randomness, and the randomness causes the current in the column where the memristor is located to have random fluctuations as the first random response; A second entropy source unit for generating a second random response, a RESET voltage is applied to the memristor in the second entropy source unit, so that the degree of breakage of the conductive filament of the memristor has randomness, and the randomness causes the current in the column where the memristor is located to have random fluctuations as the second random response; A PUF function output unit for operating the first random response and the second random response and outputting the response as the output of the dual-entropy-source physical unclonable function.

[0005] In general, the above technical solutions conceived by the present application can achieve the following beneficial effects: Very strong anti-modeling attack capability: since the final response is the result of nonlinear mixing of two independent and physically different entropy sources, the attacker needs to establish an accurate joint mathematical model for two completely different physical phenomena at the same time, which is extremely difficult in practice, greatly improving the security of the PUF.

[0006] Excellent environmental robustness: carefully selected entropy sources with complementary response characteristics to environmental factors. Through nonlinear mixing, the noise generated by a single entropy source due to environmental fluctuations can be smoothed or offset, thereby improving the consistency of the response, i.e. reliability, without relying on complex error correction codes.

[0007] Enhanced randomness and uniqueness: dual entropy sources provide a richer source of randomness, making the generated PUF response have a higher entropy value, better uniqueness and uniformity indicators, and are very suitable for generating encryption keys or device unique identifiers. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 A flowchart of a dual-entropy-source physical unclonable function implementation method provided for an embodiment of the present application.

[0009] Figure 2 A circuit diagram of a dual-entropy-source physical unclonable function device provided for an embodiment of the present application.

[0010] Figure 3 A memristor entropy source provided for an embodiment of the present application.

[0011] Figure 4 An internal circuit diagram of a dual-entropy-source physical unclonable function device provided for an embodiment of the present application. DETAILED DESCRIPTION

[0012] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0013] Embodiment 1: The present application discloses a dual-entropy-source physical unclonable function implementation method, comprising: The first entropy source unit is configured to generate a first random response, and a SET voltage is applied to the memristor in the first entropy source unit, so that the high and low resistance states of the memristor have randomness, and the randomness causes the current of the column in which the memristor is located to have random fluctuations as the first random response. The second entropy source unit is configured to generate a second random response, and a RESET voltage is applied to the memristor in the second entropy source unit, so that the degree of breakage of the conduction filament of the memristor has randomness, and the randomness causes the current of the column in which the memristor is located to have random fluctuations as the second random response. The PUF function output unit is configured to perform operation on the first random response and the second random response, and output the response as the output of the double-entropy-source physically unclonable function. Specifically, the first entropy source unit and the second entropy source unit each include a transistor and a memristor as shown in Figure 2 The first entropy source unit and the second entropy source unit can be the transistor and the memristor at any position shown in Figure 2 .

[0014] It should be noted that the memristors in the first entropy source unit and the second entropy source unit are gated and controlled by the X-axis drive unit and the Y-axis drive unit in Figure 2 . Specifically, the X-axis drive unit selects the row in which the memristor to be operated is located, and further, the Y-axis drive unit controls the column in which the memristor is located and the state of the memristor. In order to further illustrate the control process of the memristor in the entropy source unit, the following will be described in combination with Figure 2 .

[0015] As shown in Figure 2 , the hardware circuit for realizing a double-entropy-source physically unclonable function includes a memristor array, an X-axis drive unit, a Y-axis drive unit, and a PUF function output unit. The memristor array is a cross array, and the memristor array includes a plurality of entropy source units, and the entropy source units are cross arranged, and each entropy source unit includes a transistor and a memristor. The X-axis drive unit is used to gate the rows of the memristor array. Next, the connection relationship between an entropy source unit, the X-axis drive unit, the Y-axis drive unit, and the PUF function output unit will be described to describe the control process of the memristor, and the control and connection modes of other entropy source units are the same, and will not be described in detail to avoid redundancy. As shown in the figure, the gate of the transistor is connected with the X-axis drive unit, the upper electrode of the memristor is connected with the source / drain electrode of the transistor, the lower electrode of the memristor is grounded or connected with the ground terminal in the Y-axis drive unit, and the Y-axis drive unit is connected with the drain / source electrode of the transistor for controlling the state of the memristor. First, the row in which the memristor to be operated is located is selected by the X-axis drive unit, and then the selected memristor is programmed by the Y-axis drive unit. Specifically, as shown in Figure 4When memristors A and B implement a physically non-cloning function of a dual-entropy source, a voltage pulse is applied by the X-axis drive unit to ensure the transistor connected in series with memristor A is turned on. Subsequently, a negative voltage pulse is applied to the upper electrode of memristor A by the Y-axis drive unit, programming memristor A to a high-resistance state. For memristor B, a voltage pulse is applied by the X-axis drive unit to ensure the transistor connected in series with memristor B is turned on. Subsequently, a positive voltage pulse is applied to the upper electrode of memristor B by the Y-axis drive unit, programming memristor B to a low-resistance state.

[0016] Furthermore, the Y-axis drive unit applies a positive voltage pulse Vref to memristor A to perform a SET operation. Before performing a SET or RESET operation on the memristor, the X-axis drive unit selects the row containing the memristor. For example... Figure 3 As shown, when performing a SET operation on memristor A, the larger the applied voltage pulse, the greater the probability that memristor A will switch from a high-resistance state to a low-resistance state. When the voltage is the positive reference voltage Vref, the probability that memristor A will switch from a high-resistance state to a low-resistance state is 50%. If SET is successful, the generated current I in the column containing memristor A will increase; if it fails, the current in that column will remain unchanged. The current in the column containing memristor A is also called the first random response. This current signal is converted into a voltage signal by the first operational amplifier in the PUF function output unit connected to the column end, and serves as an input to the XOR gate in the PUF function output unit.

[0017] Secondly, the row containing memristor B to be operated is selected by the X-axis drive unit, and the Y-axis drive unit applies a reverse reference voltage pulse -Vref to memristor B to perform a RESET operation. After the operation is completed, due to the breakage of the conductive filament, the current in the column containing memristor B will fluctuate. This fluctuating current in the column containing memristor B is also called the second random response. This fluctuating current is processed by the second operational amplifier in the PUF function output unit. On the one hand, it is compared with the reference current Iref, and on the other hand, it is converted into a voltage signal output, which serves as another input terminal of the XOR gate in the PUF function output unit. Finally, the results of the two physical entropy sources, also known as the first random response and the second random response, are mixed by the XOR gate to obtain the output result of the PUF, which is the output result of the dual-entropy source physical non-cloning function.

[0018] It should be noted that memristors A and B can be located at any position in different rows and columns. In a hardware circuit that implements a physically non-clonable function of a dual-entropy source, the number of memristors must be at least two, while there is no limit to the number of memristors in a cross array.

[0019] Further, the first random response and the second random response can also be mixed by an operation mode other than the XOR operation, including but not limited to an encoder operation. Secondly, before the XOR gate operation, the first random response is input to the inverse input end of the first comparator in the PUF function output unit, the forward input end of the first comparator is grounded, the output of the first comparator is connected with one input end of the XOR gate, the second random response is input to the inverse input end of the second comparator in the PUF function output unit, the forward input end of the second comparator is connected with the reference current of the memristor in the second entropy source unit, the output end of the second comparator is connected with the other input end of the XOR gate, and the output of the XOR gate is taken as the XOR operation output.

[0020] It should be noted that the operation of the memristor includes SET and RESET operations, and the feedback state of the memristor is different under different operations. Figure 3 As shown in the figure, when the SET operation is performed on the memristor A, the horizontal axis is the voltage applied to the memristor A, and when the applied voltage reaches the forward reference voltage Vref, the probability is 50%; when the RESET operation is performed on the memristor B multiple times, the memristor B can achieve the maximum number of times of the on current Iref.

[0021] It should be noted that when the double-entropy source physically unclonable function is generated by selecting the memristor A and the memristor B, the initialization operation is also performed on the memristor A and the memristor B, for example, the memristor A is programmed to a high resistance state, and the memristor B is programmed to a low resistance state. For details of the process of realizing the double-entropy source physically unclonable function, reference can be made to Figure 1 As shown in the figure, after the initialization of the memristor A and the memristor B to be operated, the SET and RESET operations are performed on the memristor A and the memristor B respectively, and then the random response on the memristor A and the memristor B is read by the Y-axis driving unit, for example, the current fluctuation, and the random response is output after operation, so as to obtain the output of the double-entropy source physically unclonable function, as shown in the figure. Figure 1 As shown in the figure, the method flow chart for realizing the double-entropy source physically unclonable function is taken as an example of the memristor A and the memristor B.

[0022] Embodiment 2: The application discloses a device for realizing the double-entropy source physically unclonable function in embodiment 1, as shown in the figure. Figure 2As shown, the device of the dual-entropy source physically unclonable function comprises a memristor array, an X-axis driving unit, a Y-axis driving unit and a PUF output control unit, wherein the memristor array comprises a plurality of entropy source units arranged in a cross array, each entropy source unit comprising a transistor and a memristor, the X-axis driving unit is used to select the memristor in the row of the memristor array, and the Y-axis driving unit is used to control the state of the memristor in the column of the selected memristor, wherein when the memristor in the first entropy source unit is applied with a SET voltage, the high and low resistance states of the memristor have randomness, and the randomness causes the current in the column of the memristor in the first entropy source unit to have random fluctuations as a first random response; When the memristor in the second entropy source unit is applied with a RESET voltage, the degree of rupture of the conduction filament of the memristor has randomness, and the randomness causes the current in the column of the memristor to have random fluctuations as a second random response; and The PUF function output unit is used to operate the first random response and the second random response, and output the response as the output of the dual-entropy source physically unclonable function.

[0023] The control mode of the memristor in the specific entropy source unit is described in Embodiment 1, and will not be described in detail again.

[0024] Those skilled in the art can easily understand that the above description is only a preferred embodiment of the present application, and is not used to limit the present application, and any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for implementing a dual-entropy source physically unclonable function, characterized in that, include: The first entropy source unit is used to generate the first random response. A SET voltage is applied to the memristor in the first entropy source unit, so that the high and low resistance states of the memristor are random. The randomness causes the current in the column where the memristor is located in the first entropy source unit to fluctuate randomly as the first random response. The second entropy source unit is used to generate a second random response. A RESET voltage is applied to the memristor in the second entropy source unit, so that the degree of breakage of the conductive filament of the memristor is random. The randomness causes the current in the column where the memristor is located to fluctuate randomly, and the random fluctuation serves as the second random response. The PUF function output unit is used to perform calculations on the first random response and the second random response, and outputs the response as the output of the dual-entropy source physical non-cloning function.

2. The method for implementing a dual-entropy source physically unclonable function according to claim 1, characterized in that, It also includes programming the memristor in the first entropy source unit to a high-resistance state and programming the memristor in the second entropy source unit to a low-resistance state.

3. The method for implementing a dual-entropy source physically unclonable function according to claim 1, characterized in that, The operations include XOR operations or encoder operations.

4. The method for implementing a dual-entropy source physically unclonable function according to claim 3, characterized in that, The XOR operation includes inputting a first random response to the inverting input of a first comparator, grounding the non-inverting input of the first comparator, connecting the output of the first comparator to one input of an XOR gate, inputting a second random response to the inverting input of a second comparator, connecting the non-inverting input of the second comparator to the reference current of the memristor in the second entropy source unit, connecting the output of the second comparator to the other input of the XOR gate, and using the output of the XOR gate as the XOR operation output.

5. The method for implementing a dual-entropy source physically unclonable function according to claim 1, wherein, The first entropy source unit and / or the second entropy source unit include a transistor and a memristor, wherein the upper electrode of the memristor is connected to the source / drain of the transistor, the lower electrode of the memristor is grounded, the gate of the transistor is connected to the X-axis driving unit, and the drain / source of the transistor is connected to the Y-axis driving unit.

6. The method for implementing a dual-entropy source physically unclonable function according to claim 5, wherein, The X-axis drive unit is used to select the memristor in the row of the memristor array, and the Y-axis drive unit is used to select the memristor in the column of the memristor array.

7. The method for implementing a dual-entropy source physically unclonable function according to claim 6, wherein, The memristor array includes multiple entropy source units arranged in a cross array, and each entropy source unit includes a memristor and a transistor.

8. The method for implementing a dual-entropy source physically unclonable function according to claim 6, wherein, The Y-axis drive unit applies a voltage pulse with an amplitude equal to the positive reference voltage to the memristor to perform a SET operation; the Y-axis drive unit applies a voltage pulse with an amplitude equal to the reverse reference voltage to the memristor to perform a RESET operation.

9. The method for implementing a dual-entropy source physically unclonable function according to claim 8, wherein, Before the Y-axis drive unit performs a SET or RESET operation on the memristor, it selects the row where the memristor is located through the X-axis drive unit.

10. An apparatus for implementing the physically unclonable dual-entropy source function as described in any one of claims 1-9, characterized in that, The device includes a memristor array, an X-axis drive unit, a Y-axis drive unit, and a PUF output control unit. The memristor array includes multiple entropy source units arranged in a cross array. Each entropy source unit includes a transistor and a memristor. The X-axis drive unit is used to select the memristor in the row of the memristor array, and the Y-axis drive unit is used to control the state of the memristor in the column where the selected memristor is located. When a SET voltage is applied to the memristor in the first entropy source unit, the high and low resistance states of the memristor are random. This randomness causes the current in the column where the memristor in the first entropy source unit is located to fluctuate randomly as a first random response. When a RESET voltage is applied to the memristor in the second entropy source unit, the degree of breakage of the conductive filament of the memristor is random. This randomness causes the current in the column where the memristor is located to fluctuate randomly, and this random fluctuation serves as the second random response. The PUF function output unit is used to perform calculations on the first random response and the second random response, and outputs the response as the output of the dual-entropy source physical non-cloning function.

Images