A physically unclonable function chip and key generation method based on the statistical properties of prime number spectrum manifolds.

By mapping chip physical deviations to coordinate points on the prime number spectrum manifold and utilizing its statistical properties for deviation correction and response enhancement, the problems of unstable, insufficient uniqueness, and vulnerability to attack in PUF responses are solved, achieving stable, unique, and secure PUF response generation.

CN123069461APending Publication Date: 2026-11-10王颂斌
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611243942.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-17
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing PUF schemes are unstable in response to environmental changes, have non-ideal uniqueness distributions, lack support from first principles of number theory, are susceptible to machine learning modeling attacks, and lack applications of the statistical properties of prime number spectrum manifolds.

Method used

The chip's physical deviations are mapped to coordinate points on the prime number spectrum manifold. The statistical properties of this manifold are used for deviation correction and response enhancement, generating a stable, unique, and model-resistant PUF response. Secure keys are then generated through error correction coding and key derivation.

Benefits of technology

It maintains PUF response stability under different environmental conditions, reduces bit error rate, improves chip-to-chip uniqueness, is supported by first principles of number theory, resists modeling attacks, is low in cost and compatible with standard CMOS processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123069461A_ABST
    Figure CN123069461A_ABST
Patent Text Reader

Abstract

This invention discloses a physically unclonable function chip and key generation method based on the statistical properties of prime number spectrum manifolds, belonging to the field of cryptographic hardware and hardware security technology. This invention maps random physical deviations during chip manufacturing to coordinate points on a prime number spectrum manifold, utilizing the inherent statistical properties of the prime number spectrum manifold as a mathematical reference for deviation correction. This generates a stable, unique, unclonable, and modeling-attack-resistant chip fingerprint, and outputs a secure key through a key derivation function. This invention maintains stability (with a low bit error rate) under different temperature and voltage conditions, and the Hamming distance between chips approaches the ideal value. It is suitable for scenarios such as chip authentication, secure key generation, and IoT device identification, and can be integrated as a hardware root of trust into cryptographic hardware acceleration architectures.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of cryptographic hardware and hardware security technology, and is particularly applicable to the implementation of physically unclonable functions (PUFs) in scenarios such as chip identity authentication, security key generation, anti-counterfeiting and traceability, IoT device identity recognition, and blockchain hardware wallets. Specifically, it relates to a PUF chip and method that utilizes the statistical properties of prime number spectrum manifolds to extract random physical deviations introduced during chip manufacturing and generate stable, unique, and unclonable chip fingerprints. Background Technology

[0002] Physically unclonable functions (PUFs) utilize unavoidable process variations in chip manufacturing (such as threshold voltage fluctuations, linewidth variations, and differences in doping concentration) to generate a unique "physical fingerprint" for each chip. Existing PUF implementations have the following inherent limitations: Insufficient stability of PUF response: Existing PUFs based on ring oscillators, SRAM, and arbitrators are sensitive to temperature and voltage changes. The PUF response of the same chip under different environmental conditions may exhibit bit flipping, leading to key generation failure or authentication errors. The uniqueness distribution of PUF response is not ideal: the Hamming distance between chips in an ideal PUF should be close to 50%, but in existing PUF schemes, the response correlation of some chip pairs is high, and the uniqueness distribution is biased. The randomness of PUF response lacks support from the first principles of number theory: the randomness of existing PUFs originates from the statistical distribution of physical process deviations, lacks a correspondence with the underlying structure of number theory, and it is difficult to prove its unpredictability mathematically. PUF is threatened by machine learning modeling attacks: Arbitrator-based PUF and other schemes have been shown to be vulnerable to machine learning modeling attacks, where attackers can predict unknown responses by collecting a limited number of challenge-response pairs. There is a lack of PUF schemes based on the statistical properties of prime spectrum manifolds: the distribution of prime points on prime spectrum manifolds has been shown to have deterministic statistical properties, but it has not yet been used to enhance the stability and unpredictability of PUF responses. This invention proposes a PUF chip based on the statistical properties of prime number spectrum manifolds. It maps the chip's physical deviations to the coordinate point distribution on the prime number spectrum manifold, and uses the statistical properties of the prime number spectrum manifold as the mathematical basis for natural deviation correction and stability enhancement to generate a stable, unique, unclonable, and modeling-resistant chip fingerprint. Summary of the Invention

[0003] Purpose of the invention This invention aims to provide a physically unclonable function chip and key generation method based on the statistical properties of prime number spectrum manifolds. By mapping the physical deviation of the chip to coordinate points on the prime number spectrum manifold, the statistical distribution characteristics of the prime number spectrum manifold are used to correct the deviation and enhance the response, generating a stable and unclonable chip fingerprint supported by the first principles of number theory, which can be used as a unique identifier for the device or a secure key source. Technical solution This invention is based on the following mathematical fact: prime number spectrum manifold The distribution of prime points has definite statistical properties and can serve as an "ideal reference distribution." Random physical deviations introduced during chip manufacturing can be extracted using an on-chip sensor array and mapped to a distribution of coordinate points on the prime spectral manifold. By comparing this distribution with the ideal reference distribution, deviations caused by environmental changes can be corrected, generating a stable PUF response. This PUF chip comprises six core components: Physical Deviation Extraction Array: Configured for deployment on the chip Each physical sensor unit contains a ring oscillator (RO), SRAM cell, or delay chain to extract random physical deviations introduced during chip manufacturing. Each sensor unit outputs a characteristic value. (e.g., oscillation frequency, threshold voltage, delay time). Prime Spectrum Manifold Mapping Engine: Configured to map the eigenvalues ​​of each sensor unit Mapped to a prime spectral manifold coordinates on ,in , , This is the preset bit width. Ideal reference distributed storage unit: configured to store data from the front end. The theoretical reference distribution mapped from prime numbers The reference distribution is an intrinsic statistical distribution on the prime number spectrum manifold, and is independent of the specific chip. Deviation correction and PUF response generation module: configured to distribute the coordinate points Compared with the ideal reference distribution Compare and calculate the deviation of each coordinate point. Generate a stable binary PUF response based on the deviation. .when When it exceeds the preset threshold, ;otherwise . Error correction and auxiliary data generation module: Configured to apply error correction coding (such as BCH code, Reed-Muller code, or soft-decision LDPC code) to the PUF response, generating auxiliary data for subsequent response reconstruction, improving the stability of the response under different environmental conditions. The auxiliary data can be securely and publicly stored without disclosing the confidential information of the PUF response. Key Derivation and Output Module: Configured to convert the corrected PUF response into a fixed-length secure key or device identifier via a key derivation function (KDF, such as HKDF-SHA256), and output it through a secure interface. A random salt and context identifier are added during the derivation process to prevent the same PUF response from deriving the same key in different applications. Overall Process: Chip Power-On → Physical Deviation Extraction Array Sampling → Prime Number Spectrum Manifold Mapping → Comparison with Ideal Reference Distribution → Deviation Correction → Error Correction Coding → Key Derivation → Output Unique Chip Fingerprint PUF response stability guarantees: - After physical deviations are mapped to the prime number spectral manifold, offsets caused by environmental changes (temperature, voltage) manifest as global translations in coordinate space, rather than local perturbations; - The effects of global translations can be offset by comparison with an ideal reference distribution; - Error-correcting coding further corrects residual bit flips, reducing the bit error rate to... the following. Security Guarantees: - The randomness of the PUF response stems from the inherent statistical properties of the distribution of prime points on the prime spectrum manifold, supported by the first principles of number theory; - The irreversibility of the prime spectrum manifold mapping prevents attackers from reverse-engineering the original PUF response from publicly available auxiliary data; - Machine learning modeling attacks require training data related to the coordinate distribution of the prime spectrum manifold, which is mathematically fixed and cannot be modeled through finite sampling. Beneficial effects Bias correction based on the statistical properties of prime number spectrum manifolds ensures that the PUF response remains stable under different temperature (-40°C to 125°C) and voltage (0.8V to 1.2V) conditions, with a bit error rate lower than [missing value]. ; The Hamming distance distribution between chips approaches the ideal value. Its uniqueness is superior to traditional PUF solutions; The randomness of PUF responses is supported by the first principles of number theory, and its resistance to modeling attacks is superior to that of PUF based on arbitrators. It is compatible with standard CMOS processes, requires no additional special process steps, and has low cost. The entire process of deviation extraction, mapping, correction, and key derivation is integrated on-chip, requiring no external intervention; It can be integrated into a SoC as a standalone security chip or IP core. Detailed Implementation Example 1: PUF chip based on ring oscillator array Step R1, RO array design: Deployment on the chip Each RO unit consists of an odd number of inverters connected in series, and its oscillation frequency is affected by manufacturing process variations. Each RO unit outputs its oscillation frequency value. . Step R2, Frequency-Prime Spectrum Mapping: Map each frequency value Mapped to prime number spectrum manifold coordinates ,in , . Step R3, Reference Distribution Loading: Load the pre-computed ideal reference distribution from the on-chip ROM. (from the previous) (The coordinates of prime numbers). Step R4, Response Generation: Calculate the minimum toroidal distance between each chip coordinate and the ideal reference distribution. ,like Then output bits Otherwise, output bits. ,generate Bit PUF response. Step R5, Error Correction and Key Derivation: For The bit response is applied using BCH error correction code to generate Bit auxiliary data and A secure key. Step R6, Chip Testing: Measure the stability of the PUF response at different temperatures (-40°C, 25°C, 85°C, 125°C) and measure the uniqueness among different chips. Technical effects: The average Hamming distance between the test chips is Standard deviation Average bit flip rate of the same chip at different temperatures Error rate after error correction . Example 2: PUF chip based on SRAM power-on state (low power consumption scenario) Step S1, SRAM Array Design: A standard SRAM cell array is used, utilizing the uncertain state of the SRAM cells upon power-on (affected by threshold voltage deviation) as a source of physical entropy. Compared to a RO array, the SRAM PUF reduces power consumption by approximately [missing information]. Suitable for IoT and battery-powered scenarios. Step S2, Power-on Status Acquisition: After the chip is powered on, read... The initial state value of the SRAM cell is used as the feature value sequence. . Step S3, Prime Spectrum Manifold Mapping and PUF Response Generation: ... The SRAM state values ​​are mapped to prime spectral manifold coordinates, compared with an ideal reference distribution to generate PUF responses, and then processed by BCH error correction codes for output. A secure key. Step S4, Low Power Optimization: Each key generation process requires only one SRAM power-on read and mapping calculation, resulting in low total energy consumption. μJ. Example 3: Co-integration with RH00400 cryptographic hardware acceleration architecture Step C1, IP Integration: Integrate this PUF chip as an IP core into the RH00400 cryptographic hardware acceleration architecture, serving as the hardware root of trust for this architecture. Step C2, Key Supply: The device-unique key generated by the PUF chip is directly supplied to the key negotiation module (RH00204) and digital signature module (RH00203) of RH00400, realizing a full hardware key management link "from physical deviation to cryptographic key", without the need for external key injection. Step C3, Secure Boot: After the chip is powered on, the PUF key is first derived and used to verify the signature of the bootloader, ensuring that only authorized firmware can boot. Step C4, Comprehensive Testing: Verify the stability of the PUF key under different environmental conditions, and its functional correctness in various cryptographic operations of RH00400. Technical effect: Bit error rate of PUF key It meets the key quality requirements of each cryptographic module of RH00400 and achieves fully hardware-secure boot without the need for external key injection. Terminology Explanation

[0004] Physically Unclonable Functions (PUFs): Hardware security primitives that utilize unavoidable physical deviations during chip manufacturing to generate a unique and unclonable "physical fingerprint" for each chip. Statistical properties of the prime number spectrum manifold: prime numbers in the torus The distribution of coordinate points after upmapping has a definite statistical regularity and can be used as an ideal reference distribution for PUF response generation. Deviation correction: After mapping the chip's physical deviation to the prime number spectrum manifold, the overall offset caused by environmental changes is eliminated by comparing it with the ideal reference distribution, thereby improving the stability of the PUF response. Attached Figure Description

[0005] Figure 1 This is a block diagram of the overall architecture of the PUF chip; Figure 2 This is a schematic diagram comparing the physical deviation mapping with the reference distribution of the prime number spectrum manifold; Figure 3 Flowchart for PUF response generation and error correction coding; Figure 4 This is a comparison chart of PUF response stability tests at different temperatures.

Claims

1. A chip for physically unclonable functions based on the statistical properties of prime number spectrum manifolds, characterized in that, include: The physical deviation extraction array, comprising multiple physical sensor units, is configured to extract random physical deviations introduced during chip manufacturing. Each sensor unit outputs a feature value. ; A prime number spectrum manifold mapping engine, configured to map the eigenvalues Mapped to a prime spectral manifold coordinates on ,in , ; Ideal reference distributed storage unit, configured to store data from the front end. The theoretical reference distribution mapped from prime numbers The deviation correction and PUF response generation module is configured to compare the chip coordinate distribution with the theoretical reference distribution to generate a stable binary PUF response; the error correction and auxiliary data generation module is configured to apply error correction coding to the PUF response and generate auxiliary data for subsequent response reconstruction; the key derivation and output module is configured to convert the corrected PUF response into a fixed-length security key or device identifier through a key derivation function and output it through a secure interface.

2. The chip according to claim 1, characterized in that, The physical sensor unit is at least one of a ring oscillator unit, an SRAM unit, a delay chain unit, or a threshold voltage detection unit.

3. The chip according to claim 1, characterized in that, The theoretical reference distribution is from the previous Coordinates of prime numbers Composition, in which , , To preset irrational parameters, .

4. The chip according to claim 1, characterized in that, The deviation correction and PUF response generation module generates the PUF response by calculating the minimum toroidal distance between each chip coordinate and the theoretical reference distribution. ,like Then output bits Otherwise, output bits. ,in This is a preset threshold.

5. The chip according to claim 1, characterized in that, The error correction code is at least one of BCH code, Reed-Muller code, or soft-decision LDPC code.

6. The chip according to claim 1, characterized in that, The bit error rate of the PUF response of the chip is lower than [value missing] under different temperature (-40°C to 125°C) and different voltage (0.8V to 1.2V) conditions. .

7. The chip according to claim 1, characterized in that, The physical deviation extraction array, prime number spectrum manifold mapping engine, deviation correction and PUF response generation module are integrated on the same chip to realize full hardware PUF key derivation.

8. The chip according to claim 1, characterized in that, The auxiliary data can be stored publicly without disclosing the confidential information of the PUF response.

9. A method for generating a physically unclonable function key based on the statistical properties of prime number spectrum manifolds, applied to the chip described in claims 1-8, characterized in that... Includes the following steps: S1. Extract the physical feature value sequence of the array acquisition chip through physical deviation. ; S2. Map the eigenvalue sequence to a coordinate point distribution on a prime number spectrum manifold; S3. Load the pre-stored theoretical reference distribution of the prime number spectrum manifold; S4. Compare the coordinate point distribution with the theoretical reference distribution to generate a PUF response; S5. Apply error correction coding to the PUF response to generate auxiliary data; S6. Convert the corrected PUF response into a security key or device identifier through a key derivation function and output it.

10. The method according to claim 9, characterized in that, The security key output by the method is directly supplied to the key negotiation module or digital signature module in the cryptographic hardware acceleration architecture, and no external key injection is required.