IoT (Internet of Things) sensing layer lightweight encryption method and system based on PUF (Physical Unclonable Function) and chaotic key generation

Abstract

The invention discloses an IoT (Internet of Things) sensing layer lightweight encryption method and system based on PUF (Physical Unclonable Function) and chaotic key generation, and relates to the technical field of Internet of Things security. When equipment is powered on, an SRAM (Static Random Access Memory) PUF is utilized to generate a unique and unclonable hardware random source, a stable chaotic seed sequence is recovered in combination with BCH (Broadcast Channel) fuzzy extraction, Helper Data which does not contain sensitive information is subjected to on-chain index management, and the Helper Data is subjected to on-chain index management; realizing cross-device consistency of key parameters and device identity binding; in addition, a dynamic chaos parameter updating and exception handling mechanism is constructed, when parameter failure or side channel attack risk is detected, PUF depth resampling is triggered, and safety parameters are synchronously recovered through on-chain increments. A floating point unit is not needed in the whole process, and the high-safety and low-power-consumption real-time encryption requirements of resource-limited IoT sensing layer equipment such as medical treatment, physiological monitoring and supply chains can be met.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) security technology, and more specifically to a lightweight encryption method and system for the IoT sensing layer based on PUF and chaotic key generation. Background Technology

[0002] Currently, the Internet of Things (IoT) is being deeply applied in fields such as healthcare, finance, and industry. IoT sensing layer devices, such as physiological data sensors, logistics information terminals, and industrial data collectors, have become the core carriers for sensitive data collection. More than 70% of these devices are resource-constrained, characterized by low computing power (8-bit / 16-bit MCUs), low power consumption (microjoules / operation), and small storage (KB level). Their security protection faces significant challenges, including prominent key management risks, poor algorithm and hardware compatibility, and fragmented security systems.

[0003] While there are improvement solutions in the industry, they all have limitations: pure software chaotic encryption relies on floating-point operations, which increases energy consumption sharply; a single PUF solution is only used for identity identification and does not involve key generation; the consortium blockchain + traditional encryption fusion solution has not optimized the compatibility of energy consumption and computing power, resulting in excessively high latency for on-chain interactions.

[0004] Therefore, there is a need for a lightweight encryption scheme that can run on resource-constrained devices and combine PUF features to achieve hardware-level identity binding, while also supporting on-chain verification in necessary scenarios to reduce the risks associated with key management and improve transmission security. This is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of the above problems, this invention proposes a lightweight encryption method and system based on PUF and chaotic mapping to generate key streams. This scheme utilizes the device's existing SRAM PUF as a source of randomness and meets the practical application needs of resource-constrained devices through a lightweight SIT encryption process. It also addresses device identity binding and on-chain verification, resolving the aforementioned pain points and meeting the security and adaptability requirements of scenarios such as cross-hospital physiological data sharing in healthcare and financial supply chain logistics information transmission.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, embodiments of the present invention provide a lightweight encryption method for the IoT sensing layer based on PUF and chaotic key generation, comprising the following steps: S1. PUF Response Generation: After the IoT sensing layer device is powered on, it generates a unique and unclonable original response through the device's built-in PUF module. S2. Seed Stabilization and Auxiliary Data Management: The original response is input into the BCH fuzz extractor, which outputs a random chaotic seed sequence and auxiliary data for seed extraction; the auxiliary data is associated according to the unique identifier of the sending device and uploaded to the consortium blockchain; S3. Chaotic Keystream Generation: Map the chaotic seed to the initial parameters of the chaotic system, preset the effective domain, perform domain detection on the initial parameters, perform discrete chaotic iteration based on the compliant initial parameters, and generate the original keystream; perform bit extraction and debiasing processing on the original keystream to obtain a compliant chaotic keystream that meets the randomness requirements; S4.SIT Lightweight Encryption and Tag Generation: The compliant chaotic key stream is divided into a dynamic S-Box seed and round keys. The dynamic S-Box is used to perform multiple rounds of SIT encryption operations on the sensitive data collected by IoT sensing layer devices and output ciphertext. At the same time, a verification tag is generated and the ciphertext and verification tag are transmitted to the target node. S5. Decryption and Verification: After receiving the ciphertext, the target node retrieves the corresponding auxiliary data from the consortium blockchain index table and inputs it along with the local PUF response into the BCH fuzz extractor. Then, it executes S3 to generate a chaotic key stream consistent with the sender's, while simultaneously verifying the consistency between the verification tag and the locally generated chaotic key stream. After successful verification, it performs the SIT inverse operation based on the compliant chaotic key stream to decrypt and obtain the original sensitive data.

[0007] Furthermore, the method of S1 is as follows: The PUF module in S1 is an SRAM PUF module, and the original response R PUF bit length L PUF Determined based on the device's SRAM capacity, to meet the requirements. ,and Original response R PUF The randomness is determined by the threshold voltage V th Channel length L ch The determination, possessing the physical non-cloning property, means that for any two different IoT sensing layer devices, the probability Pr that their PUF responses are the same satisfies:

[0008] in, , These are the original PUF responses generated by the two devices, respectively.

[0009] Furthermore, the method of S2 is as follows: The specific process in S2 of inputting the original response into the BCH fuzzy extractor and outputting a random chaotic seed sequence and auxiliary data for seed extraction includes: The BCH(n,k,t) code pattern is adopted, where n is the code length, k is the information bit length, and t is the number of error correction bits, and the value of t satisfies the following conditions: Furthermore, t needs to be adapted to a noise rate of 1%-5% of the original response, i.e.:

[0010] in L represents the floor function. PUF The bit length of the original response; The chaotic seed sequence is generated by BCH encoding to produce a check bit, and then the original response R is detected based on this check bit. PUF Bit errors in the signal are corrected, and after effectively eliminating the noise in the original response, the final output is a stable and reproducible sequence. Length of auxiliary data satisfy (Unit: bits), and the auxiliary data only contains the parity information of the BCH code, with a parity bit length of [missing information]. ,Right now ; Simultaneously, the auxiliary data is associated with the unique identifier of the sending device and uploaded to the consortium blockchain. The consortium blockchain stores and records the version of the auxiliary data, including the following processing: 1) Association Binding Process: The auxiliary data obtained from BCH error correction is associated with the unique identifier Dev of the sending device. ID Hash value Dev ID Perform an XOR join to generate associated data HD. assoc : Where HD represents auxiliary data and Dev represents Dev. ID A hash value representing the unique identifier of the sending device. This represents the XOR operation; 2) On-chain storage and index management: Related data is uploaded to consortium blockchain nodes for distributed storage; each consortium blockchain node constructs an index key based on the hash value of the sending device's unique identifier, where: And record the version number of HD and its corresponding PUF response hash by index key; 3) Anti-replay control: Based on versioned on-chain index management, it ensures that the IoT sensing layer device at the receiving end can only obtain the currently valid version of HD.

[0011] Furthermore, the S3 method is as follows: In S3, the preset effective domain is: control parameter r0∈[3.80,3.99], initial value x0∈(0,1), and perturbation parameter a is a non-zero integer. Discrete chaotic iteration adopts parameter perturbation type Logistic mapping, and the iteration formula is:

[0012] in, This is the output value of the (n-1)th iteration. 'a' is the perturbation parameter obtained by mapping the chaotic seed sequence, the number of iterations N∈[32,64], and the floating-point value x is output in each iteration. n ∈(0,1), with a modulus m=2 32 Execute the floating-point value x n Quantization base y converted to a fixed-width integer n : After quantization to a 32-bit integer, bit extraction is performed; bit extraction extracts the middle 8 bits of the integer output, and the extraction rule can be expressed as:

[0013] in, This represents the floor function; The debiasing process uses the Von Neumann algorithm to process the bitstream. t is the number of groups, divided into groups of 2, and the output rules are as follows:

[0014] The uniformity of the 0 and 1 distribution of the debiased bitstream satisfies , where P r The probability of this event occurring is represented by bi′, which is a uniformly distributed set-off bit of b obtained by filtering b in groups of two using the Von Neumann algorithm. 2i and b 2i-1 It is the raw bitstream After dividing into groups of two, the two consecutive bits in the i-th group.

[0015] Furthermore, the chaotic keystream includes a chaotic seed sequence and chaotic initial parameters. Verification of the consistency between the verification label and the locally generated chaotic keystream includes: After power-on, the receiving device generates a local PUF response using its local Physically Unclonable Function (PNF) module. The target node retrieves the latest auxiliary data bound to the device from the consortium blockchain based on the unique device identifier carried in the ciphertext. and the local PUF response and Inputting the BCH fuzz extractor together recovers the chaotic seed sequence. ; chaotic seed sequence Calculate BLAKE2s hash And compared with the chaotic seed hash given by the sender in the verification label. A bit-by-bit comparison is performed; if the comparison matches, the sending device's identity is confirmed as legitimate. Simultaneously, the hash of the receiving device's unique identifier is calculated. It is then compared with the recipient binding field in the tag to confirm whether it is a legitimate authorized recipient; Verify whether the chaotic initial parameters conform to the preset valid domain of S3, and simultaneously verify the hash of the chaotic initial parameters generated by the target node. Hash of parameters in the verification label Whether they are consistent, the pass rate of this process Satisfies: Parameters have not been tampered with =1, after tampering ≈ .

[0016] Furthermore, the S4 method is as follows: In S4, the dynamic S-Box seed is used to generate a 4×4 bijective dynamic S-Box. The round key consists of multiple keys with the same length as the data blocks to be encrypted. Sensitive data collected by IoT sensing layer devices is divided into data blocks according to a preset block length of 64 bits, which is the length of the sensitive data block. It is compatible with short, sensitive data blocks collected in a single session by IoT sensing layer devices; the number of rounds of multi-round SIT encryption is 4; the length of the compliant chaotic key stream is 256 bits, of which 64 bits are used as a dynamic S-Box seed (divided into 16 segments of 4 bits each, denoted as s1, s2, ..., s). 16 Each segment is mapped to an S-Box item using modulo 16 operations, with the mapping rules as follows:

[0017] Where i,j∈[1,4], S(i,j) represents the matrix element of a 4×4 bijective dynamic S-Box, and i,j are the row and column indices of the dynamic S-Box matrix, thus obtaining a 4×4 dynamic S-Box. And it is ensured to be a bijective mapping by a no-repetition constraint. The remaining 256 bits are divided into 4 round keys of 64 bits each. , , Each corresponds to one of the four rounds of SIT encryption.

[0018] Furthermore, the S6 method is as follows: S6 system operation status anomaly monitoring: The specific aspects of S6 system operation status anomaly monitoring include: If the initial parameters output by the IoT sensing layer device do not conform to the preset valid domain described in S3 three times consecutively, the parameter compliance is determined to be abnormal; and the real-time power consumption P of the encryption operation is monitored simultaneously. real Compare it with the normal power consumption reference P normal Comparison: When the deviation exceeds the preset threshold ΔP th The power consumption behavior is judged to be abnormal at that time; When any abnormal condition is triggered, the control PUF module enters deep resampling mode, performs five raw response samplings on the SRAM PUF, calculates the noise rate of each sample, and selects the raw response with the lowest noise rate as the input for subsequent fuzzy extraction and key generation. The device sends an anomaly alert to the consortium blockchain node, including the device's unique identifier, Dev. ID and exception type Type alert After receiving the alarm, the consortium blockchain node generates new auxiliary data HD based on the latest PUF response. new and the old auxiliary data HD stored on the chain old The incremental data ΔHD is calculated using the following formula:

[0019] The consortium blockchain node will send the incremental packet ΔHD to the device that triggered the anomaly; The device performs auxiliary data synchronization locally, using the following formula:

[0020] in For the updated auxiliary data, For old auxiliary data, For the data after performing auxiliary synchronization locally, This represents the XOR operation.

[0021] Furthermore, methods such as any one of S1-S6 specifically include: Total energy consumption per encryption The energy consumption is 1-5 microjoules, and the total energy consumption consists of the energy consumption of each step, satisfying the following:

[0022] Where E PUF It is the energy consumption generated by the PUF response, E BCH This refers to the energy consumption for BCH fuzzy extraction, E. chaos It is the energy consumption of discrete chaotic iteration, E SIT It is the energy consumption of SIT encryption, and , , , Total time for a single encryption operation , (μs) (ms), For the device time window.

[0023] Furthermore, a lightweight encryption system for the IoT sensing layer based on PUF and chaotic key generation, such as S9, includes: Seed stabilization and auxiliary data management module: Inputs the original response into the fuzzy extractor and outputs a random chaotic seed sequence and auxiliary data for seed extraction; associates the auxiliary data according to the unique identifier of the sending device and uploads it to the consortium blockchain; Chaotic key stream generation module: Maps the chaotic seed to the initial parameters of the chaotic system, performs domain detection on the initial parameters, performs discrete chaotic iteration based on the compliant initial parameters, and generates the original key stream; performs bit extraction and debiasing processing on the original key stream to obtain a compliant chaotic key stream that meets the randomness requirements; The SIT lightweight encryption and tag generation module divides the compliant chaotic key stream into a dynamic S-Box seed and round keys. It uses the dynamic S-Box to perform multiple rounds of SIT encryption operations on the sensitive data collected by IoT sensing layer devices and outputs ciphertext. At the same time, it generates a verification tag and transmits the ciphertext and verification tag to the target node. Decryption and Verification Module: After receiving the ciphertext, the target node retrieves the corresponding auxiliary data from the consortium blockchain index table and inputs the local PUF response into the fuzzy extractor. Then, it executes S3 to generate a chaotic key stream consistent with the sender, while verifying the consistency between the locally generated chaotic seed sequence and the chaotic initial parameters and the verification label. After the verification is successful, it performs SIT inverse operation based on the compliant chaotic key stream to decrypt and obtain the original sensitive data. It also includes a system operation status anomaly monitoring module: when the chaos parameter verification fails or the on-chain auxiliary data version is inconsistent, the local auxiliary data is updated according to the incremental auxiliary data uploaded by the sending end.

[0024] The beneficial effects of the above-described technical solutions provided in the embodiments of the present invention include at least the following: 1) Hardware-level secure root of trust and keyless security paradigm: This invention uses the physically unclonable response generated by SRAM PUF as the unique entropy source of the encryption system, eliminating key cloning and device impersonation at the source. Combined with BCH fuzzy extraction and consortium blockchain distributed management of Helper Data, IoT devices do not need to store any static keys or sensitive intermediate data locally, completely eliminating the risk of key leakage and establishing a highly secure foundation for resource-constrained devices.

[0025] 2) Exceptional lightweight performance adapts to the stringent resource constraints of IoT sensing layer devices: The entire encryption process, including PUF response generation, chaotic key stream generation, and SIT encryption, is based on simple bitwise operations such as XOR and shifting, requiring no floating-point units and compatible with 8 / 16-bit MCUs. Through collaborative optimization of algorithms and hardware, the energy consumption of a single encryption is as low as 1-5 microjoules, with a time consumption of less than 0.5 microseconds, fully meeting the real-time and low-power requirements of sensing layer devices to complete "acquisition-encryption-transmission" within a millisecond time window.

[0026] 3) Forming a dynamic and verifiable proactive security defense system: This invention combines PUF with a chaotic system to generate a dynamic key that is unique and highly random for each session, and uses it to construct a dynamic S-Box, greatly improving the algorithm's resistance to analysis. Simultaneously, the introduced "verification tag" mechanism, in collaboration with the consortium blockchain, can perform dual verification of device identity and parameter integrity at the decryption end. The integrated anomaly handling module can proactively monitor and respond to threats such as side-channel attacks, achieving self-repair by triggering PUF resampling and on-chain security synchronization, significantly enhancing robustness in complex environments. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0028] Figure 1 This is a flowchart of a lightweight encryption method for the IoT sensing layer based on PUF and chaotic key generation, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of a lightweight encryption system for the IoT sensing layer based on PUF and chaotic key generation, provided in an embodiment of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] This invention discloses a lightweight encryption method for the IoT sensing layer based on PUF and chaotic key generation. This embodiment uses a medical wrist-worn physiological sensor as the implementation subject. It needs to encrypt the collected physiological data such as heart rate and blood oxygen saturation before transmitting it to the hospital's internal edge gateway. The specific implementation steps are as follows: S1. After the IoT sensing layer device powers on, its microcontroller (MCU) calls and executes the SRAM PUF driver instructions built into the firmware. After completing basic clock and I / O initialization, the MCU delays for 1-10 microseconds to ensure that the SRAM cells reach a stable power-on state. After the device powers on, these SRAM cells are initialized to a length of L. PUF =128 bits of raw PUF response RPUF .

[0031] This response is highly unique. For any two different IoT sensing layer devices, even those of the same model and from the same batch, their generated PUF response R will be identical. PUF1 With R PUF2 Difficult to replicate. This characteristic ensures the uniqueness of each device's root key, providing a hardware-based security foundation for the entire encryption system from the source. IoT sensing layer devices include resource- and computing-constrained embedded devices such as medical sensors and logistics terminals.

[0032] S2, After successfully generating the original PUF response R PUF Then, it is processed to remove noise and generate a reproducible key seed. The key seed obtained from S1... The input is fed into a BCH fuzzy extractor implemented in embedded software. This embodiment uses the BCH(255,131,18) code to correct errors in the input PUF response. After the extractor runs, the check bits are used to check the R... PUF The random bit errors in the sequence are corrected, and a stable, reproducible chaotic seed sequence R is finally output. seed sequence.

[0033] During the fuzzy extraction process, the system simultaneously generates crucial auxiliary data HD. This HD consists of the parity bits of the BCH code, and its length is L. HD =n k=124 bits, satisfying HD {0,1} 124 It does not involve the original response R. PUF Or chaotic seed sequence R seed Sensitive content does not require encryption when uploaded to the consortium blockchain. Next, to ensure the security and verifiability of helper data, the system initiates an on-chain management process. This process first establishes an association, linking the HD (Hardware Address) with the device's unique identifier, Dev. ID Obfuscation is performed using XOR operations to generate associated data HD. assoc :

[0034] By strongly binding Helper data to specific devices, the true HD is effectively hidden.

[0035] The HD binding will then be completed. assoc The data is uploaded to the consortium blockchain nodes for distributed storage. After receiving the data, the on-chain nodes create an index for the device: Index=H(Dev) ID (where H( (using the SHA-256 hash function), and HD assocThis index is associated with storage, and its version number V is also recorded. HD .

[0036] Finally, after confirming HD assoc After successfully and securely connecting to the blockchain, the IoT device immediately performs local data cleanup, completely erasing R from the volatile memory. PUF HD, R seed All intermediate sensitive data.

[0037] To facilitate the management of Helper data and prevent the reuse of old data, a consortium blockchain is used for its storage and version recording. The specific management process is as follows: 1) Association Binding: Associating the generated HD with the device's unique identifier, Dev. ID The associated data (HD) is generated by performing an XOR operation. assoc :

[0038] This operation associates Helper Data with the device identifier, and due to the information hiding properties of the XOR operation, HD assoc HD cannot be launched without knowing the device identifier hash value, thus preventing the leakage of original Helper data.

[0039] 2) On-chain storage and indexing: HD assoc Uploaded to consortium blockchain nodes for distributed storage. Consortium blockchain nodes are based on the device's unique identifier, Dev. ID An index table is built using the hash value, and the index key is calculated as follows:

[0040] Where H( The index is a SHA-256 hash function. This index is used to bind Helper Data version V. HD and the device PUF response hash H(R) PUF Establish a corresponding relationship. Furthermore, HD assoc It is merely a mixture of Helper Data and device identifier hash, lacking the valid information required for BCH error correction. Attackers cannot rely on HD... assoc Derive any information about the original PUF response or R seed The content can therefore be securely stored on the consortium blockchain.

[0041] 3) Anti-replay mechanism: Through the above versioning and device binding management, the use of old versions or helper data from other devices can be prevented, thereby preventing replay attacks caused by on-chain / off-chain synchronization delays.

[0042] Ultimately, after completing this operation, the IoT sensing layer device does not store any sensitive PUF response or seed information locally; all auxiliary data is securely hosted on the consortium blockchain.

[0043] S3. First, parameter mapping is performed. The system maps the chaotic seed sequence R of length 131 bits. seed The system is segmented to map the three initial parameters required for a discrete chaotic system: control parameter r0, initial value x0, and perturbation parameter a. This mapping process ensures that the initial parameters are directly derived from the physical fingerprint of the device, thus the generated keystream naturally possesses device correlation and randomness.

[0044] Secondly, domain detection and resampling are performed. Considering that some parameter combinations may lead to periodicity in chaotic iterations, this scheme performs necessary range checks on the initial parameters after mapping to avoid entering the non-chaotic region. The preset valid domain is: 1) Control the parameter r0∈[3.80,3.99] to ensure that the Logistic mapping is in a strongly chaotic region; 2) Initialize x0 ∈ (0, 1) to prevent the sequence from converging to a fixed point; 3) The disturbance parameter a is a non-zero integer, used for disturbance control parameters, and no longer participates in the modulus calculation.

[0045] If any of the above parameters does not conform to the preset valid domain, the PUF response generation module is immediately triggered to resample, and steps S1 and S2 are re-executed until a compliant initial parameter set is obtained. The floating-point value x output by the chaotic iteration is then... n Quantization base y converted to a fixed-width integer n : The modulus m is fixed at 2. 32 Used only for subsequent output quantization, outputting x in each chaotic iteration. n Afterwards, an integerization operation is performed, and it does not participate in parameter mapping or chaotic iteration.

[0046] Next, chaotic iteration and bit extraction are performed. The system uses compliant initial parameters to perform parameter-perturbed Logistic discrete mapping iteration, the formula of which is:

[0047] in, This is the output value of the (n-1)th iteration. The setting r0=3.89 ensures that the mapping is in a strongly chaotic region. The initial value x0 comes from the PUF mapping result normalized to (0,1).

[0048] Then for y n Perform the middle 8-bit extraction operation (take bits 12 to 19):

[0049] This reduces the impact of end position deviations, retaining only the middle 8 bits that are most evenly distributed in the chaotic sequence, further improving the uniformity and unpredictability of the key stream.

[0050] By discarding the bits with smaller variations in high and low bits and retaining only the more evenly distributed middle bits, it helps to reduce inter-bit correlation and improve statistical randomness. After modulus m quantization, bit extraction, and Von Neumann debiasing, the generated keystream meets common randomness test requirements in terms of zero-one distribution and autocorrelation characteristics, and can be used as input for lightweight encryption modules.

[0051] Finally, bias correction is performed. To eliminate any minor biases that may exist in the original keystream and make its 0 and 1 distribution more uniform, the system uses the Von Neumann bias correction algorithm. This algorithm processes the original bitstream... Divide the data into groups of two consecutive bits, and each group will generate one output bit or be discarded according to the following rules:

[0052] After this bias correction process, the final output is a compliant chaotic key stream K. compl Its 0 and 1 bits are uniformly distributed and satisfy This ensures the uniformity of the keystream bits, providing high-strength key material for subsequent lightweight encryption operations. Here, bi′ is a uniformly distributed set-bit sequence of 0-1 bits obtained by filtering b in groups of two using the Von Neumann algorithm. The two represent a correspondence between "original input and filtered output," and bi′ is the core unit constituting the compliant chaotic keystream. 2i and b 2i-1 It is the raw bitstream After dividing into groups of two, the two consecutive bits in the i-th group.

[0053] The sequences processed by the above method meet the commonly used randomness test requirements in terms of distribution and correlation, and can be used as a high-quality key stream input source for lightweight encryption modules.

[0054] S4. First, the compliant keystream, after modulus m quantization, bit extraction, and bias removal, is divided, and then dynamic S-Box generation is performed. The system generates the compliant chaotic keystream K with a total length of 320 bits obtained in step S3. compl Perform functional partitioning. Take the first 64 bits as the seed for generating the dynamic S-Box, and divide the remaining 256 bits directly and equally into 4 round keys.

[0055] Subsequently, based on seed S seedA 4×4 bijective S-Box is dynamically constructed specifically for this encrypted session. The construction process is as follows: A 64-bit S-Box is constructed... seed The data is divided into 16 segments, each consisting of 4 bits. These segments are then mapped to S-Box entry values ​​using modulo-16 operations, with the mapping rule defined as follows:

[0056] Where i,j∈[1,4], S(i,j) represents the matrix element of a 4×4 bijective dynamic S-Box, and i,j are the row and column indices of the dynamic S-Box matrix. After mapping according to this rule and passing the deterministic algorithm, a 4×4 S-Box matrix can be obtained, where each element is an integer between 0 and 15. This dynamic S-Box is generated in real-time from the keystream of this session, greatly enhancing the non-linear characteristics of the algorithm and effectively improving its ability to resist differential and linear cryptanalysis.

[0057] Next, multiple rounds of SIT encryption are performed. The encryption process targets sensitive data collected by the device. This embodiment uses a preset block length of 64 bits, perfectly suited for typical short data blocks such as single physiological sampling values ​​in the medical IoT or timestamps in supply chain logistics. The encryption employs a 4-round iterative structure, with each round consisting of the following three steps based on simple bitwise operations: 1) Dynamic S-Box substitution: The 64-bit data block to be encrypted is regarded as a 4×4 byte matrix. Using the aforementioned generated dynamic S-Box, S performs non-linear substitution on each byte in the matrix to achieve initial data obfuscation.

[0058] 2) Row Shift: Perform a circular left shift operation within the rows of the substituted data matrix. Specifically, the first row is not shifted, the second row is circularly shifted left by 1 byte, the third row is circularly shifted left by 2 bytes, and the fourth row is circularly shifted left by 3 bytes. This operation adjusts the byte positions of the data to further break down the input structure.

[0059] 3) Round key XOR: XOR the 64-bit round key K of the current round. r (r∈[1,4]) performs a bitwise XOR operation with the row-shifted data block to evenly spread the influence of the round key to the entire data block.

[0060] After the iterative processing described above, the original sensitive data block is converted into a corresponding 64-bit ciphertext data block.

[0061] Finally, a verification tag is appended. To ensure the integrity, authenticity, and parameter consistency of the ciphertext during transmission, and to provide verification evidence for the decryption end, the system appends a verification tag to the ciphertext. This tag consists of three parts: Chaotic seed sequence Rseed BLAKE2s hash digest: .

[0062] BLAKE2s hash digest of the initial parameter set of the chaotic system used to generate the key stream:

[0063] BLAKE2s hash digest of the receiver device's unique identifier RecID: .

[0064] Therefore, the formula for calculating the verification label is:

[0065] This tag, together with the ciphertext, constitutes the final secure data packet, enabling the target node to simultaneously verify the identity characteristics of the sending device (based on the stable seed Rseed) and the integrity of the encrypted session parameters in subsequent steps. This ensures that the decryption end can simultaneously confirm the identity of the sending device and the consistency of the parameters, effectively resisting device impersonation and data tampering attacks.

[0066] The inputs to S1-S4 all originate from the PUF response or its derived chaotic key material within the transmitter. They all occur within the transmitter device and do not depend on any data from the external network or the receiver.

[0067] S5. First, perform encrypted transmission and local key regeneration. The encrypted data packet carrying the verification tag is transmitted to the preset target node through the device's built-in wireless communication unit. After the target node device powers on, in order to decrypt the data packet, it needs to regenerate a compliant chaotic key stream locally that is completely identical to that of the sender. This process repeats S1 to S3: 1) Local PUF response generation: After the receiving end is powered on, it generates a local raw PUF response through its built-in SRAM PUF module. .

[0068] 2) Auxiliary Data Acquisition and Seed Restoration: The target node, through its consortium blockchain communication unit, uses the unique identifier Dev of the sending device to obtain the seed. ID It queries and retrieves the corresponding associated auxiliary data from the consortium blockchain's distributed storage. The original Helper Data is then reconstructed through computation. Subsequently, and Input the BCH fuzzy extractor, perform the same error correction process as the sender, and restore the stable chaotic seed sequence. .

[0069] 3) Chaotic keystream regeneration: utilizing As input, the entire process of step S3 is repeated, and finally, at the target node, a 320-bit compliant chaotic key stream K that is completely identical to that of the sender is regenerated according to the same quantization rule of modulus m, bit extraction and debiasing process. compl .

[0070] Next, the verification tag is validated. This step is crucial for security, performing double validation on the received verification tag: Stable seed feature verification: The target node reconstructs the stable seed using the local PUF response and auxiliary data, calculates its hash value, and compares it with the first part of the received tag. It also verifies whether the chaotic initial parameters conform to the preset valid domain, and simultaneously verifies the hash of the chaotic initial parameters generated by the target node. Hash of parameters in the verification label Whether they are consistent, the pass rate of this process Satisfies: Parameters have not been tampered with =1, after tampering ≈ .

[0071] chaotic seed sequence Calculate BLAKE2s hash And compared with the chaotic seed hash given by the sender in the verification label. A bit-by-bit comparison is performed; if the comparison matches, the sending device's identity is confirmed as legitimate. Simultaneously, the hash of the receiving device's unique identifier is calculated. It is then compared with the recipient binding field in the tag to confirm whether it is a legitimate authorized recipient.

[0072] Under the premise of BCH error correction coverage noise perturbation, the pass rate of legitimate pairing devices approaches 1, while the pass rate of illegitimate or unauthorized receiving devices is negligible due to receiver hash mismatch. This identity verification process does not directly compare any original PUF response, but only implements triple binding based on the repeatable recoverable chaotic seed sequence, chaotic parameter hash, and receiver identifier hash, ensuring that the sender's identity and the receiver's authorization are verified simultaneously. Only when both verifications pass will the system recognize the ciphertext as having a legitimate source and complete parameters, allowing subsequent decryption operations to be performed.

[0073] Finally, decryption is performed using the SIT inverse operation. For the verified ciphertext, the target node uses the locally regenerated compliant chaotic keystream K. compl Perform SIT decryption. This process is the reverse of encryption step S4, and involves four rounds of operations. In each round, the round key XOR, reverse shift, and dynamic S-Box inverse substitution are performed sequentially. The ciphertext data block is accurately restored to the original 64-bit sensitive plaintext data, thus completing the secure data transmission process from encryption to decryption.

[0074] S6. First, perform anomaly monitoring and judgment. The system continuously monitors two key indicators: parameter compliance anomalies and power consumption anomalies. Parameter compliance anomalies are judged when the initial parameters generated by the chaotic key stream generation module fail to meet the preset valid field of S3 for three consecutive times; power consumption anomalies are determined by the system's real-time monitoring of the power consumption during encryption operations. and compared with the preset normal power consumption benchmark A comparison was made when the real-time power consumption deviated from the normal value by more than a preset threshold. The determination of time.

[0075] When any of the above abnormal conditions are met, the system immediately triggers the dynamic parameter update process.

[0076] Secondly, deep resampling and sample optimization of PUF are performed. Upon an anomaly trigger, the PUF response generation module initiates deep resampling mode. In this mode, the number of samples increases from the usual one to five, and the module sequentially generates five independent original PUF response samples. The system calculates the noise rate between each sample and its corresponding BCH-corrected seed, and selects the sample with the lowest noise rate as the effective R for this recovery process. PUF This is used for subsequent regeneration of chaotic seed sequences.

[0077] Subsequently, on-chain alerts and Helper data updates are implemented. After completing deep resampling and sample optimization, the device immediately sends a parameter anomaly alert to its associated consortium blockchain node via the Helper data interaction module. This alert information includes the device's unique identifier, Dev. ID With exception type Type alert ,in:

[0078] After receiving an alarm, the consortium blockchain node determines that the device needs to update its security parameters. The node first checks the Dev... ID Index retrieves the legacy Helper Data HD currently stored on the device. old Subsequently, using the newly generated PUF response from the device, the new version of Helper Data HD was calculated through the BCH fuzz extractor. new Next, the consortium blockchain node generates a secure incremental update package. The incremental packet is sent to the device that initiated the alarm through the secure channel of the consortium blockchain.

[0079] Finally, local parameter synchronization and recovery are completed. After receiving and parsing the incremental packet issued by the consortium blockchain, the device performs a local Helper Data synchronization update. The synchronization formula is as follows:

[0080] This update enables local device functionality. With consortium blockchain storage Maintain consistency. Once synchronization is complete, the device immediately utilizes the new R... PUF By re-executing steps S2 and S3, a new, compliant chaotic key stream is generated, restoring the cryptographic system to a secure and usable working state. This process also updates the Helper Data, ensuring consistency between the local auxiliary data and the records on the consortium blockchain, thereby guaranteeing the reliability and continuity of subsequent key generation processes. This embodiment discloses a lightweight encryption system for the IoT sensing layer based on PUF and chaotic key generation, such as Figure 2 As shown, it includes: PUF Response Generation Module: After the IoT sensing layer device is powered on, it generates a unique and unclonable original response through the device's built-in PUF module. Seed stabilization and auxiliary data management module: Inputs the original response into the fuzzy extractor and outputs a random chaotic seed sequence and auxiliary data for seed extraction; associates the auxiliary data according to the unique identifier of the sending device and uploads it to the consortium blockchain; Chaotic key stream generation module: Maps the chaotic seed to the initial parameters of the chaotic system, performs domain detection on the initial parameters, performs discrete chaotic iteration based on the compliant initial parameters, and generates the original key stream; performs bit extraction and debiasing processing on the original key stream to obtain a compliant chaotic key stream that meets the randomness requirements; The SIT lightweight encryption and tag generation module divides the compliant chaotic key stream into a dynamic S-Box seed and round keys. It uses the dynamic S-Box to perform multiple rounds of SIT encryption operations on the sensitive data collected by IoT sensing layer devices and outputs ciphertext. At the same time, it generates a verification tag and transmits the ciphertext and verification tag to the target node. Decryption and Verification Module: After receiving the ciphertext, the target node retrieves the corresponding auxiliary data from the consortium blockchain index table and inputs the local PUF response into the fuzzy extractor. Then, it executes S3 to generate a chaotic key stream consistent with the sender, while verifying the consistency between the locally generated chaotic seed sequence and the chaotic initial parameters and the verification label. After the verification is successful, it performs SIT inverse operation based on the compliant chaotic key stream to decrypt and obtain the original sensitive data. System operation status anomaly monitoring module: When the chaotic parameter verification fails or the on-chain auxiliary data version is inconsistent, update the local auxiliary data according to the incremental auxiliary data uploaded by the sending end.

[0081] The specific implementation process of the PUF response generation module is as follows: The PUF module is an SRAM PUF module; the original response R PUF The bit length is determined based on the device's SRAM capacity, and meets the requirements. ,and ;RPUF The randomness is determined by the threshold voltage V th Channel length L ch The determination, possessing the physical non-cloning property, means that for any two different IoT sensing layer devices, the probability Pr that their PUF responses are the same satisfies:

[0082] in, , These are the original PUF responses generated by the two devices. The IoT sensing layer device is the receiving end, responsible for receiving the ciphertext and reconstructing the session key using the local PUF response to complete the decryption.

[0083] The specific implementation process of the seed stabilization and auxiliary data management module is as follows: The built-in BCH code arithmetic unit processes R through XOR and modulo-2 operations. PUF Eliminate 1%-5% of hardware noise and output a stable chaotic seed sequence R. seed The processing procedure complies with (BCH_Ext( (This is the BCH fuzzy extraction function); simultaneously, auxiliary data HD is generated, satisfying the auxiliary data length requirement. , .

[0084] It also includes an association unit and a consortium blockchain communication unit; the association unit associates auxiliary data with the device's unique identifier using an XOR operation, with the association formula being... ( (For the associated data) The consortium blockchain communication unit will associate the data. Uploaded to the distributed node storage of the consortium blockchain, and supports retrieving auxiliary data and incremental packets from the sending device from the consortium blockchain. Consortium blockchain nodes build an index table based on the device's unique identifier hash, and the indexing rules are as follows: (H( (The SHA-256 hash function) is used to bind the Helper data version. The correspondence between the device and the equipment is established to prevent replay attacks; The specific implementation process of the chaotic key stream generation module is as follows: It includes a parameter mapping unit, a parameter detection unit, and a key stream post-processing unit. The parameter mapping unit segments the chaotic seed sequence according to "control parameter r0 - initial value x0 - perturbation parameter a", and maps it to discrete chaotic initial parameters through modulo operation. The mapping rule is as follows: (SegMap( ) is a piecewise mapping function. They are respectively The segment bit length, and , (where r is the length of the chaotic seed sequence); it does not involve parameter mapping of the modulus m, which is only used for subsequent keystream diffraction processing. The control parameter r0∈[3.80,3.99], the initial value x0∈(0,1), and the perturbation parameter a are non-zero integers. If they do not meet the requirements, the PUF response generation module is triggered to resample. The keystream post-processing unit performs parameter perturbation-type Logistic iteration based on the compliant initial parameters, specifically as follows: Iteration (32-64 iterations, where) For chaotic outputs, first adjust the modulus m=2. 32 implement After unified quantization, the middle 8 bits are extracted and then Von Neumann bias removal is performed to output a compliant chaotic key stream. ,satisfy To adapt to the subsequent dynamic S-Box seed and round key partitioning requirements; The specific implementation process of the SIT lightweight encryption and tag generation module is as follows: The SIT lightweight encryption module includes a dynamic S-Box generation unit and a round encryption unit; the dynamic S-Box generation unit divides the 64-bit segment in the chaotic keystream into 4-bit segments, and then... ( Generate a 4×4 bijective dynamic S-Box; the round encryption unit divides the remaining 256-bit chaotic key stream into 4 groups of 64-bit round keys ( , (Unit: bits) Performs 4 rounds of "substitution-shift-round key XOR" on a 64-bit sensitive data block, outputting ciphertext with an appended verification tag consisting of the hash of the chaotic seed sequence, the hash of the chaotic parameters, and the hash of the receiver's unique identifier. The tag calculation formula is:

[0085] The transmission module includes a data transmission unit and a tag verification unit; the data transmission unit transmits the ciphertext and tag to the target node.

[0086] The specific implementation process of the decryption and verification module is as follows: The tag verification unit obtains auxiliary data from the consortium blockchain to verify R. seed consistency( ), Chaotic parameter validity (the parameter conforms to the preset valid domain, and ), and verify the unique identifier hash of the receiving device. Check if the data matches the recipient's bound fields in the tag; after all checks pass, perform the SIT inverse operation to decrypt and obtain the original sensitive data.

[0087] The iterative operation unit of the chaotic key stream generation module and the round encryption unit of the SIT lightweight encryption module are both implemented using combinational logic circuits, with a single operation latency in the nanosecond range; the overall system power consumption... (μJ / time), supports 24 / 7 continuous encryption operation, and the average power consumption during continuous encryption meets the requirements. (mW).

[0088] The specific implementation process of the system operation status anomaly monitoring module is as follows: Used to monitor the compliance of parameters in the chaotic key stream generation module. When the parameters are non-compliant for 3 consecutive times ( ≥3) or abnormal power consumption detected ( When ), the PUF response generation module is triggered to perform deep resampling ( =5), and sends alarms to the consortium blockchain nodes through the auxiliary data interaction module; at the same time, it receives the auxiliary data increment packet ΔHD issued by the consortium blockchain nodes, and through Complete parameter synchronization; the hardware implementation of the PUF response generation module, BCH fuzzy extraction module, and chaotic key stream generation module all adopt 8-bit / 16-bit MCU compatible logic, without the need for floating-point arithmetic units, and are adapted to the weak computing power architecture of IoT sensing layer devices (computing power less than 100MIPS).

[0089] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0090] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An IoT perception layer lightweight encryption method based on PUF and chaotic key generation, characterized in that, The method comprises the following steps: S1. PUF response generation: after the IoT perception layer device is powered on, a unique and unclonable original response is generated through a PUF module built in the device; S2. Seed stabilization and auxiliary data management: the original response is input into a BCH fuzzy extractor, and a random chaotic seed sequence and auxiliary data for assisting seed extraction are output; the auxiliary data is associated with a unique identifier of the sending device and uploaded to a consortium chain; S3. Chaotic key stream generation: the chaotic seed is mapped to an initial parameter of a chaotic system, a preset valid domain is detected for the initial parameter, and a discrete chaotic iteration is performed based on the compliant initial parameter to generate an original key stream; bit extraction and de-biasing are performed on the original key stream to obtain a compliant chaotic key stream meeting the randomness requirement; S4. SIT lightweight encryption and tag generation: the compliant chaotic key stream is divided into a dynamic S-Box seed and a round key, a multi-round SIT encryption operation is performed on sensitive data collected by the IoT perception layer device by using the dynamic S-Box, and ciphertext is output; a verification tag is generated, and the ciphertext and the verification tag are transmitted to a target node; S5. Decryption and verification: after the target node receives the ciphertext, the corresponding auxiliary data and the local PUF response are input into the BCH fuzzy extractor from the consortium chain index table, and the chaotic key stream consistent with that of the sending end is generated, and the consistency of the verification tag and the locally generated chaotic key stream is verified; After the verification passes, the original sensitive data is obtained by performing an SIT inverse operation decryption based on the compliant chaotic key stream.

2. The method of claim 1, wherein, The PU F module in the S1 is an SRAM PU F module; the original response R PUF The bit length L PUF According to the device SRAM capacity determination, meet , and ; the randomness of the original response R PUF Is determined by the threshold voltage V th , channel length L ch , has physical unclonable characteristics, and the probability Pr that the PU F responses of any two different IoT sensing layer devices are the same meets: wherein , are the raw responses generated by the two devices, respectively.

3. The method of claim 1, wherein, The specific process of inputting the original response into the BCH fuzzy extractor and outputting the random chaotic seed sequence and the auxiliary data for assisting seed extraction in S2 comprises: The BCH(n, k, t) code type is adopted, wherein n is the code length, k is the information bit length, and t is the number of error correction bits, the value of t satisfies , and t needs to adapt to the 1%-5% noise rate of R PUF , that is: wherein, represents a ceiling function, L PUF is the bit length of the original response R PUF . The supervising bits are generated by BCH encoding, and the bit errors in the original response R PUF are detected and corrected based on the supervising bits, and the chaotic seed sequence is obtained after effectively eliminating the noise of the original response. The length of the auxiliary data Satisfies , and the auxiliary data only contains the supervision bit information of the BCH code, and the supervision bit length is That is ; The auxiliary data is associated with the unique identifier of the sending device and uploaded to the consortium chain, and the consortium chain stores and records the version of the auxiliary data, including the following processing: 1) association binding processing: the auxiliary data obtained by BCH error correction is associated with the hash value Dev ID of the unique identification of the sending device to generate association data HD assoc : , where HD is expressed as auxiliary data, Dev ID is expressed as the hash value of the unique identification of the sending device, represents the XOR operation; 2) On-chain storage and index management: upload the associated data to the consortium chain node for distributed storage; the consortium chain node constructs an index key Index based on the hash value of the unique identifier of the sending device, wherein: , and records the version number of the HD and the PUF response hash corresponding thereto according to the index key; wherein, H( ) is a SHA-256 hash function; 3) Anti-replay control: based on the versioned on-chain index management, it is ensured that the IoT perception layer device at the receiving end can only obtain the current effective version of the HD.

4. The method of claim 1, wherein, In S3, the preset valid domain is: the control parameter r0∈[3.80,3.99], the initial value x0∈(0,1), and the perturbation parameter a is a non-zero integer; the discrete chaotic iteration adopts a parameter perturbation type Logistic mapping, and the iteration formula is: wherein, is the output value of the n-1th iteration, , a is a disturbance parameter mapped by a chaotic seed sequence, the iteration step number N ∈ [32, 64], and a floating point value x n ∈ (0, 1) is output each time, and the modulus value m = 2 32 , the floating point value x n is converted into a fixed-width integer quantization base y n : , after quantization to a 32-bit integer, bit extraction is performed; the bit extraction is to extract the middle 8 bits of the integer output, and the extraction rule is represented as: wherein denotes a floor function; The debiasing processing adopts Von Neumann algorithm, and the bit stream t is the number of groups, divided by 2 bits, and the output rule is: The uniformity of 0 and 1 distribution of the de-biased bit stream satisfies where P r is the probability of the event, b 2i i' is the de-biased bit of 0-1 uniform distribution obtained by screening b in groups of 2 bits by Von Neumann algorithm, b 2i-1 and b 2i are the original bit stream Two consecutive bits in the i-th group after division in groups of 2 bits.

5. The method of claim 1, wherein, In S4, the dynamic S-Box seed is used to generate a 4x4 bijective dynamic S-Box, and the round key is a plurality of keys consistent with the grouping length of the data to be encrypted; The sensitive data collected by the IoT perception layer device is divided into data blocks according to a preset grouping length, the preset grouping length is 64 bits, that is, the length of the sensitive data block , adapt to the short sensitive data block collected by the IoT perception layer device at a time; the number of rounds of the multi-round SIT encryption is 4 rounds; the length of the compliant chaotic key stream is 256 bits, of which 64 bits are used as a dynamic S-Box seed, divided into 16 segments of 4 bits each, denoted as s1, s2,..., s 16 , each segment is mapped to an S-Box entry through modulo 16 operation, and the mapping rule is: wherein i, j ∈ [1, 4], S(i, j) represents the matrix element of the 4 × 4 bijective dynamic S-Box, i, j are the row and column indexes of the dynamic S-Box matrix, and a 4 × 4 dynamic S-Box is obtained , and by ensuring that it is a bijective mapping through a non-repetition constraint, the remaining 256 bits are divided into 4 groups of round keys by 64 bits each , , , respectively corresponding to 4 rounds of SIT encryption.

6. The method of claim 4, wherein, The chaotic key stream includes a chaotic seed sequence and a chaotic initial parameter, and the consistency of the verification tag and the locally generated chaotic key stream comprises: The receiving end device generates a local PUF response of the receiving end through a local physically unclonable function module after power-on The target node obtains the latest auxiliary data bound with the device from the alliance chain according to the device unique identifier carried in the ciphertext , and inputs the and together into a BCH fuzzy extractor to recover the chaotic seed sequence ; Chaotic seed sequence Computing BLAKE2s hash And compare with the chaotic seed hash sent by the sender in the verification tag Bit by bit, the same comparison confirms that the sending device is legitimate; At the same time, the hash of the unique identification of the receiving device is calculated And compare with the receiving party binding field in the tag, which is used to confirm whether it is a legitimate authorized receiver; Check if the chaotic initial parameters are in the S3 preset valid domain, and verify the hash of the chaotic initial parameters generated by the target node and the parameter hash in the verification tag are consistent , this process passes : the parameters are not tampered with =1, after tampering ≈ .

7. The method of claim 6, wherein, Further comprising S6 system running state abnormality monitoring, and the system running state abnormality monitoring in S6 specifically comprises: IoT sensing layer devices continuously output initial parameters that do not conform to the preset valid domain described in S3, and the parameter compliance is determined to be abnormal; and simultaneously monitor the real-time power consumption P of the encryption operation real , and compare it with the normal power consumption reference P normal , when the deviation exceeds the preset threshold ΔP th , the power consumption behavior is determined to be abnormal; When any abnormal condition is triggered, the PUF module is controlled to enter a deep resampling mode, five original responses of the SRAM PUF are sampled, the noise rates of the samples are calculated, and the original response with the lowest noise rate is selected as the input for subsequent fuzzy extraction and key generation; The device sends an abnormal alarm information to the consortium chain node, which includes the device unique identifier Dev ID and the exception type Type alert . After the consortium chain node receives the alarm, it generates new auxiliary data HD new based on the latest PUF response, and calculates the incremental data ΔHD with the old auxiliary data HD old stored on the chain, and the calculation formula is: The consortium chain node sends an incremental package ΔHD to the device triggering the exception. The device performs auxiliary data synchronization locally, and the formula is: wherein is the updated assistance data, is the old assistance data, is the data after locally performing the assistance synchronization, denotes the exclusive OR operation.

8. The method according to any one of claims 1 to 7, wherein Specifically, it includes: Total energy consumption for single encryption is 1-5 microjoules, the total energy consumption is composed of the energy consumption of each step, which satisfies: where E PUF is the PUF response generation energy consumption, E BCH is the BCH blinding extraction energy consumption, E chaos is the discrete chaos iteration energy consumption, E SIT is the SIT encryption energy consumption, and , , , ; total encryption time per single , , is the device time window.

9. An IoT perception layer lightweight encryption system based on PUF and chaotic key generation, comprising: PUF response generation module: after the IoT perception layer device is powered on, a unique and unclonable original response R is generated through a PUF module built in the device PUF ; Seed stabilization and auxiliary data management module: input the original response into the fuzzy extractor, and output the random chaotic seed sequence and the auxiliary data for assisting seed extraction; associate the auxiliary data according to the unique identifier of the sending device, and upload it to the alliance chain; Chaotic key stream generation module: map the chaotic seed to the initial parameters of the chaotic system, perform domain detection on the initial parameters, perform discrete chaotic iteration based on the compliant initial parameters, and generate the original key stream; bit extraction and de-biasing are performed on the original key stream to obtain a compliant chaotic key stream meeting the randomness requirement; SIT lightweight encryption and label generation module: divide the compliant chaotic key stream into dynamic S-Box seeds and round keys, perform multi-round SIT encryption operation on the sensitive data collected by the IoT perception layer device using the dynamic S-Box, and output the ciphertext; at the same time, generate a verification label, and transmit the ciphertext and the verification label to the target node; Decryption and verification module: after the target node receives the ciphertext, the corresponding auxiliary data is obtained from the index table of the alliance chain, and the local PUF response is input into the fuzzy extractor, then the same chaotic key stream as the sending end is generated by performing S3, and the consistency of the locally generated chaotic seed sequence and chaotic initial parameters with the verification label is checked; After the verification is passed, the original sensitive data is obtained by performing SIT inverse operation decryption based on the compliant chaotic key stream.

10. The system of claim 9, wherein, Also includes, system operation state abnormality monitoring module: when the chaotic parameter verification fails or the chain auxiliary data version is inconsistent, update the local auxiliary data according to the incremental auxiliary data uploaded by the sending end.

Images