A method and system for secure transmission of dynamic key images for time-delay communication
By using an unrestricted pulse synchronization system and quantization step size to screen state samples, combined with key derivation and bidirectional diffusion techniques, the problem of unstable dynamic key generation under time-delay communication is solved, thus improving the applicability and security of image encryption.
Patent Information
- Application Number
- CN202611291499.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-25
- Publication Date
- 2026-11-17
AI Technical Summary
Existing image encryption schemes are insufficient for dynamic key consistency generation under time-delay communication conditions, and direct quantization of near-synchronous continuous states is prone to mismatch. Dynamic seed generation and high-speed image processing lack effective layering, resulting in high computational overhead and insufficient applicability.
An unrestricted pulse synchronization system is used to generate a stable sampling window. State samples are screened based on a preset quantization step size and a safety margin. A consistent dynamic session seed is generated through a key derivation function. Image encryption is performed using a non-repeating scrambling index and bidirectional diffusion technology.
It improves the applicability of dynamic key generation, avoids quantization mismatch problems, reduces computational overhead, enhances ciphertext security and robustness, and resists differential attacks.
Smart Images

Figure CN123123512A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of image secure transmission technology, and more specifically, to a dynamic key image secure transmission method and system for time-delay communication. Background Technology
[0002] Image data is widely used in remote sensing monitoring, medical imaging, industrial inspection, power line inspection and open network communication, etc. It has the characteristics of large data volume, high spatial redundancy, strong correlation between adjacent pixels and intuitive visibility of sensitive visual information. Therefore, image encryption is needed to protect the transmission security.
[0003] Existing image encryption schemes often employ a combination of scrambling and diffusion, achieving statistical concealment and resistance to differential attacks by disrupting pixel position correlations and altering pixel value distributions. Some schemes introduce neural network synchronization techniques to generate the keys or pseudo-random sequences required for encryption. Examples include schemes based on fuzzy neural network fixed-time synchronization, delayed chaotic synchronization, and inertial neural network event-triggered synchronization. These schemes utilize the synchronization state between the driving and response systems to generate chaotic sequences or key materials, which are then combined with scrambling and diffusion to achieve image encryption.
[0004] However, the aforementioned existing technologies still have the following shortcomings: First, the dynamic key consistency generation is insufficient under time-delay communication conditions. Most existing schemes assume that both communicating parties already possess a consistent key stream, failing to adequately address the problem of stable and consistent generation of dynamic key materials under sampling delay and link jitter conditions. Furthermore, existing pulse synchronization schemes typically rely on the condition that the sampling delay is less than the pulse interval, making them unsuitable for scenarios with significant communication delays. Second, direct quantization of approximately synchronized continuous states is prone to mismatch. Continuous states can usually only be approximately synchronized between communicating parties. When samples are near the quantization boundary, even small synchronization errors can lead to different discrete quantization values, which, after key derivation, can be amplified into decryption failure. Third, dynamic seed generation and high-speed image processing lack effective layering. Fixed seeds struggle to balance session dynamism and receiver reconstructability, while directly using the synchronized dynamic system as a pixel-level pseudo-random source increases computational overhead, hindering engineering implementation.
[0005] Therefore, how to stably generate consistent dynamic key materials using near-synchronous states under time-delay communication conditions and efficiently drive the image scrambling and diffusion encryption process is a technical problem that urgently needs to be solved in the field of secure image transmission. Summary of the Invention
[0006] To overcome the problem of unstable and inefficient coordination between dynamic key generation and image encryption in existing time-delay communication, this invention provides a dynamic key image secure transmission method and system for time-delay communication.
[0007] The technical solution of this invention is as follows:
[0008] A method for secure transmission of dynamic key images for time-delay communication includes the following steps:
[0009] Based on the operation of the unrestricted pulse synchronization system, when the synchronization error between the transmitter and receiver falls below a preset synchronization error threshold and remains stable, a stable sampling window is determined.
[0010] Based on a preset quantization step size, a quantization boundary set is determined. Within the stable sampling window, the transmitting end reads state samples, calculates the distance between the state samples and the quantization boundary set, and accepts the state samples only when the distance is greater than the sum of the synchronization error threshold and the safety margin. The sampling point index and state component index corresponding to the accepted state samples are then sent to the receiving end as public auxiliary information.
[0011] The receiving end reads the corresponding sample from the local synchronization state based on the publicly available auxiliary information, and performs quantization using the same preset quantization step size as the sending end to obtain the quantized sequence;
[0012] The sending and receiving ends respectively input the quantization sequence, the pre-shared master key, and the public auxiliary information into the key derivation function to generate a consistent dynamic session seed;
[0013] The sending end generates a non-repeating scrambling index, a diffusion key stream, and a feedback substitution table based on the dynamic session seed. After performing flattening, scrambling, forward feedback diffusion, and reverse feedback diffusion on the image to be transmitted, it outputs a ciphertext image.
[0014] The receiving end reconstructs the scrambling index, diffusion key stream, and feedback substitution table based on the same dynamic session seed, and sequentially performs reverse diffusion, reverse forward diffusion, and reverse scrambling to restore the original image.
[0015] Preferably, the first of the unrestricted pulse synchronization system Subsampling delay and the Subpulse interval Each satisfies ,in, This is the upper bound of the sampling delay. This is the upper bound of the pulse interval, and it is not required. The pulse interval The sampling delay is the time interval between two adjacent pulse moments. This is the time it takes for the state sample value from the sending end to reach the receiving end.
[0016] Preferably, the process of operating the transmitter and receiver based on the unrestricted pulse synchronization system includes:
[0017] The sending end is running the target neural network state The receiving end operates a coupled neural network state. and will Defined as synchronization error; the receiver maintains a historical synchronization error buffer, at the pulse moment Read and sampling delay Corresponding historical synchronization error and according to the pulse sampling controller Pulse control is implemented, where Π is the control gain matrix and δ(·) is the Dirac pulse function; the synchronization error after pulse application satisfies ,in, For continuous time variables, and These represent the times before and after the k-th pulse, respectively.
[0018] Preferably, the process of the synchronization error falling below a preset synchronization error threshold and remaining stable includes:
[0019] Continuously monitor the synchronization error between the transmitter and receiver, and when the maximum synchronization error... Continuous preset duration The error value is always less than or equal to the preset synchronization error threshold. When the continuous preset duration ends, the starting point of the stable sampling window is determined. and in A stable sampling window is formed during the period when the conditions are continuously met. When the conditions are no longer met, the extraction of dynamic key materials from the synchronization state is paused.
[0020] Preferably, the process of determining the quantization boundary set based on the preset quantization step size includes: the quantization boundary set is a set composed of integer multiples of the preset quantization step size, and the quantization adopts a uniform quantization method with rounding down;
[0021] The set of quantization boundaries is:
[0022] ;
[0023] The quantization method used is uniform quantization:
[0024] ;
[0025] in, For the set of quantization boundaries, The preset quantization step size, For a set of integers, The status sample read by the sending end.
[0026] Preferably, the distance is the absolute value of the difference between the state sample and the nearest quantization boundary in the quantization boundary set; the safe zone quantization condition is:
[0027] ;
[0028] in, The preset synchronization error threshold is... This is the safety margin.
[0029] Preferably, during the process of sending the sampling point index and state component index corresponding to the accepted state sample to the receiving end as public auxiliary information, the public auxiliary information further includes the quantization step size, the safety margin, random number, image size, algorithm identifier, and channel identifier or domain identifier.
[0030] Preferably, the derivation relationship of the dynamic session seed is as follows:
[0031] ;
[0032] ;
[0033] in, This is the dynamic session seed for the sending end. This is the dynamic session seed for the receiving end. The quantization sequence of the transmitting end, The quantization sequence for the receiving end, For the pre-shared master key, The random number, The image size is... The algorithm identifier, This refers to the channel identifier or domain identifier.
[0034] Preferably, the key derivation function is a key derivation function constructed based on a cryptographic hash function, an extensible output function, or a message authentication code; the key derivation function derives a scrambled byte stream, a forward diffusion key stream, a reverse diffusion key stream, a forward feedback substitution table, and a reverse feedback substitution table by separating tags with different fields; in one embodiment, the key derivation function is implemented based on SHAKE256, KMAC256, or HKDF.
[0035] A second aspect of the present invention also provides a dynamic key image secure transmission system for time-delay communication, comprising a synchronization state generation module, a secure zone quantization module, a quantization sequence reconstruction module, a dynamic session seed derivation module, an image encryption module, and an image decryption module, wherein:
[0036] The synchronization state generation module is used to operate the transmitter and receiver based on the unrestricted pulse synchronization system. When the synchronization error between the transmitter and receiver enters below the preset synchronization error threshold and remains stable, a stable sampling window is determined.
[0037] The safe zone quantization module is used to determine the quantization boundary set based on a preset quantization step size. Within the stable sampling window, based on the state sample read by the transmitting end, the distance between the state sample and the quantization boundary set is calculated. The state sample is accepted only when the distance is greater than the sum of the synchronization error threshold and the safe zone margin. The sampling point index and state component index corresponding to the accepted state sample are sent to the receiving end as public auxiliary information.
[0038] The quantization sequence reconstruction module is used by the receiving end to read the corresponding sample from the local synchronization state based on the publicly available auxiliary information, and to perform quantization using the same preset quantization step size as the sending end to obtain the quantization sequence;
[0039] The dynamic session seed derivation module is used to input the quantization sequence, the pre-shared master key and the public auxiliary information into the key derivation function at the sending end and the receiving end respectively to generate a consistent dynamic session seed.
[0040] The image encryption module is used by the sending end to generate a non-repeating scrambling index, a diffusion key stream, and a feedback substitution table based on the dynamic session seed, and to output an encrypted image after performing flattening, scrambling, positive feedback diffusion, and reverse feedback diffusion on the image to be transmitted.
[0041] The image decryption module is used by the receiving end to reconstruct the scrambling index, diffusion key stream, and feedback substitution table based on the same dynamic session seed, and then sequentially perform reverse diffusion, reverse forward diffusion, and reverse scrambling to restore the original image.
[0042] The advantages of the dynamic key image secure transmission method and system based on the above scheme for time-delay communication are as follows:
[0043] 1. By introducing an unrestricted pulse synchronization system, this invention does not require the sampling delay to be less than the pulse interval, and allows the sampling delay and pulse interval to each have independent upper limits. This enables it to adapt to scenarios such as large communication delays, network congestion, link jitter, and long-distance transmission, thereby improving the applicability of dynamic key generation under complex time-delay communication conditions.
[0044] 2. By setting a safe zone quantization condition, the present invention only accepts a sample when the distance between the state sample and the quantization boundary set is greater than the sum of the synchronization error threshold and the safe zone margin. This eliminates samples near the quantization boundary from the source, so that the transmitter and receiver can still be quantized to the same discrete interval even when synchronization error exists. This avoids the dynamic seed mismatch problem caused by direct quantization of near-synchronous states.
[0045] 3. This invention limits the publicly available auxiliary information to non-key materials such as sampling point index, state component index, quantization step size, security margin, random number, image size, algorithm identifier, and channel identifier or domain identifier. Synchronization state value, synchronization error value, quantization value, quantization sequence, pre-shared master key, dynamic session seed, and key stream are not transmitted in the public channel. Even if the publicly available auxiliary information is intercepted, attackers cannot reconstruct the dynamic session seed or key stream based on it, effectively protecting the security of core key materials.
[0046] 4. This invention separates the low-frequency synchronization state generation from the high-frequency image encryption process. Synchronization state generation, secure area quantization, and dynamic session seed derivation are low-frequency operations, which are executed only once when the session is established. Scrambling index generation, diffusion key stream generation, and bidirectional feedback diffusion are high-frequency operations, which are executed at the pixel level. This layered architecture avoids the computational overhead caused by directly using the synchronization dynamic system as a pixel-level pseudo-random source, reduces the difficulty of engineering implementation, and improves processing efficiency.
[0047] 5. This invention destroys pixel position correlation by scrambling without repetition, and diffuses minute changes in the plaintext image to the entire ciphertext image through positive and negative feedback diffusion. The bidirectional diffusion structure further enhances the sensitivity of the ciphertext to the plaintext and key, effectively resists differential and statistical attacks, and improves the security and robustness of the ciphertext. Attached Figure Description
[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention, 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0049] Figure 1 A schematic diagram of the dynamic key image secure transmission method for time-delay communication provided by the present invention;
[0050] Figure 2 A schematic diagram illustrating the security zone quantization guarantee for consistency between the sender and receiver in this invention.
[0051] Figure 3This is a schematic diagram of the image scrambling and bidirectional feedback diffusion process provided by the present invention. Detailed Implementation
[0052] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0053] It should be noted that when a component is referred to as "fixed," "set," or "connected" to another component, it may be located directly or indirectly on that other component. The terms "upper," "lower," "left," "right," "front," "rear," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or position based on the accompanying drawings, and are for ease of description only, and should not be construed as limiting the technical solution. The terms "first," "second," etc., are used for ease of description only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features. "Many" means two or more, unless otherwise explicitly specified. "Several" means one or more, unless otherwise explicitly specified.
[0054] This invention provides a method for secure transmission of dynamic key images for time-delay communication. See [link to relevant documentation]. Figure 1 This includes the following steps:
[0055] S1. Based on the operation of the unrestricted pulse synchronization system, the transmitting end and the receiving end are operated. When the synchronization error between the transmitting end and the receiving end enters below the preset synchronization error threshold and remains stable, a stable sampling window is determined.
[0056] S2. Determine the quantization boundary set based on the preset quantization step size. Within the stable sampling window, calculate the distance between the state sample and the quantization boundary set based on the state sample read by the transmitting end. Accept the state sample only when the distance is greater than the sum of the synchronization error threshold and the safety margin. Send the sampling point index and state component index corresponding to the accepted state sample as public auxiliary information to the receiving end.
[0057] S3. The receiving end reads the corresponding sample from the local synchronization state based on the publicly available auxiliary information, and performs quantization using the same preset quantization step size as the sending end to obtain a quantized sequence;
[0058] S4. The sending end and the receiving end respectively input the quantization sequence, the pre-shared master key and the public auxiliary information into the key derivation function to generate a consistent dynamic session seed;
[0059] S5. The sending end generates a non-repeating scrambling index, a diffusion key stream, and a feedback substitution table based on the dynamic session seed, and outputs a ciphertext image after performing flattening, scrambling, positive feedback diffusion, and reverse feedback diffusion on the image to be transmitted.
[0060] S6. The receiving end reconstructs the scrambling index, diffusion key stream, and feedback substitution table based on the same dynamic session seed, and sequentially performs reverse diffusion, reverse forward diffusion, and reverse scrambling to restore the original image.
[0061] In some embodiments, the process of performing step S1 includes:
[0062] S1.1 Operation of the unrestricted pulse synchronization system;
[0063] S1.2 Determination of stable sampling window.
[0064] Specifically, step S1.1 includes: the first step of the unrestricted pulse synchronization system. Subsampling delay and the Subpulse interval Each satisfies ,in, This is the upper bound of the sampling delay. This is the upper bound of the pulse interval, and it is not required. The pulse interval The sampling delay is the time interval between two adjacent pulse moments. This is the time it takes for the state sample value from the sending end to reach the receiving end.
[0065] Furthermore, the transmitting end runs the target neural network state. The receiving end operates a coupled neural network state. and will Defined as synchronization error; the receiver maintains a historical synchronization error buffer, at the pulse moment Read and sampling delay Corresponding historical synchronization error and according to the pulse sampling controller Pulse control is implemented, where Π is the control gain matrix and δ(·) is the Dirac pulse function; the synchronization error after pulse application satisfies ,in, For continuous time variables, and These represent the times before and after the k-th pulse, respectively.
[0066] In fact, this invention does not require that the sampling delay be less than the pulse interval, so that the synchronization system can adapt to complex communication scenarios such as large communication delays, network congestion, link jitter and long-distance transmission.
[0067] Specifically, during step S1.2, the process includes: continuously monitoring the synchronization error between the transmitting and receiving ends, and when the maximum synchronization error... Continuous preset duration The error value is always less than or equal to the preset synchronization error threshold. When the continuous preset duration ends, the starting point of the stable sampling window is determined. and in A stable sampling window is formed during the period when the conditions are continuously met. When the conditions are no longer met, the extraction of dynamic key materials from the synchronization state is paused.
[0068] Optionally, the preset synchronization error threshold and the preset duration can be set according to the actual needs and synchronization accuracy requirements of the communication system. For example, in a specific embodiment, the preset synchronization error threshold is 0.0015 and the preset duration is 0.3 seconds.
[0069] Furthermore, once the stable sampling window is determined, all subsequent state samples used to generate dynamic key materials are collected from within this window, and state samples outside the window are not used, in order to ensure the reliability and consistency of the collected samples.
[0070] In some embodiments, the process of performing step S2 includes:
[0071] S2.1 Determination of the quantization boundary set;
[0072] S2.2 Reading of state samples and selection of safe zones;
[0073] S2.3 Public release of auxiliary information.
[0074] Specifically, the execution of step S2.1 includes: determining a quantization boundary set based on a preset quantization step size. The quantization boundary set is a set consisting of integer multiples of the preset quantization step size, and the quantization adopts a uniform quantization method with rounding down.
[0075] Specifically, the set of quantization boundaries is:
[0076] ;
[0077] The quantization method used is uniform quantization:
[0078] ;
[0079] in, For the set of quantization boundaries, The preset quantization step size, For a set of integers, The status sample read by the sending end.
[0080] In one specific embodiment, the preset quantization step size Take 0.01.
[0081] Further, in step S2.2, within the stable sampling window, the transmitting end reads the state sample, calculates the distance between the state sample and the quantization boundary set, and accepts the state sample only when the distance is greater than the sum of the synchronization error threshold and the safety margin.
[0082] Specifically, the distance is the absolute value of the difference between the state sample and the nearest quantization boundary in the quantization boundary set. The safe zone quantization condition is:
[0083] ;
[0084] in, The preset synchronization error threshold is... This is the safety margin.
[0085] The safety margin These are preset system parameters used to add an extra guard distance beyond the synchronization error threshold to absorb uncertainties such as unmodeled errors, channel noise, and numerical rounding errors. Optionally, in one specific embodiment, the security margin... Take 0.0005.
[0086] Furthermore, candidate state samples that meet the above-mentioned safe zone conditions are accepted; otherwise, they are rejected. Through this safe zone screening mechanism, all accepted samples are located outside the boundary neighborhood centered on the quantization boundary, i.e., they fall within the acceptable safe zone, thereby ensuring that even if there is a synchronization error between the sender and receiver, the state samples of both parties are still quantized to the same discrete interval.
[0087] In practice, the sending end needs to quantize each accepted state sample. This involves dividing the sample's value by the quantization step size and rounding down to obtain an integer value, which is then used as the sample's quantization value. All the quantization values of the accepted samples are arranged sequentially to form the sending end's quantization sequence, which is temporarily stored locally for later use in generating dynamic session seeds. This quantization sequence is not transmitted externally.
[0088] Specifically, during the execution of step S2.3, the following steps are included: sending the sampling point index and state component index corresponding to the accepted state sample as public auxiliary information to the receiving end. The sampling point index is used to indicate the time position of the accepted state sample within the stable sampling window, and the state component index is used to indicate the state dimension to which the sample belongs in the multidimensional state system.
[0089] In fact, the publicly available auxiliary information also includes the quantization step size, the security margin, random numbers, image size, algorithm identifier, and channel identifier or domain identifier. The publicly available auxiliary information does not contain core key materials such as synchronization state values, synchronization error values, quantization values, quantization sequences, pre-shared master keys, dynamic session seeds, and keystreams. Because the publicly available auxiliary information only contains location identifiers and system parameters, and does not contain any actual key materials, even if intercepted by an attacker, it is impossible to reconstruct the dynamic session seed or keystream from it.
[0090] In some embodiments, the process of performing step S3 includes:
[0091] S3.1 Reading the corresponding sample;
[0092] S3.2 Generation of quantization sequences.
[0093] Specifically, during the execution of step S3.1, the receiving end locates and extracts the corresponding state value from the local synchronization state buffer as the corresponding sample based on the sampling point index and state component index in the publicly available auxiliary information.
[0094] In fact, since the transmitting and receiving ends have entered a near-synchronous state through the unconstrained pulse synchronization system, and the selected samples are all far from the quantization boundary (satisfying the safe zone quantization condition), although there may be slight differences between the state samples read by the transmitting end and the corresponding samples read by the receiving end, they are both within the same quantization interval.
[0095] Further, in step S3.2, the receiving end performs quantization on the corresponding sample using the same preset quantization step size and quantization method as the sending end, to obtain a quantized sequence.
[0096] In fact, since all accepted samples meet the safe zone quantization conditions and the synchronization error is controlled within the preset synchronization error threshold, the difference between the sending end state sample and the corresponding receiving end sample is less than the preset synchronization error threshold. Therefore, the sending end state sample and the corresponding receiving end sample will not be quantized to different discrete intervals, and the quantization sequences obtained by both parties are consistent.
[0097] In one specific embodiment, see Figure 2After receiving the sampling point index and the state component index, the receiving end reads the corresponding sample from the local synchronization state. And perform the same quantization. Since the accepted sample is far from the quantization boundary, under the condition that the synchronization error is controlled, and The quantized sequence will not be quantized to different intervals, therefore the quantized sequence obtained by the sender Reconstructing the quantization sequence with the receiver Maintain consistency.
[0098] In some embodiments, the process of performing step S4 includes:
[0099] S4.1 Input to the key derivation function;
[0100] S4.2 Generation of dynamic session seeds.
[0101] Specifically, during step S4.1, the sending end and the receiving end respectively input the quantization sequence, the pre-shared master key, and the public auxiliary information into the key derivation function. The derivation relationship of the dynamic session seed is as follows:
[0102] ;
[0103] ;
[0104] in, This is the dynamic session seed for the sending end. This is the dynamic session seed for the receiving end. The quantization sequence of the transmitting end, The quantization sequence for the receiving end, For the pre-shared master key, The random number, The image size is... The algorithm identifier, This refers to the channel identifier or domain identifier.
[0105] Specifically, the pre-shared master key This is confidential material pre-shared between the sender and receiver, and is not transmitted over public channels. The random number... A unique random number is generated for each image or each session to ensure the freshness of the session seed.
[0106] Further, in step S4.2, the sending end and the receiving end generate a consistent dynamic session seed based on the key derivation function.
[0107] In some embodiments, the key derivation function is a key derivation function constructed based on a cryptographic hash function, an extensible output function, or a message authentication code. In other embodiments, the key derivation function uses SHAKE256; in other embodiments, KMAC256 or HKDF may be used. When using SHAKE256, the output length is 256 bits or other preset length.
[0108] In fact, due to the quantization sequence at the sending end Quantization sequence at the receiving end Maintain consistency, and both parties use the same pre-shared master key. Random numbers Image size Algorithm identifier and channel signage Therefore, the dynamic session seed generated by both parties and Maintain consistency.
[0109] In some embodiments, during the execution of step S5, see [reference needed]. Figure 3 ,include:
[0110] S5.1 Generation of encryption parameters;
[0111] S5.2 Image preprocessing;
[0112] S5.3, Scrambling;
[0113] S5.4, positive feedback diffusion;
[0114] S5.5, Reverse Feedback Diffusion.
[0115] Specifically, during the execution of step S5.1, the sending end generates a non-repeating scrambling index, a diffusion key stream, and a feedback substitution table based on the dynamic session seed.
[0116] Specifically, the sender takes a dynamic session seed as input and, combined with different domain separation tags, derives a scrambled byte stream and a forward-diffusion key stream, respectively. Reverse diffusion key stream Positive feedback substitution table and reverse feedback substitution table Both the positive and negative feedback substitution tables are lookup tables containing 256 bytes of elements, with their inputs and outputs belonging to the integer set {0, 1, ..., 255}. The sending and receiving ends use the same dynamic session seed, domain separation tag, and substitution table generation rules to generate the same scrambling index, diffusion keystream, and feedback substitution table.
[0117] In one implementation, the positive feedback substitution table and the negative feedback substitution table are generated by Fisher-Yates shuffling driven by the corresponding domain-separated byte stream to obtain a bijective mapping on the integer set {0,1,…,255}.
[0118] Further, step S5.2 is executed to flatten the image to be transmitted. The color image to be transmitted is flattened according to the RGB channel order to a length... The byte vector, denoted as plaintext vector. ,in Image height, This is the image width. For a grayscale image, the flattened length is... .
[0119] Furthermore, step S5.3 is executed, whereby the flattened plaintext vector is scrambled based on the non-repeating scrambling index to obtain a scrambled byte vector. The scrambling operation disrupts the positional correlation between image pixels, making the ciphertext statistically difficult to analyze.
[0120] During step S5.4, a forward-diffusion keystream derived from the dynamic session seed and a forward-feedback substitution table are used to perform forward-feedback diffusion on the scrambled byte vector. The scrambled byte vector is set to... The intermediate byte vector after forward diffusion is The expression is:
[0121] ;
[0122] ;
[0123] in, For the first in the forward diffusion key stream One key byte, As a positive feedback replacement table, This is the forward propagation result of the previous byte. By mapping the forward propagation result of the previous byte through a forward feedback substitution table, it participates in the current byte operation, allowing plaintext byte changes to propagate in the forward direction.
[0124] Furthermore, during step S5.5, the reverse diffusion keystream derived from the dynamic session seed and the reverse feedback substitution table are used to perform reverse feedback diffusion on the intermediate byte vector after forward diffusion. Let the final ciphertext byte vector be... The expression is:
[0125] ;
[0126] ;
[0127] in, For the first in the backdiffusion key stream One key byte, As a substitute table for reverse feedback, This is the back-diffusion result of the next byte. By mapping the back-diffusion result of the next byte through a back-feedback substitution table and then participating it in the current byte operation, changes in the plaintext byte propagate backward. In fact, the bidirectional feedback diffusion structure allows even a tiny change in any pixel in the plaintext image to propagate throughout the entire ciphertext image, significantly enhancing the ciphertext's sensitivity to both the plaintext and the key, and effectively resisting differential and statistical attacks.
[0128] Finally, the byte vectors after backfeedback diffusion are rearranged into a ciphertext image of the same size as the original image and output.
[0129] In some embodiments, the process of performing step S6 includes:
[0130] S6.1 Reconstruction of encryption parameters;
[0131] S6.2, Reverse diffusion;
[0132] S6.3, Reverse forward diffusion;
[0133] S6.4, Reverse Scrambling.
[0134] Specifically, during the execution of step S6.1, the receiving end reconstructs the scrambling index, the diffusion key stream, and the feedback substitution table based on the same dynamic session seed.
[0135] In fact, since the receiver and the sender have the same dynamic session seed and both use the same domain separation label and pseudo-random expansion method, the receiver can reconstruct the same scrambling index, forward diffusion key stream, reverse diffusion key stream and feedback substitution table as the sender.
[0136] Further, step S6.2 is executed, performing reverse backdiffusion on the received ciphertext image. The receiving end first uses the reconstructed reverse-diffusion keystream. and reverse feedback substitution table For ciphertext byte vectors Perform reverse diffusion to restore the intermediate vector of forward diffusion. Reverse-diffusion is the inverse operation of reverse feedback diffusion, recovering the byte vector after forward diffusion using the reverse-diffusion keystream and the feedback substitution table. The expression is:
[0137] ;
[0138] ;
[0139] Further, step S6.3 is executed, performing reverse forward diffusion on the result after reverse backdiffusion. The receiving end then utilizes the reconstructed forward diffusion keystream. and positive feedback alternative table For the intermediate vector Perform reverse forward propagation to restore the scrambled byte vector. Reverse forward diffusion is the inverse operation of forward feedback diffusion, using the forward diffusion keystream and feedback substitution table to recover the scrambled byte vector. The expression is:
[0140] ;
[0141] ;
[0142] Furthermore, step S6.4 is executed to perform reverse scrambling on the de-diffused byte vector. Reverse scrambling is the inverse operation of scrambling, and it recovers the original plaintext byte vector by using the inverse mapping of the reconstructed, non-repeating scrambling index.
[0143] Finally, the recovered plaintext byte vectors are rearranged according to the channel order and size of the original image to restore the original image.
[0144] In a second aspect of the invention, a dynamic key image secure transmission system for time-delay communication is also provided, comprising a synchronization state generation module, a secure zone quantization module, a quantization sequence reconstruction module, a dynamic session seed derivation module, an image encryption module, and an image decryption module, wherein:
[0145] The synchronization state generation module is used to operate the transmitter and receiver based on the unrestricted pulse synchronization system. When the synchronization error between the transmitter and receiver enters below the preset synchronization error threshold and remains stable, a stable sampling window is determined.
[0146] The safe zone quantization module is used to determine the quantization boundary set based on a preset quantization step size. Within the stable sampling window, the transmitting end reads the state sample, calculates the distance between the state sample and the quantization boundary set, and accepts the state sample only when the distance is greater than the sum of the synchronization error threshold and the safe zone margin. The sampling point index and state component index corresponding to the accepted state sample are sent to the receiving end as public auxiliary information.
[0147] The quantization sequence reconstruction module is used by the receiving end to read the corresponding sample from the local synchronization state based on the publicly available auxiliary information, and to perform quantization using the same preset quantization step size as the sending end to obtain the quantization sequence;
[0148] The dynamic session seed derivation module is used to input the quantization sequence, the pre-shared master key and the public auxiliary information into the key derivation function at the sending end and the receiving end respectively to generate a consistent dynamic session seed.
[0149] The image encryption module is used by the sending end to generate a non-repeating scrambling index, a diffusion key stream, and a feedback substitution table based on the dynamic session seed, and to output an encrypted image after performing flattening, scrambling, positive feedback diffusion, and reverse feedback diffusion on the image to be transmitted.
[0150] The image decryption module is used by the receiving end to reconstruct the scrambling index, diffusion key stream, and feedback substitution table based on the same dynamic session seed, and then sequentially perform reverse diffusion, reverse forward diffusion, and reverse scrambling to restore the original image.
[0151] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A dynamic key image security transmission method for time-delay communication, characterized in that, Includes the following steps: Based on the operation of the unrestricted pulse synchronization system, when the synchronization error between the transmitter and receiver falls below a preset synchronization error threshold and remains stable, a stable sampling window is determined. Based on a preset quantization step size, a quantization boundary set is determined. Within the stable sampling window, based on the state sample read by the transmitting end, the distance between the state sample and the quantization boundary set is calculated. The state sample is accepted only when the distance is greater than the sum of the synchronization error threshold and the safety margin. The sampling point index and state component index corresponding to the accepted state sample are sent to the receiving end as public auxiliary information. The receiving end reads the corresponding sample from the local synchronization state based on the publicly available auxiliary information, and performs quantization using the same preset quantization step size as the sending end to obtain the quantized sequence; The sending and receiving ends respectively input the quantization sequence, the pre-shared master key, and the public auxiliary information into the key derivation function to generate a consistent dynamic session seed; The sending end generates a non-repeating scrambling index, a diffusion key stream, and a feedback substitution table based on the dynamic session seed. After performing flattening, scrambling, forward feedback diffusion, and reverse feedback diffusion on the image to be transmitted, it outputs a ciphertext image. The receiving end reconstructs the scrambling index, diffusion key stream, and feedback substitution table based on the same dynamic session seed, and sequentially performs reverse diffusion, reverse forward diffusion, and reverse scrambling to restore the original image.
2. The method for secure transmission of dynamic key images for time-delay communication according to claim 1, characterized in that, The non-restricted pulse synchronization system of the first aspect The second sampling time delay The third pulse interval The fourth pulse interval Respectively satisfy Wherein, The sampling time delay upper limit, The pulse interval upper limit, and does not require The pulse interval The time interval between the time of two adjacent pulses, and the sampling time delay The time experienced by the sending end state sampling value to reach the receiving end.
3. The dynamic key image secure transmission method for time-delay communication according to claim 2, wherein, The process of operating the transmitter and receiver based on the unrestricted pulse synchronization system includes: The sending end is running the target neural network state The receiving end operates a coupled neural network state. and will Defined as synchronization error; the receiver maintains a historical synchronization error buffer, at the pulse moment Read and sampling delay Corresponding historical synchronization error and according to the pulse sampling controller Implement pulse control; the synchronization error after pulse application satisfies ,in, For continuous time variables, and These represent the times before and after the k-th pulse, respectively. For the control gain matrix, This is the Dirac impulse function.
4. The dynamic key image secure transmission method for time-delay communication according to claim 1, wherein, The process of the synchronization error falling below a preset synchronization error threshold and remaining stable includes: The synchronization error between the sending end and the receiving end is continuously monitored. When the maximum synchronization error is always less than or equal to the preset synchronization error threshold within a continuous preset time period, the end time of the continuous preset time period is determined as the starting point of the stable sampling window, and the stable sampling window is formed only during the period when the condition is continuously met. When the condition is no longer met, the extraction of dynamic key materials from the synchronization state is suspended.
5. The dynamic key image secure transmission method for time-delay communication according to claim 1, wherein, The process of determining the quantization boundary set based on the preset quantization step size includes: the quantization boundary set is a set composed of integer multiples of the preset quantization step size, and the quantization adopts a uniform quantization method with rounding down. The set of quantization boundaries is: ; The quantization method used is uniform quantization: ; wherein, is the set of quantization boundaries, is the preset quantization step size, is the set of integers, is the state sample read by the sending end.
6. The dynamic key image secure transmission method for time-delay communication according to claim 5, wherein, The distance is the absolute value of the difference between the state sample and the nearest quantization boundary in the quantization boundary set; the safe zone quantization condition is: ; wherein, is the preset synchronization error threshold, is the safety zone margin.
7. The dynamic key image secure transmission method for time-delay communication according to claim 1, wherein, In the process of sending the sampling point index and state component index corresponding to the accepted state sample to the receiving end as public auxiliary information, the public auxiliary information also includes the quantization step size, the safety margin, random number, image size, algorithm identifier, and channel identifier or domain identifier.
8. The dynamic key image secure transmission method for time-delay communication according to claim 7, wherein, The derivation relationship of the dynamic session seed is as follows: ; ; wherein, is the sender dynamic session seed, is the receiver dynamic session seed, is the sender quantization sequence, is the receiver quantization sequence, is the pre-shared master key, is the random number, is the image size, is the algorithm identification, is the channel identification or domain identification.
9. A method for secure transmission of dynamic key images for time-delay communication according to claim 1, characterized in that, The key derivation function is a key derivation function constructed based on a cryptographic hash function, an extensible output function, or a message authentication code; the key derivation function derives a scrambled byte stream, a forward diffusion key stream, a reverse diffusion key stream, a forward feedback substitution table, and a reverse feedback substitution table by separating tags with different fields.
10. A dynamic key image secure transmission system for time-delay communication, characterized in that, It includes a synchronization state generation module, a secure area quantization module, a quantization sequence reconstruction module, a dynamic session seed derivation module, an image encryption module, and an image decryption module, among which: The synchronization state generation module is used to operate the transmitter and receiver based on the unrestricted pulse synchronization system. When the synchronization error between the transmitter and receiver enters below the preset synchronization error threshold and remains stable, a stable sampling window is determined. The safe zone quantization module is used to determine the quantization boundary set based on a preset quantization step size. Within the stable sampling window, the transmitting end reads the state sample, calculates the distance between the state sample and the quantization boundary set, and accepts the state sample only when the distance is greater than the sum of the synchronization error threshold and the safe zone margin. The sampling point index and state component index corresponding to the accepted state sample are sent to the receiving end as public auxiliary information. The quantization sequence reconstruction module is used by the receiving end to read the corresponding sample from the local synchronization state based on the publicly available auxiliary information, and to perform quantization using the same preset quantization step size as the sending end to obtain the quantization sequence; The dynamic session seed derivation module is used to input the quantization sequence, the pre-shared master key and the public auxiliary information into the key derivation function at the sending end and the receiving end respectively to generate a consistent dynamic session seed. The image encryption module is used by the sending end to generate a non-repeating scrambling index, a diffusion key stream, and a feedback substitution table based on the dynamic session seed, and to output an encrypted image after performing flattening, scrambling, positive feedback diffusion, and reverse feedback diffusion on the image to be transmitted. The image decryption module is used by the receiving end to reconstruct the scrambling index, diffusion key stream, and feedback substitution table based on the same dynamic session seed, and then sequentially perform reverse diffusion, reverse forward diffusion, and reverse scrambling to restore the original image.
