Nonlinear chaos masking encryption system and method based on dynamic inversion
By using plaintext signals as a nonlinear interference source and neural network demodulation in an optical chaotic communication system to generate chaotic carriers, the problems of easy decryption and insufficient noise resistance of the system are solved, and high-security and low-complexity optical communication is achieved.
Patent Information
- Application Number
- CN202511450840.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-11
- Publication Date
- 2026-03-03
AI Technical Summary
Existing optical chaotic communication systems are vulnerable to physical layer attacks, as their time delay parameters are easily deciphered, resulting in high system complexity and insufficient noise resistance.
A nonlinear chaotic masking encryption system based on dynamic inversion is adopted. The plaintext signal is used as a nonlinear interference source. A chaotic carrier is generated by a Mach-Zehnder modulator and an interferometer, and a neural network is used for demodulation. This simplifies the system structure and enhances security and noise resistance.
It effectively suppresses time-delay signal characteristics, reduces system complexity, and improves security and noise resistance, achieving high-security and low-complexity communication.
Smart Images
Figure CN121603191A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of optical communication technology, and more specifically, to a nonlinear chaotic masking encryption system and method based on dynamic inversion. Background Technology
[0002] Optical communication is widely used for high-speed data transmission in modern society, making its security paramount. To ensure information security, upper-layer encryption algorithms are frequently used to encrypt data; however, these algorithms are insufficient to effectively defend against all security threats. Attacks targeting the physical layer, such as fiber bending and fiber optic eavesdropping, still exist. To address this issue, various physical layer encryption methods have been proposed, including quantum stream cryptography, chaos-based secure communication, optical steganography, and hardware encryption. Among these, chaos-based secure communication is commonly used to encrypt high-speed optical signals, achieving speeds of tens of Gbps. In 2005, Argyris et al. achieved 1Gbps secure communication over 120 kilometers of optical fiber in Athens, Greece, using optical chaos. Since then, optical chaotic communication (OCC) has been extensively researched and developed, becoming one of the most promising physical layer encryption technologies.
[0003] In an OCC system, information is masked by embedding a chaotic carrier at the transmitting end. To achieve chaotic synchronization, the receiving end needs the same system architecture and parameters as the transmitting end. After synchronization, the receiving end can reconstruct the chaotic portion of the transmitted signal. Then, by comparing the encrypted signal and the chaotic carrier, the information can be directly decoded. According to Kerckhoffs' principle in modern cryptography, the OCC system architecture is public. The security of the OCC system stems from the extreme sensitivity of chaotic systems to their parameters. Because if the system parameters cannot be precisely matched, attackers cannot achieve chaotic synchronization and therefore cannot decrypt the signal. Currently, optical chaotic sources are mainly designed based on time-delayed chaotic systems, and the time delay parameter, called the Time Delay Signature (TDS), is often used as the encryption key. Once the TDS is broken, the system dynamics can be easily reconstructed. Many successful methods for extracting TDS have been proposed, such as fill factor analysis, local linear models, utilizing the autocorrelation function (ACF), and delayed mutual information (DMI).
[0004] To suppress TDS (Time-Distortion Delay), several effective solutions have been developed, broadly categorized into three types: high-dimensional coupling, post-processing, and injection of nonlinear perturbations. For example, the system proposed by Zhong Dongzhou's team couples four polarization components. Zhang Zuxing's team numerically proved an OCC (Optical Coherence Control) system composed of dispersive components and a laser. Gao Zhensen's team added an electro-optic self-feedback module to the system or introduced a chirped fiber Bragg grating for secondary encryption of the message. Nguimdo et al. integrated pseudo-random digital sequences into a phase-chaotic electro-optic delay system to introduce nonlinear perturbations. However, due to the introduction of more components, these systems inevitably suffer from additional interference, requiring the matching of more parameters, thus increasing the physical complexity of implementation. Summary of the Invention
[0005] The purpose of this invention is to provide a nonlinear chaotic masking encryption system and method based on dynamic inversion, which can reduce system complexity and enhance system security and noise resistance.
[0006] This invention provides a nonlinear chaotic masking encryption system based on dynamic inversion, comprising a transmitter and a receiver. The transmitter includes a first driving laser diode, a Mach-Zehnder modulator, an optical signal splitting module, a first time delay line, a first transmitter photodetector, a first transmitter RF amplifier, a second driving laser diode, a transmitter electro-optic phase modulator, a second time delay line, a transmitter Mach-Zehnder interferometer, a second transmitter photodetector, a second transmitter RF amplifier, and a plaintext signal mixing module. The receiver includes a first receiver photodetector, a first receiver RF amplifier, an electrical signal splitting module, a third driving laser diode, a receiver electro-optic phase modulator, a receiver Mach-Zehnder interferometer, a second receiver photodetector, a second receiver RF amplifier, and a neural network module.
[0007] The present invention also provides a nonlinear chaotic masking encryption method based on dynamic inversion, comprising: encrypting the target plaintext signal using the above-mentioned nonlinear chaotic masking encryption system based on dynamic inversion.
[0008] The implementation of the nonlinear chaotic masking encryption system and method based on dynamic inversion provided by this invention has the following beneficial effects: This invention utilizes plaintext signals As a nonlinear interference source, it can replace the pseudo-random number generator, thereby simplifying the system structure. This invention integrates the plaintext into the generation of the chaotic carrier by using a Mach-Zehnder modulator (MZM) and a Mach-Zehnder interferometer (MZI), instead of linear superposition in traditional chaotic masking, effectively suppressing the system's TDS. This invention uses a neural network for signal demodulation, reducing the requirements for high-precision matching of physical devices at both ends of the communication and significantly reducing system complexity.
[0009] This invention, based on a two-dimensional electro-optic chaotic time-delay system, designs an information loading strategy based on plaintext signal injection and a demodulation strategy based on a neural network (NN). Without increasing system complexity, it effectively enhances system security. Furthermore, simulation results show that the proposed demodulation scheme also exhibits strong noise resistance. Attached Figure Description
[0010] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a schematic diagram of the nonlinear chaotic masking encryption system based on dynamic inversion provided by the present invention. Figure 2 This invention provides system signals when there is no plaintext injection. TDS diagram; Figure 3 This is a schematic diagram of various signals in a communication system provided by the present invention, under the conditions of a plaintext rate of 2Gbps and a signal-to-noise ratio (SNR) of 7dB. Figure 4 This is a schematic diagram of BER under different plaintext rates and signal-to-noise ratios provided by the present invention. Detailed Implementation
[0011] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0012] Figure 1 A schematic diagram of the nonlinear chaotic masking encryption system based on dynamic inversion according to this embodiment is shown. In this embodiment, the nonlinear chaotic masking encryption system based on dynamic inversion includes a transmitter and a receiver. The transmitter includes a first driving laser diode, a Mach-Zehnder modulator, an optical signal splitter module, a first time delay line, a first transmitter photodetector, a first transmitter RF amplifier, a second driving laser diode, a transmitter electro-optic phase modulator, a second time delay line, a transmitter Mach-Zehnder interferometer, a second transmitter photodetector, a second transmitter RF amplifier, and a plaintext signal mixing module. The receiving end includes a first receiving end photodetector, a first receiving end radio frequency amplifier, an electrical signal splitter module, a third driving laser diode, a receiving end electro-optic phase modulator, a receiving end Mach-Zehnder interferometer, a second receiving end photodetector, a second receiving end radio frequency amplifier, and a neural network module.
[0013] In one exemplary embodiment, the first driving laser diode is used to output an optical signal to drive the Mach-Zehnder modulator; the Mach-Zehnder modulator is used as a nonlinear modulator; the optical signal splitting module is used to split the output of the Mach-Zehnder modulator into two paths, one of which is transmitted to the receiving end, and the other path is sent to the first transmitting end photodetector after passing through the first time delay line.
[0014] In one exemplary embodiment, the first transmitting photodetector is used to convert the optical signal from the first time delay line into an electrical signal; the first transmitting radio frequency amplifier is used to amplify the electrical signal from the first transmitting photodetector; and the second driving laser diode is used to drive the transmitting electro-optic phase modulator.
[0015] In one exemplary embodiment, the transmitting electro-optic phase modulator is modulated by the signal output from the first transmitting radio frequency amplifier; the output of the transmitting electro-optic phase modulator enters the transmitting Mach-Zehnder interferometer after passing through the second time delay line.
[0016] In one exemplary embodiment, the transmitting Mach-Zehnder interferometer is used to perform nonlinear and dynamic phase-intensity conversion on the signal output from the second time delay line; the second transmitting photodetector is used to convert the optical signal from the transmitting Mach-Zehnder interferometer into an electrical signal; the second transmitting radio frequency amplifier is used to amplify the electrical signal from the second transmitting photodetector; and the plaintext signal mixing module is used to mix the signal from the second transmitting radio frequency amplifier with the plaintext signal and then feed it back to the Mach-Zehnder modulator to achieve nonlinear masking.
[0017] In one exemplary embodiment, the first receiving-end photodetector is used to convert the optical signal from the Mach-Zehnder modulator into an electrical signal; the first receiving-end radio frequency amplifier is used to amplify the electrical signal from the first receiving-end photodetector; the electrical signal splitting module is used to split the electrical signal output by the first receiving-end radio frequency amplifier into two paths, one path being input to the neural network module after time delay, and the other path being input to the receiving-end electro-optic phase modulator.
[0018] In one exemplary embodiment, the third driving laser diode is used to drive the receiving-end electro-optic phase modulator; the receiving-end electro-optic phase modulator is modulated by the signal output from the first receiving-end radio frequency amplifier; the receiving-end Mach-Zehnder interferometer is used to perform phase-to-intensity conversion on the signal output from the receiving-end electro-optic phase modulator; the second receiving-end photodetector is used to convert the optical signal from the receiving-end Mach-Zehnder interferometer into an electrical signal; the second receiving-end radio frequency amplifier is used to amplify the electrical signal from the second receiving-end photodetector; the electrical signal output from the second receiving-end radio frequency amplifier is input to the neural network module; the neural network module is used to decrypt the electrical signals output from the first receiving-end radio frequency amplifier and the second receiving-end radio frequency amplifier to obtain the decrypted plaintext signal.
[0019] In one exemplary embodiment, the training process of the neural network module includes: An arbitrary binary signal is injected as a plaintext signal into the transmitting end, and the receiving end is used to obtain the electrical signals output by the first receiving end RF amplifier and the second receiving end RF amplifier. The electrical signals output from the first and second receiver RF amplifiers are used as the two-dimensional inputs to the neural network module, and the binary signal is used as the desired output. The neural network module is trained using a loss function to obtain a trained neural network module.
[0020] In one exemplary embodiment, the loss function is as follows: , in, The error signal is the loss function. This is the output vector of the neural network module; This represents the expected output of the neural network module.
[0021] This embodiment provides a nonlinear chaotic masking encryption method based on dynamic inversion, including: encrypting the target plaintext signal using the aforementioned nonlinear chaotic masking encryption system based on dynamic inversion.
[0022] In some embodiments, the above-described nonlinear chaotic masking encryption system based on dynamic inversion can also be implemented in the following ways.
[0023] In this embodiment, the working principle of the nonlinear chaotic masking encryption system based on dynamic inversion is referenced. Figure 1 .
[0024] like Figure 1 As shown in (a), the transmitting end consists of two similar, cascaded nonlinear time delay chains. Laser diode ( The output light is used to drive a Mach-Zehnder modulator (MZM) that functions as a nonlinear modulator. The output of the MZM is divided into two parts: one part is transmitted to the receiver, and the other part passes through a time delay line (…). After transmission, it is detected by a photodetector. The signal is received and converted into an electrical signal. Then, the radio frequency amplifier (RF amplifier) )enlarge The output is used to generate a signal. . Drive the electro-optic phase modulator (PM) by the signal Modulation is performed. The output of the PM is processed... The signal then enters a Mach-Zehnder interferometer (MZI) for nonlinear and dynamic (time-nonlocal) phase-intensity conversion. This conversion can be performed by a function... Description. Among them, This indicates the optical power output of the MZI. Indicates the phase of the input light. This represents the offset interference phase of the MZI. Actual interferometers have a delay. The imbalance causes the interference conditions to change over time. and The phase difference between them determines the outcome. enlarge The output generates a signal. This signal is different from the plaintext signal. The mixture is then fed back to the MZM to achieve nonlinear masking. The dynamic model at the transmitter can be described as follows:
[0025] in, For plaintext signals, and For system variables, For feedback strength, and This is the time delay parameter. This is the ratio of the low cutoff frequency to the high cutoff frequency of the corresponding link in MZM or MZI. For the offset phase of MZM, This is the offset phase of MZI.
[0026] The receiver is shown in Figure 1(b). The received signal passes directly through... and The generated electrical signal is split into two paths, one of which is delayed to obtain... And input it into the trained NN, another route PM' is provided, and the output of PM' passes through MZI'. The phase-modulated signal is then... and The signal is obtained after processing and time delay. Finally, the signal and Input a neural network NN, output the decrypted plaintext signal .
[0027] The training process of NN is as follows Figure 1 As shown in (c). This embodiment uses arbitrary binary signals. Injected into the transmitter, generating an electrical signal and As a two-dimensional input to the neural network, As the expected output, the output of NN is a vector. Error signal Used as a loss function for training, the training objective is to make waveform and Alignment. Then, the NN can learn from... and Reverse Reconstruction The nonlinear relationship is then used to demodulate the plaintext.
[0028] This embodiment provides a simulation experiment conducted in MATLAB, using a reservoir computing (RC) network to complete the modeling task. The parameters are as follows: number of neurons. Leakage rate spectral radius Regularization coefficient The parameters of the chaotic system are set to... , , , , , , The fourth-order Runge-Kutta algorithm is used to solve the problem, with a step size of [missing value]. . It is a 2Gbps plaintext signal with an amplitude of The channel has additive white Gaussian noise.
[0029] Compared with existing technologies, the advantages of this invention are: 1. It exhibits significant TDS suppression, offering high safety and low complexity: Without plaintext injection ( This embodiment uses a length of Time series from signal Extract TDS from it. For example... Figure 2 As shown in (a) and (b), in and The peak value of TDS can be clearly observed at this point, where and These represent the calculation results for ACF and DMI, respectively. Figure 2 When there is no plaintext injection, system signals TDS.
[0030] When plaintext signals are injected into the transmitter ( It is a random 0 or 1, such as Figure 3 (a) shows the generated chaotic signal like Figure 3 As shown in (b), there is obvious distortion, and the waveform resembles noise. The decrypted signal is as follows: Figure 3 As shown by the solid red line in (c), it closely matches the plaintext signal. Figure 3 (d) shows The corresponding power spectrum does not exhibit clear characteristics in the frequency spectrum. The eye diagram of a legitimate receiver is as follows: Figure 3 As shown in (e), the eye diagram is completely clear and open, and the calculated bit error rate (BER) is: For eavesdroppers, Figure 3 (f) shows the corresponding decoded eye diagram, with a calculated BER of 0.512, indicating that decryption is completely invalid. Furthermore, the system TDS was extracted using ACF and DMI, with the following results: Figure 3 As shown in (g) and (h), there is no obvious peak. Figure 3 A communication system with a plaintext rate of 2Gbps and a signal-to-noise ratio (SNR) of 7dB.
[0031] The above conclusions show that the proposed scheme enables the system to effectively suppress TDS, enhancing security without introducing additional nonlinear terms, thus reducing the system's complexity.
[0032] 2. Good communication performance and strong noise immunity: In this embodiment, additive white Gaussian noise of varying intensities is added to the channel, and the system's communication performance is verified using BER (Breakpoint Error). Figure 4 As shown, the SNR ranges from 4dB to 20dB. With a fixed SNR, the BER increases with increasing message rate; however, with a constant plaintext rate, the BER increases with decreasing SNR. Specifically, for plaintext rates of 1Gbps, 2Gbps, 3Gbps, 4Gbps, 5Gbps, 6Gbps, and 8Gbps, when the SNR satisfies the following conditions... , , , , , and Demodulation was successful at all times. Clearly, the proposed system exhibits strong noise immunity, achieving Gbps-level secure communication even at SNRs as low as 4dB. Figure 4 BER at different plaintext rates and signal-to-noise ratios.
[0033] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.
Claims
1. A nonlinear chaotic masking encryption system based on dynamic inversion, characterized in that, The system includes a transmitting end and a receiving end. The transmitting end includes a first driving laser diode, a Mach-Zehnder modulator, an optical signal splitting module, a first time delay line, a first transmitting end photodetector, a first transmitting end RF amplifier, a second driving laser diode, a transmitting end electro-optic phase modulator, a second time delay line, a transmitting end Mach-Zehnder interferometer, a second transmitting end photodetector, a second transmitting end RF amplifier, and a plaintext signal mixing module. The receiving end includes a first receiving end photodetector, a first receiving end RF amplifier, an electrical signal splitting module, a third driving laser diode, a receiving end electro-optic phase modulator, a receiving end Mach-Zehnder interferometer, a second receiving end photodetector, a second receiving end RF amplifier, and a neural network module.
2. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 1, characterized in that, The first driving laser diode is used to output an optical signal to drive the Mach-Zehnder modulator; the Mach-Zehnder modulator is used as a nonlinear modulator. The optical signal splitting module is used to split the output of the Mach-Zehnder modulator into two paths. One path is transmitted to the receiving end, and the other path is sent to the first transmitting end photodetector after passing through the first time delay line.
3. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 1, characterized in that, The first transmitting end photodetector is used to convert the optical signal from the first time delay line into an electrical signal; The first transmitting end radio frequency amplifier is used to amplify the electrical signal from the first transmitting end photodetector; The second driving laser diode is used to drive the transmitting electro-optic phase modulator.
4. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 1, characterized in that, The transmitting end electro-optic phase modulator is modulated by the signal output from the first transmitting end radio frequency amplifier; The output of the transmitting electro-optic phase modulator enters the transmitting Mach-Zehnder interferometer after passing through the second time delay line.
5. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 1, characterized in that, The transmitting end Mach-Zehnder interferometer is used to perform nonlinear and dynamic phase-intensity conversion on the signal output from the second time delay line; the second transmitting end photodetector is used to convert the optical signal from the transmitting end Mach-Zehnder interferometer into an electrical signal. The second transmitting end radio frequency amplifier is used to amplify the electrical signal from the second transmitting end photodetector; The plaintext signal mixing module is used to mix the signal from the second transmitting end RF amplifier with the plaintext signal, and then feed it back to the Mach-Zehnder modulator to achieve nonlinear masking.
6. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 1, characterized in that, The first receiving end photodetector is used to convert the optical signal from the Mach-Zehnder modulator into an electrical signal; the first receiving end radio frequency amplifier is used to amplify the electrical signal from the first receiving end photodetector. The electrical signal splitting module is used to split the electrical signal output by the first receiving end radio frequency amplifier into two paths, one of which is input to the neural network module after time delay, and the other of which is input to the receiving end electro-optic phase modulator.
7. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 1, characterized in that, The third driving laser diode is used to drive the receiving-end electro-optic phase modulator; the receiving-end electro-optic phase modulator is modulated by the signal output from the first receiving-end radio frequency amplifier; the receiving-end Mach-Zehnder interferometer is used to perform phase-to-intensity conversion on the signal output from the receiving-end electro-optic phase modulator; the second receiving-end photodetector is used to convert the optical signal from the receiving-end Mach-Zehnder interferometer into an electrical signal; the second receiving-end radio frequency amplifier is used to amplify the electrical signal from the second receiving-end photodetector. The electrical signal output by the second receiving end RF amplifier is input to the neural network module; the neural network module is used to decrypt the electrical signals output by the first receiving end RF amplifier and the second receiving end RF amplifier to obtain the decrypted plaintext signal.
8. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 1, characterized in that, The training process of the neural network module includes: An arbitrary binary signal is injected as a plaintext signal into the transmitting end, and the receiving end is used to obtain the electrical signals output by the first receiving end RF amplifier and the second receiving end RF amplifier. The electrical signals output from the first and second receiver RF amplifiers are used as the two-dimensional inputs to the neural network module, and the binary signal is used as the desired output. The neural network module is trained using a loss function to obtain a trained neural network module.
9. The nonlinear chaotic masking encryption system based on dynamic inversion according to claim 8, characterized in that, The loss function is as follows: ， in, The error signal is the loss function. This is the output vector of the neural network module; This represents the expected output of the neural network module.
10. A nonlinear chaotic masking encryption method based on dynamic inversion, characterized in that, include: The target plaintext signal is encrypted using the nonlinear chaotic masking encryption system based on dynamic inversion as described in any one of claims 1-9.