Communication method and apparatus of a semantic communication system, electronic device and medium
By introducing semantic encryption, perturbation, and signature modules into the semantic communication system, the security protection problem of the semantic communication system is solved, and stronger information transmission security and robustness are achieved.
Patent Information
- Application Number
- CN202211641019.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-20
- Publication Date
- 2025-11-04
- Estimated Expiration
- 2042-12-20
AI Technical Summary
Existing semantic communication systems lack security protection for information transmission, especially given the openness of wireless channels and the vulnerability of neural modules, making them susceptible to security threats such as eavesdropping, privacy breaches, and deception.
Deploying semantic encryption, semantic perturbation, and semantic signature modules in a semantic communication system ensures the security and robustness of information transmission by adding perturbation signals, encrypting, and generating digital signatures to the original semantic information.
It effectively mitigates the privacy risks of wireless channels, protects the privacy and integrity of semantic information, improves the robustness of semantic communication systems, and prevents eavesdropping and attacks.
Smart Images
Figure CN116248288B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of communication data processing, and particularly relates to a communication method and device of a semantic communication system, an electronic device and a medium. BACKGROUND
[0002] In the related art, semantic communication systems are continuously used with the development of various businesses. Among them, the current DLSC (DNN-based End-to-End Communication System) based on deep learning is to extract and encode semantic information by means of a deep neural network, and then realize the purpose of signal transmission by means of a wireless channel.
[0003] Further, due to the openness of the wireless channel and the vulnerability of the neural module, the semantic communication system also has serious semantic security problems such as eavesdropping, privacy leakage and fraud. In the related art, the robustness of the semantic communication system is usually strengthened by offline adversarial training of the entire system, and online semantic protection is basically not fully explored. SUMMARY
[0004] Embodiments of the present application provide a communication method and device of a semantic communication system, an electronic device and a medium. To solve the problem of lack of security protection for the information transmission process of the semantic communication system in the related art.
[0005] According to an aspect of an embodiment of the present application, a communication method of a semantic communication system is provided, which is applied to a semantic communication system including a semantic encryption module, a semantic perturbation module and a semantic signature module. The method comprises:
[0006] obtaining original semantic information to be sent;
[0007] adding a perturbation signal to the original semantic information by means of the semantic perturbation module to obtain perturbed semantic information;
[0008] encrypting the perturbed semantic information by means of the semantic encryption module to obtain encrypted semantic information, and generating a digital signature corresponding to the original semantic information by means of the semantic signature module;
[0009] sending the encrypted semantic information and the digital signature to a receiving end at a sending end of the semantic communication system.
[0010] Optionally, in another embodiment based on the above-mentioned method of the present application, the semantic encryption module comprises a semantic encryption component and a semantic decryption component.
[0011] The semantic encryption module is used to encrypt the perturbed semantic information to obtain encrypted semantic information.
[0012] The semantic encryption component in the semantic encryption module is used to encrypt the perturbed semantic information to obtain the encrypted semantic information.
[0013] The semantic encryption component is constructed by a neural structure with multiple convolution layers.
[0014] Optionally, in another embodiment based on the above method of the present application, the semantic signature module is used to generate a digital signature corresponding to the original semantic information, comprising:
[0015] A plurality of vector values are randomly generated by the semantic signature module, and the plurality of vector values are input into a generator of the semantic communication system;
[0016] Each vector value is converted into a semantic vector matching the original semantic information by the generator;
[0017] A target semantic vector most similar to the original semantic information is selected from a plurality of semantic vectors;
[0018] The target semantic vector is used as the digital signature.
[0019] Optionally, in another embodiment based on the above method of the present application, after the target semantic vector is used as the digital signature, it further comprises:
[0020] The encrypted semantic information and the corresponding key information are sent to the receiving end at the sending end of the semantic communication system; and the target semantic vector is sent to the receiving end;
[0021] The key information is the key information generated by negotiation between the sending end and the receiving end of the semantic communication system.
[0022] Optionally, in another embodiment based on the above method of the present application, before the original semantic information to be sent is obtained, it further comprises:
[0023] The semantic encryption module, the semantic perturbation module and the semantic signature module are respectively provided with start conditions, so that the corresponding modules are started when the semantic communication system detects that the start conditions are met.
[0024] According to another aspect of the embodiments of the present application, a communication device of a semantic communication system is provided, which is applied to a semantic communication system comprising a semantic encryption module, a semantic perturbation module and a semantic signature module, and the device comprises:
[0025] an obtaining module configured to obtain original semantic information to be sent;
[0026] an adding module configured to add a disturbance signal in the original semantic information by using the semantic disturbance module to obtain disturbed semantic information;
[0027] an encrypting module configured to perform encryption processing on the disturbed semantic information by using the semantic encryption module to obtain encrypted semantic information, and generate a digital signature corresponding to the original semantic information by using the semantic signature module;
[0028] a sending module configured to send the encrypted semantic information and the digital signature to a receiving end at a sending end of the semantic communication system.
[0029] According to still another aspect of the embodiments of the present application, an electronic device is provided, comprising:
[0030] a memory configured to store executable instructions; and
[0031] a display configured to execute the executable instructions with the memory to complete the operations of the communication method of any of the semantic communication systems.
[0032] According to still another aspect of the embodiments of the present application, a computer readable storage medium is provided, configured to store computer readable instructions, the instructions being executed to perform the operations of the communication method of any of the semantic communication systems.
[0033] In the present application, the semantic encryption module, the semantic disturbance module and the semantic signature module can be deployed in an initial semantic communication system, and after obtaining original semantic information to be sent, a disturbance signal is added in the original semantic information by using the semantic disturbance module to obtain disturbed semantic information; the disturbed semantic information is processed by encryption by using the semantic encryption module to obtain encrypted semantic information; and a digital signature corresponding to the original semantic information is generated by using the semantic signature module; the encrypted semantic information and the digital signature are sent to a receiving end at a sending end of the semantic communication system. By applying the technical solution of the present application, the privacy risk from a wireless channel can be reduced by using a disturbance mechanism, and the information transmission process of the semantic communication system is jointly secured in a way of correcting semantics at the sending end by encrypting the semantic information to be transmitted and by semantic signature. Thus, on one hand, the problem of lacking security protection for the information transmission process of the semantic communication system in the related art is avoided, and on the other hand, a semantic communication system with stronger robustness can be obtained.
[0034] The technical solution of the present application will be described in further detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS
[0035] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the application and, together with the description, serve to explain the principles of the application.
[0036] The application can be more clearly understood and appreciated from the following detailed description, taken in conjunction with the accompanying drawings of which:
[0037] Figure 1 A schematic diagram of a communication method of a semantic communication system according to an embodiment of the application is shown;
[0038] Figure 2 A schematic diagram of a system architecture of a semantic communication system according to an embodiment of the application is shown;
[0039] Figure 3 A flowchart of a communication method of a semantic communication system according to an embodiment of the application is shown;
[0040] Figure 4 A schematic diagram of a structure of an electronic device according to an embodiment of the application is shown;
[0041] Figure 5 A schematic diagram of a structure of an electronic device according to an embodiment of the application is shown;
[0042] Figure 6 A schematic diagram of a storage medium according to an embodiment of the application is shown. DETAILED DESCRIPTION
[0043] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangements, numerical expressions, and numerical values of components and steps set forth in these embodiments are not limiting to the scope of the present application unless otherwise specifically stated.
[0044] It should be understood, of course, that the various embodiments of the present application are not limited to the illustrations described above, but can vary the steps of the methods and the components and parameters of the systems described and / or can be arranged in different ways, all of which are intended to be embraced within the scope of the various embodiments of the present application.
[0045] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the scope of the application or its application or uses.
[0046] Techniques, methods, and devices known to those of ordinary skill in the relevant art can not be discussed in detail herein. However, where appropriate, such techniques, methods, and devices can be viewed as part of the specification and may
[0047] It should be noted that like reference numerals and letters refer to like items in the drawings and that a discussion of the same item under a different reference numeral and letter does not imply that it is a different item or that it is the only one of its type.
[0048] In addition, the technical solutions among the various embodiments of the present application can be combined with each other, but it must be based on the implementation by the ordinary skilled in the art, when the combination of technical solutions appears contradictory or unachievable, it should be considered that the combination of technical solutions does not exist, nor within the scope of protection required by the present application.
[0049] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative positional relationship, movement condition, etc. between components in a certain posture (as shown in the drawings), if the certain posture changes, the directional indications will also change accordingly.
[0050] The communication method of the semantic communication system according to the exemplary embodiments of the present application will be described below in conjunction with Figures 1-3 It should be noted that the following application scenarios are only shown for the purpose of facilitating the understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited in this respect. On the contrary, the embodiments of the present application can be applied to any applicable scenario.
[0051] The present application also proposes a communication method and device of a semantic communication system, an electronic device and a medium.
[0052] Figure 1 The flowchart of the communication method of the semantic communication system according to the embodiments of the present application is schematically shown. As Figure 1 shown, the method comprises:
[0053] S101, obtaining original semantic information to be sent.
[0054] S102, adding a disturbance signal to the original semantic information by using a semantic disturbance module to obtain disturbed semantic information.
[0055] S103, performing encryption processing on the disturbed semantic information by using a semantic encryption module to obtain encrypted semantic information, and generating a digital signature corresponding to the original semantic information by using a semantic signature module.
[0056] S104, sending the encrypted semantic information and the digital signature to the receiving end at the sending end of the semantic communication system.
[0057] Existing wireless communications are built on the classic Shannon information theory, which mainly focuses on how to accurately and efficiently transmit raw bits from the source to the destination through the channel. In the past few decades, to meet the growing demand for traffic, the transmission rate has been improved by tens of thousands of times, gradually pushing the capacity of wireless systems to the Shannon limit. At the same time, various emerging mobile applications, such as virtual reality (VR) and human-to-machine (H2M) communication, often require more intelligence in different aspects, as well as interpretation of received information to make decisions. This prompts us to rethink a more intelligent and more efficient wireless communication paradigm beyond Shannon.
[0058] In related art, a semantic communication (SC) system shows great potential in conveying semantics at the transmission source and accurately interpreting the meaning at the destination. This novel communication mode greatly promotes the above-mentioned mobile applications, such as VR and H2M. Specifically, the transmitter of the semantic communication system relies on a neural network to extract semantic information from the input, and then sends the expressed symbols to a noisy wireless channel. At the receiving end, the system reconstructs the received semantics or interprets the semantics to make intelligent decisions.
[0059] However, compared with traditional wireless communication, the semantic communication system can suffer from more challenging security problems, such as eavesdropping, tampering, and deception due to the broadcast nature of wireless communication and the vulnerability of deep neural networks.
[0060] For example, since anyone within the physical communication range of the transmitter can receive the wireless signal and possibly decode the symbols. And a semantic communication system mainly focuses on the transmission of the meaning conveyed at the source rather than the accurate raw bits, it facilitates malicious attempts by eavesdroppers to perform semantic inference attacks and adversarial attacks.
[0061] In addition, since a semantic communication system is learned from a large number of data instances, which can consist of sensitive information such as user's health status and service access history, the system can inadvertently expose sensitive information in the semantic representation to be sent to the wireless channel, which can facilitate the leakage of the input semantics.
[0062] Finally, due to the vulnerability of neural models, the semantic communication system is vulnerable to attacks by adversaries. The adversary's perturbation can mislead the system to make incorrect semantic interpretation.
[0063] Existing works focus more on improving the robustness of SC by offline adversarial training on the whole system, using adversarial examples to mitigate the blind spots of neural models. These studies can strengthen the resistance of the system to attacks, but they may not be practically deployed in real-world scenarios because it is not feasible to interrupt the communication system for retraining. In addition, these studies are mainly to defend against a single attack and cannot protect semantic communication systems from various threats.
[0064] To solve the above problems, the present application proposes a communication method of a semantic communication system, which idea is to deploy three semantic protection modules in the semantic communication system to protect the semantic communication system during data transmission. Specifically, the present application can use the disturbance mechanism triggered by the semantic disturbance module to reduce the privacy risk from the wireless channel, and through the semantic encryption module to encrypt the semantic information to be transmitted, and through the semantic signature module to perform semantic signature to correct the semantics at the sending end. The way to jointly protect the information transmission process of the semantic communication system.
[0065] As shown in Figure 2 The semantic communication system proposed by the present application includes five modules, which are a semantic encoder module, an OFDM transmitter module, an OFDM receiver module, a semantic decoder module and a classifier module. Among them,
[0066] Semantic encoder module: in one way, for image data, the semantic encoder module in the embodiment of the present application can be constructed based on a convolutional neural network model. In another way, for text data, the semantic encoder module in the embodiment of the present application can be based on a Transformer model. That is, input data is represented as X, and X is input into the semantic encoder Semantic Encoder to obtain the semantic information of the data.
[0067] OFDM transmitter module: in one way, the embodiment of the present application can realize data transmission between each client and the server through the OFDM transmitter. Further, the OFDM transmitter module uses the fast Fourier transform (IFFT) of the semantic information and adds a cyclic prefix to reduce the peak-to-average power ratio (PAPR) through signal clipping. Here, a pilot is also added for channel estimation.
[0068] Wireless channel module: in one way, the wireless transmission channel in the embodiment of the present application adopts a Rayleigh fading channel. For example, the output of the wireless transmission channel can be obtained by giving the input.
[0069] OFDM Receiver Module: In one embodiment of this application, data transmission between clients and the server can be achieved through an OFDM receiver. Furthermore, the OFDM receiver module utilizes the inverse process used by the OFDM transmitter—removing duplicate cyclic prefixes and converting the signal using a Fast Fourier Transform (FFT) to generate frequency-domain pilots and data symbols. Channel estimation and channel equalization are further incorporated to improve transmission performance.
[0070] Semantic decoder module: The semantic decoder corresponds one-to-one with the semantic encoder, and its purpose is to decode the semantic information transmitted on the channel into the original image or text data.
[0071] Classifier module: The classifier receives the data decoded by the semantic decoder and classifies it. In one embodiment, the classifier can measure the semantic information involved in the input data (i.e., images or text).
[0072] Furthermore, such as Figure 3 The diagram shown is a flowchart illustrating the communication method of the semantic communication system proposed in this application, including:
[0073] Step 1: Obtain the original semantic information to be sent.
[0074] In one embodiment, the semantic communication system proposed in this application may be equipped with three hot-swappable semantic protection modules (i.e., SemEryp semantic encryption module, SemPriv semantic perturbation module, and SemRect semantic signature module).
[0075] In one approach, the SemEryp semantic encryption module can encrypt the semantic information to be transmitted at the system's transmitter, thereby achieving the goal of ensuring semantic security on open wireless channels.
[0076] In another approach, the SemPriv semantic perturbation module aims to mitigate privacy breaches through a perturbation generator. Understandably, the perturbation generator can distort malicious adversary behavior.
[0077] In one approach, the SemRect semantic tagging module is able to ensure the interpretation of transmitted semantic information by generating a signature at the transmitter and then calibrating the semantics through the signature at the receiver.
[0078] Optionally, embodiments of this application may also set corresponding startup conditions for each semantic protection module, so that the semantic communication system starts the corresponding module after detecting that one or more startup conditions are met.
[0079] Step 2, adding a disturbance signal in the original semantic information by using a semantic disturbance module to obtain disturbed semantic information.
[0080] For the semantic disturbance module in the present application, the main goal is to protect the privacy of the input original semantic information on the semantic communication system. In one way, the eavesdropper within the physical communication range of the transmitter can receive the wireless signal and possibly decode the semantics, resulting in the privacy leakage of the legitimate user. Therefore, the present application will deploy an adversarial transformation network (ATN) in the semantic communication system to resist attribute inference attack (AIA). This is a destructive privacy attack that can infer the hidden information of the semantics.
[0081] In one way, the present application can deploy an attacker's semantic interpretation (ASI) module on the semantic communication system. It can be understood that this module is a multi-layer perceptron (MLP) and can be regarded as an eavesdropper.
[0082] Then, the semantic representation is fed back to the ATN by the semantic communication system to learn to train the system against the adversarial privacy disturbance caused by the semantic disturbance module by maximizing the cross-entropy loss of ASI, so that the disturbance generated by the semantic disturbance module can mislead the eavesdropper to make wrong predictions. However, such privacy disturbance will not affect the semantic interpretation of the legitimate user, so it can solve the problem of privacy leakage.
[0083] Step 3, encrypting the disturbed semantic information by using a semantic encryption component in the semantic encryption module to obtain encrypted semantic information.
[0084] For the semantic encryption module in the present application, the purpose is to protect the confidentiality of the input original semantic information.
[0085] It can be understood that since the malicious user may reconstruct the input data according to the semantics to perform a model inversion attack (in this attack, the attacker needs to intercept the semantic data and convert it into the original information. The attacker has two methods to obtain the original information through the semantic data, one is to train a substitute model, and the other is to steal the semantic decoder of the receiver). In order to prevent the eavesdropper from obtaining the knowledge of the semantic decoder or training a substitute semantic decoder. This type of attack may cause privacy leakage problems. For example, once the eavesdropper successfully recovers the personal face image conveyed in the wireless channel, they can use this stolen identity to break into other security systems.
[0086] In one embodiment, the semantic encryption module of the present application is composed of two components, namely a semantic encryption component (SEC) and a semantic decryption component (SDC). It should be noted that the SEC and SDS have the same neural structure, and they both use multiple convolutional layers for encryption and decryption.
[0087] The present application simultaneously generates the original semantic information of the input data and the semantic key to the opposite end to output encrypted semantic data, wherein the key is generated by negotiation between the sending end and the receiving end.
[0088] For the receiving end, the semantic decryption component can take the encrypted semantic data and the semantic key as input to obtain the original semantics.
[0089] In one embodiment, the present application can use an adversarial training method during the training of the semantic encryption module. For example, an attacker is introduced during training, and the attacker has a semantic decoder and a semantic restoration component (SRC) with the same structure as SEC and SDS, but the input of this module only has semantic information and does not have a semantic key.
[0090] Specifically, the embodiments of the present application can first train the attacker and train the semantic restoration component SRC by the attacker, so that it can restore the original semantic information without the key. Then train the semantic encryption sub-module and the semantic decryption sub-module of the communication parties, so that they can encrypt the data, so that the semantic restoration atomic module trained by the attacker is invalid.
[0091] It should be noted that the training goal of the attacker is to restore the encrypted semantic information to the original semantic information without the semantic key, while the training goal of the communication parties is to encrypt and decrypt the semantic information under the premise of sharing the semantic key, while making it impossible for the attacker to restore the original semantic information without the semantic key. Through such a training method, the present application can make the encryption sub-module and the decryption sub-module learn how to encrypt and decrypt the data so that the attacker cannot crack the encrypted data. In one embodiment, the present application can repeat the above steps multiple times to train an optimized better semantic encryption module to achieve better security protection.
[0092] Step 4, the semantic signature module randomly generates a plurality of vector values, and inputs the plurality of vector values into the generator of the semantic communication system.
[0093] Step 5, the generator converts each vector value into a semantic vector matching the original semantic information respectively.
[0094] Step 6, select the target semantic vector closest to the original semantic information from the plurality of semantic vectors as the digital signature.
[0095] The semantic signature module in the present application aims to protect the integrity of the transmitted semantic information so as to accurately interpret the transmitted semantic information at the receiving end.
[0096] In one way, the present application generates a plurality of random vectors (RVs), called semantic signatures, to calibrate such perturbations so that the influence of attacks on legitimate users can be appropriately mitigated. The detailed training process is described as follows:
[0097] Firstly, the present application can train a generative adversarial network (GAN) in a semantic communication system, and the generator of the network can generate semantic information as real distribution according to a random vector (RV). In one way, at the sending end, n random vectors can be initialized first, and then they are optimized so that the semantic information generated by the generator is as similar as possible to the original semantic information to be sent by the sender, and then an optimal target semantic vector (i.e., the target semantic vector selected from the plurality of semantic vectors that is most similar to the original semantic information) is found, which is sent to the receiving end as the digital signature of the semantic information.
[0098] It can be understood that at the receiving end, the semantic signature can be used to calibrate the transmitted semantic information. By doing so, the semantic signature module can protect the integrity of the semantic information from the influence of adversarial perturbations in the semantic communication system.
[0099] Step 7, at the sending end of the semantic communication system, the encrypted semantic information and the corresponding key information; and the target semantic vector are sent to the receiving end.
[0100] In the present application, the semantic encryption module, the semantic perturbation module and the semantic signature module can be deployed in the initial semantic communication system, and after obtaining the original semantic information to be sent, the semantic perturbation module is used to add a perturbation signal to the original semantic information to obtain perturbed semantic information; the semantic encryption module is used to encrypt the perturbed semantic information to obtain encrypted semantic information; and the semantic signature module is used to generate a digital signature corresponding to the original semantic information; at the sending end of the semantic communication system, the encrypted semantic information and the digital signature are sent to the receiving end.
[0101] By applying the technical solution of the present application, the privacy risk from the wireless channel can be mitigated by using the perturbation mechanism, and the information transmission process of the semantic communication system can be jointly secured by encrypting the semantic information to be transmitted and correcting the semantics at the sending end through the semantic signature. On the one hand, the problem of lack of security protection for the information transmission process of the semantic communication system in the related art is avoided, and on the other hand, a more robust semantic communication system can be obtained.
[0102] Optionally, in another embodiment based on the above method of the present application, the semantic encryption module comprises a semantic encryption component and a semantic decryption component;
[0103] The semantic encryption module is used to encrypt the perturbed semantic information to obtain encrypted semantic information, comprising:
[0104] The semantic encryption component in the semantic encryption module is used to encrypt the perturbed semantic information to obtain the encrypted semantic information.
[0105] The semantic encryption component is constructed by a neural structure with multiple convolution layers.
[0106] Optionally, in another embodiment based on the above method of the present application, the semantic signature module is used to generate a digital signature corresponding to the original semantic information, comprising:
[0107] The semantic signature module randomly generates a plurality of vector values and inputs the plurality of vector values into a generator of the semantic communication system;
[0108] The generator converts each vector value into a semantic vector matching the original semantic information;
[0109] The target semantic vector closest to the original semantic information is selected from a plurality of semantic vectors;
[0110] The target semantic vector is used as the digital signature.
[0111] Optionally, in another embodiment based on the above method of the present application, after the target semantic vector is used as the digital signature, it further comprises:
[0112] The encrypted semantic information and the corresponding key information are sent to the receiving end at the sending end of the semantic communication system; and the target semantic vector is sent to the receiving end;
[0113] The key information is the key information generated by negotiation between the sending end and the receiving end of the semantic communication system.
[0114] Optionally, in another embodiment based on the above method of the present application, before the original semantic information to be sent is obtained, it further comprises:
[0115] The semantic encryption module, the semantic perturbation module, and the semantic signature module are respectively provided with start conditions, so that the corresponding modules are started when the semantic communication system detects that the start conditions are met.
[0116] By applying the technical solution of the present application, the privacy risk from the wireless channel can be reduced by using the perturbation mechanism, and the information transmission process of the semantic communication system is jointly secured in the way of correcting semantics at the sending end by encrypting the semantic information to be transmitted and by semantic signature. Thus, on the one hand, the problem of lack of security protection for the information transmission process of the semantic communication system in the related art is avoided, and on the other hand, a more robust semantic communication system can be obtained.
[0117] Optionally, in another embodiment of the present application, as shown in Figure 4 the present application also provides a communication device of a semantic communication system. The device is applied to a semantic communication system comprising a semantic encryption module, a semantic perturbation module and a semantic signature module, and comprises:
[0118] The obtaining module 201 is configured to obtain original semantic information to be transmitted.
[0119] The adding module 202 is configured to add a perturbation signal to the original semantic information by using the semantic perturbation module to obtain perturbed semantic information.
[0120] The encryption module 203 is configured to encrypt the perturbed semantic information by using the semantic encryption module to obtain encrypted semantic information, and to generate a digital signature corresponding to the original semantic information by using the semantic signature module.
[0121] The sending module 204 is configured to send the encrypted semantic information and the digital signature to a receiving end at a sending end of the semantic communication system.
[0122] By applying the technical solution of the present application, the privacy risk from the wireless channel can be reduced by using the perturbation mechanism, and the information transmission process of the semantic communication system is jointly secured in the way of correcting semantics at the sending end by encrypting the semantic information to be transmitted and by semantic signature. Thus, on the one hand, the problem of lack of security protection for the information transmission process of the semantic communication system in the related art is avoided, and on the other hand, a more robust semantic communication system can be obtained. Robust semantic communication system.
[0123] In another embodiment of the present application, the obtaining module 201 is configured to perform the following steps:
[0124] The encryption module 203 is configured to encrypt the perturbed semantic information by using the semantic encryption module to obtain encrypted semantic information, and to generate a digital signature corresponding to the original semantic information by using the semantic signature module.
[0125] The encryption module 203 is configured to encrypt the perturbed semantic information by using the semantic encryption module to obtain encrypted semantic information, and to generate a digital signature corresponding to the original semantic information by using the semantic signature module.
[0126] The semantic encryption component is constructed by a neural structure with multiple convolution layers.
[0127] In another embodiment of the present application, the obtaining module 201 is configured to perform the following steps:
[0128] The semantic signature module randomly generates a plurality of vector values and inputs the plurality of vector values into a generator of the semantic communication system;
[0129] The generator converts each vector value into a semantic vector matching the original semantic information respectively;
[0130] The target semantic vector most similar to the original semantic information is selected from the plurality of semantic vectors;
[0131] The target semantic vector is taken as the digital signature.
[0132] In another embodiment of the present application, the obtaining module 201 is configured to perform the following steps:
[0133] The encrypted semantic information and the corresponding key information are sent to the receiving end at the sending end of the semantic communication system; and the target semantic vector is sent to the receiving end;
[0134] The key information is key information generated by negotiation between the sending end and the receiving end of the semantic communication system.
[0135] In another embodiment of the present application, the obtaining module 201 is configured to perform the following steps:
[0136] The semantic encryption module, the semantic perturbation module and the semantic signature module are respectively provided with starting conditions, so that the corresponding modules are started when the semantic communication system detects that the starting conditions are met.
[0137] The present application also provides an electronic device for executing the communication method of the semantic communication system described above. Please refer to Figure 5 , which shows a schematic diagram of an electronic device provided by some embodiments of the present application. As Figure 5 shown, the electronic device 3 comprises a processor 300, a memory 301, a bus 302 and a communication interface 303, the processor 300, the communication interface 303 and the memory 301 are connected through the bus 302; the memory 301 stores a computer program executable on the processor 300, and the processor 300 executes the communication method of the semantic communication system provided by any of the preceding embodiments of the present application when executing the computer program.
[0138] The memory 301 can include a random access memory (RAM) and can also include a non-volatile memory, such as at least one disk memory. The communication connection between the apparatus network element and at least one other network element is realized through at least one communication interface 303 (which can be wired or wireless), and the Internet, a wide area network, a local network, a metropolitan area network, etc. can be used.
[0139] The bus 302 can be an ISA bus, a PCI bus, an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. The memory 301 is used to store programs, and the processor 300 executes the programs after receiving execution instructions. The data recognition method disclosed in any of the embodiments of the present application can be applied to the processor 300 or implemented by the processor 300.
[0140] The processor 300 can be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above method can be completed by integrated logic circuits or instructions in the form of software in the processor 300. The processor 300 described above can be a general-purpose processor, including a processor (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a ready-to-program gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. Each method, step and logic block disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, etc. The storage medium in the art. The storage medium is located in the memory 301, and the processor 300 reads the information in the memory 301 and combines the hardware to complete the steps of the above method.
[0141] The electronic device provided by the embodiments of the present application and the communication method of the semantic communication system provided by the embodiments of the present application have the same beneficial effects as the methods they use, run or implement.
[0142] The embodiments of the present application also provide a computer readable storage medium corresponding to the communication method of the semantic communication system provided by the preceding embodiments. Please refer toFigure 6 The computer-readable storage medium shown is an optical disc 40, on which a computer program (i.e., a program product) is stored. When the computer program is run by a processor, it executes the communication method of the semantic communication system provided in any of the foregoing embodiments.
[0143] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.
[0144] The computer-readable storage medium provided in the above embodiments of this application and the data identification method provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.
[0145] It should be noted that:
[0146] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this application may be practiced without these specific details. In some instances, well-known structures and techniques have not been shown in detail so as not to obscure the understanding of this specification.
[0147] Similarly, it should be understood that, for the sake of brevity and to aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of this application, various features of this application are sometimes grouped together in a single embodiment, figure, or description thereof. However, this disclosure should not be construed as reflecting a schematic diagram in which the claimed application requires more features than expressly recited in each claim. Rather, as reflected in the following claims, inventive aspects lie in fewer than all features of a single foregoing disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of this application.
[0148] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features but not others included in other embodiments, combinations of features from different embodiments are intended to be within the scope of this application and form different embodiments. For example, in the following claims, any of the claimed embodiments can be used in any combination.
[0149] The above description is only the preferred embodiment of the present application, but the protection scope of the present application is not limited to this, any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims
1. A communication method for a semantic communication system, characterized in that, The method, applied to a semantic communication system including a semantic encryption module, a semantic perturbation module, and a semantic signature module, comprises: Obtain the original semantic information to be sent; The semantic perturbation module is used to add a perturbation signal to the original semantic information to obtain perturbed semantic information; The perturbation semantic information is encrypted using the semantic encryption module to obtain encrypted semantic information; and a digital signature corresponding to the original semantic information is generated using the semantic signature module. At the sending end of the semantic communication system, the encrypted semantic information and the digital signature are sent to the receiving end; The semantic encryption module includes a semantic encryption component and a semantic decryption component; The step of using the semantic encryption module to encrypt the perturbation semantic information to obtain encrypted semantic information includes: The perturbation semantic information is encrypted using the semantic encryption component in the semantic encryption module to obtain the encrypted semantic information; wherein, the semantic encryption component is constructed from a neural structure with multiple convolutional layers; The step of generating a digital signature corresponding to the original semantic information using the semantic signature module includes: The semantic signature module randomly generates multiple vector values and inputs these multiple vector values into the generator of the semantic communication system; the generator converts each vector value into a semantic vector that matches the original semantic information; a target semantic vector that is closest to the original semantic information is selected from the multiple semantic vectors; and the target semantic vector is used as the digital signature.
2. The method as described in claim 1, characterized in that, Following the step of using the target semantic vector as the digital signature, the method further includes: At the sending end of the semantic communication system, the encrypted semantic information, the corresponding key information, and the target semantic vector are sent to the receiving end. The key information is the key information negotiated and generated by the sending end and the receiving end of the semantic communication system.
3. The method as described in claim 1, characterized in that, Before obtaining the original semantic information to be sent, the process also includes: Startup conditions are set for the semantic encryption module, the semantic perturbation module, and the semantic signature module, so that the semantic communication system starts the corresponding module when it detects that the startup conditions are met.
4. A communication device for a semantic communication system, characterized in that, The device, applied in a semantic communication system including a semantic encryption module, a semantic perturbation module, and a semantic signature module, comprises: The acquisition module is configured to acquire the raw semantic information to be sent; The addition module is configured to add a perturbation signal to the original semantic information using the semantic perturbation module to obtain perturbed semantic information; The encryption module is configured to encrypt the perturbation semantic information using the semantic encryption module to obtain encrypted semantic information; and to generate a digital signature corresponding to the original semantic information using the semantic signature module. The sending module is configured to send the encrypted semantic information and the digital signature to the receiving end at the sending end of the semantic communication system. The semantic encryption module includes a semantic encryption component and a semantic decryption component; The encryption module is further configured to perform the encryption process on the perturbed semantic information using the semantic encryption module in the following manner to obtain encrypted semantic information: The perturbation semantic information is encrypted using the semantic encryption component in the semantic encryption module to obtain the encrypted semantic information; wherein, the semantic encryption component is constructed from a neural structure with multiple convolutional layers; The encryption module is further configured to perform the generation of a digital signature corresponding to the original semantic information using the semantic signature module in the following manner: The semantic signature module randomly generates multiple vector values and inputs these multiple vector values into the generator of the semantic communication system; the generator converts each vector value into a semantic vector that matches the original semantic information; a target semantic vector that is closest to the original semantic information is selected from the multiple semantic vectors; and the target semantic vector is used as the digital signature.
5. An electronic device, characterized in that, include: Memory, used to store executable instructions; as well as, A processor for executing the executable instructions with the memory to perform the operation of the communication method of any of the semantic communication systems of claims 1-3.
6. A computer-readable storage medium for storing computer-readable instructions, characterized in that, When the instruction is executed, it performs the operation of the communication method of any of the semantic communication systems described in claims 1-3.