CN115292726A - Semantic communication method and device, electronic equipment and storage medium - Google Patents
Semantic communication method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115292726A CN115292726A CN202210665496.9A CN202210665496A CN115292726A CN 115292726 A CN115292726 A CN 115292726A CN 202210665496 A CN202210665496 A CN 202210665496A CN 115292726 A CN115292726 A CN 115292726A
- Authority
- CN
- China
- Prior art keywords
- semantic
- encrypted
- information
- decoding
- semantic information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/602—Providing cryptographic facilities or services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/0001—Systems modifying transmission characteristics according to link quality, e.g. power backoff
- H04L1/0009—Systems modifying transmission characteristics according to link quality, e.g. power backoff by adapting the channel coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/008—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols involving homomorphic encryption
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Bioethics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- Quality & Reliability (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention provides a semantic communication method, a semantic communication device, electronic equipment and a storage medium. The method comprises the following steps: obtaining encrypted data based on the received information source signal; carrying out semantic coding on the encrypted data to obtain an encrypted semantic vector; carrying out channel coding on the encrypted semantic vector to obtain encrypted semantic information; and sending the encrypted semantic information to a receiving end. The method further comprises the following steps: receiving encrypted semantic information sent by a sending end; performing channel decoding on the encrypted semantic information to obtain first encrypted semantic information; performing semantic decoding on the first dense-state semantic information by adopting an encoding and decoding network structure constructed based on a public knowledge base to obtain second dense-state semantic information; decrypting the second dense-state semantic information to obtain plaintext semantic information; and carrying out local semantic decoding on the plaintext semantic information to obtain semantic private decoding information. The semantic communication method provided by the invention reduces the potential risk of data privacy disclosure when the semantic communication is coded and decoded by a third-party platform.
Description
Technical Field
The present invention relates to the field of communications technologies, and in particular, to a semantic communication method and apparatus, an electronic device, and a storage medium.
Background
In the existing semantic communication framework, semantic extraction, semantic coding and channel coding are performed at a sending end, and the data reaches a receiving end after being transmitted through a physical channel, and channel decoding, semantic decoding and semantic reconstruction are performed at the receiving end. The current semantic communication work is dedicated to the work of semantic expression, semantic feature extraction and reconstruction, semantic information source channel coding and the like, the privacy disclosure problem related in the semantic transmission process is rarely researched, but the privacy protection in the semantic transmission process is still one of the key problems of the semantic communication.
Compared with the traditional communication system, the semantic end-to-end communication system replaces the traditional physical layer independent communication module with the combined transceiver constructed by deep learning, however, the semantic coding and decoding and the updating of the knowledge base consume a large amount of computing and storage resources, and often need to be put to a third-party platform such as an edge server for operation, and the problem that user privacy data are exposed to an untrusted third party exists in the process; secondly, in order to better encode and decode semantic information, the third party also relates to the possibility of reversely deducing the implied semantics of the message from the private knowledge base.
Disclosure of Invention
The invention provides a semantic communication method, a semantic communication device, electronic equipment and a storage medium, which are used for solving the technical problem that in the prior art, private data are leaked in a semantic transmission process.
The invention provides a semantic communication method, which is applied to a sending end and comprises the following steps:
obtaining encrypted data based on the received information source signal;
adopting a coding and decoding network structure constructed based on a public knowledge base to carry out semantic coding on the encrypted data to obtain an encrypted semantic vector;
performing channel coding on the encrypted semantic vector to obtain encrypted semantic information;
and sending the encrypted semantic information to a receiving end.
In some embodiments, said deriving encrypted data based on the received source signal comprises:
under the condition that the data volume of the information source signal exceeds a target threshold value, performing semantic extraction on the information source signal to obtain hidden layer semantics output by a middle layer of the coding and decoding network structure;
and encrypting the hidden layer semantics to obtain the encrypted data.
In some embodiments, said deriving encrypted data based on the received source signal comprises:
and under the condition that the data volume of the information source signal does not exceed a target threshold value, encrypting the information source signal to obtain the encrypted data.
In some embodiments, the source signal is encrypted in a manner that includes any one of:
differential privacy processing, data anonymization processing, homomorphic encryption and secure multi-party computation.
In some embodiments, in the case that the encryption mode is homomorphic encryption, the codec network structure is constructed by:
setting an activation function of the coding and decoding network structure as a homomorphic operation function;
and replacing the largest pooling layer in the coding and decoding network structure with an average pooling layer.
The invention provides a semantic communication method, which is applied to a receiving end and comprises the following steps:
receiving encrypted semantic information sent by a sending end;
performing channel decoding on the encrypted semantic information to obtain first confidential semantic information;
adopting an encoding and decoding network structure constructed based on a public knowledge base to carry out semantic decoding on the first dense-state semantic information to obtain second dense-state semantic information;
decrypting the second dense-state semantic information to obtain plaintext semantic information;
and performing local semantic decoding on the plaintext semantic information by adopting a decoding network structure constructed based on a private knowledge base to obtain semantic private decoding information.
The invention also provides a semantic communication device, which is applied to a sending end and comprises the following components:
the encryption module is used for obtaining encrypted data based on the received information source signal;
the first coding module is used for carrying out semantic coding on the encrypted data by adopting a coding and decoding network structure constructed based on a public knowledge base to obtain an encrypted semantic vector;
the second coding module is used for carrying out channel coding on the encrypted semantic vector to obtain encrypted semantic information;
and the sending module is used for sending the encrypted semantic information to a receiving end.
The invention also provides a semantic communication device, which is applied to a receiving end and comprises the following components:
the receiving module is used for receiving the encrypted semantic information sent by the sending end;
the first decoding module is used for carrying out channel decoding on the encrypted semantic information to obtain first confidential semantic information;
the second decoding module is used for performing semantic decoding on the first dense-state semantic information by adopting an encoding and decoding network structure constructed based on a public knowledge base to obtain second dense-state semantic information;
the decryption module is used for decrypting the second secret semantic information to obtain plaintext semantic information;
and the third decoding module is used for carrying out local semantic decoding on the plaintext semantic information by adopting a decoding network structure constructed based on the private knowledge base to obtain semantic private decoding information.
The present invention also provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the semantic communication method as described in any one of the above when executing the program.
The invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a semantic communication method as any one of the above.
The invention also provides a computer program product comprising a computer program which, when executed by a processor, implements a semantic communication method as described in any one of the above.
According to the semantic communication method, the semantic communication device, the electronic equipment and the storage medium, the encrypted data are coded and decoded, the semantics of the original data are not directly coded and decoded, the semantic information is not directly restored by combining the public knowledge base at the receiving end, on the basis, the semantic information with the local private interpretation is restored after semantic reasoning is carried out by combining the private knowledge base at the receiving end, the potential risk of data privacy leakage when the semantic communication is coded and decoded on a third-party platform is reduced, and the local private decoding can be realized on the received semantic information.
Drawings
In order to more clearly illustrate the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is an interaction diagram of a semantic communication method provided by the present invention;
FIG. 2 is a flow chart of a semantic communication method provided by the present invention;
FIG. 3 is a schematic diagram of a communication framework for applying the semantic communication method provided by the present invention;
FIG. 4 is a second schematic diagram of a communication framework applying the semantic communication method provided by the present invention;
FIG. 5 is a second flowchart of the semantic communication method provided by the present invention;
FIG. 6 is a schematic diagram of a semantic communication device provided in the present invention;
FIG. 7 is a second schematic structural diagram of a semantic communication device according to the present invention;
fig. 8 is a schematic physical structure diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is an interaction diagram of the semantic communication method provided by the present invention. Referring to fig. 1, the overall process of the semantic communication method provided by the present invention is as follows:
the sending end encrypts the received information source signal, semantically encodes the encrypted information by combining an encoder with a public knowledge base to obtain an encrypted semantic vector, and performs channel encoding on the encrypted semantic vector;
transmitting the encrypted semantic information obtained after channel coding to a receiving end through a physical transmission channel;
a decoder at a receiving end performs channel decoding on the received encrypted semantic information to obtain first encrypted semantic information; performing semantic decoding by combining a public knowledge base to obtain second dense-state semantic information; then, the second dense-state semantic information is decrypted to obtain plaintext semantic information; and performing local semantic decoding on the plaintext semantic information by combining the private knowledge base to obtain semantic private decoding information.
Fig. 2 is a schematic flow chart of a semantic communication method according to the present invention. Referring to fig. 2, the semantic communication method provided by the present invention is applied to a sending end, and the method may include: step 210, step 220, step 230 and step 240.
and step 240, sending the encrypted semantic information to a receiving end.
It should be noted that the execution subject of the semantic communication method provided by the present invention may be an electronic device, a component in an electronic device, an integrated circuit, or a chip. The electronic device may be a mobile electronic device or a non-mobile electronic device. By way of example, the mobile electronic device may be a mobile phone, a tablet computer, a notebook computer, a palm top computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook or a Personal Digital Assistant (PDA), and the like, and the non-mobile electronic device may be a server, a Network Attached Storage (NAS), a Personal Computer (PC), and the like, and the present invention is not limited in particular.
The following describes the technical solution of the present invention in detail by taking a computer as an example to execute the semantic communication method provided by the present invention.
As shown in fig. 3, the semantic communication method provided by the present invention can be implemented based on a trained semantic communication framework.
The communication frame comprises a sending end and a receiving end. The transmitting end may include a transmitting terminal and an encoder, and the receiving end may include a receiving terminal and a decoder. The transmitting end part and the receiving end part carry out data transmission through a physical transmission channel.
In step 210, the sending terminal encrypts the received source signal to obtain an encrypted source signal, i.e. encrypted data.
The source signal may include, but is not limited to, text, image, audio, video, or a combination of the above.
In actual implementation, random perturbation type encryption or cryptography-based encryption can be adopted for the data.
The data random disturbance type encryption can modify the original information by adding random noise and the like, so that the data privacy is protected, but the disturbed data still retains the distribution information of the original data.
Data based on cryptographic encryption may be computed over encrypted data without access to the original data.
In step 220, the transmitting terminal transmits the encrypted data to an encoder, which may be configured to perform semantic coding and channel coding, respectively.
And the encoder is combined with a public knowledge base in the server and an encoding and decoding network structure constructed based on the public knowledge base to carry out semantic encoding on the encrypted data in the ciphertext state to obtain an encrypted semantic vector.
An encoder is a device that compiles, converts, and/or otherwise encodes signals or data into a form of signals that may be communicated, transmitted, and stored.
The coding and decoding network structure constructed based on the public knowledge base can comprise a coding network structure corresponding to a sending end and a decoding network structure corresponding to a receiving end.
In step 230, the encoder performs channel coding on the encrypted semantic vector to obtain encrypted semantic information.
In step 240, the encoder transmits the encrypted semantic information to a decoder of the receiving end portion through a physical transmission channel.
And after the decoder receives the encrypted semantic information sent by the sending end, the decoder respectively performs channel decoding and semantic decoding on the encrypted semantic information.
In actual execution, a decoder performs channel decoding on the received encrypted semantic information to obtain first encrypted semantic information. And the decoder is combined with a public knowledge base in the server and a coding and decoding network structure constructed based on the public knowledge base to carry out semantic decoding on the first dense-state semantic information and recover the received primary dense-state semantic information, namely the second dense-state semantic information.
And transmitting the second dense-state semantic information to a receiving terminal, and decrypting the second dense-state semantic information by the receiving terminal to obtain plaintext semantic information.
And the receiving terminal performs local semantic decoding on the received plaintext semantic information by combining with a private knowledge base of the receiving terminal, and performs semantic reconstruction to finally obtain semantic private decoding information.
On the basis of a traditional semantic communication framework, the embodiment of the invention adds encryption operation to an information source signal at a sending end, adds decryption operation to the secret semantics at a receiving end and combines local semantic interpretation operation of a private knowledge base of the receiving end.
According to the semantic communication method provided by the invention, the encrypted data is coded and decoded, the semantics of the original data are not directly coded and decoded, the semantic information is not restored by directly combining the public knowledge base at the receiving end, and on the basis, the semantic information with the local private interpretation is restored after semantic reasoning is carried out by combining the private knowledge base at the receiving end, so that the potential risk of data privacy leakage when the semantic communication is coded and decoded on a third-party platform is reduced, and the local private decoding can be realized on the received semantic information.
In some embodiments, deriving the encrypted data based on the received source signal comprises:
and under the condition that the data volume of the information source signal does not exceed the target threshold value, encrypting the information source signal to obtain encrypted data.
It should be noted that the target threshold may be set according to user requirements, and after the target threshold is set, it may be determined that the current communication scenario is a small amount of data transmission under the condition that the data amount of the source signal does not exceed the target threshold.
In practical implementation, the semantic communication method provided by the embodiment of the invention can be applied to a dense semantic communication scene of a small amount of data transmission in the following communication scene;
in a scenario of a small amount of data transmission, the bandwidth occupied by the transmission of the encrypted data between the sending terminal and the server may not be considered, and the source signal may be encrypted based on the communication framework shown in fig. 2.
In some embodiments, the source signal may be encrypted using random scrambling or cryptography-based encryption of the data.
The data random disturbance type encryption can comprise a differential privacy technology or a data anonymization technology and the like. When such encrypted data is operated on in a semantic communication codec, the usability of the data is reduced by a perturbation mechanism, and the training precision of the codec is deteriorated. Therefore, the noise threshold needs to be limited to train a semantic codec satisfying such perturbation encryption.
Among them, differential privacy technology is a means in cryptography, aiming at providing a way to maximize the accuracy of data queries while minimizing the chances of identifying their records when querying from statistical databases.
Data anonymization techniques are processes that protect private or sensitive information by eliminating or encrypting identifiers that associate individuals with stored data.
Data based cryptographic encryption may include homomorphic encryption or secure multi-party computing, etc. The operation is carried out on the encrypted data in the semantic communication codec, and a nonlinear operation process exceeding the calculation capacity of the encrypted data exists, so that the correlation function needs to be approximately replaced by utilizing linear operation in the training process, and the calculation result in a secret dimensional space can be correctly decoded.
Homomorphic encryption refers to that after homomorphic encryption is carried out on original data, a ciphertext is subjected to specific operation, and a plaintext obtained after homomorphic decryption is carried out on an obtained ciphertext calculation result is equivalent to a data result obtained by directly carrying out the same calculation on the original plaintext data.
The secure multi-party computation mainly aims at the problem of how to securely compute an appointed function under the condition of no trusted third party. The multi-party safe calculation can simultaneously ensure the input privacy and the calculation correctness, ensure that the input information of each party participating in the calculation is not exposed through a mathematical theory on the premise of no trusted third party, and simultaneously can obtain an accurate calculation result.
It is understood that different encryption schemes may also adapt or change the codec network structure.
It should be noted that the encryption method for the source signal provided in this embodiment includes, but is not limited to, encryption methods such as differential privacy processing, data anonymization processing, homomorphic encryption, secure multi-party computation, and the like, and may also be other encryption methods, which may be specifically determined according to actual needs, and is not limited specifically herein.
According to the semantic communication method provided by the invention, the information source data can be effectively protected by encrypting the information source data, and the potential risk of data leakage can be reduced.
In some embodiments, obtaining encrypted data based on the received source signal comprises:
under the condition that the data volume of the information source signal exceeds a target threshold value, performing semantic extraction on the information source signal to obtain hidden layer semantics output by a middle layer of the coding and decoding network structure;
and encrypting the hidden layer semantics to obtain encrypted data.
In practical implementation, the semantic communication method provided by the embodiment of the invention can be applied to a dense semantic communication scene of mass data transmission.
It should be noted that the target threshold may be set according to user requirements, and after the target threshold is set, the current communication scenario may be determined to be large-volume data transmission under the condition that the data amount of the source signal exceeds the target threshold.
In the scenario of mass data transmission, the communication overhead of encrypted data transmission between the terminal and the server is difficult to ignore, and the semantic communication framework in this scenario is shown in fig. 4.
It can be understood that, based on the semantic communication framework shown in fig. 4, the same encryption method as that in the above embodiment may be adopted for encrypting the source signal, and details are not described here.
In some embodiments, for the semantic communication framework shown in fig. 4, different from the semantic communication framework shown in fig. 3, semantic extraction is performed on the source signal first to obtain hidden layer semantics output by the middle layer of the coding and decoding network structure;
and encrypting the hidden layer semantics to obtain encrypted data.
Under the communication framework, the semantic communication method provided by the invention can comprise the following steps:
step 1: and the transmitting terminal performs primary semantic extraction on the received information source signal based on split learning to obtain hidden layer semantics output by the middle layer of the coding and decoding network structure. And performing homomorphic encryption or encryption in other modes on the hidden layer semantics obtained by calculation of the first layers of neural networks to obtain encrypted data.
Wherein, the middle layer can be at least one layer of neural network except the input layer and the output layer in the coding and decoding network structure.
Step 2: and the sending terminal transmits the encrypted data to the encoder, and the encoder performs semantic coding of the following layers of neural networks on the encrypted data by combining with the public knowledge base to obtain an encrypted semantic vector. And the encoder performs channel coding on the encrypted semantic vector to obtain encrypted semantic information.
And 3, step 3: the encoder transmits the encrypted semantic information to the decoder through a physical transmission channel;
and 4, step 4: the decoder performs channel decoding on the received encrypted semantic information, and performs semantic decoding by combining a public knowledge base to recover the received primary secret semantic information;
and 5: transmitting the primary secret semantic information to a receiving terminal, and decrypting the primary secret semantic information by the receiving terminal to obtain plaintext semantic information;
and performing local semantic decoding on the received plaintext semantic information by combining a private knowledge base of the self, and then performing semantic reconstruction to finally obtain semantic private decoding information.
The semantic communication method provided by the invention can adopt different communication frames and semantic communication methods under the scene of encryption transmission of information sources with different data volumes, thereby improving the applicability of semantic communication.
In some embodiments, in the case that the encryption mode is homomorphic encryption, the codec network structure is constructed as follows:
setting an activation function of a coding and decoding network structure as a homomorphic operation function;
and replacing the largest pooling layer in the coding and decoding network structure with an average pooling layer.
The training process of the coding and decoding network structure provided by the invention is illustrated by taking homomorphic encryption of the source signal as an example.
Training step 1: selecting a training data set as a public knowledge base; the type of source signal in the training dataset may be determined according to different semantic communication scenarios.
Training step 2: constructing a coding and decoding network structure based on a public knowledge base and a decoding network structure based on a private knowledge base;
in actual implementation, the activation function in the codec network structure is replaced by a homomorphic operation function. The homomorphic operation function can be an approximate function formed by homomorphic addition and homomorphic multiplication, and the maximum pooling layer in the coding and decoding network structure is replaced by an average pooling layer;
initializing model parameters of a coding and decoding network structure;
a training step 3: encrypting the information source signal by using homomorphic encryption to obtain encrypted data;
and 4, training: processing the encrypted data by using an encoder to obtain an encrypted semantic vector;
and a training step 5: processing the encrypted semantic vector by using a decoder to obtain primary secret semantic information;
training step 6: decrypting the primary dense-state semantic information by using homomorphic decryption to obtain plaintext semantic information;
training step 7: performing local semantic decoding on the plaintext semantic information by using a decoding network structure based on a private knowledge base to obtain semantic private decoding information;
a training step 8: and repeating the training steps 1-7 until model parameters of the coding and decoding network structure based on the public knowledge base and the decoding network structure based on the private knowledge base converge or the maximum training times is reached.
It can be understood that the activation function and the maximum pooling layer in the existing codec network correspond to nonlinear operation, and after homomorphic encryption is performed on the source signal, in order to adapt to the linear operation which can only be performed by homomorphic encryption, the activation function is replaced by a function of an approximate polynomial, the maximum pooling layer is replaced by an average pooling layer, and the privacy protection effect of not revealing the input source signal to a third party can be achieved.
The method can greatly reduce the risk of privacy disclosure of the data transmitted by the user on the premise of increasing a small number of parameters and the processing time of the server.
In the encryption method provided in the above embodiment, the training process of the codec network structure provided in the present invention is described by taking differential privacy processing on a source signal as an example.
Training step 1: selecting a training data set as a public knowledge base; the type of source signal in the training dataset may be determined according to different semantic communication scenarios.
Training step 2: constructing an encoding and decoding network structure based on a public knowledge base and a decoding network structure based on a private knowledge base, and initializing parameters of the encoding and decoding network structure;
a training step 3: processing the information source signals in the training data set by using a differential privacy technology, and adding noise according to the sensitivity of the data set to obtain disturbed encrypted data;
training step 4: processing the encrypted data by using an encoder to obtain an encrypted semantic vector;
training step 5: processing the encrypted semantic vector by using a decoder to obtain primary secret semantic information;
training step 6: performing local semantic decoding on the plaintext semantic information by using a decoding network structure based on a private knowledge base to finally obtain semantic private decoding information;
training step 7: and repeating the training steps 1-6 until model parameters of the coding and decoding network structure based on the public knowledge base and the decoding network structure based on the private knowledge base converge or the maximum training times is reached.
The invention provides a privacy protection communication framework facing secret semantics based on semantic communication, and integrates data encryption and decryption and a semantic local decoding module combined with a private knowledge base into a traditional semantic communication framework. Based on the semantic communication framework, the privacy protection of semantic coding and decoding data on the existing third-party platform is realized, meanwhile, the local semantic decoding of the received information is realized under the condition that a receiving end private knowledge base is not exposed to a third party, compared with the original semantic communication framework, the data of a user and the privacy of the private knowledge base are effectively protected, and the framework has wide applicability under the scene of encryption transmission of information sources with different data volumes.
Fig. 5 is a second flowchart of the semantic communication method provided in the present invention. Referring to fig. 5, the semantic communication method provided by the present invention is applied to a receiving end, and the method may include: step 510, step 520, step 530, step 540 and step 550.
and 550, adopting a decoding network structure constructed based on the private knowledge base to perform local semantic decoding on the plaintext semantic information to obtain semantic private decoding information.
It should be noted that the execution subject of the semantic communication method provided by the present invention may be an electronic device, a component in an electronic device, an integrated circuit, or a chip. The electronic device may be a mobile electronic device or a non-mobile electronic device. By way of example, the mobile electronic device may be a mobile phone, a tablet computer, a notebook computer, a palm top computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook or a Personal Digital Assistant (PDA), and the like, and the non-mobile electronic device may be a server, a Network Attached Storage (NAS), a Personal Computer (PC), and the like, and the present invention is not limited in particular.
The following describes the technical solution of the present invention in detail by taking a computer as an example to execute the semantic communication method provided by the present invention.
As shown in fig. 3, the semantic communication method provided by the present invention can be implemented based on a trained semantic communication framework.
The communication framework comprises a sending end and a receiving end. The transmitting end may include a transmitting terminal and an encoder, and the receiving end may include a receiving terminal and a decoder. The transmitting end part and the receiving end part perform data transmission through a physical transmission channel.
In this embodiment, the steps before the receiving end receives the encrypted semantic information sent by the sending end are the same as those in the above embodiments, and are not described herein again.
In step 510, the decoder at the receiving end receives the encrypted semantic information sent by the sending end.
In step 520, the decoder performs channel decoding on the encrypted semantic information to obtain a first encrypted semantic information.
In step 530, the decoder performs semantic decoding on the first dense-state semantic information by using the coding and decoding network structure constructed by the public knowledge base to obtain second dense-state semantic information. And the decoder sends the second dense semantic information to the receiving terminal.
In step 540, the receiving terminal decrypts the second dense-state semantic information to obtain the plaintext semantic information. The decryption method may be determined according to the encryption method, and is not particularly limited herein.
In step 550, the receiving terminal performs local semantic decoding on the plaintext semantic information by using a decoding network structure constructed based on the private knowledge base, and then performs semantic reconstruction to finally obtain semantic private decoding information.
According to the semantic communication method provided by the invention, the encrypted data is coded and decoded, the semantics of the original data are not directly coded and decoded, the semantic information is not restored by directly combining the public knowledge base at the receiving end, and on the basis, the semantic information with the local private interpretation is restored after semantic reasoning is carried out by combining the private knowledge base at the receiving end, so that the potential risk of data privacy leakage when the semantic communication is coded and decoded on a third-party platform is reduced, and the local private decoding can be realized on the received semantic information.
The semantic communication device provided by the invention is described below, and the semantic communication device described below and the semantic communication method described above can be referred to correspondingly.
Fig. 6 is a schematic structural diagram of a semantic communication device according to the present invention. Referring to fig. 6, the semantic communication apparatus provided by the present invention is applied to a sending end, and the apparatus may include: an encryption module 610, a first encoding module 620, a second encoding module 630, and a transmission module 640.
An encryption module 610, configured to obtain encrypted data based on the received source signal;
the first encoding module 620 is configured to perform semantic encoding on the encrypted data by using an encoding and decoding network structure constructed based on a public knowledge base to obtain an encrypted semantic vector;
a second encoding module 630, configured to perform channel encoding on the encrypted semantic vector to obtain encrypted semantic information;
a sending module 640, configured to send the encrypted semantic information to a receiving end.
The semantic communication device provided by the invention encodes and decodes the encrypted data, does not directly encode and decode the semantics of the original data, and recovers the semantic information by directly combining the public knowledge base at the receiving end, but recovers the semantic information with the local private interpretation after semantic reasoning by combining the private knowledge base of the receiving end on the basis, thereby reducing the potential risk of data privacy leakage when encoding and decoding are carried out on the semantic communication at a third-party platform, and realizing local private decoding on the received semantic information.
In some embodiments, the encryption module 610 is further configured to:
under the condition that the data volume of the information source signal exceeds a target threshold value, performing semantic extraction on the information source signal to obtain hidden layer semantics output by a middle layer of the coding and decoding network structure;
and encrypting the hidden layer semantics to obtain the encrypted data.
In some embodiments, the encryption module 610 is further configured to:
and under the condition that the data volume of the information source signal does not exceed a target threshold value, encrypting the information source signal to obtain the encrypted data.
In some embodiments, the source signal is encrypted in a manner that includes any one of:
differential privacy processing, data anonymization processing, homomorphic encryption and secure multiparty computation.
In some embodiments, in the case that the encryption mode is homomorphic encryption, the codec network structure is constructed by:
setting an activation function of the coding and decoding network structure as a homomorphic operation function;
and replacing the largest pooling layer in the coding and decoding network structure with an average pooling layer.
Fig. 7 is a second schematic structural diagram of the semantic communication device according to the present invention. Referring to fig. 7, the semantic communication apparatus provided by the present invention is applied to a receiving end, and the method may include: a receiving module 710, a first decoding module 720, a second decoding module 730, a decryption module 740, and a third decoding module 750.
A receiving module 710, configured to receive encrypted semantic information sent by a sending end;
a first decoding module 720, configured to perform channel decoding on the encrypted semantic information to obtain first dense-state semantic information;
the second decoding module 730 is configured to perform semantic decoding on the first dense-state semantic information by using an encoding and decoding network structure constructed based on a public knowledge base to obtain second dense-state semantic information;
the decryption module 740 is configured to decrypt the second dense-state semantic information to obtain plaintext semantic information;
the third decoding module 750 is configured to perform local semantic decoding on the plaintext semantic information by using a decoding network structure constructed based on the private knowledge base, so as to obtain semantic private decoding information.
The semantic communication device provided by the invention encodes and decodes the encrypted data, does not directly encode and decode the semantics of the original data any more, and recovers the semantic information by combining the public knowledge base at the receiving end, but recovers the semantic information with local private interpretation after semantic reasoning by combining the private knowledge base of the receiving end on the basis, thereby reducing the potential risk of data privacy disclosure when the semantic communication is encoded and decoded by a third-party platform, and realizing local private decoding on the received semantic information.
Fig. 8 illustrates a physical structure diagram of an electronic device, and as shown in fig. 8, the electronic device may include: a processor (processor) 810, a communication Interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication Interface 820 and the memory 830 communicate with each other via the communication bus 840. The processor 810 may call logic instructions in the memory 830 to perform a semantic communication method comprising:
obtaining encrypted data based on the received information source signal;
adopting a coding and decoding network structure constructed based on a public knowledge base to carry out semantic coding on the encrypted data to obtain an encrypted semantic vector;
performing channel coding on the encrypted semantic vector to obtain encrypted semantic information;
and sending the encrypted semantic information to a receiving end.
Or the method comprises:
receiving encrypted semantic information sent by a sending end;
performing channel decoding on the encrypted semantic information to obtain first secret semantic information;
adopting a coding and decoding network structure constructed based on a public knowledge base to carry out semantic decoding on the first dense-state semantic information to obtain second dense-state semantic information;
decrypting the second dense-state semantic information to obtain plaintext semantic information;
and performing local semantic decoding on the plaintext semantic information by adopting a decoding network structure constructed based on a private knowledge base to obtain semantic private decoding information.
In addition, the logic instructions in the memory 830 may be implemented in software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, the computer program product comprising a computer program, the computer program being storable on a non-transitory computer-readable storage medium, the computer program, when executed by a processor, being capable of executing the semantic communication method provided by the above methods, the method comprising:
obtaining encrypted data based on the received information source signal;
adopting a coding and decoding network structure constructed based on a public knowledge base to carry out semantic coding on the encrypted data to obtain an encrypted semantic vector;
performing channel coding on the encrypted semantic vector to obtain encrypted semantic information;
and sending the encrypted semantic information to a receiving end.
Or the method comprises:
receiving encrypted semantic information sent by a sending end;
performing channel decoding on the encrypted semantic information to obtain first confidential semantic information;
adopting an encoding and decoding network structure constructed based on a public knowledge base to carry out semantic decoding on the first dense-state semantic information to obtain second dense-state semantic information;
decrypting the second secret semantic information to obtain plaintext semantic information;
and performing local semantic decoding on the plaintext semantic information by adopting a decoding network structure constructed based on a private knowledge base to obtain semantic private decoding information.
In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements a semantic communication method provided by the above methods, the method comprising:
obtaining encrypted data based on the received information source signal;
adopting a coding and decoding network structure constructed based on a public knowledge base to carry out semantic coding on the encrypted data to obtain an encrypted semantic vector;
performing channel coding on the encrypted semantic vector to obtain encrypted semantic information;
and sending the encrypted semantic information to a receiving end.
Or the method comprises:
receiving encrypted semantic information sent by a sending end;
performing channel decoding on the encrypted semantic information to obtain first confidential semantic information;
adopting an encoding and decoding network structure constructed based on a public knowledge base to carry out semantic decoding on the first dense-state semantic information to obtain second dense-state semantic information;
decrypting the second secret semantic information to obtain plaintext semantic information;
and performing local semantic decoding on the plaintext semantic information by adopting a decoding network structure constructed based on a private knowledge base to obtain semantic private decoding information.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (11)
1. A semantic communication method is applied to a sending end and comprises the following steps:
obtaining encrypted data based on the received information source signal;
adopting a coding and decoding network structure constructed based on a public knowledge base to carry out semantic coding on the encrypted data to obtain an encrypted semantic vector;
performing channel coding on the encrypted semantic vector to obtain encrypted semantic information;
and sending the encrypted semantic information to a receiving end.
2. The semantic communication method according to claim 1, wherein the obtaining of the encrypted data based on the received source signal comprises:
under the condition that the data volume of the information source signal exceeds a target threshold value, performing semantic extraction on the information source signal to obtain hidden layer semantics output by an intermediate layer of the coding and decoding network structure;
and encrypting the hidden layer semantics to obtain the encrypted data.
3. The semantic communication method according to claim 1, wherein the obtaining of the encrypted data based on the received source signal comprises:
and under the condition that the data volume of the information source signal does not exceed a target threshold value, encrypting the information source signal to obtain the encrypted data.
4. The semantic communication method according to claim 2 or 3, characterized in that the source signal is encrypted according to any one of the following ways:
differential privacy processing, data anonymization processing, homomorphic encryption and secure multi-party computation.
5. The semantic communication method according to claim 4, wherein, in the case that the encryption mode is homomorphic encryption, the codec network structure is constructed by:
setting an activation function of the coding and decoding network structure as a homomorphic operation function;
and replacing the largest pooling layer in the coding and decoding network structure with an average pooling layer.
6. A semantic communication method is applied to a receiving end and comprises the following steps:
receiving encrypted semantic information sent by a sending end;
performing channel decoding on the encrypted semantic information to obtain first confidential semantic information;
adopting an encoding and decoding network structure constructed based on a public knowledge base to carry out semantic decoding on the first dense-state semantic information to obtain second dense-state semantic information;
decrypting the second dense-state semantic information to obtain plaintext semantic information;
and performing local semantic decoding on the plaintext semantic information by adopting a decoding network structure constructed based on a private knowledge base to obtain semantic private decoding information.
7. A semantic communication apparatus, applied to a sending end, comprising:
the encryption module is used for obtaining encrypted data based on the received information source signal;
the first coding module is used for carrying out semantic coding on the encrypted data by adopting a coding and decoding network structure constructed based on a public knowledge base to obtain an encrypted semantic vector;
the second coding module is used for carrying out channel coding on the encrypted semantic vector to obtain encrypted semantic information;
and the sending module is used for sending the encrypted semantic information to a receiving end.
8. A semantic communication apparatus, applied to a receiving end, comprising:
the receiving module is used for receiving the encrypted semantic information sent by the sending end;
the first decoding module is used for carrying out channel decoding on the encrypted semantic information to obtain first confidential semantic information;
the second decoding module is used for performing semantic decoding on the first dense-state semantic information by adopting an encoding and decoding network structure constructed based on a public knowledge base to obtain second dense-state semantic information;
the decryption module is used for decrypting the second secret semantic information to obtain plaintext semantic information;
and the third decoding module is used for carrying out local semantic decoding on the plaintext semantic information by adopting a decoding network structure constructed based on the private knowledge base to obtain semantic private decoding information.
9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the semantic communication method according to any one of claims 1 to 6 when executing the program.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the semantic communication method according to any one of claims 1 to 6.
11. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the semantic communication method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210665496.9A CN115292726A (en) | 2022-06-13 | 2022-06-13 | Semantic communication method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210665496.9A CN115292726A (en) | 2022-06-13 | 2022-06-13 | Semantic communication method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115292726A true CN115292726A (en) | 2022-11-04 |
Family
ID=83820527
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210665496.9A Pending CN115292726A (en) | 2022-06-13 | 2022-06-13 | Semantic communication method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115292726A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023236986A1 (en) * | 2022-06-10 | 2023-12-14 | 华为技术有限公司 | Communication method and communication apparatus |
-
2022
- 2022-06-13 CN CN202210665496.9A patent/CN115292726A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023236986A1 (en) * | 2022-06-10 | 2023-12-14 | 华为技术有限公司 | Communication method and communication apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9158925B2 (en) | Server-aided private set intersection (PSI) with data transfer | |
| US8687800B2 (en) | Encryption method for message authentication | |
| CN112182649A (en) | Data privacy protection system based on safe two-party calculation linear regression algorithm | |
| CN111510281B (en) | Homomorphic encryption method and device | |
| Peng | Danger of using fully homomorphic encryption: A look at Microsoft SEAL | |
| CN111143862B (en) | Data processing method, query method, device, electronic equipment and system | |
| CN111555880B (en) | Data collision method and device, storage medium and electronic equipment | |
| Gao et al. | A novel quantum image encryption technique based on improved controlled alternated quantum walks and hyperchaotic system | |
| US9002000B2 (en) | Method for conversion of a first encryption into a second encryption | |
| CN116561787A (en) | Training method and device for visual image classification model and electronic equipment | |
| Cheng et al. | SecureAD: A secure video anomaly detection framework on convolutional neural network in edge computing environment | |
| CN115292726A (en) | Semantic communication method and device, electronic equipment and storage medium | |
| CN114726512A (en) | Data processing method and device | |
| CN111859440B (en) | Sample classification method of distributed privacy protection logistic regression model based on mixed protocol | |
| Li et al. | A secure and efficient entropy coding based on arithmetic coding | |
| Cheng et al. | A High‐Security Privacy Image Encryption Algorithm Based on Chaos and Double Encryption Strategy | |
| CN114726580B (en) | Data processing method and device | |
| Awadh et al. | A multilayer model to enhance data security in cloud computing | |
| Peng et al. | On the security of fully homomorphic encryption for data privacy in Internet of Things | |
| CN115085897A (en) | Data processing method and device for protecting privacy and computer equipment | |
| Yassin et al. | A novel image encryption scheme based on DCT transform and DNA sequence | |
| Taka | Secure Communication by combined Diffe-Hellman key exchange Based AES Encryption and Arabic Text Steganography. | |
| CN108075889B (en) | Data transmission method and system for reducing complexity of encryption and decryption operation time | |
| CN112583823B (en) | Encoding encryption and decoding decryption processing method | |
| Jaber et al. | Application of image encryption based improved chaotic sequence complexity algorithm in the area of ubiquitous wireless technologies |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |