4G communication module data transmission method

Abstract

The invention relates to the technical field of data transmission, and particularly discloses a 4G communication module data transmission method, which adopts a deep learning-based data processing algorithm to carry out token granularity-based semantic coding and context semantic joint sensing on to-be-transmitted data. The data type of the to-be-transmitted data is intelligently identified according to the context semantic information of the to-be-transmitted data, then a corresponding encryption algorithm is matched according to the data type of the to-be-transmitted data to perform data encryption processing on the to-be-transmitted data, and the encrypted data is transmitted to a target server through the 4G communication module to perform data decryption and storage. According to the method, the type of the to-be-transmitted data can be intelligently identified, and the optimal encryption algorithm is matched according to the data type, so that the data transmission efficiency is improved while the data security is ensured.

Description

Technical Field

[0001] The present application relates to the technical field of data transmission, and more specifically, to a 4G communication module data transmission method. Background Art

[0002] With the rapid development of mobile Internet and Internet of Things (IoT), 4G communication technology, as the key link between users and the network, has been widely used in various industries. 4G communication technology not only provides high-speed data transmission services for terminal devices such as smart phones, but also supports data interaction needs in multiple fields such as intelligent transportation, telemedicine, and industrial automation.

[0003] In the process of data transmission, data security and transmission efficiency are crucial. However, the traditional 4G data transmission method mainly relies on fixed encryption algorithms. Although it can ensure data security to a certain extent, different encryption algorithms have differences in security and performance. Therefore, how to improve data transmission efficiency while ensuring data security has become an important issue facing the current 4G communication module data transmission method. In addition, for different types of data (such as personal identity information (PI), financial data, medical records, industrial control instructions, etc.), traditional transmission methods often use unified encryption and encoding methods, which may lead to insufficient protection for some types of data and excessive protection for other types of data, thereby affecting transmission efficiency.

[0004] Therefore, an optimized 4G communication module data transmission method is expected to provide differentiated encryption strategies for different types of data to improve the security and efficiency of data transmission. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a 4G communication module data transmission method, which adopts a data processing algorithm based on deep learning to perform semantic encoding and context semantic joint perception based on token granularity on the data to be transmitted, so as to intelligently identify the data type according to the context semantic information of the data to be transmitted, and then encrypt the data to be transmitted according to the data type matching the corresponding encryption algorithm, and transmit the encrypted data to the target server through the 4G communication module for data decryption and storage. The method can intelligently identify the type of data to be transmitted, and match the optimal encryption algorithm according to the data type, thereby improving the data transmission efficiency while ensuring data security.

[0006] Accordingly, according to one aspect of the present application, a 4G communication module data transmission method is provided, which includes:

[0007] Get the data to be transmitted;

[0008] Performing data type identification on the data to be transmitted to obtain a data type identification result;

[0009] Matching an encryption algorithm based on the data type identification result;

[0010] Using the encryption algorithm to encrypt the data to be transmitted to obtain encrypted data;

[0011] The encrypted data is transmitted to the target server by using a 4G communication module;

[0012] The encrypted data is decrypted on the target server and the decrypted data is stored in a database.

[0013] Compared with the prior art, the 4G communication module data transmission method provided by the present application adopts a data processing algorithm based on deep learning to perform semantic encoding and context semantic joint perception based on token granularity on the data to be transmitted, so as to intelligently identify the data type according to the context semantic information of the data to be transmitted, and then encrypt the data to be transmitted according to the data type matching the corresponding encryption algorithm, and transmit the encrypted data to the target server through the 4G communication module for data decryption and storage. The method can intelligently identify the type of data to be transmitted, and match the optimal encryption algorithm according to the data type, thereby improving the data transmission efficiency while ensuring data security. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0015] Figure 1 4G communication module data transmission method according to an embodiment of the present application.

[0016] Figure 2 This is a flowchart of step S2 in the 4G communication module data transmission method according to an embodiment of the present application.

[0017] Figure 3 Schematic diagram of data flow in step S2 of the 4G communication module data transmission method according to an embodiment of the present application.

[0018] Figure 4 Schematic diagram of data flow in step S22 of the 4G communication module data transmission method according to an embodiment of the present application.

[0019] Figure 5 Schematic diagram of data flow in step S23 of the 4G communication module data transmission method according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0021] In response to the technical problems described in the background technology, this application proposes a 4G communication module data transmission method, which uses a data processing algorithm based on deep learning to perform token-based semantic encoding and context semantic joint perception on the data to be transmitted, so as to intelligently identify its data type according to the context semantic information of the data to be transmitted, and then encrypt the data to be transmitted according to the data type matching the corresponding encryption algorithm, and transmit the encrypted data to the target server through the 4G communication module for data decryption and storage. This method can intelligently identify the type of data to be transmitted, and match the optimal encryption algorithm according to the data type, thereby improving the data transmission efficiency while ensuring data security.

[0022] Figure 1 4G communication module data transmission method according to an embodiment of the present application. Figure 1 As shown, the 4G communication module data transmission method according to the embodiment of the present application includes the following steps: S1, obtaining data to be transmitted; S2, identifying the data type of the data to be transmitted to obtain a data type identification result; S3, matching an encryption algorithm based on the data type identification result; S4, encrypting the data to be transmitted using the encryption algorithm to obtain encrypted data; S5, transmitting the encrypted data to a target server using a 4G communication module; S6, decrypting the encrypted data on the target server and storing the decrypted data in a database.

[0023] In the above 4G communication module data transmission method, the step S1 is to obtain the data to be transmitted. Specifically, in the embodiment of the present application, the data to be transmitted includes but is not limited to text data from various sources such as smart phones, Internet of Things devices, sensor networks, etc.

[0024] Specifically, for different data sources, corresponding technical means need to be used to obtain data. For smartphones, data extraction is usually achieved with the help of various application programming interfaces (APIs). For example, the location information and route planning data generated during the operation of the map navigation application in the mobile phone can be obtained through the location service API provided by the mobile phone operating system; the chat records and file transfer data generated by social applications can be collected based on the API open to the application itself.

[0025] The data acquisition of IoT devices depends on specific communication protocols and interfaces. Taking the smart home system as an example, smart cameras, smart door locks, temperature and humidity sensors and other devices, each of which follows communication protocols such as Zigbee, Bluetooth, and Wi-Fi to interact with the gateway. When acquiring data from these devices, it is necessary to deploy the corresponding driver or software module on the gateway to parse the data format sent by the device and convert the raw data such as video images, switch status, and environmental parameters collected by the device into a processable form, thereby completing data acquisition.

[0026] The data acquisition process of sensor networks is more complicated, involving the deployment of sensor nodes, data aggregation and transmission. Sensor nodes are widely distributed in the monitoring area, continuously collecting environmental data such as temperature, humidity, light intensity, soil pH, etc. These nodes transmit the collected data to the aggregation node through a self-organizing network. The aggregation node is responsible for collecting data from multiple surrounding sensor nodes, and then transmits the aggregated data to the data acquisition terminal through wireless or wired communication methods, so as to obtain the data to be transmitted from the sensor network.

[0027] In the actual process of data collection, it is necessary to ensure the consistency and accuracy of the data. Since raw data often contains noise, redundancy or errors, appropriate preprocessing measures can be introduced. Preprocessing can cover multiple levels of operations, such as data cleaning (removing invalid or incorrect entries), format conversion (unifying data from different sources into the same structure), and preliminary verification (checking whether the data conforms to the expected pattern).

[0028] It is worth noting that in this application, all actions to obtain data are carried out in compliance with the relevant data protection laws and policies of the country where the data is located, and with the authorization given by the owner of the corresponding device.

[0029] In the above-mentioned 4G communication module data transmission method, the step S2, the data type of the data to be transmitted is identified to obtain a data type identification result. It should be understood that, considering that different types of data have different security requirements, for example, for highly sensitive information such as personal privacy and financial transaction records, it is necessary to select a high-intensity encryption algorithm for encryption to protect its content from being stolen and tampered with; and for some non-sensitive data, such as public news information, ordinary system logs, etc., the security level requirements are relatively low. In this case, a simpler encryption algorithm can be selected to improve data processing efficiency and reduce encryption costs. Based on this, the present application further identifies the data type of the data to be transmitted and determines the data type to which it belongs, so as to match the appropriate encryption algorithm according to its data type.

[0030] Figure 2 This is a flowchart of step S2 in the 4G communication module data transmission method according to an embodiment of the present application. Figure 3 FIG. 4 is a schematic diagram of data flow in step S2 of the 4G communication module data transmission method according to an embodiment of the present application. Figure 2 and Figure 3 As shown, the step S2 includes: S21, performing Token segmentation on the data to be transmitted to obtain a sequence of data tokens to be transmitted; S22, performing semantic-position joint embedding coding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of semantic-position joint embedding coding vectors of data tokens to be transmitted; S23, performing context-significant semantic joint coding on the sequence of semantic-position joint embedding coding vectors of data tokens to be transmitted to obtain a context-significant semantic coding vector of data to be transmitted; S24, determining the type recognition result based on the context-significant semantic coding vector of the data to be transmitted.

[0031] Specifically, in step S21, the data to be transmitted is tokenized to obtain a sequence of data tokens to be transmitted. Since text data is usually a continuous string of characters, directly processing long sequence data not only requires a large amount of computation, but also makes it difficult to capture local detailed semantic information in the data. Therefore, the present application tokenizes the data to be transmitted to divide it into smaller semantic units to form a sequence of data tokens to be transmitted, thereby being able to more clearly reflect the semantic structure of the text and facilitate a more fine-grained semantic analysis of each Token unit.

[0032] Specifically, the step S22 performs semantic-position joint embedding coding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of semantic-position joint embedding coding vectors of the data token to be transmitted. Figure 4FIG. 2 is a schematic diagram of data flow in step S22 of the 4G communication module data transmission method according to an embodiment of the present application. Figure 4 As shown, the step S22 includes: S221, performing semantic embedding coding on each data token to be transmitted in the sequence of data token to be transmitted to obtain a sequence of data token to be transmitted semantic embedding coding vectors; S222, performing position embedding coding on each data token to be transmitted in the sequence of data token to be transmitted to obtain a sequence of data token to be transmitted position embedding coding vectors; S223, concatenating each group of corresponding data token to be transmitted semantic embedding coding vectors and data token to be transmitted position embedding coding vectors in the sequence of data token to be transmitted semantic embedding coding vectors and the sequence of data token to be transmitted position embedding coding vectors to obtain a sequence of data token to be transmitted semantic-position joint embedding coding vectors.

[0033] That is, in order to capture the semantic information of each data token to be transmitted, the present application further uses semantic embedding coding technology to process each data token to be transmitted. It should be understood that semantic embedding coding is based on the idea of ​​distributed representation, that is, by learning the co-occurrence relationship of tokens in a large amount of text data, each token is mapped to a high-dimensional vector space, represented as an embedding vector containing its semantic information, and the semantically similar tokens are closer in the vector space, while the semantically dissimilar tokens are farther away, thereby achieving effective extraction of the semantic information of each data token to be transmitted. In a specific example of the present application, the step S221 uses the Word2Vec model to perform semantic embedding coding on each data token to be transmitted in the sequence of the data token to be transmitted to obtain a sequence of semantic embedding coding vectors of the data token to be transmitted. Those of ordinary skill in the art should know that the Word2Vec model learns the vector representation of each word by pre-training on a large-scale text corpus, so that the obtained semantic embedding coding vector of the data token to be transmitted can accurately reflect the semantic features of each token, thereby providing an accurate information basis for subsequent data type identification.

[0034] Furthermore, considering that in natural language, the order of words is crucial for semantic expression. The same words arranged in different orders will form completely different sentences and express different meanings. However, word embedding technology can only generate fixed vector representations for each word in isolation, and cannot consider the position information of the word in the sentence. Therefore, in order to ensure that the network model can understand the correct semantics of the text, the present application further performs position embedding encoding on each data token to be transmitted in the sequence of data tokens to be transmitted to capture the order information of each data token to be transmitted in the original text, so as to obtain a sequence of position embedding encoding vectors of the data token to be transmitted. In an embodiment of the present application, sine-cosine position encoding technology is used to perform position embedding encoding of each data token to be transmitted. Sine-cosine position encoding utilizes the periodicity of sine and cosine functions, and by encoding the position information of each token into sine and cosine function values ​​of different frequencies, the model can understand the relative position of the word in the sentence, so as to better understand the contextual semantic information of the text.

[0035] Then, in order to simultaneously utilize the semantic information and position information of the data token to be transmitted for contextual semantic perception, the present application further concatenates and fuses each group of corresponding semantic embedding coding vectors of the data token to be transmitted and the position embedding coding vectors of the data token to be transmitted, and generates a sequence of semantic-position joint embedding coding vectors of the data token to be transmitted, so as to more accurately understand the specific meaning of each token in the overall text sequence by integrating the semantic and position information, thereby providing a more accurate information basis for subsequent data type identification.

[0036] Specifically, the step S23 performs context-significant semantic joint encoding on the sequence of semantic-position joint embedding coding vectors of the data to be transmitted to obtain the context-significant semantic coding vector of the data to be transmitted. That is, in order to further improve the contextual understanding and processing capabilities of text data, the present application proposes a context-significant semantic joint encoding based on semantic jump degree to capture the contextual dependencies between each token in the data to be transmitted, and dynamically adjust the weight of information transfer according to the semantic jump degree (i.e., the closeness of the semantic association between adjacent position tokens), thereby highlighting the key semantic information in the text and enabling the model to more accurately understand the overall semantic structure of the text.

[0037] Figure 5 FIG. 2 is a schematic diagram of data flow in step S23 of the 4G communication module data transmission method according to an embodiment of the present application. Figure 5As shown, the step S23 includes: S231, performing a semantic transfer significance measurement based on contextual semantic change perception on each data token semantic-position joint embedded coding vector in the sequence of data token semantic-position joint embedded coding vectors to be transmitted to obtain a sequence of data token semantic-position joint feature transfer significance factors to be transmitted; S232, based on the sequence of data token semantic-position joint feature transfer significance factors to be transmitted, performing feature modulation significance aggregation coding on the sequence of data token semantic-position joint embedded coding vectors to be transmitted to obtain the contextual significant semantic coding vector of the data to be transmitted.

[0038] In a specific example of the present application, the step S231 includes: first, calculating the semantic jump degree of each data token semantic-position joint embedding coding vector to be transmitted in the sequence of the data token semantic-position joint embedding coding vector to be transmitted to obtain a sequence of data token semantic-position joint feature jump degrees to be transmitted, which is expressed by the formula:

[0039] J={v1,v2,...,v i ,...,v n}

[0040]

[0041] Where J represents the sequence of the semantic-position joint embedding encoding vectors of the data token to be transmitted, v1, v2, v i and v n represent the first, second, i-th and n-th data token semantic-position joint embedding coding vectors to be transmitted in the sequence of the data token semantic-position joint embedding coding vectors to be transmitted respectively, and n is the number of vectors in the sequence of the data token semantic-position joint embedding coding vectors to be transmitted, represents the characteristic value of the jth position in the semantic-position joint embedding coding vector of the i-th data token to be transmitted, L is the characteristic scale value of the semantic-position joint embedding coding vector of the i-th data token to be transmitted, u i is the semantic strength factor of the semantic-position joint embedding encoding vector of the i-th data token to be transmitted, u i+1 represents the semantic strength factor of the semantic-position joint embedding encoding vector of the i+1th data token to be transmitted, t i Represents the semantic jump degree of the semantic-position joint embedding coding vector of the i-th data token to be transmitted.

[0042] Here, in order to capture the semantic fluctuations between each data token to be transmitted, the present application first measures the drastic degree of semantic change relative to the adjacent data token by calculating the semantic jump degree of the semantic-position joint embedding coding vector of each data token to be transmitted. It should be understood that in the process of contextual transfer encoding of semantic information of each data token to be transmitted, the semantic-position joint embedding coding vector of each data token to be transmitted performs semantic interaction in sequence according to the order of their positions in the original text. If the semantic jump degree of a data token to be transmitted is large relative to the next data token to be transmitted, that is, the semantic change is more drastic, it means that the token carries a more important semantic transition or information change in the text. Therefore, in the subsequent semantic encoding process, it should be given higher attention and weight.

[0043] Next, the semantic transfer space span of each data token semantic-position joint embedding coding vector to be transmitted in the sequence of the data token semantic-position joint embedding coding vector to be transmitted is calculated to obtain a sequence of data token semantic-position joint feature transfer space spans to be transmitted, which is expressed as follows:

[0044] s i =Count(v i →v n )

[0045] Among them, s i represents the semantic transfer space span of the semantic-position joint embedding encoding vector of the i-th data token to be transmitted, Count(v i →v n ) represents the v i and the v n The number of eigenvectors between them.

[0046] It should be understood that the spatial span of semantic transmission of each data token to be transmitted will also have an impact on its semantic importance during the semantic interactive transmission process. That is, under normal circumstances, the influence of semantic association between two tokens that are far apart is relatively weak. Therefore, while calculating the semantic jump degree, the present application further calculates the spatial span of semantic transmission of each data token to be transmitted. Here, it is considered that the token at the end of the text often carries more important global or summary information. Therefore, the present application takes the last token of the text sequence as the benchmark, and further modulates the semantic importance of each token in the text semantic transmission process by calculating the spatial span of semantic transmission of other tokens relative to the benchmark token.

[0047] Then, based on the semantic jump degree and semantic transfer space span of each data token semantic-position joint embedding coding vector to be transmitted in the sequence of the data token semantic-position joint embedding coding vector to be transmitted, the semantic transfer significance factor of each data token semantic-position joint embedding coding vector to be transmitted is calculated to obtain the sequence of the data token semantic-position joint feature transfer significance factor to be transmitted, which is expressed by the formula:

[0048]

[0049] Among them, α and β are preset weight parameters, which are used to balance the influence of semantic jump degree and semantic transfer space span. i Represents the semantic transfer significance factor of the semantic-position joint embedding coding vector of the i-th data token to be transmitted.

[0050] That is, in order to comprehensively consider the two important factors of semantic jump degree and semantic transfer space span, so as to more accurately measure the importance of the semantic-position joint embedding coding vector of each data token to be transmitted in the contextual semantic transfer process, the present application further calculates the semantic transfer significance factor based on the semantic jump degree and semantic transfer space span of the semantic-position joint embedding coding vector of each data token to be transmitted, so as to focus more on important data tokens and reduce attention to secondary information in the subsequent semantic transfer process, thereby enhancing the model's ability to understand the overall semantic structure of the text of the data to be transmitted.

[0051] In a specific example of the present application, the step S232 includes: inputting the sequence of semantic-position joint feature transfer significant factors of the data token to be transmitted into a gated transfer unit containing a softmax normalization function to obtain a sequence of semantic-position joint feature transfer significant weights of the data token to be transmitted; based on the sequence of semantic-position joint feature transfer significant weights of the data token to be transmitted, weighted aggregation is performed on the sequence of semantic-position joint embedded coding vectors of the data token to be transmitted to obtain the contextual significant semantic coding vector of the data to be transmitted, which is expressed as follows:

[0052]

[0053] Among them, softmax(·) is the normalized exponential function, mask[·] is the gated mask function, τ is the gated threshold, and w i For the v i The semantic-position joint feature of the token of the data to be transmitted transfers a significant weight, and V represents the contextual significant semantic encoding vector of the data to be transmitted.

[0054] That is, a gating mechanism is further introduced to perform gating screening on the semantic transfer significance factors of the semantic-position joint embedding coding vectors of each data token to be transmitted, and a sequence of significant weights of the semantic-position joint feature transfer of the data token to be transmitted is generated, and this is used to perform weighted aggregation on the original sequence of semantic-position joint embedding coding vectors of the data token to be transmitted, thereby generating a contextual significant semantic coding vector of the data to be transmitted containing contextual significant semantic information. In this way, by dynamically adjusting the weights of the semantic information transfer of each data token to be transmitted, the key semantic nodes in the data to be transmitted are effectively highlighted, so that the model can more accurately identify the data type on this basis.

[0055] Specifically, the step S24 determines the type recognition result based on the contextual significant semantic coding vector of the data to be transmitted. In a specific example of the present application, the step S232 includes: inputting the contextual significant semantic coding vector of the data to be transmitted into a classifier-based data type identifier to obtain the type recognition result. Specifically, the classifier-based data type identifier adopts a deep neural network architecture, and establishes a mapping relationship between data features and text data types by learning a large amount of annotated text data. In the data type recognition stage, the classifier learns the contextual semantic information of the data to be transmitted by performing multi-layer feature extraction on the contextual significant semantic coding vector of the data to be transmitted, and combines the mapping relationship learned during the training process to accurately judge and output the specific type of the data to be transmitted, such as personal identity information, financial information, news reports, academic papers, etc., thereby realizing intelligent classification processing of the data to be transmitted.

[0056] More specifically, the salient semantic coding vector of the context of the data to be transmitted is input into a data type identifier based on a classifier to obtain the type recognition result, including: using the fully connected layer of the data type identifier to fully connect the salient semantic coding vector of the context of the data to be transmitted to obtain the fully connected coding vector of the salient semantic context of the data to be transmitted; inputting the fully connected coding vector of the salient semantic context of the data to be transmitted into the Softmax classification function of the data type identifier to obtain the probability value of the salient semantic coding vector of the context of the data to be transmitted belonging to each data type label; and determining the data type label corresponding to the largest of the probability values ​​as the type recognition result.

[0057] Preferably, inputting the contextual significant semantic encoding vector of the data to be transmitted into a data type identifier based on a classifier to obtain the type identification result comprises:

[0058] Determine the label probability value p corresponding to each type recognition result label obtained by inputting the contextual significant semantic encoding vector of the data to be transmitted based on the data type identifier of the classifier j , and calculate the probability value p of each label j The square root of the sum of the squares of to get the type recognition probability value:

[0059]

[0060] Multiplying the feature mean μ of the contextual significant semantic coding vector of the data to be transmitted by the type recognition probability value to obtain a type recognition statistical field value n=μp;

[0061] Subtract one from the type identification statistical field value and divide it by the type identification statistical field value to obtain a type identification partial probability value ρ=(n-1) / n;

[0062] Calculate the power function V of each feature value in the contextual significant semantic encoding vector of the data to be transmitted, with the type recognition partial probability value as an exponent ⊙ρ and multiply it by the type identification partial probability value to obtain the type identification microscopic representation vector V1 = ρ⊙V ⊙ρ ;

[0063] After multiplying the contextual significant semantic encoding vector of the data to be transmitted by the type identification partial probability value, an exponential function with a natural constant as the base is calculated to obtain a type identification macro mapping vector V2=exp(V⊙ρ);

[0064] After calculating the base 2 logarithm of the type identification micro-representation vector, the weighted sum is performed with the type identification macro-mapping vector to obtain the optimized contextual significant semantic encoding vector of the data to be transmitted. Where V1 represents the type identification micro-representation vector, V2 represents the type identification macro-mapping vector, log2(·) represents the logarithmic function with base 2, represents point addition, ⊙ represents point multiplication, a and b are different weighting coefficients, and V' represents the optimized contextual salient semantic encoding vector of the data to be transmitted;

[0065] The optimized context-significant semantic encoding vector of the data to be transmitted is input into a classifier-based data type identifier to obtain the type identification result.

[0066] Here, since each semantic-position joint embedding coding vector of the data token to be transmitted in the sequence of semantic-position joint embedding coding vectors of the data token to be transmitted respectively represents the position and semantic joint embedding coding features of each data token to be transmitted, when performing context-significant semantic encoding, the interleaving of position and semantic encoding in the context encoding process will cause context-significant semantic encoding transmission deviation, thereby causing dynamic deviation in the mapping of features to class target probabilities when the context-significant semantic encoding vector of the data to be transmitted is input into a classifier-based data type identifier, thereby reducing the accuracy of the type recognition result obtained.

[0067] Therefore, the partial low-order derivatives of the statistical distribution field corresponding to the salient semantic coding vector of the data context to be transmitted are used as non-overlapping macro-feature representation behavior patches of the salient semantic coding vector of the data context to be transmitted, so as to strengthen the dynamic sensitivity of the long-series micro-complex information distribution of the salient semantic coding vector of the data context to be transmitted to the macro-representation behavior of class probability based on the different macro-behavior patch organization spaces under the non-isotropic backbone structure of the salient semantic coding vector of the data context to be transmitted, thereby promoting the iterative dynamic consistency of the class target between the classification target and the extracted features during the feature space-class probability mapping, so as to improve the accuracy of the type recognition result obtained by the data type identifier based on the classifier when the salient semantic coding vector of the data context to be transmitted is input.

[0068] In the above-mentioned 4G communication module data transmission method, the step S3 matches the encryption algorithm based on the data type identification result. That is, for personal identity information or financial information containing personal privacy information, a high-intensity encryption algorithm is used for encryption processing, such as RSA encryption algorithm or AES encryption algorithm, to ensure the security of these data during transmission or storage; and for data types with strong openness such as news reports or academic papers, a relatively lightweight encryption algorithm, such as Base64 encoding algorithm, can be used to facilitate data sharing and dissemination while ensuring data security.

[0069] Specifically, an encryption algorithm library can be pre-established, and after the data type is identified, the preset encryption algorithm library is automatically accessed. The encryption algorithm library contains a variety of encryption algorithms, such as RSA and AES algorithms for high-intensity encryption, Base64 encoding algorithms for lightweight encryption, etc. Each algorithm has its own characteristics in terms of security, computational complexity, encryption efficiency, etc.

[0070] For data identified as personal privacy information, such as personal identity information such as ID card number, bank card number, and financial information, high-intensity encryption algorithms are matched first. Taking the RSA algorithm as an example, it is based on the problem of large integer decomposition in number theory and has extremely high security. During the matching process, the system will determine the key length of the RSA algorithm based on the data scale and security requirements. If a large amount of highly sensitive financial data is processed, a 2048-bit or even higher key may be selected to ensure data security. Because the longer the key length, the difficulty of cracking increases exponentially, but at the same time the computational complexity will also increase. Therefore, it is also necessary to weigh computing resources and encryption time to ensure that encryption operations do not excessively affect data transmission efficiency while ensuring security.

[0071] The AES algorithm is a symmetric encryption algorithm with the advantages of fast encryption and decryption speed. When matching the AES algorithm, the real-time requirements of the data need to be considered. If you are processing financial transaction data with high real-time requirements, you can use the AES-128 or AES-256 mode. These modes can quickly complete encryption and decryption operations while ensuring security, meeting the strict requirements of transaction data for processing speed. At the same time, the system will generate a secure and reliable key for the AES algorithm and ensure the confidentiality of the key during transmission and storage to prevent key leakage from causing data to be cracked.

[0072] If the data type is highly public data such as news reports and academic papers, the system tends to match lightweight encryption algorithms, such as the Base64 encoding algorithm. Base64 encoding is mainly used to convert binary data into text format for easy transmission on the network. Although its encryption strength is relatively low, it is sufficient to meet the security requirements of such data. During the matching process, the system will optimize the Base64 encoding according to the target platform and application scenario of data transmission. If the data will be displayed on a web page, compatibility with web technologies such as HTML and CSS will be considered to ensure that the encoded data can be correctly displayed and transmitted on the web page, while avoiding data loss or errors due to encoding problems.

[0073] In the above-mentioned 4G communication module data transmission method, the step S4 uses the encryption algorithm to encrypt the data to be transmitted to obtain encrypted data. It should be understood that during the data transmission process, the data may pass through multiple network nodes and communication links, and there is a risk of being stolen, tampered with or monitored. By encrypting the data, the original data can be converted into ciphertext. Even if the data is intercepted during the transmission process, the attacker cannot obtain the real content of the data without the decryption key, thereby preventing the data from being illegally obtained and tampered with, and ensuring the confidentiality and integrity of the data.

[0074] Specifically, for symmetric encryption algorithms, a pair of identical keys needs to be created, and both the sender and the receiver use the same key for encryption and decryption operations. For asymmetric encryption algorithms, a pair of public and private keys needs to be generated, where the public key is used to encrypt data and only the corresponding private key can decrypt it. In either case, the secure storage and transmission of keys is crucial, because once the keys are leaked, the security of the entire encryption system will be seriously threatened. In order to ensure the security of the keys, a variety of technical means can be used, such as using a dedicated key management system (KMS), using a hardware security module (HSM) or software protection mechanism to ensure the security of the keys throughout their life cycle.

[0075] After the key is generated, the actual encryption process can begin. The first step is to preprocess the original data. Since most encryption algorithms can only process data blocks of fixed length, when the amount of data to be encrypted is not exactly equal to the size of these blocks, it needs to be padded to ensure that each data block can be correctly encrypted. There are many ways to fill, such as PKCS#7 filling, ANSIX.923 filling, etc. Its function is to add additional information to the last data that is less than a whole block so that its length meets the algorithm requirements. In addition, in some cases, it is also necessary to consider how to split the large data stream and break it down into smaller data blocks for encryption one by one, so as to better adapt to the network transmission characteristics.

[0076] After entering the formal encryption phase, it is necessary to call the application programming interface (API) provided by the selected encryption algorithm and pass the prepared plaintext data and the corresponding key as parameters. The encryption function will process the input data bit by bit according to specific mathematical rules. After a series of complex calculations and transformations, the final output is a ciphertext that looks random but is actually generated strictly according to the algorithm design. It is worth noting that modern encryption algorithms are often highly complex and resistant to attacks. Even if an attacker obtains part of the ciphertext fragment, it is difficult to derive any useful information from it. In order to further enhance security, you can also combine the use of message authentication code (MAC) or digital signature technology to add additional authentication information while encrypting to ensure that the received data is indeed from a trusted sender and has not been tampered with midway.

[0077] After encryption is completed, the generated ciphertext needs to be formatted to make it suitable for subsequent transmission needs. This may involve converting the binary ciphertext into a text string representation for easy transmission on the network; or adding necessary header information, checksums and other auxiliary fields to help the receiving end accurately parse the received data. At the same time, considering the bandwidth limitations and latency issues in actual applications, the encrypted data is sometimes compressed to reduce unnecessary redundancy and improve transmission efficiency.

[0078] In the above-mentioned 4G communication module data transmission method, the step S5 uses the 4G communication module to transmit the encrypted data to the target server. It should be understood that the 4G communication network has a wide coverage and a high transmission rate, which can realize the fast and stable transmission of data between different geographical locations, and provides the necessary transmission channel for subsequent data decryption and storage. Specifically, the 4G communication module follows the 4G communication protocol (such as LTE) and communicates with the base station through a radio frequency signal. First, the encrypted data is modulated in the 4G communication module to convert the digital signal into a radio frequency signal suitable for transmission in a wireless channel. Then, the radio frequency signal is sent to a nearby base station through an antenna. After the base station receives the signal, it demodulates, decodes, and other processes the signal, restores the data to a digital signal, and forwards the data to the target server through the core network.

[0079] Specifically, since the 4G network has specific requirements for the format and size of transmitted data, the encrypted data must be encapsulated in accordance with relevant specifications. This process includes adding various control information and protocol headers, which contain key content such as source address, destination address, data length, and data type. Adding source address and destination address enables data to be clearly transmitted in the network; data length information helps the receiving end determine whether the data has been received completely; data type information facilitates network equipment to perform corresponding processing during transmission.

[0080] The encapsulated data will be sent to the cache area of ​​the 4G communication module for temporary storage. The cache area plays an important role in data buffering and scheduling, coordinating the input and output rhythm of data to avoid congestion and conflict in data transmission. When there is free space in the sending queue of the 4G communication module, the data in the cache area will be taken out in sequence according to a certain scheduling strategy, ready to enter the next transmission process. The scheduling strategy usually takes into account the priority of the data. For example, for voice data with high real-time requirements or urgent control command data, the transmission will be arranged first to ensure that the data can reach the target server in time to meet the needs of actual applications.

[0081] Next, the encrypted data will be modulated in the 4G communication module. Modulation is a key step in converting digital signals into radio frequency signals suitable for transmission in wireless channels. Common modulation methods used in 4G communication modules include quadrature amplitude modulation (QAM). Taking QAM modulation as an example, it can map multiple bits of digital signals to a specific carrier amplitude and phase combination. In this way, the digital signal is loaded onto a high-frequency carrier to form a radio frequency signal that can be transmitted in a wireless channel. Different QAM modulation methods, such as 16QAM, 64QAM, 256QAM, etc., have different modulation efficiencies and anti-interference capabilities. When selecting a modulation method, the 4G communication module will dynamically adjust according to the quality of the current wireless channel. If the channel quality is good and the interference is small, the 256QAM method with high modulation efficiency will be selected to increase the data transmission rate; when the channel quality is poor and there is more interference, it will switch to the 16QAM method with stronger anti-interference ability to ensure the stability of data transmission and reduce the bit error rate.

[0082] After modulation, the RF signal will be sent out through the antenna of the 4G communication module. As a key device for wireless communication, the antenna is responsible for radiating the RF signal into the surrounding space. The antenna design of the 4G communication module needs to consider many factors, such as the gain, directivity and polarization of the antenna. Antennas with higher gain can enhance the signal transmission strength and expand the signal coverage; directional antennas can focus the signal to a specific direction, improve the signal transmission efficiency, and reduce the signal loss in other directions; the choice of polarization mode should be determined according to the actual communication environment and needs. Common polarization modes include horizontal polarization, vertical polarization and circular polarization. Different polarization modes have different transmission performance in different environments.

[0083] When the transmitted radio frequency signal propagates in the wireless channel, it will be affected by various factors, such as path loss, multipath fading and noise interference. Path loss is due to the gradual attenuation of the signal as the distance increases during the propagation process; multipath fading is because the signal will be reflected, refracted and scattered when it encounters obstacles such as buildings and terrain during the propagation process, resulting in the superposition of signals from multiple different paths at the receiving end, resulting in changes in signal strength and phase, which may cause signal distortion in severe cases; noise interference comes from various electronic devices and natural environments around, such as thermal noise and electromagnetic interference. In order to cope with these problems, 4G communication technology adopts a variety of technical means. For example, through channel coding technology, redundant information is added to the data at the sending end, and the receiving end can use this redundant information for error detection and correction to improve the reliability of data transmission; diversity technologies such as space diversity, time diversity and frequency diversity are used to increase the probability of data being correctly received by sending the same data at different spatial locations, time points or frequencies.

[0084] When the RF signal reaches the base station, the base station will perform a series of processing on the signal. The base station will first receive the signal through its own antenna, and then perform demodulation to convert the RF signal back into a digital signal. The demodulation process is the inverse process of modulation. According to the modulation method adopted by the transmitter, the received signal is demapped, filtered, and processed to restore the original digital signal. After the demodulation is completed, the base station will decode the digital signal, extract the data information, and check the integrity and correctness of the data. If an error is found in the data, the base station will perform error correction according to the channel coding rules; if the error cannot be corrected, the base station will ask the transmitter to resend the data.

[0085] The data processed by the base station will be forwarded through the core network. The core network is the backbone of the 4G communication system, responsible for connecting various base stations and accurately routing data to the target server. The core network contains a variety of network devices and protocols, such as the mobile switching center (MSC), packet data service node (PDSN) and gateway. These devices and protocols work together to select the best transmission path for the data based on the target address of the data by looking up the routing table and other methods. During the transmission process, the core network will also perform flow control and congestion management on the data to ensure the stable operation of the network and avoid network congestion causing data transmission delays or losses.

[0086] Finally, after the data reaches the target server, it will be checked for integrity and correctness again to ensure that there are no errors or losses in the data transmission process. If the data is checked correctly, the target server will store the data in the corresponding database for subsequent processing and use. If there is a problem with the data, the target server will send an error feedback message to the sender, requesting the data to be resent.

[0087] In the above-mentioned 4G communication module data transmission method, the step S6 decrypts the encrypted data on the target server and stores the decrypted data in a database. It should be understood that the target server, as the final storage and processing location of the data, needs to have corresponding decryption capabilities, that is, to master the decryption key that matches the encryption algorithm. After receiving the encrypted data, the target server first uses the decryption key to decrypt the encrypted data and restore the original data. Then, according to the data type and storage requirements, the decrypted data is stored in the corresponding database. As the core component of data storage, the database has efficient data retrieval and management capabilities, can support fast data query and analysis operations, and provides a reliable foundation for subsequent data processing and utilization.

[0088] In summary, the 4G communication module data transmission method according to the embodiment of the present application is explained, which uses a data processing algorithm based on deep learning to perform semantic encoding and context semantic joint perception based on token granularity on the data to be transmitted, so as to intelligently identify its data type according to the context semantic information of the data to be transmitted, and then encrypt the data to be transmitted according to its data type matching the corresponding encryption algorithm, and transmit the encrypted data to the target server through the 4G communication module for data decryption and storage. This method can intelligently identify the type of data to be transmitted, and match the optimal encryption algorithm according to the data type, thereby improving the data transmission efficiency while ensuring data security.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A 4G communication module data transmission method, characterized in that: include: Get the data to be transmitted; Performing data type identification on the data to be transmitted to obtain a data type identification result; Matching an encryption algorithm based on the data type identification result; Using the encryption algorithm to encrypt the data to be transmitted to obtain encrypted data; The encrypted data is transmitted to the target server by using a 4G communication module; The encrypted data is decrypted on the target server and the decrypted data is stored in a database.

2. The 4G communication module data transmission method according to claim 1, characterized in that: Performing data type identification on the data to be transmitted to obtain a data type identification result includes: Tokenize the data to be transmitted to obtain a sequence of tokens of the data to be transmitted; Performing semantic-position joint embedding coding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of semantic-position joint embedding coding vectors of the data token to be transmitted; Performing context-significant semantic joint encoding on the sequence of token semantic-position joint embedding coding vectors of the data to be transmitted to obtain a context-significant semantic coding vector of the data to be transmitted; The type recognition result is determined based on the context-significant semantic encoding vector of the data to be transmitted.

3. The 4G communication module data transmission method according to claim 2, characterized in that: Performing semantic-position joint embedding coding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of semantic-position joint embedding coding vectors of the data token to be transmitted, including: Performing semantic embedding coding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of semantic embedding coding vectors of the data token to be transmitted; Performing position embedding coding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of position embedding coding vectors of the data token to be transmitted; Each group of corresponding data token semantic embedding coding vectors to be transmitted and data token position embedding coding vectors to be transmitted in the sequence of data token semantic embedding coding vectors to be transmitted and the sequence of data token position embedding coding vectors to be transmitted are concatenated to obtain a sequence of data token semantic-position joint embedding coding vectors to be transmitted.

4. The 4G communication module data transmission method according to claim 3, characterized in that: Performing semantic embedding coding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of semantic embedding coding vectors of the data token to be transmitted, including: The Word2Vec model is used to perform semantic embedding encoding on each data token to be transmitted in the sequence of data tokens to be transmitted to obtain a sequence of semantic embedding encoding vectors of the data token to be transmitted.

5. The 4G communication module data transmission method according to claim 4, characterized in that: Performing context-significant semantic joint encoding on the sequence of the token semantic-position joint embedding coding vectors of the data to be transmitted to obtain the context-significant semantic coding vector of the data to be transmitted, including: Performing a semantic transfer significance measurement based on contextual semantic change perception on each of the data token semantic-position joint embedding coding vectors to be transmitted in the sequence of the data token semantic-position joint embedding coding vectors to be transmitted to obtain a sequence of data token semantic-position joint feature transfer significance factors to be transmitted; Based on the sequence of semantic-position joint feature transfer significance factors of the data token to be transmitted, feature modulation significant aggregation coding is performed on the sequence of semantic-position joint embedded coding vectors of the data token to be transmitted to obtain the contextual significant semantic coding vector of the data to be transmitted.

6. The 4G communication module data transmission method according to claim 5, characterized in that: The method further comprises: performing a semantic transfer saliency measurement based on contextual semantic change perception on each of the data token semantic-position joint embedding coding vectors to be transmitted in the sequence of the data token semantic-position joint embedding coding vectors to be transmitted to obtain a sequence of data token semantic-position joint feature transfer saliency factors to be transmitted, including: Calculating the semantic jump degree of each data token semantic-position joint embedding coding vector to be transmitted in the sequence of data token semantic-position joint embedding coding vectors to be transmitted to obtain a sequence of data token semantic-position joint feature jump degrees to be transmitted; Calculating the semantic transfer space span of each data token semantic-position joint embedding coding vector to be transmitted in the sequence of data token semantic-position joint embedding coding vectors to be transmitted to obtain a sequence of data token semantic-position joint feature transfer space spans to be transmitted; Based on the semantic jump degree and semantic transfer space span of each data token semantic-position joint embedding coding vector to be transmitted in the sequence of the data token semantic-position joint embedding coding vector to be transmitted, the semantic transfer significance factor of each data token semantic-position joint embedding coding vector to be transmitted is calculated to obtain the sequence of the data token semantic-position joint feature transfer significance factor to be transmitted.

7. The 4G communication module data transmission method according to claim 6, characterized in that: Based on the sequence of the semantic-position joint feature transfer significant factors of the data token to be transmitted, feature modulation significant aggregation coding is performed on the sequence of the semantic-position joint embedding coding vectors of the data token to be transmitted to obtain the contextual significant semantic coding vector of the data to be transmitted, including: Inputting the sequence of significant factors of the semantic-position joint feature transfer of the data token to be transmitted into a gated transfer unit including a softmax normalization function to obtain a sequence of significant weights of the semantic-position joint feature transfer of the data token to be transmitted; Based on the sequence of significant weights transferred by the semantic-position joint features of the data token to be transmitted, the sequence of the semantic-position joint embedded coding vectors of the data token to be transmitted is weightedly aggregated to obtain the contextual significant semantic coding vector of the data to be transmitted.

8. The 4G communication module data transmission method according to claim 7, characterized in that: Determining the type recognition result based on the contextual significant semantic coding vector of the data to be transmitted includes: The context-significant semantic encoding vector of the data to be transmitted is input into a classifier-based data type identifier to obtain the type identification result.

9. The 4G communication module data transmission method according to claim 8, characterized in that: Inputting the contextual significant semantic encoding vector of the data to be transmitted into a data type identifier based on a classifier to obtain the type identification result, including: Using the fully connected layer of the data type identifier to perform fully connected encoding on the contextually significant semantic encoding vector of the data to be transmitted to obtain the contextually significant semantic fully connected encoding vector of the data to be transmitted; Inputting the contextual significant semantic fully connected encoding vector of the data to be transmitted into the Softmax classification function of the data type identifier to obtain the probability value of the contextual significant semantic encoding vector of the data to be transmitted belonging to each data type label; The data type label corresponding to the largest probability value among the probability values ​​is determined as the type identification result.

Images