Internet of Things information service method and system based on artificial intelligence

Abstract

The invention relates to the technical field of Internet of Things and information, in particular to an Internet of Things information service method and system based on artificial intelligence, and the method comprises the steps: collecting monitoring data of an Internet of Things sensing layer; before monitoring data is transmitted to an application layer of the Internet of Things, the monitoring data is used as plaintext data to be encrypted, and initial vectors during encryption and decryption of the plaintext data in different rounds are adaptively adjusted by analyzing features of the plaintext data; performing encryption transmission on each round of plaintext data in the Internet of Things according to the initial vector during each round of plaintext data encryption and decryption; in the encryption transmission process, linear regression analysis is carried out according to the similarity between any round of key stream and all previous historical rounds of key streams, and whether an encryption party updates an original initial vector randomly generated for the first time or not before each round of encryption and decryption is re-judged. The invention aims to ensure the security and high efficiency of the Internet of Things information service and improve the quality of the Internet of Things information service.

Description

Technical Field

[0001] The present application relates to the field of Internet of Things and information technology, and specifically to an Internet of Things information service method and system based on artificial intelligence. Background Art

[0002] With the rapid development of information technology, the Internet of Things (IoT) has become one of the key forces driving global digital transformation. By connecting devices, sensors and systems in the physical world to the Internet, the IoT makes data collection, transmission and processing more convenient and efficient. From smart homes to smart cities, from industrial automation to agricultural monitoring, the application of the IoT covers almost all industry fields. Therefore, the IoT contains a large amount of privacy information. How to ensure the security of privacy data in the IoT has become a top priority for IoT information services today. In this case, cryptographic algorithms have become one of the effective measures to protect the confidentiality, integrity, permission control and authentication of IoT device data.

[0003] Serial ciphers have good application prospects due to their high security. For example, Zu Chongzhi (ZUC) cipher, as one of the commercial serial ciphers, is more suitable for application in IoT systems. However, as a serial cipher, the security of ZUC cipher lies in one-time one-key, and as a symmetric cipher, encryption and decryption parties need frequent key exchange and update, which is a huge burden for IoT systems with limited resources and affects the quality of IoT information services.

[0004] The traditional method is to use a different initial vector for each encryption and decryption to avoid key updates. Although this reduces the burden of key updates and management, and the initial vector does not require complex management, the initial vector needs to be transmitted through a secure channel each time, which is also a heavy burden on the IoT system. Summary of the invention

[0005] In order to solve the above technical problems, the present application provides an Internet of Things information service method and system based on artificial intelligence. The technical solutions adopted are as follows:

[0006] In a first aspect, an embodiment of the present application provides an Internet of Things information service method based on artificial intelligence, the method comprising the following steps:

[0007] Step 1: Collect monitoring data from the IoT perception layer;

[0008] Step 2: Before the monitoring data is transmitted to the IoT application layer, the monitoring data is encrypted as plaintext data. By analyzing the characteristics of the plaintext data, the initial vectors for encryption and decryption of plaintext data in different rounds are adaptively adjusted; specifically:

[0009] S1, calculating the plaintext similarity coefficient of the current plaintext data based on the number of character overlaps and the similarity of character sequences between the current plaintext data and all historical plaintext data;

[0010] S2, determining the number of bits for cyclic shift adjustment of the original initial vector in the next round according to the plaintext similarity coefficient of the current plaintext data; performing an XOR operation on the vector after the next round of cyclic shift adjustment and the data of the first preset number of bits in the ciphertext data obtained in the current round, to obtain the initial vector for the next round of plaintext data encryption and decryption;

[0011] Step 3: Encrypt and transmit each round of plaintext data in the Internet of Things according to the initial vector used in each round of plaintext data encryption and decryption; during the encrypted transmission process, perform linear regression analysis based on the similarity between any round of key stream and all previous historical rounds of key streams, and re-judge whether the encrypting party updates the original initial vector randomly generated for the first time before each round of encryption and decryption.

[0012] Preferably, the character overlap quantity feature is determined by an average level of the distance between the current plaintext data and a set of all types of data sequences in each historical plaintext data; the data sequence is composed of various types of monitoring data.

[0013] Preferably, the character sequence similarity feature is determined by an average level of differences between the current plaintext data and all types of data sequences in each historical plaintext data.

[0014] Preferably, the method for calculating the plaintext similarity coefficient of the current plaintext data is:

[0015] Calculate the ratio of the character overlap quantity feature and the character sequence similarity feature between the current plaintext data and each historical plaintext data;

[0016] The average level of the ratio result between the current plaintext data and all the historical plaintext data is taken as an exponential function with a natural constant as the base; and the calculation result of the exponential function is taken as the plaintext similarity coefficient of the current plaintext data.

[0017] Preferably, the method for determining the number of bits for performing cyclic shift adjustment on the initial vector randomly generated by the encryption party for the first time in the next round is:

[0018] Calculate the difference between the plaintext similarity coefficient of the current plaintext data and the number 1 as the first difference, and calculate the difference between the natural constant and the number 1 as the second difference;

[0019] Calculate the ratio of the first difference to the second difference; and use the product of the ratio and the number of bits of the original initial vector as the number of bits for cyclic shift adjustment of the original initial vector in the next round.

[0020] Preferably, the original initial vector is an initial vector randomly generated by the encryption party for the first time.

[0021] Preferably, the cyclic shift is a left cyclic shift.

[0022] Preferably, in step 3, the method for re-determining whether to update the original initial vector randomly generated by the encryption party for the first time is:

[0023] Obtain the coefficient of determination of any round obtained by linear regression analysis in step 3;

[0024] Use the determination coefficients of all rounds to predict the determination coefficient of the next round;

[0025] When the determination coefficient of the next round is greater than the preset coefficient threshold, the encryption party randomly generates an initial vector again.

[0026] Preferably, after the encryption party updates the original initial vector randomly generated for the first time, the updated initial vector is resent to the decryption party.

[0027] In the second aspect, an embodiment of the present application also provides an artificial intelligence-based Internet of Things information service system, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements the steps of any one of the above-mentioned artificial intelligence-based Internet of Things information service methods.

[0028] This application has at least the following beneficial effects:

[0029] This application first calculates the plaintext similarity coefficient based on the similarity between the plaintext data to measure the similarity of the plaintext data. This indicator, as a public information that can be obtained by both the encryption and decryption parties, is helpful for subsequent adjustment of the initial vector without interaction. Then, the initial vector in each round of encryption is adjusted using the plaintext similarity coefficient and the ciphertext of the current round. The degree of adjustment of the initial vector is controlled according to the similarity of the plaintext, which helps to ensure the unpredictability of the initial vector and improve the security of the information service of the Internet of Things system. Finally, the regularity of the key stream is analyzed by linear regression to determine whether it is necessary to update the original initial vector randomly generated by the encryption party for the first time, so as to ensure the unpredictability of the initial vector and the key stream. In this way, while ensuring the security of the information transmission of the Internet of Things, the overhead of the initial vector transmission required for each round of encryption in the ZUC encryption algorithm is reduced, which helps to improve the operating efficiency of the Internet of Things system and makes the ZUC algorithm more suitable for the Internet of Things system with limited resources; at the same time, it can also determine in real time whether the key stream is still unpredictable, and timely update the original initial vector randomly generated by the encryption party for the first time, thereby ensuring the security and efficiency of the Internet of Things information service and improving the quality of the Internet of Things information service. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 A flowchart of an artificial intelligence-based Internet of Things information service method provided in this application;

[0032] Figure 2 A flowchart of the adaptive adjustment process of the initial vector when encrypting plaintext data in different rounds provided by this application;

[0033] Figure 3 The encryption and decryption flow chart for each round provided for this application;

[0034] Figure 4 This is a block diagram of the encryption principle of the ZUC encryption algorithm provided in this application. DETAILED DESCRIPTION

[0035] In order to further explain the technical means and effects adopted by this application to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of an Internet of Things information service method and system based on artificial intelligence proposed in this application, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0036] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0037] The following is a detailed description of an artificial intelligence-based Internet of Things information service method and system provided by the present application in conjunction with the accompanying drawings.

[0038] An embodiment of the present application provides an Internet of Things information service method and system based on artificial intelligence.

[0039] Specifically, the following IoT information service method based on artificial intelligence is provided, please refer to Figure 1 , the method comprises the following steps:

[0040] Step 1: Collect monitoring data from the IoT perception layer.

[0041] The structure of the Internet of Things mainly includes the perception layer, network layer and application layer. The perception layer is responsible for connecting with the physical world and collecting various information in the physical world; the network layer transmits the information collected by the perception layer through various private networks, the Internet, wired or wireless communication networks, etc.; the application layer is the interface between the Internet of Things and users, and processes and applies the data obtained by the perception layer to specifically realize the intelligence of the entire Internet of Things system. In this process, the network layer is most vulnerable to third-party attacks, which leads to privacy data leakage.

[0042] The perception layer of the Internet of Things is composed of various sensors and perception terminals, so the corresponding monitoring data is obtained through various sensors and perception terminals in the perception layer, and the data collected by each sensor and perception terminal are respectively organized into corresponding data sequences in chronological order.

[0043] Step 2: Before the monitoring data is transmitted to the IoT application layer, the monitoring data is encrypted as plaintext data. By analyzing the characteristics of the plaintext data, the initial vectors for encryption and decryption of plaintext data in different rounds are adaptively adjusted.

[0044] After the perception layer obtains the monitoring data of the physical world, it needs to transmit the data to the application layer through the Internet for use. In this process, in order to ensure the privacy of the data, the perception layer often encrypts the data after obtaining it and transmits the ciphertext in the entire network environment.

[0045] Serial ciphers (such as ZUC ciphers) have good application prospects due to their high security. However, serial ciphers are symmetric ciphers, and their security lies in one-time one-key. Therefore, they need to change and exchange keys frequently, which increases the complexity of key management. However, due to the limited resources of IoT systems, it is difficult to bear such frequent key updates and complex key management.

[0046] In order to avoid this situation, it is usually chosen to use a different initial vector each time to increase the security of the IoT system. However, since the initial vector is needed for both encryption and decryption, it is also necessary to transmit the initial vector securely, which increases the burden on the networking system and affects the quality of IoT information services. At the same time, frequent initial vector transmission interactions will also increase the risk of being obtained by the adversary.

[0047] In this application, the process flow chart of adaptive adjustment of initial vectors during encryption of plaintext data in different rounds is as shown in the attached figure. Figure 2 As shown, specifically:

[0048] S1, calculating the plaintext similarity coefficient of the current plaintext data based on the number of character overlaps and the similarity of character sequences between the current plaintext data and all historical plaintext data;

[0049] Existing cryptographic attack methods can be roughly divided into four types: ciphertext-only attack, chosen ciphertext attack, chosen plaintext attack, and known plaintext attack. All of them infer the key by the similarity of the ciphertext after the plaintext is encrypted. In other words, the higher the similarity of the plaintext, the higher the probability of the adversary's successful inference, and the greater the possibility of a successful attack in polynomial time. In order to avoid this situation, in the ZUC encryption algorithm, a different initial vector IV is often used to generate the key stream in each encryption, thereby reducing the possibility of a successful attack by the adversary. However, this inevitably requires the use of a secure channel to send the initial vector IV, which increases the burden on the IoT system.

[0050] In order to reduce the burden of secure transmission of the initial vector IV, both the encryption and decryption parties need to be able to adjust the initial vector IV for encryption and decryption by themselves. The greater the change of the initial vector IV, the greater the unpredictability, the more difficult it is to infer the key through brute force or statistical attacks (such as known plaintext attacks), and the better the confidentiality.

[0051] Correspondingly, when the plaintext data of different encryption operations are very similar in content, the ciphertext may also show similar patterns, especially when the initial vector IV and the key stream do not change sufficiently. The adversary will attack based on this pattern, and the possibility of successful attack in polynomial time will be greater. Therefore, the similarity of plaintext data is positively correlated with the complexity of the change of the initial vector IV. The adjustment complexity of the initial vector IV can be adaptively set according to the similarity of the plaintext data, so that the randomness of the initial vector IV is increased when the plaintext has high similarity, reducing the possibility of attackers cracking the key through comparative analysis.

[0052] When conducting a password attack, the higher the frequency of the same characters in the plaintext, the more advantageous it is for the adversary; similarly, when the order of characters in the plaintext is similar, it also helps the adversary to infer, making the attack more likely to succeed. Therefore, plaintext similarity is mainly reflected in the frequency of the same plaintext characters and the similarity of the character order. The similarity of the frequency of plaintext characters can be measured by the number of overlapping characters between plaintext data, and the similarity of character order is the degree of similarity between different plaintext sequences.

[0053] For each type of data sequence in each plaintext data, a set of each type of data is obtained. Based on the above analysis, the plaintext similarity coefficient of the current plaintext data is calculated to measure the similarity of the plaintext data.

[0054] Where A is the plaintext similarity coefficient of the current plaintext data, N is the number of historical plaintext data, and J n is the number of characters that overlap between the current plaintext data and the nth historical plaintext data, En It is the character sequence similarity feature between the current plaintext data and the nth historical plaintext data. The exponential function is used to expand the plaintext similarity coefficient, thereby enhancing the distinction between different similarities.

[0055] As an implementation method of this embodiment, Where I is the number of different types of data collected by the IoT perception layer, Jc ni It is the Jaccard distance between the current plaintext data and the set of data sequences of the i-th type in the n-th historical plaintext data; In the formula, ed ni It is the edit distance between the current plaintext data and the data sequence of the ith type in the nth historical plaintext data. Among them, Jaccard distance and edit distance can be used to measure the distance between sets and the difference between sequences, and they are all well-known technologies and will not be described in detail.

[0056] It can be understood that when the similarity between the plaintext data and the historical plaintext data is high, there are more repeated characters and the character sequence of the plaintext data is close, that is, J n Larger and E n is smaller, so the plaintext similarity coefficient A is larger; conversely, when the similarity between the plaintext data and the historical plaintext data is smaller, the plaintext similarity coefficient A is also smaller.

[0057] S2, determining the number of bits for cyclic shift adjustment of the original initial vector in the next round according to the plaintext similarity coefficient of the current plaintext data; performing an XOR operation on the vector after the next round of cyclic shift adjustment and the data of the first preset number of bits in the ciphertext data obtained in the current round, to obtain the initial vector for the next round of plaintext data encryption and decryption.

[0058] When using the ZUC encryption algorithm to ensure the secure transmission of IoT data, during the first encrypted transmission, since the encryption and decryption parties do not have any interactive and shared content, and the ZUC algorithm is a symmetric encryption algorithm, the encryption and decryption parties must transmit the key K and the initial vector IV0 through a secure channel. In order to reduce the burden of secure transmission of the initial vector IV in the subsequent encryption and decryption process, the initial vector IV needs to be adjusted according to information known to both parties, and in order to ensure the unpredictability of the initial vector IV, this adjustment cannot be a fixed rule negotiated in advance, and still needs to have high randomness and difficulty in predicting. Therefore, the present application adjusts the initial vector IV of the current plaintext data during the encryption process using the ZUC algorithm through the plaintext similarity coefficient of the current plaintext data, and the plaintext similarity coefficient is information known to both parties.

[0059] First, during the first encryption transmission between the encryption and decryption parties, both parties use the key K and the initial vector IV0 to generate a key stream for encryption and decryption operations. Secondly, during the second round of encryption transmission between the two parties, since the decryption party only has one plaintext data, the plaintext similarity coefficient is not known to both parties at this time, and the key K and the initial vector IV0 are still needed to generate a key stream for encryption and decryption operations.

[0060] In the third and subsequent rounds of encryption, both the encryption and decryption parties can obtain the plaintext similarity coefficient, so the plaintext similarity coefficient can be used to adjust the initial vector IV. Take the third round as an example. At this time, the decryption party knows the plaintext of the first two rounds, so the plaintext similarity coefficient between the plaintexts of the first two rounds is the public information of the encryption and decryption parties. Since the initial vector IV of the ZUC cipher is 128 bits (bit), and the value range of the plaintext similarity coefficient is [1, e], in order to improve the unpredictability of the initial vector IV, the value range of the plaintext similarity coefficient is transformed to [0, 127], and then the initial vector IV0 is cyclically shifted according to the transformed plaintext similarity coefficient. At the same time, since the cyclic shift can only perform integer bit shifts, it is necessary to perform rounding operations. The shift number of the cyclic shift is calculated as follows:

[0061] Where B is the shift number of the initial vector cyclic shift in the third round, A2 is the plaintext similarity coefficient of the second round relative to the historical round (i.e., the first round), and e is a natural constant. is a floor function. A2-1 is recorded as the first difference, and e-1 is recorded as the second difference. It can be understood that when the similarity between the plaintext data obtained by the decryption party is higher, that is, the larger A2 is, the more complex and larger degree of transformation of the initial variable is required to enhance the unpredictability. At this time, a larger shift number is required to perform a cyclic shift transformation on the initial vector IV0.

[0062] Since the calculation of the shift number is rounded down, the shift number of the transformed plaintext similarity coefficient is the same when the integer bits are the same. This situation will reduce the security of the system. Therefore, the initial vector adjusted by the cyclic shift is XORed bit by bit with the first 128 bits of the ciphertext data obtained in the previous round to obtain the initial vector IV for the third round of encryption and decryption.

[0063] By analogy, the initial vector IV for each subsequent round can be obtained. The specific process is as follows: Figure 3 , Figure 4 As shown. Among them, Figure 3 The encryption and decryption flow chart for each round is as follows: Figure 4 This is the encryption principle block diagram of the ZUC encryption algorithm. Figure 3 In the figure, the solid-line frame is the encryption process, and the dotted-line frame is the decryption process.

[0064] Among them, the initial vectors of each round are:

[0065] IV1=IV2=IV0

[0066]

[0067] …

[0068]

[0069] Where, IV j is the initial vector for the jth round of encryption and decryption, IV0 is the initial vector randomly generated by the sender when the first encrypted transmission is performed, and c j-1 is the ciphertext of the j-1th round, c j-1 (∧128) is the ciphertext c j-1 The first 128 bits of B j is the number of bits adjusted by the j-th round of cyclic shift, is an XOR operation, and <<< is a circular shift operation. In this embodiment, <<< is a left circular shift operation.

[0070] Step 3: Encrypt and transmit each round of plaintext data in the Internet of Things according to the initial vector used in each round of plaintext data encryption and decryption; during the encrypted transmission process, perform linear regression analysis based on the similarity between any round of key stream and all previous historical rounds of key streams, and re-judge whether the encrypting party updates the original initial vector randomly generated for the first time before each round of encryption and decryption.

[0071] According to the method of step 2, the initial variable IV is gradually changed and the key stream of the corresponding round is obtained. However, since all key streams are generated based on the initial variable IV0 and the corresponding ciphertext, as the encryption rounds gradually increase, the probability of the initial variable IV0 being cracked will gradually increase. In order to ensure the unpredictability of the initial variable and the key stream, the initial variable needs to be updated. Since the adversary conducts a speculation attack by analyzing the key stream and the plaintext and ciphertext when inferring the key and the initial variable, that is, when the key streams become more and more similar, the possibility of the polynomial time adversary attack being successful will become greater and greater. When the encryption rounds are small, it is almost impossible for the adversary to succeed through the key stream attack. Therefore, when the rounds are greater than α rounds, it is necessary to consider updating the initial variable, that is, to perform key stream similarity and linear regression detection. In this embodiment, α is 10. The key stream similarity and linear regression detection, as well as the update of the initial variable are performed by the encryption party.

[0072] Taking the xth round as an example, the similarity between the key stream of the xth round and the key stream of each previous round is calculated. In this embodiment, the similarity is calculated using cosine similarity, with the corresponding round as the horizontal coordinate and the cosine similarity between rounds as the vertical coordinate, and a series of coordinate points can be obtained. Cosine similarity is a well-known technology and will not be described in detail. For example, the cosine similarity between the key stream of the xth round and the x-1th round is y x-1 , then the corresponding coordinate point is (x-1,y x-1 ). Take all the coordinate points obtained in the xth round as input, use linear regression, and output the coefficient of determination R of the linear regression model. x .

[0073] In this way, starting from the αth round, the determination coefficients of the linear regression model of each subsequent round are obtained in turn. Linear regression is a well-known technology and will not be described in detail. All the determination coefficients are taken as input, and linear regression is used to obtain the linear regression model of the determination coefficients in turn, and the linear regression model of the determination coefficients is used to predict the determination coefficient of the next round of encryption. When the predicted determination coefficient is greater than the coefficient threshold γ, it is considered that the current initial vector is no longer safe. The encryption party will initiate an initial vector update and send a new randomly generated initial vector to the decryption party through a secure channel. In this embodiment, the coefficient threshold γ is 0.4.

[0074] It is understandable that when the linear regression model of the determination coefficient fits better, it means that there is a certain predictable pattern in the change of the determination coefficient, which also reflects that the corresponding key stream has a certain similarity and predictability, which means that the initial vector and the key stream no longer meet the requirements of unpredictability, and there is a certain risk of data leakage, which means that the security and service quality of the Internet of Things information service are affected.

[0075] This secure transmission method for IoT information does not require the transmission of initial variables in each round of encryption. At the same time, the initial variables can be automatically adjusted according to the similarity between plaintext data. Since the length of the initial variables is limited, in order to improve security, the ciphertext and the adjusted initial variables are XORed to obtain the final initial variables. Then, linear regression is combined to predict the security of the key stream to determine whether the key stream is still unpredictable.

[0076] In this way, the overhead of initial variable transmission is reduced, while security is also guaranteed. The transformation of initial variables is linked to historical plaintext, which ensures the unpredictability of initial variables. Even if the adversary can decrypt one round through brute force, it still cannot crack subsequent rounds. At the same time, under the premise of ensuring the security of data transmission in the IoT system, the operating efficiency of the ZUC encryption algorithm in the IoT system is improved, the security and efficiency of IoT information services are guaranteed, and the quality of IoT information services is improved.

[0077] Based on the same inventive concept as the above method, an embodiment of the present application also provides an artificial intelligence-based Internet of Things information service system, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of an artificial intelligence-based Internet of Things information service method described in any one of the above methods are implemented.

[0078] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0079] It should be noted that, unless otherwise specified and limited, terms such as "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such article or device. In the absence of further restrictions, an element defined by the sentence "including one..." does not exclude the existence of other identical elements in the article or device including the element. In addition, the term "and\or" used herein includes any and all combinations of one or more related listed items.

[0080] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not invented by the present application.

[0081] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. An Internet of Things information service method based on artificial intelligence, characterized in that: The method comprises the following steps: Step 1: Collect monitoring data from the IoT perception layer; Step 2: Before the monitoring data is transmitted to the IoT application layer, the monitoring data is encrypted as plaintext data. By analyzing the characteristics of the plaintext data, the initial vectors for encryption and decryption of plaintext data in different rounds are adaptively adjusted; specifically: S1, calculating the plaintext similarity coefficient of the current plaintext data based on the number of character overlaps and the similarity of character sequences between the current plaintext data and all historical plaintext data; S2, determining the number of bits for cyclic shift adjustment of the original initial vector in the next round according to the plaintext similarity coefficient of the current plaintext data; performing an XOR operation on the vector after the next round of cyclic shift adjustment and the data of the first preset number of bits in the ciphertext data obtained in the current round, to obtain the initial vector for the next round of plaintext data encryption and decryption; Step 3: Encrypt and transmit each round of plaintext data in the Internet of Things according to the initial vector used in each round of plaintext data encryption and decryption; during the encrypted transmission process, perform linear regression analysis based on the similarity between any round of key stream and all previous historical rounds of key streams, and re-judge whether the encrypting party updates the original initial vector randomly generated for the first time before each round of encryption and decryption.

2. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 1, characterized in that: The character overlap quantity feature is determined by the average level of the distance between the current plaintext data and the set of all types of data sequences in each historical plaintext data; the data sequence is composed of various types of monitoring data.

3. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 2, characterized in that: The character sequence similarity feature is determined by an average level of differences between the current plaintext data and all types of data sequences in each historical plaintext data.

4. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 1, characterized in that: The calculation method of the plaintext similarity coefficient of the current plaintext data is: Calculate the ratio of the character overlap quantity feature and the character sequence similarity feature between the current plaintext data and each historical plaintext data; The average level of the ratio result between the current plaintext data and all the historical plaintext data is taken as an exponential function with a natural constant as the base; and the calculation result of the exponential function is taken as the plaintext similarity coefficient of the current plaintext data.

5. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 4, characterized in that: The method for determining the number of bits for cyclic shift adjustment of the initial vector randomly generated by the encryption party for the first time in the next round is: Calculate the difference between the plaintext similarity coefficient of the current plaintext data and the number 1 as the first difference, and calculate the difference between the natural constant and the number 1 as the second difference; calculating a ratio of the first difference to the second difference; The product of the ratio and the number of bits of the original initial vector is rounded down to an integer value as the number of bits for performing cyclic shift adjustment on the original initial vector in the next round.

6. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 5, characterized in that: The original initial vector is the initial vector randomly generated by the encryption party for the first time.

7. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 1, characterized in that: The cyclic shift is a left cyclic shift.

8. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 1, characterized in that: In step 3, the method for re-determining whether to update the original initial vector randomly generated by the encryption party for the first time is: Obtain the coefficient of determination of any round obtained by linear regression analysis in step 3; Use the determination coefficients of all rounds to predict the determination coefficient of the next round; When the determination coefficient of the next round is greater than the preset coefficient threshold, the encryption party randomly generates an initial vector again.

9. The method for providing information services for the Internet of Things based on artificial intelligence as claimed in claim 8, characterized in that: After the encryption party updates the original initial vector randomly generated for the first time, the updated initial vector is sent to the decryption party again.

10. An Internet of Things information service system based on artificial intelligence, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the artificial intelligence-based Internet of Things information service method as described in any one of claims 1 to 9 are implemented.

Images