Software data flow detection method and system based on matrix multiplication

Abstract

The invention discloses a software data flow detection method and system based on matrix multiplication, and belongs to the technical field of computers, and the method comprises the following steps: S1, obtaining a detection point matrix according to the number of data flow detection points; s2, taking the sequence of the data stream detection points as a recursive multiplication sequence, and performing recursive multiplication according to the initial matrix of the first channel in the channels and the detection point matrix to obtain a second to-be-detected result matrix of a second channel in the channels, performing recursive multiplication according to the initial matrix of the second channel and the detection point matrix to obtain a first to-be-detected result matrix of the first channel; and S3, the first channel carries out security verification through the second to-be-detected result matrix and the first expected result matrix of the first channel, the second channel carries out security verification through the first to-be-detected result matrix and the second expected result matrix of the second channel, if the security verification is passed, security data are output, and otherwise, a fault security state is entered. The technical problem that data security threats are difficult to comprehensively detect in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer technology, and particularly to a software data flow detection method and system based on matrix multiplication. BACKGROUND

[0002] The security computing software data flow usually has a high logical sequence requirement, and the wrong data flow may cause a major accident. In the security computing software, the protection and verification of the software data flow usually rely on the monitoring and management of the third-party device to ensure the accuracy of the instruction execution sequence and avoid the omission of the execution of the key security detection program. The existing technology usually detects the data flow to identify the abnormal data flow, such as the attack and abnormal data flow detection method and device for information physical system in patent No. CN115834248A, which comprises: acquiring real-time data flow in the information physical system and converting the acquired data flow into a data object set; inputting the data object set after preprocessing into the trained back propagation network to obtain the data label corresponding to the data object set; determining whether the current data is attacked or the attack type according to the data label corresponding to the data object set to quickly detect possible threats. Through the above method, once the abnormal data flow is identified, the related data is not output, thereby improving the security of the data. However, the above method converts the acquired data flow into a data object set, and inputs the data object set after preprocessing into the trained back propagation network, which only considers the abnormal change of the data flow content and does not consider the abnormal change of the data flow execution sequence, so it is difficult to detect the data flow abnormality caused by the abnormal change of the data flow execution sequence, thereby affecting the comprehensive detection of the data security threat. SUMMARY

[0003] In view of the technical problem that the existing technology is difficult to comprehensively detect the data security threat, the present application provides a software data flow detection method and system based on matrix multiplication, which obtains a second to-be-detected result matrix and a first to-be-detected result matrix through the sequence of the data flow detection point position as the matrix recursive multiplication sequence, that is, through the non-commutative property of matrix multiplication, any sequence disorder will cause the difference between the to-be-detected result matrix and the expected result matrix, and each step of matrix multiplication will transmit and accumulate the features of the previous detection point to the subsequent result, so that the content abnormality of any point in the data flow will be spread to the to-be-detected result matrix through the multiplication operation, and also cause the difference between the to-be-detected result matrix and the expected result matrix. By considering the abnormal change of the data flow execution sequence and the abnormal change of the data flow content, that is, detecting the execution sequence and the content of the program data flow, the technical problem that the existing technology is difficult to comprehensively detect the data security threat is solved.

[0004] To solve the above technical problems, the application provides a software data stream detection method based on matrix multiplication, comprising the following steps: S1: obtaining a detection point matrix of a corresponding scale according to the number of data stream detection points; S2: taking the order of the data stream detection points as a recursive multiplication order, obtaining a second to-be-detected result matrix of a second channel in the channel by recursively multiplying an initial matrix of the first channel and the detection point matrix, and obtaining a first to-be-detected result matrix of the first channel by recursively multiplying an initial matrix of the second channel and the detection point matrix; S3: the first channel performs security verification by using the second to-be-detected result matrix and a first expected result matrix of the first channel, the second channel performs security verification by using the first to-be-detected result matrix and a second expected result matrix of the second channel, and if the security verification is passed, the security data is output, otherwise, a fault security state is entered.

[0005] Preferably, in S1, the matrix elements in the detection point matrix are all different prime numbers.

[0006] Preferably, before S2, the method further comprises: obtaining a first time progression matrix of the first channel according to the characteristics of the first channel and the start time of the software running period, and obtaining the initial matrix of the first channel based on the first time progression matrix; obtaining a second time progression matrix of the second channel according to the characteristics of the second channel and the start time, and obtaining the initial matrix of the second channel based on the second time progression matrix; Wherein, the first channel and the second channel are mutual security verification channels.

[0007] Preferably, S2 comprises: S21: taking the order of the data stream detection points in the first channel as a one-time recursive multiplication order, and obtaining the second to-be-detected result matrix by one-time recursively multiplying the initial matrix of the first channel and the detection point matrix; S22: taking the order of the data stream detection points in the second channel as a two-time recursive multiplication order, and obtaining the first to-be-detected result matrix by two-time recursively multiplying the initial matrix of the second channel and the detection point matrix.

[0008] Preferably, S21 comprises: S211: taking the initial matrix of the first channel as a first process matrix; S212: multiplying the first process matrix with the detection point matrix in the first channel to obtain a first matrix, judging whether the detection point matrix in the first channel is the last detection point matrix, if yes, taking the first matrix as the second to-be-detected result matrix, otherwise, taking the first matrix as the first process matrix, and re-executing S212.

[0009] Preferably, the S22 comprises: S221: taking the initial matrix of the second channel as the second process matrix; S222: multiplying the second process matrix with the detection point matrix in the second channel to obtain a second matrix, determining whether the detection point matrix in the second channel is the last detection point matrix, if yes, taking the second matrix as the first to-be-detected result matrix, otherwise, taking the second matrix as the second process matrix, and re-executing S222.

[0010] Preferably, before the S3, the method further comprises: obtaining a first expected result matrix according to the initial matrix and the detection point matrix of the first channel, and obtaining a second expected result matrix according to the initial matrix and the detection point matrix of the second channel.

[0011] Preferably, the expression for obtaining the first expected result matrix according to the initial matrix and the detection point matrix of the first channel is: ; denotes the first expected result matrix, denotes the initial matrix of the first channel, , denotes the first detection point matrix in the detection point matrix, denotes the second detection point matrix in the detection point matrix, denotes the nth detection point matrix in the detection point matrix, and n denotes the number of data stream detection point positions; The expression for obtaining the second expected result matrix according to the initial matrix and the detection point matrix of the second channel is: ; denotes the second expected result matrix, denotes the initial matrix of the second channel.

[0012] Advantages of the present scheme: The order of the data stream detection point positions is used to obtain the second to-be-detected result matrix and the first to-be-detected result matrix through matrix recursive multiplication order, that is, the non-commutative property of matrix multiplication makes any order disorder result in differences between the to-be-detected result matrix and the expected result matrix, and each step of matrix multiplication will transmit and accumulate the characteristics of the previous detection point to the subsequent result, and the content anomaly of any point position in the data stream will be spread to the to-be-detected result matrix through multiplication operation, which will also cause differences between the to-be-detected result matrix and the expected result matrix. By considering the abnormal changes of the data stream execution order and the abnormal changes of the data stream content, the technical problem that the prior art is difficult to comprehensively detect data security threats is solved, and the data security is significantly improved; The time progression matrix is acquired through the start time of a software running period, dynamic updating of each period of the time progression matrix is realized, replay attacks and dictionary attacks are effectively resisted, and dynamic security of the system is enhanced; Through the double-channel mutual detection mechanism, each running period of each channel verifies the data stream of the other party, and the reliability and credibility of the detection process are further improved.

[0013] The application further provides a software data stream detection system based on matrix multiplication, which is suitable for the software data stream detection method based on matrix multiplication and comprises: A detection point matrix acquisition module is configured to acquire a detection point matrix of a corresponding size according to the number of data stream detection points; A to-be-detected matrix acquisition module is configured to acquire a second to-be-detected result matrix of a second channel in the channel by recursively multiplying the initial matrix of the first channel in the channel and the detection point matrix according to the sequence of the data stream detection points, and acquire a first to-be-detected result matrix of a first channel in the channel by recursively multiplying the initial matrix of the second channel in the channel and the detection point matrix. A security verification module is configured to perform security verification by the second to-be-detected result matrix of the second channel in the channel and the first expected result matrix of the first channel in the channel for the first channel, and perform security verification by the first to-be-detected result matrix of the first channel in the channel and the second expected result matrix of the second channel in the channel for the second channel, and when the security verification passes, output the security data, otherwise, prompt the software to enter a fail-safe state.

[0014] The application has the following beneficial effects: The second to-be-detected result matrix and the first to-be-detected result matrix are acquired by taking the sequence of the data stream detection points as the matrix recursive multiplication sequence, that is, the non-commutative property of the matrix multiplication makes any sequence disorder result in a difference between the to-be-detected result matrix and the expected result matrix, and each step of matrix multiplication transmits and accumulates the features of the previous detection points to the subsequent result, so that the content anomaly of any point in the data stream is diffused to the to-be-detected result matrix through the multiplication operation, and also causes a difference between the to-be-detected result matrix and the expected result matrix. By considering the abnormal change of the data stream execution sequence and the abnormal change of the data stream content, the technical problem that the existing technology cannot comprehensively detect data security threats is solved, and the security of the data is significantly improved.

[0015] The application further provides a computer device comprising a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the computer device runs, the processor and the memory communicate through the bus, and the processor executes the machine readable instructions to perform the steps of the software data stream detection method based on matrix multiplication. BRIEF DESCRIPTION OF DRAWINGS

[0016] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings. The drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0017] Figure 1 This is a schematic diagram of the first process of the software data flow detection method based on matrix multiplication of the present invention; Figure 2 This is a schematic diagram of the second process of the software data flow detection method based on matrix multiplication of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only one preferred embodiment of this invention and are only used to explain this invention. They do not limit the scope of protection of this invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0019] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations (or steps) can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it may also have additional steps not included in the figures; the process may correspond to a method, function, procedure, subroutine, subroutine, etc.

[0020] Example 1: like Figure 1 As shown, the software data flow detection method based on matrix multiplication includes the following steps: S1: Obtain a detection point matrix of the corresponding size based on the number of detection points in the data stream.

[0021] Specifically, in S1, all matrix elements in the detection point matrix are distinct prime numbers.

[0022] The detection point matrix can be an m×m matrix, where m is any integer value greater than or equal to 2. All matrix elements in the detection point matrix are distinct prime numbers, ensuring that the detection point matrix is ​​invertible and reducing the risk of brute-force attacks.

[0023] S2: in the order of the data flow detection point positions as a recursive multiplication order, obtaining a second to-be-detected result matrix of a second channel in the channel according to recursive multiplication of an initial matrix of the first channel in the channel and the detection point matrix, and obtaining a first to-be-detected result matrix of the first channel according to recursive multiplication of an initial matrix of the second channel and the detection point matrix.

[0024] As a preferred embodiment, before the S2, the method further comprises: obtaining a first time progression matrix of the first channel according to a characteristic of the first channel and a start time of the software running period, and obtaining the initial matrix of the first channel based on the first time progression matrix; obtaining a second time progression matrix of the second channel according to a characteristic of the second channel and the start time, and obtaining the initial matrix of the second channel based on the second time progression matrix; wherein the first channel and the second channel are mutual security verification channels.

[0025] As shown in Figure 2 , at the start of the security software running period, the two channels each generate a time progression matrix , is the start time of the software running period, i.e., the current time value, is a unique key variable obtained according to the characteristics of the two channels, is fixed and is a prime number, and the two channels use different key variables. The first time progression matrix of the first channel, i.e., the A channel, , represents the start time of the software running period, represents the key variable of the first channel, and the second time progression matrix of the second channel, i.e., the B channel, , represents the start time of the software running period, represents the key variable of the second channel. In the initial stage, the first channel and the second channel each select a fixed 2x1 column vector ( whose elements can be prime numbers), for example , where and are prime numbers, and the two channels can each use different values. The fixed 2x1 column vector of the A channel, , and the fixed 2x1 column vector of the B channel, wherein, , are prime numbers, and may be different, and may be different. The expression for obtaining the initial matrix of the first channel based on the first time progression matrix is: , The expression for obtaining the initial matrix of the second channel based on the second time progression matrix, representing the initial matrix of the first channel, is as follows: , This represents the initial matrix of the second channel. By obtaining the time-progression matrix through the start time of the software's runtime cycle, dynamic updates of the time-progression matrix are achieved for each cycle, effectively resisting replay attacks and credential stuffing attacks, and enhancing the dynamic security of the system.

[0026] In some embodiments, S2 includes: S21: Using the order of the data stream detection points in the first channel as the recursive multiplication order, perform a recursive multiplication based on the initial matrix of the first channel and the detection point matrix to obtain the second detection result matrix; S22: Using the order of the data stream detection points in the second channel as the order of the second recursive multiplication, perform a second recursive multiplication based on the initial matrix of the second channel and the detection point matrix to obtain the first detection result matrix.

[0027] Specifically, S21 includes: S211: Use the initial matrix of the first channel as the first process matrix; S212: Multiply the first process matrix by the detection point matrix in the first channel to obtain the first matrix. Determine whether the detection point matrix in the first channel is the last detection point matrix. If it is the last detection point matrix, use the first matrix as the second detection result matrix. Otherwise, use the first matrix as the first process matrix and re-execute S212.

[0028] Specifically, S22 includes: S221: Use the initial matrix of the second channel as the second process matrix; S222: Multiply the detection point matrix in the second channel by the second process matrix to obtain the second matrix. Determine whether the detection point matrix in the second channel is the last detection point matrix. If it is the last detection point matrix, use the second matrix as the first detection result matrix. Otherwise, use the second matrix as the second process matrix and re-execute S222.

[0029] like Figure 2 As shown, the detection point matrices of the first and second channels are consistent. During the data stream detection process in the first and second channels, the process matrix for the first detection point in the first channel is... , This represents the first detection point matrix, and the process matrix for the i-th detection point in the first channel. , This represents the process matrix for the (i-1)th detection point in the first channel. represents the i-th detection point matrix. The process matrix of the 2nd detection point in the first channel , represents the 2nd detection point matrix, the process matrix of the 3rd detection point in the first channel , represents the 3rd detection point matrix, if the number of data stream detection point positions is n, the number of detection point matrices is also n, at this time, the first to-be-detected result matrix , represents the process matrix of the n-1th detection point in the first channel, represents the n-th detection point matrix. The process matrix of the first detection point in the second channel , the process matrix of the i-th detection point in the second channel , represents the process matrix of the i-1th detection point in the second channel. The process matrix of the 2nd detection point in the second channel , the process matrix of the 3rd detection point in the second channel , the second to-be-detected result matrix , represents the process matrix of the n-1th detection point in the second channel. Through the mutual detection mechanism of the two channels, the data stream of the other party is verified by each channel in each running period, further improving the reliability and credibility of the detection process.

[0030] S3: The first channel performs security verification with the first expected result matrix of the first channel through the second to-be-detected result matrix, and the second channel performs security verification with the second expected result matrix of the second channel through the first to-be-detected result matrix, if the security verification is passed, the security data is output, otherwise, the fault safety state is entered.

[0031] Before the S3, further comprising: obtaining the first expected result matrix according to the initial matrix and the detection point matrix of the first channel, and obtaining the second expected result matrix according to the initial matrix and the detection point matrix of the second channel.

[0032] Specifically, the expression for obtaining the first expected result matrix according to the initial matrix and the detection point matrix of the first channel is: ; represents the first expected result matrix, represents the initial matrix of the first channel, , represents the first detection point matrix in the detection point matrix, represents the second detection point matrix in the detection point matrix, represents the n-th detection point matrix in the detection point matrix, n represents the number of data stream detection point positions; The expression for obtaining the second expected result matrix according to the initial matrix of the second channel and the detection point matrix is: ; represents the second expected result matrix, represents the initial matrix of the second channel.

[0033] In the verification and validation phase, when , and , it indicates that the security verification is passed, and the security data (such as the control signal) is output, when or , the fault safety state (such as output 0 or security default value) is entered, and the non-commutative property of matrix multiplication makes any order disorder result in a difference between the to-be-detected result matrix and the expected result matrix, and each step of matrix multiplication will transmit and accumulate the characteristics of the previous detection point to the subsequent result, and the abnormal content of any point in the data stream will be spread to the to-be-detected result matrix through multiplication operation, and also cause a difference between the to-be-detected result matrix and the expected result matrix. By verifying the to-be-detected result matrix and the expected result matrix, not only the abnormality caused by the sequential abnormal change of the data stream is considered, but also the abnormality caused by the abnormal change of the content of the data stream is considered, and the technical problem that the existing technology is difficult to comprehensively detect the data security threat is solved, and the security of the data is significantly improved.

[0034] Embodiment 2: The embodiment also provides a software data stream detection system based on matrix multiplication, which is suitable for the software data stream detection method based on matrix multiplication and includes: A detection point matrix acquisition module is configured to acquire a detection point matrix of a corresponding size according to the number of data stream detection points; A to-be-detected matrix acquisition module is configured to acquire a second to-be-detected result matrix of a second channel in the channel by recursively multiplying the initial matrix of the first channel and the detection point matrix according to the recursive multiplication order of the sequence of the data stream detection points, and acquire a first to-be-detected result matrix of a first channel in the channel by recursively multiplying the initial matrix of the second channel and the detection point matrix; A security verification module is configured to perform security verification on the first channel by multiplying the second to-be-detected result matrix of the second channel and the first expected result matrix of the first channel, and perform security verification on the second channel by multiplying the first to-be-detected result matrix of the first channel and the second expected result matrix of the second channel, and when the security verification is passed, output the security data, otherwise, prompt the software to enter the fault safety state.

[0035] In the embodiment, the detection point matrix of the first channel is consistent with that of the second channel. In the data stream detection process of the first channel and the second channel, the process matrix of the first detection point in the first channel is , represents the initial matrix of the second channel, represents the process matrix of the first detection point, the process matrix of the i-th detection point in the first channel is , represents the process matrix of the i-1-th detection point in the first channel, represents the process matrix of the second detection point in the first channel, , represents the process matrix of the third detection point in the first channel, , represents the process matrix of the third detection point, if the number of data stream detection points is n, the number of detection point matrix is also n, at this time, the first to-be-detected result matrix is , represents the process matrix of the n-1-th detection point in the first channel, represents the process matrix of the n-th detection point in the first channel, , represents the initial matrix of the first channel, the process matrix of the i-th detection point in the second channel is , represents the process matrix of the i-1-th detection point in the second channel. The process matrix of the second detection point in the second channel is , the process matrix of the third detection point in the second channel is , the second to-be-detected result matrix is , represents the process matrix of the n-1-th detection point in the second channel.

[0036] When , and , represents the first expected result matrix, represents the second expected result matrix, indicating that the security verification is passed, and the security data (such as control signal) is outputted, when or , then entering a failsafe state (such as outputting 0 or a safety default value), and the non-commutative property of matrix multiplication makes any sequence disorder result in a difference between the to-be-detected result matrix and the expected result matrix, and each step of matrix multiplication will transmit and accumulate the features of the previous detection point to the subsequent result, and the content abnormality of any point in the data stream will be spread to the to-be-detected result matrix through multiplication operation, and also cause a difference between the to-be-detected result matrix and the expected result matrix. By verifying the to-be-detected result matrix and the expected result matrix, not only the abnormality caused by the sequence abnormality of the data stream is considered, but also the abnormality caused by the content abnormality of the data stream is considered, and the technical problem that the prior art is difficult to comprehensively detect data security threats is solved, and the security of the data is significantly improved.

[0037] Embodiment 3 The embodiment also provides a computer device, comprising a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the computer device runs, the processor and the memory communicate through the bus, and the processor executes the machine readable instructions to perform the steps of the matrix multiplication-based software data stream detection method.

[0038] The above specific embodiments are preferred embodiments of the matrix multiplication-based software data stream detection method and system of the present application, and do not limit the specific implementation range of the present application, the range of the present application includes but is not limited to the specific embodiments, and equivalent changes made according to the shape and structure of the present application are within the protection scope of the present application.

Claims

1. A method for software data flow detection based on matrix multiplication, characterized in that, The method comprises the following steps: S1: obtaining a detection point matrix corresponding to a scale according to a number of data stream detection points; S2: taking a sequence of the data stream detection points as a recursive multiplication sequence, performing recursive multiplication on an initial matrix of a first channel and the detection point matrix to obtain a second to-be-detected result matrix of a second channel, and performing recursive multiplication on an initial matrix of the second channel and the detection point matrix to obtain a first to-be-detected result matrix of the first channel; S3: the first channel performs security verification through the second to-be-detected result matrix and a first expected result matrix of the first channel, the second channel performs security verification through the first to-be-detected result matrix and a second expected result matrix of the second channel, and if the security verification passes, security data is output, otherwise, a fault security state is entered.

2. The matrix multiplication based software data flow detection method of claim 1, wherein, In S1, the matrix elements in the detection point matrix are all different prime numbers.

3. The matrix multiplication based software data flow detection method of claim 1, wherein, Before S2, the method further comprises: obtaining a first time progression matrix of the first channel according to a characteristic of the first channel and a start time of a software running period, and obtaining the initial matrix of the first channel based on the first time progression matrix; obtaining a second time progression matrix of the second channel according to a characteristic of the second channel and the start time, and obtaining the initial matrix of the second channel based on the second time progression matrix; The first channel and the second channel are mutual security verification channels.

4. The matrix multiplication based software data flow detection method of claim 1, wherein, S2 comprises: S21: taking a sequence of the data stream detection points in the first channel as a one-time recursive multiplication sequence, performing one-time recursive multiplication on the initial matrix of the first channel and the detection point matrix to obtain the second to-be-detected result matrix; S22: taking a sequence of the data stream detection points in the second channel as a two-time recursive multiplication sequence, performing two-time recursive multiplication on the initial matrix of the second channel and the detection point matrix to obtain the first to-be-detected result matrix.

5. The matrix multiplication based software data stream inspection method of claim 4, wherein, S21 comprises: S211: taking the initial matrix of the first channel as a first process matrix; S212: multiplying the first process matrix with the detection point matrix in the first channel to obtain a first matrix, determining whether the detection point matrix in the first channel is a last detection point matrix, if yes, taking the first matrix as the second to-be-detected result matrix, otherwise, taking the first matrix as the first process matrix, and re-executing S212.

6. The matrix multiplication based software data stream inspection method of claim 4, wherein, S22 comprises: S221: taking the initial matrix of the second channel as a second process matrix; S222: multiplying the second process matrix with the detection point matrix in the second channel to obtain a second matrix, determining whether the detection point matrix in the second channel is a last detection point matrix, if yes, taking the second matrix as the first to-be-detected result matrix, otherwise, taking the second matrix as the second process matrix, and re-executing S222.

7. The matrix multiplication based software data stream inspection method of claim 1, wherein, Before S3, the method further comprises: obtaining the first expected result matrix according to the initial matrix of the first channel and the detection point matrix, and obtaining the second expected result matrix according to the initial matrix of the second channel and the detection point matrix.

8. The matrix multiplication based software data stream inspection method of claim 7, wherein, The expression for obtaining the first expected result matrix according to the initial matrix of the first channel and the detection point matrix is: ; represents a first expected result matrix, represents an initial matrix of the first channel, , represents a first detection point matrix in the detection point matrix, represents a second detection point matrix in the detection point matrix, represents an nth detection point matrix in the detection point matrix, n represents the number of data stream detection points; The expression for obtaining the second expected result matrix according to the initial matrix of the second channel and the detection point matrix is: ; denotes a second expected result matrix, denotes an initial matrix of the second channel.

9. A software data flow detection system based on matrix multiplication, adapted to the software data flow detection method based on matrix multiplication according to any one of claims 1 to 8, characterized in that, The method comprises the following steps: The detection point matrix acquisition module is configured to acquire a detection point matrix of a corresponding size according to the number of data stream detection points. The to-be-detected matrix acquisition module is configured to take the sequence of the data stream detection points as a recursive multiplication sequence, acquire a second to-be-detected result matrix of the second channel in the channel by recursively multiplying the initial matrix of the first channel in the channel and the detection point matrix, and acquire a first to-be-detected result matrix of the first channel in the channel by recursively multiplying the initial matrix of the second channel in the channel and the detection point matrix. The security verification module is configured to perform security verification on the second to-be-detected result matrix of the second channel in the first channel and the first expected result matrix of the first channel, and perform security verification on the first to-be-detected result matrix of the first channel in the second channel and the second expected result matrix of the second channel, and when the security verification is passed, output security data, otherwise, prompt the software to enter a fault safety state.

10. A computer device, comprising: The computer device comprises a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the computer device runs, the processor and the memory communicate through the bus, and the processor executes the machine readable instructions to perform the steps of the matrix multiplication based software data stream detection method according to any one of claims 1-8. ​

Images