A differential error analysis test method and system for security chips

Abstract

The present invention discloses a differential error analysis test method and system applied to a security chip, and relates to the field of signal analysis and transmission. The present invention collects characteristic data of multiple differential signal line pairs, performs superposition comprehensive analysis on multiple characteristic factors, and improves the accuracy of analysis; by constructing a final KNN classification model, a classification tool is provided for subsequent analysis, and the analysis process is simplified; by classifying the characteristic data of the security chip to be analyzed, it can be judged whether the security chip to be analyzed has a differential error, and if so, it can be adjusted in time; by adjusting the characteristic data of the security chip to be analyzed, the differential error of the security chip to be analyzed is corrected; the characteristic data of the security chip to be analyzed after adjustment is compared with the characteristic data of the security chip to be analyzed before adjustment, and the cause of the differential error of the security chip to be analyzed is analyzed.

Description

Technical Field

[0001] The present invention belongs to the field of signal analysis and transmission, and in particular, relates to a differential error analysis test method and system applied to a security chip. Background Art

[0002] Chinese patent CN107818058B discloses a semiconductor device and a memory controller for receiving a differential signal, including a differential signal phase detector and a receiver; the differential signal phase detector receives the differential signal, detects its phase and generates a mode control signal; the receiver receives the differential signal and a reference voltage, and according to the mode control signal, performs a differential signal processing operation in a differential mode, or performs a first signal and a reference voltage processing operation in a single mode; the semiconductor device can be a memory controller; data transmission can be disabled in the single mode to avoid data misrecognition caused by noise.

[0003] The cause of the differential signal error in the security chip may be due to the combined effect of multiple factors; and currently, the analysis of the cause of the differential signal error often analyzes multiple factors separately, such as the noise factor considered in the above patent, which makes it impossible to accurately analyze the cause of the differential signal error. Summary of the invention

[0004] In view of the problems in the related art, the present invention proposes a differential error analysis test method and system applied to a security chip to overcome the above technical problems existing in the existing related art.

[0005] To solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The present invention is a differential error analysis test method applied to a security chip, comprising the following steps:

[0007] S1, collecting characteristic data of multiple differential signal line pairs;

[0008] S2, collecting signal data of multiple differential signal line pairs and calculating the average value to obtain a second data mean matrix; performing a clustering operation on the second data mean matrix to obtain a differential data level label set;

[0009] S3, constructing differential data level labels of multiple differential signal line pairs according to the differential data level label set and the characteristic data of multiple differential signal line pairs to obtain an initial differential data level label data set; constructing a final KNN classification model according to the characteristic data of multiple differential signal line pairs and the initial differential data level label data set;

[0010] S4, acquiring feature data of the differential signal line pair on the security chip to be analyzed to obtain an initial comprehensive data set; using the final KNN classification model to classify the initial comprehensive data set to obtain second initial differential data level label data; and determining whether the initial comprehensive data set needs to be adjusted according to the second initial differential data level label data;

[0011] S5. analyzing the differential errors of the security chip to be analyzed according to the result of adjusting the initial comprehensive data set;

[0012] Since the differential signal is carried by the differential signal line, the differential error generated in the differential signal has a great correlation with the differential signal line; therefore, the characteristic data of multiple differential signal line pairs are collected so that the differential error generated in the differential signal can be analyzed according to the characteristic data of the differential signal line pairs; the inductance difference data, capacitance difference data, differential impedance data and signal delay data of the differential signal line pairs can intuitively reflect whether a differential error occurs in the differential signal, and provide a classification standard for the subsequent classification of the characteristic data of multiple differential signal line pairs; by clustering the second data mean matrix, similar inductance difference data, capacitance difference data, differential impedance data and signal delay data are combined into one category, which reduces the discreteness and complexity of the data and facilitates the subsequent classification of the characteristic data of multiple differential signal line pairs; differential data level labels of the characteristic data of multiple differential signal line pairs are constructed, which provides a classification standard for the subsequent construction of the KNN classification model. Class label data reduces the complexity of classification; by constructing the final KNN classification model, a classification tool is provided for the subsequent acquisition of the corresponding differential data level label data according to the feature data of the security chip to be analyzed, without the need to acquire and calculate the inductance difference data, capacitance difference data, differential impedance data and signal delay data of the security chip to be analyzed, thereby simplifying the analysis process and reducing the dependence of the analysis process on data; by classifying the feature data of the security chip to be analyzed, it can be determined whether the security chip to be analyzed has a differential error, and if so, it can be adjusted in time; by adjusting the feature data of the security chip to be analyzed, the differential error of the security chip to be analyzed is corrected to ensure that the security chip can work normally; the feature data of the security chip to be analyzed after adjustment is compared with the feature data of the security chip to be analyzed before adjustment, so as to analyze the cause of the differential error of the security chip to be analyzed.

[0013] Preferably, the S1 comprises the following steps:

[0014] S11, set a plurality of differential signal line pairs, and obtain a differential signal line pair set a = {a1, a2, ..., a i ,...,a a′}, a irepresents the set i-th differential signal line pair, a′ represents the total number of the set differential signal line pairs; the line widths of the two differential signal lines in each of the differential signal line pairs are set to be the same, and the length data and line width data of the plurality of differential signal line pairs are collected to obtain a first initial length data set and the first initial line width data set b′={b1′,b2′,...,b i ′,...,b a ″}; and Respectively represent the initial length data of the first and second differential signal lines in the i-th differential signal line pair; b i ' represents the initial line width data of the two differential signal lines in the i-th differential signal line pair;

[0015] Setting the test point spacing; taking the signal input positions of the plurality of differential signal line pairs as the starting point and evenly setting a plurality of test point pairs according to the test point spacing to obtain a first test point matrix as follows,;

[0016]

[0017] in, and Respectively represent the j-th test point on the first and second differential signal lines in the i-th differential signal line pair; b i Indicates the total number of test points on the i-th differential signal line pair;

[0018] S12, in the differential signal line pair set a={a1, a2, ..., a i ,...,a a′}, a signal source with adjustable strength and adjustable position is set around each differential signal line pair, and a first signal source set is obtained. represents the signal source set around the ith differential signal line pair; setting the first signal source set The initial intensity of each signal source is obtained by first initial intensity data set c′={c1′,c′2,...,c i ′,...,c′ a′}, c i ' represents the initial signal strength data of the signal source set around the i-th differential signal line pair;

[0019] Set the first signal source set The distance from each signal source to the corresponding differential signal line pair is obtained to obtain the first distance data set Indicates the first signal source set The distance from the i-th signal source to the corresponding differential signal line pair;

[0020] S13, for the differential signal line pair set a={a1, a2, ..., a i ,...,a a′} each differential signal line pair in the array; after the arrangement is completed, according to the first test point matrix Measure the spacing distance between each test point on each differential signal line pair and the routing angle at each test point to obtain the first initial spacing data matrix c and the first initial routing angle data matrix They are as follows:

[0021]

[0022] Among them, c ij Represents the interval data between the jth test points on the two differential signal lines in the i-th differential signal line pair; and They respectively represent the routing angles at the j-th test point on the first and second differential signal lines in the i-th differential signal line pair;

[0023] By setting a plurality of test points on each differential signal pair, the randomness of the collected characteristic data can be prevented, thereby improving the reliability of the collected characteristic data.

[0024] Preferably, S2 comprises the following steps:

[0025] S21, for the differential signal line pair set a={a1, a2, ..., a i ,...,a a′ The inductance difference data, capacitance difference data, and differential impedance data at each test point on each arranged differential signal line pair in the first differential signal line pair are measured and calculated to obtain a first differential signal line pair data matrix. as follows,

[0026]

[0027] in, Respectively represent the inductance difference data, capacitance difference data, and differential impedance data between the jth test point on the first and second differential signal lines in the i-th differential signal line pair;

[0028] Calculate the first differential signal line pair data matrix The mean values ​​of the inductance difference data, capacitance difference data, and differential impedance data at each test point in the , and the first data mean matrix d1′ is obtained; as follows,

[0029]

[0030] Among them, d1′i1 d1′ i2 and d1′ i3 Respectively represent the inductance difference mean value data, capacitance difference mean value data, and differential impedance mean value data on the i-th differential signal line pair;

[0031] S22, for the differential signal line pair set a={a1, a2, ..., a i ,...,a a′ The signal delay data at each test point on each arranged differential signal line pair in} is measured to obtain a first signal delay data matrix d; as follows,

[0032]

[0033] Among them, d ij Represents the signal delay data at the jth test point in the i-th differential signal line pair;

[0034] Calculate the mean of the signal delay data at each test point in the first signal delay data matrix d to obtain the first signal delay data mean set d2′={d2′1, d2′2, ..., d2′ i ,...,d2′ a′}, d2′ i represents the mean value of signal delay data on the i-th differential signal line pair;

[0035] The first data mean matrix d1′ and the first signal delay data mean set d2′={d2′1, d2′2, ..., d2′ i ,...,d2′ a′} are merged to obtain the second data mean matrix d3′; as follows,

[0036]

[0037] Among them, d3′ i1 d3′ i2 d3′ i3 d3′ i4 Respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data on the i-th differential signal line;

[0038] S23, clustering the second data mean matrix d3′ to obtain the final differential signal line pair classification data matrix set represents the classification data matrix of the ith differential signal line pair obtained by clustering, Indicates the total number of classified data matrices of signal line pairs obtained by clustering; as follows,

[0039]

[0040] in, They respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data on the jth differential signal line pair in the i-th differential signal line pair classification data matrix, represents the total number of differential signal line pairs in the i-th differential signal line pair classification data matrix;

[0041] S24, classifying the final differential signal line pair data matrix set Each differential signal line in the classification data matrix sets a differential data level label to obtain a differential data level label set e i Indicates the differential data level label corresponding to the i-th differential signal line pair classification data matrix;

[0042] By calculating and measuring the inductance difference data, capacitance difference data, differential impedance data and signal delay data of each test point and performing average value calculation, the accuracy of the acquired inductance difference data, capacitance difference data, differential impedance data and signal delay data is improved, thereby improving the accuracy of subsequent label construction of the characteristic data of the differential signal line.

[0043] Preferably, the calculation formulas for the inductance difference data, capacitance difference data, and differential impedance data at each test point on each arranged differential signal line pair in the differential signal line pair set in S21 are as follows:

[0044]

[0045]

[0046] Wherein, β1, β2, β3, β4, β5, β6 represent the length of the first differential signal line in the ith differential signal line pair, the length of the second differential signal line in the ith differential signal line pair, the magnetic permeability of free space, the relative magnetic permeability of the ith differential signal line pair, the distance between the jth test points on the first and second differential signal lines in the ith differential signal line pair, and the width of the first and second differential signal lines in the ith differential signal line pair; χ3 and χ4 represent the dielectric constant of free space and the relative dielectric constant of the ith differential signal line pair; γ1 and γ2 represent the distance from the first and second differential signal lines in the ith differential signal line pair to the ground; φ, represent the relative permittivity of vacuum and the relative permittivity of substrate respectively;

[0047] Preferably, the S23 comprises the following steps:

[0048] S231, randomly select a number of rows of data from the second data mean matrix d3′ to obtain an initial differential signal line pair classification center data matrix as follows,

[0049]

[0050] in, Respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data in the randomly selected i-th row of data;

[0051] S232, traverse the data in the second data mean matrix d3′ row by row, and calculate the initial differential signal line pair classification center data matrix for each row of data in the second data mean matrix d3′ The Euclidean distance between each row of data in the , and the initial Euclidean matrix is ​​obtained as follows,

[0052]

[0053] in, The second data mean matrix d3′ is the matrix of the classification center data of the initial differential signal line pair removed. The j-th row of data after the data in the initial differential signal line pair classification center data matrix The Euclidean distance between the i-th row of data;

[0054] Remove the initial differential signal line pair classification center data matrix from the second data mean matrix d3′ The data in the first row is classified into the initial Euclidean matrix. The initial differential signal line pair classification data matrix set is obtained from the classification center data corresponding to the smallest data in the jth column data. f i Represents the i-th initial differential signal line pair classification data matrix; as follows,

[0055]

[0056] Among them, f ij1 、f ij2 、f ij3 、f ij4 They respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data on the jth differential signal line pair in the i-th initial differential signal line pair classification data matrix; i′ represents the total number of differential signal line pairs in the i-th initial differential signal line pair classification data matrix;

[0057] S233. Building a honey badger population represents the i-th honey badger individual in the honey badger population, g represents the size of the honey badger population; set the honey badger population The maximum number of iterations is g1′, the current number of iterations is g′2, and the number of search space dimensions is

[0058] According to the initial differential signal line pair classification center data matrix Set the initial position of each honey badger in the honey badger population and get the initial position matrix set Represents the initial position matrix of the i-th honey badger individual in the honey badger population; as follows,

[0059]

[0060] in, Respectively , represents the position components of the inductance difference mean data, capacitance difference mean data, differential impedance mean data, and signal delay mean data in the classification center data of the j-th differential signal line pair;

[0061] S234, according to the initial Euclidean matrix Set the fitness function h′ of the honey badger population as follows:

[0062]

[0063] In the formula, h i ′ represents the sum of the Euclidean distances from each row of the ith differential signal line pair classification data matrix in the differential signal line pair classification data matrix set obtained after each round of iteration to the classification center of the corresponding differential signal line pair classification data matrix;

[0064] S235, start iterative operation; in each round of iteration, calculate the fitness value of each honey badger according to the fitness function h′ of the honey badger population, select the honey badger individual with the largest fitness value as the global best individual; the position of the global best individual is used as the global best position; update the position of the honey badger population honey badger according to the global best position;

[0065] S236. When g′2≥g1′, stop iteration and output the final global optimal position matrix as follows,

[0066]

[0067] in, Respectively Represents the position component of the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data in the classified center data of the i-th differential signal line pair; otherwise, continue to iterate;

[0068] According to the final global optimal position matrix And use S232 to get the final differential signal line pair classification data matrix set

[0069] The honey badger optimization algorithm has good optimization ability. Therefore, by using the honey badger optimization algorithm to continuously iteratively optimize multiple classification center data of the second data mean matrix, the overall discreteness of the second data mean matrix after classification is continuously reduced and finally stabilized at the minimum discreteness value. At this point, the classification of the second data mean matrix achieves the best effect.

[0070] Preferably, in S233, the center data matrix is ​​classified according to the initial differential signal line pair Set the initial position of each honey badger in the honey badger population and get the initial position matrix set The following steps are involved:

[0071] S2331, set the value range of the inductance difference mean data, the capacitance difference mean data, the differential impedance mean data and the signal delay mean data, and obtain the mean data value interval set Respectively represent the lower limits of the inductance difference mean data, the capacitance difference mean data, the differential impedance mean data and the signal delay mean data; Respectively represent the upper limits of the inductance difference mean data, the capacitance difference mean data, the differential impedance mean data, and the signal delay mean data;

[0072] S2332, based on the mean data value interval set

[0073] and the initial differential signal pair classification center data matrix set up The calculation formulas are as follows:

[0074]

[0075] In the formula, rand j1 、rand j2 、rand j3 、rand j4 For Generated random number; h(rand j1 )、h(rand j2)、h(rand j3 )、h(rand j4 ) respectively represent the j1 、rand j2 、rand j3 、rand j4 The generated tent mapping values; the function expressions are as follows,

[0076]

[0077] Where, α1, α2, α3, and α4 are tent mapping parameters, with values ​​ranging from 0 to 1;

[0078] The data generated by tent mapping has good distribution. Therefore, by using tent mapping to generate the initial position of each honey badger individual in the honey badger population, the richness of the initial solution space of the honey badger optimization algorithm is improved, thereby improving the optimization speed of the honey badger optimization algorithm.

[0079] Preferably, S3 comprises the following steps:

[0080] S31, according to the first initial length data set Calculate the differential signal line pair set a = {a1, a2, ..., a i ,...,a a′}, and obtain the first initial length difference data set represents the length difference between the two differential signal lines in the i-th differential signal line pair;

[0081] Calculate the standard deviation of each row of data in the first initial interval data matrix c to obtain the first initial interval standard deviation data set represents the standard deviation of the interval between the two differential signal lines in the ith differential signal line pair at each test point;

[0082] Select the first initial alignment angle data matrix The minimum value of each row of data in the first row is obtained to obtain the first initial minimum alignment angle data set. Indicates the minimum routing angle data on the i-th differential signal line pair;

[0083] S32, the first initial line width data set b′={b1′, b2′, ..., b i ′,...,b a ″}, the first initial length difference data set The first initial intensity data set c′={c1′,c′2,...,c i′,...,c′ a′}、First distance dataset The first initial interval standard deviation data set The first initial minimum routing angle data set Merge the data to obtain the first initial comprehensive data matrix l; as follows,

[0084]

[0085] Among them, l i1 , l i2 , l i3 , l i4 , l i5 , l i6 Respectively represent the initial line width data of the i-th differential signal line pair, the length difference data of the two differential signal lines, the initial signal strength data of the surrounding signal source, the distance data from the surrounding signal source, the standard deviation data of the interval between the two differential signal lines at each test point, and the minimum routing angle data;

[0086] S33, according to the differential data level label set e={e1, e2, ..., e i ,...,e d～} and the final differential signal line pair classification data matrix set Set the differential data level label of each row of data in the first initial comprehensive data matrix l to obtain the initial differential data level label data set l′={l1′,l2′,...,l i ′,...,l a ″},l i ' represents the differential data level label data of the i-th row data in the first initial comprehensive data matrix l;

[0087] S34, constructing an initial KNN classification model; using the initial differential data level label data set l′={l1′,l2′,...,l i ′,...,l a ″} and the first initial comprehensive data matrix l to perform training and testing operations on the initial KNN classification model; after the training and testing are completed, the final KNN classification model is obtained;

[0088] By using the initial differential data level label data set and the first initial comprehensive data matrix to train and test the initial KNN classification model, the initial KNN classification model has better recognition and classification capabilities for the characteristic data of the differential signal line pairs, providing a better model tool for subsequent classification of the characteristic data of the differential signal lines on the security chip to be analyzed.

[0089] Preferably, S4 comprises the following steps:

[0090] S41, setting the security chip to be analyzed; collecting the initial line width data of the differential signal line pair on the security chip to be analyzed, the length difference data of the two differential signal lines, the initial signal strength data of the surrounding signal sources, the distance data from the surrounding signal sources, the standard deviation data of the interval between the two differential signal lines at each test point, and the minimum routing angle data, to obtain an initial comprehensive data set Respectively representing the line width data of the differential signal line pair to be analyzed, the length difference data of the two differential signal lines, the signal strength data of the surrounding signal sources, the distance data from the surrounding signal sources, the standard deviation data of the interval between the two differential signal lines at each test point, and the minimum routing angle data;

[0091] S42, using the final KNN classification model to classify the initial comprehensive data set Classify and obtain the second initial differential data level label data m1′;

[0092] S43, setting the differential data level threshold m′2;

[0093] When m1′<m′2, for the initial comprehensive data set Adjust until m1′≥m′2; after the adjustment is completed, the final data mean set n′={n1′,n′2,n3′,n′4,n5′,n′6} is obtained;

[0094] Otherwise, there is no need to Performing adjustments; using the initial comprehensive data set as a final data mean set;

[0095] By collecting the initial line width data of the differential signal line pair on the security chip to be analyzed, the length difference data of the two differential signal lines, the initial signal strength data of the surrounding signal sources, the distance data to the surrounding signal sources, the standard deviation data of the interval between the two differential signal lines at each test point, and the minimum routing angle data, data support is provided for the subsequent differential signal line pair on the security chip to be analyzed to determine whether there is a differential error; the final KNN classification model is used to classify the initial comprehensive data set, and according to the classification results, it can be intuitively seen whether there is a differential error in the differential signal line pair on the security chip to be analyzed corresponding to the initial comprehensive data set.

[0096] Preferably, in S43, the initial comprehensive data set Adjustment is performed until m1′≥m′2. After the adjustment is completed, obtaining the final data mean set includes the following steps:

[0097] S431, setting the line width data of the differential signal line pair on the security chip to be analyzed after the signal is input, the length difference data of the two differential signal lines, the signal strength data of the surrounding signal source, the distance data from the surrounding signal source, the standard deviation data of the interval between the two differential signal lines at each test point, and the minimum routing angle data value interval, and obtaining a first value interval set Respectively representing the line width data of the differential signal line pair on the security chip to be analyzed after the signal is input, the length difference data of the two differential signal lines, the signal strength data of the surrounding signal source, the distance data from the surrounding signal source, the standard deviation data of the interval between the two differential signal lines at each test point, and the minimum lower limit of the routing angle data; Respectively representing the line width data of the differential signal line pair on the security chip to be analyzed after the signal is input, the length difference data of the two differential signal lines, the signal strength data of the surrounding signal source, the distance data from the surrounding signal source, the standard deviation data of the interval between the two differential signal lines at each test point, and the upper limit of the minimum routing angle data;

[0098] S432, start a loop operation; during each loop, the initial comprehensive data set is Update to obtain the updated second comprehensive data set The update formulas are as follows:

[0099]

[0100] In the formula, rand 1i 、rand 2i 、rand 3i 、rand 4i 、rand 5i 、rand 6i Respectively represent the random numbers generated for the line width data, the length difference data of the two differential signal lines, the signal strength data of the surrounding signal sources, the distance data from the surrounding signal sources, the standard deviation data of the interval between the two differential signal lines at each test point, and the minimum routing angle data in the i-th cycle;

[0101] S433: updating the updated second comprehensive data set The differential signal line pair on the security chip to be analyzed is set after the signal is input, so as to obtain the adjusted differential signal line pair on the security chip to be analyzed;

[0102] S434, using the final KNN classification model to classify the updated second comprehensive data set Perform classification operation to obtain the third initial differential data level label data m3′; when m3′≥m′2, stop the loop and update the second comprehensive data set As the final data mean set n′={n1′,n′2,n3′,n′4,n5′,n′6}; otherwise, loop until m3′≥m′2, and obtain the final data mean set n′={n1′,n′2,n3′,n′4,n5′,n′6}, n1′, n′2, n3′, n′4, n5′, n′6 respectively represent the final line width data, the final length difference data of the two differential signal lines, the final signal strength data of the surrounding signal sources, the final distance data from the surrounding signal sources, the final standard deviation data of the interval between the two differential signal lines at each test point and the minimum final routing angle data;

[0103] By continuously performing cyclic adjustment on the second comprehensive data set, the classification result of the second comprehensive data set according to the final KNN classification model is used as a test standard after each adjustment. When the classification result of the second comprehensive data set by the final KNN classification model meets the requirements, the cyclic adjustment of the second comprehensive data set is stopped; the adjusted second comprehensive data set is applied to the differential signal line pair on the security chip to be analyzed, thereby eliminating the differential error of the differential signal line pair on the security chip to be analyzed.

[0104] Preferably, S5 comprises the following steps:

[0105] S51, calculate the final data mean set n′={n1′, n′2, n3′, n′4, n5′, n′6} and the initial comprehensive data set The difference degree value of each data corresponding to the difference degree value set is obtained Respectively represent the final data mean set n′={n1′,n′2,n3′,n′4,n5′,n′6} and the initial comprehensive data set The difference degree values ​​of the line width data, the difference degree values ​​of the length difference data of the two differential signal lines, the difference degree values ​​of the signal strength data of the surrounding signal sources, the difference degree values ​​of the distance data from the surrounding signal sources, the difference degree values ​​of the standard deviation data of the interval between the two differential signal lines at each test point, and the difference degree values ​​of the minimum routing angle data; the calculation formulas are as follows,

[0106]

[0107] S52: the difference degree value set Each difference degree value in is sorted by numerical value; the data of the differential signal line pair corresponding to the maximum difference degree value is used as the first factor causing the differential error; the data of the differential signal line pairs corresponding to the difference degree values ​​sorted from large to small are used as the second factor, the third factor, the fourth factor, the fifth factor and the sixth factor causing the differential error in turn;

[0108] By comparing the final data mean set with the initial comprehensive data set and calculating the difference between the corresponding position data of the two, the main factors and other minor factors caused by the differential error can be reflected, thereby obtaining the analysis results.

[0109] A differential error analysis test system for a safety chip includes a first data acquisition module, a second data acquisition module, a clustering module, a label data construction module, a KNN classification model construction module, a third data acquisition module, a classification module, an adjustment module, a determination and adjustment module, and a differential error analysis module.

[0110] The present invention has the following beneficial effects:

[0111] 1. The present invention collects characteristic data of multiple differential signal line pairs and performs superimposed comprehensive analysis on multiple characteristic factors, thereby improving the accuracy of the analysis; by constructing the final KNN classification model, a classification tool is provided for subsequently obtaining corresponding differential data level label data according to the characteristic data of the security chip to be analyzed, thereby simplifying the analysis process and reducing the dependence of the analysis process on data; by classifying the characteristic data of the security chip to be analyzed, it can be determined whether the security chip to be analyzed has a differential error, and if so, it can be adjusted in time; by adjusting the characteristic data of the security chip to be analyzed, the differential error of the security chip to be analyzed is corrected; the characteristic data of the security chip to be analyzed after adjustment is compared with the characteristic data of the security chip to be analyzed before adjustment, thereby analyzing the cause of the differential error of the security chip to be analyzed.

[0112] 2. In the present invention, the honey badger optimization algorithm is used to continuously iteratively optimize the multiple classification center data of the second data mean matrix, so that the overall discreteness of the second data mean matrix after classification is continuously reduced and finally stabilized at the minimum discreteness value. At this point, the classification of the second data mean matrix achieves the best effect.

[0113] 3. In the present invention, the second comprehensive data set is continuously cyclically adjusted, and the loop is stopped when the classification result of the final KNN classification model for the second comprehensive data set meets the requirements; the adjusted second comprehensive data set is applied to the differential signal line pair on the security chip to be analyzed, thereby eliminating the differential error of the differential signal line pair on the security chip to be analyzed.

[0114] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0115] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0116] Figure 1 The present invention is a schematic diagram of a flow chart of a differential error analysis test system applied to a security chip for analyzing differential errors of a security chip. DETAILED DESCRIPTION

[0117] The following will be combined with the drawings in the embodiments of the invention to clearly and completely describe the technical solutions in the embodiments of the invention. Obviously, the described embodiments are only part of the embodiments of the invention, not all of the embodiments. Based on the embodiments in the invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the invention.

[0118] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0119] Embodiment 1

[0120] This embodiment is a differential error analysis test method applied to a security chip, comprising the following steps:

[0121] S1, collecting characteristic data of multiple differential signal line pairs;

[0122] The S1 comprises the following steps:

[0123] S11, set a plurality of differential signal line pairs, and obtain a differential signal line pair set a = {a1, a2, ..., a i ,...,a a′}, a i represents the set i-th differential signal line pair, a′ represents the total number of the set differential signal line pairs; the line widths of the two differential signal lines in each of the differential signal line pairs are set to be the same, and the length data and line width data of the plurality of differential signal line pairs are collected to obtain a first initial length data set and the first initial line width data set b′={b1′,b2′,...,b i ′,...,b a ″}; and Respectively represent the initial length data of the first and second differential signal lines in the i-th differential signal line pair; b i ' represents the initial line width data of the two differential signal lines in the i-th differential signal line pair;

[0124] Setting the test point spacing; taking the signal input positions of the plurality of differential signal line pairs as the starting point and evenly setting a plurality of test point pairs according to the test point spacing to obtain a first test point matrix As shown below:

[0125]

[0126] in, and Respectively represent the j-th test point on the first and second differential signal lines in the i-th differential signal line pair; b i Indicates the total number of test points on the i-th differential signal line pair;

[0127] S12, in the differential signal line pair set a={a1, a2, ..., a i ,...,a a′}, a signal source with adjustable strength and adjustable position is set around each differential signal line pair, and a first signal source set is obtained. represents the signal source set around the ith differential signal line pair; setting the first signal source set The initial intensity of each signal source in is obtained to obtain the first initial intensity data set c′={c1′,c′2,...,c i ′,...,c′ a′}, c i ' represents the initial signal strength data of the signal source set around the i-th differential signal line pair;

[0128] Set the first signal source set The distance from each signal source to the corresponding differential signal line pair is obtained to obtain the first distance data set Indicates the first signal source set The distance from the i-th signal source to the corresponding differential signal line pair;

[0129] S13, for the differential signal line pair set a={a1, a2, ..., a i ,...,a a′} each differential signal line pair in the array; after the arrangement is completed, according to the first test point matrix Measure the spacing distance between each test point on each differential signal line pair and the routing angle at each test point to obtain the first initial spacing data matrix c and the first initial routing angle data matrix They are as follows:

[0130]

[0131]

[0132] Among them, c ij Represents the interval data between the jth test points on the two differential signal lines in the i-th differential signal line pair; and They respectively represent the routing angles at the j-th test point on the first and second differential signal lines in the i-th differential signal line pair;

[0133] S2, collecting inductance difference data, capacitance difference data, differential impedance data and signal delay data of multiple differential signal line pairs and calculating average values ​​to obtain a second data mean matrix;

[0134] Performing a clustering operation on the second data mean matrix to obtain a differential data level label set;

[0135] The S2 comprises the following steps:

[0136] S21, for the differential signal line pair set a={a1, a2, ..., a i ,...,a a′ The inductance difference data, capacitance difference data, and differential impedance data at each test point on each arranged differential signal line pair in the first differential signal line pair are measured and calculated to obtain a first differential signal line pair data matrix. as follows,

[0137]

[0138] in, Respectively represent the inductance difference data, capacitance difference data, and differential impedance data between the jth test point on the first and second differential signal lines in the i-th differential signal line pair;

[0139] Calculate the first differential signal line pair data matrix The mean values ​​of the inductance difference data, capacitance difference data, and differential impedance data at each test point in the , and the first data mean matrix d1′ is obtained; as follows,

[0140]

[0141] Among them, d1′ i1 d1′ i2 and d1′ i3 Respectively represent the inductance difference mean value data, capacitance difference mean value data, and differential impedance mean value data on the i-th differential signal line pair;

[0142] S22, for the differential signal line pair set a={a1, a2, ..., a i ,...,a a′ The signal delay data at each test point on each arranged differential signal line pair in} is measured to obtain a first signal delay data matrix d; as follows,

[0143]

[0144] Among them, d ij Represents the signal delay data at the jth test point in the i-th differential signal line pair;

[0145] Calculate the mean of the signal delay data at each test point in the first signal delay data matrix d to obtain the first signal delay data mean set d2′={d2′1, d2′2, ..., d2′ i ,...,d2′ a′}, d2′ i represents the mean value of signal delay data on the i-th differential signal line pair;

[0146] The first data mean matrix d1′ and the first signal delay data mean set d2′={d2′1, d2′2, ..., d2′ i ,...,d2′ a′} are merged to obtain the second data mean matrix d3′; as follows,

[0147]

[0148] Among them, d3′ i1 d3′ i2 d3′ i3 d3′ i4 Respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data on the i-th differential signal line;

[0149] S23, clustering the second data mean matrix d3′ to obtain the final differential signal line pair classification data matrix set represents the classification data matrix of the ith differential signal line pair obtained by clustering, Represents the total number of differential signal line pair classification data matrices obtained by clustering; as follows,

[0150]

[0151] in, They respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data on the jth differential signal line pair in the i-th differential signal line pair classification data matrix, represents the total number of differential signal line pairs in the i-th differential signal line pair classification data matrix;

[0152] The S23 comprises the following steps:

[0153] S231, randomly select a number of rows of data from the second data mean matrix d3′ to obtain an initial differential signal line pair classification center data matrix as follows,

[0154]

[0155] in, Respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data in the randomly selected i-th row of data;

[0156] S232, traverse the data in the second data mean matrix d3′ row by row, and calculate the initial differential signal line pair classification center data matrix for each row of data in the second data mean matrix d3′ The Euclidean distance between each row of data in the , and the initial Euclidean matrix is ​​obtained as follows,

[0157]

[0158] in, The second data mean matrix d3′ is the matrix of the classification center data of the initial differential signal line pair removed. The j-th row of data after the data in the initial differential signal line pair classification center data matrix The Euclidean distance between the i-th row of data;

[0159] Remove the initial differential signal line pair classification center data matrix from the second data mean matrix d3′ The data in the first row is classified into the initial Euclidean matrix. The initial differential signal line pair classification data matrix set is obtained from the classification center data corresponding to the smallest data in the jth column data. f iRepresents the i-th initial differential signal line pair classification data matrix; as follows,

[0160]

[0161] Among them, f ij1 、f ij2 、f ij3 、f ij4 They respectively represent the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data on the jth differential signal line pair in the i-th initial differential signal line pair classification data matrix; i ′ represents the total number of differential signal line pairs in the i-th initial differential signal line pair classification data matrix;

[0162] S233. Building a honey badger population represents the i-th honey badger individual in the honey badger population, g represents the size of the honey badger population; set the honey badger population The maximum number of iterations is g1′, the current number of iterations is g′2, and the number of search space dimensions is

[0163] According to the initial differential signal line pair classification center data matrix Set the initial position of each honey badger in the honey badger population and get the initial position matrix set Represents the initial position matrix of the i-th honey badger individual in the honey badger population; as follows,

[0164]

[0165] in, Respectively The position components of the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data in the classification center data of the j-th differential signal line pair;

[0166] In S233, the center data matrix is ​​classified according to the initial differential signal line pair. Set the initial position of each honey badger in the honey badger population and get the initial position matrix set The following steps are involved:

[0167] S2331, set the value range of the inductance difference mean data, the capacitance difference mean data, the differential impedance mean data and the signal delay mean data, and obtain the mean data value interval set Respectively represent the lower limits of the inductance difference mean data, the capacitance difference mean data, the differential impedance mean data and the signal delay mean data; Respectively represent the upper limits of the inductance difference mean data, the capacitance difference mean data, the differential impedance mean data, and the signal delay mean data;

[0168] S2332, based on the mean data value interval set

[0169] and the initial differential signal pair classification center data matrix set up The calculation formulas are as follows:

[0170]

[0171] In the formula, rand j1 、rand j2 、rand j3 、rand j4 For Generated random number; h(rand j1 )、h(rand j2 )、h(rand j3 )、h(rand j4 ) respectively represent the j1 、rand j2 、rand j3 、rand j4 The generated tent mapping values; the function expressions are as follows,

[0172]

[0173] Where, α1, α2, α3, and α4 are tent mapping parameters, with values ​​ranging from 0 to 1;

[0174] S234, according to the initial Euclidean matrix Set the fitness function h′ of the honey badger population as follows:

[0175]

[0176] In the formula, h i ′ represents the sum of the Euclidean distances from each row of the ith differential signal line pair classification data matrix in the differential signal line pair classification data matrix set obtained after each round of iteration to the classification center of the corresponding differential signal line pair classification data matrix;

[0177] S235, start iterative operation; in each round of iteration, calculate the fitness value of each honey badger according to the fitness function h′ of the honey badger population, select the honey badger individual with the largest fitness value as the global best individual; the position of the global best individual is used as the global best position; update the position of the honey badger population honey badger according to the global best position;

[0178] S236. When g′2≥g1′, stop iteration and output the final global optimal position matrix as follows,

[0179]

[0180] in, Respectively represent the position components of the inductance difference mean data, capacitance difference mean data, differential impedance mean data and signal delay mean data in the classified center data of the i-th differential signal line pair; otherwise, continue to iterate;

[0181] According to the final global optimal position matrix And the final differential signal line pair classification data matrix set is obtained by S232

[0182] S24, classifying the final differential signal line pair data matrix set Each differential signal line in the classification data matrix sets a differential data level label to obtain a differential data level label set e i Indicates the differential data level label corresponding to the i-th differential signal line pair classification data matrix;

[0183] S3, constructing an initial differential data level label data set according to the differential data level label set and the differential data level labels of the characteristic data pairs of the plurality of differential signal line pairs, and obtaining the differential data level label data set;

[0184] A final KNN classification model is constructed based on the characteristic data of multiple differential signal line pairs and the initial differential data level label data set;

[0185] The S3 comprises the following steps:

[0186] S31, according to the first initial length data set Calculate the differential signal line pair set a = {a1, a2, ..., a i ,...,a a′}, and obtain the first initial length difference data set represents the length difference between the two differential signal lines in the i-th differential signal line pair;

[0187] Calculate the standard deviation of each row of data in the first initial interval data matrix c to obtain the first initial interval standard deviation data set represents the standard deviation of the interval between the two differential signal lines in the ith differential signal line pair at each test point;

[0188] Select the first initial alignment angle data matrix The minimum value of each row of data in the first row is obtained to obtain the first initial minimum alignment angle data set. Indicates the minimum routing angle data on the i-th differential signal line pair;

[0189] S32, the first initial line width data set b′={b1′, b2′, ..., b i ′,...,b a ″}, the first initial length difference data set The first initial intensity data set c′={c1′,c′2,...,c i ′,...,c′ a′}、First distance dataset The first initial interval standard deviation data set The first initial minimum routing angle data set Merging the data to obtain a first initial comprehensive data matrix;

[0190] S33, setting the differential data level label of each row of data in the first initial comprehensive data matrix according to the differential data level label set and the final differential signal line pair classification data matrix set, to obtain an initial differential data level label data set;

[0191] S34, constructing an initial KNN classification model; using the initial differential data level label data set and the first initial comprehensive data matrix to perform training and testing operations on the initial KNN classification model; after the training and testing are completed, a final KNN classification model is obtained;

[0192] S4, acquiring feature data of the differential signal line pair on the security chip to be analyzed to obtain an initial comprehensive data set; using the final KNN classification model to classify the initial comprehensive data set to obtain second initial differential data level label data; and determining whether the initial comprehensive data set needs to be adjusted according to the second initial differential data level label data;

[0193] The S4 comprises the following steps:

[0194] S41, setting a security chip to be analyzed; collecting initial line width data of a differential signal line pair on the security chip to be analyzed, length difference data of two differential signal lines, initial signal strength data of surrounding signal sources, distance data from surrounding signal sources, standard deviation data of the interval between the two differential signal lines at each test point, and minimum routing angle data, to obtain an initial comprehensive data set;

[0195] S42, using the final KNN classification model to classify the initial comprehensive data set to obtain second initial differential data level label data m1′;

[0196] S43, setting a differential data level threshold m′2; when m1′<m′2, adjusting the initial comprehensive data set until m1′≥m′2; after the adjustment is completed, obtaining a final data mean set; otherwise, the initial comprehensive data set does not need to be adjusted; the initial comprehensive data set is used as the final data mean set;

[0197] In S43, the initial comprehensive data set is adjusted until m1′≥m′2; after the adjustment is completed, obtaining the final data mean set includes the following steps:

[0198] S431, setting the line width data of the differential signal line pair on the security chip to be analyzed after the signal is input, the length difference data of the two differential signal lines, the signal strength data of the surrounding signal source, the distance data to the surrounding signal source, the standard deviation data of the interval between the two differential signal lines at each test point, and the value interval of the minimum routing angle data, to obtain a first value interval set;

[0199] S432, starting a loop operation; updating the initial comprehensive data set in each loop process to obtain an updated second comprehensive data set;

[0200] S433, setting the updated second comprehensive data set to the differential signal line pair on the security chip to be analyzed after the signal is input, to obtain the adjusted differential signal line pair on the security chip to be analyzed;

[0201] S434, using the final KNN classification model to perform a classification operation on the updated second comprehensive data set to obtain third initial differential data level label data m3′; when m3′≥m′2, stop the loop and use the updated second comprehensive data set as the final data mean set; otherwise, loop until m3′≥m′2 to obtain the final data mean set;

[0202] S5. analyzing the differential errors of the security chip to be analyzed according to the result of adjusting the initial comprehensive data set;

[0203] The S5 comprises the following steps:

[0204] S51, calculating the difference degree value between the final data mean set and each data corresponding to the initial comprehensive data set to obtain a difference degree value set;

[0205] S52. Sort each difference degree value in the difference degree value set according to the numerical value; the data of the differential signal line pair corresponding to the maximum difference degree value is used as the first factor causing the differential error; the data of the differential signal line pairs corresponding to the other difference degree values ​​sorted from large to small are used as the second factor, the third factor, the fourth factor, the fifth factor and the sixth factor causing the differential error respectively.

[0206] Embodiment 2

[0207] This embodiment discloses a differential error analysis test system applied to a security chip, the system can implement the method of the above embodiment, including a first data acquisition module, a second data acquisition module, a clustering module, a label data construction module, a KNN classification model construction module, a third data acquisition module, a classification module, an adjustment module, a determination adjustment module and a differential error analysis module;

[0208] The first data acquisition module is used to collect characteristic data of multiple differential signal line pairs;

[0209] The second data acquisition module is used to collect inductance difference data, capacitance difference data, differential impedance data and signal delay data of multiple differential signal line pairs and calculate the average value to obtain a second data mean matrix;

[0210] The clustering module is used to perform a clustering operation on the second data mean matrix to obtain a differential data level label set;

[0211] The label data construction module is used to construct the differential data level labels of the characteristic data of multiple differential signal line pairs according to the differential data level label set and the characteristic data pairs of multiple differential signal line pairs to obtain an initial differential data level label data set;

[0212] The KNN classification model building module is used to build a final KNN classification model according to the feature data of multiple differential signal line pairs and the initial differential data level label data set;

[0213] The third data acquisition module is used to acquire characteristic data of the differential signal line pair on the security chip to be analyzed to obtain an initial comprehensive data set;

[0214] The classification module is used to classify the initial comprehensive data set using the final KNN classification model to obtain second initial differential data level label data;

[0215] The adjustment module is used to adjust the initial comprehensive data set;

[0216] The determination and adjustment module is used to determine whether it is necessary to call the adjustment module to adjust the initial comprehensive data set according to the second initial differential data level label data;

[0217] The differential error analysis module is used to analyze the differential errors of the security chip to be analyzed according to the result of the adjustment of the initial comprehensive data set by the adjustment module.

[0218] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0219] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A differential error analysis test method applied to a security chip, characterized in that: The following steps are involved: S1, collecting characteristic data of multiple differential signal line pairs; wherein a plurality of differential signal line pairs are set to obtain a differential signal line pair set; S2, collecting inductance difference data, capacitance difference data, differential impedance data and signal delay data of the plurality of differential signal line pairs and calculating average values ​​to obtain a second data mean matrix; performing a clustering operation on the second data mean matrix to obtain a differential data level label set; The S2 comprises the following steps: S21, measuring and calculating the signal data at each test point on each arranged differential signal line pair in the differential signal line pair set to obtain a first differential signal line pair data matrix; calculating the mean of the same type of signal data at each test point in the first differential signal line pair data matrix to obtain a first data mean matrix; S22, measuring the signal delay data at each test point on each arranged differential signal line pair in the differential signal line pair set to obtain a first signal delay data matrix; calculating the mean of the signal delay data at each test point in the first signal delay data matrix d to obtain a first signal delay data mean set; merging the first data mean matrix and the first signal delay data mean set to obtain a second data mean matrix; S23, performing a clustering operation on the second data mean matrix to obtain a final differential signal line pair classification data matrix set; The S23 comprises the following steps: S231, randomly selecting a number of rows of data from the second data mean matrix to obtain an initial differential signal line pair classification center data matrix; S232, calculating the Euclidean distance between each row of data in the second data mean matrix and each row of data in the initial differential signal line pair classification center data matrix, to obtain an initial Euclidean matrix; removing the data in the initial differential signal line pair classification center data matrix from the second data mean matrix, to obtain a second data mean matrix after removal; classifying the j-th row of data in the second data mean matrix after removal into the classification center data corresponding to the smallest data in the j-th column of data in the initial Euclidean matrix, to obtain an initial differential signal line pair classification data matrix set; S233, constructing a honey badger population; setting the maximum number of iterations of the honey badger population to g1′ and the current number of iterations to g′2; setting the initial position of each honey badger individual in the honey badger population according to the initial differential signal line pair classification center data matrix to obtain an initial position matrix set; S234, setting the fitness function of the honey badger population according to the initial Euclidean matrix; S235, start an iterative operation; in each round of iteration, calculate the fitness value of each honey badger according to the fitness function of the honey badger population and update the position of the honey badger in the honey badger population; S236, when g′2≥g1′, stop iteration and output the final global optimal position matrix; obtain the final differential signal line pair classification data matrix set according to the final global optimal position matrix and in the manner of S232; otherwise, continue iteration; S24, setting a differential data level label for each differential signal line pair classification data matrix in the final differential signal line pair classification data matrix set to obtain a differential data level label set; S3, constructing differential data level labels of multiple differential signal line pairs according to the differential data level label set and the characteristic data of multiple differential signal line pairs to obtain an initial differential data level label data set; constructing a final KNN classification model according to the characteristic data of multiple differential signal line pairs and the initial differential data level label data set; S4, acquiring feature data of the differential signal line pair on the security chip to be analyzed to obtain an initial comprehensive data set; using the final KNN classification model to classify the initial comprehensive data set to obtain second initial differential data level label data; and determining whether the initial comprehensive data set needs to be adjusted according to the second initial differential data level label data; S5. Analyze the differential errors of the security chip to be analyzed based on the result of adjusting the initial comprehensive data set.

2. The differential error analysis test method for a security chip according to claim 1, characterized in that: The S1 comprises the following steps: S11, collecting the length data and line width data of each differential signal line pair in the differential signal line pair set to obtain a first initial length data set and a first initial line width data set; setting the test point spacing; taking the signal input positions of the plurality of differential signal line pairs as the starting point, evenly setting a plurality of test point pairs according to the test point spacing to obtain a first test point matrix; S12, setting a signal source with adjustable strength and adjustable position around each differential signal line pair in the differential signal line pair set to obtain a first signal source set; setting an initial strength of each signal source in the first signal source set to obtain a first initial strength data set; setting a distance from each signal source in the first signal source set to the corresponding differential signal line pair to obtain a first distance data set; S13, arrange each differential signal line pair in the differential signal line pair set; after the arrangement is completed, measure the spacing distance between each test point on each differential signal line pair and the routing angle at each test point according to the first test point matrix to obtain a first initial spacing data matrix and a first initial routing angle data matrix.

3. The differential error analysis test method for a security chip according to claim 2, characterized in that: The S3 comprises the following steps: S31. Calculate the length difference between the two differential signal lines in each differential signal line pair in the differential signal line pair set according to the first initial length data set to obtain a first initial length difference data set; calculate the standard deviation of each row of data in the first initial interval data matrix to obtain a first initial interval standard deviation data set; select the minimum value of each row of data in the first initial routing angle data matrix to obtain a first initial minimum routing angle data set; S32, merging the first initial line width data set, the first initial length difference data set, the first initial intensity data set, the first distance data set, the first initial interval standard deviation data set, and the first initial minimum routing angle data set to obtain a first initial comprehensive data matrix; S33, setting the differential data level label of each row of data in the first initial comprehensive data matrix according to the differential data level label set and the final differential signal line pair classification data matrix set, to obtain an initial differential data level label data set; S34, constructing an initial KNN classification model; using the initial differential data level label data set and the first initial comprehensive data matrix to perform training and testing operations on the initial KNN classification model; after the training and testing are completed, a final KNN classification model is obtained.

4. The differential error analysis test method for a security chip according to claim 3, characterized in that: The S4 comprises the following steps: S41, setting a security chip to be analyzed; collecting initial line width data of a differential signal line pair on the security chip to be analyzed, length difference data of two differential signal lines, initial signal strength data of surrounding signal sources, distance data from surrounding signal sources, standard deviation data of the interval between the two differential signal lines at each test point, and minimum routing angle data, to obtain an initial comprehensive data set; S42, using the final KNN classification model to classify the initial comprehensive data set to obtain second initial differential data level label data m1′; S43, setting the differential data level threshold m′2; when m1′<m′2, adjusting the initial comprehensive data set until m1′≥m′2; after the adjustment is completed, obtaining the final data mean set; otherwise, there is no need to adjust the initial comprehensive data set, and the initial comprehensive data set is used as the final data mean set.

5. The differential error analysis test method for a security chip according to claim 4, characterized in that: In S43, the initial comprehensive data set is adjusted until m1′≥m′2; after the adjustment is completed, obtaining the final data mean set includes the following steps: S431, setting a value interval of each characteristic data of the differential signal line pair on the security chip to be analyzed after the signal is input, to obtain a first value interval set; S432, starting a loop operation; updating the initial comprehensive data set in each loop process to obtain an updated second comprehensive data set; S433. Use the final KNN classification model to perform classification operations on the updated second comprehensive data set to obtain the third initial differential data level label data m3′; when m3′≥m′2, stop the loop and use the updated second comprehensive data set as the final data mean set; otherwise, loop until m3′≥m′2 to obtain the final data mean set.

6. The differential error analysis test method for a security chip according to claim 5, characterized in that: The S5 comprises the following steps: S51, calculating the difference degree value between the final data mean set and each data corresponding to the initial comprehensive data set to obtain a difference degree value set; S52: Analyze the differential errors of the security chip to be analyzed according to the difference degree value set.

7. A system for implementing the differential error analysis test method applied to a security chip as claimed in any one of claims 1 to 6.

Images