A random number generator trustworthiness evaluation method and system

Abstract

The application discloses a kind of random number generator credibility evaluation method and system, including obtaining the expert evaluation score and metrology evaluation score of random number generator, randomness test significance level, sample threshold value passes;If flexible content is selected to customize, input credibility evaluation comprehensive score calculation formula and upload credibility evaluation report format file, otherwise according to system default formula and format processing;Random sequence data is subjected to basic test and improved test, whether random sequence passes test is judged according to significance level, whether random sequence sample data file passes test is judged according to sample threshold value passes, record the reason of not passing, and the randomness test score is calculated;Receive randomness test score, expert evaluation score, metrology evaluation score;According to the comprehensive score of random number generator credibility evaluation calculated by self-defined formula or default formula;Receive comprehensive score and not pass reason, and output evaluation report according to content.

Description

Technical Field

[0001] This invention relates to the field of random number evaluation, and more particularly to a method and system for evaluating the reliability of a random number generator. Background Technology

[0002] Random number generators, as an important tool in modern technology, have a profound impact on their application effectiveness in numerous fields, including key applications such as public interest allocation, complex system simulation, financial risk assessment, and complex key generation. Effectively evaluating the reliability of random number generators can identify potential biases and shortcomings, guide their optimization and improvement, enhance their reliability and transparency in specific scenarios, and promote the healthy development of the market.

[0003] Assessing the reliability of a random number generator should include evaluating the randomness of its generated results, assessing the physicochemical properties of the generator, and conducting an expert evaluation of the generator's design compliance. Currently, most theoretical methods and tools for evaluating the randomness of random number generator results are only applicable to binary random number sequences, lacking methods and tools capable of directly testing random sequences in non-binary complex sample spaces. Furthermore, due to theoretical and technical complexity, existing methods for testing the randomness of non-binary random sequences often rely on binary transformations, which may introduce information loss and testing bias, failing to comprehensively assess the randomness of such generators. Existing assessments do not combine the evaluation of the randomness of the generated results with the quantitative assessment of the generator's physicochemical properties and the expert evaluation of its design compliance, thus lacking a comprehensive assessment of the random number generator's reliability.

[0004] Therefore, it is urgent to study a comprehensive evaluation method for the reliability of random number generators in non-binary complex sample spaces. Addressing the current state of technology, this invention, based on existing testing methods, designs a testing method that can directly test the randomness of non-binary sequences. Furthermore, it combines randomness testing with quantitative evaluation of the physicochemical properties of the random number generator and expert assessment of the conformity of its principle design, proposing a method and system for evaluating the reliability of random number generators. Summary of the Invention

[0005] The purpose of this invention is to provide a method for evaluating the reliability of random number generators.

[0006] To achieve the above objectives, the present invention is implemented according to the following technical solution:

[0007] This invention includes the following steps:

[0008] S1 retrieves the expert evaluation score and quantitative evaluation score of the random number generator, the significance level of the randomness test, and the sample pass threshold; if the flexible content is selected to be customized, then enter the formula for calculating the comprehensive credibility assessment score and upload the credibility assessment report format file; otherwise, it will be processed according to the system's default formula and format.

[0009] S2 receives a random sequence sample data file, a random sequence sample point file, a significance level for randomness testing, a sample pass threshold, information on whether the random sequence sample points are repeatable, and information on whether a custom comprehensive score calculation formula is used. It performs basic and improved tests on the random sequence data, determines whether the random sequence passes the test based on the significance level, determines whether the random sequence sample data file passes the test based on the sample pass threshold, records the reasons for failure, and calculates the randomness test score based on whether a custom comprehensive score calculation formula is used.

[0010] S3 receives randomness test scores, expert evaluation scores, and quantitative evaluation scores; it calculates a comprehensive score for the reliability assessment of the random number generator based on a user-defined formula or a default formula.

[0011] S4 receives the overall score and the reason for failure, and outputs an evaluation report based on the content.

[0012] Furthermore, the flexible content includes uploading random sequence sample data files and random sequence sample point files; selecting whether random sequence sample points can be repeated, whether to customize the comprehensive score calculation formula for credibility assessment, and whether to customize the credibility assessment report format.

[0013] Furthermore, S1 also includes:

[0014] If the system's default reliability assessment comprehensive score calculation formula is used, the expert evaluation score and the metrological evaluation score entered in step S1 can only be 0 or 1; an expert evaluation score of 1 indicates that the expert, based on the theoretical design principles, reviews the actual design of the random number generator and judges whether the actual design is consistent with the theoretical principles, and the final conclusion is that it passes, while 0 indicates that the conclusion is that it fails; a metrological evaluation score of 1 indicates that the final conclusion obtained after measuring the physical and chemical parameters of the random number generator through metrological means and comparing them with the actual required indicators is that it is qualified, while 0 indicates that the conclusion is that it is unqualified.

[0015] If a custom credibility assessment comprehensive score calculation formula is used, there are no restrictions on the input of expert evaluation scores and quantitative evaluation scores;

[0016] The preferred significance level for the randomness test is 0.05 or 0.01. The preferred sample threshold is calculated as follows:

[0017]

[0018] Where is the threshold for samples to pass. At the significance level, The number of samples in the random sequence sample data file;

[0019] Input parameters and file. The system checks the random sequence sample data file based on the random sequence sample point file and the information on whether the random sequence sample points are repeatable. This includes: whether the random sequence sample data file contains sample points other than those in the sample point file, and whether the repeatability information of the sample points in the random sequence conflicts with the input repeatability information. When the random sequence sample data file is found to be abnormal, the system outputs the reason for the abnormality and prompts the user to re-upload the file.

[0020] Furthermore, the method for performing basic and improved tests on random sequence data includes:

[0021] If the sample points can be repeated, the basic test includes 7 subtests, which test the uniformity of the random sequence data, the correlation of sample points, the correlation of sequences, the mean of sample points, the mean of sequences, the variance of sample points, and the variance of sequences.

[0022] If the sample points cannot be repeated, the basic test includes five subtests, which test the uniformity of the random sequence data, the mean of the sample points, the mean of the sequence, the variance of the sample points, and the variance of the sequence, respectively.

[0023] If a random sequence fails at least one of the minor tests in the basic tests, record the reason for failure, and directly determine the randomness test score as 0, then proceed to the comprehensive score calculation module.

[0024] If a random sequence passes all the subtests in the basic test, the basic test score is 1, and 14 improved tests are performed.

[0025] If the random sequence sample space has only two sample points, then each improved test is directly performed on the random sequence, and the threshold is used to determine whether each test passes.

[0026] If the random sequence sample space contains more than two sample points, each improved test includes three sub-tests: direct testing and testing after interpolation using two interpolation methods. The test using the interpolation method needs to be run three times. If at least one sub-test fails, the sub-test is deemed to have failed. If at least one sub-test fails for each improved test, the test is deemed to have failed.

[0027] Each test is scored as 1 point if it passes, and 0 points if it fails, with the reason for failure recorded.

[0028] If a custom comprehensive score calculation formula has been defined, the randomness test score is calculated as the sum of all test scores. That is, when all 15 tests are passed, the randomness test score is 15.

[0029] If no custom formula is defined for calculating the overall score, the formula for calculating the randomness test score is as follows:

[0030]

[0031] in The randomness test score, For the sample test, the first The score of each method For coefficients, when hour, ,otherwise, ;

[0032] After obtaining the randomness test score, you will be redirected to the comprehensive score calculation module.

[0033] Furthermore, the method for obtaining the comprehensive score includes:

[0034] If a custom formula for calculating the overall score has been defined, the overall score will be calculated according to the defined formula.

[0035] If no custom formula is defined for calculating the overall score, the overall score will be calculated using the following formula:

[0036]

[0037] in The overall score for the reliability evaluation of the random number generator. For expert evaluation scores, For measurement and evaluation scores, The randomness test score;

[0038] After obtaining the overall score, you will be redirected to the assessment result report generation module.

[0039] Furthermore, the method for outputting the evaluation report includes:

[0040] If a custom formula for calculating the overall score has been defined, the reliability evaluation result of the random number generator will include the overall score and the reasons for any failures in the verification.

[0041] If no custom formula is defined for calculating the overall score, the reliability evaluation result of the random number generator includes the overall score, the evaluation conclusion, and the reasons for any failures in the verification.

[0042] If a custom evaluation report format has been defined, the final evaluation report with the reliability evaluation results of the random number generator will be output according to the custom format.

[0043] If no custom evaluation report format is defined, the final evaluation report with the reliability evaluation results of the random number generator will be output in the system default format.

[0044] Secondly, a system for evaluating the reliability of a random number generator includes:

[0045] Input module inputs: Expert evaluation score and quantitative evaluation score of the random number generator, significance level of randomness test, and sample pass threshold; upload random sequence sample data file and random sequence sample point file; select whether the random sequence sample points are repeatable, whether to customize the comprehensive confidence assessment score calculation formula, and whether to customize the confidence assessment report format; if customized, input the comprehensive confidence assessment score calculation formula and upload the confidence assessment report format file; otherwise, the system default formula and format will be used.

[0046] Randomness testing module: This module receives random sequence sample data files, random sequence sample point files, randomness test significance level, sample pass threshold, information on whether random sequence sample points are repeatable, and information on whether a custom comprehensive score calculation formula is used. It performs basic and improved tests on the random sequence data, determines whether the random sequence passes the test based on the significance level, determines whether the random sequence sample data file passes the test based on the sample pass threshold, records the reasons for failure, and calculates the randomness test score based on whether a custom comprehensive score calculation formula is used.

[0047] The comprehensive score calculation module receives randomness test scores, expert evaluation scores, and quantitative evaluation scores; and calculates the comprehensive score for the reliability assessment of the random number generator based on a custom formula or a default formula.

[0048] Evaluation Result Report Generation Module: This module receives the overall score and reasons for failure, and outputs an evaluation report based on the content.

[0049] The beneficial effects of this invention are:

[0050] This invention relates to a method and system for evaluating the reliability of random number generators. Compared with existing technologies, this invention has the following technical advantages:

[0051] This invention, through the steps of calculating a comprehensive credibility assessment score, obtaining a credibility assessment report format file, performing basic and improved tests, obtaining a randomness test score, obtaining a comprehensive score, and outputting an assessment report, can directly test random sequences in non-binary complex sample spaces. This avoids information loss and testing bias caused by binary transformations, as existing technologies mostly target binary sequences or rely on binary transformations. Furthermore, this invention combines randomness testing, quantitative evaluation of physicochemical properties, and expert evaluation of principle design compliance—a comprehensive assessment lacking in existing technologies. It supports custom formulas and report formats, adapting to different needs and improving the applicability and credibility of the assessment. Attached Figure Description

[0052] Figure 1This is a flowchart illustrating the steps of a random number generator reliability evaluation method according to the present invention.

[0053] Figure 2 This is a flowchart of the module of a random number generator reliability evaluation system according to the present invention;

[0054] Figure 3 This is a schematic diagram of the evaluation process for a specific embodiment of the present invention. Detailed Implementation

[0055] The present invention will be further described below through specific embodiments. The illustrative embodiments and descriptions herein are used to explain the present invention, but are not intended to limit the present invention.

[0056] The present invention provides a method and system for evaluating the reliability of a random number generator, comprising the following steps:

[0057] like Figure 1 As shown, this embodiment includes the following steps:

[0058] S1 retrieves the expert evaluation score and quantitative evaluation score of the random number generator, the significance level of the randomness test, and the sample pass threshold; if the flexible content is selected to be customized, then enter the formula for calculating the comprehensive credibility assessment score and upload the credibility assessment report format file; otherwise, it will be processed according to the system's default formula and format.

[0059] S2 receives a random sequence sample data file, a random sequence sample point file, a significance level for randomness testing, a sample pass threshold, information on whether the random sequence sample points are repeatable, and information on whether a custom comprehensive score calculation formula is used. It performs basic and improved tests on the random sequence data, determines whether the random sequence passes the test based on the significance level, determines whether the random sequence sample data file passes the test based on the sample pass threshold, records the reasons for failure, and calculates the randomness test score based on whether a custom comprehensive score calculation formula is used.

[0060] S3 receives randomness test scores, expert evaluation scores, and quantitative evaluation scores; it calculates a comprehensive score for the reliability assessment of the random number generator based on a user-defined formula or a default formula.

[0061] S4 receives the overall score and the reason for failure, and outputs an evaluation report based on the content.

[0062] In this embodiment, the flexible content includes uploading random sequence sample data files and random sequence sample point files; selecting whether random sequence sample points can be repeated, whether to customize the comprehensive score calculation formula for credibility assessment, and whether to customize the credibility assessment report format.

[0063] In this embodiment, S1 further includes:

[0064] If the system's default reliability assessment comprehensive score calculation formula is used, the expert evaluation score and the metrological evaluation score entered in step S1 can only be 0 or 1; an expert evaluation score of 1 indicates that the expert, based on the theoretical design principles, reviews the actual design of the random number generator and judges whether the actual design is consistent with the theoretical principles, and the final conclusion is that it passes, while 0 indicates that the conclusion is that it fails; a metrological evaluation score of 1 indicates that the final conclusion obtained after measuring the physical and chemical parameters of the random number generator through metrological means and comparing them with the actual required indicators is that it is qualified, while 0 indicates that the conclusion is that it is unqualified.

[0065] If a custom credibility assessment comprehensive score calculation formula is used, there are no restrictions on the input of expert evaluation scores and quantitative evaluation scores;

[0066] The preferred significance level for the randomness test is 0.05 or 0.01. The preferred sample threshold is calculated as follows:

[0067]

[0068] in For samples to pass the threshold, At the significance level, The number of samples in the random sequence sample data file;

[0069] Input parameters and file. The system checks the random sequence sample data file based on the random sequence sample point file and the information on whether the random sequence sample points are repeatable. This includes: whether the random sequence sample data file contains sample points other than those in the sample point file, and whether the repeatability information of the sample points in the random sequence conflicts with the input repeatability information. When the random sequence sample data file is found to be abnormal, the system outputs the reason for the abnormality and prompts the user to re-upload the file.

[0070] In actual evaluation, there are no restrictions on the input of expert evaluation scores and quantitative evaluation scores: this means that the specific scale is determined by the definer; the input parameters and files can be relevant parameters and files of binary random sequences.

[0071] In this embodiment, the method for performing basic and improved tests on random sequence data includes:

[0072] If the sample points can be repeated, the basic test includes 7 subtests, which test the uniformity of the random sequence data, the correlation of sample points, the correlation of sequences, the mean of sample points, the mean of sequences, the variance of sample points, and the variance of sequences.

[0073] If the sample points cannot be repeated, the basic test includes five subtests, which test the uniformity of the random sequence data, the mean of the sample points, the mean of the sequence, the variance of the sample points, and the variance of the sequence, respectively.

[0074] If a random sequence fails at least one of the minor tests in the basic tests, record the reason for failure, and directly determine the randomness test score as 0, then proceed to the comprehensive score calculation module.

[0075] If a random sequence passes all the subtests in the basic test, the basic test score is 1, and 14 improved tests are performed.

[0076] If the random sequence sample space has only two sample points, then each improved test is directly performed on the random sequence, and the threshold is used to determine whether each test passes.

[0077] If the random sequence sample space contains more than two sample points, each improved test includes three sub-tests: direct testing and testing after interpolation using two interpolation methods. The test using the interpolation method needs to be run three times. If at least one sub-test fails, the sub-test is deemed to have failed. If at least one sub-test fails for each improved test, the test is deemed to have failed.

[0078] Each test is scored as 1 point if it passes, and 0 points if it fails, with the reason for failure recorded.

[0079] If a custom comprehensive score calculation formula has been defined, the randomness test score is calculated as the sum of all test scores. That is, when all 15 tests are passed, the randomness test score is 15.

[0080] If no custom formula is defined for calculating the overall score, the formula for calculating the randomness test score is as follows:

[0081]

[0082] in The randomness test score, For the sample test, the first The score of each method For coefficients, when hour, ,otherwise, ;

[0083] After obtaining the randomness test score, you will be redirected to the comprehensive score calculation module.

[0084] In this embodiment, the method for obtaining the comprehensive score includes:

[0085] If a custom formula for calculating the overall score has been defined, the overall score will be calculated according to the defined formula.

[0086] If no custom formula is defined for calculating the overall score, the overall score will be calculated using the following formula:

[0087]

[0088] in The overall score for the reliability evaluation of the random number generator. For expert evaluation scores, For measurement and evaluation scores, The randomness test score;

[0089] After obtaining the overall score, you will be redirected to the assessment result report generation module.

[0090] In this embodiment, the method for outputting the evaluation report includes:

[0091] If a custom formula for calculating the overall score has been defined, the reliability evaluation result of the random number generator will include the overall score and the reasons for any failures in the verification.

[0092] If no custom formula is defined for calculating the overall score, the reliability evaluation results of the random number generator include the overall score, the evaluation conclusion, and the reasons for any failed tests. The grading rules for the evaluation conclusion are shown in Table 1 below:

[0093] Table 1. Grading Rules for Evaluation Conclusions

[0094]

[0095] If a custom evaluation report format has been defined, the final evaluation report with the reliability evaluation results of the random number generator will be output according to the custom format.

[0096] If no custom evaluation report format is defined, the final evaluation report with the reliability evaluation results of the random number generator will be output in the system default format.

[0097] Secondly, a system for evaluating the reliability of a random number generator includes:

[0098] Input module inputs: Expert evaluation score and quantitative evaluation score of the random number generator, significance level of randomness test, and sample pass threshold; upload random sequence sample data file and random sequence sample point file; select whether the random sequence sample points are repeatable, whether to customize the comprehensive confidence assessment score calculation formula, and whether to customize the confidence assessment report format; if customized, input the comprehensive confidence assessment score calculation formula and upload the confidence assessment report format file; otherwise, the system default formula and format will be used.

[0099] Randomness testing module: This module receives random sequence sample data files, random sequence sample point files, randomness test significance level, sample pass threshold, information on whether random sequence sample points are repeatable, and information on whether a custom comprehensive score calculation formula is used. It performs basic and improved tests on the random sequence data, determines whether the random sequence passes the test based on the significance level, determines whether the random sequence sample data file passes the test based on the sample pass threshold, records the reasons for failure, and calculates the randomness test score based on whether a custom comprehensive score calculation formula is used.

[0100] The comprehensive score calculation module receives randomness test scores, expert evaluation scores, and quantitative evaluation scores; and calculates the comprehensive score for the reliability assessment of the random number generator based on a custom formula or a default formula.

[0101] Evaluation Result Report Generation Module: This module receives the overall score and reasons for failure, and outputs an evaluation report based on the content.

[0102] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for evaluating the reliability of a random number generator, characterized in that, Includes the following steps: S1 retrieves the expert evaluation score and quantitative evaluation score of the random number generator, the significance level of the randomness test, and the sample pass threshold; if the flexible content is selected to be customized, then enter the formula for calculating the comprehensive credibility assessment score and upload the credibility assessment report format file; otherwise, it will be processed according to the system's default formula and format. S2 receives the random sequence sample data file, random sequence sample point file, randomness test significance level, sample pass threshold, information on whether the random sequence sample points are repeatable, and information on whether a custom comprehensive score calculation formula is available. Perform basic and improved tests on random sequence data, determine whether the random sequence passes the test based on the significance level, determine whether the random sequence sample data file passes the test based on the sample pass threshold, record the reasons for failure, and calculate the randomness test score based on whether a custom comprehensive score calculation formula is used. include: If the sample points can be repeated, the basic test includes 7 subtests, which test the uniformity of the random sequence data, the correlation of sample points, the correlation of sequences, the mean of sample points, the mean of sequences, the variance of sample points, and the variance of sequences. If the sample points cannot be repeated, the basic test includes five subtests, which test the uniformity of the random sequence data, the mean of the sample points, the mean of the sequence, the variance of the sample points, and the variance of the sequence, respectively. If a random sequence fails at least one of the minor tests in the basic tests, record the reason for failure, and directly determine the randomness test score as 0, then proceed to the comprehensive score calculation module. If a random sequence passes all the subtests in the basic test, the basic test score is 1, and 14 improved tests are performed. If the random sequence sample space has only two sample points, then each improved test is directly performed on the random sequence, and the threshold is used to determine whether each test passes. If the random sequence sample space contains more than two sample points, each improved test includes three sub-tests: direct testing and testing after interpolation using two interpolation methods. The test using the interpolation method needs to be run three times. If at least one sub-test fails, the sub-test is deemed to have failed. If at least one sub-test fails for each improved test, the test is deemed to have failed. Each test is scored as 1 point if it passes, and 0 points if it fails, with the reason for failure recorded. If a custom comprehensive score calculation formula has been defined, the randomness test score is calculated as the sum of all test scores. That is, when all 15 tests are passed, the randomness test score is 15. If no custom formula is defined for calculating the overall score, the formula for calculating the randomness test score is as follows: in The randomness test score, For the sample test, the first The score of each method For coefficients, when hour, ,otherwise, ; After obtaining the randomness test score, you will be redirected to the comprehensive score calculation module; S3 receives randomness test scores, expert evaluation scores, and quantitative evaluation scores; it calculates a comprehensive score for the reliability assessment of the random number generator based on a user-defined formula or a default formula. S4 receives the overall score and the reason for failure, and outputs an evaluation report based on the content.

2. The method for evaluating the reliability of a random number generator according to claim 1, characterized in that, The flexible features include uploading random sequence sample data files and random sequence sample point files; selecting whether random sequence sample points can be repeated; whether to customize the comprehensive score calculation formula for credibility assessment; and whether to customize the credibility assessment report format.

3. The method for evaluating the reliability of a random number generator according to claim 1, characterized in that, S1 further includes: If the system's default reliability assessment comprehensive score calculation formula is used, the expert evaluation score and the metrological evaluation score entered in step S1 can only be 0 or 1; an expert evaluation score of 1 indicates that the expert, based on the theoretical design principles, reviews the actual design of the random number generator and judges whether the actual design is consistent with the theoretical principles, and the final conclusion is that it passes, while 0 indicates that the conclusion is that it fails; a metrological evaluation score of 1 indicates that the final conclusion obtained after measuring the physical and chemical parameters of the random number generator through metrological means and comparing them with the actual required indicators is that it is qualified, while 0 indicates that the conclusion is that it is unqualified. If a custom credibility assessment comprehensive score calculation formula is used, there are no restrictions on the input of expert evaluation scores and quantitative evaluation scores; The significance level for the randomness test is 0.05 or 0.

01. The threshold for the sample to pass is calculated as follows: in For samples to pass the threshold, At the significance level, The number of samples in the random sequence sample data file; Input parameters and file. The system checks the random sequence sample data file based on the random sequence sample point file and the information on whether the random sequence sample points are repeatable. This includes: whether the random sequence sample data file contains sample points other than those in the sample point file, and whether the repeatability information of the sample points in the random sequence conflicts with the input repeatability information. When the random sequence sample data file is found to be abnormal, the system outputs the reason for the abnormality and prompts the user to re-upload the file.

4. The method for evaluating the reliability of a random number generator according to claim 1, characterized in that, The method for obtaining the comprehensive score includes: If a custom formula for calculating the overall score has been defined, the overall score will be calculated according to the defined formula. If no custom formula is defined for calculating the overall score, the overall score will be calculated using the following formula: in The overall score for the reliability evaluation of the random number generator. For expert evaluation scores, For measurement and evaluation scores, The randomness test score; After obtaining the overall score, you will be redirected to the assessment result report generation module.

5. The method for evaluating the reliability of a random number generator according to claim 1, characterized in that, The method for outputting the evaluation report includes: If a custom formula for calculating the overall score has been defined, the reliability evaluation result of the random number generator will include the overall score and the reasons for any failures in the verification. If no custom formula is defined for calculating the overall score, the reliability evaluation result of the random number generator includes the overall score, the evaluation conclusion, and the reasons for any failures in the verification. If a custom evaluation report format has been defined, the final evaluation report with the reliability evaluation results of the random number generator will be output according to the custom format. If no custom evaluation report format is defined, the final evaluation report with the reliability evaluation results of the random number generator will be output in the system default format.

6. A random number generator reliability evaluation system, used to perform the method according to any one of claims 1-5, characterized in that, include: Input module inputs: expert evaluation score and quantitative evaluation score of the random number generator, significance level of randomness test, and sample pass threshold; Upload the random sequence sample data file and the random sequence sample point file; Choose whether the random sequence sample points can be repeated, whether to customize the comprehensive score calculation formula for the credibility assessment, and whether to customize the credibility assessment report format; if you choose to customize, enter the comprehensive score calculation formula for the credibility assessment and upload the credibility assessment report format file; otherwise, the system default formula and format will be used. Randomness test module: used to receive random sequence sample data files, random sequence sample point files, randomness test significance level, sample pass threshold, information on whether random sequence sample points are repeatable, and information on whether to define a custom comprehensive score calculation formula; Perform basic and improved tests on random sequence data, determine whether the random sequence passes the test based on the significance level, determine whether the random sequence sample data file passes the test based on the sample pass threshold, record the reasons for failure, and calculate the randomness test score based on whether a custom comprehensive score calculation formula is used. The comprehensive score calculation module receives randomness test scores, expert evaluation scores, and quantitative evaluation scores; and calculates the comprehensive score for the reliability assessment of the random number generator based on a custom formula or a default formula. Evaluation Result Report Generation Module: This module receives the overall score and reasons for failure, and outputs an evaluation report based on the content.

Images