JP2012141774A - Pseudo random number generation device, pseudo random number generation method and its program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a pseudo random number generation device, a pseudo random number generation method and its program for generating pseudo random numbers whose strength can be set from discrete array data.SOLUTION: The pseudo random number generation device includes: a base array data sequence conversion part 102 for defining a part of discrete array data as array data for library and the other part as data for random number determination, and for assigning a sequence with predetermined bits to each element to convert the data into a numerical sequence, and for generating an array data sequence for library and a data sequence for random number determination; a non-linear time series analysis part 103 for generating prediction data on the basis of deterministic non-linear prediction by using the array data sequence for library; and a random number generation part 105 for comparing the prediction data with the data sequence for random number determination, and for, when a non-random state continues by the predetermined number of times, evaluating that randomness has been lost with respect to the whole discrete array data, and for, otherwise, successively sending the data sequence for random number determination to the outside as the candidate of the pseudo random number sequence.

Description

  The present invention relates to a pseudo-random number generation device, a pseudo-random number generation method, and a program thereof that generate pseudo-random numbers whose strengths can be set from discrete array data.

  Conventionally, pseudo-random numbers are used for various purposes, and the use of encryption and authentication is typically known. Many conventional pseudorandom number generation techniques employ a deterministic prediction method using, for example, a linear feedback law or a nonlinear feedback law.

  In general, as the number of bits of the required pseudorandom number sequence increases, the necessary number of components for generating the pseudorandom number also increases. As a result, the generated pseudorandom number sequence is limited, and the random number performance ( It is known that the random strength is also limited. As a technique for solving this problem, a plurality of pseudo random number generation circuits composed of S stages each generate pseudo random number data of N bits (N is an integer of 2 or more), and from each of the S stage pseudo random number generation circuits. A technique is known that receives pseudo-random data (X1 to X4) and outputs a combination of arbitrary pseudo-random data among the received S pseudo-random data as a pseudo-random number sequence (see, for example, Patent Document 1).

  A technique is also known in which prediction data is generated based on deterministic nonlinear prediction for the generated pseudorandom number sequence, and the randomness of the pseudorandom number sequence is evaluated (see, for example, Patent Document 2).

JP 2010-2959 A Japanese Patent No. 3658623

  Currently, random numbers used in computer programs and the like are called pseudo-random numbers, and are known to have a certain degree of period and distribution bias. In addition, as described above, there is a problem that the required number of pseudo random number generation components increases as the number of bits of the pseudo random number sequence required for generating the pseudo random number increases. There is a problem that the pseudo-random number sequence is limited, and the random number performance (random number strength) is also limited. Furthermore, there are many methods for evaluating the random number sequence of pseudo-random numbers, and it is difficult to determine which is the best, and there are advantages and disadvantages to each method, but the generated pseudo-random number sequence There is a demand for a technique for generating a pseudo-random number while evaluating a random number sequence without limiting the random number performance (random number strength).

  Accordingly, the present invention has been made in view of the above-described problems, and provides a pseudo-random number generation device, a pseudo-random number generation method, and a program thereof that generate pseudo-random numbers whose strengths can be set from discrete array data. Objective.

In the present invention, pseudo-random numbers are evaluated while evaluating a random number sequence by using discrete array data prepared in advance, without restricting the generated pseudo-random number sequence and without limiting the random number performance (random number strength). Is generated. In particular, this discrete sequence data includes nucleobase DNA (including cDNA, cpDNA, gDNA, msDNA, mtDNA, hereinafter referred to as DNA) or RNA (mRNA (pre-mRNA / hnRNA), tRNA, rRNA, aRNA, Excellent base sequence data (hereinafter referred to as “base sequence data”) of gRNA, miRNA, ncRNA, piRNA, shRNA, siRNA, stRNA, snRNA, snoRNA, stRNA, ta-siRNA, tmRNA, and hereinafter referred to as RNA). A pseudo-random number sequence can be generated. For example, by using naturally occurring sequences such as base sequences stored in the database of base sequence data banks on the Internet (eg DDBJ: DNA Data Bank of Japan), periodicity and distribution bias While solving the problem, a pseudo-random number sequence is generated by extracting a numerical sequence of arbitrary strength from these base sequences.

  That is, the pseudo-random number generation device of the present invention is a pseudo-random number generation device that generates pseudo-random numbers from discrete array data, wherein a part of the multiple-element discrete array data is used as library array data, and the discrete Assigning a sequence of predetermined bits to each element in the library array data and the random number determination data, using another part of the target array data different from the library array data as the random number determination data By using the data sequence conversion unit that converts the sequence data sequence for the library and the random number determination data sequence, and the library sequence data sequence, the prediction data predicted based on the deterministic nonlinear prediction The non-linear time series analysis unit to be generated is compared with the prediction data and the random number determination data sequence, and the non-random state Is evaluated as randomness is lost with respect to the entire discrete array data, and otherwise, the random number determination data number sequence is set as a pseudo random number sequence candidate. And a random number generation unit that sequentially sends the pseudo-random number sequence candidates to the outside as pseudo-random numbers.

  In the pseudorandom number generator of the present invention, the discrete sequence data is composed of base sequence data.

  The pseudorandom number generator according to the present invention further includes a base sequence data download request unit that downloads base sequence data that is the discrete sequence data from a genome database deployed on the Internet through a network. To do.

  Further, in the pseudo random number generation device of the present invention, the random number predictability determination unit determines that the randomness is lost with respect to the entire discrete array data from the comparison of the prediction data and the random number determination data sequence. If evaluated, the base sequence data download request unit has an instruction to download another base sequence data from the genome database, and the base sequence data download request unit includes the random number predictability determination unit. In accordance with an instruction from the genome database, means for downloading another base sequence data from the genome database.

  Furthermore, the pseudo-random number generation method of the present invention is a pseudo-random number generation method for generating pseudo-random numbers from discrete array data, wherein a part of the multiple-element discrete array data is used as library array data, and the discrete Assigning a sequence of predetermined bits to each element in the library array data and the random number determination data, using another part of the target array data different from the library array data as the random number determination data The step of generating a sequence data sequence for library and a sequence number for determining random number data, and the step of generating prediction data predicted based on deterministic nonlinear prediction using the sequence data sequence for library data And comparing the prediction data with the random number determination data sequence, and the non-random state continues for a predetermined number of times. And evaluating the randomness of the entire discrete array data as lost, otherwise, setting the random number determination data sequence as a pseudo random sequence candidate, and the pseudo random sequence candidate Are sequentially transmitted to the outside as pseudorandom numbers.

  Further, the program of the present invention provides a computer configured as a pseudo-random number generator that generates pseudo-random numbers from discrete array data, wherein a part of the discrete array data of a plurality of elements is used as library array data, and the discrete By assigning a sequence of predetermined bits to each element in the library array data and the random number determination data, using another part of the array data different from the library array data as random number determination data Converting to a numerical sequence, generating a sequence data sequence for library and a data sequence for random number determination, and generating predicted data predicted based on deterministic nonlinear prediction using the sequence data sequence for library And comparing the prediction data and the random number determination data sequence, the non-random state is When the fixed number of consecutive times, it is evaluated that the randomness is lost with respect to the entire discrete array data, otherwise, the step of setting the random number determination data sequence as a pseudo random number sequence candidate, and the pseudo random number And a step of sequentially sending sequence candidates to the outside as pseudo-random numbers.

  According to the present invention, a pseudo-random number sequence is generated by extracting a numerical sequence of arbitrary strength from these base sequences, so that the generated pseudo-random number sequence is not limited, and random number performance (random strength) Without being restricted, it becomes possible to generate a pseudo-random number sequence while solving problems such as periodicity and uneven distribution.

It is a figure which shows the example of a system which produces | generates a pseudorandom number using the pseudorandom number generator of one Example by this invention. 1 is a block diagram of a pseudo random number generator according to an embodiment of the present invention. FIG. It is an operation | movement flowchart of the pseudorandom number generator of one Example by this invention. It is a figure which shows an example of the base sequence data used with the pseudorandom number generator of one Example by this invention. It is a figure which shows an example of the pseudorandom number sequence produced | generated with the pseudorandom number generation apparatus of one Example by this invention. It is a figure which shows the autocorrelation function of the pseudorandom number sequence produced | generated with the pseudorandom number generation apparatus of one Example by this invention. It is a figure which shows the histogram of the pseudorandom number sequence produced | generated with the pseudorandom number generation apparatus of one Example by this invention. It is a figure which shows the analysis example of the histogram of the pseudorandom number sequence produced | generated with the pseudorandom number generation apparatus of one Example by this invention.

  Hereinafter, a pseudo random number generation apparatus and method according to an embodiment of the present invention will be described.

〔System configuration〕
FIG. 1 is a diagram illustrating an example of a system that generates pseudo-random numbers using the pseudo-random number generator according to an embodiment of the present invention. The pseudorandom number generator 10 of the present embodiment can be realized using a personal computer (details will be described later). The pseudo-random number generation device 10 downloads naturally existing discrete sequence data D such as already accumulated base sequence data from the genome database 20 deployed on the Internet, and this sequence data D Is converted into a numerical sequence, and the randomness is tested regularly or periodically using a nonlinear time series analysis program (deterministic nonlinear prediction method) as disclosed in Patent Document 2, for example. The part having Regarding the determination of randomness, a part of the actual sequence data D obtained by downloading is set as library data (hereinafter referred to as “library sequence data”), and a sequence of the sequence data for the library Of the array data (hereinafter referred to as “random number determination data”) sequentially generated for random number determination in other portions of the actual array data D by using the deterministic nonlinear prediction method Compare sequences.

  Further, the pseudo-random number generator 10 determines that the non-random state continues M times from the comparison result between the prediction data predicted by the deterministic nonlinear prediction method and the random number determination data sequence sequentially extracted from the actual array data. The entire downloaded array data D is evaluated as having lost its randomness, and otherwise, the random number determination data number sequence is set as a pseudo random number sequence candidate.

  If the pseudo-random number generator 10 evaluates that the randomness of the entire downloaded sequence data D is lost by comparing the predicted data and the random number determination data sequence, it is newly separated from the genome database 20 on the Internet. The base sequence data (sequence data D) is downloaded, and randomness evaluation and random number determination data are extracted again. As a result, for example, the pseudo-random number sequence can be maintained even in applications where it is required to send the pseudo-random number sequence at all times. The pseudo-random number generation device 10 manages whether or not one or more downloaded array data D has already been evaluated. Basically, the array data D once evaluated as having lost its randomness Is not used for generating pseudo-random numbers (however, for example, a period of several tens of years or more can be determined in advance).

  Hereinafter, the configuration and operation of the pseudorandom number generator according to an embodiment of the present invention will be described more specifically.

〔Device configuration〕
FIG. 2 is a block diagram of a pseudo random number generator according to an embodiment of the present invention. The pseudorandom number generation device 10 includes a control unit 100 and a storage unit 200. The control unit 100 includes a base sequence data download request unit 101, a base sequence data sequence conversion unit 102, a non-linear time series analysis unit 103, a random number predictability determination unit 104, and a random number generation unit 105. In addition, although each function of the control part 100 which concerns on this invention is demonstrated, it should be noted that it is not intended to exclude other functions provided in the pseudorandom number generation device 10. The pseudorandom number generation device 10 can be configured as a computer, and a program describing processing contents for realizing each function of the control unit 100 and these data (base sequence data, etc.) are stored in the storage unit 200 of the computer. Each function of the pseudo-random number generation device 10 can be realized by storing the program in this area and reading and executing this program by the central processing unit (CPU) of the computer.

  The base sequence data download request unit 101 starts the random number generation from the genome database 20 deployed on the Internet through the network, and the naturally existing discrete sequence data D such as already accumulated base sequence data D The function to download. In addition, when the base sequence data download request unit 101 is evaluated as having lost randomness with respect to the entire downloaded sequence data D based on the comparison between the prediction data and the random number determination data by the random number predictability determination unit 104 , It has a function of newly downloading another base sequence data from the genome database 20 on the Internet.

  The base sequence data sequence converter 102 sets a part of sequence data (for example, the first sequence) in the actual sequence data D obtained by downloading as library sequence data, and the library sequence data and The array data sequentially taken out for random number determination in other parts (for example, the array other than the head) of the actual array data D is set as random number determination data, and each of the library array data and the random number determination data It has a function of generating a sequence data sequence for library and a data sequence for random number determination by assigning a sequence of predetermined bits to a base to convert it into a numeric sequence.

  The non-linear time series analysis unit 103 uses a non-linear time series analysis program (deterministic non-linear prediction method) as disclosed in, for example, Patent Document 2, using the sequence data sequence for the library converted into a numerical sequence. It has a function of generating predicted prediction data.

  The random number predictability determination unit 104 compares the prediction data generated by the non-linear time series analysis unit 103 with the random number determination data sequence sequentially extracted, and when the non-random state continues M times, the downloaded sequence data The entire D is evaluated as having lost its randomness, and otherwise, the random number determination data sequence is set as a pseudo random sequence candidate. Further, when the random number predictability determination unit 104 evaluates that the randomness is lost with respect to the entire downloaded sequence data D based on the comparison between the prediction data and the random number determination data, a new genome database on the Internet is used. 20 has a function of giving an instruction to the base sequence data download request unit 101 so as to download another base sequence data from 20.

  The random number generation unit 105 has a function of sequentially sending the pseudo random number sequence candidates set by the random number predictability determination unit 104 to the outside as pseudo random numbers.

[Device operation]
FIG. 3 is an operation flow diagram of the pseudo random number generation device according to an embodiment of the present invention. First, the base sequence data download request unit 101 downloads the discrete sequence data D of the base sequence data already accumulated from the genome database 20 provided on the Internet through the network (step S1). An example of the discrete sequence data D of the base sequence data to be downloaded is shown in FIG.

  Next, the base sequence data sequence converter 102 sets a part of the sequence data (for example, the first sequence) in the actual sequence data D obtained by downloading as library sequence data, and this library data The sequence data and the sequence data of the other part of the actual sequence data D (for example, the sequence other than the head) are set as random number determination data, and for each base in the library sequence data and the random number determination data, Thus, a numerical sequence of predetermined bits is assigned to be converted into a numerical sequence, and a library array data sequence and a random number determination data sequence are generated (steps S2 and S3). In addition, after all the actual array data D obtained by downloading is converted into a numerical string by assigning a predetermined bit number sequence, the library array data and random number determination data are determined and set. Of course, it may be.

  For example, the first 1000 points of the array data D can be set as library data for executing the prediction data. Further, for example, A = 0b00, G = 0b01, T = 0b10, and C = 0b11 are set, and five elements are extracted from the array data D to create a 10-bit integer value. By specifying the algorithm to be extracted separately by the user, a user-specific random number can be created even with the same base sequence data.

  Next, the nonlinear time series analysis unit 103 generates prediction data predicted by the deterministic nonlinear prediction method using the library array data converted into the numerical sequence (step S4).

  Next, the random number predictability determination unit 104 compares the prediction data generated by the non-linear time series analysis unit 103 with the random number determination data sequence sequentially extracted, and downloads when the non-random state continues M times. The entire array data D is evaluated as having lost its randomness, and otherwise, the random number determination data is set as a pseudo-random number sequence candidate (step S5). In addition, when the random number predictability determination unit 104 evaluates that the randomness is lost with respect to the entire downloaded sequence data D based on the comparison between the prediction data and the random number determination data sequence, The base sequence data download request unit 101 is instructed to download another base sequence data from the database 20, and in response to this instruction, the base sequence data download request unit 101 newly separates from the genome database 20 on the Internet. Can be downloaded, and the processing from step S1 can be repeated (step S6).

  The random number generation unit 105 sequentially sends the pseudo random number sequence candidates set by the random number predictability determination unit 104 to the outside as pseudo random numbers (step S7).

  The parameter M (≧ 1) used for determining whether or not the non-random state has continued M times is a parameter that can be arbitrarily determined by the user using the pseudorandom number generation device 10. The value of M can determine the strength of the obtained random sequence. That is, the smaller the value of M, the stronger the randomness, and the larger the value of M, the weaker the randomness of the data obtained. Therefore, in the present specification, this parameter M is referred to as “random number strength”. I will decide. As described above, the pseudorandom number generator 10 uses the genome database 20 on the Internet, and provides a pseudorandom number sequence in which the strength of randomness such as periodicity and distribution bias can be arbitrarily set by the parameter M. Can do.

  As described above, the pseudorandom number generation device 10 generates a random number sequence while evaluating randomness regularly or regularly using a non-linear time series analysis program (deterministic nonlinear prediction method) using a natural array. Therefore, it is possible to generate a pseudo-random number sequence with high reliability with respect to randomness. Further, the random number sequence desired by the user can be obtained by changing the strength of the random number with one parameter M.

  In addition, regarding the point of obtaining a random number sequence by utilizing the randomness of a naturally occurring discrete sequence such as a base sequence, a large number of sequences are stored in the storage unit 200 in advance instead of collecting and downloading such sequences. Data can be stored and retrieved from the storage unit 200 in order. In addition, regarding the algorithm for generating random numbers from the base sequence data, not only can the user's own method be implemented for versatility, but it can also be determined uniquely. Furthermore, there is a possibility that the applicable range can be expanded for data other than the base sequence.

  In addition, since it verified by another verification method whether the pseudorandom number sequence produced | generated by the pseudorandom number generation apparatus 10 has randomness, it demonstrates. FIG. 5 is a diagram showing an example of a pseudo-random number sequence generated by the pseudo-random number generator according to an embodiment of the present invention. The first 1000 points of the sequence data D (having 800,000 bases or more) are predicted. It can be set as library data for executing data. Also, for example, adenine: A = 0b00, guanine: G = 0b01, thymine: T (or uracil: U) = 0b10, cytosine: C = 0b11, 5 elements are extracted from the array data D, and 10-bit integer In this example, numerical values are generated as random numbers (the horizontal axis indicates 153 data points, and the vertical axis indicates normalization so that the maximum 10-bit integer value is 1.0).

FIG. 6 shows the result of calculating the autocorrelation function of the pseudo random number sequence shown in FIG. From the result of FIG. 6, it can be seen that the correlation coefficient R ² value = 0.0121 is obtained and there is almost no correlation. FIG. 7 shows a histogram of the pseudorandom number sequence shown in FIG. 5, and FIG. 8 shows the analysis result of the histogram. As a result of performing the chi-square test, since it does not fall within the rejection range [16.9898, ∞] with a significant level of 95% (that is, 12.42484 <16.9898), it can be said that there is a significant difference. It can be said that there is randomness.

  Thus, it was possible to objectively verify that the pseudorandom number sequence generated by the pseudorandom number generator 10 is a pseudorandom number having extremely high randomness.

  In the above-described embodiment, an example in which one pseudo-random number sequence is obtained from one array data D has been described. However, when a plurality of random number sequences are required, it can be obtained by performing the nonlinear time series analysis again.

  According to the present invention, since a pseudo-random number sequence is generated by extracting a numerical sequence of arbitrary strength from these base sequence data, the generated pseudo-random number sequence is not limited, and the random performance (random strength) ) Is not limited, and it becomes possible to generate a pseudo-random number sequence while solving problems such as periodicity and uneven distribution, which is useful for the purpose of sequentially requesting an arbitrary-length random number sequence.

DESCRIPTION OF SYMBOLS 10 Pseudorandom number generator 20 Genome database 100 Control part 101 Base sequence data download request part 102 Base sequence data sequence conversion part 103 Nonlinear time series analysis part 104 Random number predictability judgment part 105 Random number generation part 200 Storage part

Claims (6)

A pseudo-random number generator for generating pseudo-random numbers from discrete array data,
A part of the discrete array data of a plurality of elements is used as library array data, and another part of the discrete array data different from the library array data is used as random number determination data. And a data sequence conversion unit for converting each of the elements in the random number determination data into a numeric sequence by assigning a sequence of predetermined bits, and generating an array data sequence for the library and a random determination data sequence,
Using the sequence data sequence for the library, a nonlinear time series analysis unit that generates prediction data predicted based on deterministic nonlinear prediction;
Compare the prediction data and the random number determination data sequence sequentially extracted, and evaluate that the randomness is lost with respect to the entire discrete array data when a non-random state continues for a predetermined number of times, otherwise A random number predictability determination unit that sets the random number determination data number sequence as a pseudo random number sequence candidate;
A random number generator that sequentially sends the pseudo-random number sequence candidates to the outside as pseudo-random numbers;
A pseudo-random number generation device comprising:   The pseudo random number generation device according to claim 1, wherein the discrete sequence data includes base sequence data.   The pseudorandom number generation according to claim 2, further comprising a base sequence data download request unit that downloads the base sequence data as the discrete sequence data from a genome database deployed on the Internet through a network. apparatus. The random number predictability determination unit evaluates that the randomness is lost with respect to the entire discrete sequence data based on the comparison between the prediction data and the random number determination data sequence, Means for giving an instruction to the base sequence data download requesting unit to download the base sequence data;
The said base sequence data download request | requirement part has a means to download another base sequence data from the said genome database according to the instruction | indication from the said random number predictability determination part, The said base sequence data download request | requirement part has a means to download another base sequence data from the said genome database. Pseudo random number generator. A pseudo-random number generation method for generating pseudo-random numbers from discrete array data,
A part of the discrete array data of a plurality of elements is used as library array data, and another part of the discrete array data different from the library array data is used as random number determination data. And converting each of the elements in the random number determination data into a numerical sequence by assigning a predetermined bit number sequence, and generating a library array data number sequence and a random number determination data sequence,
Generating predicted data based on deterministic nonlinear prediction using the library sequence data sequence; and
Comparing the prediction data and the random number determination data sequence, when the non-random state continues for a predetermined number of times, it is evaluated that the randomness is lost with respect to the entire discrete array data, otherwise, the Setting a random number determination data sequence as a pseudo random sequence candidate;
Sequentially sending the pseudo-random number sequence candidates to the outside as pseudo-random numbers;
A pseudo-random number generation method comprising: In a computer configured as a pseudo-random number generator that generates pseudo-random numbers from discrete array data,
A part of the discrete array data of a plurality of elements is used as library array data, and another part of the discrete array data different from the library array data is used as random number determination data. And converting each of the elements in the random number determination data into a numerical sequence by assigning a predetermined bit number sequence, and generating a library array data number sequence and a random number determination data sequence,
Generating predicted data based on deterministic nonlinear prediction using the library sequence data sequence; and
Comparing the prediction data and the random number determination data sequence, when the non-random state continues for a predetermined number of times, it is evaluated that the randomness is lost with respect to the entire discrete array data, otherwise, the Setting a random number determination data sequence as a pseudo random sequence candidate;
Sequentially sending the pseudo-random number sequence candidates to the outside as pseudo-random numbers;
A program for running

Images