JPH1079031A - Method for certifying electronic signature data and device therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent signature data from being stolen and abused in each approving process by operating a certifying processing by using electronic signature data obtained by successively synthesizing inputted electronic signature data when plural signatures are necessary. SOLUTION: In a terminal A, electronic signature data (x, y, p) (a) of a user (a) and dummy electronic signature data (x, y, p) (dm) are synthesized and transmitted. In an electronic signature data collation system by plural persons, registered electronic signature data (X, Y, P) (M) are synthesized according to an approved sequence U (a, b, c,..., n), and compared with the transmitted synthesized electronic signature data (x, y, p) (a, b, c,..., n). This comparison can be operated by an already established pattern matching method, and when it is judged that they are similar, the synthesized electronic signature data are considered to be true (an legal signature). Otherwise, the data are considered to be false (an illegal signature), and the result is transmitted as a message to each terminal, and the approval is requested again.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for judging the validity of a digital signature in a computer network system which is approved by a dynamic digital signature from a plurality of signers' terminal devices.

[0002]

2. Description of the Related Art The Internet is currently receiving the most attention in computer networks. The Internet began with the ARPAnet, launched in 1969 by the U.S. Department of Defense Advanced Research Projects Agency. Initially, it was a simple computer-to-computer network, but in 1970, a network connecting four universities was established with a packet-switched network that sends data in multiple packets. The communication protocol used here, that is, the protocol is TCP / IP (Transmission Control P
rotocol / Internet Protocol).

[0003] In particular, IP is a core protocol of TCP / IP, and corresponds to the third layer (network layer) of the OSI basic reference model, and provides a service for transporting data to a designated destination (address). . The ARPANET, initially set up for military purposes, had its military components separated and became publicly available on the Internet. In particular, since the establishment of the Commercial Internet Association in 1991 and an agreement not to restrict commercial use in areas funded outside the U.S. government, it has grown dramatically in terms of business use.

Generally, in order to connect a terminal to a host computer, it is necessary to know at least four items: a protocol, a telephone number, a user ID, and a password. In particular, the user ID and the password are identifiers given to the individual user, and can be used only by the authorized individual. However, there is a risk that a user ID and a password, which are confidential personal information, are decrypted and the host computer is illegally connected.

[0005] In recent years, in addition to conventional security checks such as user IDs and passwords, security checks using biometrics have been performed in order to prevent particularly important data from being stolen over communications and abuse of computer resources. They are starting to verify their identity. Biometrics are human characteristics, such as fingerprints, eye irises, voice prints, and handwriting.

[0006] Among them, the method of personal identification by handwriting has been used in computer network systems with the advance of pattern recognition technology. Examples of tools for reading characters include an image scanner and an image OCR, and a stylus pen is generally used for handwriting.

FIG. 1 is an external view of an online handwritten character recognition device using a stylus pen. Stylus pen 1
When a character is written on the tablet 2, the signal is passed to a recognition unit (not shown), and character recognition is performed. Tablets and stylus pens are important parts of usability, so tablets have recently been replaced with LCD screens,
A wireless type is used in which the signal cable protruding from the stylus pen is eliminated. A personal computer is specifically used as the recognition unit, and character recognition is performed by software. The processing in the recognizing unit includes preprocessing / normalization, feature extraction, and identification determination.

[0008] Handwritten characters are not always constant even for the same person. There are always obstacles such as writing direction, size, and noise on the hardware. Therefore, preprocessing and normalization of the recognition unit perform normalization for removing these noises and comparing with a standard character pattern. Specifically, in the pre-processing, the removal of extra point sequences (sampling based on the relative movement amount), the removal of camera shake and random noise depending on the resolution of the tablet (smoothing by moving the load), the independence due to the malfunction of the tablet Perform data removal, etc.

After the preprocessing is completed, the size and position of the input character are normalized as shown in FIG. After the above processing, features are extracted, and identification judgment is performed. As a recognition method, there are a method in which a series of basic strokes is processed by an automaton, and a method in which pattern matching between strokes is entirely performed.

[0010] Online handwritten character recognition is roughly classified into the above two methods, but is classified as follows in terms of restrictions on the writer. (1) Only the prescribed stroke order and stroke count are accepted. (2) Give the stroke order or the number of strokes one degree of freedom. (3) Remove restrictions on stroke order and number of strokes. (4) Recognize not only square type but also broken characters and continuation characters. Although there are several recognition patterns as described above, when a signature is used as a password, the smaller the degree of freedom, the better. This is because a stricter security check can be performed as the degree of freedom is smaller.

In the basic stroke method, each stroke of an input character is approximated by one of a plurality of predetermined basic strokes, and recognition is performed using a series of basic strokes as a feature. As shown in FIG. 3A, a sequence of coordinate points obtained from the tablet is sampled at an appropriate interval, and adjacent sample points are connected by a straight line, and this is connected to a line 8 in FIG. 3B.
Or coded in 16 directions. The stroke code of the received automaton is input to the obtained feature extraction automaton and output. Then, the answer character is determined by another identification automaton using the input as an input. Although the basic stroke method is simple, it is difficult to deal with abbreviations and continuation characters.

On the other hand, the pattern matching method uses the position coordinates of strokes as features. Each stroke and the connection of a plurality of strokes are represented by a vector having coordinate values as components, and recognition is performed by pattern matching. In many cases, one stroke is approximated by 3 to 6 feature points (representative points) as shown in FIG.

In addition to the coordinates, the direction angle (tangent vector) of the movement at the writing start point and writing end point of each stroke
And the one using only the moving direction of the pen. For example, in the example of FIG. 5, the stroke strings 2, 1, 3, and 4 of the input characters that best correspond to the stroke strings A, B, C, and D in the dictionary are found, and the sum of these distances indicates the number of strokes as a variable. If the answer is within a certain threshold, the answer is determined. In other words, the stroke order is free in the collation.

[0014]

Today, the computer networking of information systems is rapidly expanding, the scale is expanding, and the number of users is increasing. As networking evolves, more and more information on the network is compromised. Therefore, the importance of security management is becoming more and more important.

[0015] There is a case where an access right that is originally permitted only to an administrator or a specific user is committed by another person. As seen in the prior art, this is mainly due to the fact that personal authentication is performed using a password. Authentication with a password is likely to be seen by others, and in fact the accident is caused by it.

In order to solve this problem, a personal authentication method using a biometric has recently been used. As a result, the accuracy of the personal authentication is further improved. However, this method is becoming more difficult than ever under the circumstances of networking. This is for the following reasons.

In the case of using biometrics, a human body characteristic (however, handwriting is limited here) which is a characteristic of the user is registered in advance, and the registered data is compared with the data to be verified inputted as a password. Then, it is determined whether the data to be verified is derived from the person.
However, the possibility that the data to be verified is stolen by networking is increasing.

In the Internet, in particular, communication information passes through many relay points, so that the danger is further increased. As mentioned in the prior art, the initial purpose of the Internet was military use, which required that any line or relay station be destroyed but still be able to communicate through alternative routes. For example, as shown in FIG. 6A, in a one-to-one connection between point A and point B, if the line in the middle is disconnected, communication becomes impossible.

On the other hand, in the case of the Internet, FIG.
When a line is used with a plurality of relay stations as shown in (b), points A and B can communicate through any path. what if
(1) is usually in the path when the communications are carried out, the fault has lost communication occurs between the relay point C ₁ and point B for some reason. In that case, if the route can only communicate in this (1), in the conventional case you can not communicate unless the wait for this period of recovery, in the case of Internet communication relay point C ₂ from the relay point C ₁ pathway Can be changed to communicate.

That is, the route (2) becomes possible as a new alternative route. Of course, if the line is congested, it is also possible to select another route for communication. That is, the optimal circuit is always selected. That is the feature and advantage of the Internet. On the other hand, passing information through many relay points has the disadvantage of increasing the risk of information being stolen along the way.

The above-mentioned Internet is used as a large-scale computer network. However, as a small-scale computer network, a LAN and a WAN connecting the LANs are installed in any company. Under such circumstances, such an electronic communication system is also used in the approval process. For example, in the past, transportation and travel expenses were billed by hand,
The process was passed to the clerk, and the approval was paid by the chief, section manager, and general manager before the expenses were paid. Recently, however, an increasing number of companies are conducting this approval process via email. The process of obtaining the approval of a plurality of individuals is particularly common in Japan, where the approval system has been established, and the electronic approval method has been used in various fields.

Regardless of the approval of a small amount of transportation expenses, when it comes to the approval of an important matter that determines the future management policy of the company, it is necessary to strictly check whether the approval has been performed by the person himself or herself. It is becoming. Therefore, this check can be performed by a biometric authentication method.

However, even if the biometric authentication method is more reliable, it does not mean that the danger of biometric data being stolen and subsequently abused has been completely eliminated. In particular, when the approval of a plurality of persons is performed by e-mail, there is a risk that the personal signature data is stolen in each approval process, and the risk that the data is later misused is further increased. Accordingly, an object of the present invention is to newly develop a method for preventing signature data from being stolen and abused in each approval process when a plurality of approvers approve with an electronic signature.

[0024]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention provides a computer network system for performing an authentication process using an electronic signature. A method and an apparatus for performing an authentication process using digital signature data sequentially synthesized. As a method of the electronic signature, a dynamic electronic signature including time-series data of relative coordinates from a writing point and writing pressure is particularly effective.

The combined digital signature data is obtained by comparing the digital signature data of the pre-registered signer with the reference composite digital signature data created by synthesizing the composite digital signature data in the same manner as the composite digital signature data. Authentication of all is performed at once.

In the conventional method, digital signature data for personal authentication checks each digital signature data individually,
The personal authentication is determined by determining whether all the electronic signature data is the same as the registered electronic signature data. On the other hand, in the present invention, the transmitted electronic signature data is one synthesized electronic signature data.

That is, the digital signature data from the terminal is not the data signed by each individual, but is synthesized in each process, and it is not known who the digital signature data is. Therefore, in the collation process on the host computer,
The electronic signature data of the individual registered in the database is synthesized, matched with the transmitted synthetic electronic signature data (combined signature data to be verified), and whether or not the transmitted signature data is all the signature of the individual To determine the validity of the approval.

On the other hand, the terminal performs an electronic signature in order to approve the document transmitted between the terminals. However, when transmitting the document to the next terminal, the electronic signature data itself is not transmitted. Instead, the digital signature data indicating that the digital signature has been approved by the previous terminal is synthesized one after another, and the individual digital signature data is transmitted in a state where it is unknown. In other words, the person who receives the digital signature data at the next terminal can steal the electronic signature data because it is not the raw electronic signature data (individual signature data), store it, and try to exploit it at another occasion. It is not possible.

The approval process of the present invention is for a plurality of persons. For this reason, if personal signature data is passed from terminal to terminal as it is, a third party who is not the person who should be authorized happens to receive the signature data at the terminal, store it, and then abuse it There is a risk. Therefore, in order to avoid this danger, synthesis of digital signature data is performed. Of course, even if the approver is not a third party, the approver himself can try to exploit it, but this can naturally be avoided.

The approval process of a plurality of persons has a flow of terminal 1 → terminal 2 → terminal 3 →... → terminal n → host computer. Normally, the digital signature data of the terminal 1 is sent to the terminal 2 as it is, although it is not raw signature data since the terminal 2 and subsequent terminals are synthetic signature data. However, the terminal 2 may be stolen.

Therefore, dummy electronic signature data is added to the electronic signature data of the first initiator (creator).
When sending from the terminal 1 to the terminal 2, dummy signature data is combined. In this way, the electronic signature data at any time has no raw electronic signature data, and there is no possibility that each electronic signature data is stolen. As a matter of course, when the personal digital signature data is synthesized by the host, the personal digital signature data is synthesized in a format that also includes the dummy data, and matching is performed with the digital signature data to be verified.

[0032]

DESCRIPTION OF THE PREFERRED EMBODIMENTS As information from a stylus pen, relative coordinates (x, y) from the start point of writing and a pen pressure p are taken out.
It is information like. Such time-series digital signature data is called dynamic digital signature data.

[0033]

[Table 1]

Assuming that the two-dimensional position coordinates of the dynamic digital signature data are (x, y) and the pen pressure is p, the sampled time-series signature data is (x (t), y (t), p (t)). Here, the sampling data is discrete time t
, The time t is t = 1, 2, 3,..., N. However, the unit differs depending on the system.

Now, the time-series digital signature data of the signer a is (x (t), y (t), p (t)) (a) (t), p (t)) (b).

The time-series digital signature data obtained by combining the digital signature data of a and b is (x (t), y (t), p (t)) (a, b). At this time, the synthesized time-series signature data differs depending on the synthesis order depending on the synthesis method. That is, (x (t), y (t), p (t)) (a, b) and (x (t), y (t), p (t)) (b, a) are different.

In the present invention, as a material for judging the order in which the transmitted electronic signature data comes, the user I
Individuals can be identified using D or a password (different from digital signature data). FIG. 7 is an example in which the user ID is used as personal identification. In the example shown in the figure, the host computer combines the registered electronic signature data according to the order U (a, b, c,..., N) in which the documents were approved. In the embodiment, the item of approval path is included in the electronic document data. We will come back to this later, but in any case, we need to know the order of the approvers.

In the electronic signature data collating system of the present invention, each person's electronic signature data is registered in advance in the database of the host computer. This database is called an electronic signature database. In the example of FIG.
When the approval data is sent from the terminal A to the terminal B, the electronic signature data of the user a who is the initiator (or simply the first approver) is sent as is. Therefore, the digital signature data of the user a may be stolen at the terminal B.
Therefore, the terminal A combines the digital signature data (x, y, p) (a) of the user a with the dummy digital signature data (x, y, p) (dm) and transmits the combined data.

That is, in this case, the digital signature data sent from the terminal A is (x, y, p) (dm, a). (X, y, p) (dm) may be used since the signature is not usually required for the originator. Naturally, dummy electronic signature data is also registered in the electronic signature database of the host computer.

In the electronic signature data collation system by a plurality of persons, the registered electronic signature data (X, Y, P) (M) is synthesized in accordance with the approved order U (a, b, c,..., N). It is compared with the transmitted synthetic digital signature data (x, y, p) (a, b, c, ..., n).

When dummy electronic signature data is also included, the composite electronic signature data synthesized by the host computer is (X, Y, P) (dm, a, b,..., N) It is compared with the digital signature data (x, y, p) (dm, a, b, ..., n).

The comparison can be performed by a pattern matching method that has already been established. If the result of the matching is determined to be similar, the combined electronic signature data is regarded as true (valid signature). Otherwise, it is considered false (illegal signature) and sent to each terminal in a message to that effect,
Request approval.

As described above, in the electronic approval process by a plurality of persons according to the present invention, individual electronic signature data is stolen and misused by synthesizing individual electronic signature data, that is, multiplexing the electronic signature data. Is not possible. In addition, since the electronic signature data is synthetic electronic signature data, even if this synthetic electronic signature data is stolen by a terminal or the like on the way and then attempted to be abused, the result is determined by the time series of the electronic signature data as a component (in this case, the order of approval). The differences make re-use difficult and the system more secure.

[0044]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described with reference to contract documents and in-house documents converted into electronic documents. First, an application program embodying the idea of the present invention performs a series of processes in an approval process using application program data. The application program data in the approval process includes approver list data, composite approval digital signature data, electronic document data, and approval status data. The above contents are summarized in FIG. This will be described in more detail. In the following, "signature data" is digital signature data,
That is, it is signature data signed by an electronic pen (such as a stylus pen).

The composite approval digital signature data is a composite of a plurality of digital signature data. The digital signature data is synthesized in the order of approval. Therefore, the signature data received by the next approver will not be the raw signature data of the previous approver, but in a composed form,
Each individual digital signature data is not seen in its raw state.
Conversely, the digital signature data is stolen and stored, and cannot be exploited on another occasion.

There are various ways to combine digital signature data.
For example, it can be considered an electronic signature data in Table 1 _{f (x t, y t,} p t) where t = 1, 2, ......, a vector of n. Each component of the time t can be arranged horizontally to form a K (= 3 × n) -dimensional vector V. Assuming that the vectors of the approvers a and b are V _a and V _b , the composite electronic signature data V can be created by V = V _a + V _b .

However, in the above method, since the corresponding components from the start point are always simply synthesized, there is a possibility that the vector becomes flat (a smooth value without unevenness) as the number of approvers increases. is there. If flat, the composite electronic signature data has no characteristic, and is not suitable for comparison.

Therefore, a DP matching combining method is generally used for such combining. Now, it is assumed that the sampling spectrum of the digital signature data of the approver a, b is as shown in FIG. In the DP matching synthesis method, the synthesis is performed based on the point where the absolute value of the spectral component is the largest, that is, the point A and the point B in FIG. In this method, since the start points S _a and S _{b of the} signature are shifted, the components of the synthesized digital signature data increase.

However, since the combined digital signature data does not become flat, the characteristics of the combined digital signature data are not lost. However, when the number of approvers is a small number of about two or three, there is no problem even if they are simply repeated from the start points S _a and S _b . In this regard, the method to be used can be arbitrarily selected depending on the system used. In other words, the synthesis method may be selected according to the needs of the system to which the present invention is applied.

Electronic document data is usually obtained by using a tool such as a word processor (word processor) or DTP.
Prepared in the initialization process. These include text data (document data) and graphic data. The approval status data includes an approval path, an approval end code, an approval time (date and time), and the like. The approval path is the approval order of the person who has approved, and the combined signature data is combined in this order.
This order does not necessarily have to match the order of the approver list, but usually the order of the approver list is the order of approval. Next, an approval process of electronic document data will be described.

When an electronic document requiring approval by a plurality of persons is drafted, the drafter uses an initialization program to execute the initialization.
Prepare electronic document data. After creating an electronic document using a word processor or DTP, the drafter creates a list of approvers and accesses the signature database of the host computer to check whether the approver is a valid member (member check). As a result, the approval status data is initialized.

In this state, of course, the composite approval digital signature data is also initialized. As the composite approval digital signature data, a numerical sequence consisting of a random number sequence may be used in an initial state in order to improve security. That is, a dummy signature is inserted with a random number. Of course, in this case, when collating, the numerical sequence used in the initialization must also be considered. Therefore, it is necessary to inform the host of which numerical sequence was used in the member check.

When the initialization processing is completed, the application program data is transmitted to the first approver. When the first approver accesses the computer after the application program data is transmitted to the terminal (usually a personal computer) of the first approver, it is understood that the data is being transmitted. The approver can use the application program to view the electronic document data, approver list data, and approval status data. As a result, it can be seen that the first approver has issued an approval request for the electronic document data. The first approver examines the content and performs an approval process using a computer. Naturally, at this time, a peripheral device (tablet, electronic pen, etc.) for performing an electronic signature must be connected to the terminal.

The digital signature data input from the tablet is processed by the standardization processing in the approval processing program. In the normalization processing, standardization of the shape and size of the digital signature data, correction of the shape in the long axis direction, standardization of the sampling number, and removal of higher-order Fourier components of each component (x, y, p) I do. The digital signature data obtained by the normalization processing is added to the composite signature data in the application program data, and the result becomes new composite signature data. Further, the approval processing program attaches a sign indicating that the first approver has approved the approval status data, transmits the application program data to the second approver, and ends the approval processing.

In the second and subsequent approval processes, the same approval process as that of the first approval process is performed. When the approval of all approvers is completed, the application program data is processed by the collation process. The composite signature data at this stage has a form in which the electronic signature data of all approvers are composited. The collation processing is controlled by a collation processing program registered in the host computer.

In the approval process, first, the electronic signature data is read from the electronic signature database in which the electronic signature data of the approver is registered, and the electronic signature data of all the approvers in the approver list are combined to generate the combined registered electronic signature data. create. Of course, at this time, if the combining method is affected by the order of the signatures, the combining is performed in consideration of the order of approval.

Next, the transmitted combined digital signature data and the combined registered digital signature data are collated by using a pattern matching method, and it is determined whether they match within a certain range of error. Effectiveness). If there is an approval signature other than the approver during the approval process, the result of the matching will be determined to be a mismatch (no similarity). Naturally, if the approver does not approve (disapprove) the document during the approval process, the approver will be notified to the originator by e-mail or telephone. In some cases, the collation process may not be performed.

FIG. 10 illustrates the above processing. The figure shows an example where there are three approvers,
No matter how many people, the process flow does not change. The application program data created (initialized) by the creator is transmitted to each approver, a signature is performed each time the approver is approved, and the signature data moves between approvers while being added to the composite electronic signature data. . The collation processing is processing in the host computer, and a signature database (DB) used in the collation processing is connected to the host computer. The host computer does not have to be a large computer,
In the case of a LAN, a personal computer server may be used. In short, it is important to have a central computer installed in a place where security is tightly controlled.

[0059]

As a communication means on the same site in a company, there is a LAN. What is networked by connecting LANs to each other is a WAN. Furthermore, the Internet is enabling the network with a worldwide spread. Computer networks have expanded computers not only in the field of computing but also in the field of communication, and have changed the means of communication in the field of communication.

In the past approval process, the original document (the document as drafted) always went to the approver, and after reading the document, a plurality of persons signed the same document for approval. For this reason, it may take one or two months for one document to be approved by all approvers.

However, with the development of computer networks, more rapid processing has been required, and the number of companies that accept digital signatures has been increasing. Approval by a computer network is dangerous, unlike the case where source documents are manually rotated as in the past. One of the risks is that the digital signature data is stolen and later misused. Since electronic approval by a plurality of persons passes through a large number of terminals, there is a high risk of signature data being stolen. The danger is even higher, especially for electronic signatures associated with commercial agreements and videoconference agreements between companies. For such a danger, the conventional method cannot avoid this danger.

On the other hand, the present invention has an advantage that individual digital signature data cannot be independently stolen because the method of synthesizing digital signature data is employed. Moreover, in the present invention, since the digital signature data is synthesized in a time-series manner, it is difficult to decompose and extract individual signature data from the synthesized signature data, and a more secure system is provided.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of an on-line handwritten character recognition device using a stylus pen according to the related art.

FIG. 2 is an explanatory diagram of character normalization in the related art.

FIG. 3 is an explanatory diagram of character recognition by a basic stroke method of the related art.

FIG. 4 is an explanatory diagram of character recognition (stroke representative point approximation) by a conventional pattern matching method.

FIG. 5 is an explanatory diagram of character recognition (feature point pattern matching) by a conventional pattern matching method.

FIG. 6 is an explanatory diagram of a conventional communication network.

FIG. 7 is a diagram showing a process of electronic approval and a flow of synthetic signature data in the present invention.

FIG. 8 is an explanatory diagram of a configuration of application program data according to the embodiment of the present invention.

FIG. 9 is an explanatory diagram of a method of synthesizing digital signature data in the embodiment of the present invention.

FIG. 10 is an explanatory diagram of a flow of signature data and a collation process in the embodiment of the present invention.

[Explanation of symbols]

 1 stylus pen 2 tablet

Claims (5)

[Claims] In a computer network system for performing an authentication process using an electronic signature, an authentication process is performed on information requiring a plurality of signatures by using electronic signature data obtained by sequentially combining input electronic signature data. A method for authenticating digital signature data, characterized in that: 2. In a computer network system for performing an authentication process using an electronic signature, electronic signature data obtained by sequentially combining input electronic signature data with information requiring a plurality of signatures is registered in advance. The signature processing is performed in such a manner that the signature digital signature data is combined with the reference composite digital signature data created by synthesizing the digital signature data of the signer in the same manner as the composite digital signature data and authentication of each signature is performed collectively. Authentication method for digital signature data. 3. The method according to claim 1, wherein the digital signature is a dynamic digital signature comprising time series data of relative coordinates from a writing point and writing pressure. 4. In a computer network system for performing an authentication process using an electronic signature, electronic signature data obtained by sequentially combining input electronic signature data with information requiring a plurality of signatures is registered in advance. Authentication processing means for verifying each signature in a lump by collating the signer's digital signature data with reference composite digital signature data created in the same manner as the composite digital signature data is combined An electronic signature data authentication system characterized by the following. 5. A computer in which, for information requiring a plurality of signatures, electronic signature data obtained by sequentially synthesizing input electronic signature data, digital signature data of a pre-registered signer is combined with synthetic electronic signature data. A medium in which a program for executing an authentication processing procedure for performing authentication of each signature collectively by collating with reference synthetic digital signature data created by synthesizing in the same manner as synthesizing the digital signature is synthesized.