An Example of Biometric Threshold Schemes

The biometric protocol application in information sharing protocols is presented on figure 4.2. It is an example of biometric threshold scheme application for the procedure of concealing information with the use of a process marking every shadow by means of individual biometric features.

Figure 4.2. Biometric threshold scheme.

Figure 4.2 presents the main idea of biometric threshold schemes using individual personal features for securing secret information. Fig. 4.2 presents two possible groups of biometric data, as well as voice parameters and fingerprint features.

Secret data in the form 'Hosei University in Tokyo' is encoded with the linguistic algorithm, then shadows – in a number determined at the stage of defining the secret information sharing algorithm – are generated. Allocating shadows is based on assigning them to individual protocol participants after their correct identification by means of biometric features specified and defined in the biometric features database. This allocation is unambiguous and it can include personal verification executed by means of one biometric feature or a set of features (their type is specified by the information protection protocol). After the stage of allocating shadows among individual protocol participants, there comes the stage of information re-creation. A required number of shadows is put together; each one is verified with a view to its origin. This is to prevent any possibility of hostile shadow takeover by unauthorised persons and so that the content of the shared information is not disclosed to unauthorised persons. It is only putting together the required number of shadows, after the stage of a proper verification of their holders, that makes it possible to recreate the shared information.

The formal procedure of coding data in biometric threshold schemes is as follows:

Step 1. B = {bs, bns}       //selection of biometric features set  

standard biometric features can be obtained by means  of a scanner, while non‐standard biometric features by  means of cognitive systems//  

Step 2. F = {f1, f2, f3, f4, f5, f6, f7}   //set of required 7 biometric features from a selected  biometrics// 

Step 3. Generating n shadows   //application  of  the  (m, n)‐threshold  scheme,  generating a shadow for every protocol participant// 

Step 4. V = {(nj, Fn)} j ϵ [0, n] – n‐th shadow owner, Fn ϵ (F0, Fn) – biometric features  of the n‐th shadow holder  

//developing  vectors  of  features,  allocation  of  biometrics  to  individual  shadows  of  the  co‐shared  secret// 

Step 5. Definition of m   //specification of the number of m shadows required  to recreate the secret// 

Step 6. Re‐creating the secret   //putting together m parts of (ni, Fn)// 

The here-described methodology of biometric threshold scheme functioning has been presented also on figure 4.3.

Figure 4.3. Structure of a biometric threshold procedure.

It is possible to use different personal data to mark shadows in biometric threshold schemes. Each of them contains individual, unique features, characteristic for a given person. Personal data can be used in the following form [75], [82]:

 as a label for selected or all shadows, required to mark shadows in the procedure of restoring the original information,

 as an additional secret shadow included in all personal data, needed to restore the original secret information.

Selection of the right form of using biometric threshold schemes depends on the type of data protection. Using biometric threshold schemes for information security guarantees the right distribution of secret parts and an unambiguous original secret reproduction.

Chapter 5

New Classes of Protocols – Linguistic and Biometric Threshold Schemes

Abstract. This chapter will discuss possible combinations of the linguistic and biometric protocols dedicated to the execution of data sharing tasks. In this context we shall define new classes of protocols for data sharing, to which linguistic-biometric solutions belong, too. The possibility to combine in one protocol two different classes of protocols enhancing the processes of data protection and securisation offers new application possibilities and at the same time it improves the security of the information protected.

In previous chapters of this dissertation we have discussed independently two new approaches to the data protection processes, as proposed by the Author. These approaches, although very different, are aimed mainly to protect data with confidential or secret character, so that unauthorised persons could not take hold of the content of the protected information. The new protocol classes presented here guarantee their appropriate protection by means of introducing new data analysis stages.

In the case of linguistic protocols, an additional stage of meaning description of the protected data has been added; it is aimed to introduce their description by means of linguistic techniques. The execution of this stage in a way imposes the need to define the semantic meaning of the protected data. This results in that it is necessary to assess its meaning in order to safeguard the protected data. In this respect it is possible to conceal not only the sets of secret (protected) data, but also their meaning and the impact they could have on the external situation. The process of linguistic data description is therefore a method guaranteeing not only a more complete assessment of the secret data, but it allows also to protect this type of secrets by applying formal grammars in the process of secret sharing.

In the case of biometric protocols an important novelty related to this solution is the possibility to define a set (or sets) of individual biometric features used for the tasks of labelling parts of the shared secret. Every secret or confidential information, subject to total protection against unauthorised access to it, can be protected by means of data sharing protocols. In the data sharing protocols it is possible to introduce additional protection by means of assigning an individual biometric feature of every participant in the data sharing

protocol by allocating it to its correct owner. In this way every part of the shared secret has a biometric label selected from the available biometric data base. This protocol guarantees a specific allocation of parts of the shared secret to their rightful owners due to that every biometrics is individual and non-repetitive. Extending the data sharing protocols by an additional stage of using biometrics in the distribution processes of individual parts of the shared secret, we enhance significantly the security of individual secret parts as well as of the entire protected, secret information. The application of biometrics in data sharing protocols serves not only to protect the shared secret parts at the stage when they are allocated; it guarantees the secret’s security also in the situation of a hostile takeover.

Taking into possession parts of the secret is useless in this situation because in the process of re-creation of the secret, the owner of such a part would not have the biometrics which was assigned to that part of the secret. In this way also biometric protocols guarantee greater security of the protected data.

Both linguistic and biometric data sharing protocols serve to enhance the security of confidential or secret data [90]. At the same time, none of these protocols excludes the application of the other solution. It is also possible to apply both these solutions at the same time. Such a solution is the main topic of this chapter of the dissertation.

5.1. Description of New Linguistic-Biometric Threshold