Authentication systems based on biometrics characteristics and data represents one of the most important trend in the evolution of the society, e.g., Smart City, Internet-of-Things (IoT), Cloud Computing, Big Data. In the near future, biometrics systems will be everywhere in the society, such as government, education, smart cities, banks etc. Due to its uniqueness, characteristic, biometrics systems will become more and more vulnerable, privacy being one of the most important challenges. The classic cryptographic primitives are not sufficient to assure a strong level of secureness for privacy. The current paper has several objectives. The main objective consists in creating a framework based on cryptographic modules which can be applied in systems with biometric authentication methods. The technologies used in creating the framework are: C#, Java, C++, Python, and Haskell. The wide range of technologies for developing the algorithms give the readers the possibility and not only, to choose the proper modules for their own research or business direction. The cryptographic modules contain algorithms based on machine learning and modern cryptographic algorithms: AES (Advanced Encryption System), SHA-256, RC4, RC5, RC6, MARS, BLOWFISH, TWOFISH, THREEFISH, RSA (Rivest-Shamir-Adleman), Elliptic Curve, and Diffie Hellman. As methods for implementing with success the cryptographic modules, we will propose a methodology which can be used as a how-to guide. The article will focus only on the first category, machine learning, and data clustering, algorithms with applicability in the cloud computing environment. For tests we have used a virtual machine (Virtual Box) with Apache Hadoop and a Biometric Analysis Tool. The weakness of the algorithms and methods implemented within the framework will be evaluated and presented in order for the reader to acknowledge the latest status of the security analysis and the vulnerabilities founded in the mentioned algorithms. Another important result of the authors consists in creating a scheme for biometric enrollment (in Results). The purpose of the scheme is to give a big overview on how to use it, step by step, in real life, and how to use the algorithms. In the end, as a conclusion, the current work paper gives a comprehensive background on the most important and challenging aspects on how to design and implement an authentication system based on biometrics characteristics.

Cloud computing offers the possibility of providing suitable access within a network for a set of resources. Many users use different services for outsourcing their data within the cloud, saving and mitigating the local storage and other resources involved. One of the biggest concerns is represented by storing sensitive data on remote servers, which can be found to be extremely challenging within different situations related to privacy. Searchable Encryption (SE) represents a particular case of Fully Homomorphic Encryption (FHE) and at the same time represents a method composed from a set of algorithms meant to offer protection for users’ sensitive data, while it preserves the searching functionality on the server-side. There are two main types of SE: Searchable Symmetric Encryption (SSE), where the ciphertexts and trapdoors for searching are performed using private key holders, and Public Key Searchable Encryption (PKSE), in which a specific number of users have the public key based on which are capable of outputting ciphertexts and giving the possibility of producing the trapdoors by using the private key from the holder. In this article, we propose a searchable encryption system that uses biometric authentication. Additionally, biometric data are used in the trapdoor generation process, such that an unauthorized user cannot submit search queries. The proposed system contains three components: classic user authentication (based on username, password, and a message with a code using short message service (SMS), biometric authentication, and the searchable encryption scheme. The first two components can be seen as two-factor authentication (2FA), and the second component represents the initialization step of the searchable encryption scheme. In the end, we show and demonstrate that the proposed scheme can be implemented with success for medium to complex network infrastructures. We have granted special attention to the trapdoor function, which generates a value that can be used to perform the search process and search function that is based on the trapdoor pair for searching within the index structure. We provide the correctness and security proof of the operations, which gives us the guarantee that the cloud servers return the correct documents. Additionally, we discuss measuring the performance of the authentication scheme in terms of performance indicators, introducing two indicators for measuring purposes—namely, cloud average number of non-legitim the user actions for cloud purposes (CANNL) and cloud average number of legitim user actions (CANLU).

Digitisation of medical records by means of Electronic Patient Record (EPR) systems promises to

Authentication systems based on biometrics characteristics and data represents one of the most important trend in the evolution of our world. In the near future, biometrics systems will be everywhere in the society, such as government, education, smart cities, banks etc. Due to its uniqueness characteristic, biometrics systems will become also vulnerable, privacy being one of the most important challenge. The classic cryptographic primitives are not sufficient to assure a strong level of secureness for privacy. The following work paper represents an effort to present the main cryptographic techniques and algorithms that can give us the possibility to raise a certain level of secureness for privacy. We will show their own challenges (strengths and weaknesses). We will demonstrate how we can use the most common and well-known techniques and algorithms in order to get a maximum efficiency and a high level in assuring the integrity of the biometrics data.

No abstract

No abstract

No abstract

No abstract