Toward a Secure and Usable User Authentication Mechanism for Mobile Passenger ID Devices for Land/Sea Border Control

Generating Datasets Based on the HuMIdb Dataset for Risk-based User Authentication on Smartphones

“…In our more recent work [13], we have provided a comprehensive work on the design of a risk-based adaptive user authentication mechanism that comprises a novel secure and usable user authentication solution. On top of that, we also modified adequately the "HuMIdb" dataset files, and we trained and tested the following most popular classification algorithms for risk-based authentication: K-NN, DT, SVM, and NB over the "HuMIdb" dataset using ten-fold cross validation.…”

“…In this way, it is suggested that risk-based user authentication mechanisms enhance the reliability of whole authentication process without interrupting the smartphone user's regular activity [9]. In our previous publications [10]- [13], we have presented: (i) a thorough review of related work on user authentication solutions for public safety and mobile devices, (ii) the security vs. usability challenge, (iii) the concept of the risk-based user authentication, as well as (iv) the HuMIdb dataset, which, to the best of our knowledge, is the most recent and publicly available dataset for behavioral user authentication [14], [15], as well as (v) a comprehensive work [13] on the design of a risk-based adaptive user authentication mechanism for mobile devices that comprises a novel secure and usable user authentication solution ensuring continuous user authentication behind-thescenes and invisible to the user.…”

“…On top of that, in [13], we also modified effectively the ''HuMIdb'' dataset files, and we trained and tested the following most popular classification algorithms for riskbased authentication: K-NN, Decision Trees (DT), Support Vector Machine (SVM), and Naïve Bayes (NB) over the ''HuMIdb'' dataset using ten-fold cross validation. Nevertheless, the evaluation results for these classification algorithms demonstrated impact of overfitting and therefore, we considered the concept of novelty detection to overcome this challenge.…”

Risk Estimation for a Secure & Usable User Authentication Mechanism for Mobile Passenger ID Devices

“…Although security and usability in mobile user authentication are often thought of as being contradictory, Risk-Based user Authentication (RBA) method has been extensively proposed in the literature to address this security vs. usability challenge enhancing security without sacrificing usability [1]- [5]. In particular, Risk-Based user Authentication (RBA) method based on behavioral biometrics appears to have the capacity to enhance the whole authentication process´s reliability without interrupting the user's normal activity by dynamically authenticating a genuine mobile user throughout their entire interaction with the mobile device, based on a risk score computed in real-time [6]- [8].…”

A Scalable Approach of Practical Byzantine Fault Tolerance Algorithms for IoMT Blockchains

“…Inside this frame of reference, the industry and research community have identified blockchain technology as a disruptive technology that can be integrated into novel security solutions for IoMT edge networks, as it can support to: a) secure IoMT devices; and b) prevent unauthorized access during data transmission (i.e., tamper proof transmission of medical data) [3]. In spite of this, resource-constrained IoMT nodes (e.g., medical sensors, IoMT Gateways), which are the primary components of IoMT-based health monitoring systems [10], cannot afford the high resource requirements, in terms of computational cost and energy consumption of Proof-of-Work (PoW) consensus algorithms which are the most popular consensus algorithms for blockchain mechanisms in the industry [11]. On the other hand, lightweight algorithms have been proposed based on Practical Byzantine Fault Tolerance (PBFT) that implement a vote system to reach consensus inside a distributed network, rather than calculate a computationally intense puzzle [3].…”