Improving the Security and QoE in Mobile Devices through an Intelligent and Adaptive Continuous Authentication System

Abstract

Continuous authentication systems for mobile devices focus on identifying users according to their behaviour patterns when they interact with mobile devices. Among the benefits provided by these systems, we highlight the enhancement of the system security, having permanently authenticated the users; and the improvement of the users’ quality of experience, minimising the use of authentication credentials. Despite the benefits of these systems, they also have open challenges such as the authentication accuracy and the adaptability to new users’ behaviours. Continuous authentication systems should manage these challenges without forgetting critical aspects of mobile devices such as battery consumption, computational limitations and response time. With the goal of improving these previous challenges, the main contribution of this paper is the design and implementation of an intelligent and adaptive continuous authentication system for mobile devices. The proposed system enables the real-time users’ authentication by considering statistical information from applications, sensors and Machine Learning techniques based on anomaly detection. Several experiments demonstrated the accuracy, adaptability, and resources consumption of our solution. Finally, its utility is validated through the design and implementation of an online bank application as proof of concept, which allows users to perform different actions according to their authentication level.

S3: An AI-Enabled User Continuous Authentication for Smartphones Based on Sensors, Statistics and Speaker Information

“…Then, it was designed the app in charge of collecting the data and, finally, the framework itself was designed as well. It is important to mention that some parts of this project rely on a previous work of our research team [ 3 , 19 ]. This allowed for the adaptation and extension of some of its software (for example, the data collection application), and also for helpful discussions that avoided duplication of research efforts, and enhanced the insight about the open research issues of the sensors and statistics based authentication problem.…”

“…The smartphone application is the first component of the S3 Platform, and its objective is to monitor data in real time, to allow the S3 platform to calculate the user’s authentication level. The design and most of the configuration parameters in the smartphone application are based on results from our previous work [ 3 , 19 ]. For the work described in this paper, a new version of the application used in such previous work has been developed.…”

“…They report that SCANet achieves 90.04% accuracy and 5.14% of EER (Equal Error Rate), using a database they built with 100 volunteers that were asked to carry out three specific tasks: reading, writing and map navigation. In another work [ 3 ], a continuous authentication system that uses sensors data and applications usage statistics was proposed. The authors used an Isolation Forest model as a user behavior anomaly detector.…”

“…Despite benefits provided by current solutions, large amounts of data may be required to generate accurate users’ models. In some cases, more than fifteen days of data are reported as necessary for obtaining an acceptable model performance [ 3 ], although in too many other published works [ 4 , 5 ] this information is simply not provided. Another problem with continuous authentication systems is that they usually suffer a high percentage of false acceptance, opening a door to illegitimate use.…”

PALOT: Profiling and Authenticating Users Leveraging Internet of Things

“…Activity recognition has been widely identified in today’s literature as human activity trackers for continuous authentication of users in IoT scenarios. Many works make use of IoT devices such as smart watches , providing gesture interaction and a permanent monitoring of physical activities [25]; wrist-worn devices for extraction of raw accelerometer data to recognize walking, standing, sitting, and lying activities [26]; smartphones to identify the owner periodically by using machine learning (ML) techniques based on anomaly detection for an adaptive continuous authentication system [27]; and wearable glasses to discriminate the real owner of the smart object from a potential impersonator using biometric features taken from touch gestures and voice commands [28]. Experiments of this last work, conducted using Google Glasses, showed above 93% detection rate using the collected features.…”