An ML and SMS remote access based model for Anti-theft protection of Android devices

Abstract

Android phones being stolen is a significant problem that causes concerns to intellectual privacy and property. Always protecting smartphones from being stolen is a problem that remains. The key findings of the survey of existing systems for theft protection are, they provide various efficient functionalities but fail when the internet is unavailable or require specialized equipment to detect thefts. Most of these solutions are not free of charge, inefficient, time-consuming, or/and inflexible. This paper puts forward a system that provides an ML-based real-time anti-theft and remote access system for android devices. It detects theft using SVM-RBF model trained on feature-set extracted from the inertial sensor’s data with an accuracy of 0.76. Whereas remote access is provided using short message services (SMS). The salient feature of this system is minimal configuration without intruding human-assisted tasks. Moreover, it will be an excellent help for authentic smartphone users to realize the theft situation and utilize the remote access features.