Estado del arte de los estudios sobre la seguridad que poseen los asistentes de voz domóticos como son Alexa o Google Home

Abstract

Currently, the management of home automation voice assistants Alexa or Google Home have had an increase due to their versatility to perform tasks, in addition to interacting with users and other devices in the home, this can generate problems that affect security. This document develops a state of the art on the security of voice assistants, through the methodologies of systematic mapping and systematic review of the literature, for the compilation of relevant studies in various digital repositories between the years 2017 and 2021, which are focused on attacks, vulnerabilities and countermeasures presented by these devices. Within the classification of attacks, aspects such as: confidentiality, integrity and availability are the most affected. Attacks such as: voice squatting, voice masquerading, eavesdropping, tampering among others. It is presented as the aggressions with the most concurrence in the review of studies carried out. On the vulnerability side, it was found that the perspectives to be taken into account are: hardware, users, authentications, software updates and market application; as the possible ways to break the security of home automation voice assistants. In the same way, methods were found such as: machine learning with 15%, attack simulation with 12%, hardware support with 19%, authentication/software with 23%, access control with 8%, recorded and protocols with 23%, which help to deal with possible attacks.