Thesis etd-08212021-125548 |
Link copiato negli appunti
Thesis type
Tesi di laurea magistrale
URN
etd-08212021-125548
Thesis title
Gait-based user's authentication: energy consumption of different software architectures for Android wearable devices
Department
INGEGNERIA DELL'INFORMAZIONE
Course of study
COMPUTER ENGINEERING
Supervisors
.
relatore Prof. Vecchio, Alessio
relatore Prof. Avvenuti, Marco
relatore Ing. Cola, Guglielmo
relatore Prof. Avvenuti, Marco
relatore Ing. Cola, Guglielmo
Keywords
- Android
- autenticazione
- authentication
- camminata
- consumo energetico
- dispositivi mobili
- energy-consumption
- gait
- mobile
- mobile devices
- smartwatch
Graduation session start date
24/09/2021
Availability
Full
Abstract (Inglese)
Abstract (Italiano)
Smartphones and smartwatches, thanks to the sensors they are equipped with, provide a suitable platform for implementing and deploying motion-based behavioral biometrics.
In this thesis, a gait-based authentication method is designed and implemented for the Android ecosystem, filling the gap between prototypical and real-world solutions. Particular attention is given to the computation, communication, and energy requirements of the system.
To this purpose, the impact of the sampling frequency on the accuracy vs consumption tradeoff is first analyzed. Then several architectural configurations are compared, to assess the benefits obtained when executing the whole method on the smartwatch, against a more distributed approach that relies also on a smartphone. Results show that a smartwatch-only solution is more energy-efficient, as transferring the data to an external device outweighs the benefits due to the decreased computational load.
In this thesis, a gait-based authentication method is designed and implemented for the Android ecosystem, filling the gap between prototypical and real-world solutions. Particular attention is given to the computation, communication, and energy requirements of the system.
To this purpose, the impact of the sampling frequency on the accuracy vs consumption tradeoff is first analyzed. Then several architectural configurations are compared, to assess the benefits obtained when executing the whole method on the smartwatch, against a more distributed approach that relies also on a smartphone. Results show that a smartwatch-only solution is more energy-efficient, as transferring the data to an external device outweighs the benefits due to the decreased computational load.
File
| Nome file | Dimensione |
|---|---|
| TesiNocerino.pdf | 1.31 Mb |
Contatta l’autore |
|