With self-provisioning of resources as premise, dew computing aims at providing computing\nservices by minimizing the dependency over existing internetwork back-haul. Mobile devices have a\nhuge potential to contribute to this emerging paradigm, not only due to their proximity to the end user,\never growing computing/storage features and pervasiveness, but also due to their capability to render\nservices for several hours, even days,without being plugged to the electricity grid. Nonetheless,misusing\nthe energy of their batteries can discourage owners to offer devices as resource providers in dew\ncomputing environments. Arguably, having accurate estimations of remaining battery would help to\ntake better advantage of a deviceâ??s computing capabilities. In this paper, we propose a model to estimate\nmobile devices battery availability by inspecting traces of real mobile device ownerâ??s activity and\nrelevant device state variables.Themodel includes a feature extraction approach to obtain representative\nfeatures/variables, and a prediction approach, based on regression models and machine learning\nclassifiers. On average, the accuracy of our approach, measured with the mean squared error metric,\noverpasses the one obtained by a relatedwork. Prediction experiments at five hours ahead are performed\nover activity logs of 23 mobile users across several months.
Loading....