Fides: A hidden market approach for trusted mobile ambient computing

We propose Fides, a computation offloading mechanism for mobile devices based on trust between them. Fides provides functionalities to the application developers in terms of (i) application splitting and (ii) metric proposal, according to which the offloading decisions will be made. A collective intelligence mechanism is learning the parameters of the ambient environment and selects the proper nearby devices (helpers) to aid the device that is asking for help (aided) via a linear complexity online algorithm. Experiments in android mobile devices show how user´s quality of experience can be improved in terms of battery consumption and delay minimisation.