Energy Compensated Cloud Assistance in mobile cloud computing

Abstract

We consider the scenario where a mobile device requires assistance from nearby devices to forward its computational tasks to a cloud server. We incentivize cooperation by allowing helper devices to conserve computational energy by offloading their own tasks to the source device's cloud, as compensation for the communication energy lost during task forwarding. We formulate an optimization problem with the objective of minimizing the cloud cost incurred by the source device due to tasks offloaded from helper devices, subject to no energy loss at the helper devices. We observe that this problem cannot be solved using a standard Lyapunov optimization approach. Instead, we construct an alternate problem that follows the standard form but has the same optimal objective value as the original problem. The resultant Energy Compensated Cloud Assistance (ECCA) algorithm does not require any statistics of the system and can be implemented distributively.