Energy-aware adaptive computational offloading for pervasive community-based cloud computing

Adaptive Computational offloading systems achieve context specific optimization on mobile and pervasive devices by offloading computational components to a resource copious remote server or cloud. However, with recent advances in the computational capacity of mobile and pervasive devices, adaptive computational offloading could facilitate the formation of ad hoc cloud-like computational environments using collections of mobile and pervasive devices, with reduced reliance on centralised infrastructures. In addition, technologies which are centric to these devices need to effectively utilize the increasingly available local resources by facilitating collaboration among themselves instead of almost invariably out-sourcing the computational tasks to the cloud.<br><br>One critical aspect of adaptive computational offloading is the decision-making process for component placement. Hence, this study formulates a decision-making strategy for adaptive computational offloading systems that enables the distribution of the application components to community-based clouds formed from multiple collaborating peers. To design an effective and efficient adaptive offloading system, the offloading decision-making algorithm must be light-weight and scalable. Thus, this study aims to improve the overall pervasive collaborative experience by extending collaboration lifetime of applications . This objective is achieved by optimising the Time to Failure (TTF) of devices due to energy depletion, while meeting application-specific performance constraints. Specifically, a max-min technique was used to maximise the minimum TTF in order to balance energy consumption across collaborating devices.<br><br>The efficacy, performance and scalability of the formulated model were evaluated. The proposed algorithm produced an optimal solution to the specified model, using integer linear programming, in affordable time and energy for a range of synthetic and real applications and collaboration sizes.