Adaptive Offloading for Pervasive Computing

Pervasive computing allows a user to access an application on heterogeneous devices continuously and consistently. However, it is challenging to deliver complex applications on resource-constrained mobile devices such as cell phones. Application-based or system-based adaptations have been proposed to address the problem, but they often require application fidelity to be significantly degraded. We believe that this problem can be overcome by dynamically partitioning the application, and by offloading part of the application execution with data to a powerful nearby surrogate. This allows the application to be delivered in a pervasive computing environment without significant fidelity degradation or expensive application rewriting. Runtime offloading needs to adapt to different application execution patterns and resource fluctuations in the pervasive computing environment. Hence, we have developed an offloading inference engine to adaptively solve two key decision-making problems in runtime offloading: (1) timely triggering of offloading, and (2) efficient partitioning of applications. Both trace-driven simulations and prototype experiments show the effectiveness of the adaptive offloading system