Exploiting Non-Causal CPU-State Information for Energy-Efficient Mobile Cooperative Computing - 2018

PROJECT TITLE :

Exploiting Non-Causal CPU-State Information for Energy-Efficient Mobile Cooperative Computing - 2018

ABSTRACT:

Scavenging the idling computation resources at the large variety of mobile devices, ranging from tiny IoT devices to powerful laptop computers, will offer a robust platform for local mobile Cloud Computing. The vision can be realized by peer-to-peer cooperative computing between edge devices, called co-computing. This Project exploits the non-causal helper's CPU-state info to design energy-efficient co-computing policies for scavenging time-varying spare computation resources at peer mobiles. Specifically, we tend to think about a co-computing system where a user offloads computation of input information to a helper. The helper controls the offloading process for the objective of minimizing the user's energy consumption based mostly on a predicted helper's CPU-idling profile that specifies the quantity of accessible computation resource for co-computing. Take into account the situation that the user has one-shot input-information arrival and also the helper buffers offloaded bits. The matter for energy-efficient co-computing is formulated as 2 sub-problems: the slave problem resembling adaptive offloading and therefore the master one to data partitioning. Given a fixed offloaded information size, the adaptive offloading aims at minimizing the energy consumption for offloading by controlling the offloading rate underneath the deadline and buffer constraints. By deriving the necessary and sufficient conditions for the optimal answer, we tend to characterize the structure of the optimal policies and propose algorithms for computing the policies. Furthermore, we tend to show that the problem of optimal data partitioning for offloading and native computing at the user is convex, admitting a straightforward answer using the sub-gradient method. Finally, the developed style approach for co-computing is extended to the situation of bursty data arrivals at the user accounting for knowledge causality constraints. Simulation results verify the effectiveness of the proposed algorithms.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here