Delay-Energy Tradeoff in Mobile Cloud Computing: An Experimental Approach (original) (raw)

Mobile applications require more and more resources to be able to execute tasks on a single device, despite the fact that mobile devices are getting better capabilities. This has been addressed through several proposals for efficient computation offloading from mobile devices to remote cloud resources or closely located computing resources known as cloudlets. In this paper we adopt a experimental driven approach to highlight the offloading tradeoffs. We show that rather than always offloading tasks to a remote machine, running particular tasks locally can be more advantageous. We propose a novel generic architecture that can be integrated to any mobile cloud computing application in order to automate the offloading decision and help these applications to improve their response time while minimizing the overall energy consumed by the mobile device.