Efficient Real -Time Mobile Computation in the Cloud Using Containers ontainers

Abstract

Mobile devices have limited resources in terms of power and bandwidth. Cloud computing offers a way to reduce the power consumption of mobile devices by offloading computation to the cloud. However, offloading computation means an increase in communication energy consumption. The trade-off between energy and network characteristics (bandwidth/latency) in a mobile device is very important. Therefore computation offloading must be done strategically. The optimum utilization of the available mobile device resources needs to be assured. In this paper, we propose an intelligent and dynamic algorithm to offload computation to the cloud utilizing containers. We focus on offloading computation based upon the communication topology, device energy and user inputs. We analyze the cost of offloading computation for different user inputs, and based on the inputs, we decide whether to offload the application to the cloud or not. Our algorithm was implemented using two different approaches, a fuzzy-logic system and a neuro-fuzzy system. These were both simulated using MATLAB?, and the results compared. Our previous work demonstrated that the fuzzy-logic system performed better than the referenced system. This paper demonstrates that the neuro-fuzzy approach, utilizing a container based hosting model, improves on our prior results.