A Two Stage Task Scheduler for Effective Load Optimization in Cloud – FoG Architectures

Abstract

In recent times, computing technologies have moved over to a new dimension with the advent of cloud platforms which provide seamless rendering of required services to consumers either in static or dynamic state. In addition, the nature of data being handled in today’s scenario has also become sophisticated as mostly real time data acquisition systems equipped with High-Definition capture (HD) have become common. Lately, cloud systems have also become prone to computing overheads owing to huge volume of data being imparted on them especially in real time applications. To assist and simplify the computational complexity of cloud systems, FoG platforms are being integrated into cloud interfaces to streamline and provide computing at the edge nodes rather at the cloud core processors, thus accounting for reduction of load overhead on cloud core processors. This research paper proposes a Two Stage Load Optimizer (TSLO) implemented as a double stage optimizer with one being deployed at FoG level and the other at the Cloud level. The computational complexity analysis is extensively done and compared with existing benchmark methods and superior performance of the suggested method is observed and reported.