Task-Driven Approach for Deadline Based Scheduling Across Sensor Networks

Tremendous evolution towards cloud computing provides changing demands to work in virtualized environment, i.e. applications such as Internet of Things (IoT), healthcare system, video streaming that are transferred to cloud, that needs responses in certain time period, that is, deadline. Moreover, cloud computing acquires substantial amount of energy whilst offering services to meet applications need, that leads to higher operational cost. In specific, bulky task to offer services to deadline based task while reducing energy consumption. However, effectual task scheduling is an attractive way to reduce energy utilization whilst guaranteeing fulfilled services for cloud users. Task scheduling in cloud is measured as multi-objective minimization crisis which comprises of make-span and reduction of energy consumption. Initially, novel learning based scheduling approach for deadline tasks has been proposed. This model is an adaptive decision making approach assists schedulers to recognize finest responses Then, decision based scheduling (DSA) algorithm is provided which investigates tasks' heterogeneity by guaranteeing time needed for scheduling tasks. Extensive simulation was carried out in MATLAB simulation environment to show applicability and effectiveness of DSA for deadline scheduling in cloud environment.