Drawer Cosine optimization enabled task offloading in fog computing

Abstract

Fog computing offers the benefit of low-latency computing thereby improving the Quality of Service (QoS) of low-latency applications. Hence, it is essential to distribute the applications in a balanced way across the different fog nodes, however, task offloading remains a challenging issue. Several existing methods are Convenient for task offloading in fog computing, but they are affected by congestion and communication delay. The foremost purpose of this work is to introduce a newly developed scheme for task offloading in fog computing named Drawer Cosine Optimization (DCO) based on multiple objectives such as makespan, cost, load, and energy. Here, DCO is designed by the unification of the Drawer Algorithm (DA) and the Sine Cosine Algorithm (SCA). Initially, the user task computation is performed and then the task is uploaded to the fog node. Every node has a local agent, which is responsible for gathering data like sensor service rate and sensor data arrival rate. But, when the fog cloud resources are constrained, task offloading requests are sent by the sensors to fog nodes, which then forward them to a master node, which is in charge of scheduling offloaded tasks to the fog nodes utilizing DCO. The developed DCO is evaluated using measures, such as load, energy, makespan, time and memory and is revealed to achieve superior values of 0.116, 0.472 J, 0.365, 3.221sec and 7.452 MB, when using task size100.