IEEE Access (Jan 2020)
Dynamic Energy Efficient Resource Allocation Strategy for Load Balancing in Fog Environment
Abstract
The Internet of Things is a flexible, emerging technology and an innovative development of the environmental trend. It is a large and complex network of devices in which fog computing plays a growing role in order to handle the information flow of such large and complex networks. Influence of their activities on carbon emissions and energy costs in unlimited results. Dynamic and efficient load balancing technology can be used to improve overall performance and reduce energy consumption. Load can be transferred or shared between computer nodes through load balancing technology. Therefore, the design of energy-efficient load balancing solutions for edge and fog environments has become the main focus. In this research work, we have proposed Dynamic Energy Efficient Resource Allocation (DEER) strategy for balancing the load in fog computing environments. In the presented strategy, initially the user submits tasks for execution to the Tasks Manager. Resource Information Provider registers resources from Cloud Data Centres. The information about the tasks and resources are then submitted to the Resource Scheduler. The resource scheduler arranges the available resources in descending order as per their utilization. The resource engine after receiving the information of tasks and resources from the resource scheduler assigns tasks to the resources as per ordered list. During execution of tasks, the information about the status of the resources is also sent to the Resource Load Manager and Resource Power Manager. The Resource Power Manager manages the power consumption through the resource On/Off mechanism. After successful execution of tasks, the resource engine returns the result to the user. Simulation results reveal that the presented strategy is an efficient resource allocation scheme for balancing load in fog environments to minimize the energy consumption and computation cost by 8.67 % and 16.77 % as compared with existing DRAM scheme.
Keywords