Efficient Multiuser Computation for Mobile-Edge Computing in IoT Application Using Optimization Algorithm.

Mobile edge computing (MEC) is a paradigm novel computing that promises the dramatic effect of reduction in latency and consumption of energy by computation offloading intensive; these tasks to the edge clouds in proximity close to the smart mobile users. In this research, reduce the offloading and latency between the edge computing and multiusers under the environment IoT application in 5G using bald eagle search optimization algorithm. The deep learning approach may consume high computational complexity and more time. In an edge computing system, devices can offload their computation-intensive tasks to the edge servers to save energy and shorten their latency. The bald eagle algorithm (BES) is the advanced optimization algorithm that resembles the strategy of eagle hunting. The strategies are select, search, and swooping stages. Previously, the BES algorithm is used to consume the energy and distance; to improve the better energy and reduce the offloading latency in this research and some delays occur when devices increase causes demand for cloud data, it can be improved by offering ROS (resource) estimation. To enhance the BES algorithm that introduces the ROS estimation stage to select the better ROSs, an edge system, which offloads the most appropriate IoT subtasks to edge servers then the expected time of execution, got minimized. Based on multiuser offloading, we proposed a bald eagle search optimization algorithm that can effectively reduce the end-end time to get fast and near-optimal IoT devices. The latency is reduced from the cloud to the local; this can be overcome by using edge computing, and deep learning expects faster and better results from the network. This can be proposed by BES algorithm technique that is better than other conventional methods that are compared on results to minimize the offloading latency. Then, the simulation is done to show the efficiency and stability by reducing the offloading latency.