Deep Learning approaches for resources scheduling in edge computing iot networks

Abstract

Technological advancement in internet-based applications leads to newlinethe proliferation of various smart applications like smart healthcare, smart newlinetransportation, smart mining, security, etc., The massive number of Internet of newlineThings (IoT) devices increases every day and continuous data collection newlinerequires an efficient computing platform for better data management. The newlinehuge computing resources of cloud computing are widely adopted in IoT newlineapplications. Moreover, these applications require large bandwidth, minimum newlinelatency, high availability, enhanced security, and low jitter. Processing all the newlinecollected data in cloud servers is unnecessary and unfeasible. Thus, an edge newlinecomputing concept was introduced which processes the data in the edge layer newlineand reduces the computation burden of cloud central servers and networks. newlinemeanwhile, the latency and bandwidth requirements of IoT applications are newlineeffectively handled by edge networks by moving the computing resources of newlinethe cloud near to the end users. newline Resource scheduling in edge computing is an essential process that newlineschedules the appropriate cloud resources for IoT network tasks. The resource newlinerequests are processed generally based on the computation requirements newlinewhich are defined by the end user. The selection of optimal resources from a newlinelarge resource pool in cloud computing is a quite challenging process. In newlinerecent times various machine learning and deep learning models are evolved newlinefor resource scheduling in edge-integrated IoT networks. However, the newlineperformances should be improved in terms of latency, response time, newlineexecution time, and efficiency to improve the overall quality of services. newlineThus, different novel deep learning techniques are developed in this research newlinework to improve the resource scheduling process in edge integrated IoT newlinenetworks. newline newline