##plugins.themes.bootstrap3.article.main##

PhD Scholar, AtulGarg

Dr. PritajYadav

Abstract

IOT device based network communication improves monitoring, reporting, business, lifestyle, etc. This directly increase load on servers, cloud hence a mediator was involved as edge computing. But large number of devices need job completion hence Edge load balancing models were required. Many of existing models provides static and dynamic solutions but none of them learn from previous job sequence. So this paper proposed a model that finds the job sequence for the IOT devices by use of modified artificial immune bio-inspired algorithm. As bio-inspired not need any prior information or training for finding the solution so this fit for the dynamic load balancing. Further paper has utilized job sequence and edge features to train LSTM model for getting initial job sequences of artificial immune algorithm. Experiment was done on different network condition by varying number of edge and IOT devices. It was found that proposed Dynamic Learning Model of Edge Load Balancing (DLM-ELB)  model has reduces the makespan time of job by 1.403% and improve edge utilization by 2.22% as compared to comparing models.

##plugins.themes.bootstrap3.article.details##