Decision Support Systems for IoT Based Infrastructures

Abstract

Computation offloading divides enormous computational tasks across numerous computer resources, circumventing hardware limits. Edge computing can employ vast quantities of data, individual preferences, and clever algorithms by offloading smart models to high-performance cloud servers. We propose a getaway-centric IoT system to enable intelligent and autonomous IoT devices at the computer infrastructure's edge. Edge computing manages IoT devices by selecting and applying the best control factor from a pool of intelligent services. The cloud-based intelligent service engine provides intelligent services by offloading intelligence and optimisation algorithms. Thus, the gateway's decision-making model may choose the best alternative. Resource virtualization-based gateway-based device management facilitates user monitoring and visualisation in the proposed IoT system. The gateway evaluates context-based profiles to enable real-time connection with intelligent services and dynamic application of the appropriate control factor to the physical device using the virtual resource. We propose two smart models to learn characteristics of a user's home environment using deep learning and build inference models for the intelligent service engine to optimise energy usage with the recommended IoT system. Inference methods forecast heater energy usage. Heaters set the environment. The decision-making algorithm also lowers the heater setting based on the two use numbers, reducing energy consumption and producing a user-desired environment.