A performance-aware dynamic scheduling algorithm for cloud-based IoT applications

Abstract

Cloud computing has been employed for supporting storage and handling of Internet of Things (IoT) data. There is an increasing demand for IoT framework to provide services with fast processing time and less delay to offer latency sensitivity real-time applications like disaster management in smart homes. IoT process is mostly comprised of scheduling techniques that makes it hard to self-adapt, self-configure to respond with performance aware on environment changes. Existing scheduling techniques of IoT applications are not based on allocating tasks through sleep modes, which unavoidably lead to more power consumption and longer time delays. Consenting sensor devices and applying separate queueing to a sensor device that varies differently in their capabilities are increasingly significant. In this work, a dynamic management framework for IoT devices in cloud (DMFIC) algorithm is proposed to evaluate and schedule requests and sensor data, which allow coordinating huge data with high time-based resolution in a cost-effectual manner through anticipating various queues for sending an appropriate notification to users. A smart home application was used to demonstrate the proposed framework. The experimental result shows that the DMFIC algorithm gives an average of 5% higher processing time and 0.2% less delay compared to other IoT services and can efficiently manage sensor data in cloud.