For providing balanced resources to the Internet of Things (IoT) cloud users, Task Scheduling (TS) plays an essential role. But, energy consumption and execution timeare increased by scheduling tasks on the cloud resources without proper analysis of task strategy.For overcoming this setback, this paper proposes the Chessboard-K-Prototype Algorithm(Ch-KPA)-based task grouping with the RICE-based Earliest Deadline First (R-EDF) scheduling. Primarily, the bag of tasks is taken as input from which attributes are extracted. Next, using fuzzy, the tasks are classified as small, large, and medium based on the task length. Subsequently, based on the task attributes, the medium and large tasks are grouped as sequential and parallel using Ch-KPA. Next, based on first in first out, the tasks are added to the queue. Subsequently, the Logistic Chaotic map-based Giant Trevally Optimization (LC-GTO) selects the single optimal Virtual Machine (VM) for the small task. Likewise, the optimal container and VMs of multi-cloud are selected for sequential and parallel tasks. Meanwhile,the availability of selected VM in the VM monitoring layer is determined by Drop-connect-Random Translation Multi-Layer Perceptron (DRT-MLP) utilizing a feature updated table. If the VM is not in the updating state, the task is mapped to that VM.Lastly, the R-EDF scheduler dynamically schedules the task to the selected VM centered on the deadline. The proposed approach’s efficiency is proved by the experimental outcomes.

Virtual Machine (VM),Chessboard-K-Prototype Algorithm (Ch-KPA), Drop-connect-Random Translation Multi-Layer Perceptron (DRT-MLP), task scheduling, Earliest Deadline First (EDF).