Host utilization prediction using Taylor Kernel Convolutional Neural Network (TKCNN) and workflow scheduling for smart irrigation cloud data centers

Abstract

One of the most significant economic sectors in these nations is agriculture, which emphasises the significance of controlling the water resources at hand to maintain the survival of this industry. Sensor and cloud computing (CC) technology are widely used in many real-time applications, particularly in agriculture. To ensure Quality of Service in the CC environment, Virtual Machine (VM) Migration as part of VM Consolidation is very important (QoS). Predicting host utilization utilizing utilization history and lowering energy consumption while satisfying Service Level Agreements (SLAs) is a challenging task as a result of the large fluctuation in the consumption of cloud resources and the dynamic workloads.CC technologies are widely used in many real-time applications, particularly in agriculture for smart irrigation cloud data centers difficult task. To solve this issue proposed work introduced a Taylor Kernel Convolutional Neural Network (TKCNN) algorithm, which trains the past history for host usage prediction in smart irrigation cloud data centers. The Optimum Energy and Resource Aware Workflow Scheduling- Host Utilization Prediction (OERES-HUP) method, which attempts to schedule the task workflow to VMs engaging in computing, was influenced by the well-known Fuzzy Multi-Verse Optimizer (FMVO) algorithm. Additionally, it focuses on dynamically deploying and underplaying VMs in accordance with job requirements. The effectiveness of scheduling algorithms is evaluated using key metrics including Resource Utilization (RU), Energy Consumption or Task (ECT), Total Energy Consumption (TEC), Makespan, and Execution Time per Task (ETT).The outcomes are a good representation of how effective the suggested algorithm is compared to current techniques. To test the efficacy of the suggested model, the cloud environment is simulated using the Cloud Sim simulator.