QC-MLQ - A Novel Quantum Computing-based Multi-Level Queue Scheduler for Hadoop-integrated Fog Computing Environment

Purpose: This paper introduces a novel Quantum Computing-based Multi-Level Queue (QC-MLQ) Task Scheduler (TS) in the Hadoop-integrated Fog Computing Environment (FCE). The QC-MLQ schedules tasks using MLQs and finds an optimal number of queues needed for best results. Performance Metrics: Metrics related to time, efficiency, and resource consumption are used to find the optimal amount of queues using QC-MLQ. Time parameters include cycle and wait time, measured in seconds (s). Efficiency metrics include the Task Completion Rate (TCR) measured in processes per microsecond (P/μs) and CPU Utilization (CPU-UTZ) percentage. Resource consumption metrics include cost in dollars ($) and Energy-UTZ measured in watts (w). Experiment: The experiment was conducted in an FCE simulator where Google cluster raw big data tasks were computing on ten Virtual Machines (VMs) using the QC-MLQ scheduler with different queues. Results: The QC-MLQ scheduler provides optimal performance using a queue size of five, half the number of VMs, providing cycle and wait time of 121005.6 s and 85405.6 s. Additionally, a TCR of 11.6928 μ/s with CPU-UTZ of 99.86% and cost of $21350 consuming 128100 w of Energy-UTZ was achieved with this queue size. Findings: The results promote using the QC-MLQ scheduler with half the number of queues concerning the number of VMs in the FCE. Using lesser or higher queues yields poor results. Future Work: As a part of future work, an architecture of Hadoop-integrated FCE with QC-MLQ with Neural Networks is presented for improved FCE for offering better performance.