Skip to content

mojtabagolab10/MHSC-Metaheuristic-Cloud-Energy-Optimization

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 

Repository files navigation

MHSC-Metaheuristic-Cloud-Energy-Optimization

Hybrid Bat-Genetic metaheuristic for energy-aware workflow scheduling in cloud datacenters (tri-objective optimization: energy, throughput, execution time).

🔗 DOI

https://doi.org/10.1016/j.future.2024.107624

Published in: Future Generation Computer Systems (Elsevier)


🌍 Overview

Cloud datacenters currently account for nearly 1% of global electricity consumption and approximately 2% of worldwide carbon emissions.

As scientific and enterprise workflows grow in scale and dependency, optimizing energy consumption without sacrificing throughput or execution time becomes a critical challenge.

MHSC introduces a sensitivity-driven hybrid metaheuristic framework that merges Genetic Algorithm (GA) and Bat Algorithm to achieve tri-objective optimization in cloud workflow scheduling.

The method simultaneously optimizes:

  • Energy Consumption
  • Execution Time
  • Throughput

🚨 The Core Challenge

Cloud workflow scheduling is:

  • NP-Complete
  • Multi-objective
  • Energy-sensitive
  • Dependency-constrained

Traditional schedulers optimize one parameter at the expense of others.

Energy ↓ often means Execution Time ↑
Throughput ↑ often means Energy ↑

A balanced solution requires intelligent global search beyond single-objective heuristics.


💡 The MHSC Innovation

MHSC integrates:

  • Hybrid Bat + Genetic search
  • Sensitivity rate computing
  • Linear programming for inertia generation
  • Intelligent threshold detection
  • Workflow input clustering
  • Deadline-aware task mapping
  • Neighborhood-based quality link prediction

The hybridization enables:

  • Faster escape from local optima
  • Reduced iteration counts
  • Improved convergence speed
  • Balanced tri-objective trade-off

📊 Experimental Improvements

Compared to baseline approaches:

  • Energy consumption improved by 6.4%
  • Execution time reduced by 8.1%
  • Throughput increased under high node dependency
  • Better scalability as workflow size grows

Performance advantage increases as task dependencies increase.


🔬 Technical Architecture

The MHSC pipeline:

  1. Input classification based on deadline constraints
  2. Hybrid Genetic distribution of workflow tasks
  3. Bat-based inertia and exploration control
  4. Sensitivity-driven threshold adaptation
  5. Multi-objective fitness evaluation

Objective Function balances:

  • Energy Model
  • Execution Time Model
  • Throughput Maximization

🌱 Why It Matters

Green computing is no longer optional.

Datacenters face:

  • Energy cost escalation
  • Carbon regulation pressure
  • SLA constraints
  • Resource scalability demands

MHSC provides a sustainable scheduling framework for:

  • Scientific workflows (astronomy, bioinformatics, physics)
  • Enterprise cloud systems
  • DevOps pipelines
  • HPC cloud platforms

🧠 Keywords

Cloud Computing • Workflow Scheduling • Metaheuristic • Bat Algorithm • Genetic Algorithm • Multi-objective Optimization • Green Computing • Energy-aware Systems • Sustainable Datacenters • High Performance Computing


📄 Publication Details

Title: MHSC: A meta-heuristic method to optimize throughput and energy using sensitivity rate computing

Authors: A. Ghorbannia Delavar
R. Akraminejad
F. Kazemipour

Journal: Future Generation Computer Systems

DOI: 10.1016/j.future.2024.107624


📚 Citation (APA)

Ghorbannia Delavar, A., Akraminejad, R., & Kazemipour, F. (2025). MHSC: A meta-heuristic method to optimize throughput and energy using sensitivity rate computing. Future Generation Computer Systems. https://doi.org/10.1016/j.future.2024.107624


📚 Citation (BibTeX)

@article{MHSC2025, title={MHSC: A meta-heuristic method to optimize throughput and energy using sensitivity rate computing}, author={Ghorbannia Delavar, Arash and Akraminejad, Reza and Kazemipour, Farhad}, journal={Future Generation Computer Systems}, year={2025}, doi={10.1016/j.future.2024.107624} }


🏫 Affiliation

Department of Computer Engineering
Payame Noor University, Tehran, Iran


⚖ License

This repository is created for academic indexing and visibility purposes.
All publication rights belong to the journal publisher.


⭐ Impact

This repository contributes to structured digital indexing of high-impact cloud optimization research (2022–2025) to enhance discoverability across GitHub, search engines, and academic ecosystems.