Skip to main content
This page provides a high-level overview of the architecture of the CloudSim-HO-Research-V2 project.

​
The Hippopotamus Optimization (HO) Algorithm

The core of this research framework is the Hippopotamus Optimization (HO) algorithm, a nature-inspired metaheuristic algorithm that models the social behavior of hippos to solve complex optimization problems.

​
Core Concepts

  • Population: A set of “hippos,” where each hippo represents a potential solution to the VM placement problem.
  • Fitness: A multi-objective function that evaluates the quality of each solution based on resource utilization, power consumption, and SLA violations.
  • Position Update: The process by which hippos move in the search space, influenced by the “leader” (the best solution) and “prey” (random solutions).

​
Implementation Details

The HO algorithm is implemented in the org.puneet.cloudsimplus.hiippo.algorithm package, with the following key classes:
  • HippopotamusOptimization.java: The main class that orchestrates the optimization process.
  • Hippopotamus.java: Represents a single solution (a hippo) in the population.
  • HippopotamusParameters.java: Encapsulates all the tunable parameters for the HO algorithm.

​
Simulation Workflow

The simulation is coordinated by the ExperimentCoordinator.java class, which manages the entire experimental workflow.

​
The Workflow Steps

  1. Configuration: The coordinator loads the experimental setup from config.properties, including the scenarios and algorithms to be tested.
  2. Scenario Generation: For each scenario, the TestScenarios.java class creates a set of VMs and hosts with specific configurations.
  3. Experiment Execution: The ExperimentRunner.java class runs the simulation for each algorithm and scenario, for a specified number of replications.
  4. Metrics Collection: During the simulation, the MetricsCollector.java class gathers all the relevant performance data.
  5. Results Writing: The CSVResultsWriter.java class writes all the collected data to CSV files in the results/ directory.
  6. Statistical Analysis: Finally, the ComparisonAnalyzer.java class performs statistical tests to compare the performance of the different algorithms.