Advances, Systems and Applications
Table 1 Comparison of state of art multi-objective optimization algorithms
From: Energy-efficient virtual machine placement in distributed cloud using NSGA-III algorithm
Algorithm | Exploration | Exploitation | Convergence Rank | Computational Complexity | Uniqueness | Cons |
|---|---|---|---|---|---|---|
Genetic Algorithm (GA) | Crossover and mutation | Selection | 4 | Exponential time complexity | Simple to implement and understand. Can handle complex problems. | Exponential time complexity may not find the optimal solution for all objectives. |
Particle Swarm Optimization (PSO) | Velocity update | Inertia weight and cognitive and social factors | 5 | Quadratic time complexity | Fast and efficient. Can find suitable solutions in a short amount of time. | Slow for small problems, may not find the optimal solution for all objectives. |
Ant Colony Optimization (ACO) | Pheromone evaporation and updating | Ant recruitment and ant selection | 6 | Quadratic time complexity | Robust and scalable. Can handle large problems. | Sensitive to the initial conditions, may not find the optimal solution for all objectives. |
Multi-objective Evolutionary Algorithm (MOEA) | Genetic operators | Selection | 3 | Polynomial time complexity | Can handle multiple objectives simultaneously. It can find the Pareto optimal set. | It can be complex to implement and understand and may not find the optimal solution for all objectives. |
Non-dominated Sorting Genetic Algorithm II (NSGA-II) | Crossover and mutation | Selection | 2 | Polynomial time complexity | Based on the concept of non-dominated sorting. Ensures that the Pareto front is always represented in the population. | It can be slow for minor problems and may not find the optimal solution for all objectives. |
Non-dominated Sorting Genetic Algorithm III (NSGA-III) | Crossover and mutation | Selection | 1 | Polynomial time complexity | It uses a niching mechanism to promote diversity. Ensures that there are enough solutions in each niche of the search space. | Can be slow for small problems, may not find the optimal solution for all objectives. |