Directionally-Enhanced Binary Multi-Objective Particle Swarm Optimisation for Load Balancing in Software Defined...

Area of Science:

• Computer Science
• Network Engineering
• Optimization Algorithms

Background:

• Internet of Things (IoT) networks require efficient task execution load balancing.
• Software-Defined Networking (SDN) offers intelligent algorithms for optimization.
• Existing load balancing methods optimize only one or two aspects (makespan, energy, cost).

Purpose of the Study:

• To propose a joint mathematical formulation for cloud computing load balancing.
• To introduce two multi-objective particle swarm optimization (MP) models: Distance Angle Multi-objective Particle Swarm Optimization (DAMP) and Angle Multi-objective Particle Swarm Optimization (AMP).
• To simultaneously optimize reliability, energy consumption, cost, and execution time in IoT task load balancing.

Main Methods:

• Developed DAMP and AMP models that probabilistically combine crowding distance and crowding angle for solution selection.
• Generated binary variants: Binary DAMP (BDAMP) and Binary AMP (BAMP).
• Compared DAMP and AMP models against standard models (BMP, BDMP, BPSO, DMP, PSO) using multi-objective optimization (MOO) functions and a proposed load balancing model.

Main Results:

• The proposed DAMP and AMP models demonstrated superior performance compared to existing state-of-the-art models.
• BDAMP and BAMP showed significant improvements in optimizing multiple load balancing aspects simultaneously.
• The study confirmed the effectiveness of meta-heuristic algorithms in cloud network management.

Conclusions:

• The developed DAMP and AMP models offer a robust solution for multi-objective load balancing in IoT networks.
• This research facilitates the integration of meta-heuristic approaches into the management layer of cloud networks.
• The findings provide decision-makers with optimized solutions and a range of candidate solutions for load balancing challenges.