Parameter Selection and Performance Comparison of Particle Swarm Optimization in Sensor Networks Localization.

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Localization is crucial for wireless sensor networks (WSNs).
• WSNs face constraints in memory, computation, and energy.
• Particle Swarm Optimization (PSO) is effective for WSN localization.

Purpose of the Study:

• To survey PSO variants and localization algorithms for WSNs.
• To determine optimal parameter and topology selections for PSO.
• To comprehensively compare PSO algorithm performance.

Main Methods:

• Surveyed popular PSO variants and WSN localization algorithms.
• Conducted extensive simulations for parameter and topology selection.
• Performed comprehensive performance comparisons of different PSO algorithms.

Main Results:

• Identified PSO with constriction coefficient and ring topology as optimal.
• This configuration outperformed other PSO variants and topologies.
• Achieved superior performance compared to the second-order cone programming algorithm.

Conclusions:

• PSO with constriction coefficient and ring topology is highly effective for WSN localization.
• Provides guidance for selecting optimal PSO parameters and topologies.
• Offers a high-performance alternative to existing localization methods.