Variable-Length Multiobjective Social Class Optimization for Trust-Aware Data Gathering in Wireless Sensor Networks

Area of Science:

• Computer Science
• Network Engineering
• Optimization Algorithms

Background:

• Data gathering in Wireless Sensor Networks (WSNs) is crucial for Internet of Things (IoT) integration.
• Large-scale WSN deployments face efficiency challenges and security threats impacting data reliability.
• Trust in data sources and routing nodes is essential for reliable data collection in WSNs.

Purpose of the Study:

• To develop a multiobjective optimization method for data gathering in WSNs.
• To jointly optimize energy consumption, travel time, cost, and trust in WSN data collection.
• To propose a modified Social Class Multiobjective Particle Swarm Optimization (SC-MOPSO) algorithm.

Main Methods:

• Introduction of a modified SC-MOPSO featuring application-dependent interclass operators.
• Incorporation of solution generation, rendezvous point management, and class-based movement strategies.
• Utilization of the Simple Additive Weighting (SAW) method from multicriteria decision-making (MCDM) to select solutions from the Pareto front.

Main Results:

• SC-MOPSO and SAW demonstrate superior performance in terms of solution domination.
• SC-MOPSO achieved a set coverage 0.06 dominant over NSGA-II, with NSGA-II showing only 0.04 dominance over SC-MOPSO.
• The proposed method exhibited competitive performance compared to NSGA-III.

Conclusions:

• The modified SC-MOPSO effectively addresses the multiobjective nature of WSN data gathering.
• The integration of trust as an optimization objective enhances data reliability in WSNs.
• The SC-MOPSO-SAW approach provides a robust solution for optimizing complex WSN data collection scenarios.