A Distributed Clustering Algorithm Guided by the Base Station to Extend the Lifetime of Wireless Sensor Networks

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Clustering algorithms are crucial for reducing energy consumption in Wireless Sensor Networks (WSNs).
• Cluster head selection significantly impacts network performance and energy balance.
• Existing centralized and distributed algorithms have limitations in efficiency and data requirements.

Purpose of the Study:

• To propose a novel distributed clustering algorithm for WSNs that enhances network lifetime.
• To dynamically form clusters and balance energy consumption among nodes.
• To introduce a hybrid approach occasionally supported by the Base Station for network reconfiguration.

Main Methods:

• A distributed clustering approach is presented, with occasional Base Station (BS) support.
• The BS sends three messages to reconfigure the 'skip' value, adapting to network status.
• Nodes use a fuzzy-logic system, specifically a Takagi-Sugeno-Kang model, to determine cluster head suitability.

Main Results:

• The proposed algorithm dynamically forms clusters and balances energy consumption.
• The fuzzy-logic system effectively manages cluster head selection.
• Simulation results demonstrate a significant extension of network operability compared to existing methods.

Conclusions:

• The novel distributed fuzzy-logic based clustering algorithm effectively extends WSN operational lifetime.
• The hybrid approach combining distributed decisions with occasional BS support offers a robust solution.
• The use of a Takagi-Sugeno-Kang fuzzy model provides an advantageous alternative to Mamdani models in this context.