Probability of Cluster Header Suitability in Wireless Sensor Networks


  • Eun Hwa Kim


Background/Objectives: Clustering techniques for wireless sensor networks are one of the technical methods for efficient energy management and regional distributed network management. The most basic approach to clustering is how to organize headers and members.
Methods/Statistical analysis: In this paper, the probability of cluster header suitability was calculated based on the coordinate value of sensor node, header candidate node and sink node. Based on this probability value, a method was proposed to be a cluster header and to form a distance-based member within the communication radius of the cluster header.
Findings: The selection and configuration of cluster headers and members was confirmed through distance-based C# simulations, and the connection to sink nodes was confirmed through cluster headers. This experiment selected cluster head with high probability of header suitability as a path close to the sink node, and formed cluster members around the cluster header. By adjusting the probability values when configuring a cluster member, the experiment confirmed that the network could form a cluster closer to a straight line. In addition, the higher the probability of cluster head suitability limit when configuring a cluster, the shorter the path from the sensor node to the sink.
Improvements/Applications: The method proposed in this paper can form a cluster independently by calculating each sensor node's probability value of header suitability based on coordinate value, and can transmit data at a short distance to sink node based on distance.