An iterative approach for optimal number of balanced clusters and placement of cluster heads in WSN with spatial constraints

G Chethana, KV Padmaja - 2019 4th International Conference …, 2019 - ieeexplore.ieee.org
2019 4th International Conference on Recent Trends on Electronics …, 2019ieeexplore.ieee.org
Sensor nodes in Wireless sensor Network collect different types of data in numerous fields
such as healthcare, precision agriculture, military, and so on. The spatial constraints due to
topographical or some other adverse conditions may restrict both the location of sensor
nodes and cluster heads inside well defined planar regions within the WSN, while some
other regions may be more amenable for easy deployment of sensors nodes and cluster
heads. A new method of determining the optimum number of balanced clusters and the …
Sensor nodes in Wireless sensor Network collect different types of data in numerous fields such as healthcare, precision agriculture, military, and so on. The spatial constraints due to topographical or some other adverse conditions may restrict both the location of sensor nodes and cluster heads inside well defined planar regions within the WSN, while some other regions may be more amenable for easy deployment of sensors nodes and cluster heads. A new method of determining the optimum number of balanced clusters and the locations of the cluster heads in a Wireless Sensor Network with spatial constraints is presented. The optimum number of balanced clusters is determined using an iterative approach. The K-means clustering algorithm is modified to obtain balanced clusters. The cluster head locations are determined using the constrained nonlinear convex optimal solver based on `pattern search' technique. In our scheme, the cluster heads are not elected from the cluster members. Instead, higher capacity communication nodes are placed at the optimal locations obtained by the optimal solver. Silhouette simulation result shows lesser the indices better the quality. Our approach is having better quality of balanced clustering with the improvement of negative indices value from 55.78% to 17%.
ieeexplore.ieee.org
以上显示的是最相近的搜索结果。 查看全部搜索结果