DACA - Disjoint path And Clustering Algorithm for self-healing WSN

Abstract

Due to their intrinsic characteristics, Wireless Sensor Networks (WSN) are prone to failure, mainly because of their energy limitations. Considering this, the use of self-healing mechanisms becomes necessary in order to create a more fault-tolerant and robust WSN. With this problem at hand, we have proposed and developed DACA, a Disjoint path And Clustering Algorithm that increases the network lifetime through network topology control and self-healing mechanisms. By using the Collection Tree Protocol (CTP) algorithm, we first create a tree using all the initials nodes of the network, having this tree as our initial communication backbone. After this, we build a set of spatial clusters using K-means and reconstruct the tree using only the Cluster Heads (CH), therefore reducing the number of active nodes in the network. With this new subset of nodes forming a tree, we apply the N-to-1 algorithm to create disjoint paths, making the network more robust to communication failures. The experiments show that DACA considerably extends the lifetime of the network by having a set of backup nodes to support the communication network when an active node dies, while still maintaining a good coverage of the area of interest.