Clustered Binary Consensus in Sensor Motes

Abstract:
In this work, we extend and adapt the binary consensus algorithm to operate using clusters in a wireless sensor network. Binary consensus is a decision making algorithm used to cause distributed entities to agree on the majority opinion held by the group when posed with a true or false question. Clustering is a common technique used in WSNs to increase the overall lifetime of the network by reducing energy consumed due to communication. Our clustered binary consensus implementation is tested using a large WSN testbed, and the experimental results show that incorporating clustering into the binary consensus algorithm reduces the time required for motes to reach consensus and lowers the number of packets sent in the network by approximately 94%.