Statistical study of energy and time costs of fault tolerance in Multilevel and EDCR protocols

Wireless Sensor Networks (WSNs) are composed usually of a huge number of sensor nodes capable of collecting and transmitting environmental data in an autonomous manner to a remote user's computer where the information may be viewed by a human, further processed and stored, or otherwise acted upon. They are generally deployed in harsh environments, where sensor nodes are prone to several failures out of which their energy depletion is the most present. Thus, the failure of a small number of nodes must not influence the network functioning. For this, including fault tolerance property in WSN routing protocols is of utmost importance. The choice of the fault tolerant protocol, and subsequently the type or the types of tolerated failures is imposed by the target WSNs application. This fault tolerance has energy and time costs that remain little studied in the scientific literature. Hence, this paper aims at studying the effect of failures density and surface density on the energy and time costs in the Multilevel and Energy aware Distributed partitioning detection and Connectivity Restoration (EDCR) protocols.