Distributed fault diagnosis with dynamic cluster-head and energy efficient dissemination model for smart city

Abstract Rapid increase in urbanization demands sophisticated monitoring for improving the lifestyle of people. Unattended supervision and data acquisition are the primary requirements of any smart city project which is fulfilled by wireless sensor networks (WSNs). However, the rigorous and rapidly changing metropolis tends to make the network unreliable. Reliability in the data is achieved through the application of fault tolerant protocols. All the literatures use the robust cluster-head (CH) as the basis for the protocol. This is unrealistic and could lead to unintended consequences upon CH node failure. In this paper, we propose a rational dynamic CH scheme, where the CH is prone to errors just like other nodes in the network. The LEACH protocol is modified to incorporate smart dynamic CH selection at the end of each round based on the residual energy and sensor readings. This ensures minimal energy waste and consolidates a green model for smart cities. An efficient dissemination scheme is proposed to facilitate the attainment of global view of the network among all the CHs at the end of r rounds. The proposed protocol gives better classification accuracy and reduced false alarm rates for varied probabilities of intermittent and permanent faults as compared to LEACH.