Fault Tolerant Operations in Wireless Sensor Networks

Abstract

Wireless Sensor Network (WSN), is a spatially scattered group of energy-consuming sensing nodes commonly known as sensors or just nodes that are supposed to monitor environmental changes. These environmental changes can be in terms of motion, sound, humidity, soil moisture, temperature, fire, etc., and thus can be utilized in various domains like intrusion detection, livestock monitoring, agriculture, home or forest fire detection, health monitoring, industrial health monitoring, logistics and many more. The advancement in wireless sensor networks has enabled the concept of smart cities and is playing a pivotal role in the ubiquitous and pervasive computations, thus enabling a smart ecosystem. newlineThough the WSN popularity has seen an upward trend, the system suffers from various limitations, that must be addressed. Energy-efficient and fault-tolerant operations require special attention. Energy-aware operations are crucial to the success and popularity of the WSNs. Also, the sensor networks are inherently fault-prone, and various algorithms have been and are being designed for the same. The thesis addresses these problems and presents an approach to deal with the same. newlineFurther, wireless sensor networks are deployed in friendly as well as hostile environments. The limitations associated with WSNs in the case of friendly deployment are relatively easier to handle when compared with hostile deployments. The basic working principle of wireless sensor networks rotates around data aggregation, transmission, and reception of data followed by processing of received data. Generally, the processing of the received data for decision making purposes is done at a centralized station which may be or may not be inside the monitored area. This makes it essential to properly utilize the available network energy, especially in the case of hostile deployments, where charging or replacing the dead or faulty sensing nodes might not be feasible. Moreover, as the sensors are prone to faults and their behavior can t be predicted, this can