Time constrained fault tolerance and management framework for k-connected distributed wireless sensor networks based on composite event detection

Abstract

Wireless sensor nodes themselves are exceptionally complex systems where a variety of components interact in a complex way. In enterprise scenarios it becomes highly important to hide the details of the underlying sensor networks from the applications and to guarantee a minimum level of reliability of the system. One of the challenges faced to achieve this level of reliability is to overcome the failures frequently faced by sensor networks due to their tight integration with the environment. Failures can generate false information, which may trigger incorrect business processes, resulting in additional costs. Sensor networks are inherently fault prone due to the shared wireless communication medium. Thus, sensor nodes can lose synchrony and their programs can reach arbitrary states. Since on-site maintenance is not feasible, sensor network applications should be local and communication-efficient self-healing. Also, as per my knowledge, no such general framework exist that addresses all the fault issues one may encounter in a WSN, based on the extensive, exhaustive and comprehensive literature survey in the related areas of research. As one of the main goals of enterprise applications is to reduce the costs of business processes, a complete and more general Fault Tolerance and management framework for a general WSN, irrespective of the node types and deployment conditions is proposed which would help to mitigate the propagation of failures in a business environment, reduce the installation and maintenance costs and to gain deployment flexibility to allow for unobtrusive installation.