Fault Analysis and Identification in Ubiquitous Networks using Cognitive Agents Approach

The ubiquitous network has set of heterogeneous networks such as wireless sensor networks, cellular networks, mobile adhoc networks, Bluetooth networks, etc. Each heterogeneous network communicates with other to perform ubiquitous computation, but there is a need of a fault tolerance mechanism. This kind of mechanism is used to provide continuous services to users in the presence of faults. The faults may be node, link, and memory level faults in the ubiquitous networks. Such levels faults are lead to the failure of the networks. However, there is a need for substantial time and sufficient resources for providing reliability in fault tolerance mechanisms. In this work, it has been proposed that cognitive agent based fault analysis and identification of types of faults using support vector machine algorithm. Also, the various types of cognitive agents are defined as network cognitive agents, heterogeneous node cognitive agents, gateway cognitive agents, ubiquitous cognitive agents, and fault analysis and identification cognitive agents. Each agent is communicating with other for analyzing the types of faults. Finally, the proposed work is simulated in terms of delay, probability analysis and identification, throughput, packet delivery ratio, and agent computation overhead.