الفهرس 
abstractOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
Wireless sensor networks (WSNs) are emerged as an important technology for different purposes. The behavior of a WSN is characterized by the type of its application Thus; in order to guarantee the network quality of service it is essential for the WSN to be able to detect failures and perform something akin to healing, recovering from events that might cause some of its parts to malfunction. Faults are likely to occur frequently and unexpectedly in sensor networks compared to traditional network. The networks must exclude the faulty sensors to ensure the network quality of service. Identify the faulty sensor nodes are not trivial at all because of the existing challenges. Sensor nodes are powered by batteries, which are considered as limited resources. It is very expensive for the base station to collect information from every sensor and identify faulty sensors in a centralized manner.
The network is said to be bi-connected; if any two nodes in a network have at least two node-disjoint paths. In a bi-connected network, if a fault happens at any node, the network is still connected, i.e., all affected communication links can be re-routed. In other words, the removal of a node and associated links from a bi-connected network does not partition the network. On the other hand, if a network is partitioned after the removal of a node and associated links, then the node is called a cut-node or articulation point so the detection of the articulation points is a serious process to keep the connectivity of the network. A graph is bi-connected if and only if it has no articulation point.
A network of sensors is considered to be connected only if there is at least one path between each pair of nodes in the network. Connectivity depends primarily on the existence of paths. Main concern in all networks is connectivity of the network nodes, so the failure of an articulation point breaks the connectivity of the network. This thesis proposes a fault detection algorithm for wireless sensor networks by detecting the articulation points before they fail and provides fault tolerance by mechanism of rotation using back propagation neural network. Traditional methods of detecting cut vertices are centralized and are very difficult.