These sensors form a network and work together to gather data and send it to the base station. The base station acts as the control center, where the sensor data is collected for further analysis and processing. In a nutshell, spatially distributed nodes use sensors to monitor physical or ecological conditions. These nodes combined with routers and gateways form a WSN system. The WSNs topology can fluctuate from a minimal star network to a sophisticated wireless mesh network. The strength of WSN is the ability to organize a large number of small nodes that accumulate and configure themselves. A typical sensor node is built up of power unit, communication unit, processing unit and sensing unit. The prime function of WSN is to gather the sensor data from the monitored area. But sensor nodes are prone to failure unpredictably due to the environmental interference and the low power essence [1]-[4]. The faulty node introduces many errors into the network and corrupts the network. Due to faults, the data collected might be wrong or the sensor data may be lost. Hence it is vital to detect events even in the existence of erroneous sensor reading and unreliable reports. Every sensor nodes can decide its own fault status and their neighboring nodes. Sensors are battery operated and thus their performance tends to deteriorate as power is exhausted. Nodes running out of power may cause topology changes in sensor networks even without mobility. For accurate sensing, the faulty node should be restored immediately. New sensors with fresh batteries may be injected to a sensor network, which are already in use, to enhance and ensure its correct operation. One of the best solutions for this problem is to have redundant deployment and to maintain a set of backup nodes. When a sensor node fails, it is replaced with a backup node. However, if the redundant sensor is a mobile node and is situated far away from the dying sensor, a protocol should be in place to locate redundant mobile sensor and to schedule the movement of mobile sensor for replacement purposes. Different applications may require fault detection to be performed in real-time mode with low latency and high throughput. Therefore, scattered restricted and distributed algorithm for each node is highly preferred in WSN. The paper is organized as follows. Related work in the fault detection area is dealt in Section 2. Section 3 discusses the proposed mechanism .The scattered restricted faulty sensor detection algorithm is proposed in Section 4. Simulation results and Performance analysis is presented in Section 5. Section 6 concludes the paper along with future scope.
II. RELATED WORKS When the local sensors chronologically forward their decisions to a fusion center in the existence of sensor faults are focused in distributed fault-tolerant decision fusion [5].Collaborative sensor fault detection (CSFD) scheme is proposed to eradicate unreliable local decisions while performing distributed decision fusion. According to the prestructured fusion rule, assuming homogeneous neighboring decision rules and fault-free environments, an upper bound is recognized on the fusion error probability. Based on this, a new Standard was found to pursuit the defective nodes. Once the fusion center identifies the defective nodes, all corresponding local decisions are