In WSN (Wireless Sensor Network), the sensor nodes may become faulty due to low battery power or some other physical defects. The faulty nodes may badly affect the network performance while trying to reach an agreement on an event. This effectively leads to deviation from the desired outcome, ineffective utilization of network bandwidth and unproductive computational overhead. In this work, we propose a CA (Cellular Automata) based scheme that can efficiently identify the faulty nodes of a WSN in operation. The scheme is developed around the SACA (Single Attractor CA) to enable diagnosis of the network region with affected nodes, without consuming massive computational overhead and bandwidth.