Habitat Monitoring using wireless sensor networks

Abstract

The deployment of wireless sensor networks in habitat monitoring is gaining importance as the manpower cost is increasing day by day. The positions of the cattle is detected and if detections at successive time intervals indicate that the position of the cattle is hardly changing, there is a chance that the cattle is sick or injured and a warning message is issued to the owner of the farm. The positions have been estimated using the Direction of Arrival estimation by maximum likelihood and MUSIC (MUltiple SIgnal Classification) algorithms. The performance of the system has been evaluated in terms of minimum root mean square error and probability of resolution. The results of direction of arrival have been improvised using the averaging process and the multimodal problem has been optimized using differential evolution. Since Direction of Arrival estimation gives only the direction and not the precise position, the phase detection of the signals is done to differentiate different positions having the same direction of arrival. Finally analysis is done regarding the movement of cattle. If it is found that they do not move and occupy the same position for a considerably large period of time, warning message is issued to the owner of the farmland.