Energy Conservation of Sensor Nodes using LMS based Prediction Model

Energy conservation is one of the most concentrated research area in wireless sensor networks. Sensor nodes deployed on large scale in an unmanned area is used for environmental monitoring. But sensor nodes work on battery power and hence are energy constraint devices. In such scenario, energy of nodes can be conserved by reducing number of transmissions occurring from sensor nodes to base station or intermediate nodes. Applications in which environmental quantity is gradually changing over time may be transmitted only if marginal changes occur. This paper proposes a data prediction model that predict data without using a priori knowledge. Least mean square algorithm is used to implement prediction model with self-adaptive step size. This work achieves improved step size and a data reduction about 94 %. This model can be used to increase lifetime of every node in the wireless sensors network.