Workload prediction of wireless sensor network based on linear prediction
lue Ik Hong, Wong Kiing Ing, Nader Nassif Barsoum · 2012
Wireless sensor network has been widely used in remote monitoring applications. However, energy efficiency of sensor nodes becomes an important issue as there are powered by batteries. This is because most of the sensor nodes are inaccessible once they have been deployed in the physical environment. Therefore, a good energy saving scheme of sensor nodes is essential in order to provide a long operational life for at least up to 2 years. Dynamic Voltage Scaling (DVS) scheme can be used to dynamically scaling the voltage level based on CPU's workload in order to save energy. In this paper, a linear prediction model is proposed to adjust the voltage level relating to its workload.