Abstract
Network protocols for wireless sensor networks should be evaluated in terms of life time in a whole system. There exists power variation node due to the manufacturing variation. In this paper, we develop a power model, in which we consider threshold-voltage variation. We implement it to QualNet in order to evaluate the impact against a life time. The simulation results show that the conventional model has overestimated the life time longer than our model when nodes are randomly deployed. In addition, the network life time is extended by 19.3% compared with the conventional model by an optimum deployment.