Wireless communication technologies are used to collect information from sensitive, hostile and inaccessible environments. They are used in both military and civil applications that include environmental, medical, military and industrial fields. Routing data to a processing center or a base station requires mechanisms for energy conservation at the end of the prolonged lifetime of the network. The simulation in this case is very constrained by the high density of the network. Existing tools cannot simulate large networks with millions of sensors. In this paper, we propose a new method using statistical regression analysis in order to predict the energy consumption and the lifetime of a wireless sensor network with hundreds or thousands of sensors by simulating smaller networks. We have validated the proposed method using a Revised LEACH protocol. Indeed, this method can be used for other protocols and other kind of simulations with the purpose of evaluating a specific parameter.