A wireless sensor network is an ad-hoc network connecting small devices equipped with sensors. Such networks are self-organized and independent of any infrastructure. The deployment of a WSN is possible in areas inaccessible to humans, or for applications with a long lifetime requirement. Indeed, devices in a wireless sensor network are usually battery-powered, tolerate failure, and may use their own communication protocols, allowing them to optimize the energy consumption. The main application of WSNs it to sense the environment at different locations and aggregate all the data to a specific node that logs it and can send alerts if necessary. This task of data aggregation is performed regularly, making it the most energy consuming. As reducing the energy consumed by sensor is the leading challenge to ensure sustainable applications, we tackle in this thesis the problem of aggregating efficiently the data of the network. Then, we study lifetime evaluation techniques and apply it to benchmark existing energy-centric protocols.