To evaluate the energy efficiency of networking protocols, network topologies, and hardware architectures for battery-powered WSNs, researchers typically rely on mathematical modeling. These models aim to accurately represent the system's dynamic behavior and estimate how long WSNs can operate autonomously without requiring human intervention for battery replacement. The definition of the "lifetime of the WSN" can vary, from the moment when the batteries of the first node are depleted (or a fraction of nodes) to the point at which all sensor nodes are no longer operational. In this work, we take a different approach: the inclusion of a device (which we name EneMeter) that enables real-time measurement of energy consumption (approximately every 50 ms).
If the investigators already have an energy model for the WSN solution, they can use the EneMeter as an optional tool to verify its accuracy. On the other hand, if an energy model for that solution is not yet available, a relatively simple semi-empirical model can be developed based on EneMeter measurements. The associated methodology will be exemplified and discussed in this work. Moreover, if nodes in a WSN have distinct roles, such as regular sensor nodes, cluster heads, or sink nodes, the researcher can connect the EneMeter device to each node type to measure it. In statistical terms, the longer the data collection period, the more accurate the resulting energy model will be. To this end, EneMeter features an SD Card that enables the device to take continuous energy measurements for hours, days, or even months.