Simulation of Mobile Wireless Networks with Accurate Modelling of Non-linear Battery Effects

For the simulation of protocols and algorithms of mobile devices, an ideal energy source, i.e. a battery with linear charge and discharge characteristics, is often assumed. However, real batteries like lithium-ion cells show nonlinear behavior, taking Rate Capacity Effect and Recovery Effect into account. The battery model presented by Rakhmatov and Vrudhula models non-linear battery behavior and can be utilized for lifetime optimization strategies since it provides a formal cost metric. A drawback of this lifetime optimization algorithm is that the load profile of the battery has to be known completely in advance. An estimation of battery lifetime at simulation runtime is not feasible. We present an algorithm for a runtime estimation of battery lifetime of lithium-ion cells. Our algorithm allows the integration of a non-linear battery model into network simulation environments for mobile devices. As an example, we describe the integration of our algorithm for battery lifetime estimation into a popular network simulation tool.