Time-based localization in Wireless Body Area Networks (WBANs), has attracted growing research interest for the last past years. Nodes positions can be estimated based on peer-to-peer radio transactions between devices. Indeed, the accuracy of the localization process could be highly affected by different factors , such as the WBAN channels where the signal is propagating through, as well as the nodes mobility that bias the peer-to-peer range estimation, and thus, the final achieved localization accuracy. The goal of this paper consists in characterizing the impact of mobility and WBAN channel on the ranging and localization estimation, based on real mobility traces acquired through a motion capture system. More specifically, the ranging error is evaluated over all the WBANs links (i.e. on-body, off-body and body-to-body links), while an impulse Radio Ultra-Wideband (IR-UWB) physical layer, as well as a TDMA-based Medium Access Control (MAC) are playing on. The simulation results show that the range measurement error can be modeled as a Gaussian distribution. To deal with the gaus-sianity observation of ranging error and to provide high positioning accuracy, an adjustable extended Kalman Filter (EKF) is proposed.