In this communication, we introduce a channel model for personal radar applications where a millimeter-wave (mm-wave) massive array is required to scan the environment and to reconstruct a map of it. The analysis is based on measurement campaigns, in a corridor and in an office room, performed using mm-wave massive arrays. In such a context, we aim at characterizing the channel from both a temporal and an angular perspective by exploiting a 2D CLEAN-like technique to extrapolate the multipath components and a K-means algorithm for clustering, where centroids statistics depend on the environment contour. The obtained channel model can be exploited for mapping algorithms based on backscattered radar measurements