Unlicensed bands offer great opportunities for numerous wireless technologies, including IEEE 802.11-based systems, 4G Licensed-Assisted-Access (LAA), and 5G New Radio Unlicensed (NR-U) networks. Achieving harmonious coexistence between these technologies requires real-time adaptation of their channel access, which can be facilitated by artificial intelligence (AI) and machine learning (ML) techniques. However, to leverage such techniques, we need to characterize the state of unlicensed wireless channel and the dynamics of the coexisting systems. In this paper, we introduce the concept of Sensing Fingerprint (SF) profile to characterize the state of coexisting networks and track their dynamics over unlicensed bands. We conduct extensive experiments to show the effectiveness of SF profile in tracking key network dynamics, including sensitivity thresholds of contending devices, their mobility, traffic loads, and other channel access parameters. AI-and ML-based controllers can utilize this tool to model the state of coexisting networks and track their dynamics.