The "machine to machine" communication paradigm will become a central element for mobile networks. This paradigm can be easily constructed by a contact-based network, notably a disruption/delay tolerant networks (DTN). To characterize a DTN, we can use the Inter-contact time among the nodes. The better understanding of inter-contact time (ICT) has practical applications on the tuning of forwarding strategies, and hence in the quality of the User Experience. Nevertheless, the fine tuning of those parameters is tight to a set of assumptions about the regularity of movement or periodicity of patterns in an usually non complete and cumbersome statistical analysis. That is why in a dynamic environment where we cannot assume any previous information the tuning of parameters is usually overestimated. In this work we study how monitoring can help to adapt those parameters to give a better understanding of both natural evolution of the network and non periodical events.