The technology of mobile communicating objects (sensors, smartphones, PDAs, netbooks, etc) is booming. We can consider the use of these objects in several areas, particularly in health field, in order to collect medical information. Most of current medical solutions rely their data collection on two approaches. The first approach is based on the assumption of constant connectivity. In such strong requirements, physiological data collection towards remote monitoring center is performed in real-time, and for an instant medical analysis by physicians. In the second approach, data are collected in stand-alone wearable storage devices, and for a postponed medical analysis. Given the mobility of subjects wearing these objects and the shortrange radio transmissions, connectivity within, and between these objects and the collection stations is not continuously ensured. This involves the use of an opportunistic communication approach to overcome the shortage issues of connectivity. Indeed, the use of Disruption-Tolerant Network protocols (or opportunistic communication) has a significant contribution to applications where access to network infrastructures is costly, difficult or impossible. In this thesis, we discuss the possibility of not having a permanent transmission means to monitor mobile persons in e-health field, but on the contrary, an opportunistic approach is possible and interesting: it can provide good results at low cost. As an application for this communication scheme, we proposed several medical scenarios ranging from easy to challenging ones. We first dealt with the medical supervision of cardiac activity of marathon runners as the reference scenario. In this scenario we collect electrocardiogram data of runners relaying on wireless technologies between their wearable sensors and many deployed base stations. The first version propose a lighter solution in terms of costs and energy consumption. It relies on ZigBee communication protocol. The second version relies on Wi-Fi and cellular networks, which the deployment is nowadays very wide. A collect system was developed, deployed, and tested in sporting events coping with outdoor conditions and harsh environments. Experiments results confirm the advantage of this new opportunistic approach in biomedical monitoring of mobile subjects, by relying on offthe- shelf devices and technologies. From the first studied scenario, many variants could be declined as general-purpose and less constrained medical scenarios. Thus, we secondly dealt with daily medical supervision of mobile subjects by relying on public Wi-Fi hotspots and cellular networks. In conclusion, different explored medical scenarios present a good introduction to the new opportunistic communications for biomedical monitoring applications.
