In this context, it is useful to highlight how the WSN approach perfectly fits the perspective of ambient intelligence in large, possibly city-wide scenarios. First of all, let us recall that wireless sensors are usually very small nodes with a limited set of capabilites. They usually incorporate by default a set of environmental sensors (temperature, light, humidity), a programmable general-purpose microcontroller, some RAM/ROM units and a radio transceiver unit. Recent implementations of wireless sensors also incorporate a serial connection (e.g., through a USB interface), which is used for programming, but can also be employed for power supply. By means of wireless communications, the sensors can be organized into a network that cooperates in tasks including, but not limited to, environmental monitoring and control, alarm dissemination, localization and tracking of moving objects, and ambient intelligence in general.

The latter definition encompasses a number of applications involving complex interactions between a user, the network, and the environment, possibly including some specific actuation devices that might be interfaced to the sensors in order to carry out a certain task. For example, a wireless sensor with a hardcoded identity footprint could communicate with the network so that its movements can be tracked: such a sensor could, e.g., be attached to a piece of luggage, to make sure that it is routed toward the correct destination by airport logistics and that, in case of errors, its route can be quickly reconstructed, in order to allow timely recovery.

In WSNs, the energy efficiency aspect is of paramount importance. In fact, it should be noted that wireless sensors are mostly powered through external batteries or battery packs. As this supply is limited, energy efficient operations and communications must be pursued, so that battery replacements on the sensors are as infrequent as possible, in order to keep power supply and maintenance manpower costs low. For this reason, networking protocol design as well as application design must carefully account for the limited processing and storage capabilities of the microcontroller: therefore, the further delay that would be caused by complex processing tasks should be explicitly considered in light of, e.g., the timing of communication protocols. In the following, we give more details on the typical configurations and peculiar features of WSNs, in order to provide a wider characterization of their uses and behaviors. In general, a WSN is characterized by being:partly stationary: the Dacomitinib vast majority of the deployed nodes are fixed and constitute the infrastructure of the network.