Towards a Tailored Sensor Network for Fire Emergency Monitoring in Large Buildings

Modern fire emergency systems are slowly moving from the traditional data-logging systems to a heterogeneous and dense network of wired/wireless sensors that can give a more complete view of the phenomenon. When the density of the sensors and/or the transmission rate start growing, standard and widely used communication protocols suffer from degradation in their performance, mostly due to the presence of simultaneous transmissions. Rather than proposing a new protocol that performs better than the standard ones in a set of network scenarios, this paper has a different aim. It attempts to draw conclusions from the nature of the sensed data itself, so that important spatial and/or temporal correlations can be revealed and, consequently, utilised for the future design of an indoor fire emergency-tailored protocol.