SOUSA, M. P.; http://lattes.cnpq.br/1494367157419683; SOUSA, Marcelo Portela.
Resumen:
Wireless sensor networks (WSNs) are becoming popular in many applications, such as monitoring,
surveillance, healthcare, among others. The performance of those systems can be
improved by the use of the cooperative diversity technique, to mitigate the fading caused by
multipath in wireless channels. By enabling a set of partner nodes to relay the received information,
this system exploits the spatial diversity by the cooperation between the distributed
antennas, located at the multiple terminals. In this Thesis, the author presents three schemes
of cooperative diversity for wireless sensor networks with cognitive characteristics, either by
the capabilities related to the dynamic spectrum management, or by the use of computational
intelligence and bio-inspired computing. The contributions range from the proposal of techniques
to novel clustering protocols, including the LF-Ant and Cognitive LF-Ant, and a novel
cooperative scheme, which involves the cooperative modulation diversity with spectrum sensing.
Regarding the proposed clustering protocols, the election of cluster-heads is inspired by
the organized behaviour of ant colonies, with the heuristic information modelled by a fuzzy inference
system. Furthermore, a new protocol, inspired by experiments with legs of desert ants,
to allocate resources in accordance to the emergency degree of patients in cognitive healthcare
networks is presented. Simulation results show the decrease of the average delay as the probability
of opportunistic access increases, that privileges the emergency reporting of patients
with higher priority to access the resources. The impact of the channel estimation errors, of the
energy factor, and of the interleaving depth is evaluated for heterogeneous networks. The information
of the network residual energy is used by energy maps to elect the vice cluster-heads
and to better distribute the energy consumption over the network, which is also benefited by the
opportunistic spectrum access.