Name: WILLIAN GARCIA VIEGA DOS SANTOS

Publication date: 21/12/2020
Advisor:

Name	Role
HELDER ROBERTO DE OLIVEIRA ROCHA	Co-advisor *
JAIR ADRIANO LIMA SILVA	Advisor *
MARCELO EDUARDO VIEIRA SEGATTO	Co-advisor *

Examining board:

Name	Role
CELSO JOSE MUNARO	Internal Examiner *
FABIO DE OLIVEIRA LIMA	External Examiner *
HELDER ROBERTO DE OLIVEIRA ROCHA	Co advisor *
JAIR ADRIANO LIMA SILVA	Co advisor *
MARCELO EDUARDO VIEIRA SEGATTO	Advisor *

Summary: With the growth in the number of devices on IoT networks, the requirements for communication reliability and efficient use of energy are increasing. In this work, a hybrid network topology for IoT sensors is proposed, using the mesh and star topologies. This hybrid network aims to improve the efficiency of energy consumption and ensure greater network coverage for the allocation of sensors. For network planning, defining the clusters of the Mesh network and points of the star network, it is proposed to merge the Genetic Algorithms and K-Medoids. For the allocation of resources in the Mesh (ZigBee) and star (LoRa) networks, Dijkstra and Genetic Algorithms are proposed, respectively, in order to minimize the energy consumption of the network. Experimental data collected in the field were used to plan the proposed network topology. A network of 380 sensors spread over Vitória-ES was connected to a gateway in Vila Velha-ES. After executing the planning and resource allocation algorithms, 26 mesh networks were defined and resources were allocated to each sensor in order to save energy and ensure communication.

Access to document