Nombre: YNGRITH SOARES DA SILVA
Tipo: MSc dissertation
Fecha de publicación: 18/12/2020
Supervisor:

Nombre	Papel
HELDER ROBERTO DE OLIVEIRA ROCHA	Advisor *
JAIR ADRIANO LIMA SILVA	Co-advisor *

Junta de examinadores:

Nombre	Papel
HELDER ROBERTO DE OLIVEIRA ROCHA	Advisor *
JAIR ADRIANO LIMA SILVA	Co advisor *
MARCIA HELENA MOREIRA PAIVA	Internal Examiner *
WANDERLEY CARDOSO CELESTE	External Examiner *

Sumario: A wireless sensor network (RSSF) is formed by sensor nodes that transmit information on a channel in which multipath transmissions occur. For this reason, simultaneous transmissions can occur, which cause interference causing power to fade. A problem faced by the WSN is the conservation of energy, which is used for communication between nodes. Generally, the power supply is made by batteries, which cannot be large, cannot be easily exchanged and have a high maintenance cost. For better use of available energy, a technique known as Topology Control can be used. Topology Control determines the transmission links between sensor nodes, keeping the network connected. In this way, it aims to obtain an efficient communication through the attribution of transmission power of the sensor nodes, consequently reducing the interference and also the retransmissions. In the present work, the GRASP, VNS and GRASP-VNS metaheuristics are proposed for five variations of the RSSF power allocation problem (minimization of power, minimization of maximum interference, minimization of total interference, minimization of power and maximum interference, minimization of power and total interference), ensuring that the graph is connected. In addition, the results obtained in these metaheuristics were compared with optimal results found by CPLEX for the same five variations of the power allocation problem. The proposed metaheuristics found the same solution as the optimal results
for most small instances (V 10 to 50 knots). For realistic instances, the GRASP-VNS metaheuristic proved to be more promising.

Acceso al documento