DCCMAPI/CCET

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
DISCENTE: THIAGO ALLISSON RIBEIRO DA SILVA
DATA: 19/03/2024
HORA: 09:30
LOCAL: Sala Virtual (Google Meet)
TÍTULO: A Smart Communication Architecture to Provide Connectivity to Applications in the Power Distribution System of Smart Grids
PALAVRAS-CHAVES: Clustering, Internet of Energy, Low-Power Wide Area Network, Modeling and Simulation of Networks, Power Distribution Systems.
PÁGINAS: 80
GRANDE ÁREA: Ciências Exatas e da Terra
ÁREA: Ciência da Computação
SUBÁREA: Sistemas de Computação
ESPECIALIDADE: Teleinformática
RESUMO: The implementation of a communication infrastructure to support the execution of services that assist in the operation of electrical power distribution systems is crucial for the evolution of the power systems. This infrastructure must be planned to meet different Quality of Service (QoS) requirements, in order to ensure the proper operation of various power grid applications. An example of such applications is the Advanced Metering Infrastructure (AMI), a platform that requires an adequate definition of the quantity and placement of Data Aggregation Points (DAPs) to receive data from Smart Meters (SMs) and forward it to utility servers. In this context, this work proposes a communication architecture to provide connectivity to applications in the power distribution system, which encompasses planning and operation aspects at the data transmission level, initially reasoned on the development of an clustering-based approach for positioning DAPs in Long Range Wide Area Network (LoRaWAN) and Narrowband Internet of Things (NB-IoT). This approach consider the use of the cluster methods K-Means (KM), K-Medoids (KMD) and Fuzzy C-Means (FCM) in an iterative heuristic process to minimize the number of DAPs, aiming at reaching values within predefined thresholds for Packet Delivery Rate (PDR). The networks have been simulated for a realistic urban scenario of 49 km2 in the Network Simulator (NS-3). The attained outcomes are related to clustering, such as number of clusters and intra-cluster distances, as well as network performance metrics, for example Signal-Noise Ratio (SNR) and delay. The preliminary results show that the NB-IoT network outperforms LoRaWAN, requiring only 4 DAPs and with a SNR 331.52% higher than the LoRaWAN infrastructure with 18 DAPs, for a minimum PDR of 95%. However, both technologies can guarantee the necessary QoS for the operation of an AMI.
MEMBROS DA BANCA:
Presidente - 005.693.753-98 - JOSE VALDEMIR DOS REIS JUNIOR
Interno - 803.957.673-34 - ANDRE CASTELO BRANCO SOARES
Externo à Instituição - JOÃO HENRIQUE KLEINSCHMIDT - UFABC