| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 4.79 MB | Adobe PDF |
Advisor(s)
Abstract(s)
A tecnologia LoRaWAN foi desenvolvida com o intuito de permitir sensores IoT de baixo custo e baixa capacidade energética, transmitir pequenas mensagens, a longas distâncias. A existência desta tecnologia e o facto de se comprovar ser bastante versátil permite, assim, o afastamento do âmbito apenas do IoT e colocá-la também em sistemas críticos, como, por exemplo, sistemas de segurança associados à infraestrutura ferroviária. No entanto, neste cenário, seria necessário tornar mais eficaz o processamento das mensagens, que, eventualmente, se conseguiria através da computação na periferia. No entanto, a arquitetura das redes lorawan impedem este tipo de acelaração do processamento. Este projeto propõe uma solução para colmatar a dificuldade de integração da tecnologia LoRaWAN com computação na periferia. Através do desenvolvimento de um sistema de coordenação entre múltiplos sistemas de computação na periferia, capazes de funcionar como gateways LoRaWAN, é possível processar as mensagens localmente, sem estas terem de ser enviadas para um elemento central na núvem. De forma a aumentar a resiliência da solução, os sistemas de computação na periferia incluem múltiplas ligações a redes móveis, desenvolvidas no sentido de encaminhar o tráfego do serviço de forma resiliente e de acordo com as prioridades definidas. Com o intuito de aumentar a segurança da solução, este trabalho propõe, também, a implementação do algoritmo K-nearest neighbors, para deteção de interferências nos sensores que monitorizam as infraestruturas críticas.
The LoRaWAN technology was designed with the objective of allowing low cost and low energetic capacity iot sensors to transmit short messages over long distances. The existence of this technology and the fact that it has proven to be quite versatile allows, thus, the deviation from the realm of just IoT and also places it in critical systems, such as, for example, security systems associated with the railway infrastructure. Although, in this scenario, it would be necessary to make message processing more efficient, which could, eventually, be achieved through computing at the edge. However, LoRaWAN network infrastructure hinder this type of processing acceleration. This project presents a solution to overcome the difficulty of integrating LoRaWAN technology with computing at the periphery. By developing a coordination system between multiple edge computing systems, capable of functioning as LoRaWAN gateways, it is possible to process messages locally, without them having to be sent to a central element in the cloud. In order to increase the resilience of the solution, computing systems at the edge include multiple connections to mobile networks, developed to route service traffic in a resilient manner and in accordance with defined priorities. In order to increase the security of the solution, this work also proposes the implementation of the knn algorithm to detect interference in sensors that monitor critical infrastructures.
The LoRaWAN technology was designed with the objective of allowing low cost and low energetic capacity iot sensors to transmit short messages over long distances. The existence of this technology and the fact that it has proven to be quite versatile allows, thus, the deviation from the realm of just IoT and also places it in critical systems, such as, for example, security systems associated with the railway infrastructure. Although, in this scenario, it would be necessary to make message processing more efficient, which could, eventually, be achieved through computing at the edge. However, LoRaWAN network infrastructure hinder this type of processing acceleration. This project presents a solution to overcome the difficulty of integrating LoRaWAN technology with computing at the periphery. By developing a coordination system between multiple edge computing systems, capable of functioning as LoRaWAN gateways, it is possible to process messages locally, without them having to be sent to a central element in the cloud. In order to increase the resilience of the solution, computing systems at the edge include multiple connections to mobile networks, developed to route service traffic in a resilient manner and in accordance with defined priorities. In order to increase the security of the solution, this work also proposes the implementation of the knn algorithm to detect interference in sensors that monitor critical infrastructures.
Description
Dissertação para obtenção do Grau de Mestre em Engenharia de Eletrónica e Telecomunicações
Keywords
LoT LoRaWAN Edge LoRa Lot Lorawan Edge
Citation
Esteves, Gonçalo Correia de Almeida da Silva - Computação na periferia na Internet das Coisas. Lisboa: Instituto Superior de Engenharia de Lisboa, 2023. Dissertação de Mestrado