Wireless networking for autonomous mobile smart camera drones
Miniatura indisponível
Data
2022
Autores
Título da revista
ISSN da revista
Título do Volume
Editora
Resumo
Dispositivos voadores em rede, tais como veículos aereos não tripulados (UAV), também chamados drones, equipados com camaras inteligentes estão cada vez mais a serem utilizados para formar sistemas autonomos cooperativos para realizar tarefas de fotogrametria e detecção, incluindo a captação de dados de vídeo e imagem em tempo real de alta qualidade para monitorização no terreno, vigilância e serviços de gestao de desastres. O aumento dos UAV tem sido principalmente atribuído a sua maleabilidade e oferta de grande autonomia no estabelecimento de uma rapida ligação sem fios, permitindo esforços coordenados na recuperação de informaçao crítica e sensivel ao tempo a partir de varios locais, incluindo locais em perigo para reduzir os factores de risco humano. Embora tenham sido desenvolvidos drones de camara inteligente em rede para produzir enormes quantidades de dados críticos de locais atingidos por catastrofes que precisam de ser transferidos e analisados em tempo real, o funcionamento em rede e relacionado com a comunicação, abrangendo o sistema de UAV único a uma grande rede ad hoc de UAVs (UAV-FANET) capaz de fornecer uma coordenação rápida de agentes para uma recuperação e disseminação adequadas de dados de vídeo e imagem, vem com desafios significativos, nomeadamente topologia dinamica e mobilidade rápida dos nós. Al em disso, as UAV-FANETs enfrentam desconexoes intermitentes e não disponíveis intermitentemente, em grande parte devido a sua estrutura 3D, impactando assim a qualidade dos pacotes disseminados. Determinada a enfrentar estes desafios, o principal objectivo desta tese e investigar novos sistemas de redes de camaras inteligentes voadoras capazes de capturar e transmitir imagens e vídeo de alta qualidade em grandes areas geográficas, concentrando-se não na reprodução de vídeo, mas na investigação de abordagens avançadas de redes sem fios para a transmissao de pacotes. Para este fim, o objetivo desta tese surge no seguinte: analisar o estudo de aplicabilidade de enxames de drones para rede e requisitos de computação necessários para o funcionamento eficiente de um sistema de enxame como um sistema de controlo em rede; investigar escolhas fiaveis de desenho de comunicaçoes sem fios capazes de fornecer apoio suficiente para a captura e disseminação adequadas de pacotes de vídeo e imagem em grandes UAV-FANETs; conceber e implementar um novo esquema baseado em clusters chamado Mecanismo de Clustering Dinamico com Balanceamento de Carga para realizar o clustering devido a mobilidade dos nós e à estrutura 3D, ao mesmo tempo que se abordam as restrições causadas por alterações topológicas, e em sincronia com uma função de entropia para lidar com a tolerância a falhas de cluster e sobrecargas de tráfego para uma disseminação adequada dos pacotes de dados FANETs; e finalmente, para investigar o encaminhamento híbrido interoperavel baseado na posição, onde o encaminhamento intra-cluster e capaz de evitar caminhos de encaminhamento inadequados com base num metodo de classificaçao simples, enquanto o encaminhamento entre cluster baseado em abordagens gananciosas e de regras de mao direita para reduzir a sobrecarga de mensagens e para escapar a situacões de nós vazios, para resolver desconexões intermitentes e limitações de escalabilidade em FANETs agrupadas. A avaliação experimental realizada utilizando a ferramenta de simulação NS-3 provam que as nossas novas estruturas de comunicação são capazes de fornecer uma disseminação adequada de pacotes de dados em comparação com outros mecanismos de última geracão. Em suma, as estruturas teórica e lógicas investigadas neste estudo fornecem a base para a concepção e implementação de sistemas sem fios adequados capazes de fornecer uma disseminação adequada de pacotes de vídeo e imagem em grandes FANETs de UAV.
Networked flying devices such as unmanned aerial vehicles (UAVs), also called drones, equipped with smart cameras are increasingly being deployed to form cooperative autonomous systems to perform photogrammetry and sensing tasks, including the capture of high-quality real-time video and image data for field monitoring, surveillance and disaster management services. The rise of UAVs has been mainly attributed to their malleability and offer of great autonomy in establishing links, allowing for coordinated efforts in retrieving critical and time-sensitive information from various locations, including endangered sites to reduce human risk factors. While networked intelligent camera drones have been developed to produce huge amounts of critical data from disaster-stricken locations that need to be transferred and analyzed in real time, the networking and communication-related operation, spanning the single UAV system to a large flying (UAV-FANET) capable of providing rapid agent coordination for proper retrieval and dissemination of video and image data, comes with significant challenges, namely dynamic topology and rapid node mobility. Moreover, UAV-FANETs face intermittent wireless disconnections and intermit tent available nodes, largely due to their 3D structure, thus impacting the quality of disseminated packets. Determined to address these challenges, the main objective of this thesis is to investigate novel flying smart camera networked systems capable of capturing and transmitting high quality images and video over large geographical areas, focusing not on video rendering, but on investigating advanced wireless networking approaches for packets transmission. To this end, the core of this thesis arises in the fol lowing: analyze applicability study of drone swarms for networking and computing requirements needed for efficient operation of a swarm system as a networked control system; investigate reliable wireless communications design choices capable of providing sufficient support for proper capture and dissemi nation of video and image packets in large UAV-FANETs; design and implement a novel cluster-based scheme called Dynamic Clustering Mechanism with Load Balancing to perform clustering based on node mobility and 3D structure while addressing constraints caused by topological changes, and in sync with an entropy function to handle cluster fault tolerance and traffic overloads for proper dissemination of data packets in FANETs; and finally, design and implement an interoperable Position-based Hybrid routing protocol, where intra-cluster routing is capable of avoiding unsuitable forwarding paths based on a simple ranking method, while inter-cluster routing based on greedy and right-hand rule approaches to reduce message overheads and to escape void node situations, to address intermittent wireless dis connections and scalability limitations in clustered FANETs. Experimental evaluations performed using the NS-3 simulation tool prove that our new communication structures are able to provide adequate data packets dissemination compared to other state-of-the-art mechanisms. In summary, the theoretical and logical frameworks investigated in this study provides the basis for the design and implementation of suitable wireless systems capable of providing adequate video and image packets dissemination in large UAV-FANETs.
Networked flying devices such as unmanned aerial vehicles (UAVs), also called drones, equipped with smart cameras are increasingly being deployed to form cooperative autonomous systems to perform photogrammetry and sensing tasks, including the capture of high-quality real-time video and image data for field monitoring, surveillance and disaster management services. The rise of UAVs has been mainly attributed to their malleability and offer of great autonomy in establishing links, allowing for coordinated efforts in retrieving critical and time-sensitive information from various locations, including endangered sites to reduce human risk factors. While networked intelligent camera drones have been developed to produce huge amounts of critical data from disaster-stricken locations that need to be transferred and analyzed in real time, the networking and communication-related operation, spanning the single UAV system to a large flying (UAV-FANET) capable of providing rapid agent coordination for proper retrieval and dissemination of video and image data, comes with significant challenges, namely dynamic topology and rapid node mobility. Moreover, UAV-FANETs face intermittent wireless disconnections and intermit tent available nodes, largely due to their 3D structure, thus impacting the quality of disseminated packets. Determined to address these challenges, the main objective of this thesis is to investigate novel flying smart camera networked systems capable of capturing and transmitting high quality images and video over large geographical areas, focusing not on video rendering, but on investigating advanced wireless networking approaches for packets transmission. To this end, the core of this thesis arises in the fol lowing: analyze applicability study of drone swarms for networking and computing requirements needed for efficient operation of a swarm system as a networked control system; investigate reliable wireless communications design choices capable of providing sufficient support for proper capture and dissemi nation of video and image packets in large UAV-FANETs; design and implement a novel cluster-based scheme called Dynamic Clustering Mechanism with Load Balancing to perform clustering based on node mobility and 3D structure while addressing constraints caused by topological changes, and in sync with an entropy function to handle cluster fault tolerance and traffic overloads for proper dissemination of data packets in FANETs; and finally, design and implement an interoperable Position-based Hybrid routing protocol, where intra-cluster routing is capable of avoiding unsuitable forwarding paths based on a simple ranking method, while inter-cluster routing based on greedy and right-hand rule approaches to reduce message overheads and to escape void node situations, to address intermittent wireless dis connections and scalability limitations in clustered FANETs. Experimental evaluations performed using the NS-3 simulation tool prove that our new communication structures are able to provide adequate data packets dissemination compared to other state-of-the-art mechanisms. In summary, the theoretical and logical frameworks investigated in this study provides the basis for the design and implementation of suitable wireless systems capable of providing adequate video and image packets dissemination in large UAV-FANETs.
Descrição
Orientação: Paulo Jorge Milheiro Mendes ; coorientação: Naercio David Pedro Magaia
Palavras-chave
DOUTORAMENTO EM INFORMÁTICA - NOVOS MEDIA E SISTEMAS UBÍQUOS, INFORMÁTICA, REDES SEM FIOS, ANÁLISE DE CLUSTERS, COMPUTER SCIENCE, WIRELESS NETWORKS, CLUSTER ANALYSIS