Natural disasters and terrorist attacks are undermining today's social and economic security. While designing the next generation of mobile system, public safety is the most debated issue that might be addressed by upcoming technological means. Despite of the increasing pervasiveness of ICT technologies in many contexts of everyday life, their adoption in rescue operations is still at its infancy. Unmanned Aerial Vehicles (UAV) have shown their practicability and feasibility while being deployed to support such extreme use cases: communication assistance in disaster areas. In this thesis, we rely on the concept of UAV-Cell deployment as the main pillar of a novel emergency orchestration in charge of i) building an on-demand cellular network able to extend or replace terrestrial facilities, ii) capturing and exploiting endangered user behaviors, iii) proactively placing UAV-Cells in order to provide sufficient Quality-of-Service (QoS) guarantees and iv) predicting intensive crowd flow migrations by means of machine-learning techniques, such as Neural Networks (NNs), to quickly react to unexpected network changes. Finally, we carry out an exhaustive simulation campaign to prove the uniqueness and validness of our framework.
Oggigiorno, disastri naturali e attacchi terroristici stanno minando la sicurezza sia sociale che economica. Nella progettazione della prossima generazione di reti mobili, la pubblica sicurezza è l'aspetto più discusso tra quelli che potrebbero essere affrontati con l'avvento di futuri strumenti tecnologici. Nonostante la sempre più pervasiva presenza di tecnologie ICT in molti contesti della vita quotidiana, la loro adozione in operazioni di soccorso è ancora ai suoi albori. Gli Aeromobili a Pilotaggio Remoto (in inglese Unmanned Aerial Vehicles (UAVs)) hanno mostrato di essere una soluzione praticabile nel supportare casi d'uso così estremi: fornire assistenza alla comunicazione in aree colpite da disastri. In questa tesi consideriamo lo schieramento intelligente di UAV-Cells come il pilastro principale di un'innovativa gestione delle emergenze, responsabile di i) costruire una rete cellulare on-demand capace di estendere o sostituire l'infrastruttura di terra, ii) cogliere e sfruttare il comportamento degli utenti in pericolo, iii) piazzare proattivamente le UAV-Cells in modo da offrire sufficienti garanzie di Qualità di Servizio (QoS) e iv) predire gli intensi flussi di folla tramite tecniche di machine-learning, come Reti Neurali, per reagire velocemente a cambiamenti inattesi nella rete. Infine, valutiamo le performance della nostra soluzione tramite delle simulazioni su campioni di traffico sintetico, dimostrando l'unicità e la praticità del nostro approccio.
An emergency orchestration solution : proactive UAV-cell deployment for supporting first responders in public safety networks
ALBANESE, ANTONIO
2017/2018
Abstract
Natural disasters and terrorist attacks are undermining today's social and economic security. While designing the next generation of mobile system, public safety is the most debated issue that might be addressed by upcoming technological means. Despite of the increasing pervasiveness of ICT technologies in many contexts of everyday life, their adoption in rescue operations is still at its infancy. Unmanned Aerial Vehicles (UAV) have shown their practicability and feasibility while being deployed to support such extreme use cases: communication assistance in disaster areas. In this thesis, we rely on the concept of UAV-Cell deployment as the main pillar of a novel emergency orchestration in charge of i) building an on-demand cellular network able to extend or replace terrestrial facilities, ii) capturing and exploiting endangered user behaviors, iii) proactively placing UAV-Cells in order to provide sufficient Quality-of-Service (QoS) guarantees and iv) predicting intensive crowd flow migrations by means of machine-learning techniques, such as Neural Networks (NNs), to quickly react to unexpected network changes. Finally, we carry out an exhaustive simulation campaign to prove the uniqueness and validness of our framework.File | Dimensione | Formato | |
---|---|---|---|
Thesis.pdf
non accessibile
Descrizione: Tesi
Dimensione
19.23 MB
Formato
Adobe PDF
|
19.23 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/142968