In recent years we have been assisted by the rapid development of small, low-power, and low-cost wireless computation/communication devices, which have served as enablers for the so-called “Internet of Things” (IoT). Within the Internet of Things, devices adopt wireless communication to cooperate to provide services and added value to users, probably in large-scale dynamic environments. This type of device-centric cooperation can be interpreted as social interaction between enabled devices instead of between people. Smartphones and tablets are clearly the most widely-known examples of such devices; however, there are also many others (e.g., home and office appliances or wearable devices) and much more will be available shortly. Devices that can connect to the Internet can exploit cloud-based services to enable interaction in an IoT scenario. However, an Internet connection might not always be available, or it might be too expensive to use. It is clear that within the Internet of Things we cannot impose the requirement of having an always-on functional Internet connection. Instead, we need to shift our focus to a new wave of interaction called “proximity-based interactions”. Applications can benefit from this type of interaction when users and devices are physically close to one another. They do not rely on an Internet connection; instead, they operate in an infrastructure-less scenario, possibly interacting in a peer-to-peer (P2P) manner. To enable proximity-based interactions, devices should be equipped with one or more P2P communication protocols such as Bluetooth, BLE, NFC, Wi-Fi Direct, 6LowPan, ZigBee, and ultrasonic. Among P2P communication protocols for mobile devices, Wi-Fi Direct has recently gained attention because it is widely available on smart devices, specifically smartphones, and also does not have the critical problems of traditional ad-hoc Wi-Fi such as speed, security, stability, scalability, and power consumption. Although many works have already tried to exploit Wi-Fi Direct in social interaction between proximal smart devices, there is not any work that attempts to operate Wi-Fi Direct in large-scale dynamic application domains. This thesis bridges this gap by proposing a middleware infrastructure, called “MAGNET”, that aims to provide reliable and stable P2P connectivity between large numbers of smart devices in a dynamic environment without user intervention. The goal is to remove the known limitations and offer a better means to exploit this technology. This goal is achieved by managing devices in several Wi-Fi Direct groups and ensuring that each device is part of a Wi-Fi Direct group in any circumstance (i.e. saturation, moving, and failing). In the second step, MAGNET oversees inter-connecting Wi-Fi Direct groups and provides multi-hop connectivity between a large number of devices. To evaluate the effectiveness of MAGNET, a Wi-Fi Direct simulator has also been developed, called “WiDiSi”. The proposed solution has been tested on real Android devices and the simulation environment. Realistic scenarios are utilized in the thesis to showcase and evaluate the key features of the proposed solution. The evaluation results illustrate the effectiveness of such a solution in providing a stable communication between a large number of mobile devices using Wi-Fi Direct.

In questi ultimi anni abbiamo assistito al rapido sviluppo di dispositivi di calcolo/comunicazione wireless di piccole dimensioni, a basso consumo e a basso costo, che sono servite da facilitatori per il cosiddetto "Internet delle cose" (IoT). I dispositivi che possono connettersi ad Internet possono sfruttare i servizi basati sul cloud per consentire un'interazione in uno scenario IoT. Tuttavia, una connessione Internet non sempre potrebbe essere disponibile, o potrebbe essere troppo costosa da usare. E' chiaro che all'interno dell'Internet delle cose non possiamo imporre il requisito di avere una connessione internet funzionale sempre attiva. Dobbiamo piuttosto spostare la nostra attenzione su una nuova tendenza di interazione denominata "interazione peer-to-peer". Tra i protocolli di comunicazione P2P per dispositivi portatili, Wi-Fi Direct si è recentemente imposto all'attenzione in quanto è ampiamente disponibile sui dispositivi smart, nello specifico sugli smartphone, e inoltre non ha i problemi critici del tradizionale Wi-Fi ad-hoc, come la velocità, sicurezza, stabilità, scalabilità, e il consumo energetico. Nonostante molte opere abbiano già cercato di sfruttare Wi-Fi Direct nell'interazione sociale tra dispositivi smart vicini, non c'è alcuna proposta che tenti di azionare Wi-Fi Direct in domini applicativi dinamici su larga scala Questa tesi colma questa lacuna proponendo un'infrastruttura middleware, chiamata "MAGNET", che mira a fornire una connettività P2P affidabile e stabile tra un gran numero di dispositivi smart in un ambiente dinamico, senza intervento da parte dell'utente. L'obiettivo è quello di rimuovere le limitazioni note e di offrire dei mezzi migliori per sfruttare questa tecnologia.

A middleware for peer-to-peer interaction of smart things

DERAKHSHAN, NASER

Abstract

In recent years we have been assisted by the rapid development of small, low-power, and low-cost wireless computation/communication devices, which have served as enablers for the so-called “Internet of Things” (IoT). Within the Internet of Things, devices adopt wireless communication to cooperate to provide services and added value to users, probably in large-scale dynamic environments. This type of device-centric cooperation can be interpreted as social interaction between enabled devices instead of between people. Smartphones and tablets are clearly the most widely-known examples of such devices; however, there are also many others (e.g., home and office appliances or wearable devices) and much more will be available shortly. Devices that can connect to the Internet can exploit cloud-based services to enable interaction in an IoT scenario. However, an Internet connection might not always be available, or it might be too expensive to use. It is clear that within the Internet of Things we cannot impose the requirement of having an always-on functional Internet connection. Instead, we need to shift our focus to a new wave of interaction called “proximity-based interactions”. Applications can benefit from this type of interaction when users and devices are physically close to one another. They do not rely on an Internet connection; instead, they operate in an infrastructure-less scenario, possibly interacting in a peer-to-peer (P2P) manner. To enable proximity-based interactions, devices should be equipped with one or more P2P communication protocols such as Bluetooth, BLE, NFC, Wi-Fi Direct, 6LowPan, ZigBee, and ultrasonic. Among P2P communication protocols for mobile devices, Wi-Fi Direct has recently gained attention because it is widely available on smart devices, specifically smartphones, and also does not have the critical problems of traditional ad-hoc Wi-Fi such as speed, security, stability, scalability, and power consumption. Although many works have already tried to exploit Wi-Fi Direct in social interaction between proximal smart devices, there is not any work that attempts to operate Wi-Fi Direct in large-scale dynamic application domains. This thesis bridges this gap by proposing a middleware infrastructure, called “MAGNET”, that aims to provide reliable and stable P2P connectivity between large numbers of smart devices in a dynamic environment without user intervention. The goal is to remove the known limitations and offer a better means to exploit this technology. This goal is achieved by managing devices in several Wi-Fi Direct groups and ensuring that each device is part of a Wi-Fi Direct group in any circumstance (i.e. saturation, moving, and failing). In the second step, MAGNET oversees inter-connecting Wi-Fi Direct groups and provides multi-hop connectivity between a large number of devices. To evaluate the effectiveness of MAGNET, a Wi-Fi Direct simulator has also been developed, called “WiDiSi”. The proposed solution has been tested on real Android devices and the simulation environment. Realistic scenarios are utilized in the thesis to showcase and evaluate the key features of the proposed solution. The evaluation results illustrate the effectiveness of such a solution in providing a stable communication between a large number of mobile devices using Wi-Fi Direct.
BONARINI, ANDREA
SCIUTO, DONATELLA
7-feb-2017
In questi ultimi anni abbiamo assistito al rapido sviluppo di dispositivi di calcolo/comunicazione wireless di piccole dimensioni, a basso consumo e a basso costo, che sono servite da facilitatori per il cosiddetto "Internet delle cose" (IoT). I dispositivi che possono connettersi ad Internet possono sfruttare i servizi basati sul cloud per consentire un'interazione in uno scenario IoT. Tuttavia, una connessione Internet non sempre potrebbe essere disponibile, o potrebbe essere troppo costosa da usare. E' chiaro che all'interno dell'Internet delle cose non possiamo imporre il requisito di avere una connessione internet funzionale sempre attiva. Dobbiamo piuttosto spostare la nostra attenzione su una nuova tendenza di interazione denominata "interazione peer-to-peer". Tra i protocolli di comunicazione P2P per dispositivi portatili, Wi-Fi Direct si è recentemente imposto all'attenzione in quanto è ampiamente disponibile sui dispositivi smart, nello specifico sugli smartphone, e inoltre non ha i problemi critici del tradizionale Wi-Fi ad-hoc, come la velocità, sicurezza, stabilità, scalabilità, e il consumo energetico. Nonostante molte opere abbiano già cercato di sfruttare Wi-Fi Direct nell'interazione sociale tra dispositivi smart vicini, non c'è alcuna proposta che tenti di azionare Wi-Fi Direct in domini applicativi dinamici su larga scala Questa tesi colma questa lacuna proponendo un'infrastruttura middleware, chiamata "MAGNET", che mira a fornire una connettività P2P affidabile e stabile tra un gran numero di dispositivi smart in un ambiente dinamico, senza intervento da parte dell'utente. L'obiettivo è quello di rimuovere le limitazioni note e di offrire dei mezzi migliori per sfruttare questa tecnologia.
Tesi di dottorato
File allegati
File Dimensione Formato  
Thesis_Naser_Derakhshan.pdf

accessibile in internet per tutti

Descrizione: Thesis text
Dimensione 10.22 MB
Formato Adobe PDF
10.22 MB Adobe PDF Visualizza/Apri

I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10589/131921