Description: This thesis is about the use of wireless techniques based on IEEE 802.11 (Wi-Fi) standard (and newest ones) in order to support industrial real-time communications. Indeed, there is a strong interest in the industry in the use of wireless devices for the communication between controllers and actuator devices (e.g., industrial robots). A technique to increase transmission determinism and dependability is to transmit a duplicated frame on more than one channel. This technique, in order to improve performance, requires some modification to the IEEE 802.11g protocol. The purpose of this thesis is the modification of the models of a network simulator in order to compare different techniques based on redundancy with respect to classical models. Eventually, a prototypal implementation may be obtained using a Linux based operating system.
Required skills: Good knowledge of the C programming language, of the Linux operating system, and basic knowledge of the most important network protocols (Ethernet, IP, TCP, UDP and possibly IEEE 802.11). Not required but welcome is the knowledge of the Python language
Keywords: IEEE 802.11, Industrial Networks, Redundancy, Wi-Fi, Wireless Networks
References: Paper 1, Paper 2
Description: The thesis focuses on the use of machine learning (ML) algorithms to estimate, in a low-cost way, the temperature in indoor environments. The thesis work consists of two main steps, that is the development of (1) a data acquisition system and (2) a temperature estimation system. For step (1), both PCs and Raspberry Pi devices will be employed. Raspberry Pis will be used for acquiring temperature sensor data while PCs will be employed to measure features related/influencing the temperature (e.g., the temperature measured by sensors in CPU, motherboard, hard disk, etc., as well as CPU usage, interrupts rate, etc.).
The final goal of the thesis work is to use the simple data acquired by the PCs (and the other features) to estimate room temperature through ML regression techniques especially based on artificial neural networks.
Required skills: Basic knowledge of the Linux operating system, good programming skills (especially Python programming language), basic knowledge of machine learning.
Keywords: Machine Learning, Artificial Neural Networks, Regression Techniques, Temperature Sensors, HVAC systems.
Description: The ability to manage Wi-Fi adapters by means of common applications executed at the user-space level is an enabling technology for many kinds of applications that require to directly manage the transmission of every single frame in the network. On the other hand, the time needed to inject the frame in the ether heavily impacts the applicability of such techniques. In this thesis, starting from the prototype implementation described in the following two references, we want to automatically estimate the time needed by the operating system to send frames on air. This thesis requires code development, analysis of methods aimed at correctly estimate this time even in disturbed environments, and experimental campaigns.
Required skills: Even if the thesis regards all the level of the Linux operating system, required skills are limited to a good knowledge of the C programming language and of the Linux operating system.
Keywords: Software-defined MAC, IEEE 802.11, message scheduling
References: Paper 1, Paper 2
Description: Experimental thesis about next-generation Wireless Sensor Networks (WSN) based on IETF 6TiSCH, a recent emerging technology for industrial automation and robotics. The thesis work regards the use of 6TiSCH nodes for collecting big data obtained from remote sensors, as defined by the Industry 4.0 paradigm, and will be implemented using brand new OpenMote B nodes ( http://www.openmote.com/ ) equipped with the OpenWSN operating system and/or commercial Analog Devices SmartMesh IP nodes.
Required skills: Good knowledge of the C programming language and basic knowledge of the Linux operating system.
Keywords: Industrial Internet of Things (IIoT), Wireless Sensor Networks (WSN), IETF 6TiSCH
Some examples of other possible arguments of theses are:
All the references can be downloaded only from the Politecnico di Torino network.
All theses will be performed in the laboratory near my office, which is located: here.
These theses are for Master Degree.
If interested contact: stefano.scanzio [ AT ] polito.it (tel. +39 011 090 5438).
Stefano Scanzio supervised the following theses:
[S10] Giuseppe Pedone, “Implementazione ed Ottimizzazione di Tecniche di Ridondanza WiFi a Livello Applicativo”, 28/07/2020, Politecnico di Torino, Master thesis.
[S9] Biru Biruk Abdissa, “Open Source Solutions for the Industrial Internet of Things”, 14/12/2018, Politecnico di Torino, Master thesis.
[S8] Cavallaro Mariano, “Ridondanza nelle reti IEEE 802.11 per l’automazione industriale”, 15/10/2015, Politecnico di Torino, Bachelor thesis.
[S7] Aguirre Selgas Blanca, “Clock Synchronization Error Measurement Techniques for RBIS”, April 2015, E.T.S.I. Industriales (UPM), Master thesis.
[S6] Vassallo Andrea, “Analisi prestazionale di reti wireless IEEE 802.11n”, 10/12/2013, Politecnico di Torino, Master thesis.
[S5] Calandri Mauro, “Verifica sperimentale di proprietà di safety in reti industriali”, 02/07/2012, Politecnico di Torino, Master thesis.
[S4] Fonio Filippo, “Piattaforma di calcolo ad alte prestazioni per il controllo industriale”, 02/07/2012, Politecnico di Torino, Master thesis.
[S3] Echeverry Tamayo Julian Felipe, “Protocolli di sincronizzazione in ambienti di controllo industriali”, 08/11/2011, Politecnico di Torino, Master thesis.
[S2] Monteleone Simone, “Sistemi di controllo open-source real-time basati su tecnologie Industrial Ethernet”, 12/04/2011, Politecnico di Torino, Master thesis.
[S1] Casalena Ernesto, “Meccanismi di sincronizzazione distribuita in sistemi operativi real-time”, 20/04/2010, Politecnico di Torino, Master thesis.
About the simulation of TSCH Wireless Sensor Networks in order to save energy and possible thesis about machine learning (see the video)
Other details on the subject: https://youtu.be/VPGPNSfRa50