| KEYWORD |
Analysis of industrial networks traffic by means of classification algorithms
keywords CLASSIFICATION ALGORITHMS, INDUSTRIAL NETWORKS, WI-FI, WIRELESS NETWORKS
Reference persons STEFANO SCANZIO
Research Groups IEIIT/CNR COMPUTER ENGINEERING AND NETWORKS GROUP
Thesis type RESEARCH THESIS
Description This thesis is about the development of a framework for the automatic analysis of the traffic of industrial networks. The traffic of this type of networks is characterized by small frames sent cyclically. These two features make the industrial traffic peculiar with respect to the classical internet traffic. Taking advantage of these two features, traffic classification systems (based on probabilistic models) for industrial networks are different with respect to classical intrusion detection systems. The classification models that will be produced in this thesis have the double function to automatically identify damages in the industrial production system (that can have more than 100000 sensors) and identify unwanted traffic. This thesis, open to more than one student, has the first target to develop a set of programs useful to extract the most important parameters from frames sent in the network, and to define a set of data for the evaluation of the performance of the proposed classification algorithms. As a second target, the thesis has the purpose to realize classification algorithms (for instance based on Gaussian Mixture Models, Neural Networks, Support Vector Machines, ...) for the automatic recognition of some specific damages or intrusion attempts.
Required skills Programmazione in linguaggi di alto livello, preferibilmente python. Conoscenza base dei piu’ importanti protocolli di rete (Ethernet, IP, TCP, UDP,...).
Notes The thesis work will be done in the IEIIT institute of the CNR (National Research Council of Italy) which is inside the Politecnico di Torino
Deadline 26/06/2022
PROPONI LA TUA CANDIDATURA