CRAN - Campus Sciences
BP 70239 - 54506 VANDOEUVRE Cedex
Tél : +33 (0)3 72 74 52 90
cran-secretariat@univ-lorraine.fr
 
 
Ph. D. Project : Malicious attacks detection and resilience in cyber-physical systems through joint dynamic resource scheduling and synthesis of adaptive control laws.
Dates : 2018/10/01 - 2021/09/30
Manager(s) CRAN: Dominique SAUTER
Other Manager(s): SONG Ye-Qiong (ye-qiong.song@loria.fr)
Full reference: The distributed monitoring and control of Cyber Physical Systems (CPS) relies on the use of information exchanged via networks between controllers, actuators and sensors. The information exchanged is therefore vulnerable to malicious attacks applied to the signals received and issued by the controllers.
The design of secure and safe Networked Control Systems (NCS) is of high importance in the control of large-scale critical infrastructures or industrial plants such as power grids, transportation systems, communication networks, oil and gas pipelines, water distribution or waste-water treatment systems and irrigation networks . Using “open” public and also wireless networks for the communication within NCS can generate severe security problems since an unauthorized access (“cyber attack”) is possible in the control system.
In this PhD thesis, to go beyond the current state of the art solutions, our original idea is to jointly design dynamic resource reallocation/scheduling and control law synthesis for drastically increasing the resilience of CPS face to malicious attacks. This can be achieved in SDN framework where physical plant controller and network controller may be merged.
The first step of the thesis will include modeling, simulation and impact analysis of different malicious attacks. A particular attention will be paid to security vulnerability and safety of the physical system.
The second step will be the design of detection & isolation filters making it possible to detect as soon as possible the occurrence of an attack.
The third step of the thesis will be the joint design for both SDN-based resource reallocation/scheduling and computing specific control signals in order to compensate the effect of the attack, in particular to prevent instability of the closed loop control system.
The forth step will focus on the study of the mutual influence of the actions of both SDN controller and plant controller. The expected results should pave the way to establish a deep understanding and characterization of the CPS behavior under the impact of this double control loops.
Keywords: Diagnostic, cyber-attacks, networked control systems, resources allocation
Conditions: 3 years, LUE (ISIT) contract, standard salary.
Expected background: control, computer science, networks.
Department(s):
Automatic Control-Identification Diagnosis
Financial aspects: LUE contract
Publications: hal-00879242v1, hal-01348183v1, hal-01094322v1, hal-00758576v1, hal-00879114v1    + CRAN - Publications