SecUAV - A Uni_ed testbed for the evaluation of secure state estimators
Tutor / director / evaluatorRey Micolau, Francesc
Document typeMaster thesis
Rights accessOpen Access
Rapidly evolving technology and tight coupling of physical sensors and actuation with the system behind it in Cyber Physical Systems (CPS) have opened up new forms of security vulnerabilities to adversaries. These new types of attacks on the physical (PHY) layer can be used to drive the system in an undesired, unstable state which could ultimately resolve in life-threatening consequences for users. This new research-area has gained traction in the CPS community recently, but a feasible solution has yet to be found. One problem with the current approach of proposing algorithms and counter-measures against such attacks is, that it can be hard to compare different works in terms of effectiveness and computational cost as every research group has different ways of providing results. This work introduces SecUAV, common testbed for the CPS community. SecUAV utilizes OpenUAV, an open-source UAV simulator, for providing a common testbed for the fair evaluation of secure state estimation algorithms. This is achieved by expanding the already existing functionality with a framework allowing to simulate different kinds of attacks and to deploy different secure CPS algorithms. The feasibility of SecUAV is demonstrated by deploying three works in it, and comparing the resulting data in a fair manner. One of the studies algorithms, ReCaP , has been developed and evaluated by SecUAV during the work on this thesis.
This TFM is focused on secure estimation, specifically on detect and counteract attacks on sensor networks by means of signal processing/estimation. The project will start with a review of state of the art secure estimators, to continue with the implementation of some algorithms in order to compare differents systems. Althought the project is done at UPC, it considers the collaboration with the Networked & Embedded Systems Laboratory (NESL) at the University of California at Los Angeles (UCLA).