· Research Leader
Oct 2017 – Now: Laboratoire LINEACT, CESI eXia
Leading research related to new challenges of big data, smart citicies, cyber physical systems, and IoT. My objective is developing techniques based on formal methods and data mining to ensure safety and functionality of complex systems including different behavioral aspects: real time, stochastic process, etc.
May 2017 – Sep 2017: Laboratoire VERIMAG, Grenoble
SUCCESS project: My main roles in this project is developing techniques to secure IoT based systems including: 1) specification and verification techniques for secure IoT components and their composition, 2) verification methods and risk assessment techniques for IoT scenarios with models of human behaviour, social interactions and human-system interactions, and 3) implementation and modeling languages with algorithms for the certification of safety, availability, secrecy, and trustworthiness across from the model to the platform.
· Research Associate
Nov 2013 – Avr 2017: University of Luxembourg, Luxembourg
UNIQUE project: as a member of the Applied Security and Information Assurance Group (APSIA) from Nov 2015, I am working on UNIQUE project (Unclonable Networks for Identification using Cholesteric Emulsions). The aim of UNIQUE is to generate encryption keys from Crystal. In two months, I developed a framework to extract the main futures of Crystal and generate keys. Currently, I am developing techniques to test the uniqueness, randomness, and reliability of the generated keys. Then in the next step, the main objective is to develop a methodology to recover the key from noisy Crystal based on the entropy metrics.
STAST project: as a member of Security and Trust of Software Systems (SaToSS) research group between Nov 2013 and Oct 2015, I worked on STAST project (Socio-Technical Analysis of Security and Trust). In STAST, my task is to propose a formal framework in which to model socio-technical components of information systems, and to develop methodologies and algorithms in order to detect, possibly in a semi-automatic or automatic way, attacks of socio-technical nature given a model of a system. I have developed a language to describe formally the socio-technical systems. Further, I added probability and cost to the possible actions and behaviors of actors in these complex system. Furthermore, I introduced the concept of attack surfaces to evaluate the weakness of the deployed system.
· Research Assistant
Jan 2009 – Oct 2013: Ericsson&Concordia University, Canada.
MOBS project: The aim of MOBS is to improve the existing verification techniques by reducing their complexities. The target is to show their efficiency on security aspect of systems. My main objective is verifying hardware and software used in system engineering by using formal methods techniques especially the probabilistic model checking. To achieve this goal, I developed a security verification framework to reduce the verification cost, specifying the system requirements, and measuring the security risk for the system under verification. The proposed framework develops efficient abstraction, symmetry reduction, and compositional verification approaches that are developed with Java to verify systems modeled in Eclipse by the OMG/INCOSE standard languages especially UML and SysML behavioral diagrams. The kernel of the verification procedure is based on the PRISM model checker. The proposed verification framework is successfully applied on real systems such as: communication protocol (Real Time Streaming Protocol and Time Triggered Ethernet Protocol), online systems, and ATM machines, etc.
· Research Assistant
Oct 2005 – June 2008: Metz University, France
I have developed a new classification framework that is based on the support vector machine. My framework is successfully applied on Biocomputing application to detect cancer in early stages. The result of this work is awarded as the best tool in the prestigious conference of Advanced Data-Mining and Application 2008. The tool is developed by C++ in VS 2005 and Cplex library.
· Research Assistant
June 1996 – Nov 1997, Sidi Bel Abbess University, Algeria
I have developed a new technique to generate automatically Fuzzy Inference System. The proposed technique is based mainly on clustering by using the similarity and Neural Networks. We applied this technique on different application area such as: systems control (arm robot), prediction (time series), and the fuzzy classification. The tool is developed by C++ and simulated in Matlab