École normale supérieure de Cachan

About: École normale supérieure de Cachan is a education organization based out in Cachan, Île-de-France, France. It is known for research contribution in the topics: Decidability & Finite element method. The organization has 2717 authors who have published 5585 publications receiving 175925 citations.

Cryptographic protocol analysis on real c code

Abstract: Implementations of cryptographic protocols, such as OpenSSL for example, contain bugs affecting security, which cannot be detected by just analyzing abstract protocols (e.g., SSL or TLS). We describe how cryptographic protocol verification techniques based on solving clause sets can be applied to detect vulnerabilities of C programs in the Dolev-Yao model, statically. This involves integrating fairly simple pointer analysis techniques with an analysis of which messages an external intruder may collect and forge. This also involves relating concrete run-time data with abstract, logical terms representing messages. To this end, we make use of so-called trust assertions. The output of the analysis is a set of clauses in the decidable class $\mathcal{H}_1$, which can then be solved independently. This can be used to establish secrecy properties, and to detect some other bugs.

Artificial Neural Networks Applied to Taxi Destination Prediction

Abstract: We describe our first-place solution to the ECML/PKDD discovery challenge on taxi destination prediction. The task consisted in predicting the destination of a taxi based on the beginning of its trajectory, represented as a variable-length sequence of GPS points, and diverse associated meta-information, such as the departure time, the driver id and client information. Contrary to most published competitor approaches, we used an almost fully automated approach based on neural networks and we ranked first out of 381 teams. The architectures we tried use multi-layer perceptrons, bidirectional recurrent neural networks and models inspired from recently introduced memory networks. Our approach could easily be adapted to other applications in which the goal is to predict a fixed-length output from a variable-length sequence.

Energy harvesting from vibration using a piezoelectric membrane

Abstract: In this paper we investigate the capability of harvesting the electric energy from mechanical vibrations in a dynamic environment through a unimorph piezoelectric membrane transducer. Due to the impedance matrices connecting the efforts and flows of the membrane, we have established the dynamic electric equivalent circuit of the transducer. In a first study and in order to validate theoretical results, we performed experiments with a vibrating machine moving a macroscopic 25 mm diameter piezoelectric membrane. A power of 1.8 mW was generated at the resonance frequency (2.58 kHz) across a 56 kQ optimal resistor and for a 2 g acceleration.