The position is part of the ANR ANTIPASTI project: Innovative architectures for distributed acoustic fiber optic sensors led by Thales and the LIP6 laboratory.
The deployment of DAS (Distributed Acoustic Sensing) systems is rapidly expanding, both for civil and military applications. The compactness, consumption and quantity of data generated are important obstacles to the miniaturization and embedding of a DAS system. The ANTIPASTI project offers an innovative and global approach to the numerical architectures in a DAS system (algorithms, tunable numerical precision and energy efficiency) as well as their integration with novel optical units.
The proposed position essentially concerns the reduction of the computing precision in order to be able to exploit embedded architectures more effectively and thus improve energy efficiency. The project will benefit from the numerical validation software CADNA (http://cadna.lip6.fr) and PROMISE (http://promise.lip6.fr) developed at the LIP6 laboratory to enable a mixed precision implementation, in order to obtain better performance while guaranteeing the numerical stability of the computation.
A complete description is given in: https://www-pequan.lip6.fr/~jezequel/offers.html
