“Random matrices in signal processing and machine learning”. Dirk Slock, Romain Couillet.

The theory of large dimensional random matrices has been a major driver of the MIMO revolution in wireless communications since the 2000’s, thanks to its capacity of harnessing the system performance of involved multivariate com- munication systems (multi-user MIMO, massive MIMO, multi-cell processing). Today, with the emerging needs for large dimensional data analytics (both in big data but also in large array processing), random matrix theory faces new challenges and needs to renew itself by providing original tools to tackle an entirely new set of random matrix models.

The main objective of the special session is to provide the attendance of SSP 2016 a glimpse on this new toolbox that will surely be the basis for future advances in the theoretical analysis of large dimensional datasets and systems. This goal will be met thanks to (a minimum of) seven poster presentations spanning a wide range of applications.