Here is an interesting architecture for sensing signals sharing much in common
We propose two compressive multiplexers for the efﬁcient sampling of ensembles of correlated signals. We show that we can acquire correlated ensembles, taking advantage of their (a priori-unknown) correlation structure, at a sub-Nyquist rate using simple modulation and ﬁltering architectures. We recast the reconstruction of the ensemble as a low-rank matrix recovery problem from generalized linear measurements. Our theoretical results indicate that we can recover an ensemble of M correlated signals composed of R M independent signals, each bandlimited to W=2 Hz, by taking O(RW logq W) samples per second, where q > 1 is a small constant.
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.