Description
Spectrum sensing is a crucial ingredient of various types of applications, such as frequency spectrum sensing for cognitive radio and angular spectrum sensing for direction of arrival estimation. A popular tool that has recently been introduced for spectrum sensing is compressive sampling. Adopting this technique, it is possible to sample the measured signal below the Nyquist rate without compromising the reconstruction error, under the condition that the measured signal is sparse in some domain (frequency, angular, etc.). Current compressive spectrum sensing techniques mainly focus on reconstructing the signal itself. However, for many applications, this is overkill and estimating the power on every frequency or angle is sufficient. In this plenary, we therefore present a new framework for reconstructing the power spectrum from compressive measurements, labeled as compressive power spectrum sensing. This allows for improved compression rates, and if designed properly, it even works without any sparsity constraints on the spectrum, i.e., it can also be used to reconstruct non-sparse spectra. In this talk, we will first introduce the concept of compressive power spectrum sensing, and then more specifically focus on reconstructing frequency and angular power spectra from compressive measurements using for instance multi-coset sampling and non-uniform antenna arrays.| Period | 16 Jun 2013 |
|---|---|
| Event title | STATOS 2013: Statistics, Optimization, and Signal Processing |
| Event type | Other |
| Location | Darmstadt, GermanyShow on map |