Sparsity Order Estimation under Constrained Budget

Mehrdad Khalaji | Mahmoud Ferdousizadeh | Arash Amini

---

URL :   http://research.shahed.ac.ir/WSR/WebPages/Report/PaperView.aspx?PaperID=43535
Date :  2017/06/05
Publish in :    (signal processing withadaptive sparse structured representation(SPARS 2017

Link :  spars2017.lx.it.pt
Keywords :Order

Abstract :

---

In many applications of compressed sensing it is either necessary or beneficial to have a priori information about the sparsity order x_0 of an unknown sparse vector x. As shown in 1, x_1/x_2 provides a more reliable measure of compressibility than the sparsity order. The latter could also be estimated by observing the vector through random linear measurements following Cauchy and Gaussian distributions. In this paper, we determine the optimal number of Cauchy and Gaussian measurements to minimize the estimation error under limited measurement budget.

spars2017.lx.it.pt