banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Sparse array design using statistical restricted isometry property
Rent this article for
Access full text Article
1. Becker, K. M. , and Preston, J. R. (2003). “The ONR Five Octave Research Array (FORA) at Penn State,” in Proceedings of OCEANS 2003, pp. 26072610.
2. Calderbank, R. , Howard, S. , and Jafarapour, S. (2010). “Construction of a large class of deterministic sensing matrices that satisfy a statistical isometry property,” IEEE J. Sel. Top. Signal. Process. 4, 358374.
3. Candes, E. J. , and Tao, T. (2005). “Decoding by linear programming,” IEEE Trans. Inf. Theory 51, 42034215.
4. Cohen, A. , Dahmen, W. , and DeVore, R. (2009). “Compressed sensing and best k-term approximation,” J. Am. Math. Soc. 22, 211231.
5. Dai, W. , and Milenkovic, O. (2009). “Subspace pursuit for compressive sensing signal reconstruction,” IEEE Trans. Inf. Theory 55, 22302249.
7. Devore, R. A. (2007). “Deterministic constructions of compressed sensing matrices,” J. Complexity 23, 918925.
8. Edelmann, G. F. , and Gaumond, C. F. (2011). “Beamforming using compressive sensing,” J. Acoust. Soc. Am. 130, EL232EL237.
9. Egidi, L. , and Manzini, G. (2011). “Spaced seeds design using perfect rulers,” Technical Report No. TR-INF-2011-06-01-UNIPMN (Computer Science Department, Universita del Piemonte Orientale), http://www.di.unipmn.it (Last viewed June 27, 2013).
View: Figures


Image of FIG. 1.

Click to view

FIG. 1.

(Color online) (a) The StRIP is plotted for the Golomb, best and worst of 200 random arrays. These arrays have length 128 with 14 elements. The Golomb array possesses a smaller StRIP than either of the two random arrays. However, the best random array has StRIP values very close to those of the Golomb array. (b) The null space property is plotted for the Golomb, worst and best random arrays over the range 2 ≤  ≤ 14, but the null space property is determined with 2 sparse components. The rank of is only 14; there are 14 directions that do not contribute to the data. (c) The StRIP is plotted for the Wichmann array, the best and worst of 200 random arrays. These arrays have lengths 124 with 19 elements. The Wichmann array has a StRIP that is never smaller than the best random array and is sometimes greater than the worst random array.

Image of FIG. 2.

Click to view

FIG. 2.

(Color online) Bearing time (BT) plots for three arrays computed from data from the Five Octave Research Array (FORA). Each BT plot shows the incoherent average from 100 to 250 Hz. (a) The conventional array shows poor resolution, especially in the regions near endfire. (b) The Golomb array displays higher resolution than the conventional in all regions. Note the resolution of the source at approximately 10°. (c) The BT plot of the Wichmann subsampled array shows remarkably worse performance than the Golomb subsampled array. The quiet source at approximately 30° is nearly absent.


Article metrics loading...



The numerical application of the statistical reduced isometry property (StRIP) and statistical null space property (SNSP) is presented and demonstrated for the design of underwater acoustic line arrays. This recent approach predicts the theoretical utility of specific subsampled arrays for compressive sensing. Three subsamplings are presented: Random, Golomb, and Wichmann. The Golomb array has no repeated spacings. The Wichmann array includes every possible interval of spacings. The SNSP is shown insensitive to the cases presented. The StRIP of the Golomb array predicts superior invertibility and is shown to perform well using at-sea data.


Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Sparse array design using statistical restricted isometry property