DOA estimation in compressed sensing framework by adaptively exploiting prior support information

Jiao Yang, Zhi Zheng, Bo Yan, Yan Ge, Kehong Liu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we propose a novel greedy pursuit algorithm for direction-of-arrival (DOA) estimation of multiple moving sources with time-variant DOAs. A prior support set of DOAs and a metric information are assumed to be available. Different from the traditional greedy pursuit algorithms which don't exploit prior support but select entries over the entire signal index set, or exploit the prior statically, we adaptively exploit the previously estimated prior support set by firstly selecting directly entries from the prior support based on the metric information. The exact number and values of unchanged DOAs are not required. Simulation results prove that the proposed method obtains more accurate estimates even for a small number of snapshots and low SNR.

Original languageEnglish
Title of host publication2016 2nd IEEE International Conference on Computer and Communications, ICCC 2016 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1895-1899
Number of pages5
ISBN (Electronic)9781467390262
DOIs
Publication statusPublished - 10 May 2017
Externally publishedYes
Event2nd IEEE International Conference on Computer and Communications, ICCC 2016 - Chengdu, China
Duration: 14 Oct 201617 Oct 2016

Publication series

Name2016 2nd IEEE International Conference on Computer and Communications, ICCC 2016 - Proceedings

Conference

Conference2nd IEEE International Conference on Computer and Communications, ICCC 2016
Country/TerritoryChina
CityChengdu
Period14/10/1617/10/16

Keywords

  • Adaptive algorithm
  • Compressed sensing
  • Direction-of-arrival (DOA) estimation
  • Prior support

Fingerprint

Dive into the research topics of 'DOA estimation in compressed sensing framework by adaptively exploiting prior support information'. Together they form a unique fingerprint.

Cite this