J4 ›› 2011, Vol. 38 ›› Issue (6): 113-116+139.doi: 10.3969/j.issn.1001-2400.2011.06.018

• Original Articles • Previous Articles     Next Articles

Small sample signals' decorrelation using the coherent accumulation vector' spatial smoothing

LI Tao1,2;LI Guolin1;XU Heng1;LIAO Huirong2
  

  1. (1. Institute 220 of 91550 PLA Troops, Dalian  116021, China;
    2. Dept. 7 of Naval Aeronautics and Astronautics Univ., Yantai  264001, China)
  • Received:2010-09-25 Online:2011-12-20 Published:2011-11-29
  • Contact: LI Tao E-mail:littff@126.com

Abstract:

When estimating the DOA of narrowband signals, if the snapshot number is small, decorrelation operation is necessary because for all the snapshots, their amplitude and phase are unchanged, which makes signals from different directions coherent. For this kind of signals, existing decorrelation algorithms always need too much computation or are sensitive to SNR. This paper proposes taking coherent accumulation operation for array outputs, and the resulting vector is used to construct a matrix by forward and backward smoothing. By taking SVD of this matrix, the signal and noise subspace is obtained then. Simulation results indicate that this method is of excellent noise-resistance performance with great computational complexity reduction in contract to existing decorrelation algorithms. It has the same performance as ESVD algorithms based on eigen-decomposition, when the SNR is higher than -14dB.

Key words: narrowband signal, DOA estimation, decorrelation, coherent accumulation

CLC Number: 

  • TN958