THE INCREMENTAL MULTI-PARAMETER ALGORITHM
John Mather · 2005
The incremental multi-parameter (IMP) algorithm is described. The algorithm may be used to analyse digitised data from an array of sensors in order to estimate parameters of incident signals. It makes use of the conventional beamformer as a key component of an iterative analysis of the data, yet is shown IO outperform alternative eigenvector-based high resolution estimators, such as MUSIC, by 10-12 dB when resolving uncorrelated signals. IMP is also shown to be capable of resolving perfectly correlated signals at signal to noise ratios lower than required by MUSIC to detect uncorrelated signals. This is achieved without the need for spatial smoothing, enabling analysis of data from non-linear arrays. Finally, results are given which illustrate the robust performance of IMP when rhe antenna calibration is in error.