Autocalibration analytique d'antenne

A new self-calibration method (i.e. without cooperating source) is presented. It is based on the fact that steering vectors have unit modulus. A first algorithm is proposed, that allows to identify steering vectors analytically, when the noise is spatially white. When the noise is colored, another algorithm is proposed, that also exploits the fact that source complex envelopes take discrete values. This second algorithm allows, after an analytical estimation of the impinging signals and a clustering step, to eventually identify the steering vectors by an explicit Least-Squares solution. These self-calibrating algorithms are illustrated with the help of the MUSIC localization method.