New radiolocation technique for intelligent antennas

This paper treats the estimation of narrowband Direction of Arrival (DoA) of far field and punctual sources using one dimensional smart antennas, in the first part we demonstrate a new expression of the propagator operator which is subspace based technique, using the orthonormal matrix Q obtained from QR factorization of data's second order statistics, next we present a new high resolution technique using the decreasing exponential operator, which is a function inspired from several physics laws, this matrix function is adjustable using minimum eigenvalue of spectral matrix. The introduced mechanism allows an approximate binarization of system's eigenvalues to obtain a projector into the noise subspace. Additionally we demonstrate that if the powers of transmitted signals are known a priori, an accurate estimation of noise power can be calculated using the diagonal components of spectral matrix. Numerical simulation is presented using a non uniform linear array, the obtained results verify that the proposed technique has the same resolution power as the eigenspace based spectral techniques.