Implementation and convergence considerations of a linearly constrained adaptive array
Abstract
Frost (1972) introduced a linearly constrained optimization algorithm that allows certain main beam properties to be preserved while good cancellation is attained. An openloop implementation of this algorithm is developed here. This implementation is shown to be equivalent to the technique developed by Jim (1977), Griffiths and Jim (1982), and Buckley and Griffiths (1982), whereby the constrained problem is reduced to an unconstrained problem. Analytical results are presented for the convergence rate when the sampled matrix inversion (SMI) or GramSchmidt (GS) algorithm are employed. It is shown that, if the SMI or GS algorithm is employed, then the transientweightingvector (TWV) solution for the constrained problem is identical to the equivalent TWV solution for the reduced unconstrained implementation.
 Publication:

IEEE Transactions on Aerospace Electronic Systems
 Pub Date:
 March 1990
 DOI:
 10.1109/7.53459
 Bibcode:
 1990ITAES..26..263G
 Keywords:

 Adaptive Filters;
 Antenna Arrays;
 Cancellation Circuits;
 Convergence;
 Interference Immunity;
 Matrices (Mathematics);
 Algorithms;
 Constraints;
 Orthogonality;
 Communications and Radar