A LINEARLY CONSTRAINED WIDEBAND ADAPTNE ARRAY ANTENNA WITH ORTHOGONAL flLTER STRUCTURE
J. Scott Goldstein · 1992
A linearly constrained wideband adaptive array sensor is described whose adaptive processor is realized with amultichannel orthogonalfilter structure. It is shown that the transform domain processor utilizing the computationally simple LMS algorithm is capable of convergence speeds which rival the least squares methods which are more qensive to implement. Furthermore. there is no additional cost in terms of the adaptive coefficient requirements (andonly a modest increase in non-adaptive computational requirements) when compared to the standard tapped-delay-line structure array utilizing the LMS algorithm. The tapped-delay-line (TDL) structure linearly constrained wideband adaptive array antenna (l) may be implemented in a partitioned form termed a Generalized Sidelobe Canceller (GSC) (2). This is depicted in figure 1 for a signal aligned linearly constrained minimum variance array composed of K sensors and J taps per fdter. The conventional beamforming matrix Wcenforces the look direction constraint. This beamforming matrix takes the form =-K The desired signal is blocked from the adaptive processor through the signal blocking matrix Ws, which for a single constraint is of dimension (K-I) x K. The full rank matrix W, is composed of rows ri, where i=l to K-I, such that ril=O (2) The block denoted Win figure 1 is composed of K-1 TDL filters of order J. The values of the adaptive coefficients of these filters are stacked by column to form the (K-l)J dimensional weight vector. It has been shown (2) that the GSC form array converges to the Wiener-Hopf solution when the weight vector is updated via the unconshained LMS (3) algorithm