Output power based partial adaptive array design

Abstract

We present an approach to partial adaptive beamforming based on a subspace selection technique. The subspace used is obtained analytically from the minimum interference output power of the fully adaptive ganrmlizrd sidrlobr canceler. We express the minimum interference output power as a function of the signal constraint, the noise autocorrelation matrix and the eigenvectors of the noise autocorrelation matrix projected onto the null space of the signal constraint.Computer simulations of this approach illustrate robust performance under a variety of adverse environments and array structures. Multiple broadband interferers can be incorporated within this framework. Simulations verify that ? amay design based on this method offers superior noise rejection as compared to other subspace methods [2,5]. The performance advantage is " s t apparent when the rank of the adaptive subspace is small. Introduction:A typical application for an array is to detect a signal s ( k ) , so that, the array output y ( k ) is an estimate of s(k). A c a m " approach to this problem is to constrain the airay. response in the direction of arrival (DOA) of Lhe S U I C~ and then minimise the D U P U~ power lu(k)12 This is a linearly constrained minimum variance (LCMV) approach to beamforming and is formulated as follows: find w s.t. n$~Iy(b)l~and C'w = f

Subspace selection for partially adaptive sensor array processing

“…However, this technique does not directly consider the MMSE performance measure, which is a function of not only the space spanned by the noise covariance matrix but also of the cross-correlation between the desired signal and the noise process. It is noted that Byerly [7] discovered that the eigenvectors corresponding to the largest eigenvalues were not necessarily the best selection, but there was no derivation provided and the approach obtained herein provides a more general solution.…”

Reduced-rank adaptive multiuser detection in hybrid direct-sequence time-hopping ultrawide bandwidth systems

“…where C C C ( ) is the spreading matrix associated with the bits transmitted before bit , which can be expressed as shown in (17). From (12), (13) and (17), it can be seen that C C C ( ) consists of the last = ((…”

“…The reduced-rank schemes are derived based on the principles of principal component (PC), cross-spectral metric (CSM) and Taylor polynomial approximation (TPA), respectively [7,[13][14][15][16][17][18][19]. The BER performance of the hybrid DS-TH UWB systems using reduced-rank RLS adaptive MUD is investigated, when communicating over UWB channels modelled by the Saleh-Valenzuela (S-V) channel model [20].…”

Short Data Record Adaptive Filtering: The Auxiliary-Vector Algorithm

“…In [14] and [15] the p columns of T were chosen to be the p maximum eigenvalue eigenvectors of the blocked data autocorrelation matrix B H RB. If, however, the columns of T have to be eigenvectors of B H RB (there is no documented technical optimality to this approach), then the best way to choose them in the minimum output variance p-rank approximation sense was presented in [16]: Select the p eigenvectors q i of B H RB, with corresponding eigenvalues λ i , that maximize |v H RBq i | 2 /λ i , i = 1, . .…”