Digital Signal Processing Reference
In-Depth Information
5
0
-5
-10
-15
-20
-25
MVDR
Beam towards -5 Degrees
Beam towards 5 Degrees
Beam towards 20 Degrees
Beam towards 27 Degrees
-30
-35
Jammer
at 27
-40
-100
-80
-60
-40
-20 0
Angles of arrival (deg)
20
40
60
80
100
FIgure 14.4
Array response performance at the eight RHCP antenna arrays with single and
multiple MVDR methods with two RHCP interferers at 27 and 50 degrees.
14.5
Combined Nulling and DOA Estimation
Interference DOA can be obtained using subspace methods such as the MUSIC algo-
rithm. Eigendecomposition is typically applied to the covariance matrix estimated from
the data collected over an observation period,
ˆ
H
ru u
=
1
Σ
1
,
(14.23)
1
where
Σ
1
is a 2
N
-by-2
N
matrix representing the eigenvalue diagonal matrix, and
u
1
is
a unitary matrix whose columns represent the corresponding eigenvectors. The signal
subspace is spanned by the eigenvectors associated with the dominant eigenvalues of
ˆ
.
The MUSIC spectrum is given by [31]
1
P
()
θ
=
()
,
(14.24)
MUSIC
H
H
a
()
θ
uua
θ
nn
where
u
n
is a matrix of the eigenvectors corresponding to the noise eigenvalues. The
MUSIC algorithm as well as other eigenstructure methods require proper sampling
of the data for covariance matrix estimation. These methods can be computationally
demanding for GPS receivers, often call for large memory, and cannot be used effec-
tively in simple analog adaptive loops. An alternative way is to build on the fact that
multiple beamformers produce different sets of weights, and each carries interference
Search WWH ::
Custom Search