User`s guide

Correlation
5-102
5Correlation
Purpose Compute the correlation along the columns of two inputs.
Library Statistics
Description The Correlation block computes the cross-correlation of corresponding columns
(channels) of the M
u
-by-N input matrix u and M
v
-by-N input matrix v. The
frame status of both inputs must be the same. The block does not accept
sample-based full-dimension matrix inputs or 2-D row vector inputs.
Frame-Based Inputs
Matrix inputs must be frame-based. The output, y, is a frame-based
(M
u
+M
v
-1)-by-N matrix whose jth column has elements
where
∗ denotes the complex conjugate. Inputs u and v are zero when indexed
outside of their valid ranges. When both inputs are real, the output is real;
when one or both inputs are complex, the output is complex.
When one input is a column vector (single channel) and the other is a matrix
(multiple channels), the single-channel input is independently cross-correlated
with each channel of the multichannel input. For example, if u is a M
u
-by-1
column vector and v is an M
v
-by-N matrix, the output is an (M
u
+M
v
-1)-by-N
matrix whose jth column has elements
Sample-Based Inputs
Matrix inputs cannot be sample based, so all sample-based inputs are column
vectors or 1-D vectors. (the block does not support 2-D row vector inputs.) If u
and v are sample-based vectors with lengths M
u
and M
v
, the Correlation block
performs the vector cross-correlation
y
ij,
u
kj,
v
kiM
v
–+()j,
*
1 iM
u
M
v
1–+()≤≤
k 1=
max M
u
M
v
,()
∑
=
y
ij,
u
k
v
ki– M
v
+()j,
*
1 iM
u
M
v
1–+()≤≤
k 1=
max M
u
M
v
,()
∑
=