NS32FV16 Datasheet, PDF (55/104 Pages) Texas Instruments – Advanced Imaging/Communication Signal Processors

3 0 Functional Description (Continued)
Syntax
EXEC VCPOLY
15
11 10
0
10000
101 0001 1000
Operation
VDIST Vector Euclidean Distance
The VDIST instruction calculates the square of the Euclide-
an distance between corresponding elements of the X and
Y complex vectors and places the result in the Z real vec-
tor
Syntax
EXEC VDIST
complex X Z
real Y
complex temp
15
11 10
0
10000
100 1111 1110
temp re 4 (real) Y 0 * X 0 re
temp im 4 0
Operation
for (n40 n ^k LENG n00)
complex X Y
real Z
Z n 4 (complex) Z n * X n01 0 temp
Z LENG re 4 (real) (Z LENG re *
X LENG01 re 0 Y 0 * temp re)
Y ADDR 4 Y 1
Note The LENG parameter for this operation must be greater than 1
VDECIDE Vector Nearest Neighbor Decision Logic
te ^{The VDECIDE instruction is used to implement nearest}
neighbor decision in Quadrature Amplitude Modulation
(QAM) modem applications The input is the X complex vec-
tor The output is placed in the Z integer vector which can
be used as an index vector to extract information from look-
up tables The indicated constant values are taken from the
Y vector
le ^Syntax
EXEC VDECIDE
15
10000
Operation
11 10
0
100 1111 0000
o _{complex X}
aligned real Y
real Z
s ^{internal register complex temp}
for (n40 n ^k LENG n00)
temp re 4 min (X n re Y 0 low)
temp im 4 min (X n im Y 0 high)
b ^{temp re 4 max (temp re Y 1 low)}
temp im 4 max (temp im Y 1 high)
X n 4 temp
Z n 4 (real) ((temp re * Y 2 low)
l ^{(extended) Y 3 low)}
((temp im * Y 2 high) (extended)
O ^{Y 3 high)}
for (n40 n ^k LENG n00)
Z n 4 (real) (X n re 1 Y n re)** 2a
(X n im 1 Y n im) **2
VFFT Vector Fast Fourier Transform
The VFFT instruction implements one pass of in-place FFT
vector update according to the radix-2 FFT method
Syntax
EXEC VFFT
15
11 10
0
10000
101 0000 0110
Operation
complex X Y Z
complex temp
for (n40 n ^k LENG
n00)
temp 4 (complex) (Z n 0 X n * Y n )
Y n 4 (complex) (Z n 1 X n * Y n )
Z n 4 temp
VESIIR Vector Extended Single-Pole IIR
This instruction is not supported by the NS32FX161
The VESIIR instruction performs a special form of an Infi-
nite-Impulse Response (IIR) filter The samples and coeffi-
cient are given as real values as well as the output result
However the accumulation is performed using extended-
precision arithmetic
Syntax
EXEC VESIIR
15
11 10
0
Note Y INCR must be specified as 1 and Y WRAP must be specified as 3
10000
101 0011 0111
54