Measuring DVB-T Signals - Digital Video and Audio Broadcasting Technology - page 410

phase response was derived, into the time domain by means of an inverse
fast Fourier transform provides the impulse response. The maximum
length of the calculable impulse response depends on the samples provided
by the channel estimation. Every third subcarrier supplies a contribution to
the channel estimation at some time, i.e. the distance between two interpo-
lation points of the channel estimation is 3 • ∆f, where ∆f corresponds to
the subcarrier spacing of the COFDM. The calculable impulse response
length is thus 1/3 ∆f, i.e. one third of the COFDM symbol period. In the
ideal case, the impulse response only consists of one main impulse at t = 0,
i.e. there is only one signal path. From the impulse response, multiple ech-
oes can be easily classified in accordance with delay and path attenuation.

*
*

RBW 10 kHz
VBW 100 kHz
SWT 10 s

Delta 2 [T1 ]
-39.71 dB
3.700000000 MHz Ref -10 dBm

*

Att 15 dB

*

-10

Marker 1 [T1 ]
-34.12 dBm
212.500000000 MHz

A

-20

Delta 1 [T1 ]
-41.17 dB
-3.700000000 MHz

1 RM
CLRWR

*

-30

1

-40

-50

PRN

-60

-70

2

1

-80

-90

-100

-110

Center 212.5 MHz 1.5 MHz/ Span 15 MHz

Date: 19.MAY.2005 13:41:17

Fig. 21.30. Spectrum of a DVB-T signal at the transmitter output before the mask
filter

21.7 Measuring the Shoulder Attenuation

The system does not utilize the full channel bandwidth, i.e. some of the 2K
or 8K subcarriers are set to zero so that no interference to adjacent chan-
nels will be caused. Due to nonlinearities, however, there are still outband
components and the effect on the spectrum and its shape has given rise to
the term 'shoulder attenuation'.
In the Standard, the permissible shoulder attenuation is defined as a tol-
erance mask. Fig. 21.30. the spectrum of a DVB-T signal at the power am-
plifier output, i.e. before the mask filter. To determine the shoulder at-
tenuation, different methods are defined and especially a relatively