Rohde & Schwarz R&S FSL3 - Frequency and Clock Offset

R&S FSL WiMAX, WiBro Measurements (Options K92/K93)
1300.2519.12 2.151 E-11
Frequency and Clock Offset
The measurement software allows selectable compensation of phase, timing and gain errors based on
pilot estimations. However, in case of subchannelization the number of pilots is decreased. Bursts with
odd subchannel indices do not provide pilots at all. The following table lists the restrictions on the
tracking ability for subchannelization:
Tracking Subchannel Index
16 (8 Pilots)
8, 24 (4 Pilots)
4, 12, 20, 28 (2 Pilots)
2, 6, 10, 14, 18, 22, 26, 30
(1 Pilot)
1, 3, 5, 7, 9, 11, 13, 15, 17, 19,
21, 23, 25, 27, 29, 31 (No
Pilot)
Phase
Available Available Available, but uses rough
frequency offset estimation
from the synchronization step
only
Timing
Available Not available Not available
Gain
Available Available Not available
While the tracking functionality has to use pilot based estimates, the actual results for frequency and
clock offset in the result summary can be data aided. In case of subchannelization the final estimation
of frequency and clock offset is done using the already decided data sequence, which gives stable
results even without pilots.
EVM
The error vector magnitude of a single constellation point is defined by
used
used
2
/2
2
/2
used
0
(, ) (, )
EVM(l,k)
1
(, )
kN
kN
k
rlk alk
alk
N
=
=
&

=
%
where ),( klr is the received constellation point and ),( kla is the transmitted constellation point at the
l
th
symbol and carrier number k .
In case of subchannelization, it is required by the standard to include the unallocated carriers
unalloc
k by
assuming
0),( =
unalloc
kla in the denominator of the EVM calculation.
Thus the EVM All Carriers result for one burst in the result summary equals
used
used
used
used
/2
2
1/2
used
0
/2
2
1/2
used
0
11
(, ) (, )
EVM_All_Carr
11
(, )
kN
L
lkN
k
kN
L
lkN
k
rlk alk
LN
alk
LN
=
==
&
=
==
&

=
%%
%%
where
L
is the number of symbols in the burst.
This definition is according to the relative constellation error defined in the IEEE 802.16–2004 standard.
Using the equations above, the error power is normalized by the average transmitted power in all 200
carriers. Please notify, that by this definition the same absolute error power leads to different EVM
results depending on the number of allocated carriers in case of subchannelization.