E.5 Estimation of time offset
36.1043GPPBase Station (BS) radio transmission and receptionEvolved Universal Terrestrial Radio Access (E-UTRA)Release 18TS
The observation period for determining the sample timing difference shall be 1 ms.
For 0.37 kHz, the observation period for determining the sample timing difference shall be 6 ms.
For 2.5 kHz, the observation period for determining the sample timing difference shall be 1 ms.
In the following represents the middle sample of the EVM window of length (defined in E.5.1) or the last sample of the first window half if is even.
is estimated so that the EVM window of length is centred on the measured cyclic prefix of the considered OFDM symbol. To minimize the estimation error the timing shall be based on the primary synchronization signal and reference signals. To limit time distortion of any transmit filter the reference signals in the 1 outer RBs are not taken into account in the timing estimation
Two values for are determined:
and
where if is odd and if is even.
When the cyclic prefix length varies from symbol to symbol (e.g. time multiplexed MBMS and unicast) then shall be further restricted to the subset of symbols with the considered modulation scheme being active and with the considered cyclic prefix length type.
E.5.1 Window length
Table E.5.1-1 and Table E.5.1-1a below specify EVM window length (W) for normal CP, the cyclic prefix length is 160 for symbols 0 and 144 for symbols 1-6.
Table E.5.1-2, Table E.5.1-2a ,Table E.5.1-2b ,Table E.5.1-2c and Table E.5.1-2d specify the EVM window length (W) for extended CP for 15 kHz, 7.5 kHz ,1.25 kHz, 2.5 kHz and 0.37 kHz sub-carrier spacing, the cyclic prefix length is 512, 1024 ,6144, 3072 and 9208 respectively.
Table E.5.1-1: EVM window length for normal CP for E-UTRA
Channel |
FFT size |
Cyclic prefix length for symbols 0 in FFT samples |
Cyclic prefix length for symbols 1‑6 in FFT samples |
EVM window length W |
Ratio of W to total CP for symbols 1‑6(Note 1) [%] |
|
1.4 |
128 |
10 |
9 |
5 |
55.6 |
|
3 |
256 |
20 |
18 |
12 |
66.7 |
|
5 |
512 |
40 |
36 |
32 |
88.9 |
|
10 |
1024 |
80 |
72 |
66 |
91.7 |
|
15 |
1536 |
120 |
108 |
102 |
94.4 |
|
20 |
2048 |
160 |
144 |
136 |
94.4 |
|
Note 1: These percentages are informative and apply to symbols 1 through 6. Symbol 0 has a longer CP and therefore a lower percentage. |
Table E.5.1-1a: EVM window length for normal CP for NB-IoT
FFT size |
Cyclic prefix length for symbols 0 in FFT samples |
Cyclic prefix length for symbols 1‑6 in FFT samples |
EVM window length W |
Ratio of W to total CP for symbols 1‑6(Note 1) [%] |
128 |
10 |
9 |
3 |
33.3 |
Note 1: These percentages are informative and apply to symbols 1 through 6. Symbol 0 has a longer CP and therefore a lower percentage. |
Table E.5.1-2 EVM window length for extended CP for 15 kHz sub-carrier spacing
Channel Bandwidth [MHz] |
FFT size |
Cyclic prefix in FFT samples |
EVM window length W |
Ratio of W to total CP (Note 1) |
1.4 |
128 |
32 |
28 |
87.5 |
3 |
256 |
64 |
58 |
90.6 |
5 |
512 |
128 |
124 |
96.9 |
10 |
1024 |
256 |
250 |
97.7 |
15 |
1536 |
384 |
378 |
98.4 |
20 |
2048 |
512 |
504 |
98.4 |
Note 1: These percentages are informative. |
Table E.5.1-2a EVM window length for extended CP for 7.5 kHz sub-carrier spacing
Channel Bandwidth [MHz] |
FFT size |
Cyclic prefix in FFT samples |
EVM window length W |
Ratio of W to total CP (Note 1) |
1.4 |
256 |
64 |
56 |
87.5 |
3 |
512 |
128 |
116 |
90.6 |
5 |
1024 |
256 |
248 |
96.9 |
10 |
2048 |
512 |
500 |
97.7 |
15 |
3072 |
768 |
756 |
98.4 |
20 |
4096 |
1024 |
1008 |
98.4 |
Note 1: These percentages are informative. |
Table E.5.1-2b EVM window length for extended CP for 1.25 kHz sub-carrier spacing
Channel Bandwidth [MHz] |
FFT size |
Cyclic prefix in FFT samples |
EVM window length W |
Ratio of W to total CP (Note 1) |
1.4 |
1536 |
384 |
336 |
87.5 |
3 |
3072 |
768 |
696 |
90.6 |
5 |
6144 |
1536 |
1488 |
96.9 |
10 |
12288 |
3072 |
3000 |
97.7 |
15 |
18432 |
4608 |
4536 |
98.4 |
20 |
24576 |
6144 |
6048 |
98.4 |
Note 1: These percentages are informative. |
Table E.5.1-2c EVM window length for extended CP for 2.5 kHz sub-carrier spacing
Channel Bandwidth [MHz] |
FFT size |
Cyclic prefix in FFT samples |
EVM window length W |
Ratio of W to total CP (Note 1) |
1.4 |
768 |
192 |
168 |
87.5 |
3 |
1536 |
384 |
348 |
90.6 |
5 |
3072 |
768 |
746 |
96.9 |
10 |
6144 |
1536 |
1500 |
97.7 |
15 |
9216 |
2304 |
2268 |
98.4 |
20 |
12288 |
3072 |
3024 |
98.4 |
Note 1: These percentages are informative. |
Table E.5.1-2d EVM window length for extended CP for 0.37 kHz sub-carrier spacing
Channel Bandwidth [MHz] |
FFT size |
Cyclic prefix in FFT samples |
EVM window length W |
Ratio of W to total CP (Note 1) |
1.4 |
5184 |
576 |
504 |
87.5 |
3 |
10368 |
1152 |
1044 |
90.6 |
5 |
20736 |
2302 |
2230 |
96.9 |
10 |
41472 |
4604 |
4498 |
97.7 |
15 |
62208 |
6906 |
6796 |
98.4 |
20 |
82944 |
9208 |
9060 |
98.4 |
Note 1: These percentages are informative. |