Artículo Recortado de:

http://lteuniversity.com/get_trained/expert_opinion1/b/nishithtripathi/archive/2015/05/29/reference-
signal-power-boosting-in-lte.aspx

Artículo publicado por:

Nishith Tripathi

Nishith
29 May 2015 7:14 PM
Reference Signal Power Boosting in LTE
Each cell in an LTE radio network sends a cell-specific reference signal (RS) from
its transmit antennas. The transmit power of a resource element (RE) carrying
such reference signal can be set to be the same as, greater than, or less than the
transmit power of an RE carrying Physical Downlink Shared Channel (PDSCH). Let’s
take a quick look at the reference signal power boosting, where the RS RE uses
more power than the PDSCH RE. The RS power boosting may or may not be
desirable from the perspective of the RF performance.

The relative transmit power levels of the RS and the PDSCH have implications on
the downlink channel estimation, the amount of downlink interference, and the
interpretation and the use of the Channel Quality Indicator (CQI) by the eNodeB.

For example, if the RS power is increased, the UE could potentially make the RS
measurements (e.g., RSRP and RSRQ) more easily and potentially quantify the
downlink channel conditions more reliably. However, the overall interference on
the RS RE for a given cell would increase due to multiple neighboring cells
transmitting more power on their own RS REs. If the signal-to-interference-plus-
noise ratio (SINR) estimated for the PDSCH degrades by a significant amount, the
CQI being reported by the UE would be lower. If the reported CQIs are relatively
lower, the eNodeB would aim for a lower target throughput by taking actions
such as the increased amount of Turbo coding in the PDSCH transmissions. The
user-experienced throughput could thus be somewhat lower when RS power is
boosted.
However, if the enhanced channel estimation and increased reliability of the
PDSCH reception lead to fewer HARQ retransmissions, throughput could actually
increase in case of the RS power boosting.
In summary, the theoretical impact of the RS power boosting on the RF
performance is not definitive.
Field testing with varying levels of RS power boosting and varying levels of traffic
loading is recommended to determine the suitability of the RS power boosting.

The eNodeB broadcasts the transmit power levels of the RS and the PDSCH in SIB
2 using the parameters referenceSignalPower, PA, and PB. The transmit power of
an RE carrying the RS (in dBm) is specified as referenceSignalPower.
PA influences a parameter called ρA, which is the ratio of the transmit power of
the PDSCH RE and the transmit power of the RS RE. ρA is applicable to the OFDM
symbols that do not carry RS.
PB establishes the relationship between ρA and ρB, where ρB is the ratio of the
transmit power of the PDSCH RE and the transmit power of the RS RE in the
OFDM symbols that carry RS.
PA ranges from 0 to 7 and corresponds to the range from -6 dB to +3 dB for ρA.
PB ranges from 0 to 3 and corresponds to the range from 5/4 to 1/2 for (ρB /ρA)[1].

Let’s take two numerical examples. Assume that a 30 W power amplifier is used
for a transmit antenna of an eNodeB and that 10 MHz downlink bandwidth is
deployed in a cell. The nominal transmit power per subcarrier is (30 W/600)= 50
mW. During an OFDM symbol where no RS is present, each subcarrier of the
PDSCH is allocated 50 mW.
Example 1:
PA =2 and PB =1 (with 3 dB power boosting for RS)

PA =2
implies
ρA = 0.5 or -3 dB.

Hence,
(Power on PDSCH RE/power on RS RE)= 0.5.

Since the transmit power allocated to PDSCH RE is the nominal power level of 50
mW, the transmit power allocated to RS RE is
(Power on PDSCH RE/0.5)= (50 mW/0.5)= 100 mW.

During an OFDM symbol carrying the RS, the number of REs carrying the RS from
one transmit antenna is
(50 Physical Resource Blocks * 2 REs/Physical Resource Block)= 100 REs.

Furthermore, a given antenna does not transmit any power on a set of 100 REs,
because such set is used by a different transmit antenna.
Hence, out of 600 REs in an OFDM symbol carrying the RS, 100 REs are subject to
RS power boosting, 100 REs have no transmit power, and remaining 400 REs have
nominal power levels.

The total transmit power during the RS-carrying OFDM symbol would be
(100 subcarriers * 100 mW per subcarrier for power-boosted RS REs) + (100 subcarriers* 0
mW for null REs) + (400 subcarriers * 50 mW per subcarrier for non-RS REs)= 30 W.

Hence, when PA =2 and PB =1, each RS RE is allocated 100 mW, while a non-RS RE
(in any OFDM symbol) is allocated 0 mW or 50 mW.

referenceSignalPower is set to
10*log10(100 mW)= 20 dBm.

Example 2:
PA =4 and PB =1 (with NO power boosting for RS)

PA =4
implies
ρA = 1 or 0 dB.

Hence,
(Power on PDSCH RE/power on RS RE)= 1.

Since the transmit power allocated to PDSCH RE is the nominal power level of 50
mW, the transmit power allocated to RS RE is
(Power on PDSCH RE/1)= (50 mW/1)= 50 mW.

Hence, out of 600 REs in an OFDM symbol carrying the RS, 100 REs are subject to
RS power level, 100 REs have no transmit power, and remaining 400 REs have
nominal power levels.

The total transmit power during the RS-carrying OFDM symbol would be
(100 subcarriers * 50 mW per subcarrier for non-power-boosted RS REs) + (100 subcarriers*
0 mW for null REs) + (400 subcarriers * 50 mW per subcarrier for non-RS REs)= 25 W.
Hence, when PA =4 and PB =1, each RS RE is allocated 50 mW, while a non-RS RE
(in any OFDM symbol) is allocated 0 mW or 50 mW.

referenceSignalPower will be set to

10*log10(50 mW)= 17 dBm.

References
1. 3GPP, TS 36.331.
2. 3GPP, TS 36.213.

[1] The number of antennas and the chosen antenna techniques influence the exact power levels of RS
and PDSCH. In the examples here, we are assuming that eNodeB has two transmit antennas and the UE
has two receive antennas and that the eNodeB will use 2-antenna transmit diversity or (2x2) single-user
multiple input multiple output (SU-MIMO) techniques. See 36.213 for more details.
• $core_v2_language.FormatString($core_v2_language.GetResource('Blog_PostQuestionAnswerVie
w_CommentsCountFormatString'), $post.CommentCount)
4g, boost, lte, sinal