comparison of 40w & 20w pa
TRANSCRIPT
-
8/4/2019 Comparison of 40W & 20W PA
1/25
-
8/4/2019 Comparison of 40W & 20W PA
2/25
2 QUALCOMM CONFIDENTIAL
Introduction
Service providers frequently come across with suggestions of using
higher power amplifier (40W instead of 20W PA) for Node B to avail of
capacity advantages.
This document discusses comparative advantages & disadvantages of
using higher Node B power (40W vis-a-vis 20W PA) based on both
theoretical (link budget) analysis and case study done with Merida for
following scenarios.
a) Interference Limited Scenario (Multi-BTS environment; 28 sites)
b) Noise Limited (Coverage Limited) Scenario (Outskirts area)
Impact of 40Watt PA on HSDPA throughput for uneven traffic
distribution has also been discussed separately.
The details of all the above are provided in the following slides.
-
8/4/2019 Comparison of 40W & 20W PA
3/25
3 QUALCOMM CONFIDENTIAL
Interference LimitedScenario
Analysis Based on Link Budget(Uniform Traffic Distribution)
-
8/4/2019 Comparison of 40W & 20W PA
4/25
4 QUALCOMM CONFIDENTIAL
Interference Limited
Scenario
Analysis Based on Link Budget (Uniform Traffic Distribution)
Analyses were carried out to ascertain max. number of CS64 users (at cell
edge) with uniform forward link loading of 70% for i) 20 Watt & ii) 40 Watt PA.
It is calculated from link budget that in order to meet same pilot and CS64service coverage targets, all channel powers need to be increased by 3 dB.
As a result, total number of users that can be served does not increase with
40 Watt PA.
Based on link budget, it can therefore be concluded that if traffic is uniformlydistributed, usage of 40 Watt PA will not provide any capacity advantage.
Summary of results along with link budgets are provided in following slide.
-
8/4/2019 Comparison of 40W & 20W PA
5/25
5 QUALCOMM CONFIDENTIAL
Interference Limited
Scenario
Analysis Based on Link Budget (Uniform Traffic Distribution)
DownlinkBudget- W- W22 22
PA TypeRequired Pilot
Power (dBm)
Required Traffic
(CS ) Channel22
Power (dBm)
Max. No. of CS Users22
at Cell Edge (with %22
PA Loading)
Watt22 22 22 .444
Watt22 22 22 .222
-
8/4/2019 Comparison of 40W & 20W PA
6/25
6 QUALCOMM CONFIDENTIAL
Interference LimitedScenario
Analysis Based On StaticSimulation in Asset
(Uniform Traffic Distribution)
-
8/4/2019 Comparison of 40W & 20W PA
7/257 QUALCOMM CONFIDENTIAL
Interference Limited Scenario
Analysis Based on Static Simulation in Asset (Uniform FL Loading)
Static coverage analyses were carried out for a cluster of 28 sites of Merida for
following scenarios to determine forward link pilot coverage improvement.
a) Static pilot coverage with 70% FL load for 20 Watt PA
b) Static pilot coverage with 70% FL load for 40 Watt PA (all common channel
powers increased by 3 dB as per link budget)
c) Static pilot coverage with 70% FL load for 40 Watt PA (common channel
power remain same as 20 Watt case)
From the statistical report, it is established that no improvement in Ec/Io isobserved in (b) with respect to (a). Also, (c) is found to be very poor in terms
of Ec/Io.
Summary of results along with detailed simulation statistics are provided in
following slide.
-
8/4/2019 Comparison of 40W & 20W PA
8/258 QUALCOMM CONFIDENTIAL
Interference Limited Scenario
Analysis Based on Static Simulation in Asset (Uniform FL Loading)
Pilot RSCP >
- dBm222
Pilot Ec/Io >
- dB44
Watt22 Static Simulation with % FL Load22 . %2222 . %4444
Static Simulation with % FL Load22
(Control Channel power increased by d2
as per link budget)
. %2222 . %2222
Static Simulation with % FL Load22
(without increasing Control Channel
power by dB)2
. %2222 . %2222
Percentage of Area with
PA Type Scenario
Watt22
Static Comparison
W W22 22
-
8/4/2019 Comparison of 40W & 20W PA
9/259 QUALCOMM CONFIDENTIAL
Interference Limited Scenario
Analysis Based on Static Simulation in Asset (Uniform FL Loading)
Conclusion
Considering the fact that forward link pilot coverage is determined by both
RSCP and Ec/Io, maximum coverage area percentage (98.95%) does notincrease by using 40W PA.
Usage of 40W PA shall necessarily have to be associated with increase of
pilot and other control channel powers by 3dB. Otherwise, significant
deterioration in Ec/Io coverage will be experienced. Usage of 40W PA however, increases pilot RSCP strength. In case of forward
link pilot RSCP limited scenario, this might help in improving coverage.
-
8/4/2019 Comparison of 40W & 20W PA
10/2510 QUALCOMM CONFIDENTIAL
Interference LimitedScenario
Analysis Based On MCSimulation in Asset
(Uniform Subs Distribution)
-
8/4/2019 Comparison of 40W & 20W PA
11/2511 QUALCOMM CONFIDENTIAL
Interference Limited Scenario
Analysis Based on MC Simulation in Asset (Uniform Subs Distribution)
MC simulations were carried out for the same cluster of Merida under
following scenarios to determine improvement in forward link blocking.
a) MC simulation with 5,000 voice users for 20 Watt PA
b) MC simulation with 5,000 voice users for 40 Watt PA (all common channelpowers increased by 3 dB as per link budget)
c) MC simulation with 5,000 voice users for 40 Watt PA (common channel
power remain same as 20 Watt case)
d) MC simulation with 5,000 voice users with 40Watt PA at FL limited sectors
only (other sectors with 20Watt PA)
The voice users have been allowed to be distributed uniformly as per
simulation process. No clutter weight has been used for this study.
Summary of results along with detailed simulation statistics are provided in
following slide.
-
8/4/2019 Comparison of 40W & 20W PA
12/2512 QUALCOMM CONFIDENTIAL
Interference Limited Scenario
Analysis Based on MC Simulation in Asset (Uniform Subs Distribution)
MC Simulation
Comparison
Low Pilot Low Eb/No
Watt22MC Simulation with ,22 2 2
voice users. %444 . %222 . %222 . %222
MC Simulation with ,22 2 2
voice users (Control Channel
power increased by dB as2
per link budget)
. %222 . %444 . %222 . %444
MC Simulation with ,22 2 2
voice users (without
increasing Control Channel
power by dB)2
. %222 . %222 . %222 . %222
Watt &22
Watt22
MC Simulation with ,22 2 2
voice users ( Watt PA used2 2
for FL limited sectors)
. %222 . %222 . %222 . %222
Percentage
of Failures
(due to RL)
Total
Percentage
of Failure
ScenarioPA Type
Watt22
Percentage of Failures
(due to FL)
-
8/4/2019 Comparison of 40W & 20W PA
13/2513 QUALCOMM CONFIDENTIAL
Interference Limited Scenario
Analysis Based on MC Simulation in Asset (Uniform Subs Distribution)
Conclusion
From the previous analyses, it is evident that increasing PA power to 40W
does not result in any improvement of total failure percentage.
For 20W scenario, 1.8% failure happened due to forward link limitations (low
pilot and low Eb/No).
Although above were almost resolved by using 40W PA, this however could
not result in any overall improvement as additional failures happened due to
reverse link limitation.
It can therefore be concluded that improvement of only forward link does not
guarantee any betterment of performance.
-
8/4/2019 Comparison of 40W & 20W PA
14/2514 QUALCOMM CONFIDENTIAL
Coverage Limited Scenario
Analysis Based On StaticSimulation in Asset
(Unloaded Scenario)
-
8/4/2019 Comparison of 40W & 20W PA
15/25
15 QUALCOMM CONFIDENTIAL
Coverage Limited Scenario
Analysis Based on Static Simulation in Asset (Unloaded Scenario)
Unloaded static coverage analyses were carried out in following scenarios for
an outskirts site of Merida to ascertain coverage improvement (if any) by
using 40Watt PA.
a) 20Watt PA with control channel powers as per link budget
b) 20Watt PA with control channel powers increased by 3dB from (a) above
c) 40Watt PA with control channel powers increased by 3 dB from (a) above
d) 40Watt PA with control channel powers increased by 6 dB from (a) above
Summary of results along with link budgets are provided in following slide.
-
8/4/2019 Comparison of 40W & 20W PA
16/25
16 QUALCOMM CONFIDENTIAL
Coverage Limited Scenario
Analysis Based on Static Simulation in Asset (Unloaded Scenario)
Pilot RSCP >
- dBm222
Pilot Ec/Io >
- dB2
Unloaded Static Simulation(CPICH= dBm)22
.444 .222
Unloaded Static Simulation
(CPICH= dBm)44.444 .2222
Unloaded Static Simulation
(CPICH= dBm)22.222 .2222
Unloaded Static Simulation
(CPICH= dBm)22.444 .2222
PA Type Scenario
Forward Link Coverage
Area (Sq. Km.) for
Watt22
Watt22
Isolated Coverage
Comparison
-
8/4/2019 Comparison of 40W & 20W PA
17/25
17 QUALCOMM CONFIDENTIAL
Coverage Limited Scenario
Analysis Based on Static Simulation in Asset (Unloaded Scenario)
Conclusion
From the previous analyses, it is evident that for same CPICH (and other
control channel) powers, pilot coverage boundary remains unchanged for
both 20W and 40W PA.
Forward link coverage continues to increase with the increase in control
channel powers.
Accordingly, for 37dBm of pilot power setting, maximum coverage is obtained
for 40Watt PA.
It may however be noted that above does not guarantee any actual coverage
improvement. Actual coverage improvement will be obtained if reverse link is
found to be strong enough to meet larger pilot coverage footprint.
-
8/4/2019 Comparison of 40W & 20W PA
18/25
18 QUALCOMM CONFIDENTIAL
Impact OnHSDPA
Qualcomm Simulation(Uneven Traffic Distribution)
-
8/4/2019 Comparison of 40W & 20W PA
19/25
19 QUALCOMM CONFIDENTIAL
Impact On HSDPA
Qualcomm Simulation (Uneven Traffic Distribution)
In order to investigate whether 40Watt PA would result in significant gain in
terms of single user application layer throughput, simulations were carried out
by Qualcomm for Cat. 8 and Cat. 12 UE.
The CS load (voice users) of neighboring cells were varied between 0% to
100% while the HSDPA serving cell was always loaded to 100% with HSDPA
traffic (20% of the power was allocated to common channels of the cell).
The following slide shows the derived average results for the single-user
application (eg. FTP) layer throughput.
-
8/4/2019 Comparison of 40W & 20W PA
20/25
20 QUALCOMM CONFIDENTIAL
Impact On HSDPA
Qualcomm Simulation (Uneven Traffic Distribution)
UE TypeNeighbor Cell
Load
Average Application
Throughput Gain from
W to W22 22
%2 . %22
%22 . %22
%222 . %22
%2 . %444
%22 . %222
%222 . %22
Cat. 22
Cat. 2
-
8/4/2019 Comparison of 40W & 20W PA
21/25
21 QUALCOMM CONFIDENTIAL
Impact On HSDPA
Qualcomm Simulation (Uneven Traffic Distribution)
Conclusion
Gain of low capability UE (Cat.12) is quite limited (max. 7.4%) as more power
does not help a limited number of codes.
On the other hand, significant gain (33%) is experienced with Cat. 8 UE due to
better utilization of 16QAM modulation (in case of no neighbor cell load).
This gain gradually diminishes with increasing neighbor cell load and virtually
no gain (0.3%) is experienced with 100% neighbor cell load.
Based on above, it can be concluded that 40Watt PA does not provide
capacity gain in an evenly loaded scenario but does provide gains in isolated
or typically indoor cell scenario.
The detailed document on above [80-W0976-1] will be released for AMX.
-
8/4/2019 Comparison of 40W & 20W PA
22/25
22 QUALCOMM CONFIDENTIAL
Conclusi
on
-
8/4/2019 Comparison of 40W & 20W PA
23/25
23 QUALCOMM CONFIDENTIAL
Conclusion
For interference-limited scenario (likely to exist in high & medium traffic areas)
usage of 40Watt PA is unlikely to provide any capacity advantage over 20Watt
PA under near-uniform radio network loading condition.
For isolated coverage scenario, usage of 40Watt PA will enable the user to
increase forward link coverage footprint while reserving adequate power for
traffic channels (compared to 20Watt PA). This however will ensure actual
coverage gain if reverse link is strong enough to match larger forward link
coverage boundary.
For uneven traffic scenario, as observed in HSDPA simulation study, capacity
gain is expected by using 40Watt PA. The gain will however largely depend onUE capability and loading of surrounding cells. As the surrounding cell loading
increases, the gain will diminish.
In view of all the above, it is recommended to opt for case-by-case decision to
select power amplifier type (20Watt or 40Watt).
-
8/4/2019 Comparison of 40W & 20W PA
24/25
24 QUALCOMM CONFIDENTIAL
Conclusion
Considering the vendor-specific cases for AMX, it is understood that for
Ericsson, NodeBs will be equipped with 40Watt PA which can be used as
either 40Watt single carrier PA or two carriers each of 20Watts. Therefore, for
Ericsson regions, it is recommended to start service with 20Watt PA and
reconfigure the site for 40Watt as per requirement.
For Nokia on the other hand, NodeBs are fitted with 40Watt PA but will require
an additional software key to activate the same. Therefore, for Nokia regions,
it is recommended to use 20Watt PA (40Watt PA without software activation)
at the launch of service and activate the software key based on site specific
requirement.
-
8/4/2019 Comparison of 40W & 20W PA
25/25
25 QUALCOMM CONFIDENTIAL
Thank
You