UMTS High-Power Capacity

Improvement Solution
Acceptance Criteria
UMTS Solution Design Dept.
www.huawei.com

Page 1
Author/ID

Co-author/ID

Approver/ID

This presentation describes the acceptance criteria for the

UMTS high-power capacity improvement solution in terms of:
Description  Coverage
 Capacity
 KPI

Page 2
 Introduction to the UMTS High-Power Capacity
Improvement Solution
Solution Positioning Processes & Key Actions
This solution is designed to address power
congestion on operator mobile networks and
Project
support the sales of existing network solutions.
kick off
By showing the benefits of this solution to influence Site survey
Scenario
operator decisions, this solution aims to turn the (hardware
discovery planning included)
configuration of using the maximum transmit power
of 40 W into a mainstream on the "infrastructure 1 2
networks".
E2E
Capacity Network high- Network
6 3
Heavy load scenario evaluation power planning
40 W/1.5 W solution
40 W/2 W 5 4
Delivery
acceptance Network Site
optimization engineering
Weak coverage 40 W/4 W Strong coverage

Scenario-specific pilot
40 W/4.5 W
power configuration This solution is an RF solution that helps increase network capacity. It
Coverage scenarios Light load
focuses on network coverage, capacity improvement, and
impacts on KPIs.

Page 3
 Acceptance Principles
For operators that do not have high expectations on solution outcomes, use the
strategy of influencing them to focus on positive outcomes. For operators that have
high expectations on solution outcomes, follow the action requirements throughout
the project in terms of application scope, bidding, and delivery.

The solution benefits and Customer strategy

impact may vary with
scenarios. It is good practice Bidding
to perform the pre-evaluation phase Principles Applicable
scope
The planning is based on
Huawei's solutions. Otherwise,
to obtain the performance need adjustments to the
expectation before bidding. Delivery phase acceptance criteria.

Evaluation results are subject to solution selection (for example, different power ratios), traffic distribution
and model, user behavior, and holiday effect. It is vital to tell operators that some sites or cells may fail to
reach the expectations. For operators who have high expectations on the solution outcomes, it is crucial to
analyze the network conditions and explain to them why the outcome is lower than the expected.

Page 4
DT-based Network Evaluation
Instructions to DT:
1. DT route: same before and after the reconstruction.
2. DT time: consistent between before and after the reconstruction.
3. UE position: same positions and orientations before and after the reconstruction.
4. Vehicle: same before and after the reconstruction.

Value Data
Dimension DT KPIs
Type Source
Coverage RSCP, Ec/Io Average Voice
Note: Adjustment to test cases in light
of field and operator demands is
CS setup success rate
CS call drop rate allowed, as long as the above
KPI
HSUPA/HSUPA single-user
Average
Voice and instructions are followed.
throughput PS
PS setup success rate
PS call drop rate

Acceptance criteria: identical with the original network, and the cluster level is not lower than that before reconstruction.

Page 5
Capacity and KPI Impact Evaluation
The capacity benefits a lot from the
solution while taking the impact on
Dimension Data Source Involved counters Value Type Period
KPIs into consideration, therefore,
the evaluation is mainly performed A week before and
PM RSCP, Ec/Io PDF or CDF after the
on the capacity and impact on reconstruction

KPIs. Coverage
RSCP, Ec/Io, and
Location-specific
Before and after
and average
MR geographic coverage the same three
mean value
visibility workdays
Coverage Coverage statistics
RSCP Ec/Io
A week before and
Scatter chart and
Capacity PM TCP, DL traffic after the
fitting curve
reconstruction
Air interface A week before and
Access, call drops, and
capability KPI PM Contrast chart after the
handover success rate
reconstruction
Capacity Capacity
TCP/RTWP Traffic/Throug
hput
MR-based coverage evaluation is conducted for versions earlier than RAN16.0 and
PM data-based coverage evaluation for RAN16.0 or later.
KPI Capacity
User awareness

Page 6
Acceptance Criteria Based on Capacity Improvement
(40 W/2 W)
Cluster-level
RSCP Same or higher
Coverage
Ec/Io 1.5 dB smaller

Capacity increase Increase by about 20%

Capacity Traffic
Traffic increased or load reduced
Load
Access, call drops, Stable (the fluctuation before and after is within the same range).
KPI and inter-RAT Note: during the same period, numerators and denominators
handover accumulated individually.

Note: 1. Compare the data one week or more before and after the reconstruction.
2. For the final impact after using the VAM, refer to the VAM criteria.
3. The criteria is prepared based on Huawei recommended methods.

Page 7
Contents

• Coverage Acceptance

• Capacity Acceptance

• KPI Acceptance

Page 8
 Coverage Assessment Methods
PM data
RSCP PDF RSCP CDF
Platform: BAM or SAU 100% 3-sector
20W2W 100%
Multi-sector
3-sector
20W2W
Period: one week before and after the 40W2W
Data Multi-sector
40W2W
reconstruction 50% 50%
Data: including the following counters collection
CDF and PDF 0% 0%
charting in Excel [-60 -70) [-70 -80) [-80 -95) [-95 -105) -110 -100 -90 -80 -70 -60
Data RSCP (dBm) RSCP (dBm)
analysis
Ec/Io PDF Ec/Io CDF
60% 100%
50% 3-sector
20W2W 20W2W
3-sector
40% Multi-sector
40W2W
30% Multi-sector
40W2W 50%
RSCP, Ec/Io 20%
10%
0% 0%
[-8 -10) [-10 -14) [-14 -18) [-18 -25) -25-23-21-19-17-15-13-11 -9 -7 -5
Average Ec/Io (dB) Ec/Io (dB)
value
Average Value RSCP (dBm) Ec/Io (dB)

20 W/2 W -71.36 -9.94

PM data
40 W/2 W -71.36 -10.9

△RSCP (dB) △Ec/Io (dB)

Average gains 40 W/2 W vs. 20 W/2 W
0 1.04

Perform the PM data-based acceptance for RAN16.0 or later and MR-based acceptance for versions earlier than RAN16.0. The same method applied to both acceptance
types.

Page 9
Acceptance for Coverage RSCP
RSCP PDF
35.00%
20W/2W
30.00%
40W/2W
25.00%
20 W/2 W/cell 20.00%
15.00%
10.00%
5.00%
Same 0.00%

coverage
40 W/2 W/cell
RSCP (dBm)

Before and After Before After Gain

In the scenario with the TCP of 40 W and pilot
power of 2 W, the same areas are covered Average RSCP (dBm) -78.39 -78.09 0.3
because the pilot power stays the same.
Comparing with that before the reconstruction, The results from the Thailand AIS project show an
RSCP remains unchanged. average RSCP increase of 0.3 dB after the
reconstruction from 20 W/2 W to 40 W/2 W.
Acceptance criteria: the cluster-level RSCP cannot be lower than that before the reconstruction.

Page 10
 Acceptance for Coverage Ec/Io
Ec/Io PDF
0.3
20 W/2 W/cell Capacity 20W/2W
0.25
improvement 40W/2W
scenario 0.2

0.15
40 W/2 W/cell
0.1

In the scenario with the TCP of 40 W and pilot power of 2 0.05

W, because there are more available power resources, the 0
suppressed traffic is released and the load is changed, the >-6 (-8,-6] (-10,-8] (-12,-10] (-14,-12] (-16,-14] (-18,-16] <=-16
Ec/Io becomes worse. Ec/Io (dB)

Ec/Io CDF
Before and After the
1 Before After Gain
20W/2W Reconstruction
40W/2W 0.8
0.6 Average Ec/Io (dB) -8.38 -9.30 -0.92
0.4
0.2
The results from the Thailand AIS project show an
0
-20 -15 -10 -5 0
average Ec/lo increase of about -0.92 dB from 20 W/2
Ec/Io (dB) W to 40 W/2 W.
Acceptance criteria: as suppressed traffic is released in varying degrees, the average cluster-level Ec/Io reduces by at most 1.5 dB in the 40
W/2 W scenario.
Page 11
Contents

• Coverage Acceptance

• Capacity Acceptance

• KPI Acceptance

Page 12
Capacity Assessment Methods
Based on PM data
Step 1. Obtain the PM data, including the following counters: VS.MeanTCP,
VS.HSDPA.MeanChThroughput.TotalBytes, and TxMaxPower.
Step 2. Calculate the TCP
TCP = VS.MeanTCP (mw)/MaxTxpower (mw) x 100%
Sort out the TCP and traffic in the ascending order of TCP.
Step3. Draw a chart with the TCP as the horizontal coordinate and the
VS.HSDPA.MeanChThroughput.TotalBytes as the vertical coordinate, then
select the curve fitting in the excel file to fit the two curves before and after the
28%
reconstruction.
Step 4. Calculate the average TCP and traffic before and after the
reconstruction, then fitted site-level traffic. The emendation factor is obtained
using the following formula: fitted site-level traffic/actual traffic.
Step 5. From the curves, obtain the fitted traffic under a load
ratio of 60% before and after the construction. Then,
calculate the actual traffic using the emendation factor with 40 W OFF 40 W ON
the following formula: Actual traffic = fitted traffic/emendation DL Load_actual 35.27% 24.01%
factor. In this way, capacity gains can be calculated after the
reconstruction. Traffic_actual 121.1 131.7

Note: Traffic_fitting 135.84 155.15

In the dual-band networks (U900 and U2100), capacity assessment is performed in
Emendation Factor 1.12 1.18
the following way: if a single carrier meets the conditions of traffic suppression
(downlink load is 60%), the traffic in the system is also suppressed, then capacity PredictiveTraffic_fitting@60% DL Load 249.75 335.55
gains are calculated under this load.
PredictiveTraffic_actual@60% DL Load 222.64 285
If the traffic is firstly suppressed in the U2100 network, its downlink load is 60%, the
downlink load in the U900 network is 54%, traffic gains can be calculated following
the above methods. Capacity Gain@60% DL Load 28%

Page 13
 Capacity Acceptance Criteria
Capacity is a key factor to measure network capability. It is defined as the maximum
throughput allowed by a network with the basic service experience ensured (KPI and
user throughput).
Cell Average Throughput VS User num
No reliable throughput can be used to
Cell throughput
2500 gain >20%
reflect the KPI. Empirically, if the
2000 throughput of 20% cell edge users is
Throughput (kbps)

1500 equal to or greater than 1 Mbit/s, the

1000
network DL load reaches to 70% (50%
20w2w HSDPA available power); when the
500
40w2w network load ratio reaches 70%, user
0
experience starts to fall short of
0 10 20 30 40 50 60
expectation. It is recommended that
Cell Edge Throughput VS User num
capacity gains are measured using
the total traffic volume under the The reconstruction is performed on a site when traffic demand unfulfillment starts based on
4000 differential expansion methods. At this point, traffic demand can still be met. After the
20w2w
Throughput (kbps)

3000 same load before and after the reconstruction, the traffic stays the same, the load reduces and user experience becomes
40w2w reconstruction. better. As the number of users increases on the site, traffic increase, load reduction, and
2000
better user experience are achieved. When the user number grows to the extent that traffic
1000
demand is badly unsatisfied, only the traffic increases.
0
0 10 20 30 40 50 60
User Number
Results form five trial sites in
The results from the Thailand project in the Thailand: perform the
trail areas (dense urban areas) show a reconstruction when traffic
demand unfulfillment starts,
capacity gain of about 28% at the load ratio
the traffic increases or
of 60%. load reduces at the site.
Criteria:
 40 W/2 W: dense urban: over 20%

Cluster-level capacity gains: accepted by the 20% of criteria. Acceptance criteria for the traffic and load: traffic increased or load reduced.

Page 14
Contents

• Coverage Acceptance

• Capacity Acceptance

• KPI Acceptance

Page 15
 KPIs to Be Checked
Collect PM data at least one week before and after the reconstruction, and use OMStar,
FMA, PRS, or other tools to analyze them.
Table 1 KPI comparison between 20 W/2 W and 40 W/2 W Note: KPIs for the same period are compared .
KPI Category KPI Name ID Name Before After

CS Radio Access Success Ratio 137511 VS.CS.Access.Success.Cell.Rate

Access
PS Radio Access Success Ratio 137509 VS.PS.Access.Success.Cell.Rate

AMR Call Drop Ratio 128483 VS.CS.AMR.Call.Drop.Cell.Rate

Call Drop
PS Call Drop Ratio 128482 VS.PS.Call.Drop.Cell.Rate

CS W2G Inter-RAT Handover Out Success Ratio

Inter-RAT
Handover
PS W2G Inter-RAT Handover Out Success Ratio

Note: the above KPIs are Huawei-defined KPIs instead of PM counters.

Perform the acceptance on the above KPIs according to the following criteria: cluster-level KPIs, site-level KPIs, or cell-level
KPIs are same as or greater than those before reconstruction.

Page 16
 U
Thank THANK YOU
www.huawei.com MTS
Copyright © 2016 Huawei Technologies Co., Ltd. All Rights Reserved.
The information contained in this document is for reference purpose only, and is subject to change or withdrawal according to specific
customer requirements and conditions.

Page 17
Appendix 1 The Counters Associated with the TCP of 40 W Scenario
Items Counters Counter ID Counter Name
73428032 VS.Periodic.MrRpt.RSCP.0
73428033 VS.Periodic.MrRpt.RSCP.1
73428034 VS.Periodic.MrRpt.RSCP.2
73428035 VS.Periodic.MrRpt.RSCP.3
73428036 VS.Periodic.MrRpt.RSCP.4
RSCP
73428037 VS.Periodic.MrRpt.RSCP.5
73428038 VS.Periodic.MrRpt.RSCP.6
73428039 VS.Periodic.MrRpt.RSCP.7
73428040 VS.Periodic.MrRpt.RSCP.8
73428041 VS.Periodic.MrRpt.RSCP.9
Coverage
73428069 VS.Periodic.MrRpt.EcNo.0
73428070 VS.Periodic.MrRpt.EcNo.1
73428071 VS.Periodic.MrRpt.EcNo.2
73428072 VS.Periodic.MrRpt.EcNo.3
73428073 VS.Periodic.MrRpt.EcNo.4
Ec/Io
73428074 VS.Periodic.MrRpt.EcNo.5
73428075 VS.Periodic.MrRpt.EcNo.6
73428076 VS.Periodic.MrRpt.EcNo.7
73428077 VS.Periodic.MrRpt.EcNo.8
73428078 VS.Periodic.MrRpt.EcNo.9
Mean TCP 67199618 VS.MeanTCP
HSDPA average user Num 67202932 VS.HSDPA.UE.Mean.Cell
Capacity
HSDPA MAC-d flow Total Bytes 67189840 VS.HSDPA.MeanChThroughput.TotalBytes
DCH user Num 73410510 VS.CellDCHUEs
CS Radio Access Success Ratio 137511 VS.CS.Access.Success.Cell.Rate
PS Radio Access Success Ratio 137509 VS.PS.Access.Success.Cell.Rate
AMR Call Drop Ratio 128483 VS.CS.AMR.Call.Drop.Cell.Rate
KPI
PS Service Drop Ratio 128482 VS.PS.Call.Drop.Cell.Rate
CS W2G Inter-RAT Handover Out Success Ratio
PS W2G Inter-RAT Handover Out Success Ratio

Page 18
Appendix 2 Analysis of Negative Impact from 20 W/2 W to 40 W/2
W
Load and Ec/Io Varies with the Number of Users in the 20
W/2 W and 40 W/2 W Scenarios
40 W/2 W/cell 100% 89.00% 90.00% 90.00% 90.00% 10 In the 40 W/2 W scenario, after
79.00%
80% 88.00%90.00% 90.00% 90.00% 6
8 performing RF optimization and
57.00%
60% 65.00% 20w/2w TCP
4
2
parameter optimization, the Ec/Io

Ec/Io (dB)
34.00% 0
40w/2w TCP reduces by at most 1.5 dB, and

TCP (%)
40% 37.00% -2
20w/2w EcIo -4
20% 20.00% 40w/2w EcIo -6 KPIs remain unchanged.
0% -8
-10
0 10 20 30 40 50 60 -12
-20%
-14
-40% -16
20 W/2 W/cell
User Number

After the reconstruction from 20 W/2 W to 40 W/2 W, Ec/Io deterioration varies with
the traffic release and downlink load.
•Under light load, the traffic keeps unchanged, the same maximum transmit power is
used, the Ec/Io stays the same.
•Under heavy load, the suppressed load is released, the capacity and maximum
transmit power increase, causing greater interference in this cell and neighboring
cells. In extreme circumstances, the transmit power is used up. The cell is fully
loaded, the Ec/Io reduces by 3 dB in the worse case.

Acceptance criteria: the average cluster-level Ec/Io reduces by 1.5 dB in the worst case.

Page 19