WCDMA RAN RF Optimization Guide
WCDMA RAN RF Optimization Guide
Kirsi Teräväinen
26.08.2014
1
Confidential
© Nokia Solutions and Networks 2014
NPO
Introduction
Introduction
Tools
RF Verification
RF Optimisation
New Site Integration
Neighbour List verification
• Initial RF optimisation is typically done when network is build and integrated but there is no commercial
traffic
• RF verification and optimisation are also needed when new sites are integrated to the network
• RF Optimization is required in the areas having poor or weak dominance
• Physical change of antenna tilt, azimuth, type and height based on scanner data
• Neighbour list (scrambling code) verification is done based on scanner data
• Including GSM/EDGE neighbours
• Make Call Performance KPI Verification with logging tool
• AMR Call setup success rate (CSSR)
• AMR Call drop rate (DCR)
• Call Setup Time (CST)
Introduction
Tools
RF Verification
RF Optimisation
New Site Integration
Neighbour List verification
• For RF optimisation and neighbour verification both Scanner (3G & 2G) and UE measurements are required
simultaneously
• Post-Processing tool is required for data analysis
• Individual call failures or drops can be analysed with drive test tools (e.g. Nemo Outdoor) but to get bigger
picture, a proper analysis tool is required
• Actix or Nemo Analyzer can be used for:
- Data analysis
- Create Maps
- Create KPI reports
• NetAct Planner can be used to verify the proposed coverage
optimisation solution e.g. Antenna tilting or Panning
Introduction
Tools
RF Verification
RF Optimisation
New Site Integration
Neighbour List verification
Missing Site –
13 20/06/2024 © Nokia Solutions and Networks 2014
No dominance
Confidential server in the area
Coverage verification (1/3)
• What impact will any planned sites (if any) have on Planned Site
CPICH_SCAN_RSCP_for SC X overlay plotted out for all three sectors of a site showing the coverage for each
sector
• Gives a good indication if a sector is radiating/overshooting too far and may need down tilt
Scan RSCP for SC 282 Scan RSCP for SC 280 Scan RSCP for SC 281
280
281
282
280 281
Plotting RSCP of each SC highlights any hardware or databuild issues where the coverage area of the cell is
different to the planned coverage
• This will include swapped feeders, incorrect azimuths/tilts or incorrect scrambling codes assigned to the cell
Sector B - SC161 Sector C - SC162
No measurements
No measurements
SC161 and SC162 in vicinity of sector C
Tx from same sector (SC = 162)
Crossed feeders can occur at:
• Antenna
• Input to BTS
17 20/06/2024 © Nokia Solutions and Networks 2014
Confidential • Within the BTS
RSCP verification
If coverage is acceptable then check for area Poor coverage areas can still
have good Ec/No if there is
of poor EcNo in the area dominating server
SC 344 displayed
over pilot pollution
This is clearly
area suffering
from pilot RSCP is good -91 dBm,
pollution EcNo is bad -10 dB
23 20/06/2024 © Nokia Solutions and Networks 2014
Confidential
Pilot Pollution (4/4)
Example
SC 272
Introduction
Tools
RF Verification
RF Optimisation
New Site Integration
Neighbour List verification
• Compare scanner measurements with planned quality and coverage thresholds agreed with
Customer
• The initial planned link budgets should be referred to in order to measure how well the coverage
and quality criteria is met
• When adapting link budgets to show measured value on the ground then its important to
remember that Slow fade margin should be removed as this is implicit in the actual measurement
12.61% 14.51%
for problem area -20
Ec/No (dB)
-10
-5
71.55% 1.34%
0
-92
5
-30 -40 -50 -60 -70 -80 -90 -100 -110 -120 -130
RSCP (dBm)
Urban DL Suburb DL Rural DL Urban DL Qual Sub DL Qual > Rural DL Qual <= 3 SHO
Coverage Coverage Coverage > -89.5 dBm 95.8 dBm > -100.8dBm servers >1
in 7dB
RSCP >= - RSCP >= -95.8 RSCP >= - Ec/Io >= -9dB Ec/Io >= -9dB Ec/Io >= -9dB
89.5dBm dBm 100.8dBm
95.7 99.0 99.8 73.3 71.3 70.7 88.3 66.7
• Check if the cause of call setup failure is CPICH RSCP and Ec/No coverage problem
• The received best server’s CPICH RSCP and Ec/No will be compared to the coverage thresholds
at the location where call setup failed and if best server’s CPICH RSCP OR Ec/No is less than the
thresholds, coverage optimisation will be performed
• The thresholds of CPICH RSCP and Ec/No depend on UE’s sensitivity:
- CPICH RSCP coverage threshold = -110 dBm
- CPICH Ec/No coverage threshold = -16 dB
• Simulation using NetAct Planner will be used to verify the proposed coverage optimisation
solution e.g. Antenna tilting or Panning
8.33% 8.33%
8.33%
UE Issue
• Categorise the failure causes for Call Setup and drop call Unknown
performance
- Drop Call analysis shows that significant proportion of 75.00%
UE Issue
• Mechanical
• The downtilt angle varies over the horizontal beamwidth
– Patterns measured ±90° from the centre of the beam have
decreasing tilt angle until there is no tilt 90° from the main
beam
• The horizontal half-power beamwidth increases with greater
downtilt angle
• The resulting gain reduction depends on azimuth direction
• Electrical
• There is uniform downtilt over the whole azimuth range
• The horizontal half-power beamwidth is independent of the
downtilt angle
• There is identical gain reduction in all azimuth directions
Mechanical
Down tilt kit
Deformation of the
horizontal pattern
Manual Remote
use use
Horizontal pattern
remains constant
40 20/06/2024 © Nokia Solutions and Networks 2014
Confidential
Antenna tilt example (1/3)
Uptilt from -6° to -4°
Before After
Impact
on Ec/No
41 20/06/2024 © Nokia Solutions and Networks 2014
Confidential
Antenna tilt example (2/3)
-1° EDT – Overshooting Cell
No dominant Server
Before After
in the area due the
missing site JS9218
Site JS9125 (SC28) from cluster 5 is Site JS9125 (SC28) was tilted by 1 deg
42 20/06/2024 © Nokia Solutions overshooting
and Networks 2014 to cluster 1 (E-tilt From 4 deg to 5 deg)
Confidential
Antenna tilt example (3/3)
-2° EDT
After antenna tilting, new drive-test should be performed in order to verify changes
Negative
changes
Before After
Introduction
Tools
RF Verification
RF Optimisation
New Site Integration
Neighbour List verification
• New sites should be integrated without degrading the performance of existing sites
• It is possible that after driving the new and surrounding site, both will require RF optimisation (tilting,
panning)
Introduction
Tools
RF Verification
RF Optimisation
New Site Integration
Neighbour List verification
• Each neighbour has a set of associated parameters e.g. CPICH measurement offset
• The post processing tool should be able to suggest appropriate neighbour lists
• Strategy for initial system deployment is to place the emphasis upon adding neighbours rather than removing them
Scrambling codes of
neighbours
and information
about thresholds
2G neighbours
and information like BSIC
50 20/06/2024 © Nokia Solutions and Networks 2014 (ncc, bcc) and BCCH
Confidential
Neighbour list verification type of approaches
There are a number of approaches that can be used to both plan and verify the neighbour plan
Neighbour Neighbour
Creation Verification
Neighbour
Manual Check Analytical
Creation
Planning Tool
Other
Manual Check
Drive Testing
Measured
Network Stats
• Analytical Neighbour Planning methods are traditionally used to generate original neighbour lists but they
can also be used for verification and new site addition
• Planning Tools – use coverage arrays and handover parameter data to determine required neighbour relations
Strongest Neighbours
Ec/Io
Ec/Io reported when High Level Process
within this range
Record drive test results with
Start
Scanner in ‘TOP N’ mode
Neighbour
Window
Analysis (within Actix Finish
Analyzer)
Time
SC Site Sector Num Of Samples Suggested NBR Additions
216 Ewer Park Stud 00161126C 97 85 (47.4%); 303 (17.5%)
349 Reading Rubgy FC 00179719B2 97 351 (28.9%)
55 20/06/2024 WOODLEY
83 and Networks
© Nokia Solutions 2014 ATE 00001085A 96 350 (34.4%); 84 (25.0%)
Confidential 436 TILEHURST ATE 00001081B 90 434 (22.2%); 283 (22.2%)
Numerical Analysis
Neighbour list verification in Actix (2/3)
Start
• Analysis provided by the post processing tool:
Run
Run Actix
Actix Analyzer
– Consider adding neighbours reported by the Analyzer
Routine
Routine
tool
– If neighbour list is full then consider Is
No Is the
the
replacing some of the existing neighbours Neighbour
Neighbour
List
List full?
full?
Yes
Consider
Consider adding
adding
• Do not remove existing neighbours without suggested
suggested
neighbours
neighbours
further investigation Consider
Consider
replacing
replacing existing
existing
neighbours
neighbours byby
those
those suggested
suggested
by
by the tool
the tool
Progress to
further additions
56 20/06/2024 © Nokia Solutions and Networks 2014
Confidential
Neighbour list verification in Actix (3/3)
Example for Reading Central (SC 280)
% Scrambling Code Samples
30.00
% of Samples
– List 20.00
33.90
15.00
18.50
10.00
13.80
9.40
5.00
3.90
3.10
2.00
1.20
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
422 423 420 283 282 286 348 177
Neighbour Scram bling Code
Does
Does list
list include
include
all
all cells
cells belonging
belonging
No
Cell for which to
to that
that site
site and
and
the neighbour the first perimeter
the first perimeter
list is being of
of cells
cells
defined
Yes
Add
Add any
any other
other
Neighbours neighbours
neighbours which
which
may
may improve
improve
coverage
coverage
Finish
58 20/06/2024 © Nokia Solutions and Networks 2014
Confidential
Summary
Summary
• Dominance, Coverage, are all issues for RF optimisation
• There are a range of tools that can be used to identify network problems
• Automated optimisation tools have been developed to help with RF optimisation
• Neighbour management is key for good RF performance