Network Coverage evaluation using MR reports

www.huawei.com

HUAWEI TECHNOLOGIES CO., LTD.

Security Level:

Introduction of Network Coverage Evaluation Based on MR

Page 2

Contents

MR based Geolocation Principle Introduction MR Basic and Location principle Traffic map value

Coverage Evaluation Introduction AGPS MR Collection Coverage Evaluation based on AGPS MR

Case Introduction

www.huawei.com

HUAWEI TECHNOLOGIES CO., LTD.

Security Level:

Page 3

MR Traffic Map Solution Overview

Geographical visualization of RSCP， ECIO， UE Power and Cell Edge based on real user information

Accurate Coverage Visualization

With configurable thresholds and parameters, auto identify spots with poor coverage, pilot pollution and overshooting problems

Coverage Problems Auto Identification

Spot with Weak Coverage

New site address is auto located according to weak coverage spots and users distribution

Smart Site Planning

Service Product Status: Available for R12, R13. Tool: OMStar

Page 4

Measurement Control

Measurement Report

Measurement Report

Periodic Trigger(MR Module)

Event Trigger (HO Module)

MR MAP : According to corresponding locating algorithm, make the information brought by UE (such as RSCP, Ec/Io, TX Power and so on) display on the map.

Traffic Map Introduction - MR Description

HUAWEI TECHNOLOGIES CO., LTD. Page 5华华华华华华华华华华华华华华，

MR Location Calculation Process

Calibrated Engineering Parameter

Feature Database

MR Data

Location Calculation Calculated Location Point

MR Location Calculation Process Nastar

Tim e IM SI RNC Cell Longitude Latitude9/16/2010 13:29:13.600 440001090253516 2305 23405 140.116544 36.123889/16/2010 13:29:23.600 440001090253516 2305 23405 140.117095 36.124334

… … … … … … … … … … … …


Locating Principle： CELLID+RTT、 Feature matching Method、 AGPS

1.CellID + RTTFrom RTT measurement results, calculates the distance from the UE to serving cell, 1 RTT result determines 1 circle, 3 results determine 3 circles, and the cross points of the 3 circles are the real location.

Signal Features

Moving Unit

2. Feature matching MethodUse the methods of DT,AGPS and coverage simulation to establish network feature database, such as RSCP, Path info, Time Delay etc. And then compare MR with the feature database. Ue location is where the bin match best.

3.AGPSRNC collect information related to GPS satellites, and the data is given to the UE needed for positioning, UE search to GPS satellite signals and estimate location.

Traffic Map Introduction - MR Location principle

The SAS is a separate device, which is connected to an RNC through the Iupc interface. The SAS is used for UE positioning


Baseline of MR Location Accuracy

Scenario not applicable to: Single sites only Site on hill /Sea / Lake

Parameters Input / Output

U / SU

Traffic map accuracy*note 1

Output 80 – 120m

Site distance Input 300 – 800mMinimum MR message collection on each position

Input 5000 MR Messages

Site Lat / Long tolerance Input < 15mAntenna azimuth tolerance (Max)

Input < 10 degree

Drive test data support (Cover major road)

Input Yes

Digital map resolution Input 5m (with building information)*Note 1: The output accuracy is subject to the input requirement as specified. Digital map resolution>50m， the impact for accuracy is about

20%~30%. No drive test : for modified feature database, the

impact for accuracy is about 30%. for propagation model ,the impact for accuracy is about 30%.

Dependences


DT/CQT MR Accuracy

Scenario AccuracyCQT 14m @ 67%DT 34m @ 67%

DT Scenario CQT ScenarioCell Quality Test


MR Location Accuracy by Using AGPS to Correct Character DB

Character DB AccuracyBefore Correction 626m @ 67%After Character DB

Correction Using

AGPS Data

153m @ 67%

Using AGPS Data to

Forming Character

DB

87m @ 67%

Specification: Before Correction: Using simulation method by propagation model to make the character DB WITHOUT any data correction.

After Character DB Correction Using AGPS Data: Based on the prediction character DB, Using the AGPS data to correct the prediction character DB.

Using AGPS Data to Forming Character DB: Directly make the character DB by AGPS data WIHOUT prediction.


Contents



Case Introduction


Traffic Map Analysis Value1.Real user RSCP coverage

2.No drive test RequiredRSCP Distribution

1.Real user Ec/Io coverage2.No drive test Required

Ec/Io Distribution

1.Real user serving cell distribution2.More precise than simulation result

Serving Cell Edge Distribution

1.Overshooting Analysis2.Handover Analysis3.Interference Analysis4.Traffic Unbalance Analysis

Remark: Max MR records cell as serving cell per lattice.


Traffic Map Analysis Value Call Drop Rate Distribution

Type UE Distribution(Example for iphone) Real user distribution/High User Spot Distribution

Handover Ratio Distribution


Contents



Case Introduction


Coverage Assessment Workflow - GeneralCoverage Perform ance Assessm ent

Field En

gineer

Nasta

r/Mapinfo/A

rcGIS

Probe A

ssistant

U-net

Omsta

r Se

rver/N

astar

V6R1

0LM

T

005.005.001

Activate M R period Reporting

Raw M R data (7 days)

· Engineering Param eter· Digital M AP

G enerate the Feature Database

003.004

005.005.005

Geo-Location Processing

· DT Data

Adjust Propagation

M odel (Optional)

005.005.004

005.005.002

Activate AGPS Reporting (Optional)

GIS Display

005.005.006

Coverage Audit

005.005.007

Interference Audit

005.005.008

005.005.009

Hotspot Audit

005.005.003

KPI M onitoring

Output Coverage Evaluation Report

005.005.010

Collect the data

002

Cell/RNC/LAC Edge Audit

005.005.0011

005.005.012

Output Topology Evaluation Report


Coverage Assessment Workflow – Based on Prediction


Coverage Assessment Workflow – Based on AGPS

Specification: Emphasis on how to collect AGPS MR data Other steps can be referred in instructions.

Mapping method is helpful to give rise and investigate to obvious RF problems


How to collect the AGPS MR data

Standalone mode: DO NOT need GPS hardware support, UE can search satellite to locate.

UE Based mode: Network sends the satellite data to UE, and help UE to search the satellite, then UE computes the location: GPS hardware can be integrated in the GCGa board, and also can be independent by using GCUa + GPS. Generally, One RNC configures one GPS. (1)AGPS function is controlled by license, so we need confirm

whether the license can be valid : LST LICENSE:FN=“*********.dat”;

(2) RNC Version ： R12 and above R12


MML Command for AGPS MR Collection

RNC：SET UMRCTRL: MRTRAFFICTYPE=TRAFFIC_AMR-1&TRAFFIC_VP-1&TRAFFIC_BE-1&TRAFFIC_OTHER-1, MRINTRAFREQPERIOD=D12, MRRPRTCFGSWITCH=MR_PRD_INTRA_FREQ_MEAS_SWITCH_1-1&MR_EVT_INTRA_FREQ_MEAS_SWITCH_2-1&MR_PRD_INTER_FREQ_MEAS_SWITCH_3-1&MR_EVT_INTER_FREQ_MEAS_SWITCH_4-1&MR_PRD_INTER_RAT_MEAS_SWITCH_5-1&MR_EVT_INTER_RAT_MEAS_SWITCH_6-1&MR_DL_BLER_MEAS_SWITCH_7-0&MR_UE_TX_POWER_MEAS_SWITCH_8-0&MR_UL_SIR_MEAS_SWITCH_9-0&MR_LCS_MEAS_SWITCH_10-1&MR_RACH_MEAS_SWITCH_11-0, MRLCSMETHODSWITCH=LCS_AGPS_UEBASED_SWITCH_1-1&LCS_AGPS_STANDALONE_SWITCH_2-0&LCS_CELLID_RTT_SWITCH_3-1, MRDLBLERPERIOD=D12, MRUETPPERIOD=D12, MRULSIRPERIOD=D12, MRLCSPERIOD=D12;

SET USMLC:SMLCMETHOD=CELLID_CENTER-1&CELLID_RTT-1&OTDOA-1&AGPS-1, AGPSMETHODTYPE=UE_BASED, AGPSADDASSDATASENDFLAG=DELIVERY, UEASSAGPSASSDATASWITCH=REFERENCE_TIME_FOR_GPS-1&REFERENCE_LOCATION_FOR_GPS-1&DGPS_CORRECT-0&GPS_NAVGMODEL-1&GPS_IONOSPHERIC_MODEL-0&GPS_UTC_MODEL-0&GPS_ALMANAC-0&GPS_ACQ_ASSIST-1&GPS_REALTIME_INTEGRITY-0, UEBASAGPSASSDATASWITCH=REFERENCE_TIME_FOR_GPS-1&REFERENCE_LOCATION_FOR_GPS-1&DGPS_CORRECT-1&GPS_NAVGMODEL-0&GPS_IONOSPHERIC_MODEL-1&GPS_UTC_MODEL-0&GPS_ALMANAC-0&GPS_ACQ_ASSIST-0&GPS_REALTIME_INTEGRITY-1, MAXGPSSATS=16, OTDOAMETHODTYPE=UE_ASSISTED, CELLIDRTTMETHODTYPE=UE_ASSISTED, EMERGLCSSWITCH=OFF, OPTIONALIESENDSWITCH=R5_POSDATA_SWITCH-0&LAST_KNOWN_SERVICE_AREA_SWITCH-0, FORCEDSHOSWITCH=OFF, LCSWORKMODE=RNC_CENTRIC;


MML Command for AGPS MR Collection

Cell:Set MR collection scope: SET UMRSCOPE:CELLID = 37851, SCOPETYPE=ByCellId, MRSCOPECTRL=ON;SET UMRSCOPE:CELLID = 37852, SCOPETYPE=ByCellId, MRSCOPECTRL=ON;……If need switch MR collection off, just transfer “MRSCOPECTRL=ON” TO “MRSCOPECTRL=OFF”

ADD USMLCCELL information:ADD USMLCCELL: RNCID = 2340, CELLID = 48143, CELLLOCCFGTYPE = CELL_ANTENNA, GCDF = DEG, MTRLGY= MET, ANTENNALATITUDEDEGREE = 29883800, ANTENNALONGITUDEDEGREE = 121870500, ANTENNAALTITUDEMETER =40, MAXANTENNARANGE=1000, ANTENNAORIENTATION=2700, ANTENNAOPENING= 1200, CELLIDRTTACTIVATEFLAG = INACTIVE, OTDOAACTIVATEFLAG = INACTIVE, AGPSACTIVATEFLAG = ACTIVE;

When MR collection finished, we need rollback the switch above.

Accessories for tools of ADD USMLCCELL AND rollback it:

ADD USMLCCELLRollback ADD USMLCCELL TOOL


Impact for Network after AGPS Switch ON Impact for SPUa Load

Smart Phone penetration rate

SPUa load enhanced

40% 5%

Impact for Connection Performance

Impact for Keeping Performance

Impact for Handover Performance

Command CS RAB Connect successful

rate Call drop rate AGPS report

rate

AGPS ON 98.10% CS: 0.478% 4.5%~5%PS: 0.76%

AGPS OFF 99.83% CS: 0.42% ——PS: 0.42%

Note: If need to degrade the impact of the connect performance further, we can switch AGPS on in batches.

1 week collection


Why Impact Network after AGPS Switch ON

When switch AGPS on, RNC will send measurement control including mass of satellite data to UE.

Sometimes the UE cannot send the response signal to RNC in time, and then cause the Iu release.

In order to solve this problem, we attempt several methods, One effective method is to close the GPS_NAVGMODEL in SET USMLC command. Because it can reduce the num of satellite data from 740 bytes to 130 bytes.

Negative impacts cannot be avoided (just reduced) To limit the impact, we can switch AGPS in batches.

HUAWEI TECHNOLOGIES CO., LTD. Page 22华华华华华华华华华华华华华华， Page 22

Contents



Case Introduction


Create Character DB

Prediction

• Dependant on propagation model and engineer parameters. Accuracy is low.

Use DT to correct

prediction DB

• Use DT data to correct the prediction DB, But just correct the DB on road. Character DB out of road will depend on prediction.

Based on AGPS

• Use AGPS MR to create character DB, not dependant on the engineer parameters and propagation model. Accuracy is highest.


Network Evaluation based on AGPS MR

Cell List

RSCPEc/IoPilot pollutionBest SeverTrafficLAC distributionEvent distributionTerminal distribution

Map Analyzer

Overshooting Cell listCell list of unreasonable LAC distribution Cell list of unbalance for UL and DLCell list with Fault AntennaCell list of unreasonable overlapping zone Cell list of pilot pollution Cell list of abnormal coverage area


Network EvaluationRSCP Coverage

For Urban Ec/Io Coverage

For Urban

We also can display RSCP, Ec/Io for each carrier to find the poor coverage area on the map.


Network EvaluationPilot Pollution Best Sever

From the pilot pollution map, get the more serious pilot pollution area.

From the Best server map, get the cell coverage range.

With this two map analyze which cell cause the pilot pollution


Network EvaluationAMR Traffic Distribution PS Traffic Distribution

Use RAB field to distinguish the service type, RAB = 1, means AMR, RAB = 4, means BE.From the pictures above, we can find the hotspot and prior to pay attention to the hotspot, to suggest network capacity expansion.


Network EvaluationLAC Distribution LAC Distribution

From the LAC distribution and the traffic distribution, we can judge Whether the LAC boundary is reasonable or not. LAC boundary cannot be divided at the dense traffic area.


Network EvaluationCall Drop Event Distribution iPhone Hotspot

Associating MR and PCHR logs, we can display the event map, such as call drop map, HO failure map, etc.We also can display the coverage map and event map of some type terminals such as iPhone.


Network Evaluation - Overshooting According to the relationship of distance and MRs to judge whether the cell is overshooting or not.

From the two pictures, we can clearly see the overshooting cells by Best Server map and MRs PDF.All the overshooting cells can be given by Nastar tool.


Network Evaluation— Faulty Antenna

NodeBID CellidFaulty

specification

3158 315821，2，3 sector

fault

3158 315831，2，3 sector

fault

3158 315811，2，3 sector

fault

3761 37616 2，3 sector fault







Rules:(1) Choose periodic MRs with AGPS.(2) Compute the angle of every AGPS MRs to Best Server azimuth(3) Stat the ratio = the angle > 90 degree/ all periodic MRs(4) If the ratio >50% and there are above two cells with different azimuth in one site


Network Evaluation— Overlapping Zone Rules:(1)Choose periodic MRs.(2) Stat the overlapping zone as couple cells(3)Compute the ratio = MRs of all Non-Best server and

best server are in active set / all MRs in best server.

Cell1 Cell2MRs of all

cell1 and cell2 in active set

MRs inCell1

Overlapping degree1

MRs inCell2

Overlapping degree2

35351 35353 3960 4082 97.01% 47351 8.36%37281 37283 1281 1337 95.81% 21195 6.04%31391 31393 2500 2976 84.01% 78696 3.18%38781 38783 1952 2574 75.84% 20303 9.61%40411 40412 10695 14633 73.09% 17296 61.84%37511 37613 2136 3128 68.29% 23296 9.17%26961 26962 3270 4820 67.84% 36321 9.00%37352 37353 14433 22436 64.33% 32855 43.93%

We can use it to check the overlapping zone of the cells , reduce the SHO rate, and remove the redundant cells.


Network Evaluation — Pilot Pollution

Pilot pollution subject is included in Nastar V6R9 or above. So we can use Nastar to analyze pilot pollution. From the result, we can see the polluted and polluter cell for CS/PS/Total.

Rules:(1) Choose periodic MRs.(2) Neighbor cells Num > 3 and top 4 cells’ RSCP > -95dBm

and Ec/Io < -12dB(3) 1st Cell’s RSCP - 4th cell’s RSCP < 5dB


Network Evaluation— LAC distribution Rules:(1)Choose periodic MRs.(2) Stat the MRs that includes the cells belonging to

different lac in active set.RNCID

CellIDMRs in different

LAC

2354 31642 612

2354 31711 788

2354 31613 847

2354 20833 745

2354 37592 618

2354 31434 408

2354 40793 856

2354 31111 153

2354 37414 1523

2354 31492 111

2354 31046 512

2354 37535 785

2354 37124 1161

The cells with more MRs in different LAC is belonging to unreasonable LAC distribution, we should change the cells’ coverage or the cells’ lac ownership.


Contents



Case Introduction


Case 1 introduction: Weak Coverage Optimization

High traffic(User num)

Few users Area

Azimuth from 130 to 170

Bad RSCP

How to improve?

Azimuth from 240 to 200

Point A B C D E

Best RSCP(dB

m)

Before -97.5 -92.5-

107.5 -87.5 -105

After -91 -92-

102.5 -84 -93Gain(dB

) 6.5 0.5 5 3.5 12

Tuning Summary


Case 2 introduction: Automatic New Added Site Based on Weak Coverage

New site address is auto located where RSCP and users distribution meet specified condition in specified amount of consecutive grids.

Automatic new added site

Automatic site planning: site address

Poor coverage area


Case 3 introduction: Pilot Pollution and Overshooting Problem

Pilot Pollution AnalysisMain

ServingPollution

Cells Suggestion

13106

21906 Control the coverage of

pollution cells

21506250066906

Lon Lat Azimuth RNCID CELLID Overshooting Distance(m)

Overshooting

Area(sq.m)113.634

1 34.7821 150 768 21952 117 1666.0309

Pilot Pollution Problem : Serving cell and Polluted cell

Overshooting Problem : overshooting distribution


Case 4: Hot Traffic Spot Finding (1/2) PS Users Num Distribution

PS EcIo Distribution

PS hot spot: MR larger than some threshold

PS hot spot: MR VS Performance

11% top cells, occupy 27% PS traffic


Case 4: Hot Traffic Spot Finding (2/2)

ResultsVerificati

on

GeographyMagnification

Comparison Analysis area

Human resource(per day)

Time (day) Hot Spot num

Before(with

traditional way ) 5km*5km

site num:70

5 7 Unsure

After(with traffic

map)2 2 14

Gain 60% 71% -

Economic

Precision Verification


Case 5 introduction: Real Coverage Distribution (No Location)

No Location MR: only MR data, Engineering Parameter, RNC Configuration Data (No Feature Database)

Thank youwww.huawei.com

Copyright©2011 Huawei Technologies Co., Ltd. All Rights Reserved.The information in this document may contain predictive statements including, without limitation, statements regarding the future financial and operating results, future product portfolio, new technology, etc. There are a number of factors that could cause actual results and developments to differ materially from those expressed or implied in the predictive statements. Therefore, such information is provided for reference purpose only and constitutes neither an offer nor an acceptance. Huawei may change the information at any time without notice.


1. Question: current accuracy can not satisfy customer’s requirement. And traffic map don’t support indoor coverage so far.

Answer: 1) locating accuracy is related with the actual geography environment for example site distance, the short distance, the higher accuracy. Modify feature database (digital map 5m is best and drive test is needed) can improve the accuracy. Another way open AGPS report, as long as UE support, the accuracy also can be improved.

2. Question: current traffic map don’t support PS throughput geography display.

Answer: The roadmap for PS throughput display will be in R14 at 2012Q2; So far we can combine user number/MR number etc conditions to indirectly recognize.

3. Question: how about the impact for RNC CPU load after open the MR switches.

Answer: In lab test based on Canada traffic model, after open MR switches, the SPU CPU Load will increase 10% relatively(50%华55%). Another side, MR switches support NodeBs and Cell’ level. It can be opened several NodeBs or cells’ switches. After collection the MR data (normally one week), then close the switches.

Network Coverage evaluation using MR reports

Documents