LTE Throughput Root

Cause Analysis
based on DT Data
Solution Overview (LTE TDD and LTE FDD)
The root cause analysis for LTE DT throughput focuses on two types of problems, sudden dropping throughput and low throughput. The analysis is performed in
multiple key dimensions, such as air interfaces, handovers, and abnormal events. The analysis supports Normal scenarios and S FN scenarios.

RAN data
DT data Association (traffic statistics and CHR data)

Evaluation
MAX

Problem
THD Top N cell identification
identification
MIN Sudden dropping Continuous low

Basic problem
Core parameter check Alarm check
check
Isolation and delimitation

Problem
Abnormal events Handover problems Insufficient scheduling Coverage problems
delimitation

Missing neighboring cell Overlap coverage No primary serving cell Handover untimely Restricted uplink

Root cause Analysis PCI confusion PCI conflict Overshoot coverage Overlapping coverage Incorrect coverage

Restricted resource capacity Resource occupied TCP/IP problem check Antenna and channel fault check

Key parameter adjustment PCI adjustment Fault check ACP antenna adjustment

Problem closure
Theme Maturity
➢ Mature of LTE Themes
RAT Theme Submodule Maturity
Common weak coverage Perform RF optimization based on GIS observation on other tools.

Coverage Overlap coverage For reference only. Preferentially handle with problematic routes with overshoot coverage in the interference analysis theme.
analysis No primary serving cell For reference only. Preferentially handle with problematic routes with no primary serving cell in the handover analysis theme.
Restricted uplink This solution cannot determine whether restricted uplink is caused by uplink interference. Manual analysis is required.
Overshoot coverage Rectify engineering parameter problems and perform RF optimization based on GIS observation.
Interference
PCI mod 3 conflict Only cell pairs with conflict are identified. Analyze root causes based on the PCI theme.
analysis
Overlapping coverage Adjust the antenna parameters or power of neighboring cells.
Instead of geographical factors, the time factor is only considered during frequent handover analysis. The length of routes where
Frequent handover frequent handovers occur is provided for reference. When a vehicle runs at a low speed (< 20 km/h), the frequent handover
submodule is restricted for use.
Handover
LTE analysis 1. Match neighboring cells based on missing configurations of neighboring cells, prepare scripts after adding policies, and then
directly deliver the scripts to the network.
Handover untimely
2. Other causes of the handover untimely problem are processed with each submodule in the coverage analysis and interference
analysis themes.
Abnormal event
E-RAB release
Abnormal Intra-frequency/Inter- 1. The abnormal event is analyzed to identify whether it is caused by downlink interference.
event frequency handover failure 2. The abnormal event is correlated with CHRs to obtain the causes for abnormal events recorded in the CHRs.
Inter-RAT handover failure
RRC reestablishment failure
Capacity When the number of RBs or grant does not reach the threshold, check the PRB usage, CCE usage, PUCCH usage, SRS usage, and
Capacity analysis
analysis average number of subscribers in cells where problem routes with low throughput exist.
 Start

Technical Process — Solution Process Output

The solution involves the following operations: Filter out top cells. Top cell list

➢ Filter out top cells: filters out cells that greatly affect the throughput on the entire
network. Input data Check basic
Improper
problems
parameters and
(parameter check
DT data alarm information
➢ Check basic problems: check core parameters and alarms for throughput. and alarm check).

Engineering
parameters Analyze the
➢ Evaluate problem route : using specific rules, identifies problem route with sudden coverage of DT
points on problem
dropping throughput or low throughput (these problem route are considered as Configuration data roads.

analysis objects). Traffic statistics Evaluate Problem roads

problem roads (dropping and low
(roads with throughput)
➢ Analyze whether coverage of problem DT points is proper: Analyzes whether the Alarm data
dropping and low
throughput).

serving cell of the DT point is deployed properly. CHR data Demarcate faults Root cause for the
and locate root problem roads
causes in problem
➢ Demarcate faults and locate the root cause for problem route : analyzes the root roads.

causes in each problem point on the problem route and obtains the main root cause
of the route throughput. Summarize root
causes of top
Root cause for the
top cells
cells.

➢ Locate root causes in top cells: calculates the ratio of root causes of problems in all
top cells, and obtains the main causes for the throughput of the cells. Provide
Optimization
suggestions for
optimization
the cells
suggestions.
➢ Provide optimization suggestions: provides optimization suggestions and directions
for different cells.
End
Technical Principles — Root Cause Overview
Technical Principles — Filtering out Top Cells
/Analyzing Coverage rationality in DT Points on Problem routes
➢ The principles top cells finding are as follows: ➢ Analyzing coverage rationality of DT points in problem routes:
1. Filter out the problem DT points whose throughput is less than the threshold, Coverage rationality analysis involves only the cells that are not
and record the number of problem points of every cell. contained in the neighboring cell table reported by the serving cell.
2. Calculate the number of problem DT points in each problem cell and sort the
problem cells by the number in descending order.
3. Accumulate the numbers of problem DT points. When the accumulated number
is greater than or equal to 80% of the total number of problem DT points, the
accumulation is complete. The cells where the 80% problem DT points locate DT point with an improper
Analyze only the cells that are not reported in the
throughput rate
are the top cells. neighboring cell table reported by the serving cell.
If the reported table indicates that a problem cell is
caused by handover problems or overshoot
coverage problems, the cell is not analyzed.
Calculate the distance (D1)
between the DT point and the
serving cell.

Filter the nearest three cells for

which D1/1.5 (the value can be
changed).

Is the DT point
Are the cells in
in the coverage of the cells Improper
the reported neighboring No Yes
(in the coverage of the main coverage
cell table?
lobe)?

Yes No

The DT point is in proper The DT point is in

coverage. proper coverage.
Technical Principles — Evaluate Problem route with Dropping and Low
Throughput continuously
➢ The procedure for identifying the problem routes with sudden ➢ The procedure for identifying problem routes with low throughput
dropping throughput is as follows: is as follows:
1. Identify the DT point for which the throughput does not reach the threshold.
2. Compare the average throughput at the DT point and its throughput in the Scan the DT points one by one to identify the point that has low throughput.
earlier 3s (the throughput in the 3s reaches the threshold). If the throughput Record the point. When the subsequent three points meet the requirement
deteriorates more than 50%, identify later points of this point until three (the DT point to be added and the previous DT point can cross a maximum
consecutive points whose throughput is greater than the threshold (which of two DT points, and the number of across DT points can be changed), the
indicates that the three points reach the threshold).
DT points are added to the same route until no subsequent three points
3. Calculate the number of DT points from the first identified point to the point
that reaches the threshold. If the number is less than 7, the route meet the requirement. If the number of DT points in a route is greater than 7
encounters the sudden dropping throughput. (the value can be changed), the route is a problem route.
DT point with sudden dropping throughput

Is the
average throughput
The DT point DT point
does not have
No
in the first 3s greater than the
configured threshold?
sudden dropping MAX
throughput.

YES
No. Go to the next DT point.
Is the The DT point THD
throughput of the DT does not have
point less than the configured No
sudden dropping
threshold? throughput.
MIN Sudden dropping Continuous low
YES Is the DT point Is the
a problem point (is the third DT point whose
Is the
The DT point
No
average throughput
of the DT point 50% greater does not have throughput less than the throughput greater than the
No
than the configured deterioration
threshold of the average
sudden dropping
throughput.
threshold)? threshold?
throughput?

YES

Identify a subsequent DT point whose

throughput is greater than the threshold Yes
and calculate the number of DT points from
the first identified DT point this DT point.

The DT point Add one to the number of Is the number of problem

Is the number less does not have
than the lower thredshold? No
sudden dropping problem roads. roads greater than seven?
throughput.

Yes

The DT point has sudden dropping

throughput.
Technical Principles — Analysis Method for Throughput Problems

➢ Sudden dropping throughput: ➢ Low throughput:

The problem is caused by early handovers due to The problem is analyzed as follows: Check whether an
improper parameter settings or a sudden change of the abnormal event occurs on the analyzed route. If an
environment. The following figure shows the detailed abnormal event occurs, locate the event cause. If no
analysis process. abnormal event occurs, check whether a handover problem
occurs (frequency or delay handover). If such a problem
DT point with a sudden
dropping throughput occurs, locate the problem cause. If no such problem
occurs, analyze the air interface. The analysis involves
Analyze the coverage of the DT point.
coverage, interference, and resources.

Does an Does frequency Does delay

Road with low Air interface
abnormal event occur No handover occur on No handover occur on No
throughput rate on the road? the road? the road? analysis
Are the coverage
cells before and after the
No Early handover
throughput suddenly drop are
the same?
Yes Yes Yes

Yes Abnormal Frequent Delay

event alalysis handover handover
analysis analysis

Check whether the RSRP, SINR, and

the number of RBs change suddenly.
Technical Principles — Coverage and Interference Analysis
➢ Coverage and Interference Analysis
The coverage analysis theme is used to identify problems related to coverage, as shown in the following figure. routes and cells with the following problems are identified: missing neighboring cells,
no primary serving cells, overlap coverage, common weak coverage, incorrect coverage, overlapping coverage, interference due to coverage overlap, PCI mod3 conflict, and restricted uplink.
Optimization suggestions are provided.

Scenario Subscenario Condition for Determining the Root Cause Optimization Suggestion
1. The level of the primary serving cell is less than the threshold for the w eak coverage level. The default value is –100 dBm.
Missing 2. The level of the neighboring cell is greater than the threshold for the strong neighboring cell level. The default value is –95 dBm.
Add missing neighboring cells.
neighboring cell 3. The neighboring cell is not in the neighboring cell list of the serving cell.
4. Over coverage does not occur.
1. The level of the primary serving cell is less than the threshold for the w eak coverage level. The default value is –100 dBm.
No primary
2. The number of neighboring cells that meet the condition is greater than 2. The condition is that the RSRP difference between the primary serving cell and the Adjust the dow ntilt of the primary serving cell.
serving cell
Weak neighboring cell is less than 3 dBm.
dow nlink 1. The level of the primary serving cell is less than the threshold for the w eak coverage level. The default value is –100 dBm.
coverage Overlap 2. The level of the neighboring cell is greater than the threshold for the strong neighboring cell level. The default value is –95 dBm. Enhance the coverage of the primary serving
coverage 3. The neighboring cell is not in the neighboring cell list of the serving cell. cell by adjusting the dow ntilt.
4. Over coverage occurs.
1. The level of the primary serving cell is less than the threshold for the w eak coverage level. The default value is –100 dBm.
Increase the coverage of the primary serving
Common w eak 2. The level of the neighboring cell is less than the threshold for the strong neighboring cell level.
cell by adjusting the dow ntilt and increasing
coverage 3. Over coverage does not exist in the primary serving cell.
the pow er of the cell.
4. Incorrect coverage does not exist.
1. The RSRP of a primary serving cell is greater than or equal to –90 dBm, and the SINR is less than 3 dB. Alternatively, the RSRP of a primary serving cell is
Overlapping greater than –100 dBm and less than –90 dBm, and the SINR is less than 0 dB. Adjust the dow ntilt, pow er, or azimuth of the
coverage 2. The number of the cell that meets the condition is greater than or equal to 2. The condition is that the level difference between the primary serving cell and the cell overlapped cell to reduce the overlapped area.
is less than 6 dBm.
Dow nlink 1. The RSRP of a primary serving cell is greater than or equal to –90 dBm, and the SINR is less than 3 dB. Alternatively, the RSRP of a primary serving cell is
Overshoot Adjust the dow ntilt of neighboring cells w ith
interference greater than –100 dBm and less than –90 dBm, and the SINR is less than 0 dB.
interference over coverage.
2. Over coverage is checked on the primary serving cell and its neighboring cell.
1. The RSRP of a primary serving cell is greater than or equal to –90 dBm, and the SINR is less than 3 dB. Alternatively, the RSRP of a primary serving cell is
PCI mod3 Adjust the antenna azimuth, pow er, or PCI of
greater than –100 dBm and less than –90 dBm, and the SINR is less than 0 dB.
conflict neighboring cells for optimization.
2. PCI MOD3 conditions are met.
Restricted Restricted 1. The uplink receive signal RSRP is good and greater than the w eak coverage threshold. Remove uplink interference or narrow
uplink uplink 2. The PCC PUCCH uplink transmit pow er is greater than 10 dB. dow nlink coverage.
Restricted 1. The ratio of DT points w ith insufficient scheduling on the problem route is greater than the low er threshold (20%). Instruct the customer to perform capacity
Insufficient sources 2. The PRB usage, CCE usage, PUCCH usage, and SRS usage of problem cells on the problem route are greater than the thresholds during the test. expansion.
scheduling Troubleshoot cell (cell name) transmission
Other problems Insufficient scheduling occurs on the problem route but the resources are not restricted.
and TCP transmission.
Technical Principles — Handover Analysis
➢ Basic Principles for Handover Analysis
 The handover analysis theme is used to identify handover-related problems. The problem types that can be identified
are shown in the following figure. After the analysis, problem routes with handover delays, frequent handovers, and
ping-pong handovers are isolated and optimization suggestions are provided.

Subtheme Determination Condition Cause Condition for Determining the Root Cause Optimization Suggestion
No primary serving cell For details, see the description for the coverage theme. For details, see the coverage theme.
Frequent handover duration
Frequent threshold (the default value is 10s) Check the handover threshold and the
Ping-pong handover N/A
handover and the number of frequent handover penalty timer.
handovers (the default value is 2)
Frequent handover N/A Check handover parameters.

Whether the PCIs of a serving cell and its neighboring cell are
When a UE reports A3/A4 events, PCI confusion Adjust the confused PCIs.
confused.
no inter-frequency or intra-
Handover frequency handover occurs if the Missing neighboring cell N/A Add missing neighboring cells.
delay configured handover delay duration
expires (the default duration is 300 Check handover-related parameters or

ms). Handover delay N/A check whether air interface signaling is

missing.
Technical Principles — Abnormal Event Analysis
➢ Basic Principles for Abnormal Event Analysis Based on DT Data
 For a route with low throughput, check whether any abnormal event occurs on the route. If any abnormal event occurs,
identify the event (whether it is a call drop, handover failure, or other events). The following figure shows the
identification process. Check whether the events are caused by coverage or interference problems on the route. If the
events are caused by coverage or interference problems, optimization suggestions are provided.
Technical Principles — Alarm Check, Resource Capability Check,
Parameter Check, and CHR Correlation Analysis
➢ Basic Principles for Alarm Check ➢ Basic Principles for Parameter Check
 If abnormal throughput rates occur in a cell or a site, all  Based on abnormal rates or problem cells with abnormal P3
alarms of the cell or the site are checked. Among all alarms, if KPIs, the related feature parameters in problem cells are
the alarms about abnormal throughput, the routes with checked. If the values of those parameters are different from
abnormal throughput are correlated with alarms, and the
baseline values, the parameter values are compared with the
impact of those alarms and optimization suggestions are
baseline values to facilitate problem locating.
provided.

➢ Basic Principles for Resource Capability Check ➢ Basic Principles for CHR Correlation Analysis
 Based on the location and occurrence time of abnormal rates  For abnormal call drops, intra-RAT handover failures, and inter-
or abnormal P3 counters, traffic KPIs about resource RAT handover failures, if air interface problems do not occur

capacities in the corresponding cells are checked. Check (the conclusion can be drawn using DT data), the CHR data can
be used to further locate the problem causes on the network.
whether resource capacity problems exist in the cell at this
This theme can associate DT data of a single user and the CHR
time. If the problems exist, the abnormal routes and resource
data on the related site, select the event which is corresponding
capacity problems are correlated and their information is
to that of the single user from the CHR data, and locate network
provided.
causes based on the event types and failure cause records.