FDD LTE Key Performance Indicator s Description Guide
Learning Objectives After having learned this training course, you should be able to understand:
Basic knowledge of FDD LTE KPIs
Categories of FDD LTE KPIs
Formulas of FDD LTE KPIs
Contents
Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs
References
Counters, Definitions, and Numbering Conventions
For more information, refer to the NetNumen M31(LTE) NE Management System eNod eB Performance Counter Reference at the following URL: http://tsm.zte.com.cn/tsm/FileCenter/File.aspx?Mode=read&FileID=30341574
Formulas
For more information, refer to the NetNumen M31(LTE) NE Management System eNod eB Key Performance Indicator Reference at the following URL: http://tsm.zte.com.cn/tsm/FileCenter/File.aspx?Mode=read&FileID=30341576
Counter Example
Counter
Example
Measurement Object
Measurement Number
Measurement Type
Counter Name
Counter Sequence Number
Cell Type
37320
Statistics of RRC Connection Establishment
Number of Successful mtAccess RRC Connection Establishment
C373200000
Counter Structure Measurement Object
Measurement Type
ID: 780-799 • • • •
Cell Type eNodeB Type eNodeB IP Link Type Cell Pair Type
Counter
ID: 373320-373399 • • • • •
Statistics of RRC connection establishment Statistics of E-RAB Statistics of time Type of a cell pair Other
ID: C + Measurement Type Number + XXX •
•
•
Number of successful mtAccess RRC connection establishment Number of unsuccessful mtAccess RRC connection establishment (timer timeout) Other
KPI/PI KPI ID: 310500-310799 • •
RRC connection establishment success rate Other
PI ID: 310500-310799 • •
Paging congestion rate Other
Reporting Procedure
This reporting procedure is described as follows: Step 1 The RNLC subsystem reports data to the OAM using the event-triggered reporting mechanism. Step 2 The CMAC, RNLU, BRS, or OSS subsystem reports data to the OAM every 10 seconds. Step 3 The OAM synchronizes data with the OMC for each measurement object every 15 minutes.
KPI Overview Description Mobile subscriber (Serviceability + Reliability): An FDD LTE can provide a mobile subscriber with high-quality, reliable, and long-term services. Mobile operator (Serviceability + Reliability + Traffic): An FDD LTE can provide as many mobile subscribers as possible with high-quality, reliable, and longterm services. Dependency System performance: UE + eNodeB + Transport + EPC Applicable environments: Bandwidth configuration, radio access, mobility speed, service type, and so on Categories FDD LTE eNodeB KPIs FDD LTE Network KPIs
Contents
Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs
FDD LTE eNodeB KPIs
Latency
C-Plane Latency
U-Plane Latency
Throughput (Mono-UE)
Peak UE Data Rate
Average UE Data Rate
Cell Edge UE Data Rate
Cell Capacity (Multi-UE)
Peak Cell Throughput
Average Cell Throughput
Cell-Edge UE Throughput
U-Plane Latency Round Trip Time Combined EPC Application Server eNodeB UE
ping
Measured Round Trip Time
Dependency Radio frequency: Under cell, mid-cell, and cell edge Neighbor cell loading: Unloaded and loaded Scheduling algorithm: Pre-scheduled and non-scheduled Ping size: 32 B, 1000 B, and 1500 B
C-Plane Latency State Transition Time
Signaling Procedure
Throughput (Mono-UE) Peak UE Data Rate
Transmission
Terminal
BS DownlinkorUplinkDataStream
Assumption
Loading: Single cell, unloaded
UE speed: Stationary mode
Location: Under cell (good RF conditions)
MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO
Dependency
Operating bandwidth
UE category
UE UL/DL RB limitation
E.g. Gateways, Routers, Firewalls
Application Server at NGMN network edge
Throughput (Mono-UE) Average UE Data Rate
Assumption
MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO
Loading: 1 UE in the serving cell, neighbor cell loaded
UE speed: Stationary mode
Location: Uniformly-distributed locations over the signal quality range
Dependency
Operating bandwidth
UE category
UE UL/DL RB limitation
UE distribution in each signal range
Throughput (Mono-UE) Cell-Edge UE Data Rate
Assumption
MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO
Loading: 1 UE in the serving cell, neighbor cell loaded
UE speed: Stationary mode
Location: Cell edge
Dependency
Operating bandwidth
UE category
UE UL/DL RB limitation
UE SINR at the cell edge
Throughput (Multi-UE) Peak Cell Throughput
Assumption
MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO
Loading: Multiple cells, loaded (70%)
UE Speed: Stationary mode
Location: Good RF conditions
Dependency
UE category: TBS limitation, 64QAM in UL
UE UL/DL RB limitation
Environment: Dense urban, urban, suburban, and rural
Scheduling algorithm
Throughput (Multi-UE) Average Cell Throughput Typical UE Distribution Model
Assumption
MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO
Loading: Multiple cells, loaded
UE speed: Stationary Mode
Location: Uniformly-distributed over the signal quality range
Dependency
UE category (TBS limitation, 64QAM in UL)
UE UL/DL RB limitation
UE distribution in each signal range, scheduling algorithm
Throughput (Multi-UE) Cell-Edge UE Throughput
Assumption
MIMO configuration: DL 2*2 MIMO, UL 1*2 SIMO
Loading: Multiple cells, loaded (70%)
UE speed: Stationary mode
Location: Cell edge
Dependency
UE category (TBS limitation, 64QAM in UL)
UE UL/DL RB limitation
Environment: Dense urban, urban, suburban, and rural
Scheduling algorithm
UE SINR at the cell edge
Contents
Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs
FDD LTE Network KPIs Accessibility RRC Establishment Success Rate E-RAB Setup Success Rate Retainability E-RAB Drop Rate Mobility Handover Success Rate Availability Cell Availability Integrity Packet Loss Rate DL PDCP SDU Latency
Accessibility Theoretical limit: 100% Ideal value in a commercial network: > 98%
Accessibility (Category)
Initial E-RAB Accessibility
KPI
RRC Connection Setup Success Rate
S1-SIG Establish Success Rate
Initial E-RAB Setup Success Rate
Added E-RAB Accessibility
Attach Success Rate
Detach Success Rate
RRC Re-Establishment Success Rate
Paging Success Rate
Air Interface
RRC Connection Setup Success Rate
S1-SIG Establish Success Rate
KPI
UE Context Setup Success Rate
Added E-RAB Setup Success Rate
E-RAB Setup Success Rate
E-RAB Block Rate (per QCI)
Call Setup Success 、 Rate, Call Barring Rate
Contention-Based PRACH Performance
Contention-Free PRACH Performance
RRC Establishment Success Rate
Description
This KPI shows the probability for a subscriber to be provided with an RRC connection upon request.
Signaling Procedure UE
EUTRAN
RRCConnectionRequest
2
RRCConnectionSetup
RRCConnectionSetupComplete
1
3
Formula
RRC Establishment Success Rate = Number of successful RRC connection establishment / (Number of successful RRC connection establishment + Number of failed RRC connection establishment) * 100%
E-RAB Setup Success Rate
Description
This KPI shows the probability for a subscriber to be provided with an E-RAB request including initial and added context setup procedures.
Signaling Procedure eNB INITIAL CONTEXT SETUP REQUEST
INITIAL CONTEXT SETUP RESPONSE
eNB
MME
1
E-RAB SETUP REQUEST
E-RAB SETUP RESPONSE
2
MME
3 4
Formula
E-RAB Setup Success Rate = (Number of successful initial E-RAB + Number of successful added E-RAB establishment) / (Number of successful initial E-RAB establishment + Number of failed initial E-R AB establishment + Number of successful added E-RAB establishment + Number of failed added E-RAB establishment) * 100%
Retainability Theoretical limit: 0% Ideal value in a commercial network: < 2%
Retainability (Category)
KPI UE E-RAB Retainability Per QCI
UE E-RAB Retainability
RRC Drop Rate
E-RAB Drop Rate (Per QCI)
How Often
E-RAB Duration Time per QCI Drop
Percentage
E-RAB UL Data Volume Per QCI Drop QoS Before Drop
E-RAB DL Data Volume Per QCI Drop
Active E-RAB Drop Rate
E-RAB Drop Rate
Description
This KPI shows the probability for an a subscriber to loss the E-RAB, such as an event b eing released by the eNodeB due to overload control.
Signaling Procedure E-RAB Released by the eNodeB)) eNB
1
MME
E-RAB RELEASE INDICATION
Formula
E-RAB Drop Rate = Number of Abnormally Released E-RAB / Number of Successfully Established E-RAB * 100%
Mobility Successful Handover Categories
Intra-Freq Intra-eNodeB
Intra-Freq Inter-eNodeB
Inter-Freq Inter-eNodeB
Inter-Freq Intra-eNodeB
LTE to UMTS
UMTS to LTE
LTE to GSM
Inter- RAT HO Success Rate X2 Based HO
Contention Based HO
S1 Based HO
All Incoming HO Per Cell
All Outgoing HO Per Cell
All Incoming HO Per Cell Pair
All Outgoing HO Per Cell Pair
Contention Free HO
Intra-RAT HO Success Rate
Troubleshooting
GSM to LTE
Mobility Handover Handover Preparation Success Rate Theoretical limit: 100% Ideal value in a commercial network 、 99% (Step 4 – Step 6)
Handover Execution Success Rate Theoretical limit: 100% Ideal value in a commercial network: 98% HO In: Step 6 – Step 11 HO Out: Step 7 – Step 17
Mobility Success Rate
Theoretical limit: 100%
Ideal value in a commercial network: 97%
MobilitySR= HO. Pr eparationSR × HO.ExecutionSR
Handover Preparation and Execution Handover
Start Point
Outgoing handover preparation
The source eNodeB decides to perform a handover.
Stop Point The source eNodeB sends the RRC Connection Reconfiguration message to the UE.
Outgoing handover execution
The source eNodeB sends the RRC Connection Reconfiguration message to the UE.
The source eNodeB receives the UE Context Release message from the destination eNodeB.
Incoming handover preparation
The destination eNodeB receives the Handover Request message from the source eNodeB.
The destination eNodeB returns the Handover Response message to the source eNodeB.
Incoming handover execution
The destination eNodeB receives the RRC Connection Reconfiguration Complete message from the UE.
The destination eNodeB sends the UE Context Release message to the source eNodeB.
Cell Handover and Cell Pair Handover
The outgoing handover preparation success rate for each handover pair measures the hando ver preparation from the serving cell to a certain neighbor cell. Here are two typical examples:
In the event of the handover from cell A to cell B, the Intra-eNodeB Intra-freq Outgoing Handover Preparation Success Rate per Cell pair is measured in cell A.
In the event of the handover from cell B to cell A, the Intra-eNodeB Intra-freq Outgoing Handover Preparation Success Rate per Cell pair is measured in cell B.
The outgoing handover success rate contains all outgoing handover preparation successes fr om the serving cell to all neighbor cells.
The measurement of these two KPIs can facilitate us in sifting two cells that suffer most hando ver preparation failures, performing neighbor cell optimization, and even deleting unusable nei ghbor cells.
Intra-eNodeB Handover Success Rate
Description
The intra-eNodeB handover success rate measures the service continuity when a subs criber is on the move. This KPI is perceptible to the subscribers, depending on system hand over processing capabilities and network planning.
Formulas
Intra-frequency handover success rate = Number of intra-frequency handover succes ses / Number of intra-frequency handover requests * 100%
Inter-frequency handover success rate = Number of inter-frequency handover succes ses / Number of inter-frequency handover requests * 100%
Intra-eNodeB Handover Success Rate
Note: If the eNodeB receives the RRC Connection Reconfiguration Complete message in step 4, it i ndicates that the handover is successful. If the eNodeB receives the RRC Connection Reestablishment Request message in step 6, it i ndicates that the handover is unsuccessful.
Inter-eNodeB X2-Interface Handover Success Rate Description The inter-eNodeB X2-interface handover success rate measures the handover successes when the UE moves between the eNodeBs over the X2 interface. This KPI is perceptible to the subscribers being o n the move, depending on system handover processing capabilities and network planning.
Formulas
Outgoing intra-frequency X2-interface handover success rate = Number of outgoing intra-freque ncy X2-interface handover successes / Number of outgoing intra-frequency X2-interface handov er attempts (serving cell) * 100%
Incoming intra-frequency X2-interface handover success rate = Number of incoming intra-freque ncy X2-interface handover successes / Number of incoming intra-frequency X2-interface handov er attempts (serving cell) * 100%
Outgoing inter-frequency X2-interface handover success rate = Number of outgoing inter-freque ncy X2-interface handover successes / Number of outgoing inter-frequency X2-interface handov er attempts (serving cell) * 100%
Incoming inter-frequency X2-interface handover success rate = Number of incoming inter-freque ncy X2-interface handover successes / Number of incoming inter-frequency X2-interface handov er attempts (serving cell) * 100%
Inter-eNodeB X2-Interface Handover Success Rate
Inter-eNodeB S1-Interface Handover Success Rate
Description When the eNodeB decides to perform a handover according to the UE measurement report and me anwhile the destination cell is not connected to the eNodeB through the X2 interface, the inter-eNod eB S1-interface handover success rate measures the S1-interface handover performed through the EPC. This KPI is perceptible to the subscribers being on the move, depending on system handover processing capabilities and network planning.
Formulas
Outgoing intra-frequency S1-interface handover success rate = Number of outgoing intra-freque ncy S1-interface handover successes / Number of outgoing intra-frequency S1-interface hando ver attempts (serving cell) * 100%
Incoming intra-frequency S1-interface handover success rate = Number of incoming intra-frequ ency S1-interface handover successes / Number of incoming intra-frequency S1-interface hand over attempts (serving cell) * 100%
Outgoing inter-frequency S1-interface handover success rate = Number of outgoing inter-frequ ency S1-interface handover successes / Number of outgoing inter-frequency S1-interface hand over attempts (serving cell) * 100%
Incoming inter-frequency S1-interface handover success rate = Number of incoming inter-frequ ency S1-interface handover successes / Number of incoming inter-frequency S1-interface hand over attempts (serving cell) * 100%
Inter-eNodeB S1-Interface Handover Success Rate
Inter-System Handover Success Rate The inter-system handover success rate consists of both incoming and outgoing handover between the LTE network and t he CDMA network, between the LTE network and the UMTS network, between the LTE network and the GSM network. Thi s example shows a handover from the CDMA network to the LTE network.
Availability Availability
Important and Demanding
Cell availability Theoretical limit: 100% Ideal value in a commercial network: > 99.995%
CellAvailability =
measurement_period- ∑ RRU.CellUnavailableTime.[cause] cause
measurement _ period
Dependency Software + Hardware Unavailable Time = Unplanned downtime only (excluding planned downtime) Physical meaning 365 * 24 * 60 * (1-99.995%) = 26.28 min cell out-of-service time
×100
Cell Availability
Description The cell availability measures the ratio of in-service time to measurement granularity time. The inservice time indicates the time interval between cell establishment and cell deletion. By counting the cell in-service time, this KPI forms a foundation for analyzing system failures and measuring system stability.
Signaling Procedure
Formulas Cell Availability = In-Service Time / Measurement Granularity Time Cell Availability = C373230700 / Measurement Granularity Time * 100%
Integrity UL/DL Packet Loss Rate
QCI
Priority
Packet Latency Budget (NOTE 1)
Packet Error Loss Rate (NOTE 2)
2
100 ms
10-2
Conversational Voice
4
150 ms
10-3
Conversational Video (Live Streaming)
3 (NOTE 3)
3
50 ms
10-3
Real Time Gaming
4 (NOTE 3)
5
300 ms
10-6
Non-Conversational Video (Buffered Streaming)
5 (NOTE 3)
1
100 ms
10-6
IMS Signalling
6
300 ms
10-6
Video (Buffered Streaming) TCP-based (e.g., www, e-mail, chat, ftp, p2p file sharing, progressive video, etc.)
7
100 ms
10-3
Voice, Video (Live Streaming) Interactive Gaming
300 ms
10-6
DL PDCP SDU Latency
Resour ce Type
1 (NOTE 3) 2 (NOTE 3)
GBR
6 (NOTE 4) 7 (NOTE 3) 8 (NOTE 5)
9 (NOTE 6)
NonGBR
8
9
Example Services
Video (Buffered Streaming) TCP-based (e.g., www, e-mail, chat, ftp, p2p file File sharing, progressive video, etc.
Downlink PDCP SDU Latency
Description This KPI indicates average downlink PDCP SDU latency based on the QCI type, from the time when a PDCP SDU reaches the eNodeB, to the time when the UE receives this PDCP SDU, th at is to say, all fragments of this PDCP SDU receives a successful HARQ response.
Signaling Procedure
Formula Average Downlink PDCP SDU Latency = Total Latency of All PDCP SDUs / Number of All PDCP SDUs
Downlink IP Packet Latency
Description This KPI indicates average downlink IP packet latency, from the time when the eNodeB receives the IP packet through the S1 or X2 interface, to the time when the first fragment of this IP packet is transmitted by the eNodeB through the air interface. It measures the time interval at which the service is processed by the eNodeB, which forms a strong foundation for network optimization.
Signaling Procedure
Formula Average Downlink IP Packet Latency (QCI 1 – 9) = Total Downlink IP Packet Latency (QCI 1 – 9 ) / N umber of All PDCP SDUs (Downlink QCI 1 – 9)
Uplink/Downlink Packet Loss Rate
Description This KPI measures the ratio of discarded PDCP SDUs to received PDCP SDUs due to the timeout of the TimeDisCard timer, when no, partial, or all fragments are transmitted through the eNodeB or air interface. It should be noted that PDCP SDUs vary from QCI to QCI (1 – 9), from uplink to downlink.
Counters Number of All PDCP SDUs Uplink: When the PDCP layer of the eNodeB receives the PDCP SDU from the UE, this counter is incremented by 1. Downlink: When the PDCP layer of the eNodeB sends the PDCP SDU to the RLC layer, this counter is incremented by 1. Number of Discarded PDCP SDUs Uplink: When the PDCP layer of the eNodeB receives the PDCP SDU from the UE, the SN is not consecutive. Downlink: When the PDCP layer of the eNodeB sends the PDCP SDU to the RLC layer, the SN is not consecutive.
Formula Air Interface or eNodeB Packet Loss Rate = Number of Discarded Packets over the Air Interface or eNodeB / Number of All Packets over the Air Interface or eNodeB
Contents
Basic Knowledge of FDD LTE KPIs FDD LTE eNodeB KPIs FDD LTE Network KPIs Other FDD LTE KPIs
Traffic Type - Number of RRC Connections
Description This KPI counts the number of RRC connections, which is used to measure the UE access to the system for call hold. It involves the following two counters:
Average number of RRC connections Maximum number of RRC connections.
Signaling Procedure
When the eNodeB receives the RRC Establishment Complete message, the number of RRC connections is in cremented by 1.
When the eNodeB triggers the RRC establishment successfully due to the handover, the number of RRC con nections is incremented by 1.
When the eNodeB releases the RRC connection, the number of RRC connections is decremented by 1.
When the eNodeB reestablishes the RRC connection in another cell, the number of RRC connections is decre mented by 1.
When the eNodeB reestablishes the RRC connection back to the serving cell, the number of RRC connection s is incremented by 1.
Formulas
Maximum Number of RRC Connections (Sampling Counter) Average Number of RRC Connections (Sampling Counter)
Traffic Type – Average Number of QCI-Based UEs
Description This KPI measures the average number of E-RAB connections, which evaluates network traffic. It s hould be noted that this KPI varies from QCI to QCI.
Signaling Procedure (Sampling Counter)
Initial E-RAB establishment success
Added E-RAB establishment success
Incoming E-RAB handover success
Incoming E-RAB modification success (changing the new QCI to the old QCI)
Outgoing E-RAB handover success
Outgoing E-RAB modification success (changing the old QCI to the new QCI)
E-RAB release
Formula
This KPI is counted by averaging the measured values of all sampling points within a specific meas urement cycle.
Traffic Type – Number of Activated QCI-Based UEs
Description This KPI counts the number of UEs in the cache during a specified time period. It can be m easured based on a specific QCI, uplink or downlink, average number or maximum number .
This KPI indicates the number of in-service UEs in the system, which forms a foundation fo r evaluating system capacity and capabilities.
Signaling Procedure N/A
Formula
This KPI is counted when any data is present in the E-RAB cache of QCI i every sampling cycle (100 ms), as defined in the 3GPP TS.
Resource Allocation Cell & System (Category) Resources
UL -C PRB Usage
DL -C PRB Usage
Total UL PRB Usage
PDCCH Usage
UL -U PRB Usage
PRACH Usage
UL PRB Usage per QCI
Paging Congestion Rate
Total D L PRB Usage
Avg CPU Load
DL -U PRB Usage
Avg DL Transmission Power
DL PRB Usage per QCI
Avg UL Interference Per PRB
PRACH Propagation Delay
Avg DSP Load
PRACH Message Load
