3rd Generation Partnership Project;

3GPP TS 28.554
Technical Specification Group Services and System
V16.4.0 (2020-03)
Aspects;Management and orchestration; Technical Specification
5G end to end Key Performance Indicators (KPI)
(Release 16)

The present document has been developed within the 3rd Generation Partnership Project (3GPP TM) and may be further elaborated for the purposes of 3GPP..
The present document has not been subject to any approval process by the 3GPP Organizational Partners and shall not be implemented.
This Specification is provided for future development work within 3GPP only. The Organizational Partners accept no liability for any use of this
Specification.
Specifications and Reports for implementation of the 3GPP TM system should be obtained via the 3GPP Organizational Partners' Publications Offices.
Release 16 2 3GPP TS 28.554 V16.34.0 (20192020-1203)

Keywords
5G,KPI,management,orchestration

3GPP

Postal address

3GPP support office address

650 Route des Lucioles - Sophia Antipolis
Valbonne - FRANCE
Tel.: +33 4 92 94 42 00 Fax: +33 4 93 65 47 16

Internet
http://www.3gpp.org

Copyright Notification

No part may be reproduced except as authorized by written permission.

The copyright and the foregoing restriction extend to reproduction in all media.

© 2020, 3GPP Organizational Partners (ARIB, ATIS, CCSA, ETSI, TSDSI, TTA, TTC).
All rights reserved.

UMTS™ is a Trade Mark of ETSI registered for the benefit of its members
3GPP™ is a Trade Mark of ETSI registered for the benefit of its Members and of the 3GPP Organizational Partners
LTE™ is a Trade Mark of ETSI registered for the benefit of its Members and of the 3GPP Organizational Partners
GSM® and the GSM logo are registered and owned by the GSM Association

3GPP
Release 16 3 3GPP TS 28.554 V16.34.0 (20192020-1203)

Contents
Foreword..........................................................................................................................................................5
1 Scope......................................................................................................................................................6
2 References..............................................................................................................................................6
3 Definitions and abbreviations.................................................................................................................6
3.1 Definitions...........................................................................................................................................................6
3.2 Abbreviations.......................................................................................................................................................6
4 End to end KPI concept and overview....................................................................................................7
5 KPI definitions template.........................................................................................................................7
6 End to end KPI definitions.....................................................................................................................7
6.1 KPI Overview......................................................................................................................................................7
6.2 Accessibility KPI.................................................................................................................................................8
6.2.1 Registered subscribers of network and network slice instance through AMF....................................................8
6.2.2 Registered subscribers of network and network slice instance through UDM....................................................8
6.2.3 Registration success rate of one single network slice instance............................................................................8
6.2.4 DRB Accessibility for UE services................................................................................................................8
6.2.5 PDU session Establishment success rate of one network slice (S-NSSAI)...................................................9
6.3 Integrity KPI........................................................................................................................................................9
6.3.1 Latency of 5G Network..................................................................................................................................9
6.3.1.0 KPI categories..........................................................................................................................................9
6.3.1.1 Downlink latency in gNB-DU..................................................................................................................9
6.3.1.2 Integrated downlink delay in RAN........................................................................................................10
6.3.2 Upstream throughput for network and Network Slice Instance...................................................................10
6.3.3 Downstream throughput for Single Network Slice Instance.......................................................................10
6.3.4 Upstream Throughput at N3 interface.........................................................................................................10
6.3.5 Downstream Throughput at N3 interface.....................................................................................................11
6.3.6 RAN UE Throughput...................................................................................................................................11
6.3.6.1 Definition...............................................................................................................................................11
6.3.6.2 Extended definition................................................................................................................................11
6.4 Utilization KPI...................................................................................................................................................12
6.4.1 Mean number of PDU sessions of network and network Slice Instance.....................................................12
6.4.2 Virtualised Resource Utilization of Network Slice Instance.......................................................................12
6.4.3 PDU session establishment time of network slice.......................................................................................13
6.4.4 Mean number of successful periodic registration updates of Single Network Slice...................................13
6.5 Retainability KPI...............................................................................................................................................13
6.5.1 QoS flow Retainability.................................................................................................................................13
6.5.1.1 Definition...............................................................................................................................................13
6.5.1.2 Extended definition................................................................................................................................14
6.5.2 DRB Retainability........................................................................................................................................14
6.5.2.1 Definition...............................................................................................................................................14
6.5.2.2 Extended definition................................................................................................................................15
6.6 Mobility KPI......................................................................................................................................................15
6.6.1 NG-RAN handover success rate..................................................................................................................15
6.6.2 Mean Time of Inter-gNB handover Execution of Network Slice................................................................15
6.6.3 Successful rate of mobility registration updates of Single Network Slice...................................................16
6.7 Energy Efficiency (EE) KPI..............................................................................................................................16
6.7.1 NG-RAN data Energy Efficiency (EE)........................................................................................................16
6.7.1.1 Definition...............................................................................................................................................16

Annex A (informative): Use cases for end to end KPIs.....................................................................17

A.1 Use case for end-to-end latency measurements of 5G network-related KPI.........................................17
A.2 Use case for number of registered subscribers of single network-slice-instance-related KPI...............17

3GPP
Release 16 4 3GPP TS 28.554 V16.34.0 (20192020-1203)

A.3 Use case for upstream/downstream throughput for one-single-network-slice-related KPI...................17

A.4 Use case for mean PDU sessions number in network slice instance.....................................................17
A.5 Use case for virtualised resource utilization of network-slice-instance related KPI..............................18
A.6 Use case for 5GS registration success rate of one single-network-slice instance-related KPI...............18
A.7 Use case for RAN UE throughput-related KPI.....................................................................................18
A.8 Use case for QoS flow retainability-related KPI...................................................................................18
A.9 Use case for DRB accessibility-related KPI.........................................................................................18
A.10 Use case for mobility KPI.....................................................................................................................19
A.11 Use case for DRB retainability related KPI..........................................................................................19
A.12 Use case for PDU session establishment success rate of one network slice (S-NSSAI) related KPI....19
A.13 Use case for integrated downlink latency in RAN................................................................................19
A.14 Use case for PDU session Establishment success rate of one single-network-slice instance-related
KPI.......................................................................................................................................................20
A.15 Use case for QoS flow retainability-related KPI...................................................................................20
A.16 Use case for 5G Energy Efficiency (EE) KPI.......................................................................................20
Annex B (informative): Change history................................................................................................22

3GPP
Release 16 5 3GPP TS 28.554 V16.34.0 (20192020-1203)

Foreword
This Technical Specification has been produced by the 3rd Generation Partnership Project (3GPP).

The contents of the present document are subject to continuing work within the TSG and may change following formal
TSG approval. Should the TSG modify the contents of the present document, it will be re-released by the TSG with an
identifying change of release date and an increase in version number as follows:

Version x.y.z

where:

x the first digit:

1 presented to TSG for information;

2 presented to TSG for approval;

3 or greater indicates TSG approved document under change control.

y the second digit is incremented for all changes of substance, i.e. technical enhancements, corrections,
updates, etc.

z the third digit is incremented when editorial only changes have been incorporated in the document.

3GPP
Release 16 6 3GPP TS 28.554 V16.34.0 (20192020-1203)

1 Scope
The present document specifies end-to-end Key Performance Indicators (KPIs) for the 5G network and network slicing.

2 References
The following documents contain provisions which, through reference in this text, constitute provisions of the present
document.

- References are either specific (identified by date of publication, edition number, version number, etc.) or
non-specific.

- For a specific reference, subsequent revisions do not apply.

- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including
a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same
Release as the present document.

[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".

[2] Void.

[3] ITU-T Recommendation E.800: "Definitions of terms related to quality of service".

[4] 3GPP TS 24.501: " Non-Access-Stratum (NAS) protocol for 5G System (5GS); Stage 3".

[5] 3GPP TS 38.331: "NR; Radio Resource Control (RRC); Protocol specification".

[6] 3GPP TS 28.552: "Management and orchestration; 5G performance measurements".

[7] 3GPP TS 23.501: " System Architecture for the 5G System; Stage 2".

[8] ETSI ES 203 228 V1.2.1 (2017-04): "Environmental Engineering (EE); Assessment of mobile
network energy efficiency".

[9] 3GPP TS 28.310: "Management and orchestration; Energy efficiency of 5G".

3 Definitions and abbreviations

3.1 Definitions
For the purposes of the present document, the terms and definitions given in 3GPP TR 21.905 [1] and the following
apply. A term defined in the present document takes precedence over the definition of the same term, if any, in 3GPP
TR 21.905 [1].

3.2 Abbreviations
For the purposes of the present document, the abbreviations given in 3GPP TR 21.905 [1] and the following apply. An
abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in
3GPP TR 21.905 [1].

EE Energy Efficiency
kbit kilobit (1000 bits)
RTT Round Trip Time

3GPP
Release 16 7 3GPP TS 28.554 V16.34.0 (20192020-1203)

4 End to end KPI concept and overview

The following KPI categories are included in the present document:

- Accessibility (see the definition in [3]).

- Integrity (see the definition in [3]).

- Utilization.

- Retainability (see the definition in [3]).

- Mobility.

- Energy Efficiency.

Editor's note: For future update of the document it will also include:

- Availability.

5 KPI definitions template

a) Name (Mandatory): This field shall contain the name of the KPI.

b) Description (Mandatory): This field shall contain the description of the KPI.
Within this field it should describe if the KPI is focusing on network or user view. This filed should also describe
the logical KPI formula to derive the KPI. For example, a success rate KPI’s logical formula is the number of
successful events divided by all events. This field should also show the KPI unit (e.g., kbit/s, millisecond) and
the KPI type (e.g., mean, ratio).

c) Formula definition (Optional):

This field should contain the KPI formula using the 3GPP defined measurement names.
This field can be used only when the measurement(s) needed for the KPI formula are defined in 3GPP TS for
performance measurements (TS 28.552 [6]).

d) KPI Object (mandatory)

This field shall contain the DN of the object instance where the KPI is applicable. The DN identifies one object
instance of the following IOC:

- NetworkSliceSubnet

- SubNetwork

- NetworkSlice

e) Remark: (Optional)
This field is for additional information reqquired for the KPI definition,
e.g. the definition of a call in UTRAN.

6 End to end KPI definitions

6.1 KPI Overview

The KPI categories defined in [2] will be reused by the present document.

3GPP
Release 16 8 3GPP TS 28.554 V16.34.0 (20192020-1203)

6.2 Accessibility KPI

6.2.1 Registered subscribers of network and network slice instance

through AMF
a) AMFRegNbr.

b) This KPI describe the total number of subscribers that are registered to a network slice instance. It is obtained by
counting the subscribers in AMF that are registered to a network slice instance. It is an Interger. The KPI type is
CUM.

c) RSSNSI   RegisteredSubNbrMean
AMF

d) SubNetwork, NetworkSlice

6.2.2 Registered subscribers of network and network slice instance

through UDM
a) UDMRegNbr.

b) This KPI describe the total number of subscribers that are registered to a network slice instance. It is
corresponding to the measurement RM.RegisteredSubUDMNbrMean that counts subscribers registered in UDM.
It is an Interger. The KPI type is CUM.

c) RSSNSI  Re gisteredSubUDMNbrMean
d) SubNetwork, NetworkSlice

6.2.3 Registration success rate of one single network slice instance

a) RegSR.

b) This KPI describes the ratio of the number of successfully performed registration procedures to the number of
attempted registration procedures for the AMF set which related to one single network slice instance and is used
to evaluate accessibility provided by the end-to-end network slice instance and network performance. It is
obtained by successful registration procedures divided by attempted registration procedures. It is a percentage.
The KPI type is RATIO.

c)

 AMF.5GSRegisSucc.Type
Type
RSR  *100%
 AMF.5GSRegisAtt.Type
Type

NOTE: Above measurements with subcounter .Type should be defined in 3GPP TS 24.501 [4].

d) NetworkSlice

6.2.4 DRB Accessibility for UE services

a) DRB Accessibility

b) This KPI describes the DRBs setup success rate, including the success rate for setting up RRC connection and
NG signalling connection. It is obtained as the succeess rate for RRC connection setup multiplied by the success
rate for NG signalling connection setup multiplied by the success rate for DRB setup. The success rate for RRC

3GPP
Release 16 9 3GPP TS 28.554 V16.34.0 (20192020-1203)

connection setup and for NG signalling connection setup shall exclude setups with establishment cause mo-
Signalling [5]. It is a percentage. The KPI type is RATIO.

c) DRBAccessibility 5QI = (∑RRC.ConnEstabSucc.Cause/∑RRC.ConnEstabAtt.Cause) *

(∑UECNTXT.ConnEstabSucc.Cause/∑ UECNTXT.ConnEstabAtt.Cause) *
(DRB.EstabSucc.5QI/DRB.EstabAtt.5QI) * 100

DRB Accessibility SNSSAI = (∑RRC.ConnEstabSucc.Cause/∑RRC.ConnEstabAtt.Cause) *

(∑UECNTXT.ConnEstabSucc.Cause/∑ UECNTXT.ConnEstabAtt.Cause) *
(DRB.EstabSucc.SNSSAI/DRB.EstabAtt.SNSSAI) * 100.

The sum over causes shall exclude the establishment cause mo-Signalling [5].

d) NetworkSlice..

6.2.5 PDU session Establishment success rate of one network slice (S-
NSSAI)
a) PDUSessionEstSR.

b) This KPI describes the ratio of the number of successful PDU session establishment request to the number of
PDU session establishment request attempts for 5G network for the SMF which related to one network slice (S-
NSSAI) and is used to evaluate accessibility provided by the end-to-end network slice and network performance.
It is obtained by the number of successful PDU session requests divided by the number of attempted PDU
session requests. It is a percentage. The KPI type is RATIO.

c)

d) NetworkSlice

6.3 Integrity KPI

6.3.1 Latency of 5G Network

6.3.1.0 KPI categories

a) E2Elatency.

b) This KPI describes the end to end packet transmission latency through the RAN, CN, and TN part of 5G network
and is used to evaluate utilization performance of the end-to-end network. It is the RTT end to end latency of UE
IP packets transmitted from UE to the N6 interface in the 5G network. The N6 interface is the reference point
between UPF and DN. It is a time interval (millisecond). The KPI type is MEAN.

c) NetworkSlice, SubNetwork

6.3.1.1 Downlink latency in gNB-DU

a) DLLat_gNB-DU.

b) This KPI describes the gNB-DU part of the packet transmission latency experienced by an end-user. It is used to
evaluate the gNB latency contribution to the total packet latency. It is the average (arithmetic mean) of the time
from reception of IP packet to gNB-DU until transmission of first part of that packet over the air interface, for a
packet arriving when there is no previous data in queue for transmission to the UE. It is a time interval
(microsecond). The KPI type is MEAN.

c) DLLat_gNB-DU = DRB.RlcSduLatencyDl or optionally DownlinkLat.QoSx = DRB.RlcSduLatencyDl.QoSx

where QOS identifies the target quality of service class.

3GPP
Release 16 10 3GPP TS 28.554 V16.34.0 (20192020-1203)

d) SubNetwork

6.3.1.2 Integrated downlink delay in RAN

a) DLLat_NR.

b) This KPI describes the average packet transmission latency through the RAN part to the UE. It is used to
evaluate latency performance of NG-RAN in downlink. The measurement is optionally split into subcounters per
QoS level (mapped 5QI or QCI in NR option 3). It is the average packets latency transmitted from RAN to UE.
It is a Time interval (microsecond). The KPI type is MEAN.

c) DLLat_NR = DRB.PdcpSduDelayDl + DRB.PdcpF1Delay + DRB.RlcSduDelayDl + DRB.AirIfDelayDl.

d) NetworkSlice, SubNetwork

e) In non-split gNB scenario, the value of DRB.PdcpF1Delay is set to zero for there are no F1-interfaces in this
scenario.

6.3.2 Upstream throughput for network and Network Slice Instance

a) UpstreanThr.

b) This KPI describes the upstream throughput of one single network slice instance by computing the packet size
for each successfully transmitted UL IP packet through the network slice instance during each observing
granularity period and is used to evaluate integrity performance of the end-to-end network slice instance. It is
obtained by upstream throughput provided by N3 interface from NG-RAN to all UPFs which are related to the
single network slice instance. The KPI unit is kbit/s and the KPI type is CUM.

c) UTSNSI   GTP.InDataOctN 3UPF

AMF

d) NetworkSlice, SubNetwork.

6.3.3 Downstream throughput for Single Network Slice Instance

a) DownstreamThr..

b) This KPI describes the downstream throughput of one single network slice instance by computing the packet size
for each successfully transmitted DL IP packet through the network slice instance during each observing
granularity period and is used to evaluate integrity performance of the end-to-end network slice instance. It is
obtained by downstream throughput provided by N3 interface from all UPFs to NG-RAN which are related to
the single network slice instance. The KPI unit is kbit/s and the KPI type is CUM.

c)

d) NetworkSlice.

6.3.4 Upstream Throughput at N3 interface

a) N3UpstreamThr.

b) This KPI describes the total number of octets of all incoming GTP data packets on the N3 interface (measured at
UPF) which have been generated by the GTP-U protocol entity on the N3 interface, during a granularity period.
This KPI is used to evaluate upstream GTP throughput integrity performance at the N3 interface. It is obtained
by measuring the GTP data upstream throughput provided by N3 interface from NG-RAN to UPF, during the
granularity period. The KPI unit is kbit/s and the KPI type is MEAN.

c) UGTPTS=SUM (GTP.InDataOctN3UPF)/timeperiod) at UPF

d) NetworkSlice

3GPP
Release 16 11 3GPP TS 28.554 V16.34.0 (20192020-1203)

6.3.5 Downstream Throughput at N3 interface

a) N3DownstreamThr.

b) This KPI describes the total number of octets of all downstream GTP data packets on the N3 interface
(transmitted downstream from UPF) which have been generated by the GTP-U protocol entity on the N3
interface, during a granularity period. This KPI is used to evaluate integrity performance at N3 interface. It is
obtained by measuring the GTP data downstream throughput provided by N3 interface from UPF to NG-RAN,
during the granularity period. The KPI unit is kbit/s and the KPI type is MEAN.

c) DGTPTS=SUM (GTP.OutDataOctN3UPF)/timeperiod) at UPF

d) NetworkSlice

6.3.6 RAN UE Throughput

6.3.6.1 Definition
a) UEDLThrp, UEULThrp.

b) This KPI that describes how NG-RAN impacts the service quality provided to an end-user.

It is obtained by payload data volume on RLC level per elapsed time unit on the air interface, for transfers
restricted by the air interface. The KPI unit is kbit/s and the KPI type is MEAN.

c) RAN UE Throughput DL = DRB.UEThpDl and

RAN UE Throughput UL = DRB.UEThpUl

or optionally RAN UE Throughput DL for single mapped 5QI or QCI = DRB.UEThpDl.QoS and

RAN UE Throughput UL for single mapped 5QI or QCI = DRB.UEThpUl.QoS

d) SubNetwork

6.3.6.2 Extended definition

To achieve a Throughput measurement (below examples are given for DL) that is independent of file size and gives a
relevant result it is important to remove the volume and time when the resource on the radio interface is not fully
utilized. (Successful transmission, buffer empty in figure 1).

3GPP
Release 16 12 3GPP TS 28.554 V16.34.0 (20192020-1203)

Successful transmission,
buffer not empty
The last slot shall
Failed transmission ( ”Block always be
error ”) ThpTimeDl removed from
calulations since it
Successful transmission,
can be impacted
buffer empty
by packet size of
User Plane (UP) packets.
No transmission, buffer not
empty (e.g. due to contention)

Data arrives to Time (slots)

empty DL buffer

First data is The send buffer is

transmitted to the UE again empty

∑
ThpVolDl = (kbits)

Total DL transferred volume = ∑ + (kbits)

UE Throughput in DL = ThpVolDl / ThpTimeDl (kbits/s)

Figure 1

To achieve a throughput measurement that is independent of bursty traffic pattern, it is important to make sure that idle
gaps between incoming data is not included in the measurements. That shall be done as considering each burst of data
as one sample.

6.4 Utilization KPI

6.4.1 Mean number of PDU sessions of network and network Slice Instance
a) PDUSesMeanNbre.

b) This KPI describes the mean number of PDU sessions that are successfully established in a network slice
instance. It is obtained by successful PDU session establishment procedures of SMFs which is related to the
network slice instance. It is an integer. The KPI type is MEAN.

c) PDUSesMeanNbr=Sum (SM.SessionNbrMean.SNSSAI) over SMFs.

d) NetworkSlice

6.4.2 Virtualised Resource Utilization of Network Slice Instance

a) VirtualResUtilizaiton.

b) This KPI describes utilization of virtualised resource (e.g. processor, memory, disk) that are allocated to a
network slice instance. It is obtained by the usage of virtualised resource (e.g. processor, memory, disk) divided
by the system capacity that allocated to the network slice instance. It is a percentage, The KPI type is Ratio.

NOTE: In the present document, this KPI is for the scenario when NF is not shared between different network
slice instances.

MeanProcessorUsage
c) VRU Processor = *100%
System Capacity Processor

3GPP
Release 16 13 3GPP TS 28.554 V16.34.0 (20192020-1203)

MeanMemoryUsage
VRU Memory = *100%
System Capacity Memory

MeanDiskUsage
VRU Disk = *100%
System Capacity Disk

d) NetworkSlice

6.4.3 PDU session establishment time of network slice

a) PDUEstTime.

b) This KPI describes the time of successful PDU session establishment which related to one single network slice
and is used to evaluate utilization provided by the end-to-end network slice and network performance. It is
obtained by measuring the time between the receipt by SMF from AMF of "
Nsmf_PDUSession_UpdateSMContext Request ", which includes N2 SM information received from (R)AN to
the SMF and the sending of a " Nsmf_PDUSession_CreateSMContext Request or
Nsmf_PDUSession_UpdateSMContext Request PDU Session Establishment Request " message from AMF to
the SMF. It is a time interval (millisecond). The KPI type is MEAN.

c) PDUEstTime = SM.PduSessionTimeMean.SNSSAI

d) NetworkSlice

6.4.4 Mean number of successful periodic registration updates of Single

Network Slice
a) RegUpdMeanNbr.

b) This KPI describes the mean number of successfully periodic registration updates in a network slice at the AMF.
It is obtained by summing successful of periodic registration updates at the AMFs which is related to the
network slice after registration accept by the AMF to the UE that sent the periodic registration update request. It
is an integer. The KPI type is MEAN.

c) RegUpdMeanNbr=Sum (AM.RegNbrMean.SNSSAI) over AMFs

d) NetworkSlice

6.5 Retainability KPI

6.5.1 QoS flow Retainability

6.5.1.1 Definition
a) QoSRetain_R1, QoSRetain_R2.

b) This KPI shows how often an end-user abnormally loses a QoS flow during the time the QoS flow is used. It is
obtained by number of QoS flows with data in a buffer that was abnormally released, normalized with number of
data session time units. The unit of this KPI is “active release / second”. The KPI type is MEAN.

c) To measure QoS flow Retainability for a single QoS level (R1) is fairly straight forward.

However to measure the QoS flow Retainability for UEs is not as straight forward. The measurement R1 is

3GPP
Release 16 14 3GPP TS 28.554 V16.34.0 (20192020-1203)

defined to look at the activity level of just one QoS level at the time, so to use this formula and measurements in
an aggregated way to get QoS flow Retainability on UE level will not be accurate (e.g. for an UE with multiple
QoS flows there might be QoS flows that are active at the same time, hence aggregating the QoS level
measurements for session time will give a larger session time than the total UE session time. See picture below).

UE session time: 10 periods (UE1:4, UE2:6)

QoS=X session time: 7 periods (UE1:4, UE2:3)

QoS=Y session time: 5 periods (UE1:0, UE2:5)

QoS Flow
UE1 QoS=X

One activity period

QoS Flow
QoS=X
UE2
QoS Flow
QoS=Y
[t]

Hence a measurement QoS flow Retainability on UE level is defined (R2) to provide a measurement for the
overall QoS flow Retainability.

d) SubNetwork, NetworkSlice

e) The definition of the service provided by 5GS is QoS flows.

6.5.1.2 Extended definition

The retainability rate is defined as:

Number of abnormally released QoS flow with data in any of the buffers
[Releases/Session time]
Active QoS flow Time

As for defining an abnormal QoS flow release with end-user impact, it shall only be considered an abnormal release of
the QoS flows if the NG-RAN considers there to be data waiting for transfer in any of the buffers.

As for defining a QoS flow as active, a QoS flow shall be considered active if there recently has been any data
transmission in any direction.

6.5.2 DRB Retainability

6.5.2.1 Definition
a) DRBRetain

b) This KPI shows how often an end-user abnormally loses a DRB during the time the DRB is active. It is obtained
by number of DRBs that were abnormally released and that were active at the time of release, normalized with
number of data session time units. The unit of this KPI is "active release / second". The KPI type is MEAN.

3GPP
Release 16 15 3GPP TS 28.554 V16.34.0 (20192020-1203)

c) DRB Retainability for a single mapped 5QI level (R1) and for a single S-NSSAI (R1) are defined as:

and

d) SubNetwork, NetworkSlice

e) The definition of the service provided by 5GS is DRBs.

6.5.2.2 Extended definition

To define (from a DRB Retainability point of view) if a DRB is considered active or not, the DRB can be divided into
two groups:
- For DRBs with bursty flow, a DRB is said to be active if any data (UL or DL) has been transferred during the
last 100 ms.

- For DRBs with continuous flow, the DRB (and the UE) is always seen as being active in the context of this
measurement, and the session time is increased from the first data transmission on the DRB until 100 ms after
the last data transmission on the DRB.

A particular DRB is defined to be of type continuous flow if the mapped 5QI is any of {1, 2, 65, 66}.

6.6 Mobility KPI

6.6.1 NG-RAN handover success rate
a) NRHoSR.

b) A KPI that shows how often a handover within NR-RAN is successful, regardless if the handover was made due
to bad coverage or any other reason. This KPI is obtained by successful handovers to the same or another gNB
divided by attempted handovers to the same or another gNB.

c)MM.HoExeInterSucc, MM.HoExeIntraSucc, MM.HoExeInterReq, MM.HoExeIntraReq, MM.HoPrepInterSucc,

MM.HoPrepIntraSucc, MM.HoPrepInterReq and MM.HoPrepIntraReq.

d) SubNetwork.

6.6.2 Mean Time of Inter-gNB handover Execution of Network Slice

a) -InterGNBHOMeanTime.

b) This KPI describes the time of successful Mean Time of Inter-gNB handover which related to one single
network slice and is used to evaluate utilization provided by the end-to-end network slice and network
performance. This KPI is obtained by measuring the time between the receipt by the Source NG-RAN from the
Target NG-RAN of a “Release Resource" and the sending of a " N2 Path Switch Request " message from Source
NG-RAN to the Target NG-RAN over a granularity period. The unit of this KPI is millisecond.

d)- NetworkSlice

6.6.3 Successful rate of mobility registration updates of Single Network Slice

a) MobilityRegUpdateSR.

3GPP
Release 16 16 3GPP TS 28.554 V16.34.0 (20192020-1203)

b) This KPI describes the successful rate of mobility registration updates in a network slice e at the AMF. This KPI
is obtained by deviding the number of successful mobility registration updates at the AMFs by number of
mobility registration update requests received by the AMFs of single network slice.

c) NetworkSlice

6.7 Energy Efficiency (EE) KPI

6.7.1 NG-RAN data Energy Efficiency (EE)

6.7.1.1 Definition
a) EE_MNDV.

b) A KPI that shows mobile network data energy efficiency in operational NG-RAN. Data Volume (DV) divided
by Energy Consumption (EC) of the considered network elements. The unit of this KPI is bit/J.

c) EEMN,DV

- for non-split gNBs;

- for split-gNBs;

d) SubNetwork

e) The Data Volume (in kbits) is obtained by measuring amount of DL/UL PDCP SDU bits of the considered
network elements over the measurement period. For split-gNBs, the Data Volume is calculated per Interface (F1-
U, Xn-U, X2-U). The Energy Consumption (in kWh) is obtained by measuring the PEE.Energy of the
considered network elements over the same period of time. The samples are aggregated at the NG-RAN node
level. The 3GPP management system responsible for the management of the gNB (single or multiple vendor
gNB) shall be able to collect PEE measurements data from all PNFs in the gNB, in the same way as the other
PM measurements.

3GPP
Release 16 17 3GPP TS 28.554 V16.34.0 (20192020-1203)

Annex A (informative):
Use cases for end to end KPIs

A.1 Use case for end-to-end latency measurements of

5G network-related KPI
The end-to-end latency is an important performance parameter for operating 5G network. In some scenarios (e.g.
uRLLC), if end-to-end latency is insufficient, the 5G network customer cannot obtain guaranteed network performance
provided by the network operator. So it is necessary to define end-to-end latency of network related measurement to
evaluate whether the end-to-end latency that network customer requested has been satisfied. A procedure is invoked by
network management system and is used:

- to update the CSMF/NSMF with the end-to-end latency parameter for monitoring;

- to inform the network customer/network operator the end-to-end latency;

- to make CSMF/NSMF aware if the end-to-end latency can meet network customer’s service requirement.

If high end-to-end latency are measured, it is also of benefit to pinpoint where in the chain from application to UE that
the latency occurs.

A.2 Use case for number of registered subscribers of

single network-slice-instance-related KPI
Number of registered subscribers of single network slice instance can be used to describe the amount of subscribers that
are successfully registered, it can reflect the usage of network slice instance, It is useful to evaluate accessibility
performance provided by one single network slice instance which may trigger the lifecycle management of the network
slice, this kind of KPI is valuable especially when network functions (e.g. AMF) are shared between different network
slice instances. This KPI is focusing on both network and user view.

A.3 Use case for upstream/downstream throughput for

one-single-network-slice-related KPI
Measuring throughput is useful to evaluate system load of end to end network slice. If the throughput of the specific
network slice instance cannot meet the performance requirement, some actions need to be performed to the network
slice instance e.g. reconfiguration, capacity relocation. So it is necessary to define the IP throughput for one single
network slice instance. This KPI is focusing on network and user view.

A.4 Use case for mean PDU sessions number in network

slice instance
It is necessary to evaluate the mean PDU session number in the network slice instance to indicate system load level. For
example, if the mean value of the PDU sessions is high, maybe the system capacity should be increased. This KPI is
focusing on network view.

3GPP
Release 16 18 3GPP TS 28.554 V16.34.0 (20192020-1203)

A.5 Use case for virtualised resource utilization of

network-slice-instance related KPI
It is necessary to evaluate the current utilization of virtualised resources (e.g. memory and storage utilization) that a
network slice instance is occupied. If the utilization is larger or smaller than the threshold, maybe some scale in/out
operations will be made by the management system. This KPI is focusing on network and user view.

A.6 Use case for 5GS registration success rate of one

single-network-slice instance-related KPI
It is necessary to evaluate accessibility performance provided by 5GS. 5GS registration for a UE is important when they
have registered to the network slice instance. If users or subscribers cannot register to the network slice instance, they
cannot access any network services in the network slice instance. This KPI is focusing on network view.

A.7 Use case for RAN UE throughput-related KPI

The UE perceived throughput in NG-RAN is an important performance parameter for operating 5G network. If the UE
throughput of the NR cell cannot meet the performance requirement, some actions need to be performed to the network,
e.g. reconfiguration or capacity increase. So it is necessary to define UE throughput KPI to evaluate whether the end-
users are satisfied. The KPI covers volume large enough to make the throughput measurement relevant, i.e. excluding
data volume of the last or only slot.

The UE throughput KPI covers also "NR option 3" scenarios. Then the gNB is "connected" towards the EPC, and not
towards 5GC.

It is proposed to allow the KPI separated based on mapped 5QI (or for QCI in case of NR option 3).

When network slicing is supported by the NG-RAN, multiple NSIs may be supported. The UL and DL UE throughput
for each NSI is then of importance to the operator to pinpoint a specific performance problem.

A.8 Use case for QoS flow retainability-related KPI

QoS flow is the key and limited resource for 5GS to deliver services. The release of the QoS flow needs to be
monitored. QoS flow retainability is a key performance indicator of how often an end-user abnormally losing a QoS
flow during the time the QoS flow is used. This key performance indicator is of great importance to estimate the end
users’ experiences.

A.9 Use case for DRB accessibility-related KPI

In providing services to end-users, the first step is to get access to the service. First after access to the service has been
performed, the service can be used.

The service provided by NG-RAN is the DRB. For the DRB to be successfully setup it is also necessary to setup an
RRC connection and an NG signalling connection.

If an end user cannot access a service, it is hard to charge for the service. Also, if it happens often that an end-user
cannot access the provided service, the end-user might change wireless subscription provider, i.e. loss of income for the
network operator. Hence, to have a good accessibility of the services is important from a business point of view.

A DRB accessibility KPI requires the following 3 measurements:

- RRC connection setup success rate.

3GPP
Release 16 19 3GPP TS 28.554 V16.34.0 (20192020-1203)

- NG signalling connection setup success rate.

- DRB setup success rate.

The success rate for RRC connection setup and for NG signalling connection setup shall exclude setups with
establishment cause mo-Signalling, since these phases/procedures occur when there is no request to setup a DRB.

This KPI is available per mapped 5QI and per S-NSSAI, and it assists the network operator with information about the
accessibility provided to their 5G network customers.

A.10 Use case for mobility KPI

When a service is used it is important that it is not interrupted or aborted. One of the fault cases in a radio system for
this is handovers/mobility.

If a mobility KPI is not considered OK, then the network operator can investigate which steps that are required to
improve the mobility towards their services.

This KPI can be used for observing the impact of mobility in NG-RAN on end-users.

A.11 Use case for DRB retainability related KPI

DRB is the key and limited resource for 5GS to deliver services. Once a QoS flow reaches a gNB it will trigger setup of
a new DRB or it will be mapped to an existing DRB. The decision on how to map QoS flows into new or existing DRBs
is taken at the CU-CP. CU-CP also defines one set of QoS parameters (one 5QI) for the DRB. If a QoS flow is mapped
to an existing DRB, the packets belonging to that QoS flow are not treated with the 5QI of the QoS flow, but they are
treated with the mapped 5QI of the DRB.

The release of the DRB needs to be monitored, so that abnormal releases while the UE is considered in an active
transfer shall be logged. DRB retainability is a key performance indicator of how often an end-user abnormally loses a
DRB during the time the DRB is actively used. This key performance indicator is of great important to estimate the end
users’ experiences. DRBs with bursty flow are considered active if any data (UL or DL) has been transferred during the
last 100 ms. DRBs with continuous flow are always seen as active DRBs in the context of this measurement. A
particular DRB is defined to be of type continuous flow if the mapped 5QI is any of {1, 2, 65, 66}.

The key performance indicator shall monitor the DRB retainability for each used mapped 5QI value, as well as for the
used S-NSSAI(s). DRBs used in 3GPP option 3 shall not be covered by this KPI. For the case when a DRB have
multiple QoS flows mapped and active, when a QoS flow is released it will not be counted as a DRB release (DRB still
active) in this KPI.

A.12 Use case for PDU session establishment success

rate of one network slice (S-NSSAI) related KPI
It is necessary to evaluate accessibility performance provided by 5GS. PDU session Establishment for a UE is important
when it has registered to the network slice. If users or subscribers cannot establish PDU sessions in slice instance, they
cannot access any network services in the network slice. This KPI is focusing on network view.

A.13 Use case for integrated downlink latency in RAN

Following figure captured in clause 4.2.3, 3GPP TS 23.501[x] illustrates the 5G system architecture. The end to end
downlink latency should be measured from Data Network to UE, of which the latency from RAN to UE is an important
part for the latency of this section is closely related to NG-RAN.

3GPP
Release 16 20 3GPP TS 28.554 V16.34.0 (20192020-1203)

NSSF NEF NRF PCF UDM AF

Nnssf Nnef Nnrf Npcf Nudm Naf

Nausf Namf Nsmf

AUSF AMF SMF SCP

N2 N4

UE (R)AN N3 UPF N6 DN

N9

The integrated downlink latency in RAN is a key performance parameter for evaluating the packet delay in RAN for
QoS monitoring. This KPI is also an important part of the end-to-end network latency for SLA assurance.

A.14 Use case for PDU session Establishment success

rate of one single-network-slice instance-related KPI
It is necessary to evaluate PDU session establishment time, it can be used to analyse the network service difference
between different RAN locations in one area, which can be used for management area division. This KPI is focusing on
network view.

A.15 Use case for QoS flow retainability-related KPI

QoS Flow is the key and limited resource for 5GS to deliver services. The release of the QoS flow needs to be
monitored. QoS Flow drop ratio is a key performance indicator of how often an end-user is abnormally losing a bearer.
This key performance indicator is of great importance to estimate the end users’ experiences.

The KPI shall be available per QoS group.

From QoS perspective it is important to focus also on call duration as in some cases wrong quality perceived by the end
user is not fully reflected by drop ratio nor retainability KPI. Typical case is when due to poor radio conditions the end
user redials (the call was terminated normally) to the same party to secure the quality. But in this case the drop ratio KPI
will not show any degradation. Secondly, although the call is dropped the end user may or may not redial depending on
dropped call duration compared to the case when the call would be normally released. It is therefore highly
recommended to monitor distribution of duration of normally and abnormally released calls.

A.16 Use case for 5G Energy Efficiency (EE) KPI

Assessment of Energy Efficiency in network is very important for operators willing to control their OPEX and, in
particular, their network energy OPEX.

Mobile Network data Energy Efficiency (EEMN,DV) is the ratio between the performance indicator (DVMN) and the
energy consumption (ECMN) when assessed during the same time frame, see ETSI ES 203 228 [8] clause 3.1 and clause
5.3.

where EEMN,DV is expressed in bit/J.

Assessment of EEMN,DV needs the collection of both Data Volumes (DV) and Energy Consumption (EC) of 5G Network
Functions (NF). How this EE KPI can be applied to NG-RAN is specified in clause 4.1 of TS 28.310 [9].

3GPP
Release 16 21 3GPP TS 28.554 V16.34.0 (20192020-1203)

Before the network operator takes any action to save network energy OPEX, the network operator needs to know the
energy efficiency of its 5G network.

This KPI needs to be used for observing the impact of NG-RAN on data energy efficiency of 5G access networks.

3GPP
Release 16 22 3GPP TS 28.554 V16.34.0 (20192020-1203)

Annex B (informative):
Change history
Change history
Date Meeting TDoc CR Rev Cat Subject/Comment New
version
2018-09 SA#81 Upgrade to change control version 15.0.0

2018-09 SA#81 EditHelp fix 15.0.1

2018-12 SA#82 SP-181041 0001 - F Align title with TS database 15.1.0

2019-03 SA#83 SP-190122 0005 2 F Update KPI subscribers of single network slice instance through 15.2.0
UDM
2019-03 SA#83 SP-190122 0011 2 F Update definition of mean number of PDU sessions KPI 15.2.0

2019-03 SA#83 SP-190111 0007 1 B Add KPI of QoS flow Retainability 16.0.0

2019-03 SA#83 SP-190111 0009 1 B Add DRB Accessibility KPI and Use Case 16.0.0

2019-06 SA#84 SP-190371 0013 - B Add KPI for NG-RAN Handover Success Rate 16.1.0

2019-06 SA#84 SP-190375 0015 1 A Correction of Throughput KPI 16.1.0

2019-09 SA#85 SP-190747 0016 2 B Add KPI for DRB Retainability 16.2.0

2019-09 SA#85 SP-190747 0017 3 B Add a new KPI definition of PDU session Establishment Success 16.2.0
Rate of one network slice (S-NSSAI)
2019-09 SA#85 SP-190751 0020 - A Correction on kbits abbreviation 16.2.0

2019-09 SA#85 SP-190747 0021 1 F Correction of Flow Retainability KPI 16.2.0

2019-09 SA#85 SP-190747 0022 1 F Correction of DRB Accessibility KPI 16.2.0

2019-09 SA#85 SP-190748 0024 2 A Correct the title of KPI 16.2.0

2019-09 SA#85 SP-190747 0025 1 B Add definition of integrated downlink latency in RAN 16.2.0

2019-09 SA#85 SP-190747 0028 2 B Add a new KPI definition of Inter-gNB handover Execution time of 16.2.0
one single network slice
2019-09 SA#85 SP-190747 0029 2 B Add a new KPI definition of PDU session Establishment Time of 16.2.0
one single network slice
2019-09 SA#85 SP-190747 0030 2 B Add new specification requirement related to extended 5QI 1 QoS 16.2.0
Flow Retainability monitoring
2019-12 SA#86 SP-191165 0032 1 B Add 5G Energy Efficiency KPI 16.3.0

2019-12 SA#86 SP-191149 0033 1 B Add a new KPI definition of Mean number of successful periodic 16.3.0
registration updates of Single Network Slice
2019-12 SA#86 SP-191149 0034 1 B Add a new description of KPI that related to successful rate of 16.3.0
mobility registration updates of Single Network Slice
2019-12 SA#86 SP-191150 0036 1 F Update the template of KPI definition for TS 28.554 16.3.0

2020-03 SA#87E SP-200163 0038 1 F Update KPI definitions to align with the new template 16.4.0

2020-03 SA#87E SP-200162 0039 - F Correction of equation color 16.4.0

3GPP