NokiaEDU

LTE Optimization Principles [FL16A]

Module 01
RF Measurement Quantities

RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

Index

• RSRP: Reference Signal Received Power

• RSSI: Received Signal Strength Indicator
• RSRQ: Reference Signal Received Quality
• SINR, CINR, SNR

6 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

RSRP

- RSRP is the linear average of received power of RS resource elements

- UE measures the rx power of multiple RS REs and takes average of them
- Reporting range -44…-140 dBm
Definition Reference signal received power (RSRP), is defined as the linear average over the power
contributions (in [W]) of the resource elements that carry cell-specific reference signals within the
• considered measurement frequency bandwidth.
For RSRP determination the cell-specific reference signals R0 according TS 36.211 [3] shall be
R0: tx antenna 1 used. If the UE can reliably detect that R1 is available it may use R1 in addition to R0 to determine
R1: tx antenna 2 RSRP.

The reference point for the RSRP shall be the antenna connector of the UE.

If receiver diversity is in use by the UE, the reported value shall not be lower than the
corresponding RSRP of any of the individual diversity branches.
Applicable for RRC_IDLE intra-frequency,
RRC_IDLE inter-frequency,
RRC_CONNECTED intra-frequency,
RRC_CONNECTED inter-frequency
Note1: The number of resource elements within the considered measurement frequency bandwidth and within the
measurement period that are used by the UE to determine RSRP is left up to the UE implementation with the limitation
that corresponding measurement accuracy requirements have to be fulfilled.

Note 2: The power per resource element is determined from the energy received during the useful part of the symbol,
excluding the CP.

11 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

 RSRP vs CPICH RSCP
• Below a measurement of 2.1GHz CPICH RSCP versus 1.8GHz/20MHz RSRP using a multi-RAT multi-band
scanner.
• Should RSRP be scaled somehow to make it comparable to RSCP?

14 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

Index

• RSRP: Reference Signal Received Power

• RSSI: Received Signal Strength Indicator
• RSRQ: Reference Signal Received Quality
• SINR, CINR, SNR

16 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

RSSI

- RSSI is internal to UE, not reported in uplink to eNB

- RSSI measures all power within the measurement bandwidth
• hence it includes thermal noise, interference and serving cell power
- Measured over those OFDM symbols that contain RS for antenna port R0

Ti

17 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

Relation between RSSI and RSRP

- RSSI = wideband power= noise + serving cell power + interference power

- Without noise and interference, 100% DL PRB activity:

RSSI=12*N*RSRP

Where:
– N is number of PRBs across the RSSI is measured and depends on the BW

- Based on the above, under 100% PRB utilization and high SNR:

RSRP (dBm)= RSSI (dBm) -10*log (12*N)

18 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

Index

• RSRP: Reference Signal Received Power

• RSSI: Received Signal Strength Indicator
• RSRQ: Reference Signal Received Quality
• SINR, CINR, SNR

21 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

RSRQ

RSRQ = N x RSRP / RSSI

Where:
• RSSI is pure wide band power measurement, including serving cell power,
interference and thermal noise
• N: RSSI measurement bandwidth in PRBs

R0: tx antenna 1
R1: tx antenna 2

22 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

Index

• RSRP: Reference Signal Received Power

• RSSI: Received Signal Strength Indicator
• RSRQ: Reference Signal Received Quality
• SINR, CINR, SNR

27 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

SINR / SNR / CINR
Definition and Measurement

SINR = S / (I+N)
• Usually SINR=SNR=CINR unless the receiver is able to separate interference from
thermal noise
• Example: IRC receiver can separate dominant interferer from noise

• The actual measurement definition must be checked for every measurement device
- which signal the ‘S’ measures?
- what is the measurement bandwidth?
- is the measured SNR instantaneous or average value?
- if the average SNR is measured, what is the number of samples averaged?

• Typically UEs measure SNR from RS

• Exact measurement method uncertain, since UE chipset vendors don’t typically
reveal technical data
• Scanners measure SNR from PSS/SSS and/or RS

28 RA47051EN16AGLA0 © Nokia Solutions and Networks 2016

NokiaEDU
LTE Optimization Principles [FL16A] Module 02

KPI Overview

RA47052EN16AGLA0 © Nokia Solutions and Networks 2016

Key Performance Indicators
KPI groups overview

• Accessibility KPIs characterize the availability of a service, or service element to its users
• Retainability (reliability) KPIs characterize Service Drop Rates or to Transport Error Rates
- that is, the ratio between erroneous or lost data units and the overall number of data units sent
• Mobility KPIs build a group of statistics related to hand-over procedures
- They could be also discussed in the Latency and Accessibility groups as well, because HO-
related KPIs are either service interrupt time or success ratio type KPIs
• Integrity KPIs identify the quality and power of the radio transmission and are one of
the most important indicators of network performance
• Usage KPIs give the information on cell resources consumption and throughput

6 RA47052EN16AGLA0 © Nokia Solutions and Networks 2016

Air interface KPIs LTE
• KPIs are categorized to distinguish different performance aspects.
• The names indicate the respective category according to 3GPP32.410 and 3GPP32.450.

Integrity:
Average CQI
Accessibility: Average Latency
RRC Setup Success Rate RLC PDU
E-RAB Setup Success Retransmission
Usage:
Rate
eNodeB1 Cell Availability eNodeB2
S1 Setup Success
Resource Block
Call Setup Success Rate
Usage
IP/PDCP/RLC Layer
Throughput
Inter RAT
X redirection
UE X2

X
Retainability:
Mobility:
RRC Drop Rate
Intra eNodeB HO
E-RAB Drop Rate
Success Rate
Inter eNodeB HO
Success Rate

7 RA47052EN16AGLA0 © Nokia Solutions and Networks 2016

Introduction
List of top level KPIs
Category Key Performance Indicator KP ID
Availability E-UTRAN Cell Availability Ratio LTE_5750a
E-UTRAN Initial E-RAB Accessibility LTE_5060i
E-UTRAN Total E-UTRAN RRC Connection Setup Success Ratio LTE_5218f
E-UTRAN Initial E-RAB Setup Success Ratio LTE_5112c
Accessibility E-UTRAN Data Radio Bearer Setup Success Ratio LTE_5003a
E-UTRAN RRC Paging Discard Ratio LTE_5031b
• These are most E-UTRAN RACH Setup Completion Success Rate LTE_5569a
E-UTRAN E-RAB Setup Success Ratio LTE_5017a
important LTE E-UTRAN E-RAB Drop Ratio, RAN View LTE_5025g
Retainability E-UTRAN Total E-UTRAN RRC Connection Re-establishment Success Ratio LTE_5143a
KPIs per each E-UTRAN E-RAB Retainability Rate, RAN View, RNL Failure with UE Lost LTE_5581b

category E-UTRAN Average PDCP SDU Delay in DL, QCI1-9

E-UTRAN RLC PDU Re-transmission Ratio Uplink
LTE_5471a - LTE_5479a
LTE_5207b
E-UTRAN RLC PDU Re-transmission Ratio Downlink LTE_5208a
E-UTRAN PDCP SDU Loss Ratio in the UL LTE_5310b
Integrity / Quality E-UTRAN PDCP SDU Loss Ratio in the DL LTE_5304b
E-UTRAN Average CQI LTE_5427b
E-UTRAN incoming IP Traffic Error Ratio LTE_5076b
E-UTRAN Average Latency Downlink LTE_5134a
E-UTRAN Average Latency Uplink LTE_5137a
E-UTRAN Total HO Success Ratio, intra eNB LTE_5043a
E-UTRAN Total HO Success Ratio, inter eNB X2 based LTE_5058c
Mobility
E-UTRAN Total HO Success Ratio, inter eNB S1 based LTE_5084b
E-UTRAN Inter RAT Total HO Success Ratio LTE_5198b
E-UTRAN average PDCP Layer Active Cell Throughput DL LTE_5292d
E-UTRAN average PDCP Layer Active Cell Throughput UL LTE_5289d
E-UTRAN Average Active UEs with data in the buffer DL LTE_5800e
E-UTRAN Average Active UEs with data in the buffer UL LTE_5801e
Usage E-UTRAN Averaged IP scheduled Throughput in DL, QCI1-9 LTE_5350a - LTE_5358a
E-UTRAN Averaged IP scheduled Throughput in UL, QCI1-9 LTE_5359a - LTE_5356a
E-UTRAN Average RRC Connected UEs LTE_5242a
E-UTRAN Average PRB usage per TTI UL LTE_5273b
E-UTRAN average PRB usage per TTI DL LTE_5276b

8 RA47052EN16AGLA0 © Nokia Solutions and Networks 2016

Timing Advance Histogram PM-counter

• These performance measurements provide feedback

about the distribution of subscribers in a radial
direction from the antenna based on instantaneous TA
(timing alignment) values d
- TA values are sent to the UE to keep the uplink timing in sync
with the downlink one d

- The TA value is provided by the Random Access response

(initial TA) and all subsequent TA commands
• New counters are stored in new measurement group: d d
- M8029 – LTE MAC
- Measurement interval parameter: PMRNL:mtMAC

d – range (in m) from the

antenna

10 RA47052EN16AGLA0 © Nokia Solutions and Networks 2016

NokiaEDU
LTE Optimization Principles [FL16A] Module 03

Physical RF Optimization

RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Content

- Detecting overshooting cells

• Timing advance counters
- Impact of interference on peak throughput
• PCI collision impact in TD-LTE
- PCI Confusion
- Impact of interference on LTE network performance – importance of physical RF
optimization

34 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

 Overshooting cells

• Cells with too large or largely distributed dominance area.

• Will cause increased interference to other cells
• Can collect excessive amount of traffic.

• How to detect overshooting cells?

• Drive tests
• Analyzing HO performance and neighbor cell measurements from drive test logs.
• Counters
• Cell pair HO analysis with inter site distance information. Indicating HOs to cells with
long inter site distance.
• Timing Advance trace
• Cell Timing Advance trace analyzes to find long distance users.

35 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Overshooting cells analysis from HO counters
Optimizer - Visualize adjacencies and HO KPIs
• To find long distance cells pairs with HO attemps ->
overshooting cell detection.

36 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Overshooting cells analysis from HO counters
Other tools to analyze Neighbor cell HO counters

• M8015 - Neighbor cell HO measurements

• Visualize long distance HO cell pairs (e.g. >10km)

37 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Content

- Detecting overshooting cells

• Timing advance counters
- Impact of interference on peak throughput
• PCI collision impact in TD-LTE
- PCI Confusion
- Impact of interference on LTE network performance – importance of physical RF
optimization

38 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

 LTE1697 Timing Advance Histogram PM-counter
Histogram configuration Each of the histogram PM-counters
• Based on an operator-configured parameter represents a bin for a certain distance range
(LNCEL:expectedCellSize) eNB is applying a pre-
defined set of PM-counter bins (TA-histogram set) #
for mapping the measured instantaneous TA-value
into a corresponding bin
- The value of the parameter is always stored in additional counter
TIMING_ADV_SET_INDEX (M8029C0)
Counter bins
• Configuration of the 'expected cell size' can be
performed
-
-
Manually by the operator or
Automatically during the system upgrade when doing the
18000
2.1 km range
Cell range
16000
database conversion  Thereby a simple configuration
5 km range
rule is applied, which picks a proper expectedCellSize 14000
depending on the configured PRACH cyclic shift (prachCS) 12000
10 km range

Range [m]
and high-speed flag (hsFlag). 15 km range (def)
10000
8000
6000
4000
2000
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
Counter bin

39 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Detecting overshooting cells

• Timing advance counters
- Impact of interference on peak throughput
• PCI collision impact in TD-LTE
- PCI Confusion
- Impact of interference on LTE network performance – importance of
physical RF optimization

45 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

 Impact of PCImod3 Collision on Tput, TD-LTE

• Case: UE at the border of two cells who have the same PCImod3, RSRP from both cells = -
67dBm in both measurement cases (only PCI changed)
• Nokia 7210 TD dongle, 2.6GHz, 10MHz bandwidth

16

14

12

10
tput, Mbps

no PCImod3 collision
8 PCImod3 collision
6

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53
seconds

46 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

PCImod3 Collision Impact, 2.3GHz@20MHz, Qualcomm TD-LTE Dongle
Example

PCI= 88/90 PCI= 87/90 (mod3 collision)

RSRP = -97dBm RSRP=-101dBm
SINR = 12dB SINR=2dB

Tput = ~21Mbps Tput = ~15Mbps

47 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Tools for Parameter Planning

- NetAct Optimizer
• PCI planning
• PRACH planning
• UL DM RS sequence planning is a future feature candidate
- Atoll
• Automatic PCI planning supported
- Asset 7
• PCI planning
- Alpha (Nokia-internal tool)
• PCI planning
• UL DM RS planning
- MUSA (Nokia internal) - post processing
- Daisy (Nokia-internal tool)
• PCI planning

48 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Detecting overshooting cells

- Timing advance counters
- Impact of interference on peak throughput
• PCI collision impact in TD-LTE
- PCI Confusion
- Impact of interference on LTE network performance – importance of
physical RF optimization

49 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

LTE1685 neighbor Relation Robustness
LTE1685-C: Warning for PCI Confusion

• Starting RL50/RL35TD: eNB reports a Warning with fault ‘neighbor cell ambiguity detected’
towards NetAct and BTS Site Manager. PCI collision is solved by manual PCI re-planning.

eNB neighbor cell ambiguity detected! PCI Collision!

• CGI of the cell which detected the collision
• Cell A
• PCI and frequency which are not unique
Cell A • PCI1, Freq1
ReportCGI: ECGI-1
PCI-1, Freq1 • List of CGIs for which the collision was detected
PCI-1, Freq1
• CGIs of CellA and NbCell1

ECGI1
PCI-1, Freq1

NbeNB1

50 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

LTE1685 neighbor Relation Robustness
Alarm
Fault name PCI confusion detected
Fault ID and name 6279: EFaultId_PciConfusionDetectedAl
Reported alarms 7655 CELL NOTIFICATION
Unit status Working
LED display Blinking yellow
Based on CGI measurements it has been detected that two or more cells operating on the
same carrier frequency use the same PCI
Alarm contains:
Meaning 1. CGI of the cell which detected the PCI confusion.
2. The CGI of the NR which is switched from available to invalid
3. The CGI of the NR which is switched from invalid to available
Unit actions after fault None
detect
The neighbor relation states 'invalid' and 'available' are exchanged between two LNREL
Detection method instances of the same cell (pointing to the same PCI and frequency) after reception of a CGI
start/transient measurement report.
Detection method cancel None
Effect Possibly increased handover failure rate for the affected cell.
Check for E-UTRAN cells that operate on the same frequency layer and use the same PCI and
Instructions that are in close vicinity. Re-assign PCIs for the discovered cells.

53 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Detecting overshooting cells

• Timing advance counters
- Impact of interference on peak throughput
• PCI collision impact in TD-LTE
- PCI Confusion
- Impact of interference on LTE network performance – importance
of physical RF optimization

54 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

RF Optimization

• Basic physical RF optimization is very important (of course..)

• Clear cell dominance areas, minimize cell overlapping
• Avoid sites shooting over large areas with other cells
• “Can’t fix bad RF by tuning parameters”

• Antenna tilting and antenna placement has big impact on other cell interference!!
• What is the impact on network performance?

55 RA47053EN16AGLA0 © Nokia Solutions and Networks 2016

NokiaEDU
LTE Optimization Principles [FL16A] Module 04

Throughput Optimization

RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

5 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Traffic throughput

PDCP layer throughput matches closest to application layer throughput.

- Calculated with 1 second averaging => typically KPIs are showing lower throughput
than experienced by users.
- Averaging results over several cells further impacts on the values, as cells with no
traffic are counted in with zero throughput.

• PDCP mean DL - LTE_5292d Average PDCP Layer Active Cell Throughput DL, kbps
• PDCP mean UL - LTE_5289d Average PDCP Layer Active Cell Throughput UL, kbps
• PDCP peak DL - LTE_291a Maximum PDCP Throughput DL, Mbps
• PDCP peak UL - LTE_288a Maximum PDCP Throughput UL, Mbps

Example values only!!

Network average from different networks (A, B, C,…)

KPI A B C …. X Y Z

AV PDCP active cell 21206 20452 20067 6994 5190 5055

DL/Cell (LTE_5292) kbits kbits ….
kbits kbits kbits kbits
AV PDCP active cell
UL/Cell (LTE_5289) 3754 3054 2436 511 502
kbits kbits …. 495 kbits
kbits kbits kbits

6 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL proactive scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

12 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
 Optimizing 2x2 Dynamic Adaptive MIMO Switching

SM SM
Upgrade Switch :
CQI If
mimoCQI > mimoSmCqiThUpOL
mimoOlCqiThU and
mimoRANK > mimoSmRiThUpOL
mimoOlCqiThD

RI Downgrade Switch::
If
mimoOlRiThU mimoCQI <= mimoDivCqiThDownOL
or
mimoOlRiThD mimoRANK <= mimoDivRiThDownOL
Time

Optimization case example:

Parameter Old New Post opt.
Pre opt.
mimoOlCqiThD 9 7
mimoOlCqiThU 11 8 12%
mimoOlRiThD 1.4 1.2
mimoOlRiThU 1.6 1.4 28%

MIMO parameter proposal has 72% 88%

resulted in DL throughput
average improvement in the
range of 2-3 Mbps (38Mbps ->
41 Mbps). SM TxDiv SM TxDiv

13 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

19 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Traffic Shaping not Activated in Transport Network

• Traffic shaping should be activated whenever there is a transition from high-speed link to a lower-speed
link (e.g. Flexi Packet Radio, Huawei NG-SDH boxes, etc etc)

Flexi Packet Hub (A2200) @ PRGOPA site: config-set interface ethernet 1/1 shaper-rate 280.0 -> 1000.0

Shaping on* Shaping off

*throughput fluctuations seem to be driven mainly by RF conditions

20 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
 Traffic Shaping not Activated in Transport Network

- Whenever feeding a low bandwidth box with high bandwidth input, check shaping in the
transport chain
• Especially if the tx buffer in the low bw box is small packet loss can occur  reduces TCP
throughput!

Example: A-2200 feeding FlexiPacket

315Mbps
A2200 1Gbps A2200/A1200
FPR
FPR
ODU
ODU

Packet loss here due to too small tx

Activate shaping buffer relative to data burstiness
here to reduce Data transfer direction
burstiness
Generalization:

thick pipe thin pipe

Box X Box Y Box Z

Be alert for packet loss here !

21 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

22 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
DL performance example features

Other features affecting DL performance

FL16
RL60
• DL SU 4x4 MIMO
• DL CA 40 MHz
• CA enhancements
• Inter frequency load balancing
• Liquid cell
RL70
• eICIC
FL16A
• DL interference shaping
• 256 QAM in DL
• DL CA 3 CC – 40 MHz
• CA-aware Idle Mode Load Balancing
• Inter frequency Load equalization
• RSRQ-based A5
• FDD DL CA 4CC
FL15A
• DL CA 3 CC – 60 MHz
• Flexible SCell Selection
• eICIC for HetNet eNodeB Configurations
• eICIC Enhancements - micro
• Additional CA Band Combinations – II
• Inter-eNodeB CA for 2 Macro eNodeBs

23 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

24 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Uplink performance – balance between user vs. cell performance

Optimization goals:
• Maximize peak throughput
• Maximize average cell throughput
• Minimize inter cell interference while maintaining cell edge throughput
• Minimize scheduling delay

Related features:
• UL power control setting
- Closed loop: SINR and RSSI target window. IAWPC
- Open loop: P0 and alpha factor.
• UL link adaptation
- LA algorithm: MCS vs. allocated PRBs.
• UL scheduler
- Proactive UL scheduling
- UL scheduler FD type: RoundRobing/Exhaustive/Mixed
- Scheduling method of the UL scheduler: CUS/IAS/CAS

25 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
• How to troubleshoot packet loss related TCP tput problems?

26 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Simulation Setup UL

Name Configuration parameters

Cluster 3GPP ISD500m
7 sites / 21 cells – wrap-around
Operation band FDD 2000 MHz
Frequency bandwidth 10MHz (50PRB)
Antenna 3GPP 14 dBi
Tx Power max 22.5 dBm
Penetration loss 20dB (outdoor to indoor)
No. of UEs 50 UEs per cell, randomly distributed
Mobility 3km/h speed
Traffic model FTP traffic: 10 MB file
Simulation time 200 s
Slow Fading On, 8 dB sigma
Fast Fading On (Typical urban)
UL Power control See slide no 2.

Scheduler TDS: PF (Proportional Fair)

FDS: Channel aware + Round Robin
MCS QPSK/16QAM (MCS0 to MCS24)
System load 10% / 25% / 50% / 100%

• load realization 10% 25% 50% 100%

• limitation of max # users scheduled per TTI 1 2 5 -
• limitation of max # PRBs per user 5 6 5 -
• total # PRBs per cell 5 12 25 50
• max # avail PRBs: 50

27 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
LTE1336: Interference aware UL power control

Interference in LTE UL plays critical role determining the throughput

UL Power control illustration

UL interference comes from neighboring cell UEs
eNodeB1 eNodeB2
- It cannot be predicted since we do not know how the UEs
will be distributed UL interference

- Contributions to interference are most prominent from

cell-border UEs
As a result the role of PC is to provide the required SINR
while controlling at the same time the interference caused to
neighboring cells
Cell- center
Looking at the UL SINR formula… Affected UE
Cell-edge
UE
UE
Useful signal is already
close to maximum
S S
SINR = =
I+N Σ(I1,I2, … In) Noise N is negligible in
interference limited Minor UL
scenarios: I >> N interference
Cell edge users inject as much to eNodeB1
interference to neighboring eNB as to
their serving eNB. Therefore
… it is clearly visible that it’s better to minimize the contributions to interference in
adjacent cells are most prominent from
interference than improve the useful signal if we want to get cell-border UEs
better SINR

31 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

36 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
UL Proactive scheduling

Benefits:
• Improves RTT by decreasing UL scheduling delay

Drawbacks:
• Increases UL interference
• Increases PDCCH usage

The feature is improves end-user performance in low loaded networks. When UL load
increases, feature settings should be optimized.

37 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Proactive scheduling – case example

- When UL Proactive Scheduling is turned off UL App. Throughput per RB is increased 67%
(drive testing at around -95dBm)
• UL Interference level is reduced 5.8dB(5.3%) and PUSCH SINR is increased 5.0dB
compared to Proactive scheduling with 50ms data.

- PUSCH RSSI is sustained but SINR is increased due to the interference

reduction : 16dB 19.9dB, 24.4%

- UL PRB utilization : 1310.7, 17.7 %

- PUSCH BLER is decreased : 8.04%3.18%, 63%

40 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

43 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
UL performance features
Features to improve UL performance in RL50:
- 4RX UL Diversity with MRC.
- 4RX UL Diversity with IRC.

• Features to improve UL performance in RL60:

- Fast ULA

• Features to improve UL performance in RL70

- UL intra eNB CoMP 4Rx
- eICIC (ABS means no UL scheduling in the eICIC active cell)

• Features to improve UL performance in FL16

- Volte uplink coverage boosting
- Configurable UL interference regions
- 64 QAM in UL
- UL Multi cluster Scheduling

• Features to improve UL performance in FL16A

- Uplink Carrier Aggregation - 2CC
- CA-aware Idle Mode Load Balancing

44 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

47 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
DL APN-AMBR Limits DL Throughput

• Example, measurement with Samsung, 20MHz, DX HLR M14.6

• APN-AMBR DL = 20Mbps in HSS, SAE-GW APN-AMBR shaping on, FTP download
# fsclish -c "show config fsClusterId=ClusterRoot
fsFragmentId=FlexiNG fsFragmentId=Internal" |grep Qos
GPRS AND LTE SUBSCRIBER DATA HANDLING COMMAND <MN_>
<
fngQosShaping: enable
fngQosPolicy: enable
MAIN LEVEL COMMAND <___>
< MNI:IMSI=244070109100221;

LOADING PROGRAM VERSION 7.9-2

PDN CONTEXT PARAMETERS

20Mbps
PHY tput, CINR
limit on
IMSI ........................
CTXID .......................
244070109100221
10 30
tput
PDN TYPE .................... BOTH
AP NAME ..................... test-ap-2
QOS CLASS IDENTIFIER ........ 6 25
PRIORITY LEVEL .............. 2
PRE-EMPTION CAPABILITY ...... DISABLED
PHY tput Megabits/sec, CINR dB

PRE-EMPTION VULNERABILITY ... DISABLED 20

VPLMN DYNAMIC ADDRE ALLOWED . YES
PGW ALLOCATION TYPE ......... DYNAMIC Data
CHARGING CHARACTERISTICS .... NORM 15 Average of Phy DL TP(Mbps)
AMBR DOWNLINK ............... 20000 Sum of SCell-CINR(Com)

AMBR UPLINK ................. 65535

10

COMMAND EXECUTED

HSS APN-AMBR
limitation 0
06/11/2010 11:31:37.393 06/11/2010 11:31:47.395 06/11/2010 11:31:57.894 06/11/2010 11:32:07.895

Time

48 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
UL APN-AMBR Limits UL Throughput

• Latest FlexiNS sw implements signaling of UL APN-AMBR from MME to UE.

• APN-AMBR parameter stored in HSS default APN profile
• When UE receives this IE in bearer setup, it shapes the UL bit rate to the UL APN-
AMBR value for non-GBR bearers (per-APN)
• Strong degrading impact on UL FTP throughput possible because UE shaping
implementation is not specified in detail by 3GPP
• Problem symptom noticed in a project: UL tput collapses after sw upgrade in MME…
- Fix: change APN-AMBR parameters in HSS

49 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP throughput measurement to check that radio and e2e
network do not limit peak throughput
- How to troubleshoot packet loss related TCP tput problems?

52 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Troubleshooting Checklist – Executive Summary!

1. Check with UDP download

• If UDP tput not ok  problem in radio parameters or maybe interference.
- Use scanner to check that there are no neighbors within >30dB of serving cell.
Do not trust RS CINR measurement!! Serving cell RSRP should be >-80dBm for
peak tput.
- Check HSS profile that there is no APN-AMBR or UE-AMBR limitation
• If UDP tput ok  problem is either i) packet loss in transport network, or ii) TCP
settings of FTP server and/or UE.
- Change to Linux FTP server and optimize UE TCP settings.
- Check that you have no packet loss in transport network.

53 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
UDP Testing – Motivation

• in >50Mbps regime FTP throughput suffers greatly from:

• Packet loss, even very little packet loss!!
• Too small TCP RX and TX window
- BTS parameters can also have misconfiguration
- If FTP throughput is bad, it is often difficult to say if this is caused by radio, transport
or TCP setting problem
- Solution: test with UDP!
- If UDP throughput is also bad  then it’s most probably a radio problem
• Either interference or parameters are wrong
• Could also be a problem with HSS profile
- If UDP throughput is good  then it’s transport or TCP problem
• Troubleshooting can be narrowed down

54 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
UDP versus TCP, Same Drive Route For Both

•
74Mbps average

50Mbps average

If you must use FTP, then having multiple

parallel FTP sessions may help.

55 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Please don’t forget to check radio interface DL buffer utilization: Qualcomm
measurement example

UE only scheduled about 600-700

subframes a second, non-full
eNB transmit buffer!! Problem
with spurious TCP tx timeouts
and/or packet loss??

56 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Index

Throughput optimization
- KPIs
- Downlink
• Impact of inter-cell interference on peak throughput
• Improving MIMO performance
• Packet drops in transport/core network
• Other DL issues
- Uplink
• User throughput vs. cell throughput
• UL power control
• UL scheduling
• Other UL issues
- QoS parameters limiting throughput
Throughput testing
- UDP tput measurement to check that radio and e2e network do not
limit peak tput
- How to troubleshoot packet loss related TCP tput problems?

57 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
How to check if the low FTP tput is caused by TCP?

1. Try with iPerf UDP server download and upload

– If UDP speed is ok, then this is the benchmark against which TCP should be
optimized
– iPerf UDP server typically used
2. Test with FTP transfer
– FTP should reach transfer speed of within a few percents of UDP speed if properly
optimized
– If the FTP download speed is much less than the UDP download, then there is most
likely a problem with TCP settings or there is packet loss in the transport network
somewhere

58 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
FTP tput problem - root causes

- There are three typical root causes for bad FTP tput
1. Windows FTP server is used in downloads (tx window too controlled)
2. UE laptop RX window too small
3. Packet loss in the e2e TCP path
- The first two problems are fixed by using Linux FTP server and optimizing laptop TCP
settings
- The third problem is:
• How to detect if packet loss is the reason for bad FTP tput?
• How to find where packet loss takes place?
• Rationale: Customer frequently needs to be convinced of the packet loss and/or
there’s a need to pinpoint which trs box causes it

59 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
TCP troubleshooting, how to find where packet loss takes place?

Analysis steps
1. Take traces from several observation points simultaneously from the same FTP
transfer. For example:
 FTP server
 EPC site router
 eNB
 UE
 Transport network (output of that suspicious box..)
2. Use tcptrace (or some other tool) to analyze the tcp flow to find out where the packet
loss happens
 Often the place of packet loss can be just guessed…

60 RA47054EN16AGLA0 © Nokia Solutions and Networks

2016
Basic Signalling

NokiaEDU

• Backup Slides

RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Index

- Handover signaling
• Measurement configuration
• Intra-LTE handover
• LTEà3G ISHO

40 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Measurement configuration investigation

Using the Nemo Log 1 RF exercise log

For the measurement report below, identify the,
frequency,
RAT type
Measurement type
What action is taken after the report is received and why?

RRC SIGNALING MESSAGE

Time: 10:01:09:704
MeasurementReport (3GPP TS 36.331 ver 10.6.0 Rel 10)
UL-DCCH-Message
message
c1
measurementReport
criticalExtensions c1
measurementReport-r8
measResults
measId : 1
measResultPCell
rsrpResult :42
RsrqResult : 12
measResultNeighCells
measResultListEUTRA
measResultListEUTRA value 1
physCellId : 408
measResult
rsrpResult : 52

41 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Index

- Handover signaling
• Measurement configuration
• Intra-LTE handover
• LTEà3G ISHO
- TDD LTE specific signaling.

42 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Inter-eNB handover, more complete (3GPP 36.300)
UE Source eNB Target eNB MME Serving Gateway

0. Area Restriction Provided

1. Measurement Control

packet data packet data

UL allocation Legend

L3 signalling
2. Measurement Reports
L1/L2 signalling
3. HO decision
User Data
4. Handover Request

Handover Preparation
5. Admission Control
6. Handover Request Ack
DL allocation
RRC Conn. Reconf. incl.
7.
mobilityControlinformation
Detach from old cell
Deliver buffered and in transit
and
packets to target eNB
synchronize to new cell

Handover Execution
8. SN Status Transfer

Data Forwarding

Buffer packets from

Source eNB
9. Synchronisation

10. UL allocation + TA for UE

11. RRC Conn. Reconf. Complete

packet data
packet data
12. Path Switch Request
13. User Plane update
request

End Marker
14. Switch DL path

Handover Completion
packet data
End Marker
15.User Plane update
response
16.Path Switch Request Ack

17. UE Context Release

18. Release
Resources

43 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Handover Preparation, step 1/3

1. The source eNB configures the UE UE Source Target MME GW

measurement procedures with an RRC
RECONFIGURATION
2. UE is triggered to send MEASUREMENT
REPORT to the source eNB. It can be event 1. RRC Reconfiguration
triggered or periodic
RRC Reconfiguration Complete
3. Source eNB makes handover decision based
on UE report + load and service 2. Measurement report
information
4. The source eNB issues a HANDOVER 3. HO
REQUEST to the target eNB decision
5. Target eNB performs admission control
4. HO request
6. Target eNB sends the HANDOVER REQUEST
ACKNOWLEDGE to the source eNB
5. Admission
control

6. HO request ack.

44 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Handover Execution, step 2/3

7. Source eNB generates the RRC UE Source Target MME GW

RECONFIGURATION towards UE informing it
to move to the target cell
Source eNB starts forwarding packets to
target eNB
7. RRC Reconfiguration
8. Source eNB sends status information to
target eNB Forward
packets to
9. UE performs the final synchronization to target
target eNB and accesses the cell via RACH 8. Status transfer
procedure
DL pre-synchronization is obtained during Buffer
packets
cell identification and measurements from
10. Target eNB gives the uplink allocation and source
9. Synchronization
timing advance information
11. UE sends RRC RECONFIGURATION 10. UL allocation and timing advance
COMPLETE to target eNB
Target eNB can begin to send data to UE 11. RRC Reconf. Complete

45 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Handover Completion, step 3/3
12.Target eNB sends a PATH SWITCH message to UE Source Target MME GW
MME to inform that the UE has changed cell
13.MME sends a USER PLANE UPDATE REQUEST
message to Serving Gateway
14.Serving Gateway switches the downlink data path 12. Path switch request
to the target side
15.Serving Gateway sends a USER PLANE UPDATE 13. User plane update
RESPONSE message to MME request
14.
16.MME confirms the PATH SWITCH message with the Switch
PATH SWITCH ACK message downlink
path
17.By sending RELEASE RESOURCE the target eNB
informs success of handover to source eNB and 15. User plane
triggers the release of resources update response
18.Upon reception of the RELEASE RESOURCE
message, the source eNB can release radio and C- 16. Path switch request ack.
plane related resources associated to the UE 17. Release resources
context
18.
Release
resources

46 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
 Handover command
• RRCConnectionReconfiguration with MobilityControlInfo (”Handover command”)
HandoverCommand : {
criticalExtensions c1 : handoverCommand-r8 : {
handoverCommandMessage {
message c1 : rrcConnectionReconfiguration : {
rrc-TransactionIdentifier 1,
criticalExtensions c1 : rrcConnectionReconfiguration-r8 : {
C-RNTI in the target
mobilityControlInfo {
targetPhysCellId 343, cell
t304 ms100,
newUE-Identity '11111001 10011010'B,
radioResourceConfigCommon {
rach-ConfigCommon {
preambleInfo {
numberOfRA-Preambles n40,
preamblesGroupAConfig {
sizeOfRA-PreamblesGroupA n4,
messageSizeGroupA b144,
messagePowerOffsetGroupB dB10
}
},
powerRampingParameters {
powerRampingStep dB4,
preambleInitialReceivedTargetPower dBm-90
},
ra-SupervisionInfo {
preambleTransMax n8,
ra-ResponseWindowSize sf10,
mac-ContentionResolutionTimer sf64

47 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Index

- Handover signaling
• Measurement configuration
• Intra-LTE handover
• LTEà3G ISHO

48 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Inter-RAT Handover to WCDMA Flow

MME
LTE, Source:CelI WCDMA, Target Cell

Measurement Report Triggered EventA2

RRC Connection Reconfiguration New MeasConfig used for IRAT Handover

RRC Connection Reconfiguration Complete

Measurement Report Triggered EventB2

Handover Preparation

Handover Preparation Successful

MobilityFromEUTRAcommand
Handover Command

HandoverToUTRANComplete

49 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Handover Preparation

LAC:2031(Hex) = 8241 Dec

TargetCellId: 6291758 = 60 01 2E (Hex)

=> 60Hex is RNC-id = 96 Dec
=> 01 1E Hex is Cell-id = 302 Dec

50 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
Handover Preparation Success

51 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
UEContextRelease

52 RA47055EN16AGLA0 © Nokia Solutions and Networks

2016
NokiaEDU
LTE Optimization Principles [FL16A] Module 06

Call Setup Optimization

RA47056EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- KPI Reference Values
- Call Setup Signaling + Counter Triggers (S1)
- Project Example

5
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
The Most Important Call Setup KPIs –
RRC Connetion Setup Success Ratio

LTE_5218f Total E-UTRAN RRC Connection Setup Success Ratio

• Formula: (Logical)
RRC Con SSR=(RRC connection setup completions / RRC connection requests)*100%

• Formula: (NetAct names)

100*sum([SIGN_CONN_ESTAB_COMP]) / sum([SIGN_CONN_ESTAB_ATT_MO_S] +
[SIGN_CONN_ESTAB_ATT_MT] + [SIGN_CONN_ESTAB_ATT_MO_D] +
[SIGN_CONN_ESTAB_ATT_DEL_TOL] + [SIGN_CONN_ESTAB_ATT_HIPRIO] +
[SIGN_CONN_ESTAB_ATT_EMG])

• Description: This KPI shows the setup success ratio for the RRC connection
establishment elementary procedure, used to set up a radio connection from a UE to an
eNB (involves SRB1 establishment).

6
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
Top Level KPIs – Category: Accessibility

LTE_5112c E-UTRAN Initial E-RAB Setup Success Ratio

• Formula: (Logical)
E-RAB ISSR=(intial E-RAB setup successes / initial E-RAB setup attempts)*100%

• Formula: (NetAct names)

100* sum([ERAB_INI_SETUP_SUCC_QCI1]+ [ERAB_INI_SETUP_SUCC_QCI2] + [ERAB_INI_SETUP_SUCC_QCI3] + [ERAB_INI_SETUP_SUCC_QCI4] +
[ERAB_INI_SETUP_SUCC_QCI5] + [ERAB_INI_SETUP_SUCC_QCI6] + [ERAB_INI_SETUP_SUCC_QCI7] + [ERAB_INI_SETUP_SUCC_QCI8] +
[ERAB_INI_SETUP_SUCC_QCI9]) /sum([ERAB_INI_SETUP_ATT_QCI1]+ [ERAB_INI_SETUP_ATT_QCI2] + [ERAB_INI_SETUP_ATT_QCI3] +
[ERAB_INI_SETUP_ATT_QCI4] + [ERAB_INI_SETUP_ATT_QCI5] + [ERAB_INI_SETUP_ATT_QCI6] + [ERAB_INI_SETUP_ATT_QCI7] +
[ERAB_INI_SETUP_ATT_QCI8] + [ERAB_INI_SETUP_ATT_QCI9])

• Description:
The KPI indicates the setup success ratio for the elementary initial E-RAB setup
procedure. It indicates the E-UTRAN's contribution to network accessibility for the end
user, not the whole end-to-end service accessibility.

Note that the probability to access a service is covered in accessibility.

7
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
Index

• Network + Field KPIs

- KPI Reference Values
- Call Setup Signaling + Counter Triggers (S1)
- Project Example

18
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
KPI Reference Values

Example values only!!

Network average from different networks (A, B, C,…)

KPI A B C D E

RRC Setup
Success Rate 99.98 99,96 99,96 99.94 99.92
[%] (LTE_5218f)

eRAB Setup
Success Ratio
99,98 99,98 99,97 99.97 99.96
[%]
(LTE_5017a)

19
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
Index

• Network + Field KPIs

- KPI Reference Values
- Call Setup Signaling + Counter Triggers (S1)
- Project Example

20
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
Connection Setup

The call setup has five phases

1. PRACH phase (also applies to handover)
2. SRB1 setup
3. NAS security setup
4. RRC security + UE Capability Enquiry (UE Cap Enquiry optional)
5. SRB2 + DRB setup

- Success of phases 1-2 must be monitored from PRACH and RRC signaling counters
- Success of phases 3-5 is monitored from S1AP and E-RAB counters  ”S1 view”

22
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
 S1 Context Setup, Successful
eNB MME
RRC Connection
Setup
M80013C43 S1_SIGN_CONN_ESTAB_ATT_UE
1 S1AP Init UE Message (NAS)

NAS security

M8006C0 EPS_BEARER_SETUP_ATTEMPTS
2 S1AP Init Ctxt Setup Req (NAS) M8013C45 UE_CTX_SETUP_ATT

RRC security

SRB2 + DRB
setup

3 S1AP Init Ctxt Setup Response

M8006C1 EPS_BEARER_SETUP_COMPLETIONS
M8013C47 UE_CTX_SETUP_SUCC

Analysis:
• LTE_5009b S1 Initial Context Setup Success Ratio
• LTE_5017a E-RAB Setup Success Ratio

23
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
 S1 Context Setup, Unsuccessful 1
eNB MME
RRC Connection
Setup

M80013C43 S1_SIGN_CONN_ESTAB_ATT_UE
1 S1AP Init UE Message (NAS)

NAS security

M8006C0 EPS_BEARER_SETUP_ATTEMPTS
2 S1AP Init Ctxt Setup Req (NAS) M8013C45 UE_CTX_SETUP_ATT

RRC security

SRB2 + DRB
setup M8013C49 UE_CTX_SETUP_FAIL_RADIO_INT
failure
3 S1AP Init Ctxt Setup Failure

Analysis:
• LTE_5522a E-UTRAN Initial Context Setup Failure Ratio due to Failed Radio Interface Procedure

24
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
 S1 Context Setup, Unsuccessful 2
eNB MME
RRC Connection
Setup

M80013C43 S1_SIGN_CONN_ESTAB_ATT_UE
1 S1AP Init UE Message (NAS)

eNB receives no response from MME for time period defined

by tEstS1Con  S1 + SRB1 released
• non-modifiable parameter <p name="tEstS1Con">16000</p>
• unit: milliseconds
• timer stopped when any subsequent S1AP message for this UE is received
• no eNB counter triggered if timer expires

Analysis:
• no counter available for the number of S1 connection setup timer expires

25
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
S1 Context Setup, Unsuccessful 3

eNB MME

RRC Connection
Setup
M80013C43 S1_SIGN_CONN_ESTAB_ATT_UE
1 S1AP Init UE Message (NAS)

NAS security
Call drops at air interface

Analysis:
• no eNB counters available for RRC layer success of NAS security messaging.

26
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
S1 Context Setup, Unsuccessful, Cell Trace Example

UE timer T3417 (fixed,

5seconds) expires after
receiving no response to
Service Request  UE
silently disappears.

tEstS1Con timer expires

after 16 sec  RRC release

27
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
Index

• Network + Field KPIs

- KPI Reference Values
- Call Setup Signaling + Counter Triggers (S1)
- Project Example

61
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
Operator A Example: RRC Performance

E-UTRAN RRC Signalling

LTE_753a RRC Connection Establishment attempts (#) LTE_5218c RRC Connection Setup Success Ratio (%) LTE_5220a RRC Success Ratio (%)

100.00 4 500 000

90.00 4 000 000

80.00
3 500 000
RRC completion success
70.00
degraded due to initial 3 000 000
context setup failures
60.00
2 500 000

50.00

2 000 000
40.00

1 500 000
30.00

1 000 000
20.00

10.00 500 000

0.00 0

62
RA47056EN16AGLA0 © Nokia Solutions and Networks 2016
NokiaEDU
LTE Optimization Principles [RL16A] Module 07

Call Drop Analysis

RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network KPIs
- KPI Reference Values
- Drop Call Causes, Theory
• UE initiated drop
• eNB initiated drop
- Project Example

This module discusses call drops and handover drops after

call setup has already succeeded. See separate module for
call setup problems.

5 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Top Level KPIs – Category: Retainability

LTE_5025g E-UTRAN E-RAB Drop Ratio, RAN View

• Formula: (Logical)
E-RAB DR RAN = (abnormal E-RAB releases from RAN point of view / all E-RAB releases)
* 100%

• Formula: (NetAct names)

100*sum([ERAB_REL_HO_PART] + [ERAB_REL_ENB] - [ERAB_REL_ENB_RNL_INA] - [ERAB_REL_ENB_RNL_RED] - [ERAB_REL_ENB_RNL_RRNA]
- [ERAB_REL_TEMP_QCI1])
/
sum([ERAB_REL_ENB] + [ERAB_REL_HO_PART] + [EPC_EPS_BEARER_REL_REQ_NORM] + [EPC_EPS_BEARER_REL_REQ_DETACH] +
[EPC_EPS_BEARER_REL_REQ_RNL] + [EPC_EPS_BEARER_REL_REQ_OTH] + [ERAB_REL_EPC_PATH_SWITCH] - [ERAB_REL_TEMP_QCI1])

• Description:
This KPI describes the ratio of abnormally released (dropped) E-RABs from RAN
point of view.

8 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network KPIs
- KPI Reference Values
- Drop Call Causes, Theory
• UE initiated drop
• eNB initiated drop
- Project Example

This module discusses call drops and handover drops after

call setup has already succeeded. See separate module for
call setup problems.

13 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

KPI Reference Values

Example values only!!

Network average from different networks (A, B, C,…)

KPI A B C …. X Y Z

eRAB Drop Ratio [%]

0.03 0.05 0.07 .… 1.4 1.66 4.27
(LTE_5025)
eRAB Drops per
PDCP Vol. [#/GB] 0.04 0.17 0.26 …. 87.08 301.31 820.28
(LTE_5812)

14 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network KPIs
- KPI Reference Values
- Drop Call Causes, Theory
• UE initiated drop
• eNB initiated drop
- Project Example

This module discusses call drops and handover drops after

call setup has already succeeded. See separate module for
call setup problems.

15 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

 RRC Re-Establishment Procedure, Successful

UE EUTRAN

RRC Re-Est triggered. UE makes cell reselection and sends Re-Est.

RRCConnectionReestablishmentRequest

RRCConnectionReestablishment

RRCConnectionReestablishmentComplete

• RRC Conn Re-Est can only succeed in a cell that has:

• short MAC-I of the source cell
• PCI of the source cell
• C-RNTI in the source cell

16 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Example: RRC Connection Reestablishment Request

RRC SIGNALING MESSAGE

Time: 9:38:13.165
RRCConnectionReestablishmentRequest (3GPP TS 36.331 ver 8.7.0 Rel 8)
UL-CCCH-Message
message
c1
rrcConnectionReestablishmentRequest
criticalExtensions
rrcConnectionReestablishmentRequest-r8
ue-Identity
c-RNTI
Bin : 4C F0 (= 19696)
physCellId : 30 This is the PCI of the cell
shortMAC-I where UE was last succesfully
Bin : AB 7D (= 43901) connected to. In HO case it’s
reestablishmentCause : otherFailure the source cell.
spare
Bin : 0 (2 bits)
Data (hex):
09 9E 01 EA B7 D8

17 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

 RRC Re-Establishment Procedure, Successful

• RRC Re-Establishment message sets up SRB1 again

RRC SIGNALING MESSAGE
Time: 16:49:06.085
RRCConnectionReestablishment (3GPP TS 36.331 ver 8.7.0 Rel 8)

DL-CCCH-Message
message
c1
rrcConnectionReestablishment
rrc-TransactionIdentifier :2
criticalExtensions
c1
rrcConnectionReestablishment-r8
SRB1 re-establishment
radioResourceConfigDedicated
in RRC Conn Re-
srb-ToAddModList
establishment
srb-ToAddModList value 1 message.
srb-Identity :1
mac-MainConfig
explicitValue
ul-SCH-Config
maxHARQ-Tx : n5
periodicBSR-Timer : infinity
retxBSR-Timer : sf2560
ttiBundling : false
drx-Config : release
-----------------KLIP-----------------------

18 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

RRC Re-Establishment Procedure, Unsuccessful

- If re-establishment is not possible eNB responds with reject  UE performs cell

reselection + TAU

UE EUTRAN

RRC Re-Est triggered. UE makes cell reselection and sends Re-Est.

RRCConnectionReestablishmentRequest

RRCConnectionReestablishmentReject
Cell reselection + TAU Req

RRC Conn Setup Complete

(NAS: Tracking Area Update Request)

19 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

RRC Re-Establishment Procedure, Unsuccessful

•Nemo log of reject case

EventId Time Subchannel Direction Message
RRCSM 09:38:13.165 CCCH Uplink RRCConnectionReestablishmentRequest
RRCSM 09:38:13.244 CCCH Downlink RRCConnectionReestablishmentReject
RRCSM 09:38:13.969 BCCH-BCH Downlink MasterInformationBlock
RRCSM 09:38:14.014 BCCH-SCH Downlink SystemInformationBlockType1
L3SM 09:38:14.014 Uplink TRACKING_AREA_UPDATE_REQUEST

•RRC Conn Re-est Request

RRC SIGNALING MESSAGE
Time: 9:38:13.165
RRCConnectionReestablishmentRequest (3GPP TS 36.331 ver 8.7.0 Rel 8)
UL-CCCH-Message
message
c1
rrcConnectionReestablishmentRequest Re-establishment cause can be:
criticalExtensions
• Other Failure (UE radio link failure)
rrcConnectionReestablishmentRequest-r8
ue-Identity • Handover Failure (T304 HO timer expiry)
c-RNTI
• Reconfiguration Failure
Bin : 4C F0 (= 19696)
physCellId : 30
shortMAC-I
Bin : AB 7D (= 43901)
reestablishmentCause: otherFailure
spare
Bin : 0 (2 bits)

20 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

RRC Re-Establishment signaling, HO Case
Source Target Serving
UE MME
eNB eNB GW

Packet data Packet data

Legend

Source C-RNTI, MAC-I

Measurement Reports L3 signalling

HO Decision

Handover Request User Data

and PCI sent to
target cell in X2AP
Admission Control and
Resource Allocation

TX2RELOCOverall
Handover Request Acknowledge
TX2RELOCexec
HO Req
.RRC Connection Reconfiguration

Deliver buffered and in

Detach from old cell transit packets to target eNB

SN Status Transfer
Data Forwarding
Failes to access target cell
T304 expires
Cell selection to source cell Buffer Packets from
Source eNB

RRC Connection
Reestablishment

ÚE can try re-

Request(Cause: handover
Failure; c-RNTI of source cell)
Accept RCR
Stop data forwarding
Handover Cancel establishment to any
cell, but only source and
RRC Connection
Reestablishment

target cell are prepared

RRC Connection
Reestablishment Complete

Reestablish SRB2 &

DRB
Optionally configure
Measurement Gaps

RRC Connection
Reconfiguration
RRC Connection
Reconfiguration Complete

Packet data Packet data

21 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

RRC Connection Re-Establishment Trigger Causes

- When UE is in RRC_CONNECTED and RRC security is active, it can trigger RRC

Connection Re-establishment
1. upon T310 expiry
2. upon reaching the maximum number of UL RLC retransmissions
3. upon handover failure (T304 expiry)
4. upon non-HO related random access problem

- If successful, RRC Conn Re-Establishment

• reconfigures SRB1 to resume data transfer of RRC msgs
• re-activates RRC security without changing algorithms
• NOTE: DRB should be re-configured after SRB1
- NOTE: if UE is in RRC_CONNECTED while RRC security is not active,
UE goes to RRC_IDLE, performs cell reselection and TAU

22 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network KPIs
- KPI Reference Values
- Drop Call Causes, Theory
• UE initiated drop
• eNB initiated drop
- Project Example

This module discusses call drops and handover drops after

call setup has already succeeded. See separate module for
call setup problems.

30 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

eNB Initiated Call Drops: Overview

• eNB can drop the call due to following triggers

• eNB-detected radio link problems

• TA timer expiry

• GTP-U failure

31 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Radio Link Problem Detection at eNB

- eNB radio link problem detection mechanisms are NSN-internally specified

- Multiple methods (called “link monitors”) are defined to detect a radio link problem in
the eNB.
- When one link monitor detects a problem, it is really a radio link problem even if other
link monitors have not yet indicated anything.
- Each link monitor has its internal criteria to decide when radio link problem is flagged
and de-flagged (radio link recovers).
- If the RLP persists longer than T_RLF, RRC+S1 release is triggered
• T_RLF = t310 + t311 (eNB-internal timer)
- Link monitors:
1. Uplink PUSCH DTX detection for scheduled uplink data
2. CQI DTX detection for periodic CQI reports in PUCCH and PUSCH
3. Uplink Ack/Nack DTX detection for transmitted downlink data
4. PDCCH Order RLF
5. SRS DTX detection (TD-LTE)

32 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

eNB Initiated Call Drops: Overview

• eNB can drop the call due to following triggers

• eNB-detected radio link problems

• TA timer expiry

• GTP-U failure

43 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

TA Timer Expiry at UE

- As UE detects Out-of-Sync status using a Timing Alignment Timer, the timer shall be
started or restarted whenever an initial TA or a TA update command is received (see
[3GPP-36.321], section 5.2). If the timer expires, the UE detects out-of-sync status.
- 3GPP TS 36.321: When timeAlignmentTimer expires at UE, UE MAC layer shall:
- flush all HARQ buffers;
- notify RRC layer to release PUCCH/SRS;
- clear any configured downlink assignments and uplink grants

- 3GPP TS 36.331: Upon receiving a PUCCH/ SRS release request from MAC layer, the
UE RRC shall:
• release periodic CQI reporting config, ie it stops CQI reporting on PUCCH
• release Scheduling Request Config

LNCEL:applyOutOfSyncState
Apply UL out-of-sync state
extendedDrxOnly (0), allDrx (1), allUEs (2)
Default: 0

45 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

eNB Initiated call Drops: Overview

• eNB can drop the call due to following triggers

• eNB-detected radio link problems

• TA timer expiry

• GTP-U failure

55 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

Release due to GTP-U Error Indication from S-GW (eNB-triggered)

- Handling of Event "GTP-U Error Indication"

- eNB may receive a “GTP-U Error Indication” on an active (single) S1 bearer (S-GW has
rejected the reception of uplink data packets). In that case eNB shall send the S1AP
message UE CONTEXT RELEASE REQUEST with cause “TNL Cause Transport Resource
Unavailable” to MME

- Example: This failure cause happens S-GW relocation attempt is not successful in X2
handover. In RL40/50, there is no support for S-GW relocation (i.e. no support of new
uplink transport layer address and uplink GTP-TEID in PATH SWITCH REQUEST
ACKNOWLEDGE message).

56 RA47057EN16AGLA0 © Nokia Solutions and Networks 2016

NokiaEDU
LTE Optimization Principles [FL16A] Module 08

Mobility and Layering

RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- Inter-layer Mobility Optimization Use Cases
• UE capabilities Feature Group Indicator
• Tracking Area optimization
• Location Area optimization
• LTE-capable UEs camping in 2G/3G
• LTE inter-frequency handover optimization
• Ping-pong mobility

5 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Drive Test KPIs for Mobility

- Use case: UE is camping in 2G/3G when LTE coverage is better.

• Analysis requires simultaneous scanner + UE logging
• Field test KPI TBD
- Use case: ping-pong cell reselection
• Field test KPI TBD

7 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- Inter-layer Mobility Optimization Use Cases
• UE capabilities Feature Group Indicator
• Tracking Area optimization
• Location Area optimization
• LTE-capable UEs camping in 2G/3G
• LTE inter-frequency handover optimization
• Ping-pong mobility

8 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

UE Capabilities Troubleshooting – FGI

- Common problem cases:

• UE does not support LTE inter-frequency handover + meas gaps
• UE does not support LTE3G ISHO

- How to find out UE capabilities?

- Why is this a problem?
• UE firmware feature details are often not known
• Most drive test tools don’t decode Feature Group Indicator in UE Capability msg

9 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

UE Capability Info, Feature Group Indicator Bitmap
….
interRAT-BandList value 5
Not decoded by all tools,
interRAT-NeedForGaps : true see TS 36.331 Annex B
interRAT-BandList value 6
interRAT-NeedForGaps : true
featureGroupIndicators Not all tools decode FGI
Bin : 5E 0D D8 80 (= 1577965696)
• Wireshark decodes.
interRAT-Parameters
• Nemo 6.1 decodes partly
utraFDD
supportedBandListUTRA-FDD
supportedBandListUTRA-FDD value 1 : bandI
supportedBandListUTRA-FDD value 2 : bandVIII
geran
supportedBandListGERAN
supportedBandListGERAN value 1 : gsm850
supportedBandListGERAN value 2 : gsm900E
supportedBandListGERAN value 3 : gsm1800
supportedBandListGERAN value 4 : gsm1900
interRAT-PS-HO-ToGERAN : false

10 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- Inter-layer Mobility Optimization Use Cases
• UE capabilities Feature Group Indicator
• Tracking Area optimization
• Location Area optimization
• LTE-capable UEs camping in 2G/3G
• LTE inter-frequency handover optimization
• Ping-pong mobility

12 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Tracking Area Optimization – Benefits

- Paging carried in
• all cells in a TA, or
• all cells in a list of TAs
- Tracking Area optimization is a trade-off between excessive paging load and
excessive TA updates
- Tracking Area optimization potential benefits
• Decrease Physical Downlink Shared Channel (PDSCH) Utilization
- Increase throughput resource available for end users
- Reduce admission control failures for VoLTE
• Increase Physical Random Access Channel (PRACH) Availability
- Improve user experience of network access and latency
• Optimize Mobility Management Entity (MME) Paging and Tracking Update
Performance
- Improve the end user system availability experience
• Lengthen UE Battery Life by Reducing Mobility Related Signaling

13 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Tracking Area Optimization – Trade-off

- Paging increases in a linear fashion

as a TA is made larger.

- TA Updates are inversely

proportional to the TA size. There is
a diminishing improvement in TA
updates as the size increases.

- The optimal size of a TA is call

model dependent. The rate at
which UE move, receive DL data
and enter idle mode influence the
optimal size.

- MME paging limitations must be

also considered!
- MME paging capacity can be the
limiting factor for TA size.
14 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016
Tracking Area Optimization – Typical Case

- Heavy traffic along the road causes tracking area updates

- Optimization solution 1: combine TAs 1-3 to one big TA
• Drawback: heavy paging load in eNBs

TAI = 3

TAI = 2
TAI = 1

Heavy traffic along road

15 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Tracking Area Optimization – Typical Case
- Heavy traffic along the road causes tracking area updates
- Optimization solution 2: put TAs 1-3 to the same TA list
• Drawback: heavy paging load in eNBs

TAI = 3

TAI = 2
TAI = 1

TA LIST:
TAI = 1
TAI = 2
Heavy traffic
TAI = 3

16 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- Inter-layer Mobility Optimization Use Cases
• UE capabilities Feature Group Indicator
• Tracking Area optimization
• Location Area optimization
• LTE-capable UEs camping in 2G/3G
• LTE inter-frequency handover optimization
• Ping-pong mobility

17 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Location Area Optimization

- CSFB capable UE makes combined EPS/IMSI attach/TAU in which MME assigns LAC to
UE
– LAC assigned based on TAI-LAI-VLR mapping table in MME
- When UE moves to 2G/3G, either because of normal cell reselection or CSFB, it will
make LAU if 2G/3G LAC is different from LTE LAC
- Question: should we set LTE LAC to be the same as 2G or 3G LAC?
– If LTE coverage is patchy, frequent LAU will occur when UE reselects to 2G/3G and
back  signaling load and “CS paging black hole”
– CSFB call setup time could increase because UE must make LAU first
– BUT: there is a danger of UE ending up being unreachable or increased paging time
if LTE LAC is the same as non-LTE LAC and UE goes camping in 2G/3G (next slide)

18 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Location Area Optimization, CS Paging Problem

- If CSFB UE does not make LAU when it reselects to 2G or 3G, MSS still assumes UE is
in LTE and pages it first from there, and only afterwards in 2G/3G (depending on core
implementation)
- Solution 1: use Gs interface so that UE makes combined LAU/RAU when it reselects
2G/3G  MSS will be informed by SGSN that UE is in 2G/3G, and there is no CS paging
delay
- Solution 2: make sure that LTE LAC is different from 2G/3G LAC (can use dummy LAC
in MME)
• Also MSS pooling, if used, may require that LTE LAC and 3G LAC are different
- Discuss options with core planning team!!
• Also Idle Mode Signaling Reduction feature might be implemented in core (not all
UEs support yet)…

19 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

 The case where LTE LAC and 3G LAC are different

Time RRC direction RRC message name Attach Accept on LTE side assigns LAC
14:32:40.432 Uplink ATTACH_REQUEST (after combined EPC/IMSI Attach)
14:32:40.432 Uplink PDN_CONNECTIVITY_REQUEST
14:32:40.514 Downlink IDENTITY_REQUEST
14:32:40.514 Uplink IDENTITY_RESPONSE
14:32:40.551 Downlink AUTHENTICATION_REQUEST
14:32:40.731 Uplink AUTHENTICATION_RESPONSE
14:32:40.731 Downlink SECURITY_MODE_COMMAND
14:32:40.731 Uplink SECURITY_MODE_COMPLETE
14:32:51.827 Downlink ATTACH_ACCEPT
14:32:51.827 Downlink ACTIVATE_DEFAULT_EPS_BEARER_CONTEXT_REQUEST
14:32:51.835 Uplink ATTACH_COMPLETE
14:32:51.835 Uplink ACTIVATE_DEFAULT_EPS_BEARER_CONTEXT_ACCEPT
Here LTEreselection or CSFB to 3G takes place, 3G cell has a different LAC than LTE cell
14:33:37.280 Uplink LOCATION_UPDATING_REQUEST
14:33:37.280 Uplink ROUTING_AREA_UPDATE_REQUEST
14:33:37.931 Downlink IDENTITY_REQUEST
14:33:38.133 Uplink IDENTITY_RESPONSE
14:33:39.359 Downlink AUTHENTICATION_REQUEST
14:33:39.359 Downlink AUTHENTICATION_AND_CIPHERING_REQUEST LA Update Request required since 3G LAC is
14:33:39.538 Uplink AUTHENTICATION_RESPONSE different from LTE LAC
14:33:39.779 Uplink AUTHENTICATION_AND_CIPHERING_RESPONSE
14:33:40.178 Downlink IDENTITY_REQUEST
14:33:40.178 Uplink IDENTITY_RESPONSE
14:33:40.840 Downlink GPRS_IDENTITY_REQUEST
14:33:40.840 Uplink GPRS_IDENTITY_RESPONSE
14:33:41.001 Downlink LOCATION_UPDATING_ACCEPT
14:33:41.059 Uplink TMSI_REALLOCATION_COMPLETE
14:33:41.059 Downlink MM_INFORMATION
14:33:41.479 Downlink ROUTING_AREA_UPDATE_ACCEPT

20 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- Inter-layer Mobility Optimization Use Cases
• UE capabilities Feature Group Indicator
• Tracking Area optimization
• Location Area optimization
• LTE-capable UEs camping in 2G/3G
• LTE inter-frequency handover optimization
• Ping-pong mobility

21 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

LTE-capable UEs Camping in 2G/3G

- LTE-capable UEs should stay in LTE if there is coverage (and not excessive interference)
- LTE uses Rel8 absolute priority based cell reselection
• In Rel8 3GLTE and 2GLTE HOs not supported
• Hence in initial phases, idle mode mobility optimization is crucial
• RU40 supports RRC Release with Redirect to LTE
- 3GLTE reselection use case:
• LTE layer(s) have higher priority than 3G layer
• LTE coverage area is a subarea of 3G coverage
- 2GLTE reselection use case
• Rural areas of 3G operators and non-3G operators
• LTE layer(s) have higher priority than 2G layer
• LTE coverage area is a subarea of 2G coverage

22 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- Inter-layer Mobility Optimization Use Cases
• UE capabilities Feature Group Indicator
• Tracking Area optimization
• Location Area optimization
• LTE-capable UEs camping in 2G/3G
• LTE inter-frequency handover optimization
• Ping-pong mobility

Idle and connected mode parameter knowledge is prerequisite for

this module. See
•LTEPAR training

Handover drops are covered in another module of this training

23 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Inter frequency cell reselection
• Idle mode strategies:
• 1. Priority layer LTE coverage
LTE capacity
• Limited coverage on capacity layer (hot-spots) Layer
Priority 3
Layer
Priority 4
• Larger bw on higher priority layer

• 2. Equal priority
• Similar or large overlapping coverage areas
• Equal bw on both layers
• Target to distribute users on both layers
• Propagation loss difference between layers must be considered

24 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Inter-frequency HO Optimization

- IFHO meas gaps should not be activated too early

• Activating IFHO measurements reduces UE tput due to meas gaps
• Activating IFHO measurements deactivates DRX  increases battery consumption
for smart phones
TCP and UDP DL tput, with and without measurement gap, 10MHz system

80

70

60
MEAS GAP ON UDP
50
tput, Mbps

MEAS GAP OFF UDP

40
MEAS GAP ON TCP
30 MEAS GAP OFF TCP

20

10

0
1 6 11 16 21 26 31 36 41 46 51 56
time, seconds

25 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Inter-frequency HO Optimization

- HO should not be triggered too early, an aggressive trigger point would be RSRP
below -115dBm @20MHz bandwidth if there is coverage from a better neighbor layer
• Otherwise end user experience suffers
• If interference present, threshold may need to be much higher
• Based on field measurements, RSRP <-115dBm results in <20Mbps tput for 20MHz
unloaded system. Anything above this is usually still ok for web browsers…
- Triggering HO at RSRP > -100dBm produces little practical user tput benefit
• TCP and internet limit practical tput of end users
• Speed test users are another story..
- Settings should be such that ping-pong is not possible for any pair of layers
• At least 5dB margin

26 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Index

- Network + Field KPIs

- Inter-layer Mobility Optimization Use Cases
• UE capabilities Feature Group Indicator
• Tracking Area optimization
• Location Area optimization
• LTE-capable UEs camping in 2G/3G
• LTE inter-frequency handover optimization
• Ping-pong mobility

28 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

Ping-Pong Idle + Connected Mode Mobility

- For every pair of layers, triggering thresholds should be checked for

– Idle-idle ping-pong
– Idle-connected ping-pong
– Connected-idle ping-pong
– Connected-connected ping-pong

- Example of idle-connected ping-pong:

1. 3GLTE idle mode reselection takes place at when LTE received level is RSRP = -
125dBm
2. In LTE, UE goes to connected mode due to always-on applications
3. A2 RRC redirect to UTRA threshold has been set to -120dBm
4. A2 is triggered and UE is redirected back to 3G
5. …and so on… go to step 1

29 RA47058EN16AGLA0 © Nokia Solutions and Networks 2016

NokiaEDU

RA47058EN16AGLA0 © Nokia Solutions and Networks 2016