13 VoLTE
13 VoLTE
Contents
5.2.1 VoLTE
eRAN
VoLTE Feature Parameter Description
Issue 05
Date 2021-11-27
HUAWEI TECHNOLOGIES CO., LTD.
and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.
All other trademarks and trade names mentioned in this document are the property of their respective holders.
Notice
The purchased products, services and features are stipulated by the contract made between Huawei and the
customer. All or part of the products, services and features described in this document may not be within the
purchase scope or the usage scope. Unless otherwise specified in the contract, all statements, information, and
recommendations in this document are provided "AS IS" without warranties, guarantees or representations of any
kind, either express or implied.
The information in this document is subject to change without notice. Every effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute a warranty of any kind, express or implied.
Website: https:/
Email: [email protected]
5.2.1 Contents
1 Change History
1.1 eRAN17.1 05 (2021-11-27)
1.2 eRAN17.1 04 (2021-09-29)
1.3 eRAN17.1 03 (2021-06-26)
1.4 eRAN17.1 02 (2021-04-30)
1.5 eRAN17.1 01 (2021-03-05)
1.6 eRAN17.1 Draft A (2020-12-29)
3 Overview
3.1 Background
3.2 Introduction
3.3 Architecture
5 Capacity Enhancement
5.1 Semi-Persistent Scheduling and Power Control
5.1.1 Principles
5.1.1.1 VoIP Semi-Persistent Scheduling
5.1.1.2 Enhanced VoIP Semi-Persistent Scheduling
5.1.1.3 Power Control in Semi-Persistent Scheduling
5.1.2 Network Analysis
5.1.2.1 Benefits
5.1.2.2 Impacts
5.1.3 Requirements
5.1.3.1 Licenses
5.1.3.2 Software
5.1.3.3 Hardware
5.1.3.4 Others
5.1.4 Operation and Maintenance
5.1.4.1 Data Configuration
5.1.4.1.1 Data Preparation
5.1.4.1.2 Using MML Commands
5.1.4.1.3 Using the MAE-Deployment
5.1.4.2 Activation Verification
5.1.4.3 Network Monitoring
5.2 ROHC
6 Coverage Optimization
6.1 TTI Bundling
6.1.1 Principles
6.1.1.1 Protocol Compliance (FDD)
6.1.1.2 TTI Bundling Entry Conditions
6.1.1.3 Data Block Transmission
6.1.1.4 TTI Bundling Exiting Conditions
6.1.1.5 Enhanced TTI Bundling Algorithm (FDD)
6.1.2 Network Analysis
6.1.2.1 Benefits
6.1.2.2 Impacts
6.1.3 Requirements
6.1.3.1 Licenses
6.1.3.2 Software
6.1.3.3 Hardware
6.1.3.4 Others
6.1.4 Operation and Maintenance
6.1.4.1 Data Configuration
6.1.4.1.1 Data Preparation
6.1.4.1.2 Using MML Commands
6.1.4.1.3 Using the MAE-Deployment (FDD)
6.1.4.2 Activation Verification
6.1.4.3 Network Monitoring
6.2 Uplink RLC Segmentation Enhancement
6.2.1 Principles
6.2.2 Network Analysis
6.2.2.1 Benefits
6.2.2.2 Impacts
6.2.3 Requirements
6.2.3.1 Licenses
6.2.3.2 Software
6.2.3.3 Hardware
6.2.3.4 Others
6.2.4 Operation and Maintenance
6.2.4.1 Data Configuration
6.2.4.1.1 Data Preparation
6.2.4.1.2 Using MML Commands
6.2.4.1.3 Using the MAE-Deployment
6.2.4.2 Activation Verification
6.2.4.3 Network Monitoring
6.3 CMR-based Voice Rate Control
6.3.1 Principles
6.3.1.1 Scenarios for Voice Rate Control
6.3.1.2 Types of Voice Rate Control
6.3.1.3 Mechanisms of Voice Rate Control
6.3.1.4 Restrictions
6.3.2 Network Analysis
6.3.2.1 Benefits
6.3.2.2 Impacts
6.3.3 Requirements
6.3.3.1 Licenses
6.3.3.2 Software
6.3.3.3 Hardware
6.3.3.4 Others
6.3.4 Operation and Maintenance
6.3.4.1 Data Configuration
6.3.4.1.1 Data Preparation
6.3.4.1.2 Using MML Commands
6.3.4.1.3 Using the MAE-Deployment
6.3.4.2 Activation Verification
6.3.4.3 Network Monitoring
6.4 MAC CE-based Voice Rate Control
6.4.1 Principles
6.4.2 Network Analysis
6.4.2.1 Benefits
6.4.2.2 Impacts
6.4.3 Requirements
6.4.3.1 Licenses
6.4.3.2 Software
6.4.3.3 Hardware
6.4.3.4 Others
6.4.4 Operation and Maintenance
6.4.4.1 Data Configuration
6.4.4.1.1 Data Preparation
6.4.4.1.2 Using MML Commands
6.4.4.1.3 Using the MAE-Deployment
6.4.4.2 Activation Verification
6.4.4.3 Network Monitoring
6.5 Inter-eNodeB VoLTE CoMP (FDD)
6.5.1 Principles
6.5.2 Network Analysis
6.5.2.1 Benefits
6.5.2.2 Impacts
6.5.3 Requirements
6.5.3.1 Licenses
6.5.3.2 Software
6.5.3.3 Hardware
6.5.3.4 Networking
6.5.3.5 Others
6.5.4 Operation and Maintenance
6.5.4.1 Data Configuration
6.5.4.1.1 Data Preparation
6.5.4.1.2 Using MML Commands
6.5.4.1.3 Using the MAE-Deployment
6.5.4.2 Activation Verification
6.5.4.2.1 Using MML Commands
6.5.4.2.2 Using Counters
6.5.4.3 Network Monitoring
6.6 Coverage-based VoLTE Experience Optimization
6.6.1 Principles
6.6.1.1 Flash SRVCC
6.6.1.2 Voice-Quality-based Inter-Frequency Handover
6.6.2 Network Analysis
6.6.2.1 Benefits
6.6.2.2 Impacts
6.6.3 Requirements
6.6.3.1 Licenses
6.6.3.2 Software
6.6.3.3 Hardware
6.6.3.4 Others
6.6.4 Operation and Maintenance
6.6.4.1 Data Configuration
6.6.4.1.1 Data Preparation
6.6.4.1.2 Using MML Commands
6.6.4.1.3 Using the MAE-Deployment
6.6.4.2 Activation Verification
6.6.4.3 Network Monitoring
6.7 VoLTE Coverage Enhancement
6.7.1 Principles
6.7.1.1 Cross-Layer Optimization for VoLTE in the Uplink
6.7.1.2 Uplink AMR Voice Frame Recovery
6.7.2 Network Analysis
6.7.2.1 Benefits
6.7.2.2 Impacts
6.7.3 Requirements
6.7.3.1 Licenses
6.7.3.2 Software
6.7.3.3 Hardware
6.7.3.4 Others
6.7.4 Operation and Maintenance
6.7.4.1 Data Configuration
6.7.4.1.1 Data Preparation
6.7.4.1.2 Using MML Commands
6.7.4.1.3 Using the MAE-Deployment (FDD)
6.7.4.2 Activation Verification
6.7.4.3 Network Monitoring
7 Quality Improvement
7.1 Voice Characteristic Awareness Scheduling
7.1.1 Principles
7.1.2 Network Analysis
7.1.2.1 Benefits
7.1.2.2 Impacts
7.1.3 Requirements
7.1.3.1 Licenses
7.1.3.2 Software
7.1.3.3 Hardware
7.1.3.4 Others
7.1.4 Operation and Maintenance
7.1.4.1 Data Configuration
7.1.4.1.1 Data Preparation
7.1.4.1.2 Using MML Commands
7.1.4.1.3 Using the MAE-Deployment (FDD)
7.1.4.2 Activation Verification
7.1.4.3 Network Monitoring
7.2 Voice Call Connection Delay Optimization
7.2.1 Principles
7.2.2 Network Analysis
7.2.2.1 Benefits
7.2.2.2 Impacts
7.2.3 Requirements
7.2.3.1 Licenses
7.2.3.2 Software
7.2.3.3 Hardware
7.2.3.4 Others
7.2.4 Operation and Maintenance
7.2.4.1 Data Configuration
7.2.4.1.1 Data Preparation
7.2.4.1.2 Using MML Commands
7.2.4.1.3 Using the MAE-Deployment (FDD)
7.2.4.2 Activation Verification
7.2.4.3 Network Monitoring
7.3 Uplink VoLTE Continuous Scheduling
7.3.1 Active Scheduling of Voice Service UEs
7.3.1.1 Principles
7.3.1.2 Network Analysis
7.3.1.2.1 Benefits
7.3.1.2.2 Impacts
7.3.1.3 Requirements
7.3.1.3.1 Licenses
7.3.1.3.2 Software
7.3.1.3.3 Hardware
7.3.1.3.4 Others
7.3.1.4 Operation and Maintenance
7.3.1.4.1 Data Preparation
7.3.1.4.2 Using MML Commands
7.3.1.4.3 Using the MAE-Deployment
7.3.1.4.4 Activation Verification
7.3.1.4.5 Network Monitoring
7.3.2 Uplink Service Status Determination
7.3.2.1 Principles
7.3.2.2 Network Analysis
7.3.2.2.1 Benefits
7.3.2.2.2 Impacts
7.3.2.3 Requirements
7.3.2.3.1 Licenses
7.3.2.3.2 Software
7.3.2.3.3 Hardware
7.3.2.3.4 Others
7.3.2.4 Operation and Maintenance
7.3.2.4.1 Data Preparation
7.3.2.4.2 Using MML Commands
7.3.2.4.3 Using the MAE-Deployment
7.3.2.4.4 Activation Verification
7.3.2.4.5 Network Monitoring
7.3.3 Enhanced Compensation Scheduling During Talk Spurts
7.3.3.1 Principles
7.3.3.2 Network Analysis
7.3.3.2.1 Benefits
7.3.3.2.2 Impacts
7.3.3.3 Requirements
7.3.3.3.1 Licenses
7.3.3.3.2 Software
7.3.3.3.3 Hardware
7.3.3.3.4 Others
7.3.3.4 Operation and Maintenance
7.3.3.4.1 Data Preparation
7.3.3.4.2 Using MML Commands
7.3.3.4.3 Using the MAE-Deployment
7.3.3.4.4 Activation Verification
7.3.3.4.5 Network Monitoring
7.4 UL Compensation Scheduling
7.4.1 Principles
7.4.2 Network Analysis
7.4.2.1 Benefits
7.4.2.2 Impacts
7.4.3 Requirements
7.4.3.1 Licenses
7.4.3.2 Software
7.4.3.3 Hardware
7.4.3.4 Others
7.4.4 Operation and Maintenance
7.4.4.1 Data Configuration
7.4.4.1.1 Data Preparation
7.4.4.1.2 Using MML Commands
7.4.4.1.3 Using the MAE-Deployment
7.4.4.2 Activation Verification
7.4.4.3 Network Monitoring
7.5 Voice-specific AMC
7.5.1 Principles
7.5.2 Network Analysis
7.5.2.1 Benefits
7.5.2.2 Impacts
7.5.3 Requirements
7.5.3.1 Licenses
7.5.3.2 Software
7.5.3.3 Hardware
7.5.3.4 Others
7.5.4 Operation and Maintenance
7.5.4.1 Data Configuration
7.5.4.1.1 Data Preparation
7.5.4.1.2 Using MML Commands
7.5.4.1.3 Using the MAE-Deployment
7.5.4.2 Activation Verification
7.5.4.3 Network Monitoring
7.6 VoLTE User Prior Access
7.6.1 VoLTE User Prior Access for Mobile-originated Calls
7.6.2 VoLTE User Prior Access for Mobile-terminated Calls
7.6.3 Network Analysis
7.6.3.1 Benefits
7.6.3.2 Impacts
7.6.4 Requirements
7.6.4.1 Licenses
7.6.4.2 Software
7.6.4.3 Hardware
7.6.4.4 Others
7.6.5 Operation and Maintenance
7.6.5.1 Data Configuration
7.6.5.1.1 Data Preparation
7.6.5.1.2 Using MML Commands
7.6.5.1.3 Using the MAE-Deployment
7.6.5.2 Activation Verification
7.6.5.3 Network Monitoring
7.7 Preferential Access of Voice Services
7.7.1 Principles
7.7.2 Network Analysis
7.7.2.1 Benefits
7.7.2.2 Impacts
7.7.3 Requirements
7.7.3.1 Licenses
7.7.3.2 Software
7.7.3.3 Hardware
7.7.3.4 Others
7.7.4 Operation and Maintenance
7.7.4.1 Data Configuration
7.7.4.1.1 Data Preparation
7.7.4.1.2 Using MML Commands
7.7.4.1.3 Using the MAE-Deployment
7.7.4.2 Activation Verification
7.7.4.3 Network Monitoring
7.8 PUSCH RB Reservation for Voice Service UEs
7.8.1 Principles
7.8.2 Network Analysis
7.8.2.1 Benefits
7.8.2.2 Impacts
7.8.3 Requirements
7.8.3.1 Licenses
7.8.3.2 Software
7.8.3.3 Hardware
7.8.3.4 Others
7.8.4 Operation and Maintenance
7.8.4.1 Data Configuration
7.8.4.1.1 Data Preparation
7.8.4.1.2 Using MML Commands
7.8.4.1.3 Using the MAE-Deployment
7.8.4.2 Activation Verification
7.8.4.3 Network Monitoring
7.9 Uplink Voice Mute Recovery
7.9.1 Principles
7.9.2 Network Analysis
7.9.2.1 Benefits
7.9.2.2 Impacts
7.9.3 Requirements
7.9.3.1 Licenses
7.9.3.2 Software
7.9.3.3 Hardware
7.9.3.4 Others
7.9.4 Operation and Maintenance
7.9.4.1 Data Configuration
7.9.4.1.1 Data Preparation
7.9.4.1.2 Using MML Commands
7.9.4.1.3 Using the MAE-Deployment
7.9.4.2 Activation Verification
7.9.4.3 Network Monitoring
8 Power Saving
9 Mobility Management
9.1 Overview
9.2 Intra-Frequency Handover
9.3 Inter-Frequency Handover
9.4 Inter-RAT Handover
9.4.1 Handover Type
9.4.2 Handover Policies
12 Parameters
13 Counters
14 Glossary
15 Reference Documents
1 Change History
This chapter describes changes not included in the "Parameters", "Counters", "Glossary", and
"Reference Documents" chapters. These changes include:
• Technical changes
Changes in functions and their corresponding parameters
• Editorial changes
Improvements or revisions to the documentation
1.1 eRAN17.1 05 (2021-11-27)
Technical Changes
None
Editorial Changes
Added descriptions of activation verification for optimized MCS selection for downlink voice
services. For details, see 4.4.2 Activation Verification.
1.2 eRAN17.1 04 (2021-09-29)
Technical Changes
None
Editorial Changes
Revised descriptions of uplink VoLTE volume estimation for dynamic scheduling. For details,
see 7.1.1 Principles.
1.3 eRAN17.1 03 (2021-06-26)
Technical Changes
None
Editorial Changes
Technical Changes
Editorial Changes
Added descriptions of carrier aggregation (CA) optimization to enhanced VoLTE user prior
access for mobile-originated calls. For details, see 7.6.1 VoLTE User Prior Access for Mobile-originated
Calls.
Technical Changes
Changed the function name "acceleration guarantee for fast-moving UEs" to "fast-moving voice
service UE guarantee" and revised the related descriptions. For details, see 10 Special Treatment of
Other Functions on Voice Services.
Technical Changes
Added Video Added the MO_VILTE_CALL_IND_SW option to the FDD • 3900 and
over LTE (ViLTE) VolteAlgoConfig.VolteOptSwitch parameter. 5900
user prior access series
for mobile- base
originated calls. stations
For details, see • DBS3900
7.6 VoLTE User Prior LampSite
Access.
and
DBS5900
LampSite
Added enhanced Added the VoLTEDci1aEnhancedSwitch option to the FDD • 3900 and
voice service CellDlschAlgo.DlEnhancedVoipSchSw parameter. 5900
scheduling in DCI series
format 1A. For base
details, see 4.1.5 stations
Dynamic Scheduling • DBS3900
and Power Control.
LampSite
and
DBS5900
LampSite
Change Parameter Change RAT Base Station Model
Description
Extended initial Added the FAST_MOVING_UE_FLAG option to the FDD • 3900 and
acceleration for QciPara.AggregationAttribute parameter. 5900
voice service Added parameters: series
UEs to • VolteAlgoConfig.AsVolteDlPwrBoostSinrThld base
acceleration stations
• AsParaGroup.AggregationAttribute
guarantee for • DBS3900
fast-moving UEs. • SpidCfg.AggregationAttribute and its option
LampSite
For details, see FAST_MOVING_UE_FLAG and
10 Special Treatment DBS5900
of Other Functions
LampSite
on Voice Services.
Asymmetric base
Bandwidth and stations
the following • DBS3900
functions: semi- LampSite
persistent and
scheduling and DBS5900
TTI bundling. For LampSite
details, see 5.1.2.2
Impacts and 6.1.2.2
Impacts.
Added enhanced Added the FAST_EPS_FALLBACK_SW option to the FDD • 3900 and
VoLTE user prior VolteAlgoConfig.VolteOptSwitch parameter. 5900
access for Added parameters: series
mobile-originated • VolteAlgoConfig.EpsFbUlActiveSchMinPeriod base
calls. For details, stations
• VolteAlgoConfig.EpsFbUlActiveSchDataVol
see 7.6 VoLTE User • DBS3900
Prior Access. • GlobalProcSwitch.MoVoiceCallUeCapbQryDelay
LampSite
• GlobalProcSwitch.EpsFbHoInUeCapbQryDelay and
DBS5900
LampSite
Added voice call Added the VOLTE_CALL_HOLD_SW option to the FDD • 3900 and
hold optimization. VolteAlgoConfig.VolteOptSwitch parameter. 5900
For details, see series
4.1.5 Dynamic base
Scheduling and stations
Power Control.
• DBS3900
LampSite
and
DBS5900
LampSite
Editorial Changes
Changed the function name "voice rate control" to "CMR-based voice rate control". For details,
see 6.3 CMR-based Voice Rate Control.
Changed the function name "load-based scheduling" to "scheduling based on TTI-level UE
number". For details, see 5.1.1.1 VoIP Semi-Persistent Scheduling.
Purpose
This document only provides guidance for feature activation. Feature deployment and feature gains
depend on the specifics of the network scenario where the feature is deployed. To achieve optimal
gains, contact Huawei professional service engineers.
Software Interfaces
Any parameters, alarms, counters, or managed objects (MOs) described in this document apply
only to the corresponding software release. For future software releases, refer to the
corresponding updated product documentation.
Trial Features
Trial features are features that are not yet ready for full commercial release for certain reasons.
For example, the industry chain (terminals/CN) may not be sufficiently compatible. However,
these features can still be used for testing purposes or commercial network trials. Anyone who
desires to use the trial features shall contact Huawei and enter into a memorandum of
understanding (MoU) with Huawei prior to an official application of such trial features. Trial
features are not for sale in the current version but customers may try them for free.
Customers acknowledge and undertake that trial features may have a certain degree of risk due to
absence of commercial testing. Before using them, customers shall fully understand not only the
expected benefits of such trial features but also the possible impact they may exert on the
network. In addition, customers acknowledge and undertake that since trial features are free,
Huawei is not liable for any trial feature malfunctions or any losses incurred by using the trial
features. Huawei does not promise that problems with trial features will be resolved in the
current version. Huawei reserves the rights to convert trial features into commercial features in
later R/C versions. If trial features are converted into commercial features in a later version,
customers shall pay a licensing fee to obtain the relevant licenses prior to using the said
commercial features. If a customer fails to purchase such a license, the trial feature(s) will be
invalidated automatically when the product is upgraded.
2.2 Applicable RAT
For details about interoperability between E-UTRAN and NG-RAN, see Interoperability Between E-
UTRAN and NG-RAN.
3 Overview
3.1 Background
There are three LTE voice solutions available. This section describes these voice solutions.
A dual-standby UE can receive or send signals on evolved universal terrestrial radio access
networks (E-UTRAN) and GSM/EDGE radio access networks or universal terrestrial radio
access networks (GERAN/UTRAN) simultaneously. Dual-standby UEs automatically use the
GERAN/UTRAN for voice services and the E-UTRAN for data services. The E-UTRAN
provides dual-standby UEs with data services only.
In the early stages of LTE network deployment, circuit switched fallback (CSFB) is used as an
interim solution to provide voice services for LTE UEs when the IP multimedia subsystem (IMS)
has not yet been deployed. Figure 3-1 illustrates the voice solution based on CSFB.
Figure 3-1 Voice solution based on CSFB
In a CSFB solution, when a UE initiates a circuit switched (CS) service in the E-UTRAN, the
MME instructs the UE to fall back to the legacy GERAN or UTRAN before the UE accesses the
service. For details about CSFB, see CS Fallback.
The Voice over LTE (VoLTE) solution is used on mature LTE networks with an IMS deployed,
as shown in Figure 3-2. With this solution, UEs can directly access voice services on an LTE
network.
If LTE coverage is not complete, UEs may leave LTE coverage areas, causing their voice
services to be discontinued. Depending on whether the packet switched (PS) domain of the
UTRAN/GERAN supports VoIP services, Huawei uses different methods to ensure voice service
continuity:
• Not supported
VoIP services are handed over to the CS domain of the UTRAN/GERAN using single
radio voice call continuity (SRVCC). For details on SRVCC, see SRVCC.
• Supported
VoIP services are handed over to the UTRAN/GERAN using PS handovers. For
details on PS handovers, see Mobility Management in Connected Mode.
Figure 3-2 Voice solution based on IMS
3.2 Introduction
VoLTE allows UEs on the E-UTRAN to access voice services. The voice services are set up
over the IP transport network between the UEs and the IMS.
Emergency services are not described in this document. For details on emergency services, see
Emergency Call.
3.3 Architecture
Network Architecture
The IMS includes multiple network elements (NEs). These NEs perform voice session control
and multimedia negotiation between the calling and called UEs.
Functional Architecture
• Basic VoLTE functions allow UEs to access voice services. For details, see 4.1
Principles in 4 Basic VoLTE Functions.
• Enhanced VoLTE functions improve voice service performance. These functions are
classified into the following categories:
▪ Capacity improvement. For details, see 5 Capacity Enhancement.
▪ Coverage improvement. For details, see 6 Coverage Optimization.
▪ Quality improvement. For details, see 7 Quality Improvement.
▪ Power saving. For details, see 8 Power Saving.
▪ Mobility management. For details, see 9 Mobility Management.
▪ Special treatment of other functions on voice services. For details, see 10
Special Treatment of Other Functions on Voice Services.
lists features and functions related to VoLTE. For the technical principles and
Table 3-1
engineering guidelines, see the corresponding feature parameter descriptions.
Table 3-1 Other VoLTE-related features and functions
Feature/Function Document
Inter-frequency handover
VoLTE uses the adaptive multi-rate (AMR) or enhanced voice services (EVS) speech codec
scheme.
AMR
AMR is an audio data compression scheme optimized for speech coding. It is widely used in
GERANs and UTRANs. There are AMR wideband (AMR-WB) and AMR narrowband (AMR-
NB) schemes. For details about AMR-WB, see 3GPP TS 26.201. For details about AMR-NB,
see 3GPP TS 26.101. The following are the voice coding rates supported by these schemes:
• Voice coding rates (kbit/s) supported by AMR-WB:
6.6, 8.85, 12.65, 14.25, 15.85, 18.25, 19.85, 23.05, and 23.85
• Voice coding rates (kbit/s) supported by AMR-NB:
4.75, 5.15, 5.9, 6.7, 7.4, 7.95, 10.2, and 12.2
In this document, AMR-NB corresponds to AMR in 3GPP specifications.
EVS
EVS, introduced in 3GPP TS 26.445 in September 2014, allows for high-definition (HD) VoLTE
speech coding and provides more significant gains than AMR-WB. There are EVS narrowband
(EVS-NB), EVS wideband (EVS-WB), EVS superwideband (EVS-SWB), and EVS fullband
(EVS-FB) schemes. The following are the voice coding rates supported by these schemes:
• Voice coding rates (kbit/s) supported by EVS-NB:
5.9, 7.2, 8.0, 9.6, 13.2, 16.4, and 24.4
• Voice coding rates (kbit/s) supported by EVS-WB:
5.9, 7.2, 8.0, 9.6, 13.2, 16.4, 24.4, 32, 48, 64, 96, and 128
• Voice coding rates (kbit/s) supported by EVS-SWB:
9.6, 13.2, 16.4, 24.4, 32, 48, 64, 96, and 128
• Voice coding rates (kbit/s) supported by EVS-FB:
16.4, 24.4, 32, 48, 64, 96, and 128
4.1.1.2 Speech Traffic Model
illustrates the speech traffic model when either AMR or EVS is used. The AMR or EVS
Figure 4-1
speech codec scheme to use is negotiated between the UEs and the IMS.
Figure 4-1 Speech traffic model
UE capability and data configurations on the MME determine whether a UE uses VoLTE.
However, VoLTE may be inappropriate for certain sites or regions. This case is termed a
VoLTE-prohibited scenario.
This section describes voice policy selection for UEs in common scenarios and VoLTE-
prohibited scenarios.
4.1.2.1 Common Scenarios
During attach or a tracking area update (TAU), the MME selects a voice policy based on the UE
capability and data configurations on the MME side. The MME then sends the UE the voice
policy using the Attach Accept or TAU Accept message. The MME selects a voice policy based
on the following principles:
• If the UE supports CSFB only, the MME selects CS Voice only (that is, CSFB).
• If the UE supports VoLTE only, the MME selects IMS PS Voice only (that is,
VoLTE).
• If the UE supports both CSFB and VoLTE, the voice policy used before negotiation
with the MME is one of the following voice policies, which is specified by operators
during UE registration:
▪ CS Voice only
CSFB is selected.
▪ IMS PS Voice only
VoLTE is selected.
▪ Prefer CS Voice with IMS PS Voice as secondary
CSFB takes precedence over VoLTE.
For details on the procedures for voice policy negotiation between the UE
and MME when this policy is used, see Annex A.2 in 3GPP TS 23.221
V9.4.0.
▪ Prefer IMS PS Voice with CS Voice as secondary
VoLTE takes precedence over CSFB.
Figure 4-2and Figure 4-3 show the procedures for voice policy negotiation
between the UE and MME when this policy is used. If no voice policy is
selected for the UE, the UE camps on the current network, which is data-
centric, or performs cell reselection to a GSM/UMTS network, which is
voice-centric. For more details on the negotiation, see Annex A.2 in 3GPP
TS 23.221 V9.4.0.
Figure 4-2 Procedure for voice policy selection (non-combined attach)
Figure 4-3 Procedure for voice policy selection (combined attach)
VoLTE mobility capability decision was introduced in 3GPP Release 11. This further helps the
MME in selecting a voice policy.
4.1.2.1.2 VoLTE Mobility Capability Decision
1. During attach, the MME sends a UE Radio Capability Match Request message to the
eNodeB to query the UE's VoLTE mobility capability. This message was introduced
in 3GPP Release 11.
2. If the eNodeB has not received the UE radio capability from the UE, it sends a UE
Capability Enquiry message to the UE.
3. The UE uses the UE Capability Information message to report its radio capability.
For details on this message, see section 5.6.3 "UE Capability Transfer" in 3GPP TS
36.331 Release 10.
4. If the eNodeB determines that the UE can ensure voice service continuity after the
UE accesses the voice service, it informs the MME using the UE Radio Capability
Match Response message.
VoLTE mobility capability decision is controlled by the setting of the
SupportS1UeCapMatchMsg option of the GlobalProcSwitch.ProtocolSupportSwitch parameter.
• If this option is selected, the eNodeB supports VoLTE mobility capability decision.
The eNodeB determines that a UE can ensure voice service continuity if the UE
supports at least one of the following:
▪ UTRAN and SRVCC from E-UTRAN to UTRAN
▪ GERAN and SRVCC from E-UTRAN to GERAN
▪ PS domain of UTRAN-FDD (that is, VoHSPA), and SRVCC from this PS
domain to the CS domain of UTRAN-FDD or GERAN
▪ PS domain of UTRAN-TDD (that is, VoHSPA), and SRVCC from this PS
domain to the CS domain of UTRAN-TDD or GERAN
• If this option is deselected, the eNodeB does not evaluate VoLTE mobility
capabilities. When the eNodeB receives a UE Radio Capability Match Request
message, it responds to the MME with an Error Indication. As a result, VoLTE is
selected as the voice policy for UEs that do not support SRVCC, in which case voice
service continuity cannot be ensured.
When making a VoLTE mobility capability decision, the eNodeB may consider the MME's SRVCC capability. The
MME informs the eNodeB of this capability using the Initial UE Context Setup message.
• After the eNodeB obtains the MME's SRVCC capability, it considers both the MME's capability and
the preceding UE capabilities. If the MME does not support SRVCC, the eNodeB replies to the MME
that the VoLTE service continuity cannot be ensured.
• If the eNodeB is not informed of the MME's SRVCC capability (for example, the UE Radio Capability
Match Request message arrives at the eNodeB earlier than the Initial UE Context Setup message), the
eNodeB does not consider the MME's capability when determining voice service continuity.
The E-UTRAN supports VoLTE after the IMS is deployed. However, in some scenarios, VoLTE
is prohibited and a non-VoLTE solution (such as CSFB) must be used.
4.1.2.2.1 Excessively Large End-to-End Delay
Description
VoLTE services have strict requirements regarding end-to-end delay. According to ITU-T
Recommendation G.114, an end-to-end delay of less than 200 ms is needed for VoLTE UEs to
be "very satisfied" with voice quality; between 200 ms to 275 ms for "satisfied"; and if the end-
to-end delay exceeds 275 ms, VoLTE UEs will start to be "dissatisfied". The ranges are shown in
Figure 4-5. In Table 6.1.7: Standardized QCI characteristics in section 6.1.7.2 "Standardized QCI
characteristics" of 3GPP TS 23.203, the recommended delay budget is 80 ms for the Uu interface
and 20 ms between the EPC and eNodeB. If the time of transmission from the eNodeB to the
EPC is longer than 20 ms, the end-to-end delay will be greater than 2 x (80 ms + 20 ms) after
VoLTE is deployed on the eNodeB. As a result, voice quality may not be satisfactory.
Figure 4-5 Relationship between the delay and voice quality
Configuration Principles
• Allocate dedicated tracking area identities (TAIs) to the specified areas, and configure
data at the associated MMEs so that CSFB instead of VoLTE will be used in the
identified tracking areas (TAs). During attach and TAUs, UEs negotiate or re-
negotiate voice policies with the MMEs. This negotiation process is transparent to
eNodeBs.
▪ For eNodeBs in the TAs with these dedicated TAIs:
If VoLTE is disabled for these TAs on the MMEs, the VoIP-related
performance indicators, such as E-RAB Setup Success Rate (VoIP), become 0 in
the cells served by these eNodeBs.
In VoLTE-prohibited scenarios, select the VoLTEOffOptSwitch option of
the CellAlgoSwitch.VoLTESwitch parameter for the cells served by these
eNodeBs. This way, if an Initial UE Context Setup, E-RAB Setup Request,
or E-RAB Modify Request message involves both a QCI 1 bearer and other
bearers, the eNodeB will no longer reject the other bearers, preventing the
E-RAB setup success rate of the other bearers from decreasing. E-RAB
setup success rate (excluding VoIP) = (L.E-RAB.SuccEst – L.E-
RAB.SuccEst.QCI.1)/[(L.E-RAB.AttEst – L.E-RAB.FailEst.X2AP) – (L.E-
RAB.AttEst.QCI.1 – L.E-RAB.FailEst.X2AP.VoIP)] × 100%
Description
Operators prefer to assign specific frequency bands, such as LTE TDD bands, only to data
services.
Configuration Principles
The configuration principles are the same as those when the end-to-end delay is excessively
large. For details, see 4.1.2.2.1 Excessively Large End-to-End Delay.
4.1.2.2.3 Voice Services Prohibited on Certain Frequencies
Description
Configuration Principles
On the eNodeB side, voice session setup includes the following procedures:
• RRC connection setup
• QCI 5 radio bearer setup
• QCI 1 radio bearer setup
Figure 4-6 shows the process of setting up a voice session between two UEs.
Figure 4-6 Voice service setup process
Radio bearer QoS management for voice services complies with the Policy and Charging Control
(PCC) architecture defined in 3GPP specifications.
Figure 4-7 shows the QoS management architecture for radio bearers of voice services.
Figure 4-7 Radio bearer QoS management architecture
QoS parameter control is performed for the dedicated bearers used by voice services, based on
the dynamic PCC rule, as follows:
1. The IMS (P-CSCF) sends QCI information to the PCRF over the Rx interface.
2. The PCRF uses received QCI information and subscription information to generate a
QoS rule, and then sends the rule to the P-GW over the Gx interface. This rule
includes the following key QoS parameters: QCI, allocation and retention priority
(ARP), guaranteed bit rate (GBR), and maximum bit rate (MBR).
3. The P-GW uses the QoS rule received from the PCRF to instruct the S-GW, MME,
and eNodeB to set up EPS bearers. Services of different QoS requirements are
carried by radio bearers with different QCIs. The QCIs for conversational voice,
conversational video, and IMS signaling are 1, 2, and 5, respectively, in compliance
with 3GPP specifications. Table 4-1 lists their QoS parameters. For details, see 3GPP
TS 23.203.
Table 4-1 QoS parameters for conversational voice, conversational video, and IMS signaling
QCI Example Resource Type Priority Packet Delay Packet Error Loss
Services Level Budget Rate
QoS parameters for QCIs 1, 2, and 5 are set in the QciPara MO. The RLC modes for
conversational voice, conversational video, and IMS signaling services are specified
by the RlcPdcpParaGroup.RlcMode parameter.
If the source cell and target cell of an inter-eNodeB handover use different RLC
modes, when the handover is complete, the voice service UE uses the RLC mode of
the source cell by default. If the UE needs to use the RLC mode of the target cell, set
the QciPara.RlcModeReconfigSwitch parameter to ON.
For details on QCIs and RLC modes, see QoS Management.
4.1.4 Admission and Congestion Control
This section describes how the admission and congestion control policy and the following basic
features work for VoLTE:
• LBFD-002023 Admission Control
• LBFD-002024 Congestion Control
For details about these features, see Admission and Congestion Control.
The eNodeB performs admission and congestion control for conversational voice (QCI 1) and
IMS signaling (QCI 5) separately.
Load Monitoring
Load monitoring provides decision references for admission and congestion control. The
eNodeB monitors various resources in a cell to obtain, for example, the usage of physical
resource blocks (PRBs), QoS satisfaction rates of GBR services, and resource insufficiency
indicators. This way, the eNodeB can know the cell status.
• Conversational voice (QCI 1)
For details about the method of calculating the QoS satisfaction rate of QCI 1 services,
see Admission and Congestion Control.
• IMS signaling (QCI 5)
QCI 5 indicates non-GBR services, and their QoS satisfaction rates do not need to be
calculated.
Admission Control
Admission control determines whether to admit a GBR service (new service or handover service)
based on the cell load reported by the load monitoring module. The cell load is represented by
the following: PRB usage, QoS satisfaction rates of GBR services, and resource insufficiency
indicators. For details, see Admission and Congestion Control.
• Conversational voice (QCI 1)
Admission control over GBR services with QCI 1 is performed based on load.
• IMS signaling (QCI 5)
▪ In FDD, admission control over non-GBR services with QCI 5 is not based
on load. If sounding reference signal (SRS) and physical uplink control
channel (PUCCH) resources are successfully allocated, these services are
directly admitted.
SRS resource allocation is considered during admission control over non-GBR services
with QCI 5 only when the eNodeB is equipped with an LBBPc board. SRS resources must
be successfully allocated before these services can be admitted.
Congestion Control
If the network is congested, the eNodeB preferentially releases low-priority GBR services to free
up resources for other services, whose QoS requirements can therefore be fulfilled. For details,
see Admission and Congestion Control.
• Conversational voice (QCI 1)
The eNodeB evaluates load status by monitoring PRB usage and the QoS satisfaction
rate. If the eNodeB determines that a cell is congested, the eNodeB rejects service
access requests and triggers congestion control to decrease the load. The congestion
threshold for QCI 1 services is specified by the CellQciPara.CongestionThreshold
parameter. For details on how to set this parameter, see Admission and Congestion Control.
• IMS signaling (QCI 5)
Congestion control is not performed for non-GBR services with QCI 5.
4.1.5 Dynamic Scheduling and Power Control
This section describes how the following features work for VoLTE:
• LBFD-00101502 Dynamic Scheduling
For details on the principles and engineering guidelines of dynamic scheduling, see Scheduling.
LampSite eNodeBs do not support high speed mobility or 1.4 MHz bandwidth. Dynamic scheduling policies
described in this document are suitable only for macro eNodeBs.
Voice services have demanding requirements on the delay. When dynamic scheduling is used for
voice services, the scheduler optimizes the handling of voice service priorities to ensure the
voice service QoS. When VoLTE is deployed, it is recommended that the enhanced proportional
fair (EPF) scheduling policy be used in both the uplink and downlink as follows:
• For the uplink
Set the CellUlschAlgo.UlschStrategy parameter to ULSCH_STRATEGY_EPF.
• For the downlink
Set the CellDlschAlgo.DlschStrategy parameter to DLSCH_PRI_TYPE_EPF.
On commercial LTE networks, the EPF scheduling policy is used in the uplink and downlink by
default.
When dynamic scheduling is using the uplink EPF scheduling algorithm, the priority of QCI 1
voice packets is lower than the priorities of the following information but is higher than the
priorities of any other initial transmission:
• HARQ retransmission
• Signaling radio bearer 1 (SRB1) and SRB2
• IMS signaling (QCI 5)
• Control signaling for public safety (QCI 69)
• Voice services for public safety (QCI 65)
The following are the policies for uplink dynamic scheduling:
• Voice-specific setting of the maximum number of uplink HARQ transmissions
▪ When a voice service is set up and the UlVoipCrosslayerOptSwitch option
of the CellUlschAlgo.UlEnhencedVoipSchSw parameter is deselected, the
maximum number of uplink HARQ transmissions for the voice service is
determined as follows:
▪ If the voice service is not in the TTI bundling state, a voice-
specific setting of the maximum number of uplink HARQ
transmissions may take effect, as controlled by the
CellQciPara.UlHarqMaxTxNum parameter.
▪ If this parameter is set to N0, this voice-specific setting
does not take effect. Instead, the maximum number of
HARQ transmissions for voice services is determined
by the CellUlschAlgo.UlHarqMaxTxNum parameter.
▪ If this parameter is set to a value other than N0, this
voice-specific setting takes effect. This function is not
recommended for cells heavily loaded with uplink
services.
▪ If the voice service is in the TTI bundling state, the maximum
number of uplink HARQ transmissions for the service is
determined by the CellUlschAlgo.TtiBundlingHarqMaxTxNum
parameter.
▪ When a voice service is set up and the UlVoipCrosslayerOptSwitch option
of the CellUlschAlgo.UlEnhencedVoipSchSw parameter is selected, the
maximum number of uplink HARQ transmissions for the voice service is
determined as described in 6.7 VoLTE Coverage Enhancement.
• Optimized uplink modulation and coding scheme (MCS) selection for voice services
When dynamic scheduling is used for voice services, it is recommended that the
CellUlschAlgo.UlVolteDeltaSinrForNack parameter be used to help ensure conservative
selection of MCSs for uplink transmission of voice services, reducing the uplink
packet loss rate of voice services and improving user experience with voice services.
▪ If this parameter is set to 0, this function does not take effect.
▪ If this parameter is set to a value other than 0, this function takes effect.
In FDD, the eNodeB lowers the uplink MCS indexes for voice service UEs
by this fixed value.
This function is not recommended for cells heavily loaded with uplink
services.
• (FDD) RB increase for first retransmission
When dynamic scheduling is used for voice services, as long as there are enough
available uplink PRBs, it is recommended that the
FIRST_RETRANS_EXPN_RB_SWITCH option of the
CellUlschAlgo.UlVoLTERetransSchStrategy parameter be selected. Selecting this option
reduces the voice packet loss rate and improves user experience with voice services. If
this option is selected for a cell, the eNodeB evaluates whether to use RB-increased
adaptive retransmission from the first uplink retransmission onwards. The evaluation
is based on the transmit power of voice service UEs.
If this option is selected, RB-increased adaptive retransmission is used from the first retransmission
onwards, including the last two retransmissions even if the UlLast2RetransSchOptSwitch option of
the CellAlgoSwitch.UlSchSwitch parameter is deselected. For details about the
UlLast2RetransSchOptSwitch option, see Scheduling.
This function does not take effect for 256QAM-enabled UEs, to prevent frequent switching between
DCI format 1A and DCI format 2A from causing the residual block error rate (RBLER) to increase.
In the voice rate control features and voice quality indicator (VQI) measurement algorithm, the
eNodeB must obtain negotiated speech codec schemes and rate sets of UEs. Currently, these are
obtained by parsing SIP messages.
4.1.6.1 Basic Concepts of IPsec
Internet Protocol Security (IPsec) is a suite of protocols and services that provide integrity
protection and security for IP networks. The IPsec protocol suite contains two security protocols
and one key exchange negotiation protocol.
The protocols are as follows:
• Security protocols: Authentication Header (AH) and Encapsulation Security Payload
(ESP)
• Key exchange negotiation protocol: Internet Key Exchange (IKE)
Item AH ESP
Protocol ID in IP header 51 50
As specified in RFC 2406, IPsec ESP in transport mode is used for SIP message transmission for
voice services. ESP can work in an encryption-only or integrity-protection-only manner.
• When the encryption algorithm is null and the integrity protection algorithm is not
null, only integrity protection is performed.
• When the integrity protection algorithm is null and the encryption algorithm is not
null, only encryption is performed. In transport mode, ESP encrypts the payload of IP
packets but does not encrypt IP headers.
The SIP message parsing procedure for voice services includes SIP message encryption check
and SIP message parsing, as shown in Figure 4-8 and Figure 4-9.
Figure 4-8 SIP message encryption check
Figure 4-9 SIP message parsing
Parsing Procedure
1. SIP message encryption check
a. After receiving an IP packet on the QCI 5 radio bearer, the eNodeB
checks the value in the encryption field of the IP packet header to
determine whether an encryption algorithm has been used.
The encryption field is protocolType or NextHeader in an IPv4 or IPv6
packet header, respectively.
• If the value is 50, an encryption algorithm has been used. The
eNodeB proceeds to encryption algorithm check.
• If the value is not 50, no encryption algorithm has been used.
The eNodeB proceeds to SIP message parsing.
b. Encryption algorithm check
The eNodeB checks whether the mandatory header field From or CSeq
can be found in individual SIP messages.
• If From or CSeq is found, the null encryption algorithm has
been used. The eNodeB proceeds to SIP message parsing.
• If From or CSeq is not found, an encryption algorithm other
than null has been used. The eNodeB does not parse the SIP
messages. Instead, it considers that SIP message parsing has
failed.
• Before determining whether IP packets received over a QCI 5 radio bearer have been
encrypted, the eNodeB filters out non-SIP messages.
• Before determining whether this encryption algorithm is the null algorithm, the eNodeB
checks the number of SIP messages after the filtering:
If there are at least four SIP messages left after the filtering, the eNodeB proceeds to
encryption algorithm check.
If there are fewer than four SIP messages after the filtering and the mandatory header
field is found in the SIP messages, the null encryption algorithm has been used. If there
are fewer than four SIP messages after the filtering and the mandatory header field is not
found in the SIP messages, the eNodeB cannot determine whether the null encryption
algorithm has been used.
a: PT in the keyword is a variable that denotes the PT value recorded in the mapping
table.
b: AMR rate set information includes mode-set.
EVS rate set information includes evs-mode-switch, bw, br, and ch-aw-recv.
If the record is fragmented, the eNodeB clears the rate set information corresponding
to the PT value in the mapping table.
3. PT value check and voice rate set determining
a. When the eNodeB receives an uplink RTP voice packet on the QCI 1
bearer, it parses the "payload type" field in the packet header to obtain the
PT value.
b. The eNodeB compares the PT value in the uplink RTP packet header
with the PT values recorded in SIP message parsing.
• If the PT value is the same as a recorded PT value, the rate set
information corresponding to the PT value is used as the
negotiated QCI 1 rate set information for the current voice
service.
• If the PT value is different from the recorded PT values, the
eNodeB considers that the negotiated QCI 1 rate set
information for the current voice service is not obtained.
Example
3. The eNodeB searches for the keyword "rtpmap" and records the corresponding PT
value and encoding scheme (AMR, AMR-WB, or EVS) in the mapping table, as
illustrated in Figure 4-11.
Figure 4-11 Search for the keyword "rtpmap"
4. The eNodeB searches for the keyword "fmtp:PT" and records the rate set information
corresponding to "fmtp:PT" in the mapping table. Specifically, the eNodeB searches
for "fmtp:110", "fmtp:100", and "fmtp:96", and updates the mapping table with their
rate set information, as illustrated in Figure 4-12.
Figure 4-12 Search for the keyword "fmtp:PT"
5. The eNodeB compares the PT value (110) in an uplink RTP packet with the
previously recorded PT values (110, 100, and 96) to find the suitable rate set
information.
As shown in Figure 4-13, the PT value in the uplink RTP packet is 110, which is the
same as a PT value recorded in the mapping table. The rate set information
corresponding to this PT value is used as the negotiated QCI 1 rate set information
for the current voice service.
Figure 4-13 PT value in an uplink RTP voice packet
In scenarios where the eNodeB cannot parse SIP messages (referred to as parsing-limited
scenarios), the eNodeB cannot obtain the accurate negotiated rate set information for the UE.
The parsing-limited scenarios are listed as follows:
• SIP messages are encrypted using an algorithm other than null.
• SIP messages are not encrypted or are encrypted using the null algorithm. However,
the SIP messages are fragmented and no fragment contains the complete coding rates
corresponding to the PT value for the current voice service.
• The eNodeB cannot determine whether the SIP messages have been encrypted using
the null encryption algorithm or an encryption algorithm other than null.
• SIP messages are out of order due to transmission faults, and the eNodeB fails to
obtain the rate set information before receiving an ACK message.
• When a UE is handed over or the RRC connection of a UE is being reestablished, the
target cell does not receive the negotiated rate set or receives an incorrect one in a
private X2 message, and the target cell cannot obtain the negotiated rate set
information by parsing SIP messages.
4.1.7 Emergency VoLTE Handling
In special scenarios such as earthquakes or tsunamis, voice capacity must be guaranteed, even if
voice quality has to suffer somewhat under specific conditions.
Emergency VoLTE handling adjusts three types of resources, described in Table 4-4, Table 4-5, and
Table 4-6, to provide voice services in the special scenarios. The intelligent optimization functions
specified by the LIOptRule.RuleID parameter can be activated or deactivated by running the
ACT LIOPTRULE or DEA LIOPTRULE command, respectively. For details, see 4.4.1 Data
Configuration. The following are the three types of resource adjustments:
38 Either of the following conditions is met: 30 or The PDCCH parameter settings are restored to those
• PDCCH_CCE_Utilization_Rate < 40 in the current database.
60%
• VOIP_CCE_UTILIZATION_RATE
< 60%
39 Both of the following conditions are met: 31 The VoLTE load is high. Parameters are
• DL_PRB_Utilization_Rate > 80% and adjusted to increase the PDSCH capacity:
• VOIP_DL_PRB_UTILIZATION_RATE 33 • The
> 80% VoipTbsBasedMcsSelSwitch
option of the
CellAlgoSwitch.DlSchSwitch
parameter is changed to be
deselected.
• The
DlRetxTbsIndexAdjOptSwitch
option of the
CellAlgoSwitch.CqiAdjAlgoSwitch
parameter is changed to be
deselected.
• The DLDeltaCqiOptSwitch option
of the
CellCqiAdjAlgo.DlVolteCqiAdjOptSw
Rule Trigger Conditions Atom Parameter Adjustment
ID Rule
ID
parameter is changed to be
deselected.
40 Either of the following conditions is met: 32 or The PDSCH parameter settings are restored
• DL_PRB_Utilization_Rate < 60% 34 to those in the current database.
• VOIP_DL_PRB_UTILIZATION_RATE
< 60%
41 Both of the following conditions are met: 35 Parameters are adjusted to increase the PUSCH
• UL_PRB_Utilization_Rate > 80% and capacity:
• VOIP_UL_PRB_UTILIZATION_RATE 37 • The UlLast2RetransSchOptSwitch
> 80% option of the
CellAlgoSwitch.UlSchSwitch parameter
is changed to be deselected.
• The UlVoipRblerControlSwitch option
of the
CellUlschAlgo.UlEnhencedVoipSchSw
parameter is changed to be deselected
• The value of the
CellUlschAlgo.UlCompenSchPeriodinSpu
parameter is changed to INTERVAL_4
• The
FIRST_RETRANS_EXPN_RB_SWITC
option of the
CellUlschAlgo.UlVoLTERetransSchStrate
parameter is changed to be deselected
• The value of the
CellUlschAlgo.UlVolteDeltaSinrForNack
parameter is changed to 0.
42 Either of the following conditions is met: 36 or The PUSCH parameter settings are restored to
• UL_PRB_Utilization_Rate < 60% 38 those in the current database.
• VOIP_UL_PRB_UTILIZATION_RATE
< 60%
For details about intelligent optimization function IDs, rule IDs, and atom rule IDs, see Automatic Congestion
Handling.
The following explains the performance indicators in the trigger conditions described in Table 4-4,
Table 4-5, and Table 4-6.
• PDCCH_CCE_Utilization_Rate
FDD: (L.ChMeas.CCE.CommUsed + L.ChMeas.CCE.ULUsed +
L.ChMeas.CCE.DLUsed)/(Number of CCEs per TTI when the maximum number of
PDCCH symbols is fixed x Number of TTIs in a measurement period)
• VOIP_CCE_UTILIZATION_RATE
(L.ChMeas.CCE.ULUsed.VoIP + L.ChMeas.CCE.DLUsed.VoIP)/(L.ChMeas.CCE.DLUsed +
L.ChMeas.CCE.ULUsed)
• DL_PRB_Utilization_Rate
L.ChMeas.PRB.DL.Used.Avg/L.ChMeas.PRB.DL.Avail
• VOIP_DL_PRB_UTILIZATION_RATE
L.ChMeas.PRB.DL.DrbUsed.Avg.VoIP/(L.ChMeas.PRB.DL.DrbUsed.Avg.VoIP +
L.ChMeas.PRB.DL.DrbUsed.Avg.QCI2 + L.ChMeas.PRB.DL.DrbUsed.Avg.QCI5 +
L.ChMeas.PRB.DL.DrbUsed.Avg.QCI6 + L.ChMeas.PRB.DL.DrbUsed.Avg.QCI7 +
L.ChMeas.PRB.DL.DrbUsed.Avg.QCI8 + L.ChMeas.PRB.DL.DrbUsed.Avg.QCI9 +
L.ChMeas.PRB.DL.DrbUsed.Avg.QCI65 + L.ChMeas.PRB.DL.DrbUsed.Avg.QCI66 +
L.ChMeas.PRB.DL.DrbUsed.Avg.ExtQci.Index0 + L.ChMeas.PRB.DL.DrbUsed.Avg.ExtQci.Index1
+ L.ChMeas.PRB.DL.DrbUsed.Avg.ExtQci.Index2 +
L.ChMeas.PRB.DL.DrbUsed.Avg.ExtQci.Index3 + L.ChMeas.PRB.DL.DrbUsed.Avg.ExtQci.Index4)
• UL_PRB_Utilization_Rate
L.ChMeas.PRB.UL.Used.Avg/L.ChMeas.PRB.UL.Avail
• VOIP_UL_PRB_UTILIZATION_RATE
L.ChMeas.PRB.UL.DrbUsed.Avg.VoIP/(L.ChMeas.PRB.UL.DrbUsed.Avg.VoIP +
L.ChMeas.PRB.UL.DrbUsed.Avg.QCI2 + L.ChMeas.PRB.UL.DrbUsed.Avg.QCI5 +
L.ChMeas.PRB.UL.DrbUsed.Avg.QCI6 + L.ChMeas.PRB.UL.DrbUsed.Avg.QCI7 +
L.ChMeas.PRB.UL.DrbUsed.Avg.QCI8 + L.ChMeas.PRB.UL.DrbUsed.Avg.QCI9 +
L.ChMeas.PRB.UL.DrbUsed.Avg.QCI65 + L.ChMeas.PRB.UL.DrbUsed.Avg.QCI66 +
L.ChMeas.PRB.UL.DrbUsed.Avg.ExtQci.Index0 + L.ChMeas.PRB.UL.DrbUsed.Avg.ExtQci.Index1
+ L.ChMeas.PRB.UL.DrbUsed.Avg.ExtQci.Index2 +
L.ChMeas.PRB.UL.DrbUsed.Avg.ExtQci.Index3 + L.ChMeas.PRB.UL.DrbUsed.Avg.ExtQci.Index4)
4.2 Network Analysis
4.2.1 Benefits
VoLTE provides voice services to the UEs in the E-UTRAN. The UEs do not need to fall back to
GERAN or UTRAN. VoLTE has the following advantages over voice services in GERAN or
UTRAN:
• Higher spectral efficiency
• Better user experience, such as lower access delay and better voice quality
When UEs are evenly distributed, the number of UEs remains unchanged, and data traffic is not
suppressed, emergency VoLTE handling is estimated to reduce the VoLTE resource (PRB and
CCE) usage by 3% to 10%.
• If the saved resources are used for bearers of new UEs, the number of served UEs is
expected to increase by 3% to 10%.
• If the saved resources are used for preferentially scheduled UEs, the perceived voice
quality for these UEs improves.
You are advised to evaluate when to use the function according to Non-Applicable Scenarios and
Applicable Scenarios to achieve optimal benefits.
Non-Applicable Scenarios
The E-UTRAN supports voice services after the IMS is deployed. However, IMS-based VoLTE
is not appropriate for the following scenarios:
• Transmission delay is too severe.
• Voice services are not allowed in certain frequency bands.
• Voice services are not allowed on certain frequencies.
To prevent UEs from initiating voice services in VoLTE-prohibited areas or to prevent VoLTE
UEs from being handed over to VoLTE-prohibited areas, it is recommended that the following
parameters or options be set: ENodeBAlgoSwitch.EutranVoipSupportSwitch, the
VoipHoControlSwitch option of ENodeBAlgoSwitch.HoAlgoSwitch,
EutranInterNFreq.VolteHoTargetInd, and the VoLTEOffOptSwitch option of
CellAlgoSwitch.VoLTESwitch. For details about the configurations, see 4.1.2.2 VoLTE-Prohibited
Scenarios.
Applicable Scenarios
• If the operator has deployed the IMS in other scenarios, it is recommended that the
ENodeBAlgoSwitch.EutranVoipSupportSwitch parameter be set to ON to enable IMS-
based VoLTE.
• If E-UTRAN coverage is discontinuous, it is recommended that SRVCC be enabled.
• If the MME supports the UE Radio Capability Match Request and UE Radio
Capability Match Response messages introduced in 3GPP Release 11, it is
recommended that the SupportS1UeCapMatchMsg option of the
GlobalProcSwitch.ProtocolSupportSwitch parameter be selected on the eNodeB side.
By doing this, the MME considers the VoLTE mobility capability of UEs during the
voice policy negotiation. In this manner, UEs that do not support SRVCC do not use
VoLTE as their voice policy, ensuring voice continuity.
• The voice quality monitoring (VQM) function is recommended for monitoring voice
quality in scenarios where VoLTE is commercially used. For details, see 11.2.3
Measurement-based Evaluation.
4.2.2 Impacts
Network Impacts
• When emergency VoLTE handling is used, the QCI 1 packet loss rate increases.
In VoLTE scenarios, Uplink Packet Loss Rate (VoIP) and Downlink Packet Loss Rate (VoIP)
increase by 10% to 30% when resource adjustment takes effect.
Function Impacts
FDD Short TTI SHORT_TTI_SW option of the Short TTI The eNodeB does
CellShortTtiAlgo.SttiAlgoSwitch parameter (FDD) not schedule a UE
in short TTI mode
if a voice bearer
has been set up
for this UE. If a
voice bearer is set
up during short TTI
scheduling, the
eNodeB
immediately stops
short TTI
scheduling for this
UE.
coverage
enhancement.
FDD UMTS and UL_SPECTRUM_SHR_PH2_SW option of UMTS and A delay may exist
LTE the SpectrumCloud.SpectrumCloudEnhSwitch LTE in the
Spectrum
Spectrum parameter Sharing communication
Sharing between UMTS
Phase 2 and LTE. The
spectrum
resources cannot
take effect
immediately after
being shared. As a
result, VoIP Semi-
persistent
Scheduling does
not take effect.
a: When repetition optimization for VoLTE is enabled, the following parameters, which are used
for VoLTE, do not take effect for eMTC UEs running voice services:
• CellUlschAlgo.UlVolteDeltaSinrForNack
• DLDeltaCqiOptSwitch option of the CellCqiAdjAlgo.DlVolteCqiAdjOptSw parameter
• CellCqiAdjAlgo.VolteNackDeltaCqi
• DlVoiceRetransOptSwitch option of the CellDlschAlgo.DlEnhancedVoipSchSw
parameter
4.3 Requirements
4.3.1 Licenses
4.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
4.3.3 Hardware
No requirements
Boards
The LBBPc board does not support retransmission scheduling preference during downlink
dynamic scheduling.
RF Modules
No requirements
4.3.4 Networking
Voice services are very sensitive to end-to-end delay. If the transmission delay between an
eNodeB and EPC exceeds 20 ms, VoLTE is not appropriate for the eNodeB.
For the requirements on transmission delay and jitter on S1 and X2 interfaces, see IP eRAN
Engineering Guide.
4.3.5 Others
FDD Eutran ENodeBAlgoSwitch.EutranVoipSupportSwitch Set this parameter to ON if you plan to use the
Voip VoLTE solution.
Support For details about when to deploy the VoLTE
Switch solution, see 4.2.1 Benefits.
• If this parameter is set to OFF, it is
recommended that the
CellHoParaCfg.InterRatMrDelayTimer
and CellHoParaCfg.VolteHoNrDelayTim
parameters be set to 0.
• If this parameter is set to ON, it is
recommended that the
CellHoParaCfg.VolteHoNrDelayTimera
parameter be set to a non-zero value
FDD E2E VQI ENodeBAlgoSwitch.E2EVQIAlgoSwitch This parameter specifies whether to enable E2E
Algorithm VQI evaluation when
Switch ENodeBAlgoSwitch.VQMAlgoSwitch is set to
different values.
To monitor E2E voice quality, set this paramete
to ON when ENodeBAlgoSwitch.VQMAlgoSwitc
is set to VQM_ALGO_SWITCH_AMR_ON or
VQM_ALGO_SWITCH_ADAPTIVE_ON.
The value ON is recommended.
a: Currently, NG-RAN does not support voice over NR (VoNR). When a UE initiates a voice
service in E-UTRAN, it may be handed over to NG-RAN before the QCI 1 bearer for the voice
service is set up. Then, the UE will fall back to E-UTRAN by EPS fallback. To prevent this
issue, set the CellHoParaCfg.VolteHoNrDelayTimer parameter to a non-zero value.
processing are
started, the timer for
delaying SRVCC MR
processing takes
effect.
Dynamic Scheduling
Table 4-11 describes the parameters used for function activation.
Table 4-11 Parameters used for activation
Use
are
Before using MML commands, refer to 4.2.2 Impacts and 4.3.2 Software and complete the parameter
configurations for related functions based on the impact and dependency relationships between
the functions, as well as the actual network scenario.
//Support of VoLTE
//Turning on the VoLTE switch
MOD ENODEBALGOSWITCH: EutranVoipSupportSwitch=ON;
MOD CELLHOPARACFG: LocalCellId=0, InterRatMrDelayTimer=3000,
VolteHoNrDelayTimer=100;
//Turning on the VQM algorithm switch and end-to-end VQI algorithm switch
MOD ENODEBALGOSWITCH: VQMAlgoSwitch=VQM_ALGO_SWITCH_ADAPTIVE_ON,
E2EVQIAlgoSwitch=ON;
//Turning on the switch for compatibility with UEs that do not support UM
MOD RLCPDCPPARAGROUP: RlcPdcpParaGroupId=0, CatType=LTE,
NonsptUmUeAdaptSwitch=ON;
//Dynamic scheduling
//Setting the uplink and downlink scheduling policies
MOD CELLULSCHALGO: LocalCellId=0, UlschStrategy=ULSCH_STRATEGY_EPF;
MOD CELLDLSCHALGO: LocalCellId=0, DlschStrategy=DLSCH_PRI_TYPE_EPF;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling VoLTE
MOD ENODEBALGOSWITCH: EutranVoipSupportSwitch=OFF;
MOD CELLHOPARACFG: LocalCellId=0, VolteHoNrDelayTimer=0;
//The following commands are optional.
//Turning off related global switches
MOD GLOBALPROCSWITCH: ProtocolSupportSwitch=SupportS1UeCapMatchMsg-0;
MOD ENODEBALGOSWITCH: VQMAlgoSwitch=VQM_ALGO_SWITCH_OFF,
E2EVQIAlgoSwitch=OFF;
//Turning off the switch for compatibility with UEs that do not support UM
MOD RLCPDCPPARAGROUP: RlcPdcpParaGroupId=0, CatType=LTE,
NonsptUmUeAdaptSwitch=OFF;
4. Check whether the following counters indicate successful voice service setup.
1526726669 L.E-RAB.SuccEst.QCI.1
1526726677 L.E-RAB.SuccEst.QCI.5
Optimized MCS selection for downlink voice services can be verified by performing the
following steps:
1. Enable a UE to access a cell and initiate a voice service.
2. Before enabling optimized MCS selection for downlink voice services, start a task on
the MAE-Access to monitor MCS-specific scheduling statistics.
a. On the MAE-Access, choose Monitor > Signaling Trace > Signaling
Trace Management.
b. In the left pane of the Signaling Trace Management window, choose
User Performance Monitoring > MCS Count Monitoring. Set the
tracing duration, to-be-traced MMEc (MME ID), and mTMSI (UE
TMSI).
c. On the MAE-Access, check the MCS-specific scheduling statistics and
record the MCS index.
3. After enabling optimized MCS selection for downlink voice services, record the
MCS index according to 2. If the MCS index decreases, this function has taken
effect.
Emergency VoLTE handling can be verified using the following methods:
• Method 1: Review SON logs on the MAE-Access.
On the MAE-Access, select LTE Automatic Congestion Handling Log to view the
SON logs. This function has taken effect if the following logs contain specific
information:
▪ Set Automatic Congestion Handling Switch
▪ Modify Cell-level Runtime Parameters
▪ Recover Cell-level Runtime Parameters
▪ Modify eNodeB-level Runtime Parameters
▪ Recover eNodeB-level Runtime Parameters
• Method 2: Run MML commands to view the parameter values in use.
Emergency VoLTE handling changes the parameter values in use, instead of the
configured values in the database. The values in use may differ from the configured
values. You can query the configured values by running the corresponding LST
commands and query the values in use as follows:
Run the DSP LIOPTRULE command, and view the values of Action Type, Action,
Local Cell ID, and Executive Status for specific intelligent optimization rules.
Expected result: The value of Action Type for some intelligent optimization rules is
MODIFY.
▪ If the value of Executive Status for an intelligent optimization rule is UN-
EXECUTED or EXECUTING, the running value of the parameter is the
same as the configured value. In this situation, run the corresponding LST
command to query the configured value, which is also the running value of
the parameter.
▪ If the value of Executive Status for an intelligent optimization rule is
EXECUTED, the parameter has been modified by ACH. In addition, if the
value of Action Type is MODIFY, you can check whether the running
value of the parameter is the same as the modified target value by running
the MML command in the Action field.
The following are descriptions of fields in the command output of DSP LIOPTRULE.
▪ Action Type: The value can be either MODIFY or RESUME. If it is
MODIFY, the intelligent optimization rule is used to modify parameter
values. If it is RESUME, the intelligent optimization rule is used to restore
parameter values to the originally configured values.
▪ Action: Indicates which and how parameters are modified when an intelligent
optimization rule applies. This field does not take effect if Action Type is set
to RESUME.
▪ Local Cell ID: Indicates the ID of the cell to which intelligent optimization
rules are applied. This field does not take effect for eNodeB-level intelligent
optimization rules.
▪ Executive Status: The value can be UN-EXECUTED, EXECUTING, and
EXECUTED. UN-EXECUTED indicates that the intelligent optimization
rule has not been executed. EXECUTING indicates that the intelligent
optimization rule is being executed. EXECUTED indicates that the intelligent
optimization rule has been executed.
Table 4-13 lists the counters used to monitor the E-RAB setup success rates for voice services.
Table 4-13 Counters used to monitor the E-RAB setup success rates of voice services
1526726668 L.E-RAB.AttEst.QCI.1
1526726676 L.E-RAB.AttEst.QCI.5
1526726669 L.E-RAB.SuccEst.QCI.1
1526726677 L.E-RAB.SuccEst.QCI.5
1526727853 L.E-RAB.AttEst.PLMN.QCI.1
1526727861 L.E-RAB.AttEst.PLMN.QCI.5
1526727854 L.E-RAB.SuccEst.PLMN.QCI.1
1526727862 L.E-RAB.SuccEst.PLMN.QCI.5
1526739741 L.E-RAB.FailEst.X2AP.VoIP
1526747657 L.E-RAB.FailEst.X2AP.VoIP.PLMN
Table 4-14 describes the counters used to monitor the handover success rates of voice services.
Table 4-14 Counters used to monitor the handover success rates of voice services
1526729526 L.HHO.IntraeNB.IntraFreq.PrepAttOut.VoIP
1526729527 L.HHO.IntraeNB.InterFreq.PrepAttOut.VoIP
1526729535 L.HHO.IntereNB.IntraFreq.PrepAttOut.VoIP
1526729536 L.HHO.IntereNB.InterFreq.PrepAttOut.VoIP
Counter ID Counter Name
1526729537 L.HHO.IntereNB.InterFddTdd.PrepAttOut.VoIP
1526729529 L.HHO.IntraeNB.IntraFreq.ExecAttOut.VoIP
1526729530 L.HHO.IntraeNB.InterFreq.ExecAttOut.VoIP
1526729538 L.HHO.IntereNB.IntraFreq.ExecAttOut.VoIP
1526729539 L.HHO.IntereNB.InterFreq.ExecAttOut.VoIP
1526729540 L.HHO.IntereNB.InterFddTdd.ExecAttOut.VoIP
1526729532 L.HHO.IntraeNB.IntraFreq.ExecSuccOut.VoIP
1526729533 L.HHO.IntraeNB.InterFreq.ExecSuccOut.VoIP
1526729541 L.HHO.IntereNB.IntraFreq.ExecSuccOut.VoIP
1526729542 L.HHO.IntereNB.InterFreq.ExecSuccOut.VoIP
1526729543 L.HHO.IntereNB.InterFddTdd.ExecSuccOut.VoIP
1526729528 L.HHO.IntraeNB.InterFddTdd.PrepAttOut.VoIP
1526729531 L.HHO.IntraeNB.InterFddTdd.ExecAttOut.VoIP
1526729534 L.HHO.IntraeNB.InterFddTdd.ExecSuccOut.VoIP
Table 4-15 describes the counters used to monitor the call drop rates of voice services.
Table 4-15 Counters used to monitor the call drop rates of voice services
1526726686 L.E-RAB.AbnormRel.QCI.1
1526726694 L.E-RAB.AbnormRel.QCI.5
Counter ID Counter Name
1526726687 L.E-RAB.NormRel.QCI.1
1526726695 L.E-RAB.NormRel.QCI.5
1526727871 L.E-RAB.AbnormRel.PLMN.QCI.1
1526727879 L.E-RAB.AbnormRel.PLMN.QCI.5
1526727872 L.E-RAB.NormRel.PLMN.QCI.1
1526727880 L.E-RAB.NormRel.PLMN.QCI.5
Table 4-16lists the counters used to monitor the uplink and downlink packet loss rates on the Uu
interface and the downlink PDCP packet discard rate for voice services.
Table 4-16 Counters used to monitor the packet loss rate
Counter ID Counter Name
1526727961 L.Traffic.UL.PktLoss.Loss.QCI.1
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
1526727934 L.Traffic.DL.PktUuLoss.Loss.QCI.1
1526727935 L.Traffic.DL.PktUuLoss.Tot.QCI.1
1526726833 L.PDCP.Tx.Disc.Trf.SDU.QCI.1
1526727889 L.PDCP.Tx.TotRev.Trf.SDU.QCI.1
1526736684 L.Traffic.UL.PktLoss.Loss.PLMN.QCI.1
1526736686 L.Traffic.UL.PktLoss.Tot.PLMN.QCI.1
1526736737 L.Traffic.DL.PktUuLoss.Loss.PLMN.QCI.1
1526736739 L.Traffic.DL.PktUuLoss.Tot.PLMN.QCI.1
Counter ID Counter Name
1526736680 L.PDCP.Tx.Disc.Trf.SDU.PLMN.QCI.1
1526736682 L.PDCP.Tx.TotRev.Trf.SDU.PLMN.QCI.1
1526745995 L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1
1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1
1526745996 L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1.Index0
1526745998 L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1.Index0
Downlink packet loss rate on the Uu interface for QCI 1 of a specific operator =
L.Traffic.DL.PktUuLoss.Loss.PLMN.QCI.1/L.Traffic.DL.PktUuLoss.Tot.PLMN.QCI.1 x 100%
Uplink packet loss rate on the Uu interface for QCI 1 in weak-coverage areas (uplink path loss
ranging within [135 dB, inf)) =
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1 x 100%
Uplink packet loss rate on the Uu interface for QCI 1 in weak-coverage areas (uplink path loss
ranging within [135 dB, 144 dB)) =
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1.Index0/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1.Index0 x 100%
1526728411 L.Voice.VQI.UL.Excellent.Times
1526728416 L.Voice.VQI.DL.Excellent.Times
1526732687 L.Voice.VQI.AMRWB.UL.Excellent.Times
1526732692 L.Voice.VQI.AMRWB.DL.Excellent.Times
Counter ID Counter Name
1526736660 L.Voice.VQI.UL.Excellent.Times.PLMN
1526736665 L.Voice.VQI.DL.Excellent.Times.PLMN
1526736670 L.Voice.VQI.AMRWB.UL.Excellent.Times.PLMN
1526736675 L.Voice.VQI.AMRWB.DL.Excellent.Times.PLMN
1526741669 L.Voice.E2EVQI.Excellent.Times
1526741674 L.Voice.E2EVQI.AMRWB.Excellent.Times
1526745678 L.Voice.VQI.EVS.UL.Excellent.Times
1526745683 L.Voice.VQI.EVS.DL.Excellent.Times
1526728412 L.Voice.VQI.UL.Good.Times
1526728417 L.Voice.VQI.DL.Good.Times
1526732688 L.Voice.VQI.AMRWB.UL.Good.Times
1526732693 L.Voice.VQI.AMRWB.DL.Good.Times
1526736661 L.Voice.VQI.UL.Good.Times.PLMN
1526736666 L.Voice.VQI.DL.Good.Times.PLMN
1526736671 L.Voice.VQI.AMRWB.UL.Good.Times.PLMN
1526736676 L.Voice.VQI.AMRWB.DL.Good.Times.PLMN
1526741670 L.Voice.E2EVQI.Good.Times
1526741675 L.Voice.E2EVQI.AMRWB.Good.Times
1526745679 L.Voice.VQI.EVS.UL.Good.Times
1526745684 L.Voice.VQI.EVS.DL.Good.Times
1526728413 L.Voice.VQI.UL.Accept.Times
1526728418 L.Voice.VQI.DL.Accept.Times
1526732689 L.Voice.VQI.AMRWB.UL.Accept.Times
Counter ID Counter Name
1526732694 L.Voice.VQI.AMRWB.DL.Accept.Times
1526736662 L.Voice.VQI.UL.Accept.Times.PLMN
1526736667 L.Voice.VQI.DL.Accept.Times.PLMN
1526736672 L.Voice.VQI.AMRWB.UL.Accept.Times.PLMN
1526736677 L.Voice.VQI.AMRWB.DL.Accept.Times.PLMN
1526741671 L.Voice.E2EVQI.Accept.Times
1526741676 L.Voice.E2EVQI.AMRWB.Accept.Times
1526745680 L.Voice.VQI.EVS.UL.Accept.Times
1526745685 L.Voice.VQI.EVS.DL.Accept.Times
1526728414 L.Voice.VQI.UL.Poor.Times
1526728419 L.Voice.VQI.DL.Poor.Times
1526732690 L.Voice.VQI.AMRWB.UL.Poor.Times
1526732695 L.Voice.VQI.AMRWB.DL.Poor.Times
1526732890 L.Voice.UL.LowQuality.Num
1526732891 L.Voice.DL.LowQuality.Num
1526736663 L.Voice.VQI.UL.Poor.Times.PLMN
1526736668 L.Voice.VQI.DL.Poor.Times.PLMN
1526736673 L.Voice.VQI.AMRWB.UL.Poor.Times.PLMN
1526736678 L.Voice.VQI.AMRWB.DL.Poor.Times.PLMN
1526741672 L.Voice.E2EVQI.Poor.Times
1526741677 L.Voice.E2EVQI.AMRWB.Poor.Times
1526745681 L.Voice.VQI.EVS.UL.Poor.Times
1526745686 L.Voice.VQI.EVS.DL.Poor.Times
1526747835 L.Voice.DL.LowQuality.EPC.Num
1526728415 L.Voice.VQI.UL.Bad.Times
1526728420 L.Voice.VQI.DL.Bad.Times
1526732691 L.Voice.VQI.AMRWB.UL.Bad.Times
1526732696 L.Voice.VQI.AMRWB.DL.Bad.Times
1526732892 L.Voice.UL.Silent.Num
1526732893 L.Voice.DL.Silent.Num
1526736664 L.Voice.VQI.UL.Bad.Times.PLMN
1526736669 L.Voice.VQI.DL.Bad.Times.PLMN
1526736674 L.Voice.VQI.AMRWB.UL.Bad.Times.PLMN
1526736679 L.Voice.VQI.AMRWB.DL.Bad.Times.PLMN
1526741673 L.Voice.E2EVQI.Bad.Times
1526741678 L.Voice.E2EVQI.AMRWB.Bad.Times
1526745682 L.Voice.VQI.EVS.UL.Bad.Times
1526745687 L.Voice.VQI.EVS.DL.Bad.Times
1526741679 L.Voice.E2EVQI.TotalValue
1526741680 L.Voice.E2EVQI.AMRWB.TotalValue
1526741681 L.Voice.VQI.UL.TotalValue
1526741682 L.Voice.VQI.DL.TotalValue
1526741683 L.Voice.VQI.AMRWB.UL.TotalValue
1526741684 L.Voice.VQI.AMRWB.DL.TotalValue
1526745688 L.Voice.VQI.EVS.UL.TotalValue
1526745689 L.Voice.VQI.EVS.DL.TotalValue
Table 4-23 lists the counters used to monitor the number of voice service UEs.
Table 4-23 Counters used to monitor the number of voice service UEs
Counter ID Counter Name
1526728456 L.Traffic.ActiveUser.DL.QCI.1
1526728446 L.Traffic.ActiveUser.UL.QCI.1
1526730601 L.Traffic.ActiveUser.DL.QCI.1.Max
1526730611 L.Traffic.ActiveUser.UL.QCI.1.Max
1526732721 L.Traffic.User.VoIP.Avg
1526732722 L.Traffic.User.VoIP.Max
1526736692 L.Traffic.User.VoIP.Avg.PLMN
1526736693 L.Traffic.User.VoIP.Max.PLMN
Table 4-24 lists the counters used to monitor the average number of PRBs occupied by voice
services.
Table 4-24 Counters used to monitor the average number of PRBs occupied by voice services
1526730883 L.ChMeas.PRB.DL.DrbUsed.Avg.VoIP
1526730884 L.ChMeas.PRB.UL.DrbUsed.Avg.VoIP
Throughput
Table 4-25and Table 4-26 list the counters used to monitor the number of PDCCH CCEs used by
voice services and total uplink/downlink traffic volume of voice services. Based on these
counters, you can calculate the average uplink/downlink throughput of voice services.
Table 4-25 Counters used to monitor the number of PDCCH CCEs used by voice services
1526736735 L.ChMeas.CCE.ULUsed.VoIP
1526736736 L.ChMeas.CCE.DLUsed.VoIP
Table 4-26 Counters used to monitor the total uplink/downlink traffic volume of voice services
Counter ID Counter Name
1526726776 L.Thrp.bits.UL.QCI.1
1526726803 L.Thrp.bits.DL.QCI.1
1526726777 L.Thrp.Time.UL.QCI.1
1526726804 L.Thrp.Time.DL.QCI.1
1526727849 L.Thrp.bits.UL.PLMN.QCI.1
1526727850 L.Thrp.Time.UL.PLMN.QCI.1
1526728050 L.Thrp.bits.DL.PLMN.QCI.1
1526728051 L.Thrp.Time.DL.PLMN.QCI.1
lists the counters used to monitor voice performance in non-standalone (NSA) dual
Table 4-27
connectivity (DC) scenarios.
Table 4-27 Counters used to monitor voice performance in NSA DC scenarios
Counter ID Counter Name
1526758973 L.NsaDc.Capable.E-RAB.AbnormRel.QCI.1
1526758974 L.NsaDc.Capable.E-RAB.NormRel.QCI.1
1526758975 L.NsaDc.Capable.Traffic.DL.PktUuLoss.Tot.QCI.1
1526758976 L.NsaDc.Capable.Traffic.DL.PktUuLoss.Loss.QCI.1
1526758977 L.NsaDc.Capable.E-RAB.SuccEst.QCI.1
1526758978 L.NsaDc.Capable.E-RAB.AttEst.QCI.1
1526758979 L.NsaDc.Capable.Traffic.UL.PktLoss.Loss.QCI.1
1526758980 L.NsaDc.Capable.Traffic.UL.PktLoss.Tot.QCI.1
1526758981 L.NsaDc.Capable.RRC.ReEst.Succ.QCI.1
QCI 1 service drop rate in NSA DC scenarios = L.NsaDc.Capable.E-
RAB.AbnormRel.QCI.1/(L.NsaDc.Capable.E-RAB.AbnormRel.QCI.1 + L.NsaDc.Capable.E-
RAB.NormRel.QCI.1) x 100%
If there are VoLTE service problems after VoLTE is deployed, troubleshoot as follows:
• FDD
Check:
1. Whether the RRC connection procedure is normal
2. Whether the attach procedure is normal
3. Whether VoLTE is registered normally in the IMS
4. Whether the VoLTE call request procedure is normal
Handle faults as follows:
▪ The results of 1 and 2 are abnormal.
The problem is caused by the E-UTRAN instead of VoLTE. In this case,
contact Huawei technical support.
▪ The results of 3 and 4 are abnormal.
Confirm the following parameter settings on the eNodeB side:
▪ RlcPdcpParaGroup.DiscardTimer for QCI 5 services is set to
DiscardTimer_Infinity.
▪ ENodeBAlgoSwitch.EutranVoipSupportSwitch is set to ON.
The subsequent procedure is as follows:
▪ If the settings are incorrect, modify the parameter settings and
perform the verification again.
▪ If the settings are correct, verify the counters in Table 4-28 to
check the packet loss rate of QCI 5 services.
▪ If QCI 5 packets are lost, contact Huawei technical
support.
▪ If QCI 5 packets are normal, the VoLTE fault is not
caused by the eNodeB. If this is the case, troubleshoot
the IMS, EPC, or transport network.
Table 4-28 Packet loss rate of QCI 5 services
Counter ID Counter Name
1526727973 L.Traffic.UL.PktLoss.Loss.QCI.5
1526727974 L.Traffic.UL.PktLoss.Tot.QCI.5
1526727946 L.Traffic.DL.PktUuLoss.Loss.QCI.5
1526727947 L.Traffic.DL.PktUuLoss.Tot.QCI.5
5 Capacity Enhancement
5.1.1 Principles
This section describes VoIP semi-persistent scheduling and the power control policies used in
this scheduling mode.
5.1.1.1 VoIP Semi-Persistent Scheduling
Introduction
When dynamic scheduling is used for voice services, the time-frequency resources or MCS is
updated through the PDCCH every 20 ms. This consumes a large volume of PDCCH resources.
Figure 5-1 shows the resource allocation for dynamic scheduling.
The eNodeB configures semi-persistent scheduling parameters for UEs that support semi-
persistent scheduling in RRC Connection Reconfiguration messages during setup of QCI 1 data
radio bearers (DRBs). The eNodeB activates uplink or downlink semi-persistent scheduling for
UEs when the UEs meet the activation conditions. The eNodeB notifies UEs of the activation
using the PDCCH Order. For details on the format of the PDCCH Order notification, see section
9.2 "PDCCH/EPDCCH validation for semi-persistent scheduling" in 3GPP TS 36.213 V12.3.0.
Effect Period
Figure 5-3 Semi-persistent scheduling effect period
• During talk spurts, semi-persistent scheduling takes effect only when all of the
following conditions are met:
▪ Switches specified by the following parameters are turned on:
▪ For the uplink, SpsSchSwitch of CellAlgoSwitch.UlSchSwitch
▪ For the downlink, SpsSchSwitch of CellAlgoSwitch.DlSchSwitch
▪ For emergency call services, CellAlgoSwitch.EmcSpsSchSwitch
and the preceding two switches
▪ The UE supports semi-persistent scheduling.
▪ If multiple bearers are set up for the UE, semi-persistent scheduling takes
effect only when there is only one voice bearer, one bearer for voice SIP
messages, and no more than one data service bearer. If there is one data
service bearer, data must not be transmitted on it.
▪ RLC segmentation is not performed in the uplink for the UE.
▪ When ROHC is enabled, the uplink or downlink ROHC is in the stable
compression state, essentially, the size of each ROHC header is relatively
stable.
• During talk spurts, dynamic scheduling takes effect only when all of the following
conditions are met:
▪ Large packets are to be scheduled during semi-persistent scheduling.
Examples of large packets include channel-associated signaling and
uncompressed packets generated when ROHC updates contexts.
▪ In the downlink, semi-persistently scheduled data is to be retransmitted.
▪ In the uplink, semi-persistently scheduled data is to be retransmitted in an
adaptive manner.
▪ The CellAlgoSwitch.EmcSpsSchSwitch parameter is set to OFF in
emergency call service scenarios.
When the UE uses semi-persistent scheduling, the highest MCS index is 15.
When activating uplink semi-persistent scheduling, the eNodeB determines the MCS and PRBs
to be allocated during the scheduling, based on the following factors:
• Voice packet size (if ROHC is disabled) or compressed voice packet size (if ROHC is
enabled)
• Adjusted wideband signal to interference plus noise ratio (SINR)
After uplink semi-persistent scheduling is activated, the UE periodically sends data and the
eNodeB periodically receives the data using the uplink semi-persistently allocated resources. The
eNodeB also checks whether the allocated MCS is suitable under the current channel conditions.
If the MCS is not suitable under the current channel conditions, the eNodeB reactivates semi-
persistent scheduling.
The eNodeB uses the logicalChannelSR-Mask-r9 IE in the RRC Connection Reconfiguration
message to instruct the UE not to send scheduling requests over the radio bearers with QCI 1.
This reduces UE power consumption. The CellUlschAlgo.SrMaskSwitch parameter controls this
function. It is recommended that this function be enabled when uplink semi-persistent scheduling
is enabled. This function takes effect only on UEs that comply with 3GPP Release 9 or later
versions.
When the number of zero-payload packets received by the eNodeB on the semi-persistently
allocated resources exceeds the value of CellUlschAlgo.SpsRelThd, the eNodeB automatically
releases these resources.
When uplink semi-persistent scheduling is used for voice services and uplink PRBs are
sufficient, it is recommended that the FIRST_RETRANS_EXPN_RB_SWITCH option of the
CellUlschAlgo.UlVoLTERetransSchStrategy parameter be selected to reduce the voice packet loss
rate and improve voice user experience. If this option is selected for a cell, the eNodeB evaluates
whether to use RB-increased adaptive retransmission from the first uplink retransmission
onwards. The evaluation is based on the transmit power of voice service UEs.
When activating downlink semi-persistent scheduling, the eNodeB assigns an MCS and PRBs to
the UE based on the following factors before the eNodeB and UE send and receive data using the
allocated resources:
• Voice packet size (if ROHC is disabled) or compressed voice packet size (if ROHC is
enabled)
• UE-reported wideband CQIs
After downlink semi-persistent scheduling is activated, the eNodeB checks whether the allocated
MCS is suitable for the channel conditions. If the MCS is not suitable for the channel conditions,
the eNodeB reactivates semi-persistent scheduling. There are two possible scenarios:
• If the periodically measured IBLER is greater than or equal to the value of
CellDlschAlgo.DlSpsMcsDecreaseIblerThd, the eNodeB lowers the MCS index and
reactivates downlink semi-persistent scheduling.
• If the periodically measured IBLER is lower than 5%, the eNodeB increases the MCS
index and reactivates downlink semi-persistent scheduling. Downlink MCS index
increase for downlink semi-persistent scheduling is controlled by the
DlSpsMcsIncreaseSwitch option of the CellAlgoSwitch.CqiAdjAlgoSwitch parameter.
In accordance with 3GPP TS 36.321 and TS 36.331, the eNodeB reserves HARQ processes for
downlink semi-persistent scheduling while configuring semi-persistent scheduling for UEs.
Before downlink semi-persistent scheduling is activated, reserved HARQ processes can be used
for dynamic scheduling to increase the number of HARQ processes available for data services.
This function is controlled by the DlSpsRevHarqUseSwitch option of the
CellDlschAlgo.DlEnhancedVoipSchSw parameter.
When the eNodeB configures downlink semi-persistent scheduling for UEs, the PUCCH requires
available downlink semi-persistent code channels for HARQ feedback.
When downlink semi-persistent scheduling is used for voice services and downlink PRBs are
sufficient, it is recommended that the DlVoiceRetransOptSwitch option of the
CellDlschAlgo.DlEnhancedVoipSchSw parameter be selected to reduce the voice packet loss rate
and improve voice user experience. If this option is selected for a cell, the eNodeB uses RB-
increased adaptive mode of retransmission in the downlink.
Voice services can use two scheduling methods: dynamic and semi-persistent (compared in Table
5-1).
For newly initiated voice services, scheduling based on TTI-level UE number allows the base
station to adaptively select dynamic or semi-persistent scheduling. The selection is based on the
number of UEs scheduled per TTI in a cell as follows:
• Adaptive selection of scheduling methods
▪ If the load is high, the eNodeB selects semi-persistent scheduling to avoid
PDCCH overload and a negative impact on voice quality and capacity.
▪ If the load is low, the eNodeB selects dynamic scheduling to provide better
voice experience and improve spectral efficiency.
• The eNodeB determines the cell load based on the number of UEs scheduled per TTI.
When there are a large number of UEs scheduled per TTI, the eNodeB determines that
the cell load is high. When there are a small number of UEs scheduled per TTI, the
eNodeB determines that the cell load is low.
• Scheduling in the uplink and downlink is controlled by the following options:
UlVoIPLoadBasedSchSwitch option of the CellUlschAlgo.UlEnhencedVoipSchSw
parameter
DlVoIPLoadBasedSchSwitch option of the CellDlschAlgo.DlEnhancedVoipSchSw
parameter
For scheduling based on TTI-level UE number, semi-persistent scheduling can be adaptively selected only when
semi-persistent scheduling is enabled.
UE Blacklist
Semi-persistent scheduling can be disabled for blacklisted UE types. Table 5-2 describes the
blacklist configurations.
Table 5-2 Blacklist configurations
For CCE-load-based scheduling, semi-persistent scheduling can be adaptively selected only when semi-
persistent scheduling is enabled.
This section describes power control policies of voice services during semi-persistent scheduling.
For details on power control, see Power Control.
If downlink semi-persistent scheduling is used, power control is not performed. Instead, the
eNodeB transmit power is evenly distributed on the RBs.
If uplink semi-persistent scheduling is used, the CloseLoopSpsSwitch option of the
CellAlgoSwitch.UlPcAlgoSwitch parameter specifies whether to enable closed-loop power control
for the PUSCH in semi-persistent scheduling mode.
• If this option is selected, the eNodeB adjusts the PUSCH transmit power based on the
measured IBLER of voice services.
• If this option is deselected, the eNodeB uses open-loop power control for the PUSCH
instead of using closed-loop power control for the PUSCH in semi-persistent
scheduling mode.
The SpsAndDrxOptSwitch option of the CellUlschAlgo.UlEnhencedVoipSchSw parameter
specifies whether to enable optimized collaboration between semi-persistent scheduling and
discontinuous reception (DRX).
• Option selected
▪ If the TPC-PUSCH-RNTI request fails, semi-persistent scheduling does not
take effect for UEs. This restriction avoids packet loss if the eNodeB cannot
send semi-persistent TPC commands.
▪ If DRX is enabled, the occasions for sending semi-persistent TPC
commands must be confined to the On Duration. If these commands are
sent outside the On Duration, dynamic scheduling instead of semi-persistent
scheduling can take effect for UEs. This restriction avoids packet loss if the
UEs cannot receive semi-persistent TPC commands from the eNodeB.
• Option deselected
▪ If the TPC-PUSCH-RNTI request fails, semi-persistent scheduling may still
take effect for UEs but the packet loss rate may increase.
▪ If DRX is enabled, semi-persistent TPC commands may be sent outside the
On Duration but the packet loss rate may increase.
If the CloseLoopSpsSwitch option is deselected, the TPC-PUSCH-RNTI corresponding to the PUSCH cannot be
requested. If the SpsAndDrxOptSwitch option is selected in this situation, uplink semi-persistent scheduling cannot
take effect for UEs.
5.1.2.1 Benefits
VoIP semi-persistent scheduling and power control in semi-persistent scheduling can be enabled
to increase capacity for voice services in capacity-limited scenarios (for example, with a
maximum CCE usage of greater than or equal to 50%).
• When semi-persistent scheduling is enabled, PDCCH resources do not hinder voice
service capacity because they are only used when semi-persistent scheduling is
initially activated or reactivated, or when semi-persistently allocated resources are
released. Semi-persistent scheduling allows the eNodeB to serve more voice service
UEs than dynamic scheduling. The VoLTE user number increases by about more than
30% in simulation case 1, according to the simulation prerequisites and capacity
evaluation method specified in appendix A of 3GPP TS 36.814.
• When enhanced VoIP semi-persistent scheduling is enabled, the Uplink Packet Loss Rate
(VoIP) decreases by up to 10% in capacity-limited scenarios, and the Downlink Packet
Loss Rate (VoIP) decreases by up to 10% in PDCCH coverage-limited scenarios (with a
PDCCH DTX proportion of greater than or equal to 5%).
• When optimized collaboration between semi-persistent scheduling under power
control and DRX is enabled, the packet loss rate decreases and voice service quality
improves for semi-persistently scheduled UEs in the DRX state.
You are advised to evaluate when to use the function according to When to Use VoIP Semi-Persistent
Scheduling and When to Use Power Control in Semi-Persistent Scheduling to achieve optimal benefits.
Cells working with 1.4 MHz bandwidth have only six PRBs. The highest MCS index that the
eNodeB can select during semi-persistent scheduling is 15. This means that UEs in the cell
center or at a place of a medium distance to the cell center consume more PRBs, and waste PRB
resources. It is recommended that uplink and downlink semi-persistent scheduling and power
control in these scheduling situations be disabled in 1.4 MHz cells.
PDCCH resources are saved because VoIP semi-persistent scheduling uses a fixed MCS and
PRBs to transmit data during talk spurts, instead of using the PDCCH to update the MCS and
PRBs at each TTI. Therefore, VoIP semi-persistent scheduling is recommended in scenarios with
low-speed UEs or other scenarios where there are no rapid changes in channel conditions. Do not
use VoIP semi-persistent scheduling in high-speed or ultra-high speed cells. These UEs move
fast and channel conditions change rapidly.
You are advised to enable this function to save CCE resources in a cell serving low-speed UEs or
having slow channel condition changes, when the following conditions are met.
• Uplink semi-persistent scheduling is recommended when all of the following
conditions are met:
▪ Cell bandwidth ≤ 10 MHz (5 MHz is recommended)
▪ L.Traffic.User.VoIP.Avg/L.Traffic.User.Avg > 5%
▪ Duration in which the maximum CCE usage is greater than 50% within one
day ≥ 5 hours
▪ Uplink Packet Loss Rate (VoIP) > 0.5%
▪ (L.ChMeas.CCE.ULUsed.TA + L.ChMeas.CCE.ULUsed.VoIP +
L.ChMeas.CCE.ULUsed.SRB)/L.ChMeas.CCE.ULUsed x 100% ≥ 30%
Network Impacts
• More RBs need to be allocated to semi-persistently scheduled UEs in the cell center,
because the highest MCS index available during VoIP semi-persistent scheduling is
only 15. If, for example, the voice rate is 23.85 kbit/s (without considering ROHC)
and MCS index 26 is selected for voice service UEs in the cell center during dynamic
scheduling, two RBs are required in the uplink and one RB in the downlink (dual-
stream mode). Under the same conditions, semi-persistent scheduling requires three
RBs in both the uplink and downlink because the maximum MCS index is 15. Semi-
persistent scheduling for voice service UEs in the cell center allocates more RBs than
dynamic scheduling. Consequently, fewer RBs are available for data services than
when VoIP semi-persistent scheduling is disabled, and the traffic volume of data
services may also decrease.
• After voice services are set up for UEs, the eNodeB reserves HARQ processes for
downlink semi-persistent scheduling, if enabled. The reserved HARQ processes
ensure that periodic data transmissions as a result of semi-persistent scheduling still
succeed even if HARQ processes are still running for retransmission of the previous
semi-persistently scheduled data when the next semi-persistently scheduled initial
transmission starts. However, these reserved HARQ processes cannot be used during
dynamic scheduling for the UEs to perform other types of services, such as data
services. When a single UE performs both voice and data services, the HARQ
processes available to data services decrease.
If there is large UE data traffic (such as full buffer services), the throughput of data
services will be lower than when dynamic scheduling is used for both voice and data
services.
If there are multiple UEs or there is not as much data traffic, reserving HARQ
processes has little impact because the scheduling opportunities for the UEs are
scattered or only a few HARQ processes are required for the data services.
• Fixed-position resource allocation is used after VoIP semi-persistent scheduling is
activated. Semi-persistent scheduling may have a longer scheduling waiting time and
higher voice packet loss rate than dynamic scheduling.
• If the conditions in the applicable scenarios described in When to Use VoIP Semi-Persistent
Scheduling are not met, uplink and downlink semi-persistent scheduling may cause the
Uplink Packet Loss Rate (VoIP) and Downlink Packet Loss Rate (VoIP) to deteriorate,
respectively.
Function Impacts
FDD High speed ProcSwitchBasedOnUserSpeed option of High Speed Uplink VoIP semi-pe
specified the CellAlgoSwitch.HighSpeedSchOptSwitch Mobility scheduling, if enable
policy parameter high speed specified
management management, takes
speed UEs or when
become moving at lo
FDD Dynamic PRB_DYNAMIC_MGMT_SW option of the Dynamic LTE cells do not use
multi-carrier NbPrbDynamicMgmt.NbPrbDynMgmtAlgoSw Multi-Carrier semi-persistent sche
Management
management parameter (FDD)
enabled for
individual
cells.
scheduling
degradation switchc
is on, downlink
semi-persistent
scheduling will not
be performed for
UEs and semi-
persistent
scheduling that has
taken effect for UEs
will be deactivated.
FDD eMTC EMTC_SWITCH option of the eMTC eMTC UEs do not support
introduction CellEmtcAlgo.EmtcAlgoSwitch uplink VoIP semi-persistent
parameter scheduling.
5.1.3 Requirements
5.1.3.1 Licenses
5.1.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
5.1.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
5.1.3.4 Others
UEs must support VoLTE, semi-persistent scheduling, and closed-loop power control in semi-
persistent scheduling. The EPC must support IMS-based voice services.
Bit 29 of the feature group indicator (FGI) field indicates UE support for semi-persistent
scheduling.
5.1.4 Operation and Maintenance
automat
releases
these
resource
SPS
Activation
Thld
UE Blacklist
Schedule semi-persistent
Switch scheduling and DRX.
With this option
selected, TPC
commands for semi-
persistently scheduled
UEs can be sent only
within the DRX-
defined On Duration. If
semi-persistent
scheduling is enabled,
optimized
collaboration between
semi-persistent
scheduling and DRX
must also be enabled.
NOTICE:
Changing the
SpsAndDrxOptSwitch
parameter setting for a cell
will cause the cell to restart.
Before using MML commands, refer to 5.1.2.2 Impacts and 5.1.3.2 Software and complete the
parameter configurations for related functions based on the impact, dependency, and mutually
exclusive relationships between the functions, as well as the actual network scenario.
//Enabling uplink semi-persistent scheduling
MOD CELLALGOSWITCH: LocalCellId=0, UlSchSwitch=SpsSchSwitch-1;
//Setting parameters related to uplink semi-persistent scheduling
MOD CELLULSCHALGO: LocalCellId=0, SpsRelThd=2, UlSpsInterval=ms20,
SrMaskSwitch=ON, UlVoLTERetransSchStrategy=FIRST_RETRANS_EXPN_RB_SWITCH-1;
//Setting parameters related to uplink robust semi-persistent scheduling
MOD VOLTEALGOCONFIG: LocalCellId=0, VolteOptSwitch=UL_ROBUST_SPS_SW-1,
LinkQltyBasedUlSpsActvThld=127, UlSpsMcsDecreaseIblerThld=20;
MOD CELLALGOSWITCH: LocalCellId=0, UlSchSwitch=SchedulerCtrlPowerSwitch-1;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling uplink semi-persistent scheduling
MOD CELLALGOSWITCH: LocalCellId=0, UlSchSwitch=SpsSchSwitch-0;
To verify uplink semi-persistent scheduling for voice services, perform the following steps:
1. Run the LST CELLALGOSWITCH command to check whether the uplink semi-
persistent scheduling switch has been turned on.
LST CELLALGOSWITCH: LocalCellId=0;
g.
h. If L.Sps.UL.ErrNum is greater than 10% of L.Sps.UL.SchNum, check whether the
CloseLoopSpsSwitch option of the CellAlgoSwitch.UlPcAlgoSwitch parameter has
been selected.
• If this option is deselected, select this option.
• If this option is selected, contact Huawei technical support.
To verify downlink semi-persistent scheduling for voice services, perform the following steps:
1. Run the LST CELLALGOSWITCH command to check whether the downlink
semi-persistent scheduling switch has been turned on.
LST CELLALGOSWITCH: LocalCellId=0;
To check whether IBLER values can converge when closed-loop power control in PUSCH semi-
persistent scheduling is enabled, perform the following steps:
1. Run the LST CELLALGOSWITCH command to check whether closed-loop power
control in PUSCH semi-persistent scheduling has been enabled.
2. Use a UE to access the cell and initiate an uplink voice service.
3. Start a task on the MAE-Access to monitor IBLER values.
a. On the MAE-Access, choose Monitor > Signaling Trace > Signaling
Trace Management.
b. In the left pane of the displayed window, choose User Performance
Monitoring > BLER Monitoring. Set the tracing duration and MMEc
(MME ID). Start uplink IBLER monitoring.
c. On the MAE-Access, check whether the IBLER values converge, as
shown in Figure 5-7. If the values of Uplink IBLER (Permillage) are less
than 100 (that is, the IBLER values are less than 10%), the IBLER values
have converged. If the UE is close to the eNodeB, the IBLER values are
relatively small because of favorable channel quality. If the UE is far
from the eNodeB, the IBLER values are relatively large because of
limited power and may not converge.
Figure 5-7 Uplink IBLER monitoring results
2. Observe the following counters to check whether the effect of adaptive switching
between dynamic scheduling and semi-persistent scheduling is proper.
Semi-persistent scheduling reduces PDCCH consumption, increases voice service capacity, but
causes a slight decrease in voice service quality.
You can compare counters in the following table given the same number of UEs in a cell, UE
location, and traffic model. If the values of these counters decrease after semi-persistent
scheduling is enabled, PDCCH consumption has been reduced.
1526728304 L.ChMeas.CCE.ULUsed
1526728305 L.ChMeas.CCE.DLUsed
For monitoring on voice service capacity and voice quality, see 4.4.3.4 Voice Capacity and 4.4.3.3
Voice Quality.
Power control in semi-persistent scheduling ensures the voice quality of voice service UEs. For
details, see 4.4.3.3 Voice Quality.
Power control in semi-persistent scheduling must be used together with semi-persistent scheduling.
5.2 ROHC
Robust header compression (ROHC) provides an efficient header compression mechanism for
data packets transmitted on radio links. ROHC helps reduce the impact of high bit error rates
(BERs) and the long round trip time (RTT). ROHC helps reduce header overheads and packet
loss, and shortens the response time.
Currently, ROHC is used to compress the headers of only voice packets, as shown in Figure 5-8.
ROHC reduces the packet size and PRB consumption. When PRBs are insufficient, ROHC helps
increase system capacity.
Figure 5-8 ROHC for voice services
After deploying VoLTE, operators can enable or disable ROHC using the
CellAlgoSwitch.RohcSwitch parameter. ROHC is an extensible framework consisting of different
profiles for data streams compliant with different protocols. Profiles define the compression
modes for streams with different types of protocol headers. Voice services use profiles 0x0001
and 0x0002.
The size of compressed voice packets varies because the ROHC compression efficiency depends
on the ROHC operating mode and the change in the dynamic part of packet headers at the
application layer. A header can be compressed to a size as small as 1 byte, which efficiently
reduces the voice packet size.
For more details about ROHC principles and engineering guidelines, see ROHC.
6 Coverage Optimization
6.1.1 Principles
TTI bundling enables a data block to be transmitted in four consecutive TTIs, which are bound
together and treated as the same resource. Different HARQ redundancy versions of the same data
block are transmitted in different TTIs. TTI bundling makes full use of HARQ combining gains
and reduces the number of retransmissions and RTT.
When a cell-edge UE suffers from poor channel quality and is allocated limited transmit power,
TTI bundling increases the cell edge coverage of the PUSCH. Only in FDD mode, TTI bundling,
as defined in 3GPP Release 12 (referred to as R12 TTI bundling), further increases the cell edge
coverage of the PUSCH. Specifically, this function improves the uplink quality of cell-edge
voice service UEs with limited transmit power. The gains produced by this function can be
observed when voice quality is maintained at a certain level, for example, when the mean
opinion score (MOS) is 3.
TTI bundling is controlled by the TtiBundlingSwitch option of the CellAlgoSwitch.UlSchSwitch
parameter. When this option is selected, the eNodeB determines whether a UE enters the TTI
bundling state based on the channel conditions. When a UE enters the TTI bundling state, the
eNodeB determines the number of PRBs and selects an MCS based on the channel quality and
the amount of data to be transmitted.
6.1.1.1 Protocol Compliance (FDD)
The sub-functions supported by TTI bundling vary with the protocol release and depend on the
CellTtiBundlingAlgo.R12TtiBundlingSwitch parameter setting:
• Set to OFF
▪ According to section 8.6.1 "Modulation order and redundancy version
determination" in 3GPP TS 36.213 of Release 10, when TTI bundling is
enabled, the modulation scheme must be QPSK and the resource allocation
size is restricted to a maximum of three PRBs.
▪ The highest MCS index allowed is 10, and the TBS can be up to 504 bits.
Voice services are delay-sensitive. If higher-layer data is not transmitted
within the specified delay budget, voice quality deteriorates.
• Set to ON
▪ In accordance with 3GPP TS 36.213 CR0463 of Release 12, when TTI
bundling is enabled, the modulation scheme must be QPSK but the resource
allocation size is no longer limited to three PRBs.
▪ In accordance with 3GPP TS 36.213 CR0463 of Release 12, the new
HARQ feedback time sequence shown in Figure 6-1 is supported, and the
number of HARQ processes decreases from four to three. This function is
referred to as R12 TTI bundling eHARQ in the following sections.
▪ The highest MCS index allowed is no longer 10.
6.1.1.2 TTI Bundling Entry Conditions
The UE transmits identical data within a bundle of four consecutive TTIs and performs HARQ
retransmission also within a bundle of four consecutive TTIs. The four uplink subframes
correspond to only one piece of feedback on the PHICH. The retransmission interval changes
from 8 TTIs (normal HARQ RTT) to 16 TTIs (bundle HARQ RTT).
Assume that the last TTI is numbered N among the TTIs bundled for transmitting a data block.
The eNodeB sends an ACK or NACK as feedback to the UE in the (N + 4)th TTI. Based on the
feedback, the UE determines whether a retransmission is required. If it is required, the
retransmission policy varies depending on the CellTtiBundlingAlgo.R12TtiBundlingSwitch
parameter setting:
• If this parameter is set to ON, the UE retransmits identical data blocks in the (N + 9)th
through (N + 12)th TTIs (the green arrow in Figure 6-1). In this situation, the maximum
number of uplink HARQ retransmissions is specified by the
CellTtiBundlingAlgo.R12TtiBHarqMaxTxNum parameter.
• If this parameter is set to OFF, the UE retransmits identical data blocks in the (N +
13)th through (N + 16)th TTIs (the orange arrow in Figure 6-1). In this situation, the
maximum number of uplink HARQ retransmissions is specified by the
CellUlschAlgo.TtiBundlingHarqMaxTxNum parameter.
Figure 6-1 TTI bundling
When a UE enters the TTI bundling state, the number of RLC segments of a given voice packet
cannot exceed the value of CellUlschAlgo.TtiBundlingRlcMaxSegNum. In the example shown in
Figure 6-2, the value of this parameter is 4.
When the UE is located at the cell edge, RLC segmentation in collaboration with TTI bundling
produces fewer RLC segments than RLC segmentation without TTI bundling. This reduces
PDCCH overheads.
When TTI bundling is enabled, it is recommended that the
RlcPdcpParaGroup.ENodeBReorderingTimerAdapt parameter be set to ON. This setting reduces
the voice packet loss caused by improper configurations of the reordering timer at the eNodeB,
the number of additional RLC retransmissions, and the number of RLC status reports, saving
radio resources.
6.1.1.4 TTI Bundling Exiting Conditions
The eNodeB instructs the UE to exit the TTI bundling state when the following exiting
conditions are met:
• FDD
▪ If voice services have not been released:
The policy for exiting the TTI bundling state varies depending on the
CellTtiBundlingAlgo.SinrThdToTrigTtib parameter setting:
▪ If this parameter is set to a value other than 255, the eNodeB
sends the UE an RRC Connection Reconfiguration message,
instructing the UE to exit TTI bundling when the measured SINR
is greater than the sum of the target SINR and the value of
CellUlschAlgo.HystToExitTtiBundling for a number of times equal
to CellUlschAlgo.StatisticNumThdForTtibExit.
▪ If this parameter is set to 255, the eNodeB evaluates exit from
TTI bundling only after it receives voice frames during talk
spurts. If the measured SINR is greater than the sum of the target
SINR and the value of CellUlschAlgo.HystToExitTtiBundling for a
number of times equal to the value of
CellUlschAlgo.StatisticNumThdForTtibExit, the eNodeB sends the
UE an RRC Connection Reconfiguration message, instructing the
UE to exit the TTI bundling state.
▪ If voice services have been released, the eNodeB compares the measured
SINR with the smaller value between 6 dB and the sum of the target SINR
and the value of CellUlschAlgo.HystToExitTtiBundling. If the measured SINR
is greater than that smaller value for a number of times equal to the value of
CellUlschAlgo.StatisticNumThdForTtibExit, the eNodeB sends the UE an RRC
Connection Reconfiguration message, instructing the UE to exit the TTI
bundling state.
The TTIB_EXIT_BASED_ON_QCI_SW option of the
CellTtiBundlingAlgo.TtiBundlingAlgoSw parameter specifies whether to enable exit
from the TTI bundling state based on QCIs. If this option is selected and a UE
accesses both voice and data services, the UE exits the TTI bundling state after the
voice service is released. It no longer uses TTI bundling for the data services.
Selecting this option in weak-coverage scenarios may cause a decrease in the RRC
connection reestablishment success rate and an increase in the service drop rate.
After a UE enters the TTI bundling state, the eNodeB does not instruct the UE to exit this state when any of the
following occurs:
• A data service is transmitted on the default bearer.
• A new dedicated bearer is set up.
• The QCI 1 voice service is released.
Instead, the eNodeB instructs the UE to exit the TTI bundling state when either of the following occurs:
• The UE meets the exiting conditions.
• The UE experiences a handover, service drop, or RRC connection reestablishment.
In weak-coverage scenarios, after the UE exits the TTI bundling state due to a handover or RRC
connection reestablishment, the eNodeB must evaluate entry into TTI bundling again before the
UE can return to the TTI bundling state.
The enhanced TTI bundling algorithm is enabled if the
TTIBUNDLING_ALGO_ENHANCE_SW option of the
CellTtiBundlingAlgo.TtiBundlingAlgoSw parameter is selected. With this algorithm enabled:
• During a handover or RRC connection reestablishment, the TTI bundling state in the
source cell applies to the target cell. Essentially, if a UE is in the TTI bundling state
before a handover or RRC connection reestablishment, the UE enters the TTI bundling
state after the handover or RRC connection reestablishment without another TTI
bundling evaluation, reducing reconfiguration signaling and improving voice quality.
This function requires that TTI bundling be enabled in the target cell of handover or RRC connection
reestablishment.
6.1.2.1 Benefits
Scenario Description
L.Traffic.User.VoIP.Avg/L.Traffic.User.Avg> 5%
• The percentage of uplink voice packets from the cell edge meets
the following requirement:
L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1/L.Traffic.UL.PktLoss.Tot.QCI.1 > 5%
• The loss rate of uplink voice packets from the cell edge meets the
following requirement:
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1
> 2%
1526727378 L.Traffic.User.Avg
Sites with a high traffic volume of voice services at cell edges 1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.Q
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
Cells with a high packet loss rate for voice service UEs at the cell 1526745995 L.Traffic.UL.FarUE.PktLoss.Loss
edge
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1
1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.Q
6.1.2.2 Impacts
Network Impacts
▪ A UE in the TTI bundling state uses a high MCS index. This increases the
probability of DTX for PDSCH HARQ feedback and increases Downlink
Packet Loss Rate (VoIP).In this situation, you are advised to change the value
of CellUciOnPuschPara.DeltaOffsetAckIndexForTtiB from its default value 9
to 11 so that Downlink Packet Loss Rate (VoIP) does not increase much.
▪ TTI bundling uses a maximum of three PRBs, a modulation scheme of
QPSK, and an MCS index no higher than 10. When TTI bundling is
enabled, no more than 504 bits can be transmitted within each TB, which
restricts the uplink throughput of UEs in the TTI bundling state. As
signaling and voice services are given precedence over data on logical
channels, the UEs will preferentially transmit signaling and voice services,
further restricting the uplink throughput of data services.
• (FDD) The enhanced TTI bundling algorithm has the following impacts on network
performance:
▪ The highest MCS index is no longer confined to 10. This may further
increase the probability of DTX for PDSCH HARQ feedback. As a result,
the Downlink Packet Loss Rate (VoIP) increases.
▪ After RRC connections are reestablished, UEs stay in the TTI bundling
state, during which a high MCS index is used. This increases the probability
of DTX for PDSCH HARQ feedback. As a result, the Downlink Packet Loss
Rate (VoIP) increases.
Enhanced TTI bundling has better coverage performance than common TTI
bundling. However, enhanced TTI bundling uses a high MCS index, which
increases the probability of DTX for PDSCH HARQ feedback. Therefore,
enhanced TTI bundling is recommended only when the uplink packet loss
rate is high and the downlink packet loss rate is low.
• (FDD) When TTI bundling is enabled, setting
CellTtiBundlingAlgo.R12TtiBundlingSwitch to ON has the following impacts on
network performance:
The highest MCS index is no longer confined to 10. This may further increase the
probability of DTX for PDSCH HARQ feedback. As a result, the Downlink Packet Loss
Rate (VoIP) increases.
Function Impacts
CellDrxPa
CellDrxPa
parameter
• If a UE is in
instructs th
after activa
when the U
measurem
CellDrxPa
CellDrxPa
parameter
FDD High speed ProcSwitchBasedOnUserSpeed option of High Speed TTI bundling, if enabl
specified the CellAlgoSwitch.HighSpeedSchOptSwitch Mobility specified policy mana
policy parameter low-speed UEs or wh
management from fast to slow.
FDD Cell radius Cell.CellRadius Extended Cell Extended cell range leads to the
greater than Range coverage limitation in the uplink.
15 km TTI bundling enhances PUSCH
coverage and increases the
MCS indexes in uplink weak
coverage areas, enhancing
uplink coverage and improving
coverage quality.
FDD Cell radius Cell.CellRadius Extended Cell Extended cell range leads to the
greater than Range coverage limitation in the uplink.
100 km TTI bundling enhances PUSCH
coverage and increases the
MCS indexes in uplink weak
coverage areas, enhancing
uplink coverage and improving
coverage quality.
• TTI bundling
takes
precedence for
individual UEs.
FDD Breathing BreathingPilotSwitch option of the Breathing If UEs are working in TTI
Pilot CellDlschAlgo.BreathingPilotAlgoSwitch Pilot bundling mode, they
parameter cannot enter DRX mode,
lowering the gains offered
by Breathing Pilot.
enabled f
individual
▪ These two
functions
take effec
simultane
for individ
UEs.
FDD QCI-specific The LowDelayServiceOptSwitch Air Interface RRC messages are required to
TTI bundling option of the Latency
Optimization the entry and exit of QCI-spec
CellAlgoSwitch.ServiceDiffSwitch bundling. Therefore, the RRC
parameter and the signaling in the cell increases.
QCI_TTI_BUNDLING_SWITCH Enabling both VoLTE and QCI
option of the TTI bundling (including 3GPP
CellQciPara.QciAlgoSwitch 12-compliant TTI bundling) pro
parameter the following impacts:
• When a UE enters t
VoLTE TTI bundling
and then performs a
latency service, or w
UE enters the QCI-s
TTI bundling state a
performs a VoLTE s
the TTI bundling sta
retained. In this cas
eNodeB determines
whether the UE exit
TTI bundling state b
the VoLTE TTI bund
conditions, unless th
VoLTE service has
released.
• When a UE that is n
TTI bundling state
RAT Function Function Switch Reference Description
Name
simultaneously perfo
low-latency service
VoLTE service, the
determines whether
enters the TTI bund
state based on the V
TTI bundling entry
conditions.
FDD Video TTI TtiBundlingForVideoSwitch option Video This function requires RRC me
bundling of the CellAlgoSwitch.UlSchSwitch Experience to trigger it, which increases R
Optimization
parameter signaling interactions in cells. E
VoLTE TTI bundling and video
bundling (including 3GPP Rele
compliant TTI bundling) produ
following impacts:
• If a UE enters the V
TTI bundling state a
performs video serv
remains in the VoLT
bundling state. The
conditions for video
bundling are used fo
bundling evaluation
after the VoLTE ser
are released and the
exits the VoLTE TTI
bundling state.
• If the UE enters the
TTI bundling state a
performs VoLTE ser
remains in the video
bundling state. The
conditions for VoLTE
RAT Function Function Switch Reference Description
Name
FDD eMTC EMTC_SWITCH option of the eMTC eMTC UEs do not support
introduction CellEmtcAlgo.EmtcAlgoSwitch TTI bundling.
parameter
6.1.3 Requirements
6.1.3.1 Licenses
6.1.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
6.1.3.3 Hardware
FDD:
• 3900 and 5900 series base stations
• DBS3900 LampSite and DBS5900 LampSite
Boards
RF Modules
No requirements
6.1.3.4 Others
• R8 TTI bundling requires UEs to be compliant with 3GPP Release 8. Bit 28 of FGI in
the UE-EUTRA-Capability IE of the UECapabilityInformation message indicates this
UE capability.
• R12 TTI bundling requires UEs to be compliant with 3GPP Release 12. The e-HARQ-
Pattern-FDD-r12 and noResourceRestrictionForTTIBundling-r12 IEs in the UE-
EUTRA-Capability IE of the UECapabilityInformation message indicate this UE
capability.
The e-HARQ-Pattern-FDD-r12 IE indicates whether a UE supports R12 TTI bundling
eHARQ, and the noResourceRestrictionForTTIBundling-r12 IE indicates whether the
resource allocation size supported by the UE is not limited to three PRBs.
6.1.4 Operation and Maintenance
• The value
OFF is
recommended
for normal
operation
scenarios.
• The value ON
is
recommended
if channel
quality for the
UEs is poor
and their
transmit
power is
limited.
FDD TTI Bundling CellUlschAlgo.TtiBundlingTriggerStrategy This parameter specifies the policies for
Trigger triggering TTI bundling.
Strategy • If TTI bundling is used for
VoLTE services only, set this
parameter to SERVICE_VOIP.
• If TTI bundling is used for
VoLTE services or a
combination of VoLTE and
data services, set this
parameter to
SERVICE_MULTIAPP.
CellTtiBundlingA
parameter is set to
In FDD, the defaul
recommended.
CellUlschAlgo
before the eNod
exit the TTI bun
default value is
Before using MML commands, refer to 6.1.2.2 Impacts and 6.1.3.2 Software and complete the
parameter configurations for related functions based on the impact and mutually exclusive
relationships between the functions, as well as the actual network scenario.
//Setting R12 TTI bundling switches
MOD CELLALGOSWITCH: LocalCellId=0, UlSchSwitch=TtiBundlingSwitch-1;
MOD CELLTTIBUNDLINGALGO: LocalCellId=0, R12TtiBHarqMaxTxNum=n20,
R12TtiBundlingSwitch=ON, SinrThdToTrigR12TtiB=3;
//Setting the parameters related to exit from TTI bundling and enhancement
to TTI bundling algorithm
MOD CELLULSCHALGO: LocalCellId=0, StatisticNumThdForTtibExit=N20_TTIB_EXIT,
HystToExitTtiBundling=5;
MOD CELLTTIBUNDLINGALGO: LocalCellId=0,
TtiBundlingAlgoSw=TTIB_EXIT_BASED_ON_QCI_SW-1&TTIBUNDLING_ALGO_ENHANCE_SW-1;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling TTI bundling
MOD CELLALGOSWITCH: LocalCellId=0, UlSchSwitch=TtiBundlingSwitch-0;
To verify TTI bundling for UEs far from the eNodeB, perform the following steps:
1. On the MAE-Access, create and start a Uu interface tracing task for the test cell.
2. Enable a UE to access the cell and initiate an uplink voice service.
3. Move the UE away from the eNodeB until the RRC_CONN_RECFG and
RRC_CONN_RECFG_CMP messages are present in the Uu tracing result. Check
the IE mac-MainConfig > ul-SCH-Config > ttiBundling in the RRC_CONN_RECFG
message. The value TRUE (as shown in Figure 6-3) indicates that the UE is in the TTI
bundling state for the uplink voice service.
Figure 6-3 RRC_CONN_RECFG message (indicating that the UE has entered the TTI bundling state)
5. On the MAE-Access, query the values of the counters listed in Table 6-8 to check the
status of TTI bundling. If these values are not 0, TTI bundling has taken effect.
Table 6-8 Counters for verifying TTI bundling
Counter ID Counter Name
1526728496 L.Traffic.User.TtiBundling.Avg
1526728911 L.Signal.Num.TtiBundling.Enter
1526728912 L.Signal.Num.TtiBundling.Exit
1526746002 L.Traffic.User.R12TtiBundling.Avg
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
6.2 Uplink RLC Segmentation Enhancement
6.2.1 Principles
This section describes how uplink RLC segmentation enhancement works for VoLTE.
The number of uplink RLC segments is dependent on the TBS determined by uplink scheduling.
The smaller the TBS, the larger the number of uplink RLC segments. If channel quality is poor,
uplink power is limited and a small TBS results in a large number of uplink RLC segments,
which causes:
• Long delay of VoLTE packets
• Loss of uplink VoLTE packets (because VoIP packets wait in the UE buffer so long
that they cannot be scheduled in a timely manner and the packet discard timer expires
at the PDCP layer on the UE side)
• Large overhead of RLC and MAC headers
• Large consumption of CCEs and RBs by uplink dynamic scheduling of VoLTE
services
Uplink RLC segmentation enhancement restricts the TBS in uplink dynamic scheduling to
control the number of uplink RLC segments for VoLTE packets. This restriction improves voice
quality when channel quality is poor.
The CellUlschAlgo.UlVoipRlcMaxSegNum parameter has been introduced to control the maximum
number of uplink RLC segments for UEs not in the TTI bundling state as follows:
• If the number of uplink RLC segments is less than or equal to this parameter value,
uplink RLC segmentation enhancement does not take effect.
• If the number of uplink RLC segments is greater than this parameter value, uplink
RLC segmentation enhancement takes effect. The minimum TBS is guaranteed in
each uplink dynamic scheduling based on the VoLTE packet size and the configured
maximum number of RLC segments. This way, the number of uplink RLC segments
of a VoLTE packet will not exceed its upper limit.
Uplink RLC segmentation enhancement does not take effect when any of the following
conditions is met:
• The CellUlschAlgo.UlVoipRlcMaxSegNum parameter is set to 0.
• The GlobalProcSwitch.LcgProfile parameter is set to LCG_PROFILE_1.
• UEs are in the TTI bundling state.
6.2.2.1 Benefits
Uplink RLC segmentation enhancement improves voice quality when the channel conditions for
VoLTE services are poor.
When the CellUlschAlgo.UlVoipRlcMaxSegNum parameter is set to its recommended value, this
function increases the MOS by about 0.3 for VoIP UEs not supporting TTI bundling in uplink
weak coverage areas.
6.2.2.2 Impacts
Network Impacts
Uplink RLC segmentation enhancement increases the uplink MCS index, IBLER, and RBLER
for VoLTE UEs that are located in uplink weak-coverage areas but do not stay in the TTI
bundling state.
Function Impacts
6.2.3 Requirements
6.2.3.1 Licenses
Prerequisite Functions
None
None
6.2.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
6.2.3.4 Others
No requirements
6.2.4 Operation and Maintenance
6.2.4.1 Data Configuration
Before using MML commands, refer to 6.2.2.2 Impacts and complete the parameter configurations
for related functions based on the impact relationships between the functions, as well as the
actual network scenario.
//Setting parameters related to uplink RLC segmentation enhancement
MOD CELLULSCHALGO: LocalCellId=0, UlVoipRlcMaxSegNum=20,
TtiBundlingRlcMaxSegNum=4;
MOD GLOBALPROCSWITCH: LcgProfile=LCG_PROFILE_0;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling uplink RLC segmentation enhancement
MOD CELLULSCHALGO: LocalCellId=0, UlVoipRlcMaxSegNum=0;
After uplink RLC segmentation enhancement is activated with the Max Number of UL RLC
Segments for VoIP parameter set to a non-zero value, the MCS index should be increased for a
cell-edge UE to control the number of uplink RLC segments. To verify this, perform the
following steps:
1. Run the LST CELLULSCHALGO command to check whether Max Number of
UL RLC Segments for VoIP is set to 0. If it is set to 0, go to 2. If it is set to a non-
zero value, run the following command to change its value to 0:
MOD CELLULSCHALGO: LocalCellId=0, UlVoipRlcMaxSegNum=0;
2. Use a UE that does not support TTI bundling to access the cell and initiate an uplink
VoLTE service.
3. Start a task on the MAE-Access to monitor MCS-specific scheduling statistics.
a. On the MAE-Access, choose Monitor > Signaling Trace > Signaling
Trace Management.
b. In the left pane of the Signaling Trace Management window, choose
User Performance Monitoring > MCS Count Monitoring. Set the
tracing duration, to-be-traced MMEc (MME ID), and mTMSI (UE
TMSI).
c. On the MAE-Access, observe MCS-specific uplink scheduling statistics,
including the number of scheduling times and the number of RBs. Move
the UE away from the eNodeB until the MCS index for uplink scheduling
is 0 and there are three or fewer RBs available for uplink scheduling.
This indicates that the UE has entered an area with poor signal quality.
4. Run the MOD CELLULSCHALGO command with Max Number of UL RLC
Segments for VoIP set to a non-zero value.
To ensure satisfactory voice quality, you are advised to set the Max Number of UL RLC Segments
for VoIP parameter to its recommended value. However, to facilitate verification, you can set this
parameter to a smaller non-zero value so that the selected MCS index increases more noticeably.
1526727961 L.Traffic.UL.PktLoss.Loss.QCI.1
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
6.3 CMR-based Voice Rate Control
6.3.1 Principles
Scenario Overview
In this document, the EVS-SWB codec scheme is implemented only in the full-header format composed
of the Codec_Mode_Request (CMR) byte.
1. The eNodeB checks whether voice rate control is enabled for a MOCN cell based on
the CellAlgoSwitch.UlAmrcMode parameter setting.
• If this parameter is set to ULRATECTRL_OFF, the eNodeB does not
execute voice rate control.
• If this parameter is set to a value other than ULRATECTRL_OFF, the
eNodeB proceeds to 2.
2. The eNodeB checks whether voice rate control is enabled for the operator of the UE
based on the CellOp.UlAmrcMode parameter setting.
• If this parameter is set to ULRATECTRL_OFF, the eNodeB does not
execute voice rate control.
• If this parameter is set to INVALID, the eNodeB executes voice rate
control based on the setting of the cell-level parameter
CellAlgoSwitch.UlAmrcMode.
• If this parameter is set to any other value, the eNodeB executes voice rate
control based on specific combinations of the operator-level and cell-level
parameter settings.
Table 6-10 summaries how operator-specific voice rate control takes effect.
Table 6-10 Rules of validating operator-specific voice rate control
Value of Cell-Level Parameter Value of Operator-Level Parameter Effective Voice Rate Control Descrip
CellAlgoSwitch.UlAmrcMode CellOp.UlAmrcMode Pattern
not
intersec
with on
another
the
eNodeB
does no
execute
voice ra
control.
Uplink- • The TBS of the UE is greater than • The TBS of the UE is less than
quality- TbsUpTh for two consecutive rate TbsDownTh for two consecutive rate
based increase evaluation periods. decrease evaluation periods.
voice • The Uplink Packet Loss Rate (VoIP) of the • The Uplink Packet Loss Rate (VoIP) of the
rate UE is less than UE is greater than
control VoiceAmrControl.PlrThdForIncreasingAmr VoiceAmrControl.PlrThdForDecreasingA
for two consecutive rate increase for two consecutive rate decrease
evaluation periods. evaluation periods.
Load- • The uplink GBR PRB usage of the cell is The uplink cell load over the Uu interface meets
based less than or equal to both of the following conditions:
voice CellRacThd.UlRbThdforVolteLoadAmrc, • The uplink GBR PRB usage is greater
rate and the GBR CCE usage is less than or than
control equal to CellRacThd.UlRbThdforVolteLoadAmrc, o
CellRacThd.CceThdforVolteLoadAmrc. the GBR CCE usage is greater than
• The TBS of the UE is greater than CellRacThd.CceThdforVolteLoadAmrc.
TbsUpTh for two consecutive rate • RLC segmentation is performed for upli
increase evaluation periods. voice packets.
• The Uplink Packet Loss Rate (VoIP) of the
UE is less than
VoiceAmrControl.PlrThdForIncreasingAmr
for two consecutive rate increase
evaluation periods.
SINR- • The measured uplink SINR of the UE is The uplink SINR of the UE measured in a rate
based greater than the value of decrease evaluation period of 60 ms is less than o
voice equal to CellUlschAlgo.SinrThdForVoLTERateCtrl.
Type Rate Increase Conditions Rate Decrease Conditions
rate CellUlschAlgo.SinrThdForVoLTERateCtrl
control plus 5 dB. NOTE:
• The TBS of the UE is greater than The SINR-based voice rate control function decreases the
TbsUpTh for two consecutive rate voice rate to the lowest supported rate to prevent low voice
increase evaluation periods. quality, such as voice skipping and discontinuity.
• The Uplink Packet Loss Rate (VoIP) of the
UE is less than
VoiceAmrControl.PlrThdForIncreasingAmr
for two consecutive rate increase
evaluation periods.
Note that:
• Evaluation switches
▪ Uplink-quality-based voice rate control
The increased coding rate can exceed the initial coding rate of a call only if
the UlAmrcExceedingInitialSw or UlEvsExceedingInitialSw option of
the CellAlgoSwitch.AmrcAlgoSwitch parameter is selected.
▪ Load-based voice rate control
If the VoLTELoadAmrcSw option of the CellAlgoSwitch.AmrcAlgoSwitch
parameter is selected, load-based voice rate control is enabled.
▪ SINR-based voice rate control
If the CellUlschAlgo.SinrThdForVoLTERateCtrl parameter is set to a value
other than -100, SINR-based voice rate control is enabled.
• Evaluation periods
▪ Uplink-quality-based and load-based voice rate control
The rate decrease evaluation period is specified by the
CellUlschAlgo.AmrcDecreasingPeriod parameter. The rate increase
evaluation period is automatically calculated by the eNodeB. To prevent
ping-pong switching between rate increases and rate decreases, the eNodeB
uses the policy of the rate increase evaluation period being longer than or
equal to the rate decrease evaluation period. This lifts the probability of rate
decreases and lowers the probability of rate increases.
▪ SINR-based voice rate control
The rate decrease evaluation period is 60 ms. The rate increase evaluation
period is automatically calculated by the eNodeB. To prevent ping-pong
switching between rate increases and rate decreases, the eNodeB uses the
policy of the rate increase evaluation period being longer than the rate
decrease evaluation period. This lifts the probability of rate decreases and
lowers the probability of rate increases.
• Evaluation items (The evaluation periods for the following items are subject to the rate
control type.)
▪ TbsUpTh and TbsDownTh are automatically calculated based on the
VoiceAmrControl.RsnThdForIncreasingAmr and
VoiceAmrControl.RsnThdForDecreasingAmr parameters, respectively. The
larger the parameter values, the smaller the values of TbsUpTh and
TbsDownTh.
▪ Uplink Packet Loss Rate (VoIP) =
L.Traffic.UL.PktLoss.Loss.QCI.1/L.Traffic.UL.PktLoss.Tot.QCI.1 x 100%
Signaling Procedure
During voice rate control, the eNodeB adjusts the voice rates of UEs, as described in Table 6-12.
Table 6-12 Parameter settings for voice rate control
ADAPTIVE_ENB_CONTROL AMR
EVS-SWB
Adjustment Process
Depending on the rates supported by a UE, the voice rate of the UE can be adjusted between the
high voice rate mode and the low voice rate mode. Table 6-13 lists the rates supported by the two
modes.
Table 6-13 Rates supported by the high and low voice rate modes
Rate Scheme High Voice Rate Mode Low Voice Rate Mode
(VoiceAmrControl.HighAmrCodingMode) (VoiceAmrControl.LowAmrCodingMode)
AMR-NB 12.2 kbit/s and 7.4 kbit/s 7.4 kbit/s and 4.75 kbit/s
AMR-WB 23.85 kbit/s and 12.65 kbit/s 12.65 kbit/s and 6.6 kbit/s
EVS-SWB 24.4 kbit/s and 13.2 kbit/s 13.2 kbit/s and 9.6 kbit/s
Voice rate control only supports rate adjustment within one rate scheme.
For the EVS-SWB scheme, if the channel-aware mode (CAM) is supported, the voice rate must be 24.4 kbit/s or
13.2 kbit/s. The CAM capability information is acquired during SIP message parsing.
6.3.1.4 Restrictions
Application Conditions
If IPsec with the null algorithm is used, the eNodeB can still perform AMR or EVS rate control through rate
adjustment.
AMR or EVS rate control can be enabled or disabled for specified UE types through a whitelist
and a blacklist. Table 6-14 and Table 6-15 describe the whitelist and blacklist configurations.
Table 6-14 Whitelist and blacklist configurations for AMR rate control
Table 6-15 Whitelist and blacklist configurations for EVS rate control
6.3.2.1 Benefits
• Uplink-quality-based voice rate control
▪ When uplink quality is favorable:
▪ Using a high AMR voice rate increases the MOS by 0.01 to 0.2.
▪ Using a high EVS voice rate increases the MOS by 0.2 to 0.5.
▪ When uplink quality is unfavorable, a low voice rate results in a 5% to 20%
decrease in the uplink packet loss rate of cell-edge UEs and a 5% to 20%
decrease in the proportion of single-UE MOSs less than 3. In addition, the
uplink voice coverage improves by 0.5 dB to 1 dB.
Uplink packet loss rate of cell-edge UEs =
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1.Index0/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1.I
ndex0
Applicable Scenarios
1526728303 L.ChMeas.CCE.CommUsed
1526728304 L.ChMeas.CCE.ULUsed
1526728305 L.ChMeas.CCE.DLUsed
1526728765 L.ChMeas.CCE.Avail
1526730884 L.ChMeas.PRB.UL.DrbUsed.Avg.VoIP
1526742156 L.ChMeas.PRB.UL.DrbUsed.Avg.QCI2
1526728434 L.ChMeas.PRB.UL.Avail
Non-Applicable Scenarios
• High-speed cells
• Ultra-high-speed cells
6.3.2.2 Impacts
Network Impacts
Function Impacts
6.3.3 Requirements
6.3.3.1 Licenses
6.3.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
6.3.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
6.3.3.4 Others
EVS rate control requires UEs to support EVS-SWB header-full (with CMR).
6.3.4 Operation and Maintenance
FDD Uplink CellAlgoSwitch.UlAmrcMode For the setting notes of the parameter values,
Rate see 6.3.1.3 Mechanisms of Voice Rate Control.
Control
Mode
FDD Uplink CellOp.UlAmrcMode This parameter specifies the mode of voice rate
Rate control for a specific operator.
Control
Mode
CellAlgoSwitch.AmrcAlgoSwitch
parameter is selected.
If this parameter is set to 10 or a
smaller value, the threshold
does not take effect.
When AMR or EVS rate control is enabled, if no voice rate control parameter group is configured for a specific
coding scheme, the eNodeB selects one of the three AMR-WB rates, one of the three AMR-NB rates, or one of the
three EVS-SWB rates by default. In addition, the following parameters are set to the values recommended in Table
6-18:
• VoiceAmrControl.PlrThdForDecreasingAmr
• VoiceAmrControl.PlrThdForIncreasingAmr
• VoiceAmrControl.RsnThdForDecreasingAmr
• VoiceAmrControl.RsnThdForIncreasingAmr
Table 6-19 Recommended values for Packet Loss Rate Thd for Decreasing
High Rate Coding Mode Low Rate Coding Mode Recommended Value
AMR_WB_23_85kbps AMR_WB_12_65kbps 6
AMR_WB_12_65kbps AMR_WB_6_6kbps 16
AMR_NB_12_2kbps AMR_NB_7_4kbps 6
AMR_NB_7_4kbps AMR_NB_4_75kbps 10
AMR_WB_23_85kbps AMR_WB_6_6kbps 24
AMR_NB_12_2kbps AMR_NB_4_75kbps 20
EVS_SWB_24_4kbps EVS_SWB_13_2kbps 6
High Rate Coding Mode Low Rate Coding Mode Recommended Value
EVS_SWB_13_2kbps EVS_SWB_9_6kbps 16
EVS_SWB_24_4kbps EVS_SWB_9_6kbps 24
Before using MML commands, refer to 6.3.2.2 Impacts and 6.3.3.2 Software and complete the
parameter configurations for related functions based on the impact, dependency, and mutually
exclusive relationships between the functions, as well as the actual network scenario.
//Enabling AMR or EVS rate control
MOD CELLALGOSWITCH: LocalCellId=0, UlAmrcMode=ADAPTIVE_ENB_CONTROL,
AmrcAlgoSwitch=UlAmrcExceedingInitialSw-1&UlAmrCheckSw-
1&VoiceCodingModeMeasSw-1&UlEvsExceedingInitialSw-1&VoLTELoadAmrcSw-1;
//Enabling operator-level AMR or EVS rate control (The cell-level voice rate
control function must also be enabled.)
MOD CELLOP: LocalCellId=0, TrackingAreaId=0,
UlAmrcMode=ADAPTIVE_ENB_CONTROL;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling AMR or EVS rate control
MOD CELLALGOSWITCH: LocalCellId=0, UlAmrcMode=ULRATECTRL_OFF;
//Disabling AMR or EVS rate control for a specific operator
MOD CELLOP: LocalCellId=0, TrackingAreaId=0, UlAmrcMode=ULRATECTRL_OFF;
• For UEs that have accessed cells by initial access, rate adjustment information is not
transferred. The counters listed in Using Counters can be used to check whether voice
rate control has taken effect.
• For UEs undergoing X2-based handovers, X2 interface tracing can be used to check
whether voice rate control has taken effect. If the tracing result indicates the
information similar to that shown in Figure 6-12, the source cell has sent such rate
adjustment information (including the UE's initial rate and the rate set negotiated with
the UE) to the target cell by using a UlAmrc or UlEvs IE. The following figure uses a
UlAmrc IE as an example.
Figure 6-12 UlAmrc sent for a UE undergoing an X2-based handover
Using Counters
This section provides counters that can be used to check whether voice rate control has taken
effect.
• AMR rate control
On the MAE-Access, check the values of the counters listed in Table 6-21. AMR rate
control has taken effect if any value is greater than 0.
Table 6-21 Counters used to monitor AMR rate control
1526741685 L.Voice.UL.AMRNB.Increase.Times
1526741686 L.Voice.UL.AMRNB.Decrease.Times
1526741687 L.Voice.UL.AMRWB.Increase.Times
1526741688 L.Voice.UL.AMRWB.Decrease.Times
1526745691 L.Voice.UL.EVSSWB.Increase.Times
1526745692 L.Voice.UL.EVSSWB.Decrease.Times
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
Observe the following function subsets:
• Number of Times the Uplink Speech Coding Rate Changes
This function subset measures the number of rate increases and decreases for uplink
voice rate control in a cell.
• Speech Coding Scheme Distribution
This function subset measures the distribution of the numbers of uplink and downlink
voice packets in different voice coding modes in a cell.
6.4 MAC CE-based Voice Rate Control
6.4.1 Principles
A new voice rate control mechanism, RAN-assisted codec adaptation, was introduced in 3GPP
Release 14. The eNodeB uses access network bitrate recommendation (ANBR) and access
network bitrate recommendation query (ANBRQ) for UEs to increase or decrease voice rates in
the uplink and downlink.
MAC CE-based voice rate control indirectly adjusts the uplink and downlink voice coding rates
by adjusting the rates at the physical layer. (MAC CE is short for Media Access Control control
element.) The relationship between physical-layer rates and voice coding rates is described in
Supported Voice Rate Indexes.
During voice rate control, the eNodeB only recommends physical-layer rates to UEs, and the calling and called UEs
further negotiate based on the recommended rates.
MAC CE-based voice rate control can be implemented through the following four methods.
• Uplink rate decrease: The eNodeB triggers the ANBR procedure, and the calling and
called UEs complete end-to-end negotiation of the rate decrease.
• Uplink rate increase: The eNodeB triggers the ANBR procedure, the peer UE triggers
the ANBRQ procedure, and the calling and called UEs complete negotiation of the
rate increase.
• Downlink rate decrease: The eNodeB triggers the ANBR procedure, and the calling
and called UEs complete end-to-end negotiation of the rate decrease.
• Downlink rate increase: The eNodeB triggers the ANBR procedure, the peer UE
triggers the ANBRQ procedure, and the calling and called UEs complete negotiation
of the rate increase.
• The peer UE cannot trigger the ANBRQ procedure in any of the following scenarios:
▪ The VolteAlgoConfig.VolteAnbrqProhibitTimer parameter is set to DISABLE for the
eNodeB serving the peer UE.
▪ The peer UE does not support ANBRQ.
• For details, see section 10.7 "Access network bitrate recommendation" in 3GPP TS 26.114 V16.7.0.
Table 6-23 lists the switches and trigger conditions for the four types of voice rate control.
Table 6-23 MAC CE-based voice rate control types and trigger conditions
Control Switch Trigger Condition
Type
Uplink rate ANBR: Select the The eNodeB dynamically triggers the
decrease VOLTE_ANBR_UL_DECREASE_SW corresponding control type based on the
optiona. physical-layer rateb and air interface
ANBRQ: None scheduling capabilityc.
• The eNodeB triggers uplink or
Uplink rate ANBR: Select the downlink rate decrease when
increase VOLTE_ANBR_UL_INCREASE_SW it detects the air interface
optiona. scheduling capability is lower
ANBRQ: The than or equal to the physical-
VolteAlgoConfig.VolteAnbrqProhibitTimer layer rate.
parameter is set to a value other than • The eNodeB triggers uplink or
DISABLE. downlink rate increase when
it detects the air interface
Downlink ANBR: Select the scheduling capability is higher
rate VOLTE_ANBR_DL_DECREASE_SW than the physical-layer rate.
decrease optiona.
ANBRQ: None
c: The eNodeB determines the air interface scheduling capability based on the channel state. For
downlink rate decrease or increase, the eNodeB determines the air interface scheduling
capability based on the VolteAlgoConfig.VolteAnbrDlSchRbNumThld parameter value.
MAC CE-based voice rate control involves ANBR and ANBRQ procedures.
• ANBR procedure
The ANBR procedure can be triggered when the switch and trigger conditions of the
corresponding control type are met.
1. The eNodeB delivers a recommended uplink or downlink physical-layer
rate to the UE through the downlink MAC CE.
2. After receiving the recommended rate, the UE negotiates with the peer UE
to determine whether to adjust the voice coding rate.
Figure 6-13 ANBR procedure
• ANBRQ procedure
The ANBRQ procedure can be triggered after the peer UE triggers an uplink or
downlink rate increase.
1. The eNodeB delivers the bitRateQueryProhibitTimer-r14 IE to the UE
through an RRC message.
The bitRateQueryProhibitTimer-r14 IE contains the minimum interval
(specified by the VolteAlgoConfig.VolteAnbrqProhibitTimer parameter) at
which the UE sends a request for the recommended physical-layer rate to
the eNodeB.
2. The UE sends a Recommended Bit Rate MAC Control Element message
to the eNodeB, requesting it to query the recommended physical-layer
rate.
3. The eNodeB delivers the recommended uplink or downlink physical-layer
rate to the UE through the downlink MAC CE.
4. After receiving the recommended rate, the UE negotiates with the peer UE
to determine whether to increase the voice coding rate.
Figure 6-14 ANBRQ procedure
For details about MAC CE, see section 6.1.3.13 "Recommended bit rate MAC Control Element"
in 3GPP TS 36.321 V16.2.0.
Supported Voice Rate Indexes
MAC CE-based voice rate control supports rate adjustment between two sets of rate indexes, as
listed in Table 6-24.
Table 6-24 Rate indexes supported by MAC CE-based voice rate control
High-Rate Index 9 12 13
Low-Rate Index 6 10 12
6.4.2.1 Benefits
• ANBR
▪ When rate increase is enabled and the uplink or downlink air interface
capability is high, the eNodeB recommends a higher voice coding rate to
UEs, improving the uplink or downlink voice service experience.
▪ When the VOLTE_ANBR_UL_DECREASE_SW option of the
VolteAlgoConfig.VolteOptSwitch parameter is selected and the uplink air
interface capability is low, the eNodeB recommends a lower physical-layer
rate to UEs, reducing the uplink voice packet loss rate at the cell edge.
▪ When the VOLTE_ANBR_DL_DECREASE_SW option of the
VolteAlgoConfig.VolteOptSwitch parameter is selected and the downlink air
interface capability is low, the eNodeB recommends a lower physical-layer
rate to UEs, improving the downlink voice service capacity of the cell.
• ANBRQ
When the VolteAlgoConfig.VolteAnbrqProhibitTimer parameter is set to a value other
than DISABLE, the eNodeB recommends a physical-layer rate to UEs based on the
air interface scheduling capability, preventing unnecessary rate increases initiated by
UEs.
To achieve optimal benefits, you are advised to evaluate when to use the function based on Table
6-25 and monitor the counters listed in Table 6-26 to determine the sites and cells that the function
is applicable to. If the conditions described in Table 6-25 are not met, there will be limited
benefits, but no negative impacts.
Table 6-25 When to use
Scenario Description
1526727378 L.Traffic.User.Avg
Sites with a high traffic volume of voice services at cell edges 1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.Q
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
1526745995 L.Traffic.UL.FarUE.PktLoss.Loss
Site/Cell Type Counter ID Counter Name
Cells with a high packet loss rate for voice service UEs at the cell 1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.Q
edge
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1
6.4.2.2 Impacts
Network Impacts
• ANBR
▪ When rate increase is enabled and the uplink or downlink air interface
capability is high, the voice quality improves while the cell capacity
decreases.
▪ When downlink voice rate decrease is triggered with the
VOLTE_ANBR_DL_DECREASE_SW option selected, the cell capacity
increases while the voice quality deteriorates.
• ANBRQ
▪ In scenarios where the uplink or downlink rate increases, if a peer UE does
not trigger the ANBRQ procedure, the calling and called UEs will directly
perform end-to-end negotiation of the rate increase, affecting the uplink
performance of the peer UE.
▪ If the VolteAlgoConfig.VolteAnbrqProhibitTimer parameter is set to a value
other than DISABLE: A larger value of this parameter results in less
frequent reporting of the "Recommended Bit Rate MAC Control Element"
IE from a UE to the eNodeB and lower uplink overhead. A smaller value of
this parameter results in the opposite effects.
Function Impacts
6.4.3 Requirements
6.4.3.1 Licenses
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
CMR-based voice rate control adjusts the voice coding rate by modifying CMR packets at the application layer. This
way, the eNodeB can accurately control the uplink voice rate.
MAC CE-based voice rate control indirectly adjusts the uplink and downlink voice coding rates by adjusting the
rates at the physical layer. The relationship between physical-layer rates and voice coding rates is described in
Supported Voice Rate Indexes.
Therefore, if CMR-based voice rate control cannot be used, MAC-CE-based uplink voice rate control is
recommended.
6.4.3.3 Hardware
No requirements
Boards
RF Modules
No requirements
6.4.3.4 Others
Before using MML commands, refer to 6.4.2.2 Impacts and 6.4.3.2 Software and complete the
parameter configurations for related functions based on the impact and mutually exclusive
relationships between the functions, as well as the actual network scenario.
//Enabling MAC CE-based uplink voice rate decrease
MOD VOLTEALGOCONFIG: LocalCellId=0,
VolteOptSwitch=VOLTE_ANBR_UL_DECREASE_SW-1;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling MAC CE-based uplink voice rate decrease
MOD VOLTEALGOCONFIG: LocalCellId=0,
VolteOptSwitch=VOLTE_ANBR_UL_DECREASE_SW-0;
1. On the MAE-Access, create and start a Uu interface tracing task for the test cell.
2. Enable a UE to access the cell and initiate an uplink voice service.
3. Check the RRC_CONN_RECFG message for setting up the QCI 1 bearer in the
traced Uu messages. Check whether the bitRateQueryProhibitTimer-r14 field is
contained in the IE drb-ToAddModList > DRB-ToAddMod > LogicalChannelConfig
in the RRC_CONN_RECFG message, as shown in Figure 6-15. If this field is
contained, MAC CE-based voice rate control has taken effect.
Figure 6-15 MAC CE-based voice rate control
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
6.5 Inter-eNodeB VoLTE CoMP (FDD)
6.5.1 Principles
This function was enhanced in eRAN11.1. In versions earlier than eRAN11.1, it was controlled by the
UlJROverRelaxedBHSw option of the ENodeBAlgoSwitch.OverBBUsSwitch parameter. If the pre-enhancement
function is required in eRAN11.1 or later versions, this option needs to be selected. For details, see eRAN8.1 UL
CoMP.
6.5.2.1 Benefits
Inter-eNodeB VoLTE CoMP reduces the number of retransmissions for VoLTE UEs, RBLER,
and number of TBs unsuccessfully transmitted during initial uplink transmission. This function
also decreases the QCI 1 packet loss rate for cell-edge UEs by 0%–20%.
• In macro-macro or micro-micro scenarios, VoLTE coverage increases by 0 dB to 2
dB.
• In macro-micro scenarios, VoLTE coverage increases by 0 dB to 3 dB.
• If the original one-way transmission delay is short, this function will produce a large
gain. If the original one-way transmission delay is longer than 8 ms, this function does
not yield noticeable effect.
You are advised to evaluate when to use the function according to Applicable Scenarios and
Information Collection to achieve optimal benefits. If the conditions described in Applicable Scenarios
are not met, there will be limited benefits, but no negative impacts.
Applicable Scenarios
Inter-eNodeB VoLTE CoMP is recommended when all of the following conditions are met:
• There are VoLTE services on the network.
• VoLTE service coverage needs to be improved.
• The inter-BBU one-way delay is less than or equal to 8 ms.
• Packets are lost for cell-edge UEs accessing QCI 1 services.
QCI 1 packet loss rate for cell-edge UEs =
L.Traffic.UL.BorderUE.PktLoss.Loss.QCI.1/L.Traffic.UL.BorderUE.PktLoss.Tot.QCI.1
Other deployment suggestions are the same as those for LOFD-001066 Intra-eNodeB UL CoMP.
For details, see UL CoMP.
If inter-eNodeB VoLTE CoMP and LOFD-001066 Intra-eNodeB UL CoMP are both enabled, both of them yield
satisfactory gains. Therefore, it is recommended that they are enabled simultaneously.
Information Collection
Before enabling inter-eNodeB VoLTE CoMP, collect the information described in Table 6-28.
Table 6-28 Information to be collected
Inter-BBU time Verify that time synchronization between BBUs has a deviation of less than
synchronization 3 μs. If this requirement is not met, the eNodeB automatically disables this
accuracy function.
This time synchronization accuracy requirement can be fulfilled by using a
GNSS or IEEE1588 V2 clock. For details, see Synchronization.
Physical cell PCI conflicts between intra-frequency neighboring cells have a negative
identifier (PCI) impact on UL CoMP performance. You are advised to turn on the PCI
conflict conflict alarm switch so that PCI collision and confusion can be reported.
Resource audit Contact Huawei technical support for resource audit. This function shares
system resources with SFN, CA, and CSPC.
6.5.2.2 Impacts
Network Impacts
When a BBP is restarted or reset, cells are reestablished. In this situation, the binding between
cells and BBPs may change and UL CoMP performance may be affected.
Function Impacts
individual
UEs.
a: If inter-eNodeB VoLTE CoMP is disabled, the event A3 offset for UL CRA is controlled by
the UlCsAlgoPara.UlCsA3Offset parameter. If inter-eNodeB VoLTE CoMP is enabled, the event
A3 offset for UL CRA is controlled by the CellUlCompAlgo.UlCompA3OffsetForRelaxedBH
parameter.
6.5.3 Requirements
6.5.3.1 Licenses
6.5.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
6.5.3.3 Hardware
• 3900 and 5900 series base stations (This function is not supported in scenarios where
both macro base stations using four antennas and micro base stations using two
antennas are deployed.)
• DBS3900 LampSite and DBS5900 LampSite
Boards
Each cell's UL CoMP capabilities depend on the BBPs that have been configured and on the
receive antenna configurations:
• This function is incompatible with 4R cells served by LBBPc boards.
• All other LBBP and UBBP boards support UL CoMP for VoLTE services between
two or three cells. 4R cells do not support 3-cell UL CoMP.
RF Modules
To ensure UL CoMP performance, it is recommended that the distance between the RRUs of the
serving and cooperating cells not exceed 1000 m.
6.5.3.4 Networking
You are advised to bind intra-frequency cells served by RRUs installed on the same pole or
tower to BBPs in the same connection set.
This function requires inter-BBU cells to be planned to ensure that inter-BBU routes are
reachable. eX2 interfaces between BBUs are involved in the planning. For details, see eX2 Self-
Management.
6.5.3.5 Others
This function runs over an IP RAN. If the transmission delay is too long, cooperating cell data
will arrive at the RLC layer too late and the data frames may fail to be acknowledged. Therefore,
when this function is enabled, it is recommended that the RlcPdcpParaGroup.UlRlcSnSize
parameter be set to RlcSnSize_size10.
IP Performance Monitor (PM) sessions of forward activation type are recommended for delay
detection. eNodeBs will automatically create IP PM sessions to detect link status of eX2
interfaces. If an IP PM session of bidirectional activation type is configured at either end of an
eX2 interface, it may conflict with an automatically created session. Therefore, the eX2 interface
fails to work properly.
6.5.4 Operation and Maintenance
and Table 6-30 describe the parameters used for function activation and optimization,
Table 6-29
respectively.
Table 6-29 Parameters used for activation
Parameter Name Parameter ID Option Setting Notes
For other optimization suggestions, see the parameters for basic optimization, downlink RSRP optimization, and UL
CoMP enhancement functions of LOFD-001066 Intra-eNodeB UL CoMP described in UL CoMP.
Before using MML commands, refer to 6.5.2.2 Impacts and 6.5.3.2 Software and complete the
parameter configurations for related functions based on the impact and mutually exclusive
relationships between the functions, as well as the actual network scenario.
//Enabling inter-eNodeB VoLTE CoMP
MOD ENODEBALGOSWITCH: OverBBUsSwitch=UlVoiceJROverRelaxedBHSw-1;
Commands for activating a specific function of UL CoMP vary with the networking mode, antenna configuration,
measurement mechanism, and other factors. This section provides an example.
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling inter-eNodeB VoLTE CoMP
MOD ENODEBALGOSWITCH: OverBBUsSwitch=UlVoiceJROverRelaxedBHSw-0;
6.5.4.1.3 Using the MAE-Deployment
Run the DSP CELLULCOMPCLUSTER command to query the cooperating candidate list of a
cell. If Serving Sector Equipment(Group) eNB ID and Coordinating Cell eNodeB ID have
different values, inter-eNodeB VoLTE CoMP has taken effect.
6.5.4.2.2 Using Counters
Observe the counters listed in Table 6-31 to evaluate whether inter-eNodeB VoLTE CoMP has
taken effect in a cell.
• With inter-eNodeB data CoMP enabled
If the values of the counters in Table 6-31 increase, inter-eNodeB VoLTE CoMP has
taken effect in the cell. This increase is small if there are few voice service UEs in the
cell.
• With inter-eNodeB data CoMP disabled
If the values of the counters in Table 6-31 are not 0, inter-eNodeB VoLTE CoMP has
taken effect in the cell.
Table 6-31 Counters related to inter-eNodeB VoLTE CoMP
Counter ID Counter Name
1526737762 L.ChMeas.ULRelaxedBHCoMP.PRB.Avg
1526737763 L.ULCoMP.ULRelaxedBHCoMP.User.Avg
Monitor the counters listed in Table 6-32 to check whether inter-eNodeB VoLTE CoMP has
improved the coverage for voice service UEs and decreased the packet loss rate.
• Uplink Packet Loss Rate (VoIP) = L.Traffic.UL.PktLoss.Loss.QCI.1/L.Traffic.UL.PktLoss.Tot.QCI.1
• Uplink QCI 1 packet loss rate for cell-edge UEs =
L.Traffic.UL.BorderUE.PktLoss.Loss.QCI.1/L.Traffic.UL.BorderUE.PktLoss.Tot.QCI.1
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
1526727961 L.Traffic.UL.PktLoss.Loss.QCI.1
1526741762 L.Traffic.UL.BorderUE.PktLoss.Tot.QCI.1
1526741761 L.Traffic.UL.BorderUE.PktLoss.Loss.QCI.1
6.6.1 Principles
Overview
With flash SRVCC, uplink quality is evaluated to determine whether to set up a dedicated voice
bearer. An eNodeB will refuse the request for setting up a dedicated voice bearer for a UE if the
eNodeB identifies that the UE is in a weak-coverage area. The IMS then requests that the UE
retry a CSFB procedure, maximizing the probability of a successful voice call.
This function is controlled by the CellHoParaCfg.FlashSrvccSwitch parameter.
Principles
After this function is enabled, the eNodeB calculates the uplink path loss based on the power
headroom report (PHR) sent by a UE and measures the uplink sounding reference signals (SRSs)
or demodulation reference signals (DMRSs) to obtain the uplink SINR.
The eNodeB determines that a UE is in a weak-coverage area if either of the following
conditions is met:
• Path loss > CellHoParaCfg.UlPoorCoverPathLossThd
• SINR < CellHoParaCfg.UlPoorCoverSinrThd
Figure 6-16 Signaling procedure for UEs in weak-coverage areas
When the UE is in idle mode, CS Paging Notification is sent to the UE through a Paging
message. When the UE is in connected mode, CS Paging Notification is sent to the UE
through a DL NAS message.
UE Whitelist and Blacklist
A whitelist and a blacklist can be configured on the eNodeB to allow flash SRVCC to be enabled
or disabled for specified UE types. Table 6-33 describes the whitelist and blacklist configurations.
Table 6-33 Whitelist and blacklist configurations
Overview
This function allows the eNodeB to identify UEs experiencing poor voice quality based on the
uplink or downlink packet loss rate for QCI 1 services. The eNodeB triggers inter-frequency
measurements and inter-frequency handovers for these UEs to decrease the packet loss rate and
increase the percentage of LTE UEs on the LTE network.
This function takes effect when the InterFreqVolteQualityHoSwitch option of the
CellHoParaCfg.VoLTEQualityHoAlgoSwitch parameter is selected.
Inter-Frequency Measurement
Before delivering the measurement configuration related to event A4, the eNodeB checks
whether measurement gaps have been set up. If they have been set up and include other
measurement gaps, the eNodeB does not reconfigure the measurement gaps. Otherwise, the
eNodeB set up the measurement gaps.
The eNodeB adheres to common rules for filtering neighboring cells and frequencies. If the
EutranInterNFreq.VoLTEQualityIfHoTargetInd parameter of a frequency is set to
NOT_ALLOWED, the eNodeB filters out this frequency.
Inter-Frequency Handover
Table 6-34describes the parameters that specify the thresholds of event A4 for voice-quality-based
inter-frequency handovers. The other parameters are the same as those for event A4 for
triggering coverage-based inter-frequency handovers.
Table 6-34 Thresholds of event A4 for voice-quality-based inter-frequency handovers
6.6.2.1 Benefits
Network Impacts
Function Impacts
FDD VoLTE over CellEmtcAlgo.EmtcVolteSupportSwitch eMTC eMTC voice service UEs do not
eMTC support coverage-based VoLTE
experience optimization.
FDD PRB- CellMLB.MlbTriggerMode being set to Intra-RAT When either of the MLB functions is
usage- PRB_ONLY and Mobility enabled and the HoAdmitSwitch
Load
based InterFreqMlbSwitch option of the Balancing option of the
connected CellAlgoSwitch.MlbAlgoSwitch CellMlbHo.MlbMatchOtherFeatureMode
mode load parameter parameter is selected, handovers to
equalization a cell will not be admitted if an MLB
procedure is triggered in the cell.
FDD User- CellMLB.MlbTriggerMode being set to Intra-RAT This decreases the handover
number- UE_NUMBER_ONLY and Mobility
Load preparation success rate. Therefore,
based InterFreqMlbSwitch option of the Balancing it is recommended that the
connected CellAlgoSwitch.MlbAlgoSwitch HoAdmitSwitch option of the
mode load parameter CellMlbHo.MlbMatchOtherFeatureMode
equalization parameter be deselected.
6.6.3 Requirements
6.6.3.1 Licenses
6.6.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
6.6.3.3 Hardware
For FDD, the following base stations are compatible with this function:
• 3900 and 5900 series base stations
• DBS3900 LampSite and DBS5900 LampSite
Boards
No requirements
RF Modules
No requirements
6.6.3.4 Others
This function requires coordination between Huawei eNodeBs and the IMS. UEs must be able to
perform CSFB after receiving an SIP message (SIP 500/380/503).
6.6.4 Operation and Maintenance
Flash SRVCC
Before using MML commands, refer to 6.6.2.2 Impacts and 6.6.3.2 Software and complete the
parameter configurations for related functions based on the impact and dependency relationships
between the functions, as well as the actual network scenario.
//Enabling flash SRVCC and voice-quality-based inter-frequency handover
MOD CELLHOPARACFG: LocalCellId=0, FlashSrvccSwitch=ON,
UlPoorCoverPathLossThd=125, UlPoorCoverSinrThd=0,
VoLTEQualityHoAlgoSwitch=InterFreqVolteQualityHoSwitch-1,
VolteQualPktLossPeriod=2, VolteQualIfHoQci1PlrThld=20,
VolteQualRecoveryQci1PlrThld=5;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling flash SRVCC and voice-quality-based inter-frequency handover
MOD CELLHOPARACFG: LocalCellId=0, FlashSrvccSwitch=OFF,
VoLTEQualityHoAlgoSwitch=InterFreqVolteQualityHoSwitch-0;
1. On the MAE-Access, check the value of the counter listed in Table 6-37. If the value is
greater than 0, flash SRVCC has taken effect.
Table 6-37 Counter for verifying flash SRVCC
Counter ID Counter Name
1526742168 L.E-RAB.FailEst.PoorCover.VoIP
2. On the MAE-Access, check the values of the counters listed in Table 6-38. If any value
is greater than 0, voice-quality-based inter-frequency handover has taken effect.
Table 6-38 Counters for verifying voice-quality-based inter-frequency handover
Counter ID Counter Name
1526745864 L.HHO.InterFreq.VolteQual.PrepAttOut
Counter ID Counter Name
1526745865 L.HHO.InterFddTdd.VolteQual.PrepAttOut
1526745866 L.HHO.InterFreq.VolteQual.ExecAttOut
1526745867 L.HHO.InterFddTdd.VolteQual.ExecAttOut
1526745868 L.HHO.InterFreq.VolteQual.ExecSuccOut
1526745869 L.HHO.InterFddTdd.VolteQual.ExecSuccOut
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
Additionally, you can monitor the counters listed in Table 6-39 to check whether the counter
values are less than those measured before function activation.
Table 6-39 Counters for verifying flash SRVCC
Counter ID Counter Name
1526728403 L.IRATHO.SRVCC.E2G.PrepAttOut
1526728400 L.IRATHO.SRVCC.E2W.PrepAttOut
Monitor the counters listed in Table 6-38. If any value is greater than 0, voice-quality-based inter-
frequency handover has taken effect.
6.7 VoLTE Coverage Enhancement
6.7.1 Principles
Overview
This function increases the tolerable delay over the Uu interface, decreases the probability of
uplink packet loss caused by congestion over the Uu interface, and improves the uplink coverage
for voice service UEs as follows:
• Optimizes the following QCI 1 radio bearer-related items that the eNodeB sends to
UEs:
▪ PDCP-layer discard timer
▪ Maximum number of uplink HARQ transmissions
• Optimizes the timer for reordering at the receiver eNodeB in AM/UM.
• Selects the optimal TBS based on the volume of VoLTE data to be scheduled.
Description
This function is controlled by the UlVoipCrosslayerOptSwitch option of the
CellUlschAlgo.UlEnhencedVoipSchSw parameter.
Table 6-40 lists the timer and HARQ configurations used after this function is enabled.
After this function is enabled, the eNodeB automatically adjusts the configurations, requiring no manual
intervention.
Configuration Value When This Function Takes Effect Value Before This Function Takes Effect
Item
Maximum 8 CellUlschAlgo.UlHarqMaxTxNum
number of
uplink HARQ
transmissions
(in the non-
TTI bundling
state)
6.7.2.1 Benefits
• Cross-layer optimization for VoLTE in the uplink is recommended when all of the
following conditions are met:
▪ The number of voice service UEs meets the following requirement:
L.Traffic.User.VoIP.Avg/L.Traffic.User.Avg > 5%
▪ The percentage of uplink voice packets from the cell edge meets the
following requirement:
L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1/L.Traffic.UL.PktLoss.Tot.QCI.1 > 5%
▪ The loss rate of uplink voice packets from the cell edge meets the following
requirement:
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1 > 2%
• Uplink AMR voice frame recovery is recommended when VoLTE is deployed on the
live network.
Information Collection
Before enabling VoLTE coverage enhancement, collect the information described in Table 6-41.
Table 6-41 Site/Cell collection
1526727378 L.Traffic.User.Avg
Sites with a high traffic volume of voice services at cell edges 1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.Q
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
Cells with a high packet loss rate for voice service UEs at the cell 1526745995 L.Traffic.UL.FarUE.PktLoss.Loss
edge
L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1/L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1
1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.Q
6.7.2.2 Impacts
Network Impacts
Function Impacts
6.7.3 Requirements
6.7.3.1 Licenses
6.7.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
6.7.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
6.7.3.4 Others
No requirements
6.7.4 Operation and Maintenance
Before using MML commands, refer to 6.7.2.2 Impacts and 6.7.3.2 Software and complete the
parameter configurations for related functions based on the impact and dependency relationships
between the functions, as well as the actual network scenario.
//Setting parameters related to cross-layer optimization for VoLTE in the
uplink and uplink RLC segmentation enhancement
MOD CELLULSCHALGO: LocalCellId=0,
UlEnhencedVoipSchSw=UlVoipCrosslayerOptSwitch-1, UlVoipRlcMaxSegNum=23,
TtiBundlingRlcMaxSegNum=5;
//Enabling uplink AMR voice frame recovery
MOD VOLTEALGOCONFIG: LocalCellId=0, VolteOptSwitch=VOICE_FRAME_RECOVERY_SW-
1;
1. On the MAE-Access, create and start a Uu interface tracing task for the test cell.
2. Enable a UE to access the cell and initiate an uplink voice service.
3. Check the RRC_CONN_RECFG message for setting up the QCI 1 bearer among the
traced Uu messages. Check the IE drb-ToAddModList > DRB-ToAddMod > pdcp-
Config > discardTimer and the IE mac-MainConfig > explicitValue > ul-SCH-
Config > maxHARQ-Tx in the RRC_CONN_RECFG message, as shown in Figure 6-
17. If the values are as shown in Figure 6-17, cross-layer optimization for VoLTE in the
uplink has taken effect.
Figure 6-17 Verification of cross-layer optimization for VoLTE in the uplink
Observe the L.Voice.Packet.Recovery counter. If its value is not 0, uplink AMR voice frame
recovery has taken effect.
6.7.4.3 Network Monitoring
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
7 Quality Improvement
7.1.1 Principles
Uplink delay-based dynamic scheduling allows the eNodeB to assign different scheduling
priorities based on the data scheduling waiting time and service types during uplink dynamic
scheduling. This balances the scheduling waiting time for UEs, improving voice quality.
Specifically, voice quality improves for cell-edge UEs with poor channel quality. This function
improves user experience with voice services when the voice service load is high.
The uplink delay-based scheduling policy depends on the setting of the
CellUlschAlgo.UlDelaySchStrategy parameter:
• Set to VOIP_DELAYSCH
The eNodeB increases the scheduling priority of VoIP services whose scheduling
waiting time is longer than 25 ms, as shown in Figure 7-1.
• Set to VOIP_AND_DATA_DELAYSCH
▪ (FDD) The MCS index is decreased during scheduling request (SR)
scheduling of voice services but not during SR scheduling of data services.
Figure 7-2 shows the scheduling priorities.
• Set to NO_DELAYSCH
The delay-based scheduling policy is not used for uplink dynamic scheduling.
Figure 7-1 Scheduling priority when UlDelaySchStrategy is set to VOIP_DELAYSCH
Uplink VoLTE volume estimation for dynamic scheduling allows the eNodeB to estimate the
uplink service volume, against the fact that the eNodeB can obtain an accurate downlink service
volume but cannot obtain an accurate uplink service volume. The estimation improves the
scheduling efficiency.
The eNodeB performs uplink estimation (for QCI 1 services only) based on the VoLTE model
and uplink scheduling intervals:
• During talk spurts, the eNodeB estimates the number of voice packets in the UE buffer
based on their uplink scheduling intervals. Based on the voice packet size, the eNodeB
then calculates the VoLTE data volume for dynamic scheduling.
• During silent periods, the eNodeB uses the voice packet size as the uplink VoLTE data
volume for dynamic scheduling.
Uplink VoLTE volume estimation for dynamic scheduling makes the calculation of uplink
VoLTE data volume for dynamic scheduling more accurate, and shortens the additional delay of
voice packets in case the calculated volume is less than the actual volume. This function
improves voice quality when a cell is heavily loaded and DRX is enabled. Figure 7-3 shows the
principles.
Figure 7-3 Uplink VoLTE volume estimation for dynamic scheduling
If a called UE does not answer, the eNodeB releases the calling UE when the UE inactivity timer
expires. This may prolong the call delay or interrupt the session. Table 7-1 lists the effective
values of the UE inactivity timer.
Table 7-1 Effective values of the UE inactivity timer
Setting of the DynDrxSwitch Setting of the Effective Value of the UE Inactivity
Option of the CellAlgoSwitch.UEInactiveTimerQCI1Switch Timer
CellAlgoSwitch.DynDrxSwitch Parameter
Parameter
Selected ON QciPara.UeInactivityTimerDynDrxQci
parameter value for QCI 1
Deselected ON QciPara.UeInactiveTimerForQci
parameter value for QCI 1
a: If there are multiple bearers with the highest priority, the maximum QciPara.UeInactivityTimerDynDrxQci
or QciPara.UeInactiveTimerForQci parameter value is used as the effective value of the UE inactivity timer.
This function limits the RSRP of the PUSCH so that there will not be excessive interference. By
default, voice service UEs and data service UEs use the same PUSCH RSRP upper limit. This
limit is controlled by the following parameters:
• CellPcAlgo.PuschRsrpHighThd
• ParaAutoOptCfg.PUSCHRsrpHighThd4AutoOpt
• UlCsAlgoPara.UlCoPcRbkRsrpThd
For details, see the descriptions of RSRP upper limit in PUSCH closed-loop power control in
Power Control.
In heavy-load scenarios, if voice service UEs and data service UEs use the same PUSCH RSRP
upper limit, the UE transmit power will be unable to reach its theoretical maximum. To resolve
this, the PUSCH RSRP upper limit is increased, for voice services, to a value equal to that of
data services plus an offset specified by CellPcAlgo.PuschRsrpHighThdOffsetVoIP. If this
parameter is set to 255, there is no PUSCH RSRP upper limit for voice services.
The scheduling request indicator (SRI) period adaption function extends the SRI period of voice
or data services to 40 ms or even 80 ms based on the same principles in the high or heavy SRI
load state. This increases the uplink delay of voice packets and affects voice user experience.
For details on the SRI period adaption function for the high and heavy SRI load states, see Physical Channel
Resource Management.
When the voice-based SRI period adaption optimization function is enabled, an SRI period of 20
ms can be used for QCI 1 services and the SRI period of data services remains unchanged only if
the number of voice service UEs is less than 40 in a heavy-load cell.
The voice-based SRI period adaption optimization function is controlled by the
SriPeriodOptForVoipSW option of the CellPucchAlgo.SriAlgoSwitch parameter. This function
takes effect only when both of the following conditions are met:
• PUCCHCfg.SriPeriodAdaptive is set to QCIADAPTIVE.
• CellPucchAlgo.SriReCfgInd is set to FALSE.
In accordance with 3GPP TS 26.190, bits of an AMR-WB voice frame are sequenced according
to their importance, and then divided into classes A and B. If RLC fragments of class A are all
received and only RLC fragments of class B are lost, then the eNodeB compensates for the lost
fragments. The assembled voice frame is error-tolerable and can be used for subsequent
decoding. This way, the decoding performance is better than if the frame is discarded, and voice
user experience improves.
Figure 7-4 Smart AMR voice frame recovery for uplink VoLTE services
In FDD, when VoLTE coverage enhancement is enabled, smart AMR voice frame recovery
improves uplink MCS scheduling and voice user experience. Therefore, the smart AMR voice
frame recovery function requires VoLTE coverage enhancement to be enabled.
Smart AMR voice frame recovery is controlled by the AmrVoiceFrameSmartCoverySw option
of the CellUlschAlgo.UlEnhencedVoipSchSw parameter.
This function does not take effect if any of the following conditions is met:
• The speech codec mode in use is not AMR-WB.
• The speech codec mode has not been identified. For example, in parsing-limited
scenarios (see 4.1.6.3 Parsing-Limited Scenarios), the eNodeB fails to identify the speech
codec mode from the first received SID frame.
• The voice coding rate has not been identified. For example, when RTP packets are
encrypted, the eNodeB fails to identify the voice coding rate.
Feedback on downlink HARQ is usually sent over the PUCCH. If the PUCCH is exposed to
interference, the PUCCH demodulation performance is affected, causing voice user experience
to deteriorate.
After a QCI 1 bearer is set up for a UE, this function enables the eNodeB to trigger one-time
uplink scheduling of the UE while scheduling this UE in the downlink. Feedback on downlink
HARQ is then transmitted over the PUSCH, improving feedback demodulation performance.
VoLTE HARQ feedback reporting over the PUSCH is enabled if the
VOLTE_PUSCH_ACK_OPT_SW option of the VolteAlgoConfig.VolteOptSwitch parameter is
selected.
7.1.2 Network Analysis
7.1.2.1 Benefits
The low-value range refers to the path loss range that corresponds to average MOSs less
than 3.5 before smart AMR voice frame recovery is enabled.
▪ If ROHC is enabled, the overall packet loss rate of QCI 1 services or the
uplink packet loss rate of UEs far from the cell center is decreased by 5% to
10%.
Function Description
Uplink delay- This function is recommended if all of the following conditions are met:
based dynamic • L.Traffic.User.VoIP.Avg > 0
scheduling • L.Traffic.User.Avg > 400
Function Description
Smart AMR voice This function is recommended if all of the following conditions are met:
frame recovery • The percentage of AMR-WB services in voice servicesb is
greater than 20%.
• In FDD, VoLTE coverage enhancement is enabled. (This
condition is mandatory for this function.)
• ROHC is enabled.
• Some UEs are at the cell edge.
L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1 > 0
VoLTE HARQ This function is recommended if all of the following conditions are met:
feedback • Average uplink interference on the PUCCHc > –112 dBm
reporting over the • Average uplink interference on the PUCCHc >
PUSCH L.UL.Interference.Avg
• Uplink Resource Block Utilizing Rate < 60%
Separate This function is recommended if all of the following conditions are met:
configuration of • L.Traffic.User.VoIP.Avg > 0
the PUSCH • CCE usagea > 70%
RSRP upper limit
• Uplink RB usage > 80%
(FDD)
Voice-based SRI This function is recommended if both of the following conditions are met:
period adaption • L.Traffic.User.VoIP.Avg > 0
optimization • L.Traffic.User.Avg > 250
(FDD)
Network Impacts
Function Impacts
7.1.3 Requirements
7.1.3.1 Licenses
7.1.3.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
7.1.3.3 Hardware
For FDD, the following base stations are compatible with this function:
• 3900 and 5900 series base stations
• DBS3900 LampSite and DBS5900 LampSite
Boards
In FDD, do not use LBBPc for the smart AMR voice frame recovery function. This board does
not support this function.
RF Modules
No requirements
7.1.3.4 Others
No requirements
7.1.4 Operation and Maintenance
Before using MML commands, refer to 7.1.2.2 Impacts and 7.1.3.2 Software and complete the
parameter configurations for related functions based on the impact and dependency relationships
between the functions, as well as the actual network scenario.
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling uplink delay-based dynamic scheduling
MOD CELLULSCHALGO: LocalCellId=0, UlDelaySchStrategy=NO_DELAYSCH;
1. On the MAE-Access, check the counter values listed in Table 7-10 and Table 7-11
before and after functions are enabled.
• If the value of L.Thrp.Time.UL.QCI.1 decreases, uplink VoLTE volume
estimation for dynamic scheduling has taken effect.
Table 7-10 Counter for verifying uplink VoLTE volume estimation for dynamic
scheduling
1526726777 L.Thrp.Time.UL.QCI.1
1526728411 L.Voice.VQI.UL.Excellent.Times
1526728412 L.Voice.VQI.UL.Good.Times
1526728413 L.Voice.VQI.UL.Accept.Times
1526728414 L.Voice.VQI.UL.Poor.Times
1526728415 L.Voice.VQI.UL.Bad.Times
After voice characteristic awareness scheduling is enabled, the uplink VoLTE packet loss rate is
expected to decrease. You can verify that by observing the following counters.
1526727961 L.Traffic.UL.PktLoss.Loss.QCI.1
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
1526745995 L.Traffic.UL.FarUE.PktLoss.Loss.QCI.1
1526745997 L.Traffic.UL.FarUE.PktLoss.Tot.QCI.1
7.2.1 Principles
The eNodeB performs the following types of optimization to reduce the uplink scheduling delay
and improve user experience in the call setup phase:
• In the call setup phase, more data is scheduled for cell-center voice service UEs as
triggered by scheduling requests (SRs).
• When QCI-based smart preallocation is enabled for QCI 5 services, smart
preallocation takes effect only for QCI 5 services.
However, these two types of optimization reduce the uplink spectral efficiency.
They are controlled by the CellUlschAlgo.VolteCallSetupSchEnhTimer parameter.
• If this parameter is set to 0, neither SR-based scheduling optimization nor smart
preallocation optimization takes effect.
• If this parameter is set to a value in the range of (0, 10], only smart preallocation
optimization takes effect.
• If the value of this parameter is greater than 10, both SR-based scheduling
optimization and smart preallocation optimization take effect. The duration of SR-
based scheduling optimization is specified by the
CellUlschAlgo.VolteCallSetupSchEnhTimer parameter.
7.2.2 Network Analysis
7.2.2.1 Benefits
Function Description
Call-based DRX This function is recommended if all of the following conditions are met:
exit optimization • DRX or dynamic DRX is enabled.
• SIP messages are not encrypted, or the null encryption algorithm
is used for SIP messages.
• Normal preallocation and smart preallocation are both disabled.
• DRX is enabled for QCI 5, and the long DRX cycle is greater
than 160 ms.
7.2.2.2 Impacts
Network Impacts
▪ L.Cdrx.Active.TtiNum/L.Cdrx.Sleep.TtiNum
Function Impacts
7.2.3 Requirements
7.2.3.1 Licenses
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
RAT Function Function Switch Reference Description
Name
None
7.2.3.3 Hardware
For FDD, the following base stations are compatible with this function:
• 3900 and 5900 series base stations
• DBS3900 LampSite and DBS5900 LampSite
Boards
No requirements
RF Modules
No requirements
7.2.3.4 Others
No requirements
7.2.4 Operation and Maintenance
Before using MML commands, refer to 7.2.2.2 Impacts and 7.2.3.2 Software and complete the
parameter configurations for related functions based on the impact and dependency relationships
between the functions, as well as the actual network scenario.
//Enabling call-based DRX exit optimization
MOD CELLALGOSWITCH: LocalCellId=0, VoLTESwitch=VoLTECallConnDelayOptSw-1;
Start SIP-INVITE message tracing and Uu interface tracing. If the tracing results indicate that
the release IE in drx-Config as shown in Figure 7-6 was sent after a SIP message had been sent,
call-based DRX exit optimization has taken effect.
Figure 7-6 release IE in drx-Config
After voice call connection delay optimization is enabled, the average delay of processing
downlink packets for QCI 5 DRBs in a cell decreases.
Average delay of processing downlink packets for QCI 5 DRBs in a cell =
L.Traffic.DL.PktDelay.Time.QCI.5/L.Traffic.DL.PktDelay.Num.QCI.5
During uplink VoLTE continuous scheduling, VoLTE UEs are continuously scheduled in the
uplink during talk spurts to reduce the uplink scheduling delay. This improves voice quality. It
decreases the uplink packet delay and the jitter for VoLTE services.
The subfunctions described in 7.3.1 Active Scheduling of Voice Service UEs, 7.3.2 Uplink Service Status
Determination, and 7.3.3 Enhanced Compensation Scheduling During Talk Spurts are recommended when
DRX is enabled or inter-frequency networking is used.
7.3.1 Active Scheduling of Voice Service UEs
7.3.1.1 Principles
Active scheduling of voice service UEs enables the eNodeB to estimate the size of the voice
packets to be transmitted by UEs, after the UEs send SRs to the eNodeB. The eNodeB then
continuously schedules the UEs within the subsequent uplink subframes until all the voice
packets are transmitted.
This function reduces the delay from the time the eNodeB receives the BSR to the time the UL
Grant is delivered, which reduces the uplink scheduling delay and lowers the packet loss rate due
to timeout.
In FDD, this function takes effect for voice service UEs that are within the network coverage and
use dynamic scheduling. It is controlled by the UlVoLTEContinuousSchSw option of the
CellUlschAlgo.UlEnhencedVoipSchSw parameter. For this function to take effect, the
UlVoipSchOptSwitch option of this parameter must also be selected and uplink semi-persistent
scheduling must not be activated for VoLTE UEs.
7.3.1.2 Network Analysis
7.3.1.2.1 Benefits
Active scheduling of voice service UEs improves voice quality because it reduces the uplink
scheduling delay for voice service UEs and minimizes the probability of packet loss caused by
timeout at the PDCP layer.
7.3.1.2.2 Impacts
Network Impacts
The number of uplink PRBs occupied by VoLTE UEs may slightly increase. Consequently, there
will be less resources available for data services, slightly decreasing the peak throughput of data
services.
Function Impacts
7.3.1.3 Requirements
7.3.1.3.1 Licenses
7.3.1.3.2 Software
Prerequisite Functions
None
None
7.3.1.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
7.3.1.3.4 Others
No requirements
7.3.1.4 Operation and Maintenance
NOTE:
For active scheduling of voice
service UEs to take effect, select
the UlVoipSchOptSwitch option
of this parameter and disable
uplink semi-persistent scheduling
for VoLTE UEs.
Before using MML commands, refer to 7.3.1.2.2 Impacts and complete the parameter
configurations for related functions based on the impact relationships between the functions, as
well as the actual network scenario.
//Enabling active scheduling of voice service UEs
MOD CELLULSCHALGO: LocalCellId=0,
UlEnhencedVoipSchSw=UlVoLTEContinuousSchSw-1&UlVoipSchOptSwitch-1;
• In FDD:
If the value of Uplink Packet Loss Rate (VoIP) decreases, this function has taken effect.
Uplink Packet Loss Rate (VoIP) = L.Traffic.UL.PktLoss.Loss.QCI.1/L.Traffic.UL.PktLoss.Tot.QCI.1
7.3.1.4.5 Network Monitoring
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
7.3.2 Uplink Service Status Determination
7.3.2.1 Principles
7.3.2.2.1 Benefits
In gap-assisted measurement scenarios, uplink service status determination reduces the packet
delay and jitter of voice services, improving voice quality.
7.3.2.2.2 Impacts
Network Impacts
Function Impacts
7.3.2.3 Requirements
7.3.2.3.1 Licenses
Prerequisite Functions
None
None
7.3.2.3.3 Hardware
Boards
No requirements
RF Modules
No requirements
7.3.2.3.4 Others
No requirements
7.3.2.4 Operation and Maintenance
Before using MML commands, refer to 7.3.2.2.2 Impacts and complete the parameter
configurations for related functions based on the impact relationships between the functions, as
well as the actual network scenario.
//Enabling uplink service status determination
MOD CELLULSCHALGO: LocalCellId=0,
UlEnhencedVoipSchSw=UlVoLTEContinuousSchSw-1&UlVoipSchOptSwitch-1;
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
7.3.3 Enhanced Compensation Scheduling During Talk Spurts
7.3.3.1 Principles
7.3.3.2.1 Benefits
If DRX is used, enhanced compensation scheduling during talk spurts reduces the packet delay
and jitter of voice services, improving voice quality.
7.3.3.2.2 Impacts
Network Impacts
In DRX scenarios, enhanced compensation scheduling during talk spurts increases the
consumption of PDCCH CCEs and PUSCH RBs for voice service UEs. If there are a large
number of voice service UEs, the cell throughput will decrease.
Function Impacts
simultaneously for
individual UEs.
• These two functions
can both be enabled
for individual cells.
7.3.3.3 Requirements
7.3.3.3.1 Licenses
Prerequisite Functions
None
None
7.3.3.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
7.3.3.3.4 Others
No requirements
7.3.3.4 Operation and Maintenance
Before using MML commands, refer to 7.3.3.2.2 Impacts and complete the parameter
configurations for related functions based on the impact relationships between the functions, as
well as the actual network scenario.
//Enabling enhanced compensation scheduling during talk spurts
MOD CELLULSCHALGO: LocalCellId=0,
UlEnhencedVoipSchSw=UlVoLTEContinuousSchSw-1&UlVoipSchOptSwitch-1;
If the value of Uplink Packet Loss Rate (VoIP) decreases, this function has taken effect.
Uplink Packet Loss Rate (VoIP) = L.Traffic.UL.PktLoss.Loss.QCI.1/L.Traffic.UL.PktLoss.Tot.QCI.1
7.3.3.4.5 Network Monitoring
For details, see 4.4.3.2 Voice QoS and 4.4.3.3 Voice Quality.
7.4 UL Compensation Scheduling
7.4.1 Principles
Uplink data transmission relies on the SRs sent by UEs. If an eNodeB fails to detect SRs, the
scheduling may be delayed. This may increase the voice packet wait time or even cause timeout-
triggered packet loss. Figure 7-7 shows the situation before UL compensation scheduling is
enabled.
Figure 7-7 Uplink compensation scheduling not enabled
UL compensation scheduling is a technique where the eNodeB identifies voice service UEs and
measures the length of time a UE is not scheduled in the uplink. If the duration reaches a certain
threshold, the eNodeB actively sends a UL Grant to the UE to ensure that uplink voice packets
are transmitted in a timely manner. In this way, this function shortens the voice packet waiting
time and reduces the number of packets discarded because of the expiry of PDCP Discard Timer.
Figure 7-8 shows the situation after UL compensation scheduling is enabled.
7.4.2.1 Benefits
UL compensation scheduling reduces the value of Uplink Packet Loss Rate (VoIP) when there is
interference on the PUCCH, shortens voice packet delay, and improves voice quality.
7.4.2.2 Impacts
Network Impacts
Function Impacts
7.4.3 Requirements
7.4.3.1 Licenses
Prerequisite Functions
None
None
7.4.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
7.4.3.4 Others
No requirements
7.4.4 Operation and Maintenance
• The
UlVoi
option
determ
VoLTE
Before using MML commands, refer to 7.4.2.2 Impacts and complete the parameter configurations
for related functions based on the impact relationships between the functions, as well as the
actual network scenario.
//Setting parameters related to UL compensation scheduling
MOD CELLULSCHALGO: LocalCellId=0, UlEnhencedVoipSchSw=UlVoipSchOptSwitch-
1&UlVoipServStateEnhancedSw-1, UlCompenSchPeriodinSpurt=INTERVAL_20,
UlCompenSchPeriodinSilence=INTERVAL_160;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling UL compensation scheduling
MOD CELLULSCHALGO: LocalCellId=0, UlEnhencedVoipSchSw=UlVoipSchOptSwitch-0;
1. On the MAE-Access, start performance monitoring tasks for the counters listed in
Table 7-18.
After UL compensation scheduling is enabled, you can monitor the following counters to check
whether the uplink VoLTE packet loss rate decreases.
1526727961 L.Traffic.UL.PktLoss.Loss.QCI.1
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
7.5.1 Principles
For details on the principles and engineering guidelines of adaptive modulation and coding
(AMC), mainly MCS selection for uplink dynamic scheduling, see Scheduling.
The eNodeB adjusts SINRs for UEs to be dynamically scheduled in the uplink, based on the
specified uplink target IBLER, and then preliminarily selects MCSs for the UEs.
To reduce the VoLTE packet loss rate and VoLTE packet delay, operators can set different target
IBLERs for voice, video, and data services, as shown in Table 7-19.
Table 7-19 Parameters related to the target IBLER
Typical Application Parameter Description
Video service UEs and voice CellQciPara.InitDlTargetIbler Specifies the target IBLER
service UEs for the downlink.
• For voice and video service UEs, the parameters listed in Table 7-19 take effect irrespective of the
target IBLER adaptation function. For data service UEs, the parameters listed in Table 7-19 take effect
only when the target IBLER adaptation function is disabled. For details about target IBLER adaptation,
see Scheduling.
• The CellQciPara.SinrAdjustTargetIbler and CellQciPara.InitDlTargetIbler parameters take effect
only when services are set up, modified, or released.
Voice-specific AMC and uplink RLC segmentation enhancement can be enabled simultaneously.
Uplink RLC segmentation enhancement takes precedence if its requirements are met.
7.5.2 Network Analysis
7.5.2.1 Benefits
This function improves VoLTE service quality by reducing the VoLTE packet loss rate and
VoLTE packet delay.
7.5.2.2 Impacts
Network Impacts
Function Impacts
7.5.3 Requirements
7.5.3.1 Licenses
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
7.5.3.3 Hardware
Base Station Models
No requirements
Boards
No requirements
RF Modules
No requirements
7.5.3.4 Others
No requirements
7.5.4 Operation and Maintenance
Before using MML commands, refer to 7.5.2.2 Impacts and 7.5.3.2 Software and complete the
parameter configurations for related functions based on the impact and dependency relationships
between the functions, as well as the actual network scenario.
//Setting parameters related to voice-specific AMC
MOD CELLQCIPARA: LocalCellId=0, Qci=1, SinrAdjustTargetIbler=10,
InitDlTargetIbler=10;
On the MAE-Access, check the values of the counters listed in Table 7-21. If the uplink IBLER of
uplink voice services converges on the configured target IBLER, this function has taken effect.
Uplink IBLER of uplink voice services = (L.Traffic.UL.SCH.QPSK.ErrTB.Ibler.QCI.1 +
L.Traffic.UL.SCH.16QAM.ErrTB.Ibler.QCI.1 +
L.Traffic.UL.SCH.64QAM.ErrTB.Ibler.QCI.1)/(L.Traffic.UL.SCH.QPSK.TB.QCI.1 +
L.Traffic.UL.SCH.16QAM.TB.QCI.1 + L.Traffic.UL.SCH.64QAM.TB.QCI.1) x 100%
1526737730 L.Traffic.UL.SCH.QPSK.ErrTB.Ibler.QCI.1
1526737731 L.Traffic.UL.SCH.16QAM.ErrTB.Ibler.QCI.1
1526737732 L.Traffic.UL.SCH.64QAM.ErrTB.Ibler.QCI.1
1526737724 L.Traffic.UL.SCH.QPSK.TB.QCI.1
1526737725 L.Traffic.UL.SCH.16QAM.TB.QCI.1
1526737726 L.Traffic.UL.SCH.64QAM.TB.QCI.1
After voice-specific AMC is enabled, you can observe the following counters to monitor whether
the uplink VoLTE packet loss rate is affected.
1526727961 L.Traffic.UL.PktLoss.Loss.QCI.1
1526727962 L.Traffic.UL.PktLoss.Tot.QCI.1
Overview
VoLTE user prior access for mobile-originated calls enables the eNodeB to identify calling UEs
and perform differentiated handling. Calling UEs include UEs accessing LTE voice services,
voice service UEs performing redirection-based EPS fallback, and voice service UEs performing
handover-based EPS fallback. This function allows preferential access of voice service UEs over
data service UEs and increases the voice service setup success rate. It includes basic and
enhanced functions which are applicable to different types of calling UEs. Table 7-22 lists the
function switches and types of calling UEs.
Table 7-22 Overview of VoLTE user prior access for mobile-originated calls
Basic Function
The mo-VoiceCall-v1280 information element (IE) has been added to the RRC Connection
Request message, indicating a cause of RRC connection setup, in 3GPP Release 12. In heavy-
load scenarios, basic VoLTE user prior access for mobile-originated calls enables the eNodeB to
identify a UE as a calling UE based on the mo-VoiceCall-v1280 IE in the RRC Connection
Request message from the UE and take the following measures to allow preferential access of
this voice service UE:
• Optimizing admission control, congestion control, and flow control
• Raising the preemption priority of voice service UEs
As shown in Figure 7-9 and Figure 7-10, the eNodeB sends a SIB2 carrying the voice service cause
indication. After receiving an RRC Connection Request message from a UE, the eNodeB
identifies the UE as a calling UE based on the IE in this message.
For VoLTE UEs, the eNodeB sends a SIB2 carrying the voiceServiceCauseIndication-r12 IE. For details, see
section 5.3.3 "Actions related to transmission of RRCConnectionRequest message" in 3GPP TS 36.331 V16.4.0.
For ViLTE UEs, if the MO_VILTE_CALL_IND_SW option of the VolteAlgoConfig.VolteOptSwitch parameter
is selected, the eNodeB sends a SIB2 carrying the videoServiceCauseIndication-r14 IE. For details, see section 5.3.3
"Actions related to transmission of RRCConnectionRequest message" in 3GPP TS 36.331 V16.4.0.
Figure 7-9 Major messages for identifying a calling VoLTE UE
After identifying a calling UE, the eNodeB takes the following measures to increase the call
setup success rate and improve voice service UE performance:
• DRX optimization
DRX does not take effect on the default bearer of each identified voice service UE.
This eliminates the impact of sleep time on the SIP messages carried on the QCI 5
bearer and increases scheduling probability for the SIP messages. DRX does not take
effect until the QCI 1 bearer is set up.
• Admission and congestion control optimization
The eNodeB increases the ARP priority level of each identified voice service UE. This
allows the identified voice service UE to preempt the resources of UEs with low ARP
priority levels when the permissible number of UEs is limited, which ensures
preferential access of this UE.
• CA optimization
To minimize the impact of gap-assisted measurements on voice service performance,
the eNodeB reduces the probability of performing CA on each identified voice service
UE. For details on CA, see Carrier Aggregation.
CA optimization takes effect for identified UEs accessing LTE voice services when
any of the following conditions is met:
▪ The CaMgtCfg.VolteCaA2RsrpThld parameter is set to 0.
▪ The CaMgtCfg.VolteCaA2RsrpThld parameter is set to 255 and the
VolteSupportCaInterFreqMeasSw option of the
CaMgtCfg.CellCaAlgoSwitch parameter is selected.
CA optimization takes effect for identified voice service UEs performing redirection-
based EPS fallback when all of the following conditions are met:
▪ The CaMgtCfg.VolteCaA2RsrpThld parameter is set to 0.
▪ The EPSFB_CARR_CONFIG_INHERIT_SW option of the
VolteAlgoConfig.VolteOptSwitch parameter is selected.
• Flow control optimization
The eNodeB increases the priority for each identified voice service UE in flow control
over RRC connection request messages for initial access, reducing the number of
voice service UEs under flow control.
DRX, CA, and flow control are optimized by default. Optimization on admission and congestion
control is controlled by the CellRacThd.VolteArpOverride parameter.
Enhanced Function
Enhanced VoLTE user prior access for mobile-originated calls enables the eNodeB to identify
calling UEs based on the following IEs and perform corresponding optimization to ensure
preferential access of these voice service UEs:
• The mo-VoiceCall-v1280 IE contained in the RRC Connection Request message when
the MO_VOICE_CALL_IND_SW option of the VolteAlgoConfig.VolteOptSwitch
parameter is selected for UEs accessing LTE voice services and voice service UEs
performing redirection-based EPS fallback
• The iMSvoiceEPSfallbackfrom5G IE contained in the "Source eNB to Target eNB
Transparent Container" field in the handover request for voice service UEs performing
handover-based EPS fallback
After identifying a calling UE, the eNodeB takes the following measures to increase the voice
service setup success rate and improve voice service UE performance:
• DRX optimization
DRX does not take effect on the QCI 5 bearer of each identified voice service UE.
This eliminates the impact of sleep time on the SIP messages carried on the QCI 5
bearer and increases scheduling probability of the SIP messages. DRX does not take
effect until the QCI 1 bearer is set up.
• Continuous scheduling
Continuous scheduling takes effect for identified voice service UEs. The eNodeB
proactively sends uplink scheduling indications to the UEs. The interval at which
continuous scheduling is performed is specified by the
VolteAlgoConfig.EpsFbUlActiveSchMinPeriod parameter, and the volume of data
transmitted in continuous scheduling is specified by the
VolteAlgoConfig.EpsFbUlActiveSchDataVol parameter. This optimization remains in
effect until the QCI 1 bearer is set up.
• Optimized uplink scheduling based on UE categories
Optimized uplink scheduling based on UE categories takes effect for identified voice
service UEs to optimize the TBS for scheduling.
▪ For UEs accessing LTE voice services or voice service UEs performing
redirection-based EPS fallback
If an IE indicating a UE of category 0 is contained in Msg3, data scheduling
is performed based on the TBS of UE category 0.
If no IE indicating a UE of category 0 is contained in Msg3, data scheduling
is performed based on the TBS of UE category 1.
▪ For voice service UEs performing handover-based EPS fallback
Data scheduling is performed based on the TBS corresponding to the UE
category indicated in Msg3.
• Delayed UE capability query
Delayed UE capability query takes effect for identified voice service UEs, shortening
the call delay.
▪ UE E-UTRAN capability
For a UE accessing LTE voice services or a voice service UE performing
redirection-based EPS fallback, an E-UTRAN capability query is initiated
without delay.
For a voice service UE performing handover-based EPS fallback, the E-
UTRAN capability is synchronized with the E-UTRAN capability
information sent from the NG-RAN through the core network, and no E-
UTRAN capability query is required.
▪ UE GERAN/UTRAN capability
For an identified voice service UE, a GERAN/UTRAN capability query is
initiated without delay.
▪ UE NG-RAN capability
For a UE accessing LTE voice services or a voice service UE performing
redirection-based EPS fallback, the NG-RAN capability query is delayed
for a period specified by the GlobalProcSwitch.MoVoiceCallUeCapbQryDelay
parameter.
For a voice service UE performing handover-based EPS fallback, the NG-
RAN capability query is delayed for a period specified by the
GlobalProcSwitch.EpsFbHoInUeCapbQryDelay parameter.
• CA optimization
To minimize the impact of gap-assisted measurements on voice service performance,
the eNodeB reduces the probability of performing CA on each identified voice service
UE. For details on CA, see Carrier Aggregation.
CA optimization takes effect for identified UEs accessing LTE voice services when
any of the following conditions is met:
▪ The CaMgtCfg.VolteCaA2RsrpThld parameter is set to 0.
▪ The CaMgtCfg.VolteCaA2RsrpThld parameter is set to 255 and the
VolteSupportCaInterFreqMeasSw option of the
CaMgtCfg.CellCaAlgoSwitch parameter is selected.
CA optimization takes effect for identified voice service UEs performing redirection-
or handover-based EPS fallback when all of the following conditions are met:
▪ The CaMgtCfg.VolteCaA2RsrpThld parameter is set to 0.
▪ The EPSFB_CARR_CONFIG_INHERIT_SW option of the
VolteAlgoConfig.VolteOptSwitch parameter is selected.
7.6.2 VoLTE User Prior Access for Mobile-terminated Calls
Overview
VoLTE user prior access for mobile-terminated calls requires VoLTE marking in PS paging. The
eNodeB selects called UEs from identified voice service UEs based on the mo-VoiceCall-v1280
IE sent from the UEs. The eNodeB then performs differentiated handling to improve voice user
experience.
1. The eNodeB sends SIB1 with an extension capability indicator to the UE, informing
the UE that this function is supported.
2. The eNodeB identifies the VoLTE paging priority of the voice service in the S1 PS
paging message. Then, the eNodeB informs the UE that this is a voice service, by
sending an RRC paging message over the radio interface.
3. The UE accesses the network by sending an RRC Connection Request message that
contains the mo-VoiceCall-v1280 IE. After interpreting this IE, the procedure
continues the same way as that for VoLTE user prior access for mobile-originated
calls.
Figure 7-11 Major messages for identifying a called UE
7.6.3.1 Benefits
This function increases the voice call setup success rate and improves VoLTE UE performance.
This function increases the priorities of identified voice service UEs in flow control over RRC
Connection Request messages, reducing the number of voice service UEs under flow control and
increasing the call setup success rate.
This function reduces the scheduling waiting time for SIP messages and increases the PUSCH
RB usage.
To achieve optimal benefits, you are advised to evaluate when to use the function according to
Table 7-23 and Table 7-24. If the conditions described in Table 7-23 are not met, there will be limited
benefits, but no negative impacts.
Table 7-23 When to use
Scenario Description
1526732721 L.Traffic.User.VoIP.Avg
1526727379 L.Traffic.User.Max
1526729949 L.RRC.SetupFail.ResFail.UserSpec
1526728490 L.RRC.SetupFail.Rej.FlowCtrl
1526736867 L.Cell.UserLic.Limit.Num.PLMN
1526736869 L.Cell.UserLic.Limit.Num
7.6.3.2 Impacts
Network Impacts
Function Impacts
7.6.4 Requirements
7.6.4.1 Licenses
VoLTE user prior access for mobile-originated calls requires the licenses listed in the following
table.
VoLTE user prior access for mobile-terminated calls is not under license control.
7.6.4.2 Software
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
FDD VoLTE user VoLTEMoPrefSwitch option of the VoLTE VoLTE user prior access for
prior access CellAlgoSwitch.VoLTESwitch parameter mobile-terminated calls can
for mobile- take effect only if basic
originated VoLTE user prior access for
calls mobile-originated calls is
enabled.
FDD Device-pipe SpecUeIdentifySwitch option of the Specified VoLTE user prior access for
identification UeCooperationPara.SpecUserCooperationSwitch User mobile-terminated calls can
Coordinated
parameter Scheduling take effect only if device-
RAT Function Function Switch Reference Description
Name
pipe identification is
enabled.
None
7.6.4.3 Hardware
For FDD, the following base stations are compatible with this function:
• 3900 and 5900 series base stations
• DBS3900 LampSite and DBS5900 LampSite
Boards
• Fast EPS fallback: Only the LBBPc is incompatible with this function.
• Other functions: no requirements
RF Modules
No requirements
7.6.4.4 Others
UE
in
Before using MML commands, refer to 7.6.3.2 Impacts and 7.6.4.2 Software and complete the
parameter configurations for related functions based on the impact and dependency relationships
between the functions, as well as the actual network scenario.
//(For VoLTE UEs) Enabling basic VoLTE user prior access for mobile-
originated calls
MOD CELLALGOSWITCH: LocalCellId=0, VoLTESwitch=VoLTEMoPrefSwitch-1;
//(For ViLTE UEs) Enabling basic VoLTE user prior access for mobile-
originated calls
MOD CELLALGOSWITCH: LocalCellId=0, VoLTESwitch=VoLTEMoPrefSwitch-1;
MOD VOLTEALGOCONFIG: LocalCellId=0, VolteOptSwitch=MO_VILTE_CALL_IND_SW-1;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling VoLTE user prior access
MOD CELLALGOSWITCH: LocalCellId=0, VoLTESwitch=VoLTEMoPrefSwitch-0;
MOD VOLTEALGOCONFIG: LocalCellId=0, VolteOptSwitch=FAST_EPS_FALLBACK_SW-0;
To verify activation of basic VoLTE user prior access for mobile-originated calls, perform the
following steps:
1. On the MAE-Access, start a Uu interface tracing task. In the tracing results, verify
that SIB2 contains "voiceServiceCauseIndication-r12:true" and
rrcConnectionRequest contains "establishmentCause:mo-VoiceCall-v1280". Figure 7-
12 and Figure 7-13 show examples.
Figure 7-12 voiceServiceCauseIndication-r12:true
2. On the MAE-Access, start performance monitoring tasks for the counters listed in
Table 7-26.
1526745657 L.RRC.ConnReq.Att.MoVoiceCall
1526745658 L.RRC.ConnReq.Succ.MoVoiceCall
Counter ID Counter Name
1526745656 L.RRC.ConnReq.Msg.MoVoiceCall
To verify activation of enhanced VoLTE user prior access for mobile-originated calls, perform
the following steps:
1. On the MAE-Access, start a Uu interface tracing task. In the tracing results, observe
mo-VoiceCall-v1280 for UEs accessing LTE voice services and voice service UEs
performing redirection-based EPS fallback. Alternatively, start an S1 interface
tracing task on the MAE-Access. In the tracing results, observe
iMSvoiceEPSfallbackfrom5G:true in the S1AP_HANDOVER_REQ message for
voice service UEs performing handover-based EPS fallback. Figure 7-14 and Figure 7-15
show examples.
Figure 7-14 mo-VoiceCall-v1280
2. Observe signaling messages to verify whether the following optimizations have taken
effect.
• DRX optimization
DRX does not take effect on the QCI 5 bearer but takes effect after the
QCI 1 bearer is set up. After the QCI 1 bearer is set up, observe the IEs
related to DRX. The example shown in Figure 7-16 indicates that the
optimization has taken effect.
Figure 7-16 DRX-related IEs after the QCI 1 bearer is set up
This section describes how to monitor VoLTE user prior access for mobile-originated calls.
After this function is activated, the VoLTE call setup success rate increases. You can monitor the
following counters to evaluate network performance:
RRC Setup Success Rate (Service) x E-RAB Setup Success Rate (VoLTE) = RRC Setup Success Rate
(Service) x (L.E-RAB.SuccEst.QCI.1/L.E-RAB.AttEst.QCI.1) x 100%
If congestion occurs, the VoLTE call setup success rate and the number of VoLTE UEs increase.
You can monitor L.Traffic.User.VoIP.Avg and L.Traffic.User.VoIP.Max to evaluate network
performance.
7.7 Preferential Access of Voice Services
7.7.1 Principles
Overview
With this function, the eNodeB reserves UE resources for VoLTE UEs. When the number of
online UEs in a cell exceeds the difference between the maximum allowed number of UEs in the
cell and the reserved number of VoLTE UEs, VoLTE UEs use the reserved resources or preempt
data service resources to ensure their access. The eNodeB releases data service UEs or redirects
these UEs to inter-frequency or inter-RAT cells.
The VoltePrefAdmissionSwitch option of the CellAlgoSwitch.RacAlgoSwitch parameter
specifies whether to preferentially admit voice services. The CellRacThd.VolteReservedNumber
parameter specifies the reserved number of VoLTE UEs.
Voice services refer to VoLTE services, but not CSFB or over the top (OTT) voice services.
UE Identification
For UEs that do not support the mo-VoiceCall-v1280 IE (introduced in 3GPP Release 12), the
eNodeB cannot identify service types during RRC connection setup. Currently, VoLTE UEs can
be identified only during E-RAB setup.
Only UEs that have been identified as voice service UEs are allowed to occupy the resources
reserved for VoLTE. The eNodeB identifies voice service UEs involved in RRC connection
setup, RRC connection reestablishment, or intra-RAT incoming handovers according to the
following principles.
• For UEs involved in RRC connection setup
▪ If the RRC connection setup cause is "MO-data", "MT-access", or "MO-
Voice" and a QCI 1 bearer is successfully set up for a UE before the timer
specified by the CellRacThd.VoltePrefAdmissionTimer parameter expires, the
eNodeB identifies the UE as a voice service UE.
▪ If the RRC connection setup cause is "MO-data", "MT-access", or "MO-
Voice" but a QCI 1 bearer has not been successfully set up for a UE before
the timer specified by the CellRacThd.VoltePrefAdmissionTimer parameter
expires, the eNodeB identifies the UE as a data service UE.
▪ If the RRC connection setup cause is not "MO-data", "MT-access", or
"MO-Voice" for a UE, the eNodeB identifies the UE as a data service UE.
• For UEs involved in RRC connection reestablishment or intra-RAT incoming
handovers
▪ If a QCI 1 bearer is successfully set up for a UE before the timer specified
by the CellRacThd.VoltePrefAdmissionTimer parameter expires, the eNodeB
identifies the UE as a voice service UE.
▪ If a QCI 1 bearer has not been successfully set up for a UE before the timer
specified by the CellRacThd.VoltePrefAdmissionTimer parameter expires, the
eNodeB identifies the UE as a data service UE.
UE Access Processing
When preferential access of voice services is enabled and the number of online UEs in a cell
exceeds the difference between the maximum allowed number of UEs in the cell and the
reserved number of VoLTE UEs, the base station handles voice and data service UEs as follows:
• If the VoLTEPreemptionSwitch option of CellAlgoSwitch.RacAlgoSwitch is
deselected:
▪ VoLTE UEs that newly access or are handed over to the cell use the UE
resources reserved for VoLTE services to access voice services.
▪ The base station redirects newly admitted data service UEs to inter-
frequency or inter-RAT cells or directly releases them.
• If the VoLTEPreemptionSwitch option of CellAlgoSwitch.RacAlgoSwitch is selected:
▪ Newly admitted VoLTE UEs preempt resources allocated to low-priority
data service UEs. The base station redirects those data service UEs to inter-
frequency or inter-RAT cells or directly releases them. If the preemption
fails, these VoLTE UEs use the UE resources reserved for VoLTE UEs to
access voice services. For details on preemption, see Admission and Congestion
Control.
▪ The base station redirects newly admitted data service UEs to inter-
frequency or inter-RAT cells or directly releases them.
The ENodeBAlgoSwitch.RedirectSwitch parameter specifies whether to redirect or release data
service UEs. It is recommended that inter-frequency or inter-RAT redirection be used. If there is
no inter-frequency or inter-RAT neighboring cell, the eNodeB directly releases data service UEs.
When the UE resources reserved for VoLTE services are used up, VoLTE UEs cannot preferentially access the cell.
7.7.2 Network Analysis
7.7.2.1 Benefits
If preferential access of voice services is enabled, more VoLTE UEs can gain preferential access
to the network using reserved resources. This increases the voice call setup success rate, the cell
capacity in terms of voice service UEs, and the number of online VoLTE UEs, and improves
VoLTE UE performance.
You are advised to evaluate when to use the function according to Applicable Scenarios and Non-
Applicable Scenarios to achieve optimal benefits.
Applicable Scenarios
This function is recommended when there is a small chance that the maximum number of UEs in
a cell (indicated by L.Traffic.User.Max) approaches the maximum allowed value. In other scenarios,
this function brings limited benefits but no negative impacts.
When the number of online UEs frequently approaches the maximum allowed number of UEs in a cell, capacity
expansion or hardware upgrade is recommended.
1526727379 L.Traffic.User.Max
1526726767 L.Traffic.DRB.QCI.1
Non-Applicable Scenarios
Network Impacts
• If preferential access of voice services is enabled, data service UEs cannot use
reserved UE resources. If no VoLTE UEs access the cell later, the reserved resources
may be wasted, decreasing the system capacity.
• This function affects the number of online data service UEs, the intra-RAT handover
success rate, and the service drop rate, depending on the setting of the
CellRacThd.VolteReservedNumber parameter:
▪ If this parameter is set to a smaller value, fewer VoLTE UEs will access a
cell using reserved UE resources. When the reserved resources are used up,
VoLTE UEs cannot preferentially access the given cell. This has a slight
impact on the number of data service UEs, the intra-RAT handover success
rate, and the service drop rate.
▪ If this parameter is set to a larger value, there are fewer data service UEs,
the intra-RAT handover success rate decreases, and the service drop rate
increases. When there are a large number of VoLTE UEs at the same time,
cell throughput decreases.
Function Impacts
FDD Cell outage None Cell Outage Cell outage detection and
detection and Detection and compensation and preferential
Compensation
compensation access of voice services can be
enabled for the same cell at the
same time.
Voice services are given
preferential treatment when the
number of online UEs in a cell
exceeds the difference between
the maximum allowed number
of UEs in the cell and the
reserved number of VoLTE
UEs. When this happens, the
eNodeB releases RRC
connections or rejects RRC
connection requests of non-
VoLTE UEs. If this status lasts
for a long time, the cell may be
incorrectly identified as an
outage cell. Therefore, when
preferential access of voice
services takes effect, the
eNodeB does not process the
cell outage results reported by
RAT Function Function Switch Reference Description
Name
FDD RAN sharing None RAN Sharing RAN sharing and preferential
access of voice services can be
enabled at the same time.
For RAN sharing features, the
number of RRC_CONNECTED
UEs can be licensed on a per
operator basis. However, this
number is not cell-specific.
Therefore, when both RAN
sharing and preferential access
of voice services are enabled,
reserved UE resources for
preferential access of voice
services are not on a per
operator basis.
7.7.3 Requirements
7.7.3.1 Licenses
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
7.7.3.3 Hardware
Base Station Models
No requirements
Boards
No requirements
RF Modules
No requirements
7.7.3.4 Others
Preferential access of voice services requires the precondition function on the IMS. With the
precondition function, a QCI 1 bearer is set up before a voice session starts. This shortens the
time for identifying voice service UEs.
7.7.4 Operation and Maintenance
and Table 7-29 describe the parameters used for function activation and optimization,
Table 7-28
respectively.
Table 7-28 Parameters used for activation
RAT Parameter Parameter ID Option Setting Notes
Name
Before using MML commands, refer to 7.7.2.2 Impacts and 7.7.3.2 Software and complete the
parameter configurations for related functions based on the impact and dependency relationships
between the functions, as well as the actual network scenario.
//Enabling preferential access of voice services
MOD CELLALGOSWITCH: LocalCellId=0, RacAlgoSwitch=VoltePrefAdmissionSwitch-
1&VoltePreemptionSwitch-1;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling preferential access of voice services
MOD CELLALGOSWITCH: LocalCellId=0, RacAlgoSwitch=VoltePrefAdmissionSwitch-0;
5. Enable the new UE to perform data services such as web browsing in the cell.
In the S1 interface tracing result on the MAE-Access, verify that the eNodeB has sent
the UE CONTEXT RELEASE REQUEST message to release a data service UE,
with the cause value of reduce-load-in-serving-cell.
Figure 7-22 UE CONTEXT RELEASE REQUEST
6. On the MAE-Access, view the live network counters listed in Table 7-30. This
function has taken effect if the values of these counters increase.
Table 7-30 Counters
Counter ID Counter Name
1526739725 L.VoiceUser.VoIPPref.Succ
1526739726 L.HHO.Prep.FailIn.AdmitFail.VoIPPref
1526739727 L.HHO.InterFddTdd.Prep.FailIn.AdmitFail.VoIPPref
1526739728 L.RRC.ReEstFail.VoIPPref
1526739730 L.UECNTX.Rel.eNodeB.VoIPPref.PreEmpSucc
1526739731 L.UECNTX.Rel.eNodeB.VoIPPref.Fail
After preferential access of voice services is enabled, data service UEs experience UE context
release, RRC connection reestablishment failures, and incoming handover failures. The values of
the counters listed in Table 7-31 are expected to increase.
Table 7-31 Counters
1526739725 L.VoiceUser.VoIPPref.Succ
1526739726 L.HHO.Prep.FailIn.AdmitFail.VoIPPref
1526739727 L.HHO.InterFddTdd.Prep.FailIn.AdmitFail.VoIPPref
1526739728 L.RRC.ReEstFail.VoIPPref
Counter ID Counter Name
1526739730 L.UECNTX.Rel.eNodeB.VoIPPref.PreEmpSucc
1526739731 L.UECNTX.Rel.eNodeB.VoIPPref.Fail
7.8.1 Principles
Basic PUSCH RB reservation for voice service UEs is recommended in heavy-traffic scenarios.
This function allows a certain number of RBs at a specified position to be reserved for voice
service UEs on the PUSCH. Voice service UEs can preferentially use reserved RB resources. If
the reserved RB resources are used up, voice service UEs can use non-reserved RB resources.
However, non-voice service UEs cannot use reserved RB resources.
Basic PUSCH RB reservation for voice service UEs is enabled when both of the following
conditions are met:
• The UlVoipRbRsvSwitch option of the CellAlgoSwitch.UlSchExtSwitch parameter is
selected.
• The VolteAlgoConfig.VoiceRbRsvRsrpThld parameter is set to 0.
The CellUlschAlgo.UlVoipRsvRbStart parameter specifies the start position of reserved RBs, and
the CellUlschAlgo.UlVoipRsvRbNum parameter specifies the number of reserved RBs.
Enhanced PUSCH RB reservation for voice service UEs is recommended when there is heavy
traffic and interference to the PUSCH.
This function allows a certain number of RBs at a specified position on the PUSCH to be
reserved for cell-edge coverage-limited voice service UEs.
• If the uplink RSRP of a voice service UE in the cell is less than or equal to the value
of the VolteAlgoConfig.VoiceRbRsvRsrpThld parameter, the UE can preferentially use
the reserved RBs.
In case the reserved RBs are insufficient, the eNodeB allocates only the remaining
reserved RBs to the UE. If the reserved RBs have been exhausted, non-reserved RBs
can be used.
• If the uplink RSRP of a voice service UE in the cell is greater than the value of the
VolteAlgoConfig.VoiceRbRsvRsrpThld parameter, the UE can use only non-reserved
RBs.
• Non-voice-service UEs in the cell can use only non-reserved RBs.
Enhanced PUSCH RB reservation for voice service UEs is enabled when both of the following
conditions are met:
• The UlVoipRbRsvSwitch option of the CellAlgoSwitch.UlSchExtSwitch parameter is
selected.
• The VolteAlgoConfig.VoiceRbRsvRsrpThld parameter is set to a value within the range
of -140 to -43.
The CellUlschAlgo.UlVoipRsvRbStart parameter specifies the start position of reserved RBs, and
the CellUlschAlgo.UlVoipRsvRbNum parameter specifies the number of reserved RBs.
7.8.2 Network Analysis
7.8.2.1 Benefits
Applicable PUSCH RB reservation for voice service UEs is recommended when all of
scenarios the following conditions are met:
• L.UL.Interference.Avg ≥ –110 dBm
• Uplink Resource Block Utilizing Rate > 50%
Scenario Conditions
Average L.UL.Interference.Avg
interference and
noise detected on
each uplink PRB
The number of voice service UEs in a cell and number of RBs used by voice service UEs are used to determine the
number of reserved RBs in a cell. The interference received by PRBs on the PUSCH helps determine the position of
reserved RBs in a cell.
7.8.2.2 Impacts
Network Impacts
• When RB resources are reserved, fewer resources are available for uplink data and
interference increases accordingly. This decreases the uplink traffic volume and uplink
cell throughput, increases the uplink bit error rate of common data service UEs, and
decreases the uplink MCS indexes for data service UEs.
• When PUSCH RB reservation for voice service UEs is enabled in heavy-load
scenarios, the KPIs of data service UEs may deteriorate. The impact on data service
UEs increases as more RBs are reserved.
Function Impacts
FDD eMTC EMTC_SWITCH option of the eMTC If the RBs reserved in PUSCH
introduction CellEmtcAlgo.EmtcAlgoSwitch RB reservation for voice
parameter service UEsa are in conflict
with the RBs being used by
eMTC services, voice service
UEs cannot use these RBs
before eMTC data
transmission is complete.
After eMTC data transmission
is complete, voice service
UEs can use these RBs.
Before PUSCH RB
reservation for voice service
UEs is disabled, these RBs
cannot be used by eMTC
services.
a: "PUSCH RB reservation for voice service UEs" refers to both basic and enhanced PUSCH RB
reservation for voice service UEs.
7.8.3 Requirements
7.8.3.1 Licenses
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
7.8.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
7.8.3.4 Others
No requirements
7.8.4 Operation and Maintenance
The relevant parameter settings must be identical between the serving cell and its intra-frequency neighboring cells
to ensure that voice service UEs use the same RBs in the frequency domain.
Before using MML commands, refer to 7.8.2.2 Impacts and 7.8.3.2 Software and complete the
parameter configurations for related functions based on the impact and mutually exclusive
relationships between the functions, as well as the actual network scenario.
//Enabling PUSCH RB reservation for voice service UEs
MOD CELLALGOSWITCH: LocalCellId=0, UlSchExtSwitch=UlVoipRbRsvSwitch-1;
MOD VOLTEALGOCONFIG: LocalCellId=0, VoiceRbRsvRsrpThld=-120;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling PUSCH RB reservation for voice service UEs
MOD CELLALGOSWITCH: LocalCellId=0, UlSchExtSwitch=UlVoipRbRsvSwitch-0;
View the following performance indicator to monitor the running status of basic PUSCH RB
reservation for voice service UEs:
• Counters related to the average interference and noise (L.UL.Interference.Avg.PRB0
through L.UL.Interference.Avg.PRB99)
If the average interference and noise on the reserved RBs are expected to be
significantly lower than those on the non-reserved RBs, this function is running
properly.
• Uplink Packet Loss Rate (VoIP), which is equal to L.Traffic.UL.PktLoss.Loss.QCI.1 divided by
L.Traffic.UL.PktLoss.Tot.QCI.1
The uplink packet loss rate of cell-edge voice service UEs is expected to decrease.
7.9 Uplink Voice Mute Recovery
7.9.1 Principles
7.9.2.1 Benefits
This function shortens the duration of voice mute and improves user experience with voice
services.
7.9.2.2 Impacts
Network Impacts
Function Impacts
None
7.9.3 Requirements
7.9.3.1 Licenses
Before activating this function, ensure that its prerequisite functions have been activated and
mutually exclusive functions have been deactivated. For detailed operations, see the relevant
feature documents.
Prerequisite Functions
None
7.9.3.3 Hardware
No requirements
Boards
No requirements
RF Modules
No requirements
7.9.3.4 Others
No requirements
7.9.4 Operation and Maintenance
Before using MML commands, refer to 7.9.3.2 Software and complete the parameter configurations
for related functions based on the dependency relationships between the functions, as well as the
actual network scenario.
//Enabling uplink voice mute recovery
MOD CELLULSCHALGO: LocalCellId=0,
UlEnhencedVoipSchSw=UlCallMuteRecoverSwitch-1;
The following provides only deactivation command examples. You can determine whether to
restore the settings of other parameters based on actual network conditions.
//Disabling uplink voice mute recovery
MOD CELLULSCHALGO: LocalCellId=0,
UlEnhencedVoipSchSw=UlCallMuteRecoverSwitch-0;
On the MAE-Access, start performance monitoring tasks for the counters listed in Table 7-36. If
the value of any counter listed in Table 7-36 increases, uplink voice mute recovery has taken
effect.
Table 7-36 Counters for verifying uplink voice mute recovery
1526742144 L.HHO.IntraCell.CallMute.ExecAttOut
1526742148 L.UECNTX.Rel.eNodeB.CallMute
1526742151 L.RRCRedirection.IntraLTE.CallMute
1526742152 L.RRCRedirection.InterFddTdd.CallMute
You can view the counters listed in Table 7-37 to monitor the running status of the uplink voice
mute recovery function.
Table 7-37 Counters used to monitor uplink voice mute recovery
1526742146 L.HHO.IntraCell.CallMute.ExecSuccOut
1526742144 L.HHO.IntraCell.CallMute.ExecAttOut
1526742148 L.UECNTX.Rel.eNodeB.CallMute
1526742151 L.RRCRedirection.IntraLTE.CallMute
1526742152 L.RRCRedirection.InterFddTdd.CallMute
8 Power Saving
This chapter describes how the basic feature LBFD-002017 DRX works for VoLTE.
Overview
As shown in Figure 8-1, DRX allows UEs to enter the sleep state to save power when data is not
being transmitted. DRX is typically used for services with consecutive small packets that are
transmitted periodically, such as voice services. The CellDrxPara.DrxAlgSwitch parameter
specifies whether to enable DRX.
Figure 8-1 DRX working for voice services
Enabling DRX for QCI 1 can reduce the battery consumption of voice services, but affects voice
quality. For example, the packet loss rate increases. For details on DRX and its impacts, see DRX
and Signaling Control.
• In FDD
Similar to the default bearer for a data service, the default bearer for a VoLTE service
(that is, the signaling bearer with a QCI of 5) always exists as long as the UE is in
RRC_CONNECTED mode, regardless of whether the UE has a QCI 1 voice bearer.
Therefore, it is recommended that the DRX parameters for QCI 5 be set to the same
values as those for a default bearer used in a data service. In addition, as a long DRX
cycle leads to a long delay for setting up a QCI 1 voice bearer, to provide fast access
for VoLTE UEs, it is recommended that the long DRX cycle for QCI 5 not exceed 320
ms.
False detection of the PDCCH may cause voice packet loss. Because of this, there is a low probability that voice
quality will deteriorate. Preallocation can be used to reduce the impact of false detection. For details on how DRX
works with preallocation, see DRX and Signaling Control.
When a voice service UE is performing both a gap-assisted intra-RAT measurement and DRX,
the voice quality of the UE deteriorates. Setting CellDrxPara.VolteGapDrxExclusiveSwitch to ON
can eliminate the impact and improve voice quality. This parameter controls mutual
exclusiveness between gap-assisted measurement and DRX.
With this function enabled in DRX scenarios, the eNodeB measures the uplink or downlink
packet loss rate of each UE performing QCI 1 services. If either the uplink or downlink packet
loss rate of a UE is greater than the CellDrxPara.VoltePlrThldForExitingDrx parameter value, the
eNodeB instructs the UE to exit DRX, reducing the impact of DRX on voice quality.
This function reduces the QCI 1 packet loss rate, but the following will also decrease: the
amount of power saved, the number of UEs in the DRX state, and the duration in which VoLTE
UEs stay in the DRX-defined active state. If a UE in a weak-coverage area fails to receive a
signaling message for exiting the DRX state, the QCI 1 call drop rate will increase.
In FDD, if exit from DRX based on voice quality is activated after the
CellDrxPara.DrxSrDetectOptSwitch parameter is set to ON, the downlink BLER increases.
When a voice service UE in the DRX state experiences low uplink SINR, the eNodeB instructs
the UE to exit DRX. This function reduces the packet loss rate and delay of voice service UEs
and improves voice quality. This function is controlled by the
CellDrxPara.SinrThldForVolteDrxCtrl parameter.
• If this parameter is set to -100, this function does not take effect.
• If this parameter is set to a value other than -100, this function takes effect. When the
uplink SINR of a voice service UE in the DRX state is less than the value of the
CellDrxPara.SinrThldForVolteDrxCtrl parameter, the eNodeB instructs the UE to exit
DRX.
After SINR-based exit from DRX is enabled, downlink scheduling delay decreases because
downlink voice packets are no longer bundled.
Downlink Packet Bundling for Voice Service UEs in the On Duration (FDD)
In scenarios with a high probability of false SR detection, the eNodeB may schedule the QCI 1
services of UEs during the DRX-defined sleep time in the downlink, because of inconsistent
DRX states between the eNodeB and the UEs. This scheduling increases the downlink packet
loss rate of QCI 1 services and affects voice quality.
To relieve this impact, downlink packet bundling for voice service UEs in the On Duration has
been introduced. It is controlled by the CellDrxPara.DrxSrDetectOptSwitch parameter. When this
parameter is set to ON, the eNodeB does not perform downlink scheduling for QCI 1 services
upon detecting an SR during the DRX-defined sleep time. Downlink scheduling for QCI 1
services is resumed only after the eNodeB enters the DRX-defined On Duration.
This function reduces the impacts of false SR detection on voice quality, but it increases the
downlink packet delay of QCI 1 services. If a QCI 1 bearer is accompanied by other bearers
where data is transmitted, the downlink packet delay of those bearers will also increase.
9 Mobility Management
9.1 Overview
This chapter describes the mobility of voice services. It is recommended that coverage-based
intra-RAT and inter-RAT handovers for voice services be enabled by default to ensure voice
service continuity.
Voice and data service handovers are classified into intra-frequency, inter-frequency, and inter-
RAT handovers. Their basic handover procedures are the same. However, certain handover
parameters can be set on a per QCI basis.
• During a coverage-based, distance-based, or UL-quality-based handover, if the UE is
running services with different QCIs, the handover parameters mapped onto the
highest-priority QCI take effect.
The QCI priority is specified by the CellQciPara.QciPriorityForHo parameter. A smaller
value of this parameter indicates a higher priority.
If different QCIs have the same priority specified by the CellQciPara.QciPriorityForHo
parameter, the eNodeB uses the handover parameters mapped onto the 3GPP-defined
highest-priority QCI for the UE. For details, see section 6.1.7 "Standardized QoS
characteristics" in 3GPP TS 23.203 V10.3.0.
• For details on handovers caused by other reasons, see Mobility Management in Connected
Mode.
VoLTE is generally deployed in existing LTE networks, which are data networks. Mobility
parameters in existing LTE networks have been optimized constantly and can satisfy the KPI
requirements of data services.
• The QoS requirements differ depending on voice and data services. It is recommended
that inter-frequency or inter-RAT mobility parameters be separately configured for
voice and data services after VoLTE is deployed.
For data services, the inter-frequency or inter-RAT mobility parameters optimized in
the existing network are recommended. For voice services, the default inter-frequency
or inter-RAT mobility parameters are recommended. These recommendations
minimize the impact on the KPIs of data services in future network optimization on
VoLTE services.
• For intra-frequency mobility, relative thresholds are used and can be separately
configured for voice and data services. However, separate configurations are not
recommended.
Like the default bearer for a data service, the signaling bearer (QCI 5) for a VoLTE service
always exists as long as the UE is in RRC_CONNECTED mode. Therefore, mobility parameters
for QCI 5 can be set to the same values as those for the default bearer used in a data service.
9.2 Intra-Frequency Handover
This section lists the handover parameters that are set based on QCIs in different intra-frequency
handover scenarios. These parameters are as follows:
• IntraFreqHoGroup.IntraFreqHoA3Hyst
• IntraFreqHoGroup.IntraFreqHoA3Offset
• IntraFreqHoGroup.IntraFreqHoA3TimeToTrig
For details on the principles for intra-frequency handovers, see Mobility Management in Connected
Mode.
This section describes the handover parameters that are set based on QCIs in different inter-
frequency handover scenarios.
For details on the principles for inter-frequency handovers, see Mobility Management in Connected
Mode.
• InterFreqHoGroup.InterFreqHoA5Thd1Rsrp
• InterFreqHoGroup.InterFreqHoA5Thd1Rsrq
When a UE initiates a voice service, the eNodeB delivers the A4 measurement configuration to the UE,
instructing the UE to measure the frequency identified by the ServiceIfDlEarfcnGrp.DlEarfcn
parameter.
• Set the CnOperatorQciPara.Qci parameter to define the mapping between the inter-
frequency handover frequency policy and QCI 1 of an operator.
• Set the CellQciPara.QciPriorityForHo parameter to assign the highest handover priority
to QCI 1.
• Initial planning and configuration are recommended for neighboring cells on
frequencies that carry only VoLTE services.
The following handover parameters related to events A1 and A2 can be set based on QCIs:
• InterFreqHoGroup.FreqPriInterFreqHoA1ThdRsrp
• InterFreqHoGroup.FreqPriInterFreqHoA1ThdRsrq
• InterFreqHoGroup.FreqPriInterFreqHoA2ThdRsrp
• InterFreqHoGroup.FreqPriInterFreqHoA2ThdRsrq
Similar to service-based inter-frequency handover, frequency-priority-based inter-frequency
handover uses certain A4-related parameters that can be specified based on QCIs. For details, see
Service-based Inter-Frequency Handover.
This section describes the handover parameters that are set based on QCIs in different inter-RAT
handover scenarios.
For details about inter-RAT handovers, see Mobility Management in Connected Mode.
The following common inter-RAT handover parameters can be set based on QCIs:
• InterRatHoCommGroup.InterRatHoA1A2Hyst
• InterRatHoCommGroup.InterRatHoA1A2TimeToTrig
• InterRatHoCommGroup.InterRatHoA1ThdRsrp
• InterRatHoCommGroup.InterRatHoA1ThdRsrq
• InterRatHoCommGroup.InterRatHoA2ThdRsrp
• InterRatHoCommGroup.InterRatHoA2ThdRsrq
The handover parameters listed in the following table can be set based on QCIs for each RAT.
Table 9-1 Parameters related to coverage-based inter-RAT handover
UTRAN • InterRatHoUtranGroup.InterRatHoUtranB1ThdEcn0
• InterRatHoUtranGroup.InterRatHoUtranB1ThdRscp
• InterRatHoUtranGroup.InterRatHoUtranB1Hyst
• InterRatHoUtranGroup.InterRatHoUtranB1TimeToTrig
• InterRatHoUtranGroup.LdSvBasedHoUtranB1ThdEcn0
• InterRatHoUtranGroup.LdSvBasedHoUtranB1ThdRscp
GERAN • InterRatHoGeranGroup.InterRatHoGeranB1Hyst
Target RAT Parameter ID
• InterRatHoGeranGroup.InterRatHoGeranB1Thd
• InterRatHoGeranGroup.InterRatHoGeranB1TimeToTrig
• InterRatHoGeranGroup.LdSvBasedHoGeranB1Thd
CDMA2000 • InterRatHoCdmaHrpdGroup.InterRatHoCdmaB1Hyst
HRPD • InterRatHoCdmaHrpdGroup.InterRatHoCdmaB1ThdPs
• InterRatHoCdmaHrpdGroup.InterRatHoCdmaB1TimeToTrig
• InterRatHoCdmaHrpdGroup.Cdma2000HrpdB2Thd1Rsrp
• InterRatHoCdmaHrpdGroup.Cdma2000HrpdB2Thd1Rsrq
With service-based inter-RAT handover, voice services are handed over from the E-UTRAN to
the GERAN or UTRAN in the service setup phase.
If VoLTE has been deployed, to prevent the eNodeB from immediately triggering the service-
based inter-RAT handover procedure after UEs set up voice services, the
UtranServiceHoSwitch and GeranServiceHoSwitch options of the
ENodeBAlgoSwitch.HoAlgoSwitch parameter must be set in one of two ways:
• Both options are deselected.
• Both options are selected, and the ServiceIrHoCfgGroup.InterRatHoState parameter
for QCI 1 and QCI 5 is not set to MUST_HO.
The eNodeB supports distance-based inter-RAT handovers only to the GERAN or UTRAN.
The handover parameters listed in the following table can be set based on QCIs for each RAT.
These parameters also apply to coverage-based inter-RAT handovers.
Table 9-2 Distance-based inter-RAT handover parameters
UTRAN • InterRatHoUtranGroup.InterRatHoUtranB1ThdEcn0
• InterRatHoUtranGroup.InterRatHoUtranB1ThdRscp
• InterRatHoUtranGroup.InterRatHoUtranB1Hyst
• InterRatHoUtranGroup.InterRatHoUtranB1TimeToTrig
• InterRatHoUtranGroup.LdSvBasedHoUtranB1ThdEcn0
• InterRatHoUtranGroup.LdSvBasedHoUtranB1ThdRscp
GERAN • InterRatHoGeranGroup.InterRatHoGeranB1Hyst
• InterRatHoGeranGroup.InterRatHoGeranB1Thd
Target RAT Parameter ID
• InterRatHoGeranGroup.InterRatHoGeranB1TimeToTrig
• InterRatHoGeranGroup.LdSvBasedHoGeranB1Thd
Certain handover parameters for UL-quality-based inter-RAT handover can be configured based
on QCIs, similar to the situation for coverage-based inter-RAT handovers. For details, see
Coverage-based Inter-RAT Handover.
The eNodeB evaluates whether to use SRVCC or PS handover for executing the preceding inter-
RAT handovers so that the CS or PS domain of the target network can carry voice services. The
eNodeB selects a handover policy based on conditions such as UE capability and whether the
target network can carry IMS-based voice services.
SRVCC
With SRVCC, the eNodeB hands over voice services from E-UTRAN to the CS domain of the
GERAN or UTRAN to ensure voice call continuity. For details, see SRVCC.
PS Handover
With PS handovers, the eNodeB hands over voice services from the E-UTRAN to the PS domain
of the UTRAN to ensure voice call continuity.
If the UTRAN supports the IMS-based VoHSPA voice solution, voice services can be handed
over to VoHSPA of the UTRAN using inter-RAT PS handovers. In this situation, the mechanism
of inter-RAT PS handovers for voice services is the same as that for data services. For details,
see the descriptions of handover as a handover policy in Mobility Management in Connected Mode.
ANR
If ANR is enabled, the eNodeB acts as follows when selecting UEs to perform fast ANR
measurement:
• The eNodeB does not select the UEs with QCI 1 bearers when selecting UEs that have
accessed the cell through initial access or handovers.
• For UEs that have been selected for fast ANR measurement, the eNodeB uses a
measurement configuration policy depending on the value of the
GlobalProcSwitch.VoipWithGapMode parameter.
▪ When this parameter is set to ENABLE, the eNodeB does not delete the
fast ANR measurement configuration for the UEs when QCI 1 bearers are
set up for the UEs later. The continual gap-assisted measurements may
affect the quality of QCI 1 services.
▪ When this parameter is set to DISABLE, the eNodeB sends RRC
Connection Reconfiguration messages to delete fast ANR measurements
when QCI 1 bearers are set up for the UEs later.
• UEs read cell global identifications (CGIs) during DRX-defined sleep time. The sleep
time stops when a data packet arrives. This has an impact on the CGI reading success
rate. This success rate is even lower for VoLTE services because voice packets are
scheduled at a fixed interval and it is more probable that the sleep time stops. After a
CGI is acquired, the identified cell is automatically configured as a neighboring cell.
Event-triggered ANR can be triggered at coverage-based handover measurements or MLB-based
measurements. Gap-assisted measurement will be started for voice service UEs to avoid service
drops when coverage-based handovers are triggered. Voice service UEs are not selected for
User-number-based MLB.
To ensure voice quality, it is recommended that GlobalProcSwitch.VoipWithGapMode be set to
DISABLE when deploying VoLTE.
After the ANR-based proactive PCI conflict detection function of the LOFD-002007 PCI
Collision Detection and Self-Optimization feature is enabled, the eNodeB selects UEs without
QCI 1 bearers for proactive PCI conflict detection. In addition, the eNodeB sends A3 and A4
measurement configurations to these UEs, instructing them to measure their serving frequencies
and neighboring E-UTRAN frequencies. For UEs that have been selected, the eNodeB uses a
measurement configuration policy depending on the value of the
GlobalProcSwitch.VoipWithGapMode parameter.
• When this parameter is set to ENABLE, the eNodeB does not delete the measurement
configuration for the UEs when QCI 1 bearers are set up for the UEs later. The
continual gap-assisted measurements may affect the quality of QCI 1 services.
• When this parameter is set to DISABLE, the eNodeB sends RRC Connection
Reconfiguration messages to delete the measurements if QCI 1 bearers are set up for
the UEs later.
• When a voice service is initiated for a UE with carriers aggregated, the UE may or
may not stay in the CA state, depending on the setting of the
CaMgtCfg.VolteCaA2RsrpThld parameter.
▪ If this parameter is set to 0, the UE exits the CA state.
▪ If this parameter is set to 255, the UE can stay in the CA state and the voice
service is scheduled only in the primary cell (PCell).
▪ If this parameter is set to a value within the range of -140 to -43 and:
▪ If the RSRP of the PCell is less than or equal to the value of
CaMgtCfg.VolteCaA2RsrpThld, the UE exits the CA state.
▪ If the RSRP of the PCell is greater than the value of
CaMgtCfg.VolteCaA2RsrpThld, the UE can stay in the CA state
and the voice service is scheduled only in the PCell.
• A UE with a voice service ongoing may or may not enter the CA state, depending on
the setting of the CaMgtCfg.VolteCaA2RsrpThld parameter.
▪ If this parameter is set to 0, the UE will not enter the CA state.
▪ If this parameter is set to 255, the UE may or may not enter the CA state, as
described in the following table.
▪ If this parameter is set to a value within the range of -140 to -43 and:
▪ If the RSRP of the PCell is less than or equal to the value of
CaMgtCfg.VolteCaA2RsrpThld, the UE will not enter the CA state.
▪ If the RSRP of the PCell is greater than the value of
CaMgtCfg.VolteCaA2RsrpThld, the UE may or may not enter the
CA state, as described in the following table.
Scenario Description
GSM and LTE Buffer Zone Optimization (FDD), MDT and Periodic Measurement Reporting
For GSM and LTE buffer zone optimization (FDD), Minimization of Drive Tests (MDT), or
periodic measurement reporting, the eNodeB selects UEs for measurement. The eNodeB uses a
measurement configuration policy for voice service UEs depending on the value of the
GlobalProcSwitch.VoipWithGapMode parameter:
• Set to ENABLE:
The eNodeB randomly selects UEs. When a voice service UE is selected, continual
gap-assisted measurements may affect the voice quality of the UE.
• Set to DISABLE:
▪ If a voice service has been initiated for a UE, the eNodeB does not select
the UE for periodic inter-frequency or inter-RAT measurement
configurations. The eNodeB will not deliver these configurations to the UE
after the voice service is released. The UE will be evaluated again next time
it is selected.
▪ If no voice service has been initiated for a UE, the eNodeB sends periodic
inter-frequency or inter-RAT measurement configurations to the UE. If a
voice service is initiated for the UE later, the eNodeB deletes these
measurement configurations while sending new configurations to the UE.
The eNodeB will not deliver inter-frequency or inter-RAT measurement
configurations to the UE after the voice service is released. The UE will be
evaluated again next time it is selected.
MLB
After LOFD-001109 DL Non-GBR Packet Bundling is enabled, the scheduling priority of voice
services is no longer the highest. This feature is not recommended in VoLTE-enabled cells when
there are many voice service UEs. That is because activating this feature may slightly affect
voice quality.
AC Barring Skip
When AC barring skip (FDD) is enabled, the eNodeB restricts the access of UEs depending on
their access causes and service types in congestion scenarios. This helps increase the VoLTE
access success rate. For details, see Access Class Control.
The following formula is recommended for calculating the average uplink interference on the PUCCH:
Average uplink interference on the PUCCH = (–121 x L.UL.Interference.PUCCH.Index0 – 120 x
L.UL.Interference.PUCCH.Index1 – 119 x L.UL.Interference.PUCCH.Index2 – 118 x
L.UL.Interference.PUCCH.Index3 – 117 x L.UL.Interference.PUCCH.Index4 – 116 x
L.UL.Interference.PUCCH.Index5 – 115 x L.UL.Interference.PUCCH.Index6 – 114 x
L.UL.Interference.PUCCH.Index7 – 113 x L.UL.Interference.PUCCH.Index8 – 112 x
L.UL.Interference.PUCCH.Index9 – 108 x L.UL.Interference.PUCCH.Index10 – 104 x
L.UL.Interference.PUCCH.Index11 – 100 x L.UL.Interference.PUCCH.Index12 – 96 x
L.UL.Interference.PUCCH.Index13 – 92 x L.UL.Interference.PUCCH.Index14 – 91 x
L.UL.Interference.PUCCH.Index15)/(L.UL.Interference.PUCCH.Index0 +
L.UL.Interference.PUCCH.Index1 + L.UL.Interference.PUCCH.Index2 +
L.UL.Interference.PUCCH.Index3 + L.UL.Interference.PUCCH.Index4 +
L.UL.Interference.PUCCH.Index5 + L.UL.Interference.PUCCH.Index6 +
L.UL.Interference.PUCCH.Index7 + L.UL.Interference.PUCCH.Index8 +
L.UL.Interference.PUCCH.Index9 + L.UL.Interference.PUCCH.Index10 +
L.UL.Interference.PUCCH.Index11 + L.UL.Interference.PUCCH.Index12 +
L.UL.Interference.PUCCH.Index13 + L.UL.Interference.PUCCH.Index14 +
L.UL.Interference.PUCCH.Index15)
For details about the RSRP threshold for PUCCH power control and PUCCH SINRTarget, see Power
Control.
If the channel condition changes dramatically on the radio interface or there is false CQI
detection, the MCS index selected for voice service UEs suddenly changes, causing voice
packets to be lost.
Optimized CQI adjustment is implemented through the CellCqiAdjAlgo.VolteRptCqiDiffLimit
parameter. This parameter is used to set the CQI change amount in case of sudden CQI change.
• If this parameter is set to 0, this function does not take effect.
• If this parameter is set to a value other than 0, this function takes effect. This
parameter limits CQI increase amount so that moderate MCS indexes are selected for
voice service UEs, reducing voice packet loss caused by the use of high MCS indexes.
In FDD, this limit does not take effect for eMTC UEs running voice services, if the
VOLTE_REPETITION_OPT_SWITCH option of the
CellEmtcAlgo.EmtcEnhancedVolteAlgoSw parameter is selected.
Table 10-1 describes some of CQI adjustment algorithms that do not take effect on voice service
UEs.
Table 10-1 CQI adjustment algorithms ineffective for voice service UEs
The CQI adjustment algorithms controlled by the preceding options are used for data service UEs. These algorithms
are intended for increasing throughput and enabling the BLER to approach its target value. They may alter the
BLER. Therefore, voice service UEs that are highly sensitive to the BLER will be affected by these algorithms.
This function enables the eNodeB to increase uplink power for VoLTE services when the uplink
RBLER meets a certain condition. It prevents the VoLTE packet losses caused by delayed uplink
power control in abrupt interference scenarios. In FDD, this function also enables the eNodeB to
select initial SINR adjustment values for UEs based on the cell-level adjustment value, making
MCS selection more reliable.
Uplink AMC enhancement for VoLTE is controlled by the EnhancedUlVoipAmcSw option of
the CellUlschAlgo.UlEnhencedVoipSchSw parameter.
• If this option is deselected, this function does not take effect.
• If this option is selected, this function takes effect. The eNodeB performs the
following operations for voice service UEs with QCI 1 bearers:
▪ Increases power when detecting exceptions.
The eNodeB increases the uplink power by 3 dB each time it detects that
the power headroom is not reported within a certain period or that residual
block errors (errors that exist after the maximum number of retransmissions
is reached) occur twice in the uplink within a certain period.
▪ Changes the power headroom reporting period to 100 ms.
▪ (FDD) Selects more accurate UE-level initial SINR adjustment values
(denoted by SinrAdj), based on the cell-level SinrAdj, for UEs that access
the cell after this option is selected.
▪ (FDD) Reduces the MCS index based on the amount of data to be
scheduled, after the previous MCS adjustment is complete, to adapt to the
volume of voice data.
▪ (FDD) Changes the SINR to be compared with the PUSCH SINR upper
limit from the average SINR (denoted by AverageSinr) to (AverageSinr +
Min(SinrAdj, 0)) if the cell-level SinrAdj is less than –10 dB. The purpose is
to prevent incorrect power decreases.
When uplink AMC enhancement for VoLTE is enabled:
• RBLER-triggered uplink power increase is a more timely action of power control. It
helps reduce the uplink packet loss rate of QCI 1 services.
• (FDD) MCS selection is more reliable due to initial SINR adjustment value selection
for voice service UEs based on the cell-level adjustment value.
• The increase in power raises the interference to neighboring cells and causes a slight
decrease in the user-perceived uplink data rate.
• (FDD) If the ENodeBAlgoSwitch.HoWithSccCfgAddBlindSwitch parameter is set to
OFF, the SCell configuration success rate increases.
This function improves call experience for UEs that would enter idle mode because of UE
inactivity timer expiration when experiencing a services drop at the application layer. With this
function, the eNodeB postpones UE context release so that these UEs can initiate or receive calls
as soon as possible.
VoLTE call delay optimization is controlled by the
RrcConnStateTimer.VoiceUeReleaseDelayTimer parameter.
• If this parameter is set to 0, this function does not take effect.
• If this parameter is set to a value other than 0, this function takes effect.
For UEs that would enter idle mode because of UE inactivity timer expiration, the
eNodeB does not release the UE context until the timer defined by the
RrcConnStateTimer.VoiceUeReleaseDelayTimer parameter expires. This prolongs the
duration for the UEs to stay in connected mode.
This function enables the base station to implement corresponding UE guarantee functions for
voice service UEs that are identified as fast-moving UEs.
• Fast-moving UE identification
The eNodeB identifies fast-moving UEs using any of the following methods:
▪ SC-based identification (SC is short for service classification)
The eNodeB can identify a UE as a fast-moving UE when the
FAST_MOVING_UE_FLAG option of the
AsParaGroup.AggregationAttribute parameter for the acceleration guarantee
parameter group (specified by the AsParaGroup.AsParaGroupID parameter)
corresponding to the UE is selected.
▪ QCI-based identification
The eNodeB can identify a UE as a fast-moving UE when the
FAST_MOVING_UE_FLAG option of the QciPara.AggregationAttribute
parameter for the QCI (specified by the QciPara.Qci parameter)
corresponding to the UE is selected.
▪ SPID-based identification (SPID is short for subscriber profile ID)
The eNodeB can identify a UE as a fast-moving UE when the
FAST_MOVING_UE_FLAG option of the SpidCfg.AggregationAttribute
parameter for the SPID (specified by the SpidCfg.Spid parameter)
corresponding to the UE is selected.
• UE guarantee functions
The eNodeB uses the following functions to improve user experience of voice services
for fast-moving UEs only when QCI 1 bearers are set up for these UEs:
▪ Downlink power increase
This function is controlled by the
VolteAlgoConfig.AsVolteDlPwrBoostSinrThld parameter.
▪ If this parameter is set to 127, this function does not take effect.
▪ If this parameter is set to a value other than 127, this function
takes effect. When downlink coverage is limited in non-massive
MIMO scenarios and the downlink SINR for a fast-moving voice
service UE is less than the parameter value, the eNodeB
proactively increases the downlink transmit power for the UE. In
this case, the PDCCH for the UE uses the remaining power after
scheduling, and the PDSCH power is increased during
scheduling.
The LBBPc does not support this function. eMTC UEs and UEs involved in
joint transmission do not support this function.
After this function takes effect, the increase in the downlink power
accelerates and the downlink coverage performance improves for voice
service UEs. However, the cell capacity decreases and the downlink
interference to neighboring cells increases.
▪ Uplink lightweight semi-persistent scheduling
This function is controlled by the
VolteAlgoConfig.FastMovingUeUlLwSpsDataVol parameter.
▪ If this parameter is set to 255, this function does not take effect.
▪ If this parameter is set to a value other than 255, this function
takes effect. The semi-persistent scheduling interval is 640 ms,
and the amount of data that can be scheduled in semi-persistent
scheduling is determined by the
VolteAlgoConfig.FastMovingUeUlLwSpsDataVol parameter. When
the number of zero-payload packets consecutively received by
the eNodeB on the semi-persistently allocated resources is greater
than or equal to eight, the eNodeB automatically releases these
resources.
The LBBPc does not support this function.
For individual UEs, if uplink lightweight semi-persistent scheduling and
uplink semi-persistent scheduling (controlled by the SpsSchSwitch option
of the CellAlgoSwitch.UlSchSwitch parameter) are both enabled, the former
function takes precedence. Other impacted functions and mutually exclusive
functions of uplink lightweight semi-persistent scheduling are the same as
those of uplink semi-persistent scheduling. For details, see 5.1.2.2 Impacts and
5.1.3.2 Software.
▪ L.Sps.UL.SchNum
RB resources are insufficient when the RB resources available for fast-moving voice
service UEs cannot satisfy downlink semi-persistent scheduling in specific scenarios, for
example, where RB blocking (controlled by CellRbReserve.RbRsvMode) is enabled, the
proportion limitation for downlink semi-persistent scheduling (specified by
SpectrumCloud.DlSpsRestrictRatio) is configured, or compact bandwidth (controlled by
Cell.CustomizedBandWidthCfgInd) is enabled. In this case, downlink lightweight semi-
persistent scheduling cannot take effect, and downlink semi-persistent scheduling cannot
take effect either if enabled.
For details about RB blocking, see Scheduling. For details about the proportion limitation
for downlink semi-persistent scheduling, see LTE FDD and NR Spectrum Sharing. For
details about compact bandwidth, see Compact Bandwidth (FDD).
After the status of downlink lightweight semi-persistent scheduling is
changed from disabled to enabled, the values of the following counters
change from zero to a non-zero value:
▪ L.Sps.DL.ErrNum
▪ L.Sps.DL.SchNum
When UE whitelist control is enabled, this function reduces the impact of the procedures specific
to voice service UEs that are capable of EN-DC on voice services, increasing the QCI 1 bearer
setup success rate in NSA networking.
Section 6.1.7 in 3GPP TS 23.203 of Release 10 provides the QoS requirements for services with
standardized QCIs. Table 11-1 lists the QoS requirements for QCI 1 services.
Table 11-1 QoS requirements for QCI 1 services
Resource Type Priority Packet Delay Packet Error Typical Service
Budget (ms) Loss Rate
• Packet Delay Budget indicates the threshold for the delay between the UE and PDN
gateway (P-GW). The corresponding user satisfaction rate is 98%.
• Packet Error Loss Rate indicates the maximum allowed percentage of data-link-
layer SDUs at the TX end that are not successfully sent to the corresponding upper
layer of the RX end.
The uplink and downlink air interface packet loss rates and the downlink PDCP packet loss rate
of QCI 1 services can be measured using counters. For details about related counters, see 4.4.3.2
Voice QoS.
The mean opinion score (MOS) is an important indicator for evaluating voice quality. MOS-
based evaluation involves subjective evaluation, objective evaluation, and measurement-based
evaluation.
11.2.1 Subjective Evaluation
MOS-based subjective evaluation is a process in which users listen to raw speech materials and
processed speech materials, whose voice quality must have degraded, and then score their
quality. These scores are the basis for calculating MOSs.
Table 11-2 lists the MOS standards defined in ITU-T G.107.
Table 11-2 MOS standards
MOS Quality Level Auditory Distortion Extent Required Auditory Effort
Overview
Using a third-party drive test tool to evaluate speech quality is time-consuming and expensive,
and cannot monitor speech quality in real time. To monitor speech quality in VoLTE networks,
vendors have developed their own measurement-based evaluation methods based on the key
factors that determine voice quality. Huawei introduced the Voice Quality Monitoring (VQM)
algorithm.
VQM is used for activities including network monitoring, network optimization, VIP guarantee,
and user complaint handling. This algorithm reduces the necessity of drive tests for obtaining
voice quality. Table 11-3 lists the voice quality indicators (VQIs) provided by VQM and the
corresponding parameters.
Table 11-3 VQIs provided by VQM and the corresponding parameters
VQI Parameter Description
VQM Principles
1. The eNodeB monitors the voice coding rate and performance indicators such as the
frame error rate (FER) of voice packets. If the delay jitter for voice packets with QCI
1 exceeds the VQMAlgo.ULDelayJitter parameter value, the eNodeB considers that
packet loss has occurred.
2. The eNodeB inputs the monitoring results to the Huawei proprietary VQI model.
After inputting the results into the VQI model, the eNodeB produces fitted VQI
values at intervals specified by VQMAlgo.VQMAlgoPeriod. The VQI model is based
on the MOS specified in ITU-T P.863 and uses mathematical formulas for MOS
fitting.
3. The VQI values are stored in call history records (CHRs). Once stored, they are used
to collect the statistics on cell-level voice quality counters and monitor user-level
performance.
The voice quality monitoring results (including statistics about cell-level voice quality counters and
user-level performance monitoring results) and CHRs do not contain users' privacy information.
VQI is a P.863 standard-based estimation method using mathematical formulas fitting. It differs from
tool-based objective evaluation because they have different working principles and measurement
scopes. The MOSs provided by those tools are subject to factors such as test equipment, UE codec
mechanism, and jitter buffer algorithm and may fluctuate greatly. However, VQI values are immune
to UE factors. VQI and MOS have different principles and measurement scopes, and therefore they
are different.
VQM takes effect only in RLC UM mode because voice services are delay-sensitive and generally
work in RLC UM mode.
The eNodeB evaluates voice quality based on the VQI values provided by the VQM algorithm
and voice quality level thresholds. In addition, the eNodeB collects the statistics of counters
related to voice quality listed in Table 11-4.
Table 11-4 Voice quality evaluation rules and related counters
Application Limitations
The VQM algorithm supports AMR-NB, AMR-WB, EVS-WB primary mode (5.9 kbit/s to 24.4
kbit/s), and EVS-SWB primary mode (9.6 kbit/s to 24.4 kbit/s).
VQI values are not calculated in the following scenarios because the eNodeB cannot identify the
voice coding rates of UEs:
• Voice packets are encrypted by UEs.
• Voice packets are fragmented.
• There are no voice packets.
• One RTP packet carries multiple voice frames.
• The eNodeB cannot determine whether SIP messages have been encrypted using the
null algorithm or any other encryption algorithm.
• In scenarios where the eNodeB cannot parse the SIP messages for voice services (see
4.1.6.3 Parsing-Limited Scenarios for details), the eNodeB fails to identify the voice coding
mode in use, or the eNodeB determines that EVS is used, based on the first SID frame.
• During a handover, the source cell does not provide the rate set information to the
target cell. In this case, VQIs are not calculated for EVS calls.
• The 13.2 kbit/s coding rate is changed between the channel aware mode (CAM) and
the non-CAM mode. In this case, VQIs cannot be differentiated and are measured
based on the initial coding rate.
• Multiple ACK messages are received, and the encoding scheme negotiated based on
the current ACK message is different from that based on the previous ACK message.
In this scenario, the encoding scheme negotiated based on the previous ACK message
is used to calculate the VQI value.
When the number of incoming voice packets (excluding SID frames) is less than 60, VQI values
are not calculated because an excessivley small number of samples may cause large errors.
The measurement results of the VQM algorithm used in SIP encryption scenarios may be
inaccurate because voice packets without conversational contents (for example, packets carrying
ring back tone services) are also included in the QCI 1 statistics. As a result, values of counters
related to poor voice quality and mute voice increase.
The voice service capacity can be evaluated based on the performance statistics related to the
number of UEs performing QCI 1 services. For related counters, see 4.4.3.4 Voice Capacity.
Voice services are delay-sensitive. When there are no other alarms on a BBP, VQIs deteriorate if
the average downlink voice processing delay in a cell (ms) or downlink voice packet discard rate
in a cell (%) continuously exceeds their respective thresholds. To address this issue, cell capacity
expansion is recommended, for example, by replacing the BBP with a larger-capacity one.
The performance of voice services can be evaluated using the following KPIs, which are
described in 4.4.3.1 Voice KPIs:
• E-RAB Setup Success Rate (VoIP)
12 Parameters
The following hyperlinked EXCEL files of parameter documents match the software version
with which this document is released.
• Node Parameter Reference: contains device and transport parameters.
• eNodeBFunction Parameter Reference:contains all parameters related to radio access
functions, including air interface management, access control, mobility control, and
radio resource management.
• eNodeBFunction Used Reserved Parameter List: contains the reserved parameters that are in
use and those that have been disused.
You can find the EXCEL files of parameter reference and used reserved parameter list for the software version used
on the live network from the product documentation delivered with that version.
FAQ 1: How do I find the parameters related to a certain feature from parameter
reference?
1. Open the EXCEL file of parameter reference.
2. On the Parameter List sheet, filter the Feature ID column. Click Text Filters and
choose Contains. Enter the feature ID, for example, LOFD-001016 or TDLOFD-
001016.
3. Click OK. All parameters related to the feature are displayed.
FAQ 2: How do I find the information about a certain reserved parameter from the used
reserved parameter list?
1. Open the EXCEL file of the used reserved parameter list.
2. On the Used Reserved Parameter List sheet, use the MO, Parameter ID, and BIT
columns to locate the reserved parameter, which may be only a bit of a parameter.
View its information, including the meaning, values, impacts, and product version in
which it is activated for use.
13 Counters
The following hyperlinked EXCEL files of performance counter reference match the software
version with which this document is released.
• Node Performance Counter Summary: contains device and transport counters.
• eNodeBFunction Performance Counter Summary:
contains all counters related to radio access
functions, including air interface management, access control, mobility control, and
radio resource management.
You can find the EXCEL files of performance counter reference for the software version used on the live network
from the product documentation delivered with that version.
FAQ: How do I find the counters related to a certain feature from performance counter
reference?
1. Open the EXCEL file of performance counter reference.
2. On the Counter Summary(En) sheet, filter the Feature ID column. Click Text
Filters and choose Contains. Enter the feature ID, for example, LOFD-001016 or
TDLOFD-001016.
3. Click OK. All counters related to the feature are displayed.
14 Glossary
15 Reference Documents
• 3GPP TS 23.401, "General Packet Radio Service (GPRS) enhancements for Evolved
Universal Terrestrial Radio Access Network (E-UTRAN) access"
• 3GPP TS 23.203: "Policy and charging control architecture"
• 3GPP TS 36.321: "Medium Access Control (MAC) protocol specification"
• 3GPP TS 26.114: "Multimedia Telephony; Media handling and interaction"
• 3GPP TS 36.331: "E-UTRA; Radio Resource Control (RRC)"
• 3GPP TS 23.221: "Architectural requirements"
• 3GPP TS 26.445: "EVS; Detailed Algorithmic Description"
• 3GPP TS 26.201: "Adaptive Multi-Rate - Wideband (AMR-WB) speech codec; Frame
structure"
• 3GPP TS 36.213: "E-UTRA; Physical layer procedures"
• ITU-T G.107, "The E-model: a computational model for use in transmission planning"
• ROHC
• Scheduling
• Power Control
• QoS Management
• CS Fallback
• CSPC
• eMTC
• SRVCC
• Emergency Call
• Connection Management
• IPsec
• RAN Sharing
• Carrier Aggregation
• Turbo Receiver
• Relay
• UL CoMP
• Synchronization
• Cell Management
• eX2 Self-Management
• 3D Beamforming (FDD)
• MIMO
• DL CoMP (FDD)
• eMIMO (FDD)
• Breathing Pilot