-
eRAN Feature Documentation
Product Version: eRAN8.1
Library Version: 01
Date: 03/23/2015
For any question, please contact us.
Copyright Huawei Technologies Co., Ltd. 2015. All rights
reserved.
Flow Control
Contents
3.6.3 Flow Control
eRAN
Flow Control Feature Parameter Description
Issue 01
Date 2015-03-23
HUAWEI TECHNOLOGIES CO., LTD.
Copyright Huawei Technologies Co., Ltd. 2015. All rights
reserved.No part of this document may be reproduced or transmitted
in any form or by any means without prior written consent of Huawei
Technologies Co., Ltd.
Trademarks and Permissions
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.
NoticeThe 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.
Huawei Technologies Co., Ltd.
Address: Huawei Industrial Base Bantian, Longgang Shenzhen
518129 People's Republic of China
Website: http://www.huawei.com
Email: [email protected]
3.6.3 Contents
1 About This Document
1.1 Scope
1.2 Intended Audience
1.3 Change History
1.4 Differences Between eNodeB Types
2 Overview
2.1 Introduction
2.2 Benefits
2.3 Application Scenarios
2.4 Basic Principles of Flow Control
2.4.1 Control-Plane Data Flows and Protocols
2.4.2 User-Plane Data Flows and Protocols
2.4.3 OAM Data Flows
2.4.4 Logical Architecture of an eNodeB
2.5 Flow Control Mechanism
3 Control-Plane Flow Control
3.1 MME Overloadtriggered Flow Control
3.1.1 Objective
3.1.2 Principle
3.1.3 Monitoring
3.2 Flow Control for Random Access Preamble
3.2.1 Objective
3.2.2 Principle
3.2.2.1 Flow Control Points
3.2.2.2 Flow Control Actions
3.2.3 Monitoring
3.3 Initial Access Request Control
3.3.1 Objective
3.3.2 Principle
3.3.2.1 Flow Control Points
3.3.2.2 Flow Control Actions
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
1 trong 13 5/19/2015 6:28 PM
-
3.3.3 Monitoring
3.4 Paging Message Flow Control
3.4.1 Objective
3.4.2 Principle
3.4.2.1 Flow Control Points
3.4.2.2 Flow Control Actions
3.4.3 Monitoring
3.5 RRC Reject Wait time
3.5.1 Objective
3.5.2 Principle
3.5.3 Monitoring
3.6 AC Control
3.6.1 Objective
3.6.2 Principle
3.6.3 Monitoring
3.7 Uplink Synchronized UE Number Control
3.7.1 Objective
3.7.2 Principle
3.7.3 Monitoring
4 User-Plane Flow Control
4.1 Downlink User-Plane Flow Control
4.1.1 Objective
4.1.2 Principle
4.1.2.1 Flow Control Points
4.1.2.2 Flow Control Actions
4.1.3 Monitoring
4.2 Uplink User-Plane Flow Control
4.2.1 Objective
4.2.2 Principle
4.2.2.1 Flow Control Points
4.2.2.2 Flow Control Actions
4.2.3 Monitoring
5 OAM Flow Control
5.1 OAM Flow Control
5.1.1 Objective
5.1.2 Principle
5.1.3 Monitoring
6 CPU Core Deployment and CPU Usage Monitoring
6.1 LMPT
6.1.1 CPU Core Deployment
6.1.2 CPU Usage Monitoring
6.1.2.1 Counters Related to the CPU Usage on the LMPT
6.1.2.2 Alarm Related to CPU Overload on the LMPT
6.1.2.3 CPU Usage Monitoring on the U2000 or LMT
6.2 UMPT
6.2.1 CPU Core Deployment
6.2.2 CPU Usage Monitoring
6.2.2.1 Counters Related to the CPU Usage on the UMPT
6.2.2.2 Alarm Related to CPU Overload on the UMPT
6.2.2.3 CPU Usage Monitoring on the U2000 or LMT
6.3 LBBP
6.3.1 CPU Core Deployment
6.3.2 CPU Usage Monitoring
6.3.2.1 CPU Usage Counters
6.3.2.2 Alarm Related to CPU Overload on the LBBP
6.3.2.3 CPU Usage Monitoring on the U2000 or LMT
6.4 UBBP
6.4.1 CPU Core Deployment
6.4.2 CPU Usage Monitoring
6.4.2.1 CPU Usage Counters
6.4.2.2 Alarm Related to CPU Overload on the UBBP
6.4.2.3 CPU Usage Monitoring on the U2000 or LMT
7 Parameters
8 Counters
9 Glossary
10 Reference Documents
1 About This Document Scope
This document describes eNodeB flow control, including its
technical principles, related features, network impact, and
engineering guidelines.
Any managed objects (MOs), parameters, alarms, or counters
described herein correspond to the software release delivered with
this document. Any future updates will be described in the product
documentation delivered with future software releases.
This document applies only to LTE FDD. Any "LTE" in this
document refers to LTE FDD, and "eNodeB" refers to LTE FDD
eNodeB.
This document applies to the following eNodeBs.
eNodeB Type Model
Macro 3900 series eNodeBs
LampSite DBS3900 LampSite
Intended Audience
This document is intended for personnel who:
Need to understand the features described herein
Work with Huawei products
Change History
This section provides information about the changes in different
document versions. There are two types of changes:
Feature change
Changes in features and parameters of a specified version
Editorial change
Changes in wording or addition of information and any related
parameters affected by editorial changes
eRAN8.1 01 (2015-03-23)
This issue includes the following changes.
Change Type Change Description Parameter Change Affected
Entity
Feature change None None -
Editorial change Added the following sections:
2.4 Basic Principles of Flow Control
2.5 Flow Control Mechanism
3.2 Flow Control for Random Access Preamble
3.3.2.1 Flow Control Points
3.3.2.2 Flow Control Actions
3.4.2.1 Flow Control Points
3.4.2.2 Flow Control Actions
4.1.2.1 Flow Control Points
4.1.2.2 Flow Control Actions
4.2.1 Objective
4.2.2 Principle
4.2.3 Monitoring
6 CPU Core Deployment and CPU Usage Monitoring
Added the following parameters:
RSCGRPALG.TCSW
RSCGRPALG.CTTH
RSCGRPALG.CCTTH
-
eRAN8.1 Draft A (2015-01-15)
Compared with Issue 01 (2014-04-26) of eRAN7.0, Draft A
(2015-01-15) of eRAN8.1 does not include any changes.
Differences Between eNodeB Types
The features described in this document are implemented in the
same way on macro and LampSite eNodeBs.
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
2 trong 13 5/19/2015 6:28 PM
-
2 Overview Introduction
With flow control, an eNodeB controls input and output flow to
prevent overload and improve equipment stability. Flow control
includes control-plane, user-plane, OAM data flow control, and
uplink synchronized UE number control.
Flow control is implemented in the following ways:
An eNodeB controls input flow to prevent overload of the eNodeB
and ensure the eNodeB's processing capability when services
increase dramatically.
An eNodeB controls output flow to prevent overload of the peer
network element (NE).
Benefits
When services increase dramatically, flow control decreases the
possibility of eNodeB reset and deteriorating access and handover
success rates, thereby improving eNodeB stability.
Application Scenarios
As shown in Figure 2-1, three types of data flows exist in LTE
networks: control-plane, user-plane, and OAM data flows.
Figure 2-1 eNodeB data flows
According to data flow and load, there are eight load points in
LTE networks, as shown in Table 2-1.
Table 2-1 Load points in LTE networks
Data Flow Type Data Flow Load Point
Control-plane data flow Uplink signaling data flows from UEs to
an eNodeB Load point 5: An eNodeB is overloaded if UEs send too
much signaling to the eNodeB.
Uplink signaling data flows from an eNodeB to an MME Load point
1: An MME is overloaded if an eNodeB sends too much signaling to
the MME.
Downlink signaling data flows from an MME to an eNodeB Load
point 3: An eNodeB is overloaded if an MME sends too much signaling
to the eNodeB.
Downlink signaling data flows from an eNodeB to a UE N/A
Signaling data flows between eNodeBs Load point 6: An eNodeB is
overloaded if a peer eNodeB sends too much signaling to the
eNodeB.
User-plane data flow Uplink service data flows from UEs to an
eNodeB Load point 5: An eNodeB is overloaded if UEs send too much
service data to the eNodeB.
Uplink service data flows from an eNodeB to an S-GW Load point
2: An S-GW is overloaded if an eNodeB sends too much service data
to the S-GW.
Downlink service data flows from an S-GW to an eNodeB Load point
4: An eNodeB is overloaded if an S-GW sends too much service data
to the eNodeB.
Downlink service data flows from an eNodeB to a UE N/A
Service data flows between eNodeBs Load point 6: An eNodeB is
overloaded if a peer eNodeB sends too much service data to the
eNodeB.
OAM data flow Uplink OAM data flows from an eNodeB to the
operations support system(OSS)
Load point 7: The OSS is overloaded if an eNodeB sends too much
OAM data to the OSS.
Downlink OAM data flows from the OSS to an eNodeB Load point 8:
An eNodeB is overloaded if the OSS sends too much OAM data to the
eNodeB.
This document describes flow control at the preceding eight load
points.
Basic Principles of Flow Control
Three types of data flows exist in LTE networks:
Control-plane data flow
User-plane data flow
Operation, administration and maintenance (OAM) data flow
Flow control includes intra-eNodeB flow control and flow control
between eNodeBs and other NEs.
The basic principles of flow control include:
Controls UE access and reduces data transmission rates based on
the received back pressure indicator when a module is
overloaded.
Identifies and suppresses low-priority services based on service
priorities to ensure the high-priority services are performed
preferentially.
2.4.1 Control-Plane Data Flows and Protocols
Figure 2-2 shows control-plane data flows:
Uplink signaling data flows from a UE to an eNodeB
Uplink signaling data flows from an eNodeB to an MME
Downlink signaling data flows from an MME to an eNodeB
Downlink signaling data flows from an eNodeB to a UE
Signaling data flows between eNodeBs
Figure 2-2 Control-plane data flows
Figure 2-3 and Figure 2-4 show the protocol stacks related to
the control plane.
Figure 2-3 Control-plane protocol stacks between the UE and the
MME
Figure 2-4 Control-plane protocol stacks between eNodeBs
2.4.2 User-Plane Data Flows and Protocols
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
3 trong 13 5/19/2015 6:28 PM
-
Figure 2-5 shows user-plane data flows:
Uplink service data flows from a UE to an eNodeB
Uplink service data flows from an eNodeB to an S-GW
Downlink service data flows from an S-GW to an eNodeB
Downlink service data flows from an eNodeB to a UE
Service data flows between eNodeBs
Figure 2-5 User-plane data flows
Figure 2-6 shows the protocol stacks related to the user
plane.
Figure 2-6 User-plane protocol stacks
2.4.3 OAM Data Flows
Figure 2-7 shows OAM data flows:
Uplink OAM data flows from an eNodeB to the operations support
system (OSS)
Downlink OAM data flows from the OSS to an eNodeB
Figure 2-7 OAM data flows
2.4.4 Logical Architecture of an eNodeB
Figure 2-8 shows the functions of an eNodeB.
The main processing and transmission unit (MPT) provides the
following functions:
Control plane functions, including:
RRC functions over the Uu interface
S1AP/X2AP functions over the S1 and X2 interfaces
User plane functions, which are GTP-U functions implemented by
the TRAN module, including S1 and X2 user plane functions and
transmission functions
OAM plane functions
The baseband processing unit (BBP) provides the following
functions:
Control plane functions, including the Uu interface resource
allocation function implemented by the CELLM module
User plane functions, including PDCP, RLC, and MAC functions
OAM plane functions
Figure 2-8 Logical architecture of an eNodeB
Flow Control Mechanism
Figure 2-9 shows the flow control mechanism, which includes the
following procedures:
Measurement
The eNodeB collects statistics about CPU usage and number of
received signaling messages or data volumes, and then determines
whether to perform flow control based on the statistics.
Flow control actions
Open-loop control. For example, the eNodeB performs flow control
based on the number of received signaling messages or data volumes,
including:
Random Access Preamble
RRC Connection Request
Handover Request
RRC Connection Re-establishment Request
Paging
Downlink data volume
Closed-loop control. For example, the eNodeB performs flow
control on received signaling messages or user-plane data based on
CPU usage, including rejecting handover or access requests and
discarding messages based on priorities.
NOTE:
Physical downlink control channel (PDCCH), physical uplink
control channel (PUCCH), and sounding reference signal (SRS)
resources are reserved based on their own specifications to support
the maximum number of UEs that the eNodeB can handle. Therefore,
flow controldoes not need to be designed for these resources. For
details, see Physical Channel Resource Management Feature Parameter
Description.
Monitoring
Intra-eNodeB flow control effect can be monitored through
performance counters.
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
4 trong 13 5/19/2015 6:28 PM
-
Figure 2-9 Flow control mechanism
Both the MPT and BBP are multi-core processing systems. Flow
control is triggered by the load of each CPU core or thread, not
the average load of all CPUs. The purpose is to ensure that flow
control can be performed in time when overload occurs and therefore
ensure system reliability.
Table 2-2 lists the CPU cores required for diffident flow
control actions. The description of each type of CPU is as
follows:
CP Mng CPUs
CPUs for the CELLM module and OAM module
CP Traffic CPUs
CPUs for the RRC, S1AP, and X2AP modules
UP Mng CPUs
CPUs for user-plane management at the cell or eNodeB level
UP Traffic CPUs
CPUs for user-plane protocol stack processing at the UE
level
Table 2-2 CPU cores required for different flow control
actions
Flow Control Action Board Type CPU Type
Control plane: Paging MPT CP Mng CPUs
BBP LBBP: CP Mng CPUs
UBBP: CP Mng CPUs and CP Traffic CPUs
Control plane: Random Access Preamble BBP LBBP: CP Mng CPUs
UBBP: CP Mng CPUs and CP Traffic CPUs
Control plane: RRC Connection Request/RRC Connection
Re-establishment Request/HandoverRequest
MPT CP Traffic CPUs
User plane BBP UP Traffic CPUs
OAM plane MPT CP Mng CPUs
BBP LBBP: CP Mng CPUs
UBBP: CP Mng CPUs and CP Traffic CPUs
For details about the CPU cores and CPU usage monitoring for MPT
and BBP boards, see 6 CPU Core Deployment and CPU Usage
Monitoring.
3 Control-Plane Flow ControlThis chapter describes the
objectives, principles, and monitoring methods of control-plane
flow control.
MME Overloadtriggered Flow Control
3.1.1 Objective
The objective of MME overload-triggered flow control is to
relieve the impact of MME overload caused by a large number of UEs
requesting access to the network. MME overload-triggered flow
control corresponds to load point 1 in Figure 2-1.
3.1.2 Principle
When an MME is overloaded, the MME sends an OVERLOAD START
message to eNodeBs, instructing the eNodeBs to start flow control.
The eNodeBs then limit the number of UEs accessing the network.
When the MME is no longer overloaded, the MME sends an OVERLOAD
STOP message tothe eNodeBs, instructing them to stop flow control.
For details about MME overload-triggered flow control, see 3GPP TS
36.413 Release 9 or Release 10. Figure 3-1 shows the signaling
exchange between an MME and an eNodeB in MME overload-triggered
flow control.
Figure 3-1 MME overloadtriggered flow control
After receiving an OVERLOAD START message, the eNodeB processes
access requests according to 3GPP TS 36.413 Release 9 or Release 10
as follows:
If the Overload Action IE in the Overload Response IE within the
OVERLOAD START message is set to-"reject RRC connection
establishments for non-emergency mobile originated data transfer"
(i.e., reject traffic corresponding to RRC cause "mo-data" and
"delayTolerantAccess" in TS 36.331 [16]), or-"reject RRC connection
establishments for signalling" (i.e., reject traffic corresponding
to RRC cause "mo-data", "mo-signalling" and "delayTolerantAccess"
in TS 36.331 [16]), or-"only permit RRC connection establishments
for emergency sessions and mobile terminated services" (i.e., only
permit traffic corresponding to RRC cause "emergency" and
"mt-Access" in TS 36.331 [16]), or-"only permit RRC connection
establishments for high priority sessions and mobile terminated
services" (i.e., only permit traffic corresponding to RRC cause
"highPriorityAccess" and "mt-Access" in TS 36.331 [16]), or-"reject
only RRC connection establishment for delay tolerant access" (i.e.,
only reject traffic corresponding to RRC cause
"delayTolerantAccess" in TS 36.331 [16]).
The eNodeB shall ensure that only signalling traffic
corresponding to permitted RRC connections is sent to the MME.
3.1.3 Monitoring
Table 3-1 lists the counters related to MME overloadtriggered
flow control. For details, see eNodeB Performance Counter
Reference.
Table 3-1 Counters related to MME overloadtriggered flow
control
Counter Name Counter Description
L.RRC.ConnReq.Msg.disc.FlowCtrl Number of times the RRC
Connection Request message is discarded due to flow control
L.RRC.SetupFail.Rej.FlowCtrl Number of times the eNodeB sends an
RRC Connection Reject message to the UE due to flow control
L.HHO.PrepAttIn.disc.FlowCtrl Number of times the HANDOVER
REQUEST message is discarded over the S1 or X2 interface due to
flow control (without returning apreparation failure message)
L.HHO.Prep.FailIn.FlowCtrl Number of times the target eNodeB
sends a handover preparation failure message for an
intra-duple-mode handover over the S1 or X2interface to the source
eNodeB due to flow control
Flow Control for Random Access Preamble
3.2.1 Objective
Flow control on the Random Access Preamble messages is to
protect an eNodeB from being overloaded due to the random access of
a large number of UEs. A large number of UEs' access requests lead
to high load and even reset of the eNodeB. Flow control on Random
Access Preamblecorresponds to load point 5 shown in Figure 2-1.
3.2.2 Principle
3.2.2.1 Flow Control Points
The flow control point for the contention-based preambles is
located on the CELLM module, which is marked 1 shown in Figure
3-2.
The eNodeB does not perform flow control on non-contention-based
preambles.
Figure 3-2 Flow control points for Random Access Preamble
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
5 trong 13 5/19/2015 6:28 PM
-
3.2.2.2 Flow Control Actions
Flow control on the Random Access Preamble message is
implemented by controlling the number of preambles to be processed.
The CELLM module adaptively adjusts the number of preambles to be
processed based on the CPU usage of the BBP control plane. Table
3-2 provides the detailed flowcontrol actions.
Table 3-2 Flow control actions
CPU Usage (N) Flow Control Action
N 95% The eNodeB discards all preambles to prevent the BBP from
breaking down.
85% N < 95% The eNodeB adjusts the preamble processing
capability to 100 times per second and discards additional
preambles to prevent the BBP from breaking down.
80% N < 85% The eNodeB adjusts the preamble processing
capability to 300 times per second.
N < 80% The eNodeB adjusts the preamble processing capability
to 400 times per second.
3.2.3 Monitoring
Table 3-3 describes the counters related to flow control on
Random Access Preamble messages. For details, see eNodeB
Performance Counter Reference.
Table 3-3 Counters related to flow control on Random Access
Preamble messages
Counter Name Description
L.RA.GrpA.Att Number of times the contention-based preamble in
group A is received
L.RA.GrpA.Resp Number of times a cell sends a Random Access
Response message after receiving a preamble in group A
L.RA.GrpB.Att Number of times the contention preamble in group B
is received
L.RA.GrpB.Resp Number of times a cell sends a Random Access
Response message after receiving a preamble in group B
Initial Access Request Control
3.3.1 Objective
The objective of initial access request control is to relieve
the impact of eNodeB overload caused by a large number of UEs'
access requests. A large number of UEs' access requests lead to
heavy load and even causes the eNodeB to reset. Initial access
request control corresponds to load points5 and 6 in Figure
2-1.
3.3.2 Principle
Initial access requests include RRC connection requests, S1
handover requests, and X2 handover requests. An initial access
request is the start of an access procedure. After an eNodeB
accepts an initial access request, a lot of subsequent processing
is required, causing numerous overheads.Therefore, initial access
request control can reduce the eNodeB load from the beginning.
To ensure high-priority services, initial access request control
processes access requests based on priorities. Access requests with
the following causes are prioritized in descending order:
Emergency1.
High priority2.
Handover3.
CS paging4.
PS paging5.
Mobile-terminated access6.
Mobile-originated signaling7.
Mobile-originated data8.
Delay Tolerant Access9.
When an eNodeB is overloaded, the eNodeB rejects or discards
some initial access requests based on the CPU usage of the main
control board or baseband board as follows:
When the CPU usage of the main control board or baseband board
is equal to or greater than 80% and less than 90% (not
configurable, with 5s as the smooth filtering value), and the CPU
usage is increasing, the eNodeB starts initial access request
control based on the priorities of initialaccess requests. The
eNodeB sends an RRC Connection Reject message to a UE if the eNodeB
rejects the access request.
When the CPU usage of the main control board or baseband board
is equal to or greater than 90%, the eNodeB discards initial access
request messages of the rejected UEs to prevent a breakdown and the
eNodeB does not send RRC Connection Reject messages to these
UEs.
When the CPU usage of the main control board or baseband board
drops to less than 80%, the eNodeB stops initial access request
control based on the priorities of initial access requests.
When the CPU usage of the main control board or baseband board
is equal to or greater than 80% and less than 90%, and the CPU
usage is decreasing, the eNodeB sends an RRC Connection Reject, S1
HANDOVER FAILURE, or X2 HANDOVER FAILURE message to a UE if
theeNodeB rejects the access request. The RRC Connection Reject
message contains the RRC Reject Wait time IE. For details, see 3.5
RRC Reject Wait time.
If the eNodeB load is heavy for a long time, the eNodeB controls
the number of signaling messages received from peer NEs to reduce
the load as follows:
The eNodeB decreases the SCTP buffer threshold to reduce the
number of signaling messages sent from the MME to the eNodeB,
thereby reducing the load in the downlink. For details, see SCTP
Congestion Control Feature Parameter Description.
The eNodeB reduces the frequency of UEs' access to the network
using Access Class (AC) Barring to reduce the load in the uplink.
For details about AC Control, see 3.6 AC Control.
3.3.2.1 Flow Control Points
The RRC Connection Request and Handover Request messages have
the same flow control points (located on the RRC processing module
on the MPT control plane), which are marked 2 shown in Figure 3-3
and 1 shown in Figure 3-4.
As shown in Figure 3-3, there are two flow control points for
the RRC Connection Request message. One is located on the MAC
module of the BBP (marked 1), and the other is located on the RRC
processing module on the MPT control plane (marked 2).
The flow control point for the Handover Request message is
located on the RRC S1AP/X2AP module of the MPT. The S1 Handover
Request and X2 Handover Request messages have the same flow control
priority.
The flow control point for the RRC Connection Re-establishment
Request message is also located on the RRC processing module on the
MPT control plane, which is marked 1 shown in Figure 3-4.
Figure 3-3 Flow control points for the RRC Connection Request
message
Figure 3-4 Flow control point for the Handover Request and RRC
Connection Re-establishment Request messages
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
6 trong 13 5/19/2015 6:28 PM
-
3.3.2.2 Flow Control Actions
Flow Control Actions on the BBP
The BBP performs the open loop flow control.
The MAC module of the BBP measures the number of RRC Connection
Request messages received per second. If the number exceeds a
predefined threshold, the BBP discards the exceeded messages and
increases the L.RRC.ConnReq.Msg.disc.FlowCtrl counter value.
The purpose of limiting the number of RRC Connection Request
messages on the BBP is to prevent excessive messages from being
transferred to the MPT and consuming too much load processing
capability of the eNodeB.
Table 3-4 lists the current message processing capabilities of
different BBPs.
Table 3-4 Message processing capabilities of different BBPs
Board Type Maximum Number of RRC Connection Request Messages
That Can Be Processed
LBBPc 80 per second
LBBPd 150 per second
UBBPd3/d4 360 per second
UBBPd5/d6 410 per second
Flow Control Actions on the MPT
The MPT performs flow control for RRC Connection Request and
Handover Request messages. Table 3-5 lists the current message
processing capabilities of different MPTs.
Table 3-5 Message processing capabilities of different MPTs
Board Type Maximum Number of RRC Connection Request and Handover
Request Messages That Can Be Processed
LMPT 130 per second
UMPT 300 per second
The MPT also performs flow control for RRC Connection
Re-establishment Request messages. Table 3-6 lists the current
message processing capabilities of different MPTs.
Table 3-6 Message processing capabilities of different MPTs
Board Type Maximum Number of RRC Connection Re-establishment
Request Messages That Can Be Processed
LMPT 130 per second
UMPT 225 per second
The MPT performs flow control for RRC Connection Request and
Handover Request messages based on CPU usage of the control plane
and signaling message priorities. The following lists the message
priorities in descending order:
RRC Connection Request for emergency services
RRC Connection Request for high-priority services
Handover Request
RRC Connection Request for mobile-terminated (MT) access
RRC Connection Request for mobile-originated (MO) signaling
RRC Connection Request MO data
RRC Connection Request for delay tolerant access
Table 3-7 lists the flow control actions on the MPT based on CPU
usage.
Table 3-7 Flow control actions on the MPT based on CPU usage
CPU Usage (N) Flow Control Action
N 90% The MPT discards RRC Connection Request and Handover
Request messages.
80% N < 90% The MPT starts flow control and responds with an
RRC Connection Reject or Handover Failure message for rejected UEs
based on the priorities.
N < 80% The MPT does not perform flow control on RRC
Connection Request or Handover Request.
3.3.3 Monitoring
For details, see 3.1 MME Overloadtriggered Flow Control.
Paging Message Flow Control
3.4.1 Objective
The objective of paging message flow control is to relieve the
impact of eNodeB overload caused by a large number of paging
messages. A large number of paging messages lead to heavy load and
even causes the eNodeB to reset. Paging message flow control
corresponds to load point 3 inFigure 2-1.
3.4.2 Principle
A paging message is the start of a procedure. A large number of
successfully processed paging messages lead to a large number of
UEs' access to the network and excessive signaling overheads on the
eNodeB. Therefore, paging message flow control reduces the eNodeB
load from thebeginning.
To guarantee the user experience for high-priority services,
paging message flow control takes different actions to process
paging messages with different causes. 3GPP Release 8, Release 9,
and Release 10 define paging causes as follows:
According to 3GPP Release 8 and Release 9, the CN domain IE is
used to distinguish CS and PS services. The following table
describes the CN domain IE.
IE/Group Name Presence Range IE type and reference Semantics
description
CN Domain M - ENUMERATED(PS, CS) -
According to 3GPP Release 10, the Paging Priority IE is used to
distinguish between CS and PS services. According to 3GPP Release
10, an eNodeB determines the paging priorities of only CS fallback
(CSFB) and IP multimedia subsystem enhanced multimedia priority
service(IMS-eMPS), which have high paging priorities to ensure user
experience for CS services. Paging priorities of PS services need
to be planned by operators and configured on the core network. The
following table describes the Paging Priority IE.
IE/Group Name Presence Range IE type and reference Semantics
description
Paging Priority M - ENUMERATED (PrioLevel1, PrioLevel2,
PrioLevel3, PrioLevel4, PrioLevel5,PrioLevel6, PrioLevel7,
PrioLevel8, )
Lower value code point indicates higher priority
3.4.2.1 Flow Control Points
As shown in Figure 3-5, there are two flow control points for
paging messages. One is located on the S1AP module of the MPT
(marked 1), and the other is located on the CELLM module of the BBP
(marked 2).
Figure 3-5 Flow control points for paging messages
3.4.2.2 Flow Control Actions
Table 3-8 and Table 3-9 describe the flow control actions for
paging messages on the MPT and BBP, respectively.
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
7 trong 13 5/19/2015 6:28 PM
-
Table 3-8 Flow control actions for paging messages on the
MPT
CPU Usage Flow Control Action
85% The MPT starts flow control and gradually decreases its
paging message processing capability to the minimum value (150
times per second). If the number of paging messages to be processed
exceeds theprocessing capability, the MPT discards paging messages
based on their priorities and discards paging messages for PS
services first.
< 85% The MPT does not perform flow control on paging
messages and processes paging messages using the maximum processing
capability.
Note:
Maximum number of paging messages processed by the LMPT per
second: 1800
Maximum number of paging messages processed by the UMPT per
second: 2400
Table 3-9 Flow control actions for paging messages on the
BBP
CPU Usage Flow Control Action
90% The BBP discards paging messages.
< 90% The BBP does not discard paging messages.
3.4.3 Monitoring
Table 3-10 describes the counters related to paging message flow
control. For details, see eNodeB Performance Counter Reference.
Table 3-10 Counters related to paging message flow control
Counter Name Description
L.Paging.S1.Rx Number of received paging messages over the S1
interface in a cell
L.Paging.UU.Att Number of UEs contained in paging messages
transmitted over the Uu interface in a cell
L.Paging.UU.Succ Number of successful paging responses from the
UE in a cell
L.Paging.Dis.Num Number of discarded paging messages from the
MME to UEs
RRC Reject Wait time
3.5.1 Objective
The objective of sending the RRC Reject Wait time IE is to
prevent signaling bursts caused by UEs' frequent access to the
network. When an eNodeB rejects a UE's access request, the eNodeB
includes the RRC Reject Wait time IE in the RRC Connection Reject
message to inform the UE of thewait time before the UE initiates
another access request. The function of sending the RRC Reject Wait
time IE corresponds to load point 5 in Figure 2-1.
3.5.2 Principle
When an eNodeB is overloaded, the eNodeB includes the RRC Reject
Wait time IE in the RRC Connection Reject message to inform the UE
of the wait time before the UE initiates another access request.
The UE processes the RRC Reject Wait time IE according to 3GPP TS
36.331. Figure 3-6shows the signaling exchange between the eNodeB
and UE.
Figure 3-6 Signaling exchange between the eNodeB and UE
When the CPU usage of the main control board or baseband board
is equal to or greater than 80% (not configurable, with 5s as the
smooth filtering value), the eNodeB starts initial access request
control and includes the RRC Reject Wait time IE in the RRC
Connection Reject message.
The value of the RRC Reject Wait time IE is configurable. You
can run the MOD RRCCONNSTATETIMER command to set the
RrcConnStateTimer.T302 parameter. For details, see eNodeB MML
Command Reference.
3.5.3 Monitoring
For details, see 3.1 MME Overloadtriggered Flow Control.
AC Control
3.6.1 Objective
As defined in 3GPP TS 36.331, access class (AC) control is a
method used to control the UE access to the network. The objective
is to relieve the impact of eNodeB overload caused by a large
number of UEs requesting access to the network. AC control
corresponds to load point 5 in Figure 2-1.
3.6.2 Principle
With AC control, the eNodeB broadcasts AC control parameters to
UEs in a cell through System Information Block 2 (SIB2) messages.
UEs then determine whether they can access the cell based on the
received AC control parameters. Figure 3-7 shows the signaling
exchange between the eNodeBand UE for AC control.
Figure 3-7 Signaling exchange between the eNodeB and UE for AC
control
AC control can be classified in to static AC control and
intelligent AC control:
Static AC control
With static AC control, after AC control parameters are
configured on the OSS by an operator, the eNodeB broadcasts
parameters to UEs through system information (SI) update, without
considering the current network condition. The eNodeB implements
static AC control on thefollowing objects:
Emergency
Mobile-originated data
Mobile-originated signaling
Multimedia telephony voice
Multimedia telephony video
CSFB
You can adjust the frequencies of UEs' access to the network by
running the MOD CELLACBAR command on the eNodeB side and
configuring the parameters CellAcBar.AcBarringFactorForCall,
CellAcBar.AcBarringFactorForSig,
CellAcBar.AcBarFactorForMVoice,CellAcBar.AcBarFactorForMVideo, and
CellAcBar.AcBarFactorForCsfb.
Intelligent AC control
With intelligent AC control, the eNodeB automatically triggers
or cancels AC control by adjusting the mobile-originated signaling
and mobile-originated data access frequencies based on the load of
a cell. For details, see Access Class Control Feature Parameter
Description.
3.6.3 Monitoring
For details, see 3.1 MME Overloadtriggered Flow Control.
Uplink Synchronized UE Number Control
3.7.1 Objective
The objective of uplink synchronized UE number control is to
relieve the impact of eNodeB overload caused by too many uplink
synchronized UEs in a short time in special scenarios such as
festivals. Uplink synchronized UE number control corresponds to
load point 5 in Figure 2-1.
3.7.2 Principle
When the number of UEs in a cell is too large, the eNodeB
selects some uplink synchronized UEs that have not transmitted or
received data for a period and changes the status of these UEs to
out of synchronization. This releases physical uplink control
channel (PUCCH) and sounding referencesignal (SRS) resources for
UEs attempts to access or be handed over to the cell.
You can run the MOD ENODEBFLOWCTRLPARA command to set the
eNodeBFlowCtrlPara.AdaptUnsyncUserNumThd and
eNodeBFlowCtrlPara.AdaptUnsyncTimerLen parameters. The
eNodeBFlowCtrlPara.AdaptUnsyncUserNumThd parameter indicates the
number of uplink synchronizedUEs, and the
eNodeBFlowCtrlPara.AdaptUnsyncTimerLen parameter indicates the
duration in which no data is transmitted. For details, see eNodeB
MML Command Reference.
3.7.3 Monitoring
Table 3-11 lists the counters related to uplink synchronized UE
number control. For details, see eNodeB Performance Counter
Reference.
Table 3-11 Counters related to uplink synchronized UE number
control
Counter Name Description
L.Traffic.User.Avg Average number of users in a cell
L.Traffic.User.Max Maximum number of users in a cell
L.Traffic.User.Ulsync.Avg Average number of UL synchronized
users in a cell
4 User-Plane Flow ControlThis chapter describes the objectives,
principles, and monitoring methods of user-plane flow control.
Downlink User-Plane Flow Control
4.1.1 Objective
The objective of downlink user-plane flow control is to ensure
eNodeB reliability when the amount of downlink user-plane data
exceeds the eNodeB's processing capability. Downlink user-plane
flow control corresponds to load points 4 and 6 in Figure 2-1.
4.1.2 Principle
Downlink user-plane flow control reduces the number of downlink
packets, including packets carried by guaranteed bit rate (GBR) and
non-GBR bearers. An eNodeB preferentially performs flow control on
packets carried by non-GBR bearers. If congestion persists after
the eNodeB has performedflow control on all packets carried by
non-GBR bearers, the eNodeB performs flow control on packets
carried by GBR bearers until the eNodeB is no longer congested.
4.1.2.1 Flow Control Points
Figure 4-1 shows the flow control points for downlink user
plane, which are marked 1 and 2.
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
8 trong 13 5/19/2015 6:28 PM
-
Generally, the data processing capability of the TRAN module is
not a bottleneck that affects traffic specifications. For details
about the flow control mechanism of the TRAN module, see Transport
Resource Management Feature Parameter Description.
The downlink protocol stack processing module (PDCP/RLC/MAC)
becomes overloaded when the data volume on the user plane is
excessively large. Therefore, downlink user plane flow control is
performed based on the CPU usage of the user plane.
Figure 4-1 Flow control points for downlink user plane
The downlink open loop flow control is performed on the TRAN
module of the MPT to ensure the total throughput for a BBP is below
the 1.2 times the BBP peak throughput specifications. Additional
data flows will be discarded. During closed-loop flow control, the
BBP sends a flow control indicatorto the MPT based on its load, and
the TRAN module of the MPT dynamically adjusts the data rate.
When the PDCP/MAC/RLC stack processing module is overloaded, the
module sends a back pressure to the TRAN module to request a
decrease in the data volume. On the contrary, the module sends a
back pressure cancelation to TRAN module. The measurements are
performed on a persecond basis. The following table lists the
related CPU usage thresholds.
Board Type UP Overload Threshold UP Idle Threshold
LBBPc 90% 85%
LBBPd1/d2 80% 74%
LBBPd3 74% 68%
After receiving a back pressure indicator, the TRAN module
decreases the downlink data volume to reduce the data load being
processed by the BBP. After receiving a back pressure cancelation
indicator, the TRAN module gradually restores the downlink data
volume. The flow control algorithmimplements optimal matching
between the data volume and the processing capability of the
BBP.
4.1.2.2 Flow Control Actions
The TRAN module of the MPT performs flow control based on the
back pressure indicator.
In the congestion state, the MPT decreases the allowed data
volume by 10% based on the currently allowed volume at an interval
of 10s. When receiving the back pressure indicator, the MPT does
not discard data. Instead, it buffers the data that cannot be sent
in time.
In the normal state, the MPT increases the allowed data volume
by a semi-persistent increment based on the currently allowed
volume. This increment increases twice at an interval of 10s.
The TRAN module decreases or increases the packet transmission
rates of both GBR and non-GBR bearers. During the packet
transmission rate decrease, the TRAN module preferentially
decreases the packet transmission rate of non-GBR bearers. If the
CPU usage of the BBP cannot bereleased even when no packets on all
non-GBR bearers are allowed to be delivered, the TRAN module
decreases the packets on GBR bearers.
4.1.3 Monitoring
Table 4-1 describes the counters related to downlink user-plane
flow control. For details, see eNodeB Performance Counter
Reference.
Table 4-1 Counters related to downlink user-plane flow
control
Counter Name Description
L.PDCP.Tx.Disc.Trf.SDU.QCI.1 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 1 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.2 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 2 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.3 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 3 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.4 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 4 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.5 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 5 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.6 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 6 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.7 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 7 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.8 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 8 in a
cell
L.PDCP.Tx.Disc.Trf.SDU.QCI.9 Number of downlink traffic SDUs
discarded by the PDCP layer for services with the QCI of 9 in a
cell
L.Traffic.Board.UPlane.CPULoad.MAX
L.Traffic.Board.UPlane.CPULoad.AVG
The usages of data processing CPUs are sampled every second and
the maximum one is taken as the sampling result. The average of
every five consecutive samplingresults is used as the filtered CPU
usage.
The average of these filtered CPU usages is indicated by
L.Traffic.Board.UPlane.CPULoad.AVG, and the maximum of these
filleted CPU usages is indicated
byL.Traffic.Board.UPlane.CPULoad.MAX.
Uplink User-Plane Flow Control
For details about uplink user-plane flow control, seeTransport
Resource Management Feature Parameter Description. Uplink
user-plane flow control corresponds to load point 2 in Figure
2-1.
4.2.1 Objective
The objective of uplink user-plane flow control is to ensure the
eNodeB reliability when the amount of uplink user-plane data
exceeds the eNodeB processing capability.
4.2.2 Principle
4.2.2.1 Flow Control Points
Uplink user-plane flow control on the BBP is performed on the
PDCP/RLC/MAC module based on the usage of UP Traffic CPUs, which is
marked 1 in Figure 4-2. Uplink user-plane flow control on the MPT
is performed on the TRAN module, which is marked 2 in Figure
4-2.
Figure 4-2 Flow control points for uplink user plane
Uplink user-plane flow control is performed based on the CPU
usage of the protocol stack processing module and the buffered time
of packets over the S1/X2 interface. The following table lists the
CPU usage thresholds for the uplink user-plane flow control.
Board Type UP Overload Threshold UP Idle Threshold
LBBPc 95% 95%
LBBPd1/d2 82% 82%
LBBPd3 75% 75%
When the CPU usage of the user plane exceeds the overload
threshold, the MAC scheduler reduces the total amount of uplink
data that is allowed to be transmitted to the BBP over the Uu
interface. In this way, UEs send less uplink data to the eNodeB,
thereby reducing the CPU usage on the userplane.
When the CPU usage of the user plane is lower than the idle
threshold, the MAC scheduler increases the total amount of uplink
data that is allowed to be transmitted to the BBP over the Uu
interface.
When the buffered time of packets over the S1/X2 interface meets
congestion or congestion cancelation conditions, the MPT sends the
related congestion or congestion cancelation indicator to the BBP.
The BBP then decreases or increases the maximum uplink data volume
in UE level to accomplishthe flow control.
4.2.2.2 Flow Control Actions
Flow Control for the Overload on the PDCP/RLC/MAC Module of the
BBP
As described in 4.2.2.1 Flow Control Points, PDCP/RLC/MAC flow
control is based on CPU usage. The BBP monitors its overload state
at an interval of 100 ms. If the BBP is overloaded, the BBP sends
an indicator to instruct the uplink scheduler to decrease the
maximum transport block size (TBS)for the BBP by 1%. If the BBP is
not overloaded, it sends an indicator to the uplink scheduler to
increase the maximum TBS for the BBP by 1%.
Flow Control for S1/X2 Interface Congestion
The back pressure algorithm is controlled by RSCGRPALG.TCSW. The
algorithm takes effect only when the switch is turned on.
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
9 trong 13 5/19/2015 6:28 PM
-
If the buffered time of packets over the S1/X2 interface is
greater than the value of the congestion threshold parameter
RSCGRPALG.CTTH, the MPT is overloaded, and sends a back pressure
indicator to the BBP.
If the buffered time of packets over the S1/X2 interface is less
than the value of RSCGRPALG.CCTTH, the MPT enters the idle state
and sends a back pressure cancelation indicator to the BBP.
The uplink flow control algorithm for S1/X2 interface congestion
is controlled by the UlUuFlowCtrlSwitch option of the
ENODEBALGOSWITCH.TrmSwitch parameter. Uplink user-plane flow
control for the MAC module of the BBP takes effect only when this
option is turn on. When the uplinktransmission is congested, the
MPT sends an indicator to the uplink scheduler to limit the uplink
UE rate to resolve congestion. For details, see Transport Resource
Management Feature Parameter Description.
In the congestion state, the BBP decreases the target TBS of UEs
by 10%, and the uplink scheduler of the MAC performs related
scheduling of the low data volume. At the same time, the BBP stops
transmitting the data for all non-real-time services.
In the normal state, the BBP increases the target TBS of the
UEs, and the uplink scheduler of the MAC performs related
scheduling of the high data volume. At the same time, the BBP
restarts the transmission of the data for all non-real-time
services.
4.2.3 Monitoring
Whether uplink user-plane flow control is performed cannot be
observed based on throughput-related counters because the
throughput on the user plane is changed according to various
factors. However, the counter of CPU usage of the BBP can be used
to evaluate whether uplink user-plane flowcontrol is performed. For
details, see eNodeB Performance Counter Reference.
Table 4-2 Counters related to uplink user-plane flow control
Counter Name Description
L.Traffic.Board.UPlane.CPULoad.MAX
L.Traffic.Board.UPlane.CPULoad.AVG
The usages of data processing CPUs are sampled every second and
the maximum one is taken as the sampling result. The average of
every five consecutive samplingresults is used as the filtered CPU
usage.
The average of these filtered CPU usages is indicated by
L.Traffic.Board.UPlane.CPULoad.AVG, and the maximum of these
filleted CPU usages is indicated
byL.Traffic.Board.UPlane.CPULoad.MAX.
5 OAM Flow ControlThis chapter describes the objectives,
principles, and monitoring methods of OAM flow control.
OAM includes software upgrades, file downloads and uploads,
logs, MML commands, alarms, performance counters, and performance
monitoring data. OAM requires a large amount of data transmission
and CPU usage, which negatively impacts the control-plane and
user-plane data transmissionduring busy hours. The priorities of
most OAM tasks are lower than those of control-plane and user-plane
data transmission. OAM flow control releases CPU resources for
control-plane and user-plane data transmission.
OAM Flow Control
5.1.1 Objective
The objective of OAM flow control is to prevent OAM tasks from
occupying too many resources and negatively affecting control-plane
and user-plane data transmission. OAM flow control corresponds to
load points 7 and 8 in Figure 2-1.
5.1.2 Principle
When services are busy, an eNodeB performs flow control to tasks
of log recording, software management, and performance monitoring
based on service priorities to ensure that the key performance
indicators (KPIs) are maintained.
The priorities of control-plane, user-plane, and OAM services
are described as follows in descending order:
High-priority OAM tasks, including tasks related to MML
commands, alarms, performance counters, and high-priority
logs1.
Control-plane and user-plane data transmission2.
Medium-priority OAM tasks, including tasks related to
medium-priority logs3.
Low-priority OAM tasks, including tasks related to low-priority
logs, performance monitoring, and software download4.
NOTE:
High-, medium-, and low-priority logs are defined as
follows:
High-priority log: including security, running, user operation,
and statistics logs
Medium-priority log: call history record (CHR)
Low-priority log: debug log
When an eNodeB is overloaded, the eNodeB performs OAM flow
control as follows:
The eNodeB does not perform flow control to high-priority OAM
tasks.
When the CPU usage of the main control board or baseband board
is equal to or greater than 70% (not configurable, with 5s as the
smooth filtering value), the eNodeB rejects new performance
monitoring tasks. The eNodeB also stops ongoing performance
monitoring tasks.
The eNodeB stops log recording based on log priorities. When the
CPU usage of the main control board or baseband board is equal to
or greater than 70%, the eNodeB stops low-priority log recording.
When the CPU usage of the main control board or baseband board is
equal to orgreater than 80%, the eNodeB stops medium-priority log
recording.
When the CPU usage of the main control board or baseband board
is less than 70%, the eNodeB accepts new performance monitoring
tasks.
When the CPU usage of the main control board or baseband board
is less than 80%, the eNodeB restarts medium-priority log
recording. Similarly, when the CPU usage of the main control board
or baseband board is less than 70%, the eNodeB restarts
low-priority log recording.
5.1.3 Monitoring
OAM flow control interrupts performance monitoring tasks and a
message indicating that tasks are stopped because of flow control
is displayed on the U2000.
6 CPU Core Deployment and CPU Usage MonitoringThis chapter
describes the CPU core deployment and CPU usage monitoring for MPT
and BBP boards.
LMPT
6.1.1 CPU Core Deployment
LMPT CPUs comprise of the following types:
CP Mng CPUs
The load of CP Mng CPUs is mainly caused by OAM, paging message
processing, and SCTP message processing. Therefore, flow control on
OAM and paging messages is performed based on the usage of CP Mng
CPUs. Generally, usage of these CPUs is not related to the number
ofUEs and traffic volumes. Therefore, no counters are designed for
CP Mng CPUs usage reporting. However, the usage of CP Mng CPUs can
be monitored on the U2000 or LMT.
CP Traffic CPUs
The loads of CP Traffic CPU are mainly caused by the signaling
messages over the S1, Uu, and X2 interfaces. Flow control on RRC
messages and handover messages is performed based on the usage of
CP Traffic CPUs. Some counters are designed for them. For details
about relatedcounters, see 6.1.2.1 Counters Related to the CPU
Usage on the LMPT.
6.1.2 CPU Usage Monitoring
6.1.2.1 Counters Related to the CPU Usage on the LMPT
Counter Name Description
VS.BBUBoard.CPULoad.Max The usages of the CP Traffic CPUs are
sampled every second and the maximum one is taken as the sampling
result. The average of every five consecutive sampling results is
used as the filteredCPU usage.
The average of these filtered CPU usages is indicated by
VS.BBUBoard.CPULoad.Mean, and the maximum of these filleted CPU
usages is indicated by
VS.BBUBoard.CPULoad.Max.VS.BBUBoard.CPULoad.Mean
6.1.2.2 Alarm Related to CPU Overload on the LMPT
If the usage of any CPU among CP Traffic CPUs and CP Mng CPUs is
greater than 90% in 30 consecutive seconds, the eNodeB reports
ALM-26202 Board Overload.
6.1.2.3 CPU Usage Monitoring on the U2000 or LMT
The usage of CP Mng CPUs is reported. The reporting period can
be configured to a value within the range of 5s to 30s on the U2000
or LMT.
UMPT
6.2.1 CPU Core Deployment
UMPT CPUs comprise of the following types:
CP Mng CPUs
CP Traffic CPUs
6.2.2 CPU Usage Monitoring
6.2.2.1 Counters Related to the CPU Usage on the UMPT
Counter Name Description
VS.BBUBoard.CPULoad.Max The usages of CP Traffic CPUs and CP Mng
CPUs are sampled every second and the maximum one is taken as the
sampling result. The average of every five consecutive
samplingresults is used as the filtered CPU usage.
The average of these filtered CPU usages is indicated by
VS.BBUBoard.CPULoad.Mean, and the maximum of these filleted CPU
usages is indicated by
VS.BBUBoard.CPULoad.Max.VS.BBUBoard.CPULoad.Mean
6.2.2.2 Alarm Related to CPU Overload on the UMPT
If the usage of any CPU among CP Traffic CPUs and CP Mng CPUs is
greater than 90% in 30 consecutive seconds, the eNodeB reports
ALM-26202 Board Overload.
6.2.2.3 CPU Usage Monitoring on the U2000 or LMT
The Average usage of CP Mng CPUs and CP Traffic CPUs is
reported. The reporting period can be configured to a value within
the range of 5s to 30s on the U2000 or LMT.
LBBP
6.3.1 CPU Core Deployment
LBBP CPUs comprise of the following types:
CP Mng CPUs
UP Mng CPUs
These CPUs serve as resources pools to be scheduled between
cells.
UP Traffic CPUs
These CPUs mainly serve as the resource pool for processing
PDCP, RLC, and MAC. The CPU usage increases with the number of
users and the load of services.
6.3.2 CPU Usage Monitoring
6.3.2.1 CPU Usage Counters
Counter Name Description
VS.BBUBoard.CPULoad.Max The usages of the CP Mng CPUs are
sampled every second and the maximum one is taken as the sampling
result. The average of every five consecutive sampling results is
used as the filtered
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
10 trong 13 5/19/2015 6:28 PM
-
Counter Name Description
CPU usage.
The average of these filtered CPU usages is indicated by
VS.BBUBoard.CPULoad.Mean, and the maximum of these filleted CPU
usages is indicated by
VS.BBUBoard.CPULoad.Max.VS.BBUBoard.CPULoad.Mean
L.Traffic.Board.UPlane.CPULoad.MAX The usages of the UP Traffic
CPUs are sampled every second and the maximum one is taken as the
sampling result. The average of every five consecutive sampling
results is used as the filteredCPU usage.
The average of these filtered CPU usages is indicated by
L.Traffic.Board.UPlane.CPULoad.AVG, and the maximum of these
filleted CPU usages is indicated by
L.Traffic.Board.UPlane.CPULoad.MAX.L.Traffic.Board.UPlane.CPULoad.AVG
6.3.2.2 Alarm Related to CPU Overload on the LBBP
If the usage of any CPU among CP Traffic CPUs and CP Mng CPUs is
greater than 90% in 30 consecutive seconds, the eNodeB reports
ALM-26202 Board Overload.
6.3.2.3 CPU Usage Monitoring on the U2000 or LMT
The average CP Mng CPUs usage is taken as the result of CPU
Usage Monitoring. The CPU usage reporting period can be set to 5s
to 30s on the U2000 or LMT.
UBBP
6.4.1 CPU Core Deployment
UBBPd CPUs comprise of the following types:
CP Mng CPUs
CP Traffic CPUs
UP Mng CPUs
UP Traffic CPUs
6.4.2 CPU Usage Monitoring
6.4.2.1 CPU Usage Counters
Counter Name Description
VS.BBUBoard.CPULoad.Max The usages of CP Traffic CPUs and CP Mng
CPUs are sampled every second and the maximum one is taken as the
sampling result. The average of every five consecutive sampling
results isused as the filtered CPU usage.
The average of these filtered CPU usages is indicated by
VS.BBUBoard.CPULoad.Mean, and the maximum of these filleted CPU
usages is indicated by
VS.BBUBoard.CPULoad.Max.VS.BBUBoard.CPULoad.Mean
L.Traffic.Board.UPlane.CPULoad.MAX The usages of the UP Traffic
CPUs are sampled every second and the maximum one is taken as the
sampling result. The average of every five consecutive sampling
results is used as the filteredCPU usage.
The average of these filtered CPU usages is indicated by
L.Traffic.Board.UPlane.CPULoad.AVG, and the maximum of these
filleted CPU usages is indicated by
L.Traffic.Board.UPlane.CPULoad.MAX.L.Traffic.Board.UPlane.CPULoad.AVG
6.4.2.2 Alarm Related to CPU Overload on the UBBP
When the average usage of CP Mng CPUs and CP Traffic CPUs
exceeds 90% for 30 consecutive seconds, the eNodeB reports
ALM-26202 Board Overload.
6.4.2.3 CPU Usage Monitoring on the U2000 or LMT
The average usage of CP Traffic CPUs and CP Mng CPUs is taken as
the result of CPU Usage Monitoring. The CPU usage reporting period
can be set to 5s to 30s on the U2000 or LMT.
7 ParametersTable 7-1 Parameter description
MO Parameter ID MML Command Feature ID Feature Name
Description
RrcConnStateTimer T302 MODRRCCONNSTATETIMER
LST RRCCONNSTATETIMER
LBFD-002007 / TDLBFD-002007
RRC ConnectionManagement
Meaning: Indicates the length of timer T302. T302 specifies the
time during which a UE whose RRC connection request is rejected
hasto wait before the UE can initiate a request again. This timer
is started when the UE receives an RRCConnectionReject message
andstopped when the UE enters the RRC_CONNECTED mode or performs
cell reselection.
GUI Value Range: 1~16
Unit: s
Actual Value Range: 1~16
Default Value: 4
CellAcBar AcBarringFactorForCall MOD CELLACBARLST CELLACBAR
LBFD-002009 / TDLBFD-002009
Broadcast of systeminformation
Meaning: Indicates the access probability factor for
mobile-originated calls. A mobile-originated call is granted access
if the randomnumber generated by the UE is less than this access
probability factor; otherwise, the access request is rejected.
According to 3GPPTS 36.331, if any of the parameters
AC11BarforCall, AC12BarforCall, AC13BarforCall, AC14BarforCall, and
AC15BarforCall is set toBOOLEAN_TRUE, the eNodeB sends UEs P00 as
the access probability factor for mobile-originated calls in the
system informationblock type 2 (SIB2), regardless of the actual
setting of the AcBarringFactorForCall parameter.
GUI Value Range: P00(0%), P05(5%), P10(10%), P15(15%), P20(20%),
P25(25%), P30(30%), P40(40%), P50(50%), P60(60%),P70(70%),
P75(75%), P80(80%), P85(85%), P90(90%), P95(95%)
Unit: %
Actual Value Range: P00, P05, P10, P15, P20, P25, P30, P40, P50,
P60, P70, P75, P80, P85, P90, P95
Default Value: P95(95%)
CellAcBar AcBarringFactorForSig MOD CELLACBARLST CELLACBAR
LBFD-002009 / TDLBFD-002009
Broadcast of systeminformation
Meaning: Indicates the access probability factor for signaling.
Signaling from a UE is granted access if the random number
generatedby the UE is less than this access probability factor;
otherwise, the access request is rejected. According to 3GPP TS
36.331, if any ofthe parameters AC11BarForSig, AC12BarForSig,
AC13BarForSig, AC14BarForSig, and AC15BarForSig is set to
BOOLEAN_TRUE,the eNodeB sends UEs P00 as the access probability
factor for signaling in the system information block type 2 (SIB2),
regardless ofthe actual setting of the AcBarringFactorForSig
parameter.
GUI Value Range: P00(0%), P05(5%), P10(10%), P15(15%), P20(20%),
P25(25%), P30(30%), P40(40%), P50(50%), P60(60%),P70(70%),
P75(75%), P80(80%), P85(85%), P90(90%), P95(95%)
Unit: %
Actual Value Range: P00, P05, P10, P15, P20, P25, P30, P40, P50,
P60, P70, P75, P80, P85, P90, P95
Default Value: P95(95%)
CellAcBar AcBarFactorForMVoice MOD CELLACBARLST CELLACBAR
LBFD-002009 / TDLBFD-002009
Broadcast of systeminformation
Meaning: Indicates the access probability factor for multimedia
telephony (MMTEL) voice services. An MMTEL voice service isgranted
access if the random number generated by the UE is less than this
access probability factor; otherwise, the access request
isbarred.
GUI Value Range: P00(0%), P05(5%), P10(10%), P15(15%), P20(20%),
P25(25%), P30(30%), P40(40%), P50(50%), P60(60%),P70(70%),
P75(75%), P80(80%), P85(85%), P90(90%), P95(95%)
Unit: %
Actual Value Range: P00, P05, P10, P15, P20, P25, P30, P40, P50,
P60, P70, P75, P80, P85, P90, P95
Default Value: P95(95%)
CellAcBar AcBarFactorForMVideo MOD CELLACBARLST CELLACBAR
LBFD-002009 / TDLBFD-002009
Broadcast of systeminformation
Meaning: Indicates the access probability factor for multimedia
telephony (MMTEL) video services. An MMTEL video service isgranted
access if the random number generated by the UE is less than this
access probability factor; otherwise, the access request
isbarred.
GUI Value Range: P00(0%), P05(5%), P10(10%), P15(15%), P20(20%),
P25(25%), P30(30%), P40(40%), P50(50%), P60(60%),P70(70%),
P75(75%), P80(80%), P85(85%), P90(90%), P95(95%)
Unit: %
Actual Value Range: P00, P05, P10, P15, P20, P25, P30, P40, P50,
P60, P70, P75, P80, P85, P90, P95
Default Value: P95(95%)
CellAcBar AcBarFactorForCsfb MOD CELLACBARLST CELLACBAR
LBFD-002009 / TDLBFD-002009
Broadcast of systeminformation
Meaning: Indicates the access probability factor for CS fallback
(CSFB) services. If the random number generated by a UE is less
thanthe parameter value, the eNodeB accepts the CSFB service access
request; otherwise, the eNodeB rejects the access request.
GUI Value Range: P00(0%), P05(5%), P10(10%), P15(15%), P20(20%),
P25(25%), P30(30%), P40(40%), P50(50%), P60(60%),P70(70%),
P75(75%), P80(80%), P85(85%), P90(90%), P95(95%)
Unit: %
Actual Value Range: P00, P05, P10, P15, P20, P25, P30, P40, P50,
P60, P70, P75, P80, P85, P90, P95
Default Value: P95(95%)
eNodeBFlowCtrlPara AdaptUnsyncUserNumThd
MODENODEBFLOWCTRLPARA
LSTENODEBFLOWCTRLPARA
None None Meaning: Indicates the threshold of the ratio of
uplink-synchronized UEs in a cell or a baseband processing unit. If
the ratio of uplink-synchronized UEs in a cell or a baseband
processing unit exceeds this threshold, the eNodeB adaptively
enables some UEs that do nottransmit or receive data for a period
of time greater than the AdaptUnsyncTimerLen parameter value to
enter the asynchronizationstate. As a result, PUCCH resources can
be released and used by other UEs.
GUI Value Range: 50~100
Unit: %
Actual Value Range: 50~100
Default Value: 100
eNodeBFlowCtrlPara AdaptUnsyncTimerLen MODENODEBFLOWCTRLPARA
LSTENODEBFLOWCTRLPARA
None None Meaning: Indicates whether an adaptive
asynchronization UE is selected when the adaptive asynchronization
function is enabled. If aUE does not transmit or receive data for a
period of time that is longer than the AdaptUnsyncTimerLen
parameter value, the UE is acandidate adaptive asynchronization UE.
When this parameter value is larger than or equal to the
UeInactiveTimer parameter value, theadaptive asynchronization
function does not take effect.
GUI Value Range: 1~20
Unit: s
Actual Value Range: 1~20
Default Value: 4
8 CountersTable 8-1 Counter description
Counter ID Counter Name Counter Description Feature ID Feature
Name
1526726833 L.PDCP.Tx.Disc.Trf.SDU.QCI.1 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 1 in a
cell
Multi-mode: None
GSM: None
Radio Bearer Management
Radio Bearer Management
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
11 trong 13 5/19/2015 6:28 PM
-
Counter ID Counter Name Counter Description Feature ID Feature
Name
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Basic Scheduling
Basic Scheduling
1526726839 L.PDCP.Tx.Disc.Trf.SDU.QCI.2 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 2 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726845 L.PDCP.Tx.Disc.Trf.SDU.QCI.3 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 3 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726851 L.PDCP.Tx.Disc.Trf.SDU.QCI.4 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 4 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726857 L.PDCP.Tx.Disc.Trf.SDU.QCI.5 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 5 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726863 L.PDCP.Tx.Disc.Trf.SDU.QCI.6 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 6 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726869 L.PDCP.Tx.Disc.Trf.SDU.QCI.7 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 7 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726875 L.PDCP.Tx.Disc.Trf.SDU.QCI.8 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 8 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726881 L.PDCP.Tx.Disc.Trf.SDU.QCI.9 Number of downlink PDCP
SDUs discarded forservices carried on DRBs with a QCI of 9 in a
cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002008
TDLBFD-002008
LBFD-002025
TDLBFD-002025
Radio Bearer Management
Radio Bearer Management
Basic Scheduling
Basic Scheduling
1526726884 L.Paging.S1.Rx Number of received paging messages
over the S1interface in a cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002011
TDLBFD-002011
Paging
Paging
1526726885 L.Paging.UU.Att Number of UEs contained in paging
messagestransmitted over the Uu interface in a cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002011
TDLBFD-002011
Paging
Paging
1526726886 L.Paging.UU.Succ Number of Successful Paging
Responses from theUE in a Cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002011
TDLBFD-002011
Paging
Paging
1526727212 L.Paging.Dis.Num Number of discarded paging messages
from theMME to UEs
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002011
TDLBFD-002011
Paging
Paging
1526727215 L.RA.GrpA.Att Number of times the contention preamble
in group Ais received
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002010
TDLBFD-002010
Random Access Procedure
Random Access Procedure
1526727216 L.RA.GrpA.Resp Number of times a cell sends a Random
AccessResponse message after receiving a preamble ingroup A
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002010
TDLBFD-002010
Random Access Procedure
Random Access Procedure
1526727218 L.RA.GrpB.Att Number of times the contention preamble
in group Bis received
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002010
TDLBFD-002010
Random Access Procedure
Random Access Procedure
1526727219 L.RA.GrpB.Resp Number of times a cell sends a Random
AccessResponse message after receiving a preamble ingroup B
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002010
TDLBFD-002010
Random Access Procedure
Random Access Procedure
1526727379 L.Traffic.User.Max Maximum number of users in a cell
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002007
TDLBFD-002007
RRC Connection Management
RRC Connection Management
1526728333 L.Traffic.User.Ulsync.Avg Average number of UL
synchronized users in a cell Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002007
TDLBFD-002007
RRC Connection Management
RRC Connection Management
1526728489 L.RRC.ConnReq.Msg.disc.FlowCtrl Number of times the
RRC Connection Requestmessage is discarded due to flow control
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002007
TDLBFD-002007
RRC Connection Management
RRC Connection Management
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
12 trong 13 5/19/2015 6:28 PM
-
Counter ID Counter Name Counter Description Feature ID Feature
Name
1526728490 L.RRC.SetupFail.Rej.FlowCtrl Number of times the
eNodeB sends an RRCConnection Reject message to the UE due to
flowcontrol
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-002007
TDLBFD-002007
RRC Connection Management
RRC Connection Management
1526728491 L.HHO.PrepAttIn.disc.FlowCtrl Number of times the
HANDOVER REQUESTmessage is discarded over the S1 or X2
interfacebecause of flow control (without returning apreparation
failure message)
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-00201801
LBFD-00201802
TDLBFD-00201801
TDLBFD-00201802
Coverage Based Intra-frequency Handover
Coverage Based Inter-frequency Handover
Coverage Based Intra-frequency Handover
Coverage Based Inter-frequency Handover
1526728492 L.HHO.Prep.FailIn.FlowCtrl Number of times that the
target eNodeB sends ahandover preparation failure message for an
intra-duple-mode handover over the S1 or X2 interface tothe source
eNodeB because of flow control
Multi-mode: None
GSM: None
UMTS: None
LTE: LBFD-00201801
LBFD-00201802
TDLBFD-00201801
TDLBFD-00201802
Coverage Based Intra-frequency Handover
Coverage Based Inter-frequency Handover
Coverage Based Intra-frequency Handover
Coverage Based Inter-frequency Handover
1526737817 L.Traffic.Board.UPlane.CPULoad.AVG Average
control-plane CPU usage of a board in aneNodeB
None None
1526737818 L.Traffic.Board.UPlane.CPULoad.MAX Maximum
control-plane CPU usage of a board in aneNodeB
None None
9 GlossaryFor the acronyms, abbreviations, terms, and
definitions, see Glossary.
10 Reference Documents3GPP TS 36.413, "Evolved Universal
Terrestrial Radio Access Network (E-UTRAN); S1 Application Protocol
(S1AP)"1.
3GPP TS 36.331, "Evolved Universal Terrestrial Radio Access
(E-UTRA); Radio Resource Control (RRC); Protocol
specification"2.
Transport Resource Management Feature Parameter
Description3.
SCTP Congestion Control Feature Parameter Description4.
HEEX Startpage file:///C:/Users/tuyennd1483/Desktop/eRAN Feature
Documentation ...
13 trong 13 5/19/2015 6:28 PM