ZTE UMTS RAN Equipment Redundancy Feature Guide

Operator Logo
RAN Equipment Redundancy

WCDMA RAN
Feature Guide
RAN Equipment Redundancy Feature Guide

Version
V4.5
Date
2010-10-15
Author
LouDeGang
Approved By
YuFeng
Remarks
Not open to the Third Party
2010 ZTE Corporation. All rights reserved.

ZTE CONFIDENTIAL: This document contains propriet ary information of ZTE and is not to be
disclosed or used without the prior written permission of ZTE.
Due to update and improvement of ZTE products and technologies, information in this document
is subjected to change without notice.
ZTE Confidential Proprietary
TABLE OF CONTENTS
1
Functional Attribute ............................................................................................ 1
2
2.1
Overview .............................................................................................................. 1
RAN Equipment Redundancy............................................................................... 1
3
3.1
3.1.1
3.1.2
3.1.3
3.1.4
3.1.5
3.2
3.2.1
3.2.2
3.2.3
3.2.4
3.3
3.3.1
3.3.2
3.3.3
3.3.4
Technical Description......................................................................................... 1
Board Redundancy Protection of RAN ................................................................. 2
Brief Introduction to Board Redundancy Protection ............................................. 2
General Process of Board Switching.................................................................... 3
Board Fault Detection ........................................................................................... 7
Redundancy Protection Solution for RNC Boards ............................................... 8
Redundancy Protection Solution for Node B Boards ......................................... 28
Port Redundancy Protection............................................................................... 31
Backup of Ethernet Port...................................................................................... 31
Backup of E1/T1 Port .......................................................................................... 33
Backup of SDH Port ............................................................................................ 33
Backup of Level-1 Switch Optical Port ............................................................... 33
Redundancy Protection of Communication Link ................................................ 36
ATM Link Backup ................................................................................................ 36
IP Link Backup .................................................................................................... 36
SCTP Link Backup .............................................................................................. 36
Data Division Transmission Backup ................................................................... 36
4
4.1
4.1.1
4.1.2
4.1.3
4.2
4.2.1
4.2.2
4.2.3
Parameter Description ..................................................................................... 36

Parameter List..................................................................................................... 36
Configuration Information on Board Redundancy Protection ............................ 36
Configuration Information on Port Redundancy Protection ............................... 37
Configuration Information on Communication Link Redundancy Protection ..... 37
Parameter Configuration..................................................................................... 37
Configuration Information on Board Redundancy Protection ............................ 37
Configuration Information on Port Redundancy Protection ............................... 39
Configuration Information on Communication Link Redundancy Protection ..... 39
5
5.1
5.2
Counter and Alarm............................................................................................ 39

Counter List ......................................................................................................... 39
Alarm List ............................................................................................................ 39
Glossary ............................................................................................................. 42
II
FIGURES
Figure 3-1 Principle of Hybrid 1+1 Backup.............................................................................. 3
Figure 3-2 Active/Standby Switching Process......................................................................... 6
Figure 3-3 Active/Standby Interlock Circuit ............................................................................. 7
Figure 3-4 Connection of Major Boards in the RAN ................................................................ 9
Figure 3-5 Before Switching .................................................................................................. 10
Figure 3-6 After Switching of Optical Port ............................................................................. 11
Figure 3-7 After Switching Because of the Fault of Network Port of Internal Media Plane .. 11
Figure 3-8 After Switching Because the Active Board is Offline ........................................... 12
Figure 3-9 DTA Before Switching .......................................................................................... 12
Figure 3-10 DTA After Switching ........................................................................................... 13
Figure 3-11 DTA Hybrid 1+1 Backup Signal Flow................................................................. 13
Figure 3-12 DTI Before Switching ......................................................................................... 15
Figure 3-13 DTI After Switching............................................................................................. 16
Figure 3-14 DTA Hybrid 1+1 Backup Signal Flow................................................................. 16
Figure 3-15 Before GIPI4 Switching ...................................................................................... 17
Figure 3-16 Before EIPI Switching ........................................................................................ 18
Figure 3-17 SDTA2(1:1backup)before switching .................................................................. 23
Figure 3-18 SDTA2(1:1backup)After Switching of Optical Port Abnormal ........................... 23
Figure 3-19 SDTA2(1:1backup)After switchover caused by internal media plane Ethernet
port fault .................................................................................................................................... 23
Figure 3-20 SDTA2(1+1 Backup) Before switching of Internal Media Plane Ethernet Port
Failure ....................................................................................................................................... 24
Figure 3-21 SDTA2(1+1backup)After switching of Internal Media Plane Ethernet Port
Failure ....................................................................................................................................... 24
Figure 3-22 SDTA2 After switching of active board off-line .................................................. 24
Figure 3-23 SDTI before switching of optical port ................................................................. 25
Figure 3-24 SDTI After switching of optical port.................................................................... 26
Figure 3-25 SDTI(Hybrid 1+1backup)Before switching of internal media plane Ethernet Port
failure ........................................................................................................................................ 26
Figure 3-26 SDTI(Hybrid 1+1backup)After switching of Internal Media Plane Ethernet Port
Failure ....................................................................................................................................... 26
Figure 3-27 After switching of active board is offline............................................................. 27
Figure 3-28 Typical Configuration of B8200 .......................................................................... 28
III
Figure 3-29 FE Routing Interconnection Between UIM and UIM.......................................... 32

Figure 3-30 Trunk Interconnection of Control Plane ............................................................. 33
Figure 3-31 GUIM-GLI Connection........................................................................................ 35
Figure 3-32 Double-Shelf Interconnection Between GUIM Boards ...................................... 35
TABLES
Table 3-1 Backup Modes of RNC Boards ............................................................................... 8
Table 3-2 describes the backup mode of each board in the B8200. .................................... 28
Table 3-3 Backup Modes of B8200 Boards........................................................................... 28
Table 3-4 FE Routing Ports and Trunk Ports Supported by Various Boards ....................... 31
Table 3-5 GE Ports Supported by the GUIM/GLI .................................................................. 34
IV
Functional Attribute
System version: [RNC V3.09, Node B V4.09, OMMR V3.09, OMMB V4. 09]
Attribute: [Optional]
Involved NEs:
UE
-
Node B
RNC
MSCS
-
MGW
-
SGSN
-
GGSN
-
HLR
-
Note:
*-: The NE is not involved.
*: The NE is involved.
Dependency: [None]
Mutual-exclusion function: [None]
Remarks: [None].
Overview
2.1
RAN Equipment Redundancy

The 1+1 backup mode is configured for the key boards of the RNC and N ode B, for
example, the system clock board, operation & maintenance board, and control plane
processing board. The load sharing backup mode (also called the resource redundancy
pool) is configured for the user plane processing board and various interface b oards.
Thanks to the redundancy configuration, the functions and performance of the whole
system are not affected in case one board is faulty. The interface board that processes
fiber transmission is configured with automatic protection switching (APS), th us ensuring
the reliability of high-speed lines, especially optical port transmission.
Technical Description
There are three types of redundancy protections in the RA N, that is, board redundancy
protection, port redundancy prot ection, and communication link redundancy protection.
3.1
Board Redundancy Protection of RAN

The RA N system provides diversified board backup modes, so as to ensure normal
system operation in case a board is faulty.
3.1.1
Brief Introduction to Board Redundancy Protection

In the RA N system, the board redundancy backup mode can be configured through the
Back UpMode paramet er. The Back UpMode paramet er can be configured to
BOARD_BACKUP _MODE_NONE (no backup) or the following values:
1+1 backup (B OARD_BA CKUP _MODE_MAS TE RSLAVE)

For the mutual -backup boards, only one board carries services (including access
and processing) at a time and the standby board does not carry any service.
Hybrid 1+1 backup (BOA RD_BACK UP_MODE_MAS TE RSLAVE)

Hybrid 1+1 backup is one type of 1+1 backup, which only has 1+1 backup at
processing boards but not interface boards.
1:1 backup (B OARD_BA CKUP _MODE_1TO1)

Only one board processes data at a time. The interface units can be distributed in
the mutual-backup boards.
Load sharing (B OARD_BA CKUP_MODE_MULTI_ONE )

All boards in t he resource pool are active to share the load. In the redundancy
configuration mode, the system is not affected in case one board is faulty.
The comparison bet ween the 1+1 mode and 1:1 mode is as follows:
Common feature
The two mutual-backup boards share one logical address (compared wit h physical
address, a logical address indicates a logical node for service. When function is
implemented, the active/standby board configuration is not concerned) and one
suite of bearer resources (for example, the circuit resource an d internal media
stream resource)
Difference
1+1 backup: Only one board is running in active status at a time.
1:1 backup: Res ourc es and interfaces are divided into groups, and protection
switching is implemented in each group. The two mutual-backup boards work
cooperatively. The groups in a board can be in different working states. However,
make sure that if one group in a board is active, it must be standby in the other
board For example, in one board, Group A is active and Group B is standby. Then,
in the other board, Group A is standby and Group B is active. As a result, the two
boards may both have active resources at the same time and can process the
access service. For the SDTA 2, SDTI and APBE boards, the corresponding optical
ports in the active board and standby board compos e an APS group. The
active/standby state of a board is independent from the active/standby state of the
external optical port of the board.
Principle of Hybrid 1+ 1 backup:
As it is shown in : The coming data stream A and B are

distributed into standby and active board respectively. Data stream A on standby
board will form as internal data stream after processing and connects to active
board through dedicated channel between active and standby board. Active board
will process the data stream from standby board and combines it with the data
stream B into new data stream C. Data stream C is a logical channel, the up level
processing only needs to process the data stream itself but n ot needs to know
the exact access point. Compare to the multi boards load sharing, hybrid 1+1
backup doesnt need extra internal switching network.
Figure 3-1
Principle of Hy brid 1+1 Backup
Standby Board
Data stream
A
Cross
board
Channe
l
Access
Data stream B
A,B combine into C by

load sharing
Active Board
3.1.2
General Process of Board Switching

For the boards or devices in the 1+1 backup mode or 1:1 backup mode, the system not
only reports an alarm, but also takes preventive and self-healing measures (for example,
switching) in case a board is faulty. When the active board is faulty, the standby board
takes over its work, thus preventing the service functions from being affected for a long
time.
In addition, the system also provides some means to adjust the active board manually.
To verify whet her the current standby board can work normally, manual switching can be
initiated to switch the standby board to the active state.
The RAN system provides the basic common switching reasons for the boards in 1+1
backup mode and in 1:1 backup mode. For different boards, the RAN provides different
switching reasons for different devices and functions on the boards. The following
section describes the general switching reasons and switching process.
3.1.2.1
Reason for Board Switching

The general switching reasons are as follows:
1
Man-machine Command
The system allows user to initiate a switching to the specified board from the OMC
by command.
Key-press switching
Each board supports the switching that is initiated through manual intervention from
a panel key-press. Press E XCH key in the panel of the active board, and then user
can initiate the active/standby switching.
Abnormal switching due to reset of active board

This is an abnormality proc essing flow after the active board is reset. If the active
board is reset due to manual intervention or its abnormal operation, the original
standby board automatically initiates the abnormal switching process, thus switches
itself into the active board. In this switching mode, the active board is suddenly
reset, so some service data cannot be synchronized to the standby board. After the
standby board takes over the work, some service dat a is lost.
Abnormality of network interfac e of the control plane

Boards work on the basis of messages of the control plane. When the net work
interface of the control plane is abnormal, the normal operation of boards is surely
affected. If detecting that the net work interface of the control plane does not receive
any packet, the system resets the board and thus attempts to restart the net work
interface. Before the reset, the system instructs the standby board to initiate the
switching process through the primary channel so t hat the standby board can tak es
over the ongoing services.
Removal of board
Each board is configured with an unplugging switch. Before a b oard is removed,
the system initiates the active/standby switching through an interrupt notice. The
standby board takes over the work first before the active board is out of position.
Meanwhile, the system reports the switch-on alarm and lights up. During the
switching, the E NUM indicator flashes quickly at a frequency of 5 Hz. Upon
completion of the switching, the ENUM indicator flashes at a frequency of 1 Hz.
Before the ENUM indicat or flashes at a frequency of 1 Hz, do not remove the board.
Otherwise, the system triggers the abnormal switching process due to the reset of
the active board, as described in 3).
Besides the common reasons described above, other reasons, such as operation faults
of board, may trigger the active/standby switching.
3.1.2.2
Switching Priority Control

Board switching is intended to restore the system from the faulty state. If the standby
board has severer problem than the active board , the switching proc ess is of no use. To
avoid this problem, user can configure the switching reason according to the fault
severity.
Except for the switching initiat ed by manual intervention, other types of switching are
initiated by the system automatically, that is, each switching reason corresponds to a
specific board fault. If detecting a fault in the acti ve board, the system initiates the
active/standby switching. If detecting a fault in the standby board, the system sets a fault
flag (hereinafter referred t o as the switching barring reason). According to the severity of
faults, the system sets some levels for the switching reas ons and switching barring
reasons. The higher the level is, the severe the fault is. When the active board initiates
the active/standby switching, the standby board compares the level of the switching
reason with the level of the switching barring reason. If the level of the switching reason
is higher, it indicates that the fault of the active board is severer and thus the system
allows the active/standby switching. Otherwise, the fault of the standby board is severer
and thus the system bars the active/standby switching.
To avoid frequent active/standby switching and switching -back, the system sets a lowlevel 5-minute switching barring every time a switching process is complete.
3.1.2.3
Switching Flow
When both the active and standby boards work normally, every functional module in the
active board synchronizes the key information to the standby board at real time.
A normal switching process comprises three phases including preliminary switching,
standby-to-active (SA) switching, and active-to-standby AS) switching, thus switching the
board process and communication link, as shown in Figure 3-2.
Figure 3-2
Active/Standby Switching Process
Active board
Starting switch
Standby board
Pre
Sync data
Begin
communication
link switch
SA
AS
End
Hardware
communication
active/standby
link switch
switch
back
After receiving a switching request, a board first makes necessary judgment, for
example, whether the current board is an active board, and whether the standby board
is in position. If the s witching conditions are met, the active board sends a preliminary
switching command. The standby board compares the level of the switching barring
reason with the level of the switching reason. If the level of the switching barring reason
is higher, the system bars the active/standby switching. Ot herwise, the system
implements the preliminary switching operation, that is, instruct the active board and
standby board of the registered process to perform necessary active/standby data
synchronization. In this phas e, the active board is always in active state and can
implement normal service functions. After the preliminary switching is complete, the
system implements the active/standby switching bet ween the active process and
standby process. The active process and standby proc ess are in different boards, so the
ongoing services are affected transiently. To minimize the impact, the system
implements link switching upon completion of the preliminary switching, that is, instruct
another service board to cache messages and send the cached messages to the new
active board upon completion of the active/standby switching.
In the SA(Slave to Active) phase and AS(Active to Slave) phase, the system instructs
the registered processes to implement the active/standby switching. The detailed
switching is implemented in each process respectively .
After the SA switching and AS switching are complete, the system switches to the
original link and sends the messages cached by another servic e board to the new active
board.
All normal active/standby switching processes follows the above flow. For the a bnormal
switching due t o reset of the active board, the system cannot follow the above switching
flow becaus e the active board is not in position. After receiving the interrupt message
because the active board is reset or out of position, the original stan dby board
immediat ely implements the SA switching and link switching and then takes over the
services of the original active board.
3.1.3
Board Fault Detection

The RA N provides diversified mechanisms for checking whether a board works normally.
The following section describes some fault det ection mechanisms common to multiple
boards. In addition, different functional boards also provide other fault detection
mechanisms.
3.1.3.1
Partner Board In-Position Detection

Through an interlock circuit, the active and standby boards implement the active/standby
competition and check whether the partner board is in position. Figure 3-3 shows the
interlock circuit.
Figure 3-3
Active/Standby Interlock Circuit
A A
Board
BB
Board
MASCT
MASCT
O_MS-VIE O_MS-VIE
I_MS-VIE
I_MS-VIE
MASCT (reverses the phase of output signals or MAS CT inside the logic) is the
active/standby competition signal controlled by software, and is used for the
active/standby competition bet ween boards. I_MS -VIE and O_MS -VIE are the interlock
signals between boards. If the actual level of MASCT is a high level, it indic ates that the
board is competing for the active state. If the actual level of MAS CT is a low level, it
indicates that the board does not compete for the active state. If the actual level of
O_MS-V IE (I_MS-V IE signal of the peer board) is a low level, it indicates that the board
is in active state. Otherwise, it indicates that the board is in standby state.
If the active board detects that the standby board is in position, the active board
synchronizes its data to the standby board through the primary channel. If the standby
board detects that the active board is offline, the standby board is switched to an active
board.
3.1.3.2
Communication Check for the Internal Media Plane

If a board (for example, APBE, RUIB, and EUIP) has the int ernal media plane
communication capability, the board can also provide a mechanism of checking the
communication in the internal media plane. Through this mechanism, a board can trigger
the active/standby switching if detecting the abnormal communication in its internal
media plane.
The port check mechanism in the int ernal media plane is as follows: The board regularly
sends a check packet to the switching board GUIM from the specified media plane port .
After rec eiving the check packet, the GUIM loops back the check packet. If the board
does not rec eive the looped check packet within a certain period, the board determines
that the communication of the media plane is abnormal.
After detecting the abnormal communication of the media plane, the active board
triggers the switching.
3.1.3.3
Internal HW Communication Detection

If a board (for example, the IMAB and EUIP) has the internal HW communication
capability, the board c an also provide a mechanism of checking the internal HW
communication. The check mechanism is as follows: The board sends a check packet to
the switching board GUIM at Timeslot 0 of HW. The GUIM loops back Timeslot 0. The
board checks whether the received loop back data is consistent with the sent data. If not,
the board determines that the HW link is faulty.
The active board triggers the switching process after detecting the fault of the HW link.
3.1.4
Redundancy Protection Solution for RNC Boards

The table describes the backup mode of each board in the RNC.
Table 3-1
Backup Modes of RNC Boards
Functional
Board
Backup Mode
Remarks
APBE
1:1 backup
When the active board is
switched to a standby
board, the original active
board is not reset.
The 1:1 backup mode must be configured if

APS between boards is required.
Support the APS of t wo pairs of S TM -1
interfaces in a board or one pair of S TM-1
interfaces bet ween boards.
DTA
Hybrid 1+1 backup, 1:1
None
Functional
Board
Backup Mode
Remarks
backup
DTI
Hybrid 1+1 backup, 1:1

backup
None
SDTA2
Hybrid 1+1 backup
None
SDTI
Hybrid 1+1 backup
None
GIP I4
1+1, load sharing
None
GLI
Load sharing
None
GUIM
1+1 backup
None
ICM
1+1 backup
None
THUB
1+1 backup
None
PSN
Load sharing
None
PWRD
No backup
None
RCB
1+1 backup
None
ROMB
1+1 backup
None
RUB
Load sharing
None
SBCX
1+1 backup
None
UIMC
1+1 backup
None
Figure 3-4 shows the connections of the main boards in the table.
Figure 3-4
Connection of Major Boards in the RA N

Switch Shelf BPSN
Control Shelf BCTC

ICM
PSN
UIMC
GPS
Antenna
UIMC
ROMB
RCB
GLI
GLI
CHUB
RCB
Resource Shelf BGSN
GUIM
GUIM
Control Plane
User Plane
APBE
STM-1
GIPI
IMAB
DTB
E1
GE
Iu/Iur/Iub Port
Control Plane Data Stream
CircuitHW
EIPI
RUB
GIPI
FE
OMC-B
FE
OMC-R
Interface Unit
User Plane Data Stream
Switch Unit
Circuit Data Stream
Processing Unit
3.1.4.1
APBE Troubleshooting Process

The APBE provides four external S TM -1 optical ports. The APBE boards support the 1:1
backup between boards. In this case, APS should be configured for the optical ports
between the APBE boards. When the inter-board APS is configured, 4 optical port APS
protection groups bet ween boards are supported. The protection of optical ports of the
APBE boards complies with G.841[1], and supports the 1:1 bidirectional protection of
APS. For details on APS, refer to ZTE UMTS ATM Transmission Feature Guide.
The switching of the APBE boards can be triggered by the following reasons:
1
The active optical port is abnormal.
The communication to the internal media plane is abnormal.
The active board is offline.
Figure 3-5 shows the switching triggered by the fault of the active optical port. The
colored lines indicate the connection that data transits currently. The colored blocks
indicate the processing units that data transits currently. Before switching:
Figure 3-5
Before S witching
Active Board
Control
Plane
Processing
Unit
Workig Optical
Port
Media
Plane
Processing
Unit
Inner Control Plane

Ethernet Port
Inner Media Plane

Ethernet Port
Inner Control Plane Standby Board

Ethernet Port
Control
Plane
Processing
Unit
Inner Media Plane

Ethernet Port
Media
Plane
Processing
Unit
Protection Optical
Port
Through the APS mechanism for optical ports, the system initiates the switching after
detecting that the frame of the active optical port is lost. Meanwhile, the media plane
processing units of the faulty APBE are switched to the APBE where the protected
optical port is located. After switching:
10
Figure 3-6
After S witching of Optical Port

Active Board
Control
Plane
Processing
Unit
Working
Optical Port
Media
Plane
Processing
Unit
Inner Control Plane

Ethernet Port
Inner Media Plane

Ethernet Port
Inner Control Plane

Ethernet Port
Standby Board
Control
Plane
Processing
Unit
Media
Plane
Processing
Unit
Inner Media Plane

Ethernet Port
Protection Optical
Port
After the active board detects that the communication to the internal media plane is
abnormal (for details, see Section 3.1.3.2Communication Check for the Internal Media
Plane), the active board is switched over. After switching:
Figure 3-7
Plane
After S witching Because of the Fault of Network Port of Int ernal Media
Active Board
Control
Plane
Processing
Unit
Workin Optical
Port
Media
Plane
Processing
Unit
Inner Control Plane

Ethernet Port
Inner Media Plane

Ethernet Port
Inner Control Plane

Ethernet Port
Standby Board
Inner Media Plane

Ethernet Port
Control
Plane
Processing
Unit
Media
Plane
Processing
Unit
Protection Optical
Port
When the active board is reset or is removed, the standby board detects that the active
board is offline through the interlock circuit (for details, see Section 3.1.3.1Part ner
Board In-Position Det ection). Then, the standby board is switched to the active board,
and completely functions as the active board. After switching:
11
Figure 3-8
After S witching Because the Active Board is Offline
Active Board
Control
Plane
Processing
Unit
Working
Optical Port
3.1.4.2
Media
Plane
Processing
Unit
Inner Control Plane

Ethernet Port
Inner Media Plane

Ethernet Port
Inner Control Plane

Ethernet Port
Standby Board
Inner Media Plane

Ethernet Port
Control
Plane
Processing
Unit
Media
Plane
Processing
Unit
Protection Optical
Port
DTA Trouble shooting process

DTA board directly provides E1/T1 interfaces. It supports inter-boards 1:1 backup and
Hybrid 1+1backup.
The following reasons may cause faulty switch of DTA:
Communication to internal media plane is abnormal
Active board is offline
The switching process of DTA board is shown in the following figures. The lines in color
depict the active path where dat a is currently carried on. Whilst, the blocks in color
shows the active unit where dat a is currently processed. Before switching:
Figure 3-9
DTA Before Switching
Active Board
Control
Plane
Processi
ng Unit
External
DT
access
Media
Plane
Processi
ng Unit
Internal Control
Plane Ethernet
Port
Internal Control
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Cross Board
HW
DT
Access
Standby Board
Control
Plane
Processin
g Unit
Media
Plane
Processin
g Unit
DT
Access
After switching caused by any reas ons:
12
Extern
al DT
access
Figure 3-10
DTA A fter Switching
Active Board
Control
Plane
Processi
ng Unit
External
DT
access
Media
Plane
Processi
ng Unit
Internal Control
Plane Ethernet
Port
Internal Control
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Standby Board
Cross Board
HW
DT
Access
Control
Plane
Processin
g Unit
Media
Plane
Processin
g Unit
DT
Access
Extern
al DT
access
When active board detecting abnormal communication of internal media plane, its will
initiate switching process. Detection mechanism refer to 3.1. 3.2.
When det ecting active board offline, standby board will switch over to active board and
take over all of active boards function.
Principle of DTA hybrid 1+1 backup is shown below:
E1/T1 signal input through DTA interfac e is processed into internal HW and converged
into active board through TDM switching chip. TDM signal is send to IMA processing
chip which supports IMA group. EMA link messages from multiple E 1 belong to one
logical path, IMA chip will make load sharing on active IMA links. At the back plane,
message of IMA group link connects to NP through SPI.
Figure 3-11
DTA Hybrid 1+1 Backup Signal Flow
32 E1/T1
access
8*8MHW
Access processing
m*n MHW
TDM Swith
Link
processing
Load sharing
processing
SPI
Message
processing
To
Internal
Switch
Network
232MHW
32 E1/T1
access
8*8MHW
Access processing
m*n MHW
TDM Switch
Link
processing
Load sharing
processing
SPI
Message
Processing
To
Internal
Switch
Network
Totally there are 64 E 1/T1 in the unit, the first 32 E1/ T1 connect to the left board in order,
the second 32 E1/ T1 connect to the right board in order. When configuring link
group(IMA link group, MP), equably distribute the PPP/IMA links to the right and left
board in order to get the best protection.
13
Physically the 2 *32 M HW is multiplied on one pair of LVDS (Low Voltage Differentiate
Signal). This LVDS only exists in BGSN frame, so this backup scheme can only be
applied to BGSN.
The connectivity logic of left and right board is same, but the connectivity result is
different due to different mapping relationships. For example: Suppose link group C is
composed of link group A and B, A is configured at No.1 E1 (first port of left board), B is
configured at No. 33 E1 (first port of right board). When left board is active, it connects
No.1 E1 to A, and the first E1 on right board will connect to B through HW. When right
board is active, it connects its No.1 E1 to B, and left board will connect its No.1 E1 to A
through HW.
3.1.4.3
THUB Troubleshooting Process

In the RAN system, the THUB is the center of data exchange bet ween control planes in
different shelves. The THUB boards work in the 1+1 backup mode. The port of the
control plane undergoes high-resistance multiplexing. The communication of the control
plane is implemented through the active po rt of the active board. When a port of the
THUB is abnormal, the control plane communication is interrupted between the shelf
connected to this port and other shelves and thus the shelf cannot work normally. When
the THUB is abnormal, the control plane communication is interrupted among all shelves
and thus the system cannot work normally. To avoid the problem, the THUB boards
provide port redundancy backup and board redundancy backup. There are the following
backup scenarios:
1
The control plane communication between shelves is abnormal

In the TRUNK interc onnection mode, the system checks the link of the control
plane according the hardware status of the control plane and packet
receiving/transmitting status. If detecting that one or more ports in a TRUNK gro up
in the active board are abnormal for at least one minute, the system switches over
the board, so as to ensure normal communication of the control plane.
GE port protection inside BCTC shelf

The THUB is interconnected to the UIMC in the same shelf throug h the GE port of
the backplane. If the link is interrupted, the control plane communication between
other shelves and the other servic e boards in this shelf is interrupted, thus causing
a great impact. The THUB checks the receiving/transmitting status of t he packets
between the THUB and the GE port of the backplane of the UIMC. If det ecting that
data receiving/transmitting is abnormal, the THUB performs board switching and
thus attempts to recover itself.
For further information on THUB backup, see Section 3. 2.1.
14
3.1.4.4
DTI Troubleshooting Process

DTI board directly provides E1/ T1 interfaces. It supports inter-boards 1:1 backup and
hybrid 1+ 1backup.
The following reasons may cause faulty switch of DTI:
The switching proc ess of DTI board is shown in the following figures. The lines in color
depict the active path where dat a is currently carried on. Whilst, the blocks in color
shows the active unit where dat a is currently processed.
Before switching:
Figure 3-12
DTI Before Switching
Active Board
Control
Plane
Processi
ng Unit
External
DT
access
Media
Plane
Processi
ng Unit
DT
Access
Internal Control
Plane Ethernet
Port
Internal Control
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Cross Board
HW
Standby Board
Control
Plane
Processin
g Unit
Media
Plane
Processin
g Unit
DT
Access
Extern
al DT
access
After switching caused by any reas ons:
15
Figure 3-13
DTI After Switching
Active Board
Control
Plane
Processi
ng Unit
External
DT
access
Media
Plane
Processi
ng Unit
Internal Control
Plane Ethernet
Port
Internal Control
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Standby Board
Control
Plane
Processin
g Unit
Media
Plane
Processin
g Unit
Cross Board
HW
DT
Access
Extern
al DT
access
DT
Access
When active board detecting abnormal communication of internal media plane, its will
initiate switching process. Detection mechanism refer to 3.1. 3.2.
When det ecting active board offline, standby board will switch over to active board and
take over all of active boards function.
Principle of DTI hybrid 1+ 1 backup is shown below:
E1/T1 signal input through DTI interfac e is processed into internal HW and c onverged
into active board through TDM s witching chip. TDM signal is send to NP (Network
Processing) after framing processing of HDLC. According to configuration, one MP (Multi
PPP) link may be formed by multiple PPP links. Under hybrid 1+1 backup, those P PP
will be distributed at active and standby board. NP only needs to process the MP links
and doesnt need to care about the PPP itself which form the MP and where is the
bottom signal of the PPP.
Figure 3-14
DTA Hybrid 1+1 Backup Signal Flow
32 E1/T1
access
8*8MHW
Access processing
m*n MHW
TDM Swith
Link
processing
Load sharing
processing
SPI
Message
processing
To
Internal
Switch
Network
232MHW
32 E1/T1
access
8*8MHW
Access processing
m*n MHW
TDM Switch
Link
processing
Load sharing
processing
SPI
Message
Processing
To
Internal
Switch
Network
Totally there are 64 E 1/T1 in the unit, the first 32 E1/ T1 connect to the left board in order,
the second 32 E1/ T1 connect to the right board in order. When configuring link
16
group(IMA link group, MP), equably distribute the PPP/IMA links to the right and left
board in order to get the best protection.
Physically the 2 *32 M HW is multiplied on one pair of LVDS (Low Voltage Differentiate
Signal). This LVDS only exists in BGSN frame, so this backup scheme can only be
applied to BGSN.
The connectivity logic of left and right board is same, but the connectivity result is
different due to different mapping relationships. For example: Suppose link group C is
composed of link group A and B, A is configured at No.1 E1 (first port of left board), B is
configured at No. 33 E1 (first port of right board). When left board is active, it connects
No.1 E1 to A, and the first E1 on right board will connect to B through HW. When right
board is active, it connects its No.1 E1 to B, and left board will connect its No.1 E1 to A
through HW.
3.1.4.5
GIPI4 Troubleshooting Process

The GIPI4 provides an external 1000/100M GE port directly. The GIPI4 boards support
the 1+1 backup mode and load sharing mode.
The failures of the GIPI4 boards that may trigger switching are as followings:
1
The external Ethernet port is faulty.
Within 2 seconds, the external Ethernet port receives more than 128 bad frames.
The communication to the internal media plane is abnormal.
Figure 3-15 shows the GIP I4 switching process. The colored lines indicate the
connection that data transits currently. The colored blocks indicate the processing units
that data transits currently. Before switching:
Figure 3-15
Before GIPI4 Switching
Active Board
Control
Plane
Processing
Unit
Outer Ethernet
Port
Media
Plane
Processing
Unit
Inner Control Plane

Ethernet Port

Ethernet Port
Control
Plane
Processing
Unit
Inner Media Plane

Ethernet Port
Inner Media Plane

Ethernet Port
Media
Plane
Processing
Unit
Outer Ethernet
Port
Figure 3-16 shows the result after the switching triggered by any reason:
17
Figure 3-16
Before EIP I Switching

Active Board
Inner Control Plane

Ethernet Port

Ethernet Port
Control
Plane
Processing
Unit
Inner Media Plane

Ethernet Port
Inner Media Plane

Ethernet Port
Control
Plane
Processing
Unit
Outer Ethernet
Port
Media
Plane
Processing
Unit
Media
Plane
Processing
Unit
Outer Ethernet
Port
The external Ethernet faults are judged according t o standard Ethernet circuit signals. In
case frames are lost, the system determines that the Ethernet is faulty and thus triggers
the switching process.
Bad frame detection is implemented according to the check result of the Ethernet frames
received by the bottom-layer chip. If more than 128 bad Ethernet frames are received
within 2 seconds, the system determines that the Ethernet works abnormally and thus
triggers the switching process.
After the active board detects that the communication of the internal media plane is
abnormal, the system initiates the s witching. For details on t he check mechanism, see
Section 3.1.3.2Communication Check for the Internal Media Plane .
When the active board is reset or is removed, the standby board detects that the active
board is offline through the interlock circuit (for details, see Section 3.1.3.1Part ner
Board In-Position Det ection"). Then, the standby board is switched to the active board,
and completely functions as the active board.
3.1.4.6
GLI Troubleshooting Process

The GLI is the exchange center of the media plane dat a between shelves. The board
backup mode is load sharing.
The GLI has four media-plane GE optical ports. In a Level-1 switching system, the
corresponding GE interfaces of the GLI work in the 1+1 backup mode. The
active/standby states of the GE interfaces are independent from each ot her. A whole
GLI has four independent active/standby GE modules. Two GLIs in the load sharing
mode may both have active GE modules. If one ac tive GE module is abnormal, the
associated services are affected. If the whole GLI is abnormal, none of the GE modules
in the GLI can work normally, and the related services are interrupt ed. The redundancy
backup of the GLI and GE interfac es can prevent this problem effectively without
affecting the services. For details, see Section 3.2.4Backup of Level-1 Switch Optical
Port.
18
In addition, the GLI is interconnected to the PSN through the HSSL. The switching chip
in the FLI can identify the status of the HSSL. If the HSSL is interrupted, no packets are
sent from the HSSL to the PS N for switching.
3.1.4.7
GUIM Troubleshooting Process

The GUIM is the switching cent er for media plane communication, control plane
communication, and HW communication in the resourc e shelf (BGS N). During inter-shelf
communication, the GUIM has the routing function for the control plane and media plane.
The GUIM boards work in the 1+1 backup mode. The rear insertion card provides a pair
of external routing interfaces and one pair of TRUNK interfaces for interconnection of the
control plane. The panel provides two external GE int erfaces. Like the THUB and GLI,
the system takes the redundancy backup measure without affecting the system when
the control plane link or media plane link in the GUIM is abnormal.
The GUIM is protected in following res pects:
3.1.4.8
Redundancy protection for the control plane link, including board protection and port
protection. For details, see Section 3.1.2General Process of Board Switching and
Section 3.2.1B ackup of Ethernet Port.
Media plane link protection: The media plane of the GUIM forwards the
active/standby state of the GE ports together with the active/standby state of boards.
The system checks a link by receiving and transmitting the link check packet. If
detecting that the link of the optical port in the active board is faulty, the system
switches over the optical module or the board. For details, see Section
3.2.4Backup of Level-1 Switch Optical Port.
For the net work interface of the internal control plane, the system periodically
queries the receiving/transmitting statistics of the network interface of the control
plane. If detecting that the network interface of the control plane is abnormal, the
system switches over the board and thus the original standby board takes over the
work.
For the network interface of the internal media plane, the board in the shelf wit h the
media plane communication capability checks whether the link is normal through
packet interaction. If a faulty board cannot heal itself, the board wit h the media
plane communication capability sends a message to instruct the GUIM to switch
over the faulty board, thus removing the fault of the net work interface in the active
GUIM.
The GUIM periodically checks the status of the HW communication links used in the
shelf. If all HW links are faulty, the system switches over the GUIM.
ICM Troubleshooting Process

The ICM provides the clock management and output function, and works in the 1+1
backup mode. In case a clock is faulty, the system has an extremely high requirement
19
for the recovery time. Therefore, the ICM is automatically switched over in hardware,
and software switching follows hardware switching. After software receives the interrupt
caused by hardware s witching, the switching flow is the same as described in Section
3.1.2General Process of Board Switching. The ICM mainly provides the following
protection mechanisms:
The ICM checks the GPS module periodically. In cas e the active ICM is abnormal, the
system initiates the switching process. In case the standby ICM is abnormal, the system
sets switching barring.
The hardware in the ICM c hecks the status of the active and standby boards and
compares hardware signals. If the active board is abnormal but the standby board is
normal, the hardware automatically initiates hardware switching. After the hardware
switching is complete, the interrupt informs the software that the hardware switching is
complete. A fter receiving the interrupt, the software implements software switching, thus
keeping the active/standby consistency between the soft ware and hardware.
The ICM supports the input of multiple clock benchmarks, thus providing redundancy
protection bet ween clock benchmarks. The ICM checks the validity of each access clock
benchmark. If det ecting that the clock benchmark is invalid, the ICM gives an alarm and
switches over the clock benchmark as configured.
3.1.4.9
PSN Troubleshooting Process

The PSN mainly exchanges the media stream data between the GLIs. The PSN boards
work in the load sharing mode. If two PSN boards are normal, the media streams are
shared by the two boards for forwarding. In case one PSN cannot work normally or the
HSSL is interrupt ed, the switching chip in the GLI automatically stops sending packets in
the HSSL. Instead, packets are sent only in a normal HSSL.
In case one PSN board is abnormal, the maximum switching capacity is halved. Based
on the current design, one PSN board is enough to satisfy the needs of service
switching.
3.1.4.10
RCB Troubleshooting Process

The RCB and ROMB are both a MP board. Except the functions and orientations, their
troubleshooting modes are similar to each other. The following section describes the
RCB troubleshooting process in detail.
The RCB is connected to the switching unit, and is responsible for protocol processing of
the control planes of Iu, Iub, and Iur. The active RCB is connected to the standby RCB
through a 100M Ethernet. The RCB boards work in the 1+1 backup mode.
The software running in the active RCB is the same as that running in the standby RCB
except for their working state (active or standby). During the initial power-on operation,
20
the two RCB boards compete for the working state. Hardware interlocking is configured
between t he active RCB and standby RCB (for details, see Section 3.1.3.1Part ner
Board In-P osition Detection"), thus avoiding the dual -active or dual -standby inclination.
After the active/standby competition is complete, the active and standby RCB boards are
up in different capacities respectively. The active RCB needs to read t he configuration
data and receives the requests of other boards. The standby RCB actively interacts with
the active RCB, so as to ensure the consistency of software versions and static data. If
the data of the active RCB is changed during the normal operation, the system
synchronizes the changing data to the standby RCB through regular synchronization or
real-time synchronization. During the normal operation, the standby RCB receives the
synchronized data of the active RCB and keeps the consistency of static data between
the active RCB and active RCB, thus taking over the work of t he active RCB smoot hly in
real time.
In case the active RCB fails (for details on the reason, see S ection 3.1.2.1Reason for
Board Switching), the standby RCB is automatically switched int o an active RCB, its
logical address becomes the logical address of the active RCB, and the bottom -layer
communication link is also switched, thus ensuring normal communication bet ween the
new active RCB and other boards. The original active RCB is automatically reset. After
the rest, the original active RCB becomes a standby RCB.
3.1.4.11
ROMB Troubleshooting Process

The ROMB processes a global process, controls the operation and maintenance of the
entire system (including the operation and maint enance proxy), connects itself to the
OMC through a 100M Ethernet, and isolates the internal net work segments from the
external network segments. The ROMB also serves as the core of operation and
maintenance of the ZXW R RNC. It monitors and manages the boards in the system
directly or indirectly. The ROMB boards work in the 1+1 backup mode.
The ROMB troubleshooting process is similar to the RCB troubleshooting process.
3.1.4.12
RUB Troubleshooting Process

The backup mode of the RUB boards is load sharing, or called the resource redundancy
pool.
The RNC can be configured with multiple resource shelves, each resource shelf is
configured wit h multiple RUB boards, and each RUB board is configured with one CPU
and 14 (V TCD board) OR 15 (V TCD_2) DSPs. For RUB troubleshooting, the minimum
resource unit is a DSP.
When configuring a RCP module (the RcpModule parameter) through the OMC, you
need to configure the subsystem number (the SubSystem parameter) corresponding to
the preferred shelf for the RCP module. To obtain the RUB resources for new service
setup in the RCP, the RUB boards in the preferred res ourc e shelf are polled. If a RUB
21
board is available, the RUB board is selected for service set up. Ot herwise, a RUB board
is selected from another alt ernative shelf.
After the RCP selects a RUB board, a DSP will be selected in the RUB according to a
certain strategy. After the DSP is selected, the RUB board notifies the DSP number to
the RCP.
When a DSP in a RUB board is faulty, the status of the DSP is Unavailable. After
receiving the Unavailable notice, the RCP releases the DSP-related service, thus
avoiding resource deadlock. When a subsequent service is set up, the faulty DSP in the
RUB board is not selected for service setup until the DSP becomes normal again. If all
DSPs in the RUB board are faulty or the CP U of the RUB board is faulty, the RCP does
not select the RUB board for setting up a new service.
3.1.4.13
SDTA2 Troubleshooting Process

SDTA2 provides 4 CS TM-1 optical ports, and supports 246 E1 (each optical port support
maximum 63 E1) and maximum 168 IMA groups.
SDTA2 supports inter-board 1:1 backup and hybrid 1+1 backup, and 4 groups of inter
board optical ports APS protection are supported in this case. APS supports the
following protection types: 1+1 uni-direction, 1+1 bi-direction, 1:1 uni-direction and 1:1
bi-direction. All E1s accessed to the active optical port of backup board are transferred
to active board for processing (IMA links are also processed on active board). In case of
hybrid 1+1 backup, there is no backup on optical ports, all E1s access to standby board
are transferred to active board for processing.
SDTA2 faulty switching includes the following reasons:
Active optical port is abnormal
The switching process of SDTA 2 board is shown in the following figures. The lines in
color depict the active path where data is currently carried on. Whilst, the blocks in col or
shows the active unit where dat a is currently processed.
For 1:1 backup, diagram of before switching is shown below:
22
Figure 3-17
SDTA 2(1: 1backup)before switching
Active board
Control
Plane
Processing
unit
Working Port
Media
Plane
Processing
unit
Standby Board
Internal Control
Plane Ethernet Port
Internal Control
Plane Ethernet Port
Internal Media Plane

Ethernet Port

Ethernet Port
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Protection Port
For 1: 1 board backup mode, when det ecting active optical port abnormal, active board
switching over will be triggered. Status after switching over is shown below:
Figure 3-18
SDTA 2(1: 1backup)A fter Switching of Optical Port Abnormal
Active board
Control
Plane
Processin
g unit
Working Port
Media
Plane
Processin
g unit
Internal Control
Plane Ethernet
Port
Internal Control
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Standby Board
Main Data
Path
Control
Plane
Processin
g unit
Media
Plane
Processin
g unit
Protection
Port
In this case, when S DTA det ecting abnormal communication to internal media plane, it
will trigger switchover. Diagram aft er switchover is shown below:
Figure 3-19 SDTA 2(1: 1backup)A fter switchover caused by internal media plane
Ethernet port fault
Active Board
Control
Plane
Processing
unit
Working port
Media
Plane
Processing
Unit
Internal control
plane Ethernet port
Internal Media plane

Ethernet Port
Internal control
Plane Ethernet port

Ethernet Port
Path for active and

standby data
Standby Board
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Protection Port
For 1+1 hy brid backup, before switching is shown below:
23
Figure 3-20 SDTA 2(1+ 1 Backup) Before switching of Internal Media Plane Ethernet
Port Failure
Active board
Standby Board
Control
Plane
Processing
unit
Working Port
Media
Plane
Processing
unit
Internal Control
Plane Ethernet Port
Internal Control
Plane Ethernet Port

Ethernet Port

Ethernet Port
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Active-standby Data Channel
Working Port
For 1+ 1 hybrid backup, when active board det ecting internal media plane
communication abnormal (refer to 3.1.3. 2), switching will be triggered. After switching
status is shown below:
Figure 3-21
Failure
SDTA 2(1+ 1backup)A fter switching of Int ernal Media Plane Ethernet Port
Active board
Standby Board
Control
Plane
Processing
unit
Working Port
Media
Plane
Processing
unit
Internal Control
Plane Ethernet Port
Internal Control
Plane Ethernet Port

Ethernet Port

Ethernet Port
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Working Port
When active board resets or it is removed, standby board will detect active board offline(refer to 3.1.3.1) through inter-lock circuit and switch to active board to take over all
active boards function.
Figure 3-22
SDTA 2 After switching of active board off-line

Active Board
Control
Plane
Processing
unit
Working port
Internal control
plane Ethernet port

Ethernet Port
Internal control
Plane Ethernet port
Standby Board

Ethernet Port
Media
Plane
Processing
Unit
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Working Port
24
3.1.4.14
SDTI Troubleshooting Process

SDTI provides 2 groups of CS TM-1 optical ports, and each group has two optical
ports(W,P).
SDTI supports inter-board 1:1 backup and hybrid 1+1 backup, and media plane and
control plane units switch separately. APS prot ection configuration is independent from
board configuration, it maximum 2 groups of inter board APS. APS supports the
following protection types: 1+1 uni-direction, 1+1 bi-direction, 1:1 uni-direction and 1:1
bi-direction. The switching of t wo groups APS is separated. The reason for optical switch
includes: LOS, LOF, SF, SD.
SDTI faulty switching includes the following reasons:
Active board is offline or reset.
APS switching includes the following reasons:
LOS, LOF, SF, SD on optical ports.
The switching process of S DTI board caused by the first optical port is shown in the
following figures. The lines in color depict the active path where data is currently carried
on. Whilst, the blocks in color shows the active unit where data is currently processed.
Before switching:
Figure 3-23
SDTI before switching of optical port
Active board
Control
Plane
Processing
unit
Working Port
Media
Plane
Processing
unit
Standby Board
Internal Control
Plane Ethernet Port
Internal Control
Plane Ethernet Port

Ethernet Port

Ethernet Port
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Protection Port
When detecting frame loss on optical port through APS protection mechanism, it will
initiate optical port switching and all media plane processing units will switch to standby
board. After switching:
25
Figure 3-24
SDTI After switching of optical port
Active board
Control
Plane
Processin
g unit
Working Port
Media
Plane
Processin
g unit
Internal Control
Plane Ethernet
Port
Internal Control
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Internal Media
Plane Ethernet
Port
Standby Board
Control
Plane
Processin
g unit
Media
Plane
Processin
g unit
Main Data
Path
Protection
Port
In case of hybrid 1+1 backup, before switching:
Figure 3-25 SDTI(Hybrid 1+1backup)Before switching of internal media plane Ethernet

Port failure
Active board
Control
Plane
Processing
unit
Working Port
Media
Plane
Processing
unit
Standby Board
Internal Control
Plane Ethernet Port
Internal Control
Plane Ethernet Port

Ethernet Port

Ethernet Port
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Working Port
For 1+ 1 hybrid backup, when active board det ecting internal media plane
communication abnormal (refer to 3.1.3. 2), switching will be triggered. After switching
status is shown below:
Figure 3-26 SDTI(Hybrid 1+1backup)After switching of Internal Media Plane Ethernet

Port Failure
Active board
Control
Plane
Processing
unit
Working Port
Media
Plane
Processing
unit
Standby Board
Internal Control
Plane Ethernet Port
Internal Control
Plane Ethernet Port

Ethernet Port

Ethernet Port
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Working Port
26
When active board resets or it is removed, standby board will detect active board offline(refer to 3.1.3.1) through inter-lock circuit and switch to active board to take over all
active boards function. Status after switching is showing below:
Figure 3-27
After switching of active board is offline
Active board
Standby Board
Control
Plane
Processing
unit
Working Port
3.1.4.15
Media
Plane
Processing
unit
Internal Control
Plane Ethernet Port
Internal Control
Plane Ethernet Port

Ethernet Port

Ethernet Port
Control
Plane
Processing
unit
Media
Plane
Processing
unit
Working Port
UIMC Troubleshooting Process

The UIMC board is the switching board of the control plane communicatio n in the control
shelf (B CTC) and switching shelf (BPS N). The UIMC provides the shelf interconnection
interfaces (including the FE routing interface and trunk interface) through a rear ins ertion
card, and provides the control plane network interface int ern ally.
The UIMC boards are under the redundancy protection in the following respects:
1
Redundancy protection for the control plane link between shelves, including board
protection and port protection.
For the net work interface of the internal control plane, the system periodically
queries the receiving/transmitting statistics of the network interface of the control
plane. If detecting that the network interface of the control plane is abnormal, the
system switches over the board and thus the original standby board takes over the
work.
For details, see Section 3.1.2General Process of B oard S witching and Section
3.2.1Backup of Ethernet Port.
3.1.4.16
SBCX Troubleshooting Process

When acting as OMM server, SBCX managers all the NE of RNC and adopts 1+1
backup to avoid OMM service single point failure.
The reason of SBCX failure switching is shown below:
Active SBCX is off-line
27
Active SBCX abandons active status

The rear insertion board of SBCX is off-line
Active OMM software is fault
Active and standby SBCX have the same soft ware running on them, including OmmHost,
OMMservice and Oracle dat abase soft ware. Two SBCX boards will compete for active
and standby during starting up. There is a hardware inter -lock mechanism(refer to
3.1.3.1) which will prevent system from double active or double standby. After the active
and standby competitive status end up, active board and standby boards work in its
status respectively. Only the OMM server soft ware starting up with active status will work
and manage RNC, OMM in standby board only runs active -standby process to ensure
real-time dat a synchronization bet ween active and standby boards.
When active board is faulty, standby board will automatically switch to active status.
Original communication links between former active OMM and RNC as well as MINOS
will automatically switch to new active board. Former active board will start up into
standby status.
3.1.5
Redundancy Protection Solution for Node B Boards

Figure 3-28
Typical Configuration of B8200
PM
FS
FAM
SA
CC
Table 3-2 describes the backup mode of each board in the B8200.
Table 3-2
Backup Modes of B8200 Boards
Functional Board
Backup Mode
CC
1+1 backup
FS
None
PM
1+1 backup
SA
None
BPC
Resource pool
Remarks
28
3.1.5.1
CC Troubleshooting Process
Primary control
The active and standby CC boards work in the 1+1 backup mode. The active and
standby CC boards have completely the same hardware, run the same software,
and implement NBAP processing and the processin g of the IP/ATM protoc ol stack
between the control plane and user plane. At a time, only one CC board is active.
The active CC board carries services, including access and processing. The
standby CC board does not carry any service, but receives the synchr onized data
of the active CC board and keeps the same running conditions as the active CC
board.
There exists an independent communication channel between the active CC board
and standby CC board. The active and standby CC boards run the same software,
but are in different modes. During the initial power-on operation, the two CC boards
compete for the active/standby state. Upon completion of the competition, the
standby CC board actively interacts with the active CC board, so as to ensure the
consistency of soft ware versions and static data. During the normal operation, the
active CC board synchronizes the changing dat a (for example, database dat a,
control information of the control plane and user plane of the current access,
software parameters, and software versions) to t he standby CC board through
regular synchronization or real-time synchronization. During the normal operation,
the standby CC board rec eives the synchroniz ed data of the active CC board and
keeps the consistency of static data between the active CC board and active CC
board, thus taking over the work of the active CC board smoothly in real time.
The system can initiate the active/standby switching in the OMC. Alternatively, the
active/standby switching is triggered by a severe fault of the active CC board. In
this case, the current active CC board switches the current hardware and software
to the standby state, and only processes the synchronized data of t he new active
CC board without processing any service. The original standby CC board swi tches
its software and hardware to the active state and takes over the work of the original
active CC board by using the synchronized static data and real-time operation dat a,
thus minimizing the impact of the fault of the original CC board on the entire s ystem.
Clock
In cas e the active CC board is faulty, the active CC board is automatically to the
standby CC board.
Clock extraction redundancy design for multiple E1/ T1 links: In case an E1/ T1 link
is faulty, the system can automatically switch the faulty E 1/T1 link to another
normal one for clock extraction.
Configuration of multiple clock reference sources: In case the current clock
reference source is faulty, the current clock referenc e source is automatically
switched to the standby clock referenc e source.
29
Clock holdover: When all clock reference sources are faulty, the system can keep
working normally for at least 90 days. (the BB8200 can be held over for at least 180
days)
3.1.5.2
PM Troubleshooting Process
The PM troubleshooting process is divided int o two parts:
1
Active/standby switching mechanism of the PM board itself (including the PM

hardware/software processing)
Processing in the CC board
Active/standby switching of the PM board itself

The PM boards work in the active/standby mode. Both PM boards have an voltage
over a channel (in the 8200 device, every slot has a separate power supply channel)
to be powered. In terms of final selection, the hardware mechanism ensures that
the voltage of only one PM board acts on a specific power supply channel. In case
of the active/standby PM switching, switching is implemented immediately, thus
avoiding power-off of the boards.
Processing in the CC board

In the CC, the active/standby PM backup mode has three actions:
3.1.5.3
When the active PM board detects that a new board is inserted to a specific
slot, the active PM board immediately sends a message to the CC board. If
the CC board decides to supply power for the newly inserted board, the CC
board sends a power supply command to both the active and standby PM
boards, thus ensuring that both PM boards enable the out put of the
corresponding channel.
The CC board sends the active/standby state reported by a PM board to the

OMC in real time, so that the OMC can display the active/standby state of
each PM board correctly.
The CC board periodically checks whether each PM board enables each

power supply channel, and synchronizes the result returned by the active PM
board to the standby PM board, thus ensuring that the active power supply
and standby power supply in the same power supply state.
BPC Troubleshooting Process

The cells set up in a baseband pool share all the baseband resources in the baseband
pool.
30
In case a baseband board is faulty, Node B assigns the cell setup and radio link setup
initiated by the RNC to anot her available baseband board in the baseband pool.
For the existing UEs in the faulty baseband board, Node B sends a reset request to the
RNC, requesting the RNC to res et the UEs in the faulty baseband board. The public
resources (for example, cells) in the faulty baseband board are deleted, and Node B
notifies the deletion result to the RNC through an audit response. The RNC initiates cell
setup and radio link setup again, and Node B assigns them to another available
baseband board in the baseband pool.
3.2
Port Redundancy Protection
3.2.1
Backup of Ethernet Port

Ethernet port backup mainly refers to the backup of the Ethernet ports between the
interconnected shelves. The shelves are interconnected to each other by different
means, mainly routing interconnection and TRUNK int ercon nection.
1
Routing int erconnection: A pair of networks cables is used in t wo ends respectively.

At every end, one network cable is active and the other network cable is standby. In
case a working port is faulty, the faulty port is switched to another port for
continuous communication.
TRUNK interconnection: Through LACP protocol defined in IEEE802.3, the four

physical ports of the same attributes are aggregated, thus increasing bandwidth and
redundancy.
Port backup is also accompanied by board backup. The board backup mode is 1+1
backup. At a time, the Ethernet port in the active board receives and transmits data. In
case the port of the active board is faulty, the active/standby switching is triggered. The
Ethernet port of the new active board is active for transmitting and receiving data.
Table 3-3lists the FE routing ports and trunk ports supported by various boards.
Table 3-3
FE Routing Ports and Trunk Ports Supported by Various Boards
Board
Routing Port (t wo FE ports in each

pair)
Trunk Port (four FE ports in each

group)
UIMC
In the B CTC shelf and BPSN: 3
In the B CTC shelf: 1
GUIM
THUB
In a minor environment, the FE routing interc onnection bet ween UIM (including UIMC
and GUIM) boards can be used for networking. In a major environment, a THUB board
needs to be interconnected to other shelves for networking. In 0, a pair of routing ports
or a group of trunk ports can be inte rconnected between t wo shelves.
31
3.2.1.1
Routing Interconnection
Routing interconnection is also called FE interc onnection. The UIM (including UIMC and
GUIM) board and UIM/ THUB board can each provide one pair of FE ports for
interconnection. The boards work in the 1+1 backup mode. The Ethernet between the
active board and standby board communicates with the external world through high
resistance. 0 shows the connection mode. The blue channels indic ate the availabl e
communication links, and the read channels indicate the high-resistance channels. The
two available communication links include an active link and a standby link. All control
plane communication streams are in the active link. When the active link is faulty, the
system switches the faulty link to the standby link. In this communication mode, the
traffic is only 100M because only one channel is really used.
Figure 3-29
FE Routing Interconnection Bet ween UIM and UIM
UIM
3.2.1.2
UIM
Active
Active
Standby
Standby
TRUNK Interconnection
The routing scheme only supports the t raffic of 100M, and cannot satisfy the needs of
high traffic. Therefore, the inter-shelf cascaded boards are interconnected through a
trunk, so as to increase the inter-shelf cascading traffic (theoretically, the maximum
traffic is as high as 400M). Trunk interconnection complies with the link aggregation
protocol. The physical links with the same transmission media type and transmission
rate are bundled together. Logically, these bundled physical links appear to be one link.
Link aggregation is also c alled t runking, which allows the peer physical link between
switches or between a switch and a server to multiply its bandwidth. Therefore, trunking
is an important technology whereby to inc reas e the link bandwidth and ensure the
flexibility and redundancy of link transmission.
0 shows the connection mode. The blue channels indicate the available communication
links, and the red channels indicate the high -resistance channels. The four available
communication links are aggregated into a group of trunk ports. In case one link is faulty,
communication is implemented through another link.
32
Figure 3-30
3.2.2
Trunk Int erconnection of Control Plane
UIM/THUB
UIM/THUB
Active
Active
Standby
Standby
Backup of E1/T1 Port

In the A TM mode, multiple E1 ports, as IMA links, are bundled in an AIM group. When
some E1 links are interrupt ed, the services carried o ver the E 1 ports are transferred to
other E1 links.
In the IP mode, multiple E1 ports, as PPP links, are bundled in a MLPPP group. When
some E1 links are interrupt ed, the services carried over the E 1 ports are transferred to
other E1 links.
3.2.3
Backup of SDH Port

SDH ports can be backed up in three modes: port backup plus board backup, APS, and
load sharing of optical port.
Port backup plus board backup: The board backup mode is 1+1 backup. At a time, the
optical port in the active board receives and transmits data. In case the port of the active
board is faulty, the active/standby switching is triggered. The optical port of the new
active board is active for transmitting and receiving data.
APS: Support the APS for the optical ports in a board and the optical ports between
boards. For details on APS, refer to ZTE UMTS ATM Transmission Feature Guide.
Load sharing of optical port: The board backup mode is 1:1 backup. At a time, the
optical ports of the active and standby boards both transmit and receive data. The data
received by the optical port of the standby board is forwarded to the active board.
3.2.4
Backup of Level-1 Switch Optical Port

The Level-1 switch system is an IP packet switching system, which mainly comprises
GUIM, GLI, and PSN boards. The GLI and the PSN are cross-connected through a
33
HSSL. The PSN boards work in the load sharing mode, that is, two PSN boards switch
packets at the same time. The port backup mode is applied bet ween GLI boards. The
GLI boards are cross-connected to the GUIM boards in another shelf respectively. At
each end, every GE port has two optical modules for backup. Through active/standby
port backup and optical module backup, the backup proportion bet ween physical links
and logical links is 4:1, thus providing a reliable link assuranc e for the access of IP
packets to the switching center.
The GUIM has two GE ports. Active/standby GE port backup is accompanied by board
backup. E very GE port has two optical modules. The GLI has four GE ports, and every
GE port has two optical modules. Th e two GE ports wit h the same GLI sequence
number have an active/standby relation. The active/standby relation between t he GE
ports with different sequence numbers is independent from each other. 0 lists the GE
ports supported by the GUIM/GLI.
Table 3-4
GE Ports Supported by the GUIM/GLI
Board
Number of GE Ports
Number of Optical Modules per GE Port
GUIM
GLI
E very GE port has two optical modules, thus enabling the double-t ransmitting & onereceiving mechanism through hardware. Both optical modules transmit data to the
external world. At a time, only the active optical port receives data. The hardware also
provides the switching function for the packet-receiving optical module.
The following section describes the inter -shelf fiber interconnection mode in det ail.
3.2.4.1
GUIM-GLI Interconnection
In the GUIM, there are t wo optical ports interconnected to the GLI. The active/standby
relation bet ween the two optical ports is also accompanied by board backup. In normal
cases, the two optical ports both forward the media streams. Traffic balance is attained
between t he two GE ports in the GUIM through a c ertain algorithm. The system
compares the number of the available GE ports in the boards. The system prefers the
GUIM with a larger number of available optical ports as the active end. In case the GE
port of the active GUIM is faulty, board switching is initiated. 0 shows the detailed
connection mode.
34
Figure 3-31
GUIM-GLI Connection
GUIM
#1
G
L
I
#
1
4
1
GUIM
#2
2
1
G
L
I
#
2
3.2.4.2
Interconnection between GUIM Boards in Two Shelves

In a small switch office, the GLI-PSN switch mode is not needed. Instead, the GUIMGUIM double-s helf interconnection mode is enough for the service traffic. In this case,
the following double-shelf interc onnection mode can be used. Unlike the GUIM-GLI
interconnection, only one optical port is provided for interconnection in this mode, as
shown in Figure 3-32.
Figure 3-32
Double-S helf Interconnection Between GUIM Boards
GUIM#1
GUIM#3
GUIM#2
GUIM#4
35
3.3
Redundancy Protection of Communication Link
3.3.1
ATM Link Backup

ATM link backup is configured through the bottom-lay er redundant link. According to the
link status, the upper-layer application selects an available link for backup. The link
status is fed back by monitoring the on/ off state of the physical layer, and is obtained
through the OAM check of the A TM.
3.3.2
IP Link Backup
IP link backup is configured through route redundancy. According to the link status, the
upper-layer application selects an available route for backup. The link status can be
obtained by monitoring the on/off state of the physical layer or through the IP layer
detection prot ocol (for example, BFD). In case an external interface of the system is
faulty, the route related to the faulty interface also fails. When sending data, an
application selects another route. If the BFD mechanism det ects that the corresponding
route is not available, the application does not select the corresponding route, either.
Instead, the application selects anot her normal route.
For details about BFD, refer to ZTE UMTS IP UTRA N Feat ure Guide.
3.3.3
SCTP Link Backup

SCTP uses the Multihomed S CTP mechanism, that is, support the multi -access mode.
In a SCTP couple, there are multiple pairs of IP addresses that support the transmission
of SCTP data. In case a pair of IP channels is interrupted, the status of the SCTP couple
is not affected.
3.3.4
Data Division Transmission Backup

When multiple interfaces (of the same type or different types) support the transmission
of the same service, the data transmitted through a faulty interfac e is migrated to
another normal interface.
Parameter Description
4.1
Parameter List
4.1.1
Configuration Information on Board Redundancy Protection

No.
Abbreviated name
Parameter name
36
4.1.2
BackUpMode
Backup Mode
RcpModule
RCP Module
SubSystem
SubSystem
Configuration Information on Port Redundancy Protection

For details on the parameters related to APS of optical ports, refer to ZTE UMTS ATM
Transmission Feature Guide.
4.1.3
Configuration Information on Communication Link Redundancy

Protection
For details on the BFD parameters of the IP link, refer t o ZTE UMTS IP UTRAN Feat ure
Guide.
4.2
Parameter Configuration
4.2.1
Configuration Information on Board Redundancy Protection
4.2.1.1
Backup Mode
OMC Path
Path: View->Configuration Management ->RNC NE ->RNC Ground Resource
Management-> RA CK-> Create->Board->A vailable Board
Relat ed description
None
Interface parameter description:
The parameter indicates the backup mode of boards. Its value is as follows:
0
No backup
1 1+1 backup (the name is BOARD_BACKUP_MODE_MASTERSLAVE; only one

board is running normally at any time)
2 Load sharing (load sharing or port backup only applies to the boards in adjacent
slots)
37
Back-to-back backup (not used currently)
N+1 backup
1:1 backup
Note: The backup mode is specified when the corresponding boards are configured.
Recommendation: Configure t he backup mode according to the field networking
mode
4.2.1.2
Rcp Module
OMC Path
Management-> RA CK-> Create->Board-> RCP -> Module 1 / Module 2
None
Interface parameter description:
The parameter indicates the module number of the RCP. Its value is 3 to 127.
Recommendation: none
4.2.1.3
Sub System
OMC Path
Management-> RA CK-> Create->Board-> RCP -> Module 1 / Module 2
None
Parameter description: none
The parameter indicates the subsystem number of the resource shelf of the RCP.
Its value is 0 to 19. The invalid value is 0xFF. When the Iub int erface is carried in
the ATM, the available SubSystem [0] of the RCP must be the resource shelf that is
38
accessed by Node B under the jurisdiction of the RCP, and the number of available
shelves is 1.
4.2.2
Configuration Information on Port Redundancy Protection

N/A.
4.2.3
Configuration Information on Communication Link Redundancy

Protection
N/A.
5
5.1
Counter and Alarm

Counter List
N/A.
5.2
Alarm List
Alarm Code
198005376
Description
Internal Port of B oard Is Down
Severity Level
Level 2
Caus e
1. The number of int ernal ports configured for this board and
the neighboring slots exceeds the number of physical slots.
2. Pin Healthy of this slot on the backplane is damaged.
Alarm Code
198066003
Description
Cont rol plane communication abnormal between board and

its home module
Severity Level
Level 2
Caus e
1. The board is configured in the database but not powered

on.
2. The board is powered on but the control plane link to its
home MP is broken.
3. The board is powered on and DIP switch ENUM is
switched on.
Alarm Code
198005378
Description
Board HW Error
39
Severity Level
Level 2
Caus e
1.The NE clock is abnormal.

2. The TDM connection chip is abnormal.
4. The board hardware has faults.
5. The back board hardware has faults.
Alarm Code
198005379
Description
Board Doesn't Exist/ CPU is in Reset Status for a Long Time
Severity Level
Level 2
Caus e
1. The board configured in the database is not on position.

2. An error occurs during board res etting.
3. Pin Healthy of this slot on the backplane is damaged.
Alarm Code
198066004
Description
Cont rol plane communication abnormal between MP and

OMP
Severity Level
Level 2
Caus e
1. The MP board is configured in the database but not

powered on.
2. The MP board is powered on but the cont rol plane link to
the OMP is broken.
3. The MP board is powered on and DIP switch ENUM is
switched on.
Alarm Code
198005381
Description
Input Clock Abnormal
Severity Level
Level 1
Caus e
The clock board runs exceptionally, and is not in the normal

working status. No clock reference is input to the c lock board
or the clock reference is exceptional.
Alarm Code
198005395
Description
External Port of Board Is Down
Severity Level
Level 2
Caus e
1.The network interface at the local end or the peer end is

faulty.
2.The work modes of network interfaces at both ends do not
match.
3. Physical connection has faults.
4.The GEMA C on this board does not support the 100M
speed of the negotiate mode.
40
Alarm Code
198005399
Description
DSP Resource Unavailable
Severity Level
Level 2
Caus e
1. DSP does not succeed in loading the version and can not
work normally.
2. DSP has HPI interface exception.
3. The media plane loop back fails.
4. Lan91 link is broken.
5. The link of 5328 port is broken.
6. DSP running halts.
7. The system initialization of the Tone resource DSP fails.
Different formats of tone resources or even no tone resource
exists on the board.
Alarm Code
198026127
Description
Clock reference source lost (Level 3 alarm)
Severity Level
Level 3
Caus e
Clock reference lost is detected for the first time.
Alarm Code
198026128
Description
Severity Level
Level 2
Caus e
Clock reference lost has been detected for more than 10

minutes.
Alarm Code
198026129
Description
Severity Level
Level 1
Caus e
Clock reference lost has been detected for more than 24

hours.
Alarm Code
198005122
Description
Communication Link between Active and Standby Boards Off
Severity Level
Level 3
Caus e
1. Acti ve and standby boards are configured on OMC, but no

standby board is physically plugged.
2. The communication link between active and standby boards is
off.
41
Alarm Code
198005189
Description
Notification of Active/Standby Changeover
Severity Level
Notification Level 4
Caus e
It is hardware system fault. The active and standby boards switch

over due to artificial operation.
Alarm Code
198005660
Description
Physical link down between active port and standby port
Severity Level
Level 4
Caus e
Physical Link Down Between Acti ve Port and Standby Port
Alarm Code
198066058
Description
The rate of error code in a mate link is high at GUIM
Severity Level
Level 3
Caus e
The rate of error code in a mate link is high. It is adventive
Glossary
AC
Access Control
APBI
ATM Process Board Interface
APBE
ATM Process Board Enhanced Version
APS
Automatic Protection Switching
ATM
Asynchronous Transfer Mode
BHCA
Busy-Hour Call Attempts
BME
Base station Multiplex Equipment
BM-SC
Broadcast Multicast Service Centre
BRS
Bearer Subsystem
CBC
Cell Broadcast Centre
THUB
Trunk HUB
CN
Core Network
DASF
Dynamic Radio Bearer Cont rol
42
DBS
Data Base Subsystem
DDN
Digital Data Network
43

ZTE UMTS RAN Equipment Redundancy Feature Guide

Uploaded by

Copyright:

Available Formats

ZTE UMTS RAN Equipment Redundancy Feature Guide

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ZTE UMTS RAN Equipment Redundancy Feature Guide

Uploaded by

Copyright:

Available Formats

Operator Logo

RAN Equipment Redundancy

RAN Equipment Redundancy Feature Guide

RAN Equipment Redundancy Feature Guide

2010 ZTE Corporation. All rights reserved.

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Functional Attribute ............................................................................................ 1

Parameter Description ..................................................................................... 36

Counter and Alarm............................................................................................ 39

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Figure 3-29 FE Routing Interconnection Between UIM and UIM.......................................... 32

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

RAN Equipment Redundancy

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Board Redundancy Protection of RAN

Brief Introduction to Board Redundancy Protection

1+1 backup (B OARD_BA CKUP _MODE_MAS TE RSLAVE)

Hybrid 1+1 backup (BOA RD_BACK UP_MODE_MAS TE RSLAVE)

1:1 backup (B OARD_BA CKUP _MODE_1TO1)

Load sharing (B OARD_BA CKUP_MODE_MULTI_ONE )

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Principle of Hybrid 1+ 1 backup:

As it is shown in : The coming data stream A and B are

Principle of Hy brid 1+1 Backup

A,B combine into C by

General Process of Board Switching

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Reason for Board Switching

Abnormal switching due to reset of active board

Abnormality of network interfac e of the control plane

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Switching Priority Control

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Active/Standby Switching Process

ZTE Confidential Proprietary

2010 ZTE Corporation. All rights reserved.

RAN Equipment Redundancy Feature Guide

Board Fault Detection

Partner Board In-Position Detection

Active/Standby Interlock Circuit

ZTE Confidential Proprietary