lte transport network configuration

Transport Network Configuration

USER DESCRIPTION

39/1553-HSC 105 50/1 Uen S

Copyright

© Ericsson AB 2009–2012. All rights reserved. No part of this document may bereproduced in any form without the written permission of the copyright owner.

Disclaimer

The contents of this document are subject to revision without notice due tocontinued progress in methodology, design and manufacturing. Ericsson shallhave no liability for any error or damage of any kind resulting from the useof this document.

Trademark List

All trademarks mentioned herein are the property of their respective owners.These are shown in the document Trademark Information.

39/1553-HSC 105 50/1 Uen S | 2012-11-21

Contents

Contents

1 Introduction 1

2 LTE RAN Transport Overview 3

3 Layer 2 Networks 5

4 Layer 3 Networks 7

5 O&M IP Networks 9

6 Managed Object Model 11

7 QoS 13

7.1 Layer 3 QoS, DSCP 13

7.2 Layer 2 QoS, p-bit 16

7.3 Egress Mapping Process 16

8 MTU 19

9 SCTP Configuration 21

10 IP Addressing 25

10.1 Layer 2 Considerations 25

10.2 Layer 3 Considerations 25

10.3 O&M Considerations 26

11 S1 and X2 Configuration 27

12 Mul Configuration 29

13 Configuration Examples 31

13.1 RPU Configuration 31

13.2 DU Configuration 32

13.3 S1 Interface Configuration 33

13.4 X2 Interface Configuration 37

13.5 Mul Interface Configuration 37

13.6 Network Synchronization Configuration 39

39/1553-HSC 105 50/1 Uen S | 2012-11-21


39/1553-HSC 105 50/1 Uen S | 2012-11-21

Introduction

1 Introduction

This document describes configuration aspects of the LTE RAN transportnetwork. The relevant Managed Object (MO) and attribute values are listedalong with example configurations.

This document details the following information related to configuring theInternet Protocol (IP) transport network:

• S1 and X2 configuration

• IP addressing

• Virtual Local Area Network (VLAN) configuration

• Signaling or Stream Control Transmission Protocol (SCTP) configuration

• Quality of Service (QoS) configuration

• Mul configuration

This document describes the configuration of transport-network-relatedparameters for the RBS, Ericsson's implementation of the Evolved UTRANNodeB (eNodeB) specified by 3GPP.

The RBS Autointegration feature reduces the amount of data that must beentered on site when configuring an RBS. A description of how to configure IPtransport network functions in an RBS using autointegration can be found inRBS Autointegration.

The Integrated IP Security feature provides a security solution for S1 and X2user and control plane traffic, including protection of traffic for Synchronizationover IP (SoIP). A description of how to configure transport networks withsecurity can be found in IP Security.

The IPv6 feature provides functionality to use IPv6 in LTE RAN transportnetworks. A description of how to configure transport networks with IPv6 canbe found in IPv6.

A description of how to migrate from a transport network with IPv4 to a transportnetwork with IPv6 can be found in IP Transport.

139/1553-HSC 105 50/1 Uen S | 2012-11-21


2 39/1553-HSC 105 50/1 Uen S | 2012-11-21

LTE RAN Transport Overview

2 LTE RAN Transport Overview

This section provides a brief overview of the LTE RAN architecture and theIP-based transport network. Further information can be found in TransportNetwork Functions.

The LTE RAN transport network provides connectivity between the RBSs andthe RBS and the core network using the X2 and S1 interfaces. Figure 1 showsthe LTE RAN architecture.

L0000076A

RBS

Mul

S1 X2

Evolved Packet Core

LTE RAN S1 CP S1 UP Mul

RBS

MME

OSS-RC

Mul S1 X2

SGW

RBS

MulS1

X2

Typically acommonphysicalconnection

IP/Ethernet transport

Figure 1 LTE RAN Architecture

The basic transport network options for transporting S1 and X2 traffic are Layer2 and Layer 3 topologies. A mix of basic topologies can be used to build thetransport network.

339/1553-HSC 105 50/1 Uen S | 2012-11-21


4 39/1553-HSC 105 50/1 Uen S | 2012-11-21

Layer 2 Networks

3 Layer 2 Networks

In a Layer 2 network, all nodes are connected to one another without the use ofLayer 3 routers. The Layer 2 technology predominantly used in LTE networks isEthernet. In an Ethernet network, all nodes share the same Ethernet broadcastdomain in this topology variant, so the broadcast domain can be quite large.

One way to reduce the broadcast domain size is to configure separate VLANsfor a group of RBSs. Broadcast domain size should be considered whendetermining the number of RBSs in a Layer 2 topology. Figure 2 shows anexample of a Layer 2 network.

L0000241A

Switch

RBS

DULayer 2TransportNetwork

S1_X2_SoIP subnet /m

Mul subnet /n

Figure 2 RBS Connectivity to Layer 2 Transport Network

Ethernet Services, such as E-Line, E-LAN, and E-Tree services as defined byMetro Ethernet Forum (MEF), can also be used to provide Layer 2 connectivitybetween LTE sites without the need for the network to be a pure Ethernetnetwork.

539/1553-HSC 105 50/1 Uen S | 2012-11-21


6 39/1553-HSC 105 50/1 Uen S | 2012-11-21

Layer 3 Networks

4 Layer 3 Networks

Layer 3 topologies are another option for LTE RAN transport networks. Layer 3routers are used in this topology variant, which allows the use of a separatesub-network for each RBS. The transport network used should support Layer3 routing between the RBS and the core network.

A Layer 3 router is found at each site or at a hub site servicing a number ofRBSs. VLAN tagging may be used to separate S1, X2, and SoIP traffic ontoone VLAN and Operation and Maintenance (O&M) traffic onto a separateVLAN. If VLANs are used, VLAN separation must be supported over Layer 3hops in the transport network.

A number of RBSs can be grouped in the same, larger sub-network. Thisconfiguration has an impact on the VLAN configuration and sub-network size.Figure 3 shows an example of a Layer 3 network.

L0000242A

Router

RBS

DULayer 3TransportNetwork

S1_X2_SoIP subnet /m

Mul subnet /n

Figure 3 RBS Connectivity to Layer 3 Transport Network

739/1553-HSC 105 50/1 Uen S | 2012-11-21


8 39/1553-HSC 105 50/1 Uen S | 2012-11-21

O&M IP Networks

5 O&M IP Networks

O&M traffic is carried over the Mul interface that shares a common physicallink with S1 and X2 traffic in the RBS. A separate VLAN can be configured forO&M. The O&M traffic can be routed to the Operation and Support System -Radio and Core (OSS-RC).

939/1553-HSC 105 50/1 Uen S | 2012-11-21


10 39/1553-HSC 105 50/1 Uen S | 2012-11-21

Managed Object Model

6 Managed Object Model

The RBS is implemented using the Ericsson Connectivity Packet Platform(CPP). The Managed Object Model (MOM) used in the RBS shares some MOsused in Wideband Code Division Multiple Access (WCDMA), but some MOsare not relevant for LTE. A list of MOs used for IP transport can be found inIP Transport. Full details of the MOM structure can be found in the ManagedObject Model (MOM) User Guide.

1139/1553-HSC 105 50/1 Uen S | 2012-11-21


12 39/1553-HSC 105 50/1 Uen S | 2012-11-21

QoS

7 QoS

QoS configuration is described in several documents. QoS for data transportover the air interface is described in Radio Bearer and Quality of Service. Thissection describes QoS configuration in the RBS for S1, X2 and Mul traffic.

QoS for the LTE RAN is based on the use of Differentiated Services CodePoint (DSCP) and priority bits (p-bits) that enable the intermediate transportnetwork to prioritize traffic. This is especially important when the capacity in thetransport network is limited.

7.1 Layer 3 QoS, DSCP

Egress IP packets from the RBS are tagged with DSCP values. A DSCP valuedefines a Per-Hop Behavior (PHB) that will be given to an IP packet, that is, aset of packet forwarding properties when the packet is transported through anetwork.

User plane traffic is transported over the S1 and X2 interfaces throughE-UTRAN Radio Access Bearers (E-RABs). Each E-RAB has a QoS ClassIdentifier (QCI) assigned to it when the E-RAB is set up with S1-AP or X2-APsignaling. The QCI defines QoS characteristics for data transported overthe E-RAB. It is possible to define DSCP values to use for each QCI byconfiguration. This is done using the QciProfilePredefined MO. The QCIvalue is then specified using the qci attribute, and the corresponding DSCPvalue is specified using the dscp attribute.

There are no QCI values for control plane traffic. Instead DSCP values to usecan be configured using the MOs for configuration of the control plane. DSCPvalues to use for O&M traffic over Mul can be configured using the MOs forconfiguration of O&M communication.

The recommended DSCP values for different traffic types are listed in Table 1.

Note: The recommended DSCP values may differ from the default values inthe used MOs.

Table 1 Recommended QCI, PHB and DSCP Values

Traffic Type PHB DSCP(1)

Network Synchronization LU 54 (46)

Routing, Network Control CS6 48

QCI1 - GBR Conversational Voice EF 46

QCI2 - GBR Conversational Voice (LiveStreaming)

AF42 36

1339/1553-HSC 105 50/1 Uen S | 2012-11-21


Traffic Type PHB DSCP(1)

QCI3 - GBR Real Time Gaming AF41 34

QCI4 - GBR Non-Conversational Video(Buffered Streaming)

AF43 38

QCI5 - IMS Signaling CS5 40

QCI6 - Non-GBR TCP Specific Services AF31 26

QCI7 - Non-GBR Voice/Video/InteractiveGaming

AF11 10

QCI8 - Non-GBR TCP Premium Bearer AF12 12

QCI9 - Non-GBR TCP Default Bearer AF13 14

S1AP/X2AP - Inter-Node Signaling CS5 40

O&M Access (including O&M Bulk Data)(2) CS2 16

(1) Default decimal values are shown in parentheses.(2) Currently, O&M Bulk Data cannot be separated from general O&M Access. Otherwise,O&M Bulk Data should be given lower priority.

MOs for configuring DSCP values for control plane traffic include the following:

• SCTP association endpoint configured using the dscplabel attribute inthe ENodeBFunction MO

• IP sync configured using the ntpDscp attribute in the IpAccessHostEtMO

All other O&M communication (for example, all performance managementdata) coming from the RBS transport interface applies the DSCP value setin the dscp attribute in the Ip MO class. Complete details can be found inManaged Object Model RBS.

The DSCP values set in outgoing IP packets are used for mapping to egressqueues and to Layer 2 p-bit values in the RBS. They may also be used by anyLayer 3 equipment in the intermediate transport network.

Figure 4 shows mapping of QoS values.

14 39/1553-HSC 105 50/1 Uen S | 2012-11-21

QoS

UserPlane

ControlPlaneand OAM

L0000361B

For use in the egress interface (backhaul) For use in thebackhaul

Configurable Configurable Configurable Configurable

QCI DSCP

NTPSCTPOAM

123456789

12345678

DSCP p-bitPrio

Highest

Strictprioritysche-duling

Lowest

Queue

Not

vis

ible

to th

e op

erat

orp-bit is not relevantfor egress handling

Figure 4 Mapping of QoS Values

Ericsson does not foresee issues with recommended mapping due to capacitylimitations when the transport connectivity is over-dimensioned. However, it isimportant to ensure that QoS marking is implemented and that the transportnetwork supports separation and prioritization when capacity bottlenecks occur.

QCI-to-DSCP mapping in the RBS can be configured. DSCP-to-egress queuemapping is fixed.

The mapping of DSCP values to egress queues on the Digital Unit (DU) is aslisted in Table 2.

Configuration of DSCP-to-egress queue mapping is possible using the EgressIP Traffic Shaping feature as described in Egress IP Traffic Shaping.

Table 2 DSCP-to-DU Egress Queue Mapping

DSCP Range DU Egress Queue

56-63 1 (highest priority)

48-55 2

40-47 3

32-39 4

24-31 5

16-23 6

8-15 7

0-7 8 (lowest priority)

1539/1553-HSC 105 50/1 Uen S | 2012-11-21


7.2 Layer 2 QoS, p-bit

DSCP values are mapped to p-bit values for all outgoing IP packets inaccordance with IEEE 802.1q. Mapping is configurable and is done in theGigaBitEthernet MO using the setDscpPbit action, which changes thedscpPbitMap attribute. Recommended mappings between DSCP values andp-bits are shown in Table 3. DSCP values for which no mapping is configuredwill get p-bit 0.

It is also possible to configure the p-bit values that shall be used for the layer twoARP and gratuitous ARP messages. This is done using the configPbitArpattribute in the GigabitEthernet MO.

Table 3 Recommended DSCP to p-bit Mapping

DSCP p-bit

0 1

10, 12, 14 2

18, 20, 22 3

8, 16, 26, 28 4

34, 36, 38, 46 5

24, 32, 40, 48 6

51, 54 7

7.3 Egress Mapping Process

A marker sets DSCP values in outgoing IP packets. For user plane traffic(E-RABs), the marker uses QCI values to select DSCP values as described inRadio Bearer and Quality of Service. For control plane and O&M traffic, themarker uses the DSCP values defined by configuration. The classifier sorts themarked packets into egress queues (up to eight queues), based on the DSCPvalues. The queues are served with a strict priority scheme. Queue 1 is alwaysserved first if a packet is waiting to be sent. Weighted priority is possible usingthe feature Egress IP Traffic Shaping as described in Egress IP Traffic Shaping.

The p-bit is set after queuing. The setting is independent of the queues fromwhich the packet came. The p-bit setting is based on the DSCP value asspecified in the configuration process explained in Section 7.2 on page 16. Thep-bit does not affect packet handling in the RBS. The p-bit is meant to be usedonly by the backhaul nodes.

The flow from QCI to p-bit is shown in Figure 5.

16 39/1553-HSC 105 50/1 Uen S | 2012-11-21

QoS

UE request E-RAB from MME

QCI from E-RAB set up by MME

DSCPmarkingbased on QCI

Classificationto queuebased onDSCP

p-bit may beadded, based on DSCP1

User data flows

MME assigned QCI or configured DSCP

Bearer-level traffic Aggregation Backhaul network

Control plane andOAM flows, DSCPconfigured via MOs

QCI 1

QCI 9

Lowest prio

Marker

Classifier

Marker

DSCP=36 Queue 1

Queue 2

Queue 3

Queue 4Egress

Load max ½ of downlink,i.e. queues 1–N neverdrop packets

p-bit marking

Queue N

Strict prio

QCI12

9

DSCP4636

14

Queue34

7

p-bit55

2

QCI table example

2221

4

Highest prio

L0000362B

Figure 5 Egress QoS Process in RBS

1739/1553-HSC 105 50/1 Uen S | 2012-11-21


18 39/1553-HSC 105 50/1 Uen S | 2012-11-21

MTU

8 MTU

The size of the largest Protocol Data Unit (PDU) that can be sent over atransport protocol is called the Maximum Transmission Unit (MTU). Figure 6shows MTUs for different protocol layers that are used for user data transportthrough the LTE RAN. There is an end-user IP packet MTU for IP packets thatare transported between the UE and the Application Server. There are alsoIP packet MTUs for IP transport between the PDN-GW and the S-GW, andbetween the S-GW and the eNodeB. Further, there are Ethernet frame MTUsfor Ethernet transport between the PDN-GW and the S-GW, and between theS-GW and the eNodeB.

L0000523A

L1

PDCP

TCP/UDP Appl

L1

PDCP GTP-U UDP

GTP-U UDP

GTP-U UDP

TCP/UDP Appl

GTP-U

UDP

L1 L1

RLC MAC

RLC MAC

IP IP IP

IP IP IP IP Eth/L1 Eth/L1 Eth/L1 Eth/L1

UE eNodeB S-GW PDN-GW Application Server

Eth Eth

Air Interface

Mobile Backhaul

Packet Core

Packet Core /

Internet

End-user IP packet MTU

Transport IP packet MTU Transport Eth Frame MTU

Figure 6 Transport Network MTU

To send IP packets that exceed the Ethernet MTU, the IP layer must firstfragment the packets to align with the Ethernet MTU. Fragmentation of packetshas a negative impact on network performance since it introduces delays. Thisbecause fragments are buffered by the receiving node until all fragments havearrived, after which the fragments are arranged in order and reassembled.These delays will impact the maximum throughput that can be achieved aswell as potentially impacting the quality of experience for delay sensitiveapplications.

The Ethernet MTU depends on the type of frames that are used by Layer 2.For Ethernet 802.3 frames the MTU is 1492 bytes. For Ethernet v2 the MTUis 1500 bytes. For Ethernet Jumbo frames eNodeB supports the MTU 2000bytes. Jumbo frame support in the LTE RAN nodes and transport network isrecommended as a long-term solution as a means to avoid fragmentation.For jumbo frames to work, support is required from all LTE RAN nodes andthe transport network.

The Ethernet MTU size for an IP interface of a DU can be configured via theattribute mtu in the MO class IpInterface. The actual MTU used canbe lower than the configured MTU dependent on what frame types the DUhardware supports.

1939/1553-HSC 105 50/1 Uen S | 2012-11-21


20 39/1553-HSC 105 50/1 Uen S | 2012-11-21

SCTP Configuration

9 SCTP Configuration

Some of the SCTP parameters must be configured to ensure correct operationof the SCTP layer. The recommended values for these parameters aredescribed in this section.

The recommended values occasionally differs from the default values becausethe MOM is shared with WCDMA nodes such as the Radio Network Controller(RNC) and RBS (NodeB). The recommended values are based on a SCTPassociation between the eNodeB and the MME, where the eNodeB acts as asingle homed SCTP endpoint and the MME acts as a either a single homedendpoint or a multi-homed endpoint.

Some LTE networks have legacy requirements from a WCDMA network,requiring aligned settings of the recommended values to achieve the samecharacteristics in both networks. For this case, a specific set of recommendedvalues is specified.

The operator can define their own recommendations. This may be required tofulfill operator specific network characteristics.

The IP address for SCTP is configured using the ipaddress attribute in theIpAccessHostEt MO as control plane shares the same IP address as userplane traffic.

Table 4 describes the recommended SCTP configuration.

Table 4 Recommended SCTP Configuration

Attribute DefaultValue

RecommendedValue ineNodeB

Recommended Valuein eNodeB ConsideringLegacy WCDMANetwork

Comment

associationMaxRtx 8 20 12 Unit: One attempt

bundlingActivated 1 1 1 0: DISABLED

1: ENABLED

bundlingTimer 10 0 10 Unit: 1 millisecond

heartbeatInterval 30 2 1 Unit: 1 second

This attribute must be greaterthan or equal to the value of theheartbeatPathProbingIntervalattribute.

A short heart beat interval gives rapidfault detection, but network load shouldbe considered.

heartbeatMaxBurst 0 1 1 If set to 0, path probing is not used.

2139/1553-HSC 105 50/1 Uen S | 2012-11-21






Comment

heartbeatPathProbingInterval

500 100 70 Unit: 0.01 second

This attribute must be less thanor equal to the value of theheartbeatInterval attribute.

This attribute must be greater than orequal to the maximumRto attribute.

heartbeatStatus true true true true: enable heartbeats

false: disable heartbeats

initialAdRecWin 32768 16384 16384 Unit: 1 byte

initialRto 20 20 20 Unit: 0.01 second

maximumRto 40 40 40 Unit: 0.01 second

This attribute must be less thanor equal to the value of theheartbeatPathProbingIntervalattribute.

maxIncomingStream 17 2 2 This is the allowed maximum number ofincoming streams for an association.

maxOutgoingStream 17 2 2 This is the allowed maximum number ofoutgoing streams for an association.

maxUserDataSize 1480 556 556 Unit: 1 byte

Prevents situations where fragmentationof IP/SCTP packets will occur in thenetwork (maxUserDataSize and IPheader in total will not exceed theMTU(1)) size. A higher value (up to1480 bytes without IPsec protection and1416 bytes with IPsec protection) canbe used if the MTU size used in thenetwork is known and it can be ensuredthat fragmentation is avoided.

mBuffer 256 64 64 Unit: 1 kilobyte.

This attribute must be greater than thenThreshold attribute.

minimumRto 10 10 10 Unit: 0.01 second

numberOfAssociations

- 320 320 Assuming an MME(2) pool size of 64(maximum size) for S1 connectivity andup to 256 RBS neighbors that requireX2 connectivity.

nPercentage 85 85 85 Unit: 1%

nThreshold 192 48 48 Unit: 1 kilobyte

pathMaxRtx 4 10 for multi-homed MME

20 for single-homed MME

6 for multi-homed MME

12 for single-homedMME

Unit: One attempt

This setting is specified for either amulti-homed host or a single-homedhost in the MME. However for X2, thehost in the other RBS is single-homed,which has the effect of reducing theinterruption tolerance time by one half.It is deemed that the S1 connection ismore important, hence the value of thissetting.

22 39/1553-HSC 105 50/1 Uen S | 2012-11-21

SCTP Configuration





Comment

pathSelection 0 1 1 0: SCTP_PATHS

1: IP_PATHS

potentiallyFailedMaxRtx

4 0 0 Setting this attribute to 0, together withsetting the attribute pathSelection to1 means that a new path will be selectedafter the first failed retransmission.

tSack 4 1 4 Unit: 0.01 second

(1) Maximum Transfer Unit(2) Mobility Management Entity

2339/1553-HSC 105 50/1 Uen S | 2012-11-21


24 39/1553-HSC 105 50/1 Uen S | 2012-11-21

IP Addressing

10 IP AddressingThe IP address plan is based on a number of factors including the following:

• Number of nodes in the sub-network

• VLAN separation

• Future expansion

IP addressing in a transport network based on IPv4 is described in Table 5.

Information about IP addressing in a transport network based on IPv6 can befound in IP Transport.

Table 5 RBS IP Addressing

Traffic TypeNumber of IP

Addresses Comment

User plane

Control plane

Network synchronization,if applicable

1 IP address for networksynchronization only appliesif the NTP-based solution isused.

O&M 1 At least one IP addressmust be defined for O&Mconnectivity. Additional IPaddresses can be allocatedfor Site LAN.

Default Gateway 1-3

10.1 Layer 2 Considerations

In a Layer 2 topology, large sub-networks are typically configured. Thesub-network size depends on the number of RBSs in the same sub-networkor VLAN. Sub-network and broadcast addresses are shared by all RBSs in asub-network. Similarly, the same default gateways can be used for all RBSs inthe sub-network. With a large sub-network size the amount of broadcast trafficin the sub-network becomes large, which should be considered.

10.2 Layer 3 Considerations

Layer 3 topologies allow a separate sub-network to be allocated for each nodewhen the node connects to a Layer 3 device, such as a router. This allowseach RBS to have a dedicated sub-network. Each RBS then requires a uniquesub-network and broadcast address. To keep down the number of addressesused it is recommended to let several RBSs share the same subnetwork.

2539/1553-HSC 105 50/1 Uen S | 2012-11-21


10.3 O&M Considerations

When a separate VLAN is used for O&M, a sub-network is also required for theO&M related IP addresses for the RBS.

26 39/1553-HSC 105 50/1 Uen S | 2012-11-21

S1 and X2 Configuration

11 S1 and X2 Configuration

This section provides a general guideline about how to configure the interfacesfor S1 and X2. It describes configuration of parts that are common for theS1 and X2 interfaces.

The MOs that must be defined are listed in Table 6.

Table 6 S1 and X2 MOs

MO Description

ExchangeTerminalIp One instance is defined for each DU.

GigaBitEthernet One instance is defined for each DU.

IpInterface One instance is defined for each sub-network or VLAN. Whenusing VLANs, two IpInterfaces MOs are required: onefor S1/X2 and SoIP and one for O&M.

IpAccessHostEt One instance is defined for each DU. With IPsec, oneadditional instance must be defined as described in IPSecurity.

IpAccessSctp One instance is defined for each RBS.

Sctp One instance of the Sctp MO is defined for each RBS.

Figure 7 shows the relationship between MOs.

L0000238A

MO ClassPlugInUnit

MO ClassExchange TerminalIp

MO ClassGigabitEthernet

MO ClassIpInterface

MO ClassIpAccessHostEt

MO ClassIpAccessSctp

MO ClassSctp

Figure 7 Configuration Overview for S1 and X2 Interfaces

Examples of complete configurations of the S1 and X2 interfaces can be foundin Section 13.3 on page 33 and Section 13.4 on page 37 respectively

2739/1553-HSC 105 50/1 Uen S | 2012-11-21


Information on configuration of S1 and X2 with IPsec can be found in IPSecurity.

28 39/1553-HSC 105 50/1 Uen S | 2012-11-21

Mul Configuration

12 Mul Configuration

This section contains general guidelines about configuring the Mul interface.

The ExchangeTerminalIp and GigabitEthernet Mul MO instances arecommon with those defined for S1 and X2 in Section 11 on page 27.

In addition, the MOs listed in are also defined for Mul.

Table 7 MOs for Mul

MO Description

IpInterface Mul can be configured using a different VLAN. In this case,another IpInterface MO, separate from that defined for S1and X2, must be defined.

IpOam One instance is defined for each RBS.

Ip One instance is defined for each RBS.

IpHostLink One instance is defined for each RBS.

Dhcp One instance is defined for each RBS.

A configuration example can be found in Section 13.5 on page 37.

More information on O&M can be found in IP O&M.

Information on configuration of Mul with IPsec can be found in IP Security.

2939/1553-HSC 105 50/1 Uen S | 2012-11-21


30 39/1553-HSC 105 50/1 Uen S | 2012-11-21

Configuration Examples

13 Configuration Examples

This section contains examples for configuring transport-related MOs. Onlyattributes without default values and attributes that should have valuesdeviating from the default value are included in the examples. Other attributescan be left as default. Setting the attributes must take into account the currenttransport network configuration. A complete list of attributes for the relevantMOs can be found in the MOM.

The configuration examples are described in this section using figures andtables. The figures show the relevant MO instances and their relations.

The relations can be one of the following types:

• Containment (parent child)

• MO reference (attribute value)

The tables in this section provide the parent relation and attributes (includingthose containing references to other MOs) for each MO instance.

13.1 RPU Configuration

A Reliable Program Uniter (RPU) is an entity providing a common referencingmethod for an active and a standby pair of a reliable programs. A reliableprogram is one that can move to and be restarted on another processor, ifthe current processor fails. For each reliable program, one RPU MO instanceexists for each active and standby pair of Main Processors (MPs) that thereliable program can be run on.

MO instances representing certain transport layer protocol terminations musthave an rpuId attribute that refers to an RPU MO. The RPU MO defines thepair of processors on which the protocol termination is configured to operate.All RPU MOs are configured during the initial software configuration and theirnames must be known at the later configuration of transport layer protocolterminations.

The RPU used in the RBS is described in Table 8.

3139/1553-HSC 105 50/1 Uen S | 2012-11-21


Table 8 Example of Configuration of RPU in RBS

Position inMO createsequence.

<Type> Parent:<Type>=<name> <attribute>=<value> Comment

ReliableProgramUniterId=sctp_host

Load module containing the SCTPprotocol software

reliableProgramLabel=sctp_host

1 ReliableProgramUniter

Parent: SwManagement=1

admActiveSlot=<MO Reference> Slot for MP with active protocol forSCTP

13.2 DU Configuration

Table 9 provides a generic example of relevant MOs and attributes whenconfiguring the DU.

Table 9 Example of Configuration of DU in RBS

Position in MO createsequence.

<Type> Parent:<Type>=<name>

<attribute>=<value> Comment

1 ExchangeTerminalIp

Parent: PlugInUnit

ExchangeTerminalIpId=1

2 GigaBitEthernet

Parent: ExchangeTerminalIp in Position 1

GigaBitEthernetId=1

IpInterfaceId=1

defaultRouter0=<IPAddress>

In accordance with the IPaddress plan

defaultRouter1=<IPAddress>

Used for redundant router

mtu =1500 If Ethernet Jumbo frames aresupported in the network mtucan be set to 2000.

networkPrefixLength=<n>

Sub-network mask; validvalues 20–30

ownIpAddressActive=<IPAddress>

Only configured if RPS used

rps=true or false Router Path Supervisionenabled or disabled

vid=<n> VLAN ID

Valid IDs are 0-4094; valid ID0 only has priority tag, othershave priority tag and VLAN ID.

3 IpInterface

Parent: MO in Position 2

vLan=true or false VLAN tagging enabled ordisabled

32 39/1553-HSC 105 50/1 Uen S | 2012-11-21


13.3 S1 Interface Configuration

This section provides an example of MOs and relevant attributes that mustbe configured for the S1 interface. An overview of S1 configuration is shownin Figure 8.

L0000239A

ENodeBFunction

8

Sctp

7

IpAccessSctp

6

IpAccessHostEt5

PlugInUnit

1

ExchangeTerminalIp

2

GigabitEthernet

3

IpInterface

4

Figure 8 S1 Interface Configuration Overview

Table 10 provides an example the S1 interface configuration in the RBS.

Table 10 Example of S1 Configuration in RBS

Number inFigure 8 /position inMO createsequence. <Type> Parent:<Type>=<name> <attribute>=<value> Comment

1 PlugInUnit PlugInUnitId=1

2 ExchangeTerminalIp

Parent: PlugInUnit in Position 1

ExchangeTerminalIpId=1

GigaBitEthernetId=1

dscpPbitMap Set as specified in Table 3.

3 GigaBitEthernet

Parent: ExchangeTerminalIp inPosition 2

configPbitArp Use default value.

3339/1553-HSC 105 50/1 Uen S | 2012-11-21




IpInterfaceId=1

defaultRouter0=<IP Address> In accordance with the IP addressplan

defaultRouter1=<IP Address> Used for redundant router

mtu =1500 If Ethernet Jumbo frames aresupported in the network mtu canbe set to 2000.

networkPrefixLength=<n> Sub-network mask; valid values20–30

ownIpAddressActive=<IPAddress>

Only configured if RPS used

rps=true or false RPS enabled/disabled

vid=1100 Suggested VLAN ID for SoIP, S1,and X2

4 IpInterface

Parent: GigabitEthernet inposition 3

vLan=true VLAN tagging enabled

IpAccessHostEtId=1 One instance for each RBS

ipAddress=<IP Address> IP host address

ipInterfaceMoRef=<MOReference>

MO in position 4

ntpDscp=54 Only applicable if SoIP used

5 IpAccessHostEt

Parent: IpSystem=1

ntpServerMode=DISABLED RBS is Synch client when SoIPused

IpAccessSctp=1 One instance for each RBS

ipAccessHostEtRef1=<MOReference>

MO in position 5

6 IpAccessSctp

Parent: IpSystem=1

ipAccessHostEtRef2=<MOReference>

Not used as only one DU in the RBS

34 39/1553-HSC 105 50/1 Uen S | 2012-11-21




SctpId=1 One instance for each RBS

ipAccessSctpRef=<MOReference>

MO in position 6

rpuId=<MO Reference> Reference to ReliableProgramUniter MO for SCTP (see Table 8)

associationMaxRtx=20 Maximum number of consecutiveretransmissions to a remote peer(on all the destination IP addressesof the peer)

bundlingActivated=ENABLED

bundlingTimer=0 Value 0 means that SCTP bundlesand sends available data directly

heartbeatInterval=2 Amount of time added to the RTO(1)

of a specific address when settingup the period of time betweensending heartbeats

heartbeatMaxBurst=1 Maximum number of heartbeats thatcan be sent at the same time duringthe path probing procedure

heartbeatPathProbingInterval=100

Interval between consecutiveheartbeats during the path probing

heartbeatStatus=true Enables and disables hearbeatsupervision for SCTP associations.

initialAdRecWin=16384 Initial value of the advertisedreceiver window credit

initialRto=20 Initial value that the RTO takesprior to the first measurement of theRTT(2)

minimumRto=10 (default value) Minimum value for RTO

maximumRto=40 (default value) Maximum value for the RTO

maxUserDataSize=556 Maximum number of bytes of an IPdatagram that can be transferred ina single unit over a specific path inan IP network

If an IP datagram exceeds thePMTU(3), it is either fragmented ordropped.

7 Sctp

Parent: TransportNetwork=1

mBuffer=64 Sets the buffer size used for storinguser data pending to be sent orretransmitted in an association

3539/1553-HSC 105 50/1 Uen S | 2012-11-21




nPercentage=85 This attribute is set as a percentageof the attribute nThreshold. Acongestion ceased indication is sentto the SCTP user when the use ofthe SCTP send buffer drops belownPercentage.

nThreshold=48 Sets the threshold value used to askthe SCTP user to stop the deliveryof data on an association

A congestion indication is sent tothe user when the SCTP send bufferuse is above the value of attributenThreshold.

numberOfAssociations=320 Maximum number of associationsthat can be handled by this SctpMO

pathMaxRtx=10 Maximum number of consecutiveretransmissions to a remotetransport address

pathSelection=IP_PATHS Type of path selection algorithm

potentiallyFailedMaxrtx=0 Maximum number ofretransmissions over a pathbefore it is considered as failed.

7 Continued Sctp

Parent: TransportNetwork=1

tSack=1 Sets the time delay for sendingSelective Acknowledgement(SACK)

ENodeBFunctionId=1 One instance for each RBS

dnsLookupOnTai=ON (defaultvalue)

dnsLookupTimer=0 (defaultvalue)

If set to 0, no DNS(4) lookupperiodically

A DNS lookup can be done manuallyusing the updateMMEConnectionaction.

dscpLabel=40 Applies to SCTP transport only

eNBId=<RBS ID>

8 ENodeBFunction

sctpRef=<MO Reference> MO in position 7

Reference to the Sctp instance thatis used for S1 and X2 connections.

TermPointToMmeId=<n> The automatic MME discoveryfunction uses naming convention asdescribed in RBS Autointegrationwhen creating this MO.

ipAddress1=<IP Address>

- TermPointToMme

Parent: ENodeBFunction inposition 8

ipAddress2=<IP Address>

(1) Retransmission Time Out(2) Round-Trip Time(3) Path Maximum Transfer Unit(4) Domain Name System

36 39/1553-HSC 105 50/1 Uen S | 2012-11-21


A description of how to configure an S1 interface with security can be found inIP Security.

13.4 X2 Interface Configuration

The X2 interface configuration is common with that of the S1 interface. Table10 provides details for the common MO instances. Furthermore, the MOs listedin Table 11 must also be configured for an X2 interface.

Table 11 Example of X2 Configuration in RBS

Positionin MOcreatesequence.


1 EUtranNetwork

Parent:ENodeBFunction inposition 8 in Table 10

EUtranNetworkId=1

ExternalENodeBFunctionId=<n>

Up to 256 MO instances foreachEUtranNetwork.

A maximum of 256 neighbor RBSsand X2 connections are possible.

2 ExternalENodeBFunction

Parent: EUtranNetwork in position1

eNBId=<n> ID of other RBS within the PLMN(1)

TermPointToENBId=13 TermPointToENB

Parent: ExternalENodeBFunction in position 2

ipAddress=<IP Address> IP address to the neighboring RBS

(1) Public Land Mobile Network

A description of how to configure an X2 interface with security can be found inIP Security.

13.5 Mul Interface Configuration

An overview of Mul interface configuration is shown in Figure 9.

3739/1553-HSC 105 50/1 Uen S | 2012-11-21


L0000412B

IpOam5

Ip6

IpHostLink7

Dhcp8

PlugInUnit1

ExchangeTerminalIp2

GigabitEthernet3

IpInterface4

Figure 9 Mul Interface Configuration Overview

O&M uses the same physical interface as S1 and X2 traffic, consequently theconfiguration of certain MOs are common. Table 10 provides details for thecommon MO instances.Table 12 provides an example of Mul configuration.

Table 12 Example of Mul (O&M) Configuration in RBS

Number inFigure 9 /position inMO createsequence.


1 PlugInUnit Same MO in position 1 in Table 10

2 ExchangeTerminalIp Same MO in position 2 in Table 10

3 GigabitEthernet Same MO in position 3 in Table 10

IpInterfaceId=2

defaultRouter0=<IP Address>

mtu=1500

networkPrefixLength=<n>

vid=1900 Suggested VLAN ID for O&M

4 IpInterface

Parent: GigabitEthernet MO inposition 3

vLan=true

5 IpOam IpOamId=1

38 39/1553-HSC 105 50/1 Uen S | 2012-11-21


Table 12 Example of Mul (O&M) Configuration in RBS

Number inFigure 9 /position inMO createsequence.


IpId=1

defDomainName=<default domainname>

Default domain name

dnsServerAddresses=<IPAddresses>

List of IP addresses (maximum ofthree) of DNS servers

dscp=16 Applies to O&M transport only

isDefDomainName=true Specifies whether the defaultdomain name is present

nodeInterfaceName=lh0 Specifies the link that defines theOAM IP address for the node

Before the node hasbeen configured, thenodeInterfaceName attributehas the value ‘‘le0’’ (defaultvalue). The IP address of theIpHostLink MO must be the O&MIP Address for the node, Thereforethe nodeInterfaceName attributemust be set to the value of theinterfaceName attribute in theIpHostLink MO in position 7.

workingMode=HOST_MODE Routing not supported usingIpHostLink

6 Ip

Parent: IpOam MO in position 5

useHostFile=true (default value) Specifies whether the host file used

The hostFile is used by theresolver and chosen as a source ofinformation before querying a DNSserver.

IpHostLinkId=1

ipAddress=<IP Address> O&M IP address for RBS

ipInterfaceMoRef=<MOReference>

Reference to IpInterface MO inposition 4

If the RBS was configured usingthe Integrate RBS application, thisattribute points to the IpInterfaceMO with the IpInterfaceIdattribute = 1.

7 IpHostLink

Parent: Ip MO in position 6

interfaceName=lh0

DhcpId=18 Dhcp

Parent: IpOam MO in position 5 dhcpServerAddresses=<IPAddress>

13.6 Network Synchronization Configuration

An example of NTP-based synchronization configuration of the RBS is providedin this section.

3939/1553-HSC 105 50/1 Uen S | 2012-11-21


An overview of Network Synchronization configuration is shown in Figure 10.

L0000240A

TransportNetwork7

Synchronization

syncReference[1]syncReference[2]

syncReference[8]

8IpSyncRef

6aIpSyncRef

6bIpInterface

4

GigabitEthernet

3

ExchangeTerminalIp

2

PlugInUnit

1

IpAccessHostEt5

Figure 10 Network Synchronization Configuration Overview

Network Synchronization uses the same MO instances as that for S1 and X2.Table 10 provides details for the common MO instances.

Table 13 provides an example NTP-based network synchronizationconfiguration.

Table 13 Example of NTP-Based Network Synchronization Configuration in RBS


1 PlugInUnit Same MO in position 1 in Table 10

2 ExchangeTerminalIp Same MO in position 2 in Table 10

3 GigabitEthernet Same MO in position 3 in Table 10

4 IpInterface Same MO in position 4 in Table 10

5 IpAccessHostEt Same MO in position 5 in Table 10

40 39/1553-HSC 105 50/1 Uen S | 2012-11-21


Table 13 Example of NTP-Based Network Synchronization Configuration in RBS


IpSynchRefId=16a IpSynchRef

Parent: IpAccessHostEt MO inposition 5

ntpServerAddress=<IP Addressor Domain Name>

IP address or domain name ofprimary network synchronizationNTP server, for example,.morSymmetricon SSU 2000

IpSynchRefId=26b IpSyncRef

Parent: IpAccessHostEt MO inposition 5

ntpServerAddress=<IP Addressor Domain Name>

IP address or domain name ofsecondary network synchronizationNTP server, for example, .Symmetricon SSU 2000

Two IpSyncRef MOs belonging tothe same IpAccessHostEtMO must have differentntpServerIpAddress values.

7 TransportNetwork TransportNetworkId=1

SynchronizationId=1 Created automatically and cannotbe deleted

syncReference[1]=<MOReference>

Same MO in position 6a

syncRefPriority[1]=<priority> Value 1 highest priority and value 8lowest priority

syncReference[2]=<MOReference>

Same MO in position 6b

8 Synchronization

Parent: TransportNetwork MOin position 7

syncRefPriority[2]=<priority> Value 1 highest priority and value 8lowest priority

4139/1553-HSC 105 50/1 Uen S | 2012-11-21

lte transport network configuration

Documents

configuration of transport

x2 configuration ip

configuration aspects

configuration examples

rpu configuration

ip transport network

x2 interface configuration

s1 interface configuration