Transmission Planning Handbook v0.3

Network structure and dimensioning

NOTE: THIS DOCUMENT CONTAINS VERY SENSITIVE AND STRATEGIC INFORMATION. THIS DOCUMENT SHALL NOT BE DISTRIBUTED OUTSIDE OF GROUP3G.

Transmission Planning Handbook

Authors:File Name:Transmission Planning Handbook

Area:Network Planning & EngineeringReference:

Date:Version:0.3

Document history

VERSIONCHANGES TO THE PREVIOUS VERSION

0.1Integration of fixed line and microwave handbooks, amendments

0.2Filling with content

0.3Review after Regions comments

Index

61.INTRODUCTION

61.1.DEFINITIONS

61.2.RELATED DOCUMENTS

71.3.INTRODUCTION

71.4.SPLIT OF RESPONSABILITIES

71.4.1.Group3G and Supplier split of responsibilities

71.4.2.Head Quarters and Regional split of responsibilities

92.GENERAL ASPECTS

92.1.GENERIC TRANSMISSION NETWORK TOPOLOGY

92.1.1.Definition of Backbone, Regional and Access Layers

102.1.2.Planning Guidelines for Access and Regional Layers for Supplier Ericsson

102.1.2.1.Access Layer

102.1.2.1.1.Topology

132.1.2.1.2.Unavailability

132.1.2.1.3.Capacities

142.1.2.2.Regional Layer




162.1.3.Planning Guidelines for Access and Regional Layers for Supplier Nortel

162.1.3.1.Access Layer




202.1.3.2.Regional Layer




232.1.4.Backbone Layer for both Suppliers

232.1.4.1.Topology

262.1.4.2.Unavailability

262.1.4.3.Capacities

262.1.5.Selection of Transmission Concentration Points 1st and 2nd order locations

272.1.6.Dimensioning of Iu, Iub and Iur interfaces

282.1.6.1.Iub Interface

282.1.6.1.1.Provisioning of VPC

292.1.6.1.2.Dimensioning of the interface

302.1.6.1.3.Physical interfaces

322.1.6.1.4.Infrasharing case

322.1.6.2.Iur Interface





352.1.6.3.Iu Interface





382.1.7.Selection of RNC locations

392.1.7.1.Definition of RNC area

412.1.7.2.Definition of exact RNC location

422.1.8.RNC configuration regarding transmission

422.1.9.Node B configuration regarding transmission

422.1.10.Redundancy concept

422.1.10.1.Path Redundancy

422.1.10.1.1.General definitions:

452.1.10.1.2.Access Network Protection

452.1.10.1.3.Regional Network Protection

452.1.10.1.3.1.SDH Technology in Regional Networks: case of Nortel

472.1.10.1.3.2.ATM Technology in Regional Networks: case of Ericsson

492.1.10.1.4.Backbone Network

492.1.10.2.Equipment Redundancy

492.1.10.2.1.Card Protection

502.1.11.Quality planning objectives

502.2.INFRA SHARING CONTRACT WITH E-PLUS

512.2.1.General aspects

532.2.2.PoPC for Operator Designated Traffic

542.2.2.1.UPoNRC and PoISC

552.2.2.2.PoPC for Infra Sharing

562.2.3.Pricing Model

562.3.SYNCHRONIZATION NETWORK PLANNING

562.4.MONITORING AND REPORTING OF PERFORMANCE PARAMETERS

572.5.COMMON TRANSMISSION PLANNING TOOLS

572.5.1.Radio Relay Planning Tool

582.5.2.Transmission Planning Tool

592.5.2.1.Interfaces to other Tools in Group3G:

602.5.2.2.Definition of identifier.

602.6.PROCESSES

602.7.DOCUMENTATION

613.CRITERIA FOR DECIDING ON THE TRANSMISSION MEDIUM TO BE USED

613.1.General Medium concept

633.2.Available technologies

633.2.1.Digital Subscriber Line (DSL)

643.2.2.Microwave

643.2.3.Wireless Local Loop / Local Multipoint Distribution System (WLL/LMDS)

653.2.4.Dense Wavelength Division Multiplexing (DWDM)

663.2.5.Laser link

673.2.6.Leased Lines (LL)

683.2.6.1.Deutsche Telekom

693.2.6.2.Other providers

703.2.7.Summary of available technologies

723.3.Factors to evaluate when selecting a transmission medium

723.3.1.Access Layer

733.3.2.Regional Layer

743.3.3.Backbone Layer

754.PLANNING HANDBOOK FOR FIXED LINES

754.1.QTB (QUAM TRANSMISSION BACKBONE)

754.1.1.General view. Points of presence.

774.1.2.Equipment

774.1.3.Routing tables

784.2.LEASED LINES

784.2.1.Carrier evaluation process

784.2.2.DTAG

784.2.3.Frame contracts signed with carriers

794.3.EQUIPMENT

794.3.1.Vendor Alcatel

794.3.2.Vendor nortel

794.4.Planning Guidelines for Distribution Frames

805.PLANNING HANDBOOK FOR MICROWAVE LINKS

805.1.PLANNING & ENGINEERING

805.1.1.Group3G Frequencies

865.1.2.Hop distances

865.1.2.1.RegTP requirements

865.1.2.2.Group3G requirements

865.1.3.Transmission Concentration Points

865.1.4.Line of Sight and Near End Clearance

875.1.5.Fade Margin

875.1.6.Frequency Planning and Interference

875.1.6.1.General

875.1.6.2.PDH dual polarized hops

875.1.6.3.SDH dual polarized hops

885.1.6.4.Distances to other systems

885.1.6.5.RegTP Processes

885.2.MICROWAVE EQUIPMENT

885.2.1.Vendor Ericsson

885.2.1.1.PDH systems MINI-LINK E

935.2.1.2.SDH systems

935.2.1.3.Waveguides

945.2.2.Vendor Y

955.2.2.1.PDH systems

955.2.2.2.SDH systems

955.2.2.3.Waveguides

955.2.3.Racks, Cabinets and Fans

955.2.3.1.Heat Capacity

955.2.3.2.Examples for Cabinet and Rack layouts

955.3.ISA PROCESSES AND MIGRATION TO GROUP3G

95ACCESS AND REGIONAL LAYERS

1. INTRODUCTION

1.1. DEFINITIONS

Access Layer of the Transmission Network (=Access Layer), the Transmission Network between the NodeB and the corresponding Traffic Concentration Point (TCP) of 1st Order or RNC or the Regional Layer

ISA, Infra Sharing Contract with E-Plus

Operator Designated Traffic, means the Other Partys traffic that is collected by the Responsible Party in a certain (Sub-) Region in case of Area Sharing as well as in case of Infra Sharing.PoISC, means Point of Infra Sharing Connection. It is a port at an RNC of the Responsible Party. It carries the Iu-, Iur- Mur- and Mu- Operator Designated Traffic of the Other Party. A PoISC is a port at an RNC of the Responsible Party in the case of logically split RNC. If the RNC is physically separated the PoISC is not necessarily in the RNC.

PoPC, means Point of Physical Connection. This is the physical connection point (DF) for the Operator Designated Traffic. PoPCs have to be defined for Area Sharing (=National Roaming), for Infra Sharing and for CS and PS traffic.

Regional Layer of the Transmission Network (=Regional Layer), means the transmission network connecting the Access Layer with the Backbone Layer.

Subregion, a geographically connected part of a Region for which the Parties have agreed to provide coverage with an agreed network quality through one of their UMTS Networks at a single Agreed Date. The Subregions bear the name of a city.

TCP, Traffic Concentration Point. TCP1 means TCP 1st Order. TCP2 means TCP 2nd Order.

Trunk, a physical link that carries traffic from two or more different sites.

1.2. RELATED DOCUMENTS

In order to get a better understanding of this document, it is convenient to read the following related documents:

GSM/GPRS Network Plan

ISA Contract with E-Plus, including the Supplements Start-up Notice (Supplier Agreement)

Acquisition of QTB, contract with Alcatel

DDF/ODF Planning Guidelines

Synchronization Planning Guidelines

Group3Gs Frame Contract with Carriers1.3. INTRODUCTION

The following document deals with Group3G preferred UMTS Transmission network topology. It is very important to define common transmission planning criteria for the whole Germany, to be taken into account by the different Regional Offices when planning the transmission network. A common criteria is the key for a smooth and homogeneous quality of the overall transmission network.

The main topics discussed in this document are the relationship and processes between the Regional Offices and Headquarters, the topology and redundancy concepts, capacity in the very beginning, preferred media, quality and performance objectives and standard configuration of the UMTS equipment regarding transmission.

1.4. SPLIT OF RESPONSABILITIES 1.4.1. Group3G and Supplier split of responsibilitiesTo be developed after Group3G receives the answer for the Start-up Notice from the Suppliers.

1.4.2. Head Quarters and Regional split of responsibilities

According to the agreement reached between P&E in the Headquarters and the Regional Directors, the work split between the Headquarters and the Regions will follow this rough scheme:

The Headquarter is responsible for:

the definition of the planning principles that apply to the whole transmission network

signing frame agreements with vendors and carriers

the planning of the Backbone Layer.

And the Regions are responsible for:

the finding of the anchor TCP and RNC locations, that will be later the base for the planning of the Access and Regional Layers

the planning of the Access and Regional Layers.

In a graphical way, the split looks like this:

2. GENERAL ASPECTS

2.1. GENERIC TRANSMISSION NETWORK TOPOLOGY

Information about the General Network and Medium Concept can be found in section 3 of this document Criteria for deciding on the Transmission Medium to be used.2.1.1. Definition of Backbone, Regional and Access LayersAccess Layer of the Transmission Network (=Access Layer) means the Transmission Network between the NodeB and the corresponding Transmission Concentration Point (TCP) of 1st Order or RNC or the Regional Layer

Regional Layer of the Transmission Network (=Regional Layer) means the transmission network connecting the Access Layer with the Backbone Layer.

Backbone Layer of the Transmission Network (= Backbone Layer) means the Transmission Network that connects the Core Network nodes among them and/or that connects different Regional Layers and (Sub-)Regions. The core network elements are the border for the backbone network.

Subregion means a geographically connected part of a Region for which the Parties have agreed to provide coverage with an agreed network quality through one of their UMTS Networks at a single Agreed Date. The Subregions bear the name of a city.

TCP means Traffic Concentration Point. TCP1 means TCP 1st Order. TCP2 means TCP 2nd Order.

Access Layer:

Node B

- Node B

Node B

- TCP-2nd Order

TCP 2nd Order- TCP-1st Order or RNC

Regional Layer:

TCP-1st Order- RNC

RNC

-RNC

RNC

- Core Network Node (2G/3G)

Backbone Layer:

Core Network Node (2G/3G) - Core Network Node (2G/3G)

2.1.2. Planning Guidelines for Access and Regional Layers for Supplier EricssonThe solution presented here is the standard solution. Especial cases shall allow Group3G Regional Planner to change the final topology or the number of network elements. The final planned topology must be approved by the Supplier in that Region.

2.1.2.1. Access Layer 2.1.2.1.1. Topology

Node B

- Node B

Node B

- TCP-2nd Order


The right part of the following figure describes the Access Layer:

Structures for the Access Network shall be:

Chain and Tree configurations, especially for the connection NodeB NodeB.Star and Chain configurations, especially for connection TCP 2.order NodeB.

Double star configurations, especially for TCP 1.order - TCP 2.order.Media:

Node B

- NodeB and

Node B

- TCP-2nd Order:

PDH Ericsson MiniLinkE radio relay point-to-point system with the following frequencies shall be used:

38 GHz (2. priority); 28 GHz (1. priority); 26 GHz; 23 GHz; (exceptional 18GHz and 15GHz )

PDH Ericsson MiniLinkE radio relay point-to-point systems with 8E1 or 17E1 in 1+0 configuration preferred without waveguides can be used.

Point-to-multipoint radio relay systems shall not be used.

Media:

TCP 2.order- TCP-1.order or RNC

PDH Ericsson MiniLinkE radio relay point-to-point system with dual polarized antennas and the following frequencies can be used (in general 2 x 17E1 and 2+0 configuration):

38 GHz (2. priority); 28 GHz (1. priority); 26 GHz; 23 GHz; (exceptional 18GHz)

Point-to-to-multipoint radio relay systems shall not be used.

Ericsson MiniLinkE radio relay systems can be used for TCP 2.order TCP-1order or RNC connections.

Because of lack of frequency allocations or line of sight or the business case, leased lines shall be used to connect some Node Bs. The decision of the media shall always be responsibility of the Regional Transmission Planner from Group3G (depending on section 3 of this document Criteria for deciding on the Transmission Medium to be used).

Depending in the contract with the city Carriers (if it is a flat price not depending in distances) the Node B could directly be connected to the RNC location from all parts of the city or handed over at an transfer point with e.g. 1xSTM-1

The RBS3000 ATM switching functionality at TCP 2nd Order can be used as ATM/AAL2 traffic aggregator if the number of Node Bs in chain to the RNC do not exceed delay requirements (3GPP TR 25.853 or newest). The limit according Ericsson are 5 Node Bs in chain which using the RBS3000 ATM switching functionality.

There are three network topology possibilities to connect Node Bs to TCP 2nd Order with radio relay systems. The following describes the possibilities taking into account the capacity for the last Node B (8E1 with UTRAN Sharing partner), the maximum number of Node Bs allowed in chain which using the RBS3000 ATM switching functionality (max. 5 in chain) and the space for transmission equipment in the RBS (max. space for 6 units, means Ericsson MiniLinkE 1xAMM-4U). A mixture of this three topologies (A, B and C) shall be build in the Access Layer. The decision which to build shall do the Group3G planner.

A) A transmission rack shall only be installed at TCP 2nd Order if the number of Node Bs exceeds 4 to connect to the TCP 2nd Order. The maximum number of Node B connected to the TCP 2nd Order in chain is 2. Three chains, every chain consisting of 2 Node Bs, can be connected to the TCP 2nd Order, means 6 Node Bs connected to the TCP 2nd Order.

Due to delay reasons the last Node Bs in chain can be connected physical to the transmission equipment at the previous Node B not using the RBS integrated ATM-switch. This means the last Node B in chain is logical connected in star topology to the RBS interfaces of TCP 2nd Order due to delay reasons. This is valid for the TCP 2nd Order which are connected to the TCP1.oder. In this case the TCP 1st Order are the nearest to the RNC in the 4 site Regional layer ring.

The capacity between TCP 2nd Order and TCP 1st Order shall be 17xE1 and the configuration is max. 2+0 for the Ericsson MiniLinkE radio relay system. Dual polarized antennas shall be used and the second 17xE1 in 1+0 configuration shall be also engineered for later installation.B) Only 3 Node Bs in chain shall be connected to a TCP 2nd Order without transmission rack. The requirements for a transmission rack shall be prepared for future transmission space needs. The Ericsson MiniLinkE radio relay system capacity for the last Node B in chain shall be 1x 8E1 and for the other 2 radio relay systems connected in chain to the TCP 2nd Order 1x17E1.

Due to delay reasons the last 3 Node Bs connected in chain shall be connected physical to the transmission equipment at the previous Node B not using the RBS integrated ATM-switch. This means the Node Bs in chain are logical connected in star topology to the RBS interfaces of TCP 2nd Order.

The capacity between TCP 2nd Order and TCP 1st Order shall be Ericsson MiniLinkE 17xE1 and max. 2+0 radio relay system. Dual polarized antennas shall be used and the second 17xE1 shall be also engineered for later installation.

C) If there is no space for a transmission rack at TCP 2nd Order available, then max. 3 Node Bs can be connected with 34 Mbps radio relay systems to TCP 2nd Order. Only one ET-M3 board at TCP 2nd Order and at the Node B is needed. The capacity between TCP 2nd Order and TCP 1st Order shall be 34 Mbps radio relay system.

The possibility C) shall only apply in an exceptional case.

The transmission rack will always be delivered from another Supplier. The PDH radio-relay-system is MiniLinkE and shall be delivered by Ericsson.

Delays in the Transmission Network (refer to ITU-T G.114 and 3GPP TR 25.853 or newest) shall be taken into account and minimized by planning the topology and the Network elements. Ericsson shall check this Access Network Topology regarding delay restrictions for any kind of service (e.g. Node B number in chain connected to RNC). The limitation according Ericsson information are 5 Node Bs in chain with RBS ATM switching functionality.

2.1.2.1.2. Unavailability

NodeB - TCP 2.order-TCP 1.order

Threshold

BER=10E-6

Pua= 262min

(0,05%) / year

The unavailable time shall be subdivided equally between the following events:

UA due to climatic conditions

UA due to equipment

UA due to human failures, power supply and other reasons.

SDH hops shall be planned with 0,003% Unavailability for Vertical polarization and 0,005% for Horizontal polarization.The Unavailability for Leased Lines can be different and depends on the contracts with

the Leased Line Carrier (e.g. SLA).2.1.2.1.3. Capacities

For Leased Lines (e.g. Carrier Line and SDSL): 1E1 for the NodeB

The decision to connect the Leased Line NodeB to the nearest NodeB or direct to the TCP1.order or RNC depends on the Leased Line costs. A cost comparison shall be done by the Regional Planner.

For Microwave Links : 8E1 for the last NodeB (4E1 for Group3G and 4E1 for E-Plus)

In the very beginning it is only required to connect 1xE1 for a NodeB to the RNC. Later on it is

possible to benefit from the statistical gain of the TCP1. and 2.order. The Ericsson NodeB (RBS) has got an ATM switching functionality.

2.1.2.2. Regional Layer

2.1.2.2.1. Topology

TCP-1st Order- RNC

RNC


RNC

-RNC

The Regional Layer of the Transport Network connects the Access Layer to the Backbone Layer and carries the Iub-traffic from concentration points (TCP 1.order, Access Layer) to the RNC. Depending on the locations of the RNC also Iur-traffic might run over the Regional Layer as well.

For the Regional Layer, redundant links between the sites are required. In general it is beneficial to set up rings with a low degree of meshes. These rings connect the Access Layer via TCP 1st order to locations of the Backbone Layer in that region.

The TCP 1.order is connected in a ring based on Vendor X SDH radio relay systems (max. 2xSTM-1, 2+0 configuration) and Leased Line to the RNC. This ring consists of max. 4 sites (including the RNC) due to delay reasons when using RBS or RXI ATM switching functionality. The limitation due to delay reasons according Ericsson information are 5 Node Bs in chain with RBS ATM switching functionality.

Occasionally also single Node B can be connected to the TCP 1st Order of the Regional Layer.

Medias for the connections:

TCP-1st Order- TCP-1st Order

TCP-1st Order- RNC

RNC


A media mix (Leased Lines and radio relay systems) shall be used depending on the capacity and due to redundancy reasons in the preferred ring topology.

The distance between two radio relay systems in a ring shall be > 5km and the angle between two SDH radio relay systems at one site for the same ring shall be >60. Two SDH radio relay systems at one site should be avoided. At one site in the Regional layer two media shall be used to get redundancy with media diversity.

nxE3 and nxSTM-1 Leased Lines can be used.

Vendor X SDH radio relay point-to-point system with dual polarized antennas and the

following frequencies shall be used (in general 2 x STM-1 and 2+0 configuration):

38 GHz; exceptional case 26 GHz; 23 GHz; 18 GHz; 13 GHz

In general the Group3G Planner shall decide depending on the business case (depending on section 3 of this document Criteria for deciding on the Transmission Medium to be used), the capacity forecast (Input from Radio Planning, in HQ responsibility) and the Network Design for the media, capacity and technology to use.

At TCP 1.order a number of transmission links (in general 4, means about 20 NodeBs) from TCP 2.order and additional Node Bs in the neighbourhood are terminated.

In each TCP 1.order a grooming/multiplexing/cross-connecting device might be used to be able to multiplex the nxE1, nxE3 or nxSTM-1 that are coming from the Access layer into the nxE3 or nxSTM-1 that make up the ring topology among the TCP 1.order and the RNC. In the beginning the RBS3000 ATM switching functionality at TCP 1st Order can be used as ATM/AAL2 traffic aggregator.

In general the Regional Transmission Planners of Group3G will make a business case taking into account the transmission topology planned and investment for SDH-radio-relay-systems, the traffic forecast (Input from Radio Planning, in HQ responsibility) in every covered city to make an decision for using the RBS3000 ATM switching functionality or RXI with more slots than RBS3000 at TCP 1st Order. The future requirements for a TCP 1st Order to install an RXI additional space shall be taken into account.

An additional transmission rack is needed at TCP1. order to get enough space for the SDH and PDH radio relay systems. The rack will be delivered from another Supplier.

The specific ET-board configuration for an TCP 1st Order is an Group3G Transmission Planners decision.

Delays in the Transmission Network (refer to ITU-T G.114 and 3GPP TR 25.853) shall be taken into account and minimized by planning the topology and the Network elements. The supplier shall check this Regional Network Topology regarding delay restrictions for any kind of service (e.g. Node B number in chain connected to RNC).


TCP 1 RNC Core Network nodeRNC-RNC

Threshold BER =10E-6

Pua= 210min (0,04 %) / yearThreshold BER =10E-6

Pua= 210min (0,04 %) / year



UA due to equipment


SDH hops shall be planned with 0,003% Uavailability for Vertical polarisation and 0,005% for Horizontal polarisation.

The Unavailability for Leased Lines can be different and depends on the contracts with the Leased Line Carrier (e.g. SLA).

2.1.2.2.3. CapacitiesIn the beginning 1xSTM-1 for the regional ring is enough capacity to get also full redundancy for the Node Bs.

A media mix (Leased Lines and radio relay systems) shall be used depending on the capacity and due to redundancy reasons in the preferred ring topology.

The distance between two radio relay systems in a ring shall be > 5km and the angle between two SDH radio relay systems at one site for the same ring shall be >60. Two SDH radio relay systems at one site should be avoided. At one site in the Regional layer two media shall be used to get redundancy with media diversity.

nxE3 and nxSTM-1 Leased Lines can be used.

2.1.3. Planning Guidelines for Access and Regional Layers for Supplier NortelThe solution presented here is the standard solution. Especial cases shall allow Group3G Regional Planner to change the final topology or the number of network elements. The final planned topology must be approved by the Supplier in that Region.

2.1.3.1. Access Layer

2.1.3.1.1. Topology

Node B

- Node B

Node B

- TCP-2nd Order


The Access Layer will carry mainly the Iub interface from the Node Bs towards the RNC.

The standard topology allows up to four (4) TCP 2nd Order connected to each TCP 1st Order in most of the cases. This restriction is not due to space problems in the TCP 1st Order, since an additional transmission rack is already installed, but it is due to capacity problems in the STM-1 ring topology. This capacity shall be enough to support all the n x E1 interfaces coming from the Access network. For example, if a STM-1 Regional ring is installed and 5 TCP 1st Order included in the ring, in the worst case only 16 E1s per TCP 1st Order are possible to drop into the ring.

The TCP 2nd Order will be connected to TCP 1st Order with 1 x 17E1 PDH radio relay system, but it shall be prepared for the immediate upgrade to a 2 x 17E1 radio relay system with dual polarized antennas (even the link engineering shall be done in advance). The PDH radio relay systems shall be from Ericsson (MiniLinkE).

The TCP 2nd Order is only able to support the Node B installed in this location. The reason is because the Node B outdoor cabinet only has space for a three high unit transmission equipment, and it is already occupied with the radio relay system towards the TCP 1st Order.

Therefore, in order to make the topology suitable and the capacity in the Regional ring/s worth, additional outdoor transmission cabinet shall be installed in some of the TCP 2nd Order locations. The locations with most possibilities of becoming a high concentration point shall be chosen to host an additional outdoor cabinet. This locations shall have a minimum number of 3 Node Bs connected (including the Node B in the TCP itself), so the investment in the additional rack is worth it, and a maximum of 6 Node Bs connected (once again due to capacity reasons in the SDH Regional ring).

This way, some of the TCP 2nd Order will be simple Node Bs directly connected to the TCP 1st Order and some others will be real concentration points. Group3Gs Regional Transmission Planners shall have the responsibility to select the TCP 2nd Order locations in a case by case situation.

The following picture shows an example of how the Access network will look like with the defined standard solution:

With the standard solution, up to 90 Node Bs can be connected with one Regional ring. Of course, this is taking into account that all Node Bs will be connected to TCP via radio relay systems. Of course, this will not always be true. Because of lack of frequency allocations or line of site, leased lines shall be used to connect some Node Bs. The decision of the media shall always be responsibility of the Regional Transmission Planner from Group3G.

Depending in the contract with the city Carriers (if it is a flat price not depending in distances) the Node B could directly be connected to the RNC location from all parts of the city.

Media:

Node B

- NodeB

Node B

- TCP-2nd Order

As explained before, PDH radio relay point-to-point systems shall be used as much as possible. This equipment shall be from Ericsson.

Point-to-multipoint radio relay systems will not be used.


Depending on the capacity on the trunk TCP 2nd order TCP 1st order, the medias used will be:

PDH radio relay point-to-point systems will be used as much as possible (2x17E1, dual polarized antenna). This equipment shall be from Ericsson.

SDH radio relay point-to-point systems will be used as the next option when the capacity requires it. This equipment shall be from Vendor X.

Point-to-to-multipoint radio relay systems will not be used.

nxE3 and nxSTM-1 Leased Lines are a good alternative solution compared to radio relay systems. The decision for a media shall be taken by Regional Transmission Planner from Group3G and depending on section 3 of this document Criteria for deciding on the transmission medium to be used.

Delays in the Transmission Network (refer to ITU-T G.114 and 3GPP TR 25.853) shall be taken into account and minimized by planning the topology and the Network elements. Nortel shall check the Access Network Topology regarding delay restrictions for any kind of service (e.g. NodeB number in chain connected to RNC).


NodeB - TCP 2.order-TCP 1.order

Threshold

BER=10E-6

Pua= 262min

(0,05%) / year



UA due to equipment


SDH hops shall be planned with 0,003% Uavailability for Vertical polarisation and 0,005% for Horizontal polarisation.The Unavailability for Leased Lines can be different and depends on the contracts with


For Leased Lines (e.g. Carrier Line and SDSL): 1E1 for the NodeB. It is possible to cascade 2 NodeBs with 1E1 due to fractional E1.

The decision to connect the Leased Line NodeB to the nearest NodeB or direct to the TCP1.order or RNC depends on the Leased Line costs. A cost comparison shall be done by the Regional Planner.

For Microwave Linkes : 8E1 for the last NodeB (4E1 for Group3G and 4E1 for E-Plus).In the very beginning it is only required to connect physical 1xE1 for a NodeB to the RNC. Later on it is possible to benefit from the statistical gain of the TCP1. and 2.order. The Ericsson NodeB (RBS) has got an ATM switching functionality inside.

2.1.3.2. Regional Layer

2.1.3.2.1. Topology

The Transmission Access Layer is defined as the connections:

TCP-1st Order- TCP-1st Order

TCP-1st Order- RNC

RNC

- Core Network Node (2G/3G)The Regional Layer of the Transport Network connects the Access Layer to the Backbone Layer and carries the Iub-traffic from concentration points (TCP 1st order, Access Layer) to the RNC. Depending on the locations of the RNC also Iur-traffic might run over the Regional Layer as well.

For the Regional Layer, the standard solution consists of a ring topology between the TCP of 1st Order. For cities with a small number of Node Bs only one Regional ring is required, but if the topology and the number of Node Bs justifies it, several ring structures shall be settled up.

These rings shall be based on SDH technology, because the traffic and the number of Node Bs do not justify a higher investment in the very beginning. The SDH equipment shall be from Alcatel(1641, 1651 and 1661 ADM family). One SDH multiplexer shall be installed in each of the TCP 1st Order.

In the very beginning, the bandwidth supported in these Regional ring/s shall be STM-1 (also in some very special cases n x E3), depending on the number of Node Bs connected to the ring. The type of ADM shall be chosen by the Regional Transmission Planner from Group3G in a later step.

These SDH multiplexers shall have n x E1 tributaries (and in very few cases n x E3) to collect the traffic from the Node Bs in the Access network.

An outdoor cabinet for transmission equipment shall be installed in each of the TCP 1st Order to host the SDH multiplexers and some other additional transmission (PDH and SDH radio relay systems) and non-transmission equipment (batteries, etc.). The Supplier for these additional cabinets is ongoing.

The number of TCP 1st Order in the Regional ring shall be up to 5. This number shall be reviewed once the Nominal Transmission plan is ready, to take into account delays and the maximum number of Node Bs possible in order to not over exceed the capacity of the ring (based in STM-1 in the beginning).

In the case that the RNC is located in the same city as the Node Bs, the location of the RNC shall be part of the Regional ring/s and shall also be a TCP 1st Order. Therefore, also a SDH multiplexer shall be located in the RNC location. In this case, the RNC is located in Group3Gs backbone site of that city.

In case that the RNC is located in a different city than the Node Bs, the Backbone location in that city shall be part of the Regional ring/s and shall also be a TCP 1st Order.

The SDH RNC from Nortel only accepts STM-1 Clear Channel (unstructured) interfaces. Therefore, in front of each Nortels RNC a Passport 7480 is needed to be able to connect the E1s coming from the Access layer to the RNC. This Passport 7480 shall be part of the transmission network as PP7k in front of an SDH RNC and part of the Transmission Preside Management System.

The Passport 7k should be totally redundant in the main boards (Control Processor, power supply, etc.) and in the interfaces connecting the PP to the RNC.

Media:

A media mix (Leased Lines and radio relay systems) shall be used to achieve redundancy reasons in the regional ring/s topology. The media to be chosen is a decision of Group3Gs Regional Transmission Planner in a case by case situation.

In the very beginning, SDH STM-1 radio relay point-to-point systems will be used when the capacity requires it. It shall be installed as 1+0, but prepared for an upgrade to 2+0. This equipment will be delivered from another Supplier.

The following picture gives a general view of the Regional Layer standard topology:

Delays in the Transmission Network (refer to ITU-T G.114 and 3GPP TR 25.853) shall be taken into account and minimized by planning a suitable topology and number of elements in the network. Nortel shall check the Regional Network topology regarding delay restrictions for any kind of service (e.g. Node B number in chain connected to RNC).

The solution presented here is the standard solution. Special cases shall allow Group3G Regional planner to change the topology or the number of network elements.


TCP 1 RNC Core Network nodeRNC-RNC

Threshold BER =10E-6


Pua= 210min (0,04 %) / year



UA due to equipment


SDH hops shall be planned with 0,003% Unavailability for Vertical polarisation and 0,005% for Horizontal polarisation.The Unavailability for Leased Lines can be different and depends on the contracts with


In the beginning 1xSTM-1 for the regional ring is enough capacity to get also full redundancy for the Node Bs.

A media mix (Leased Lines and radio relay systems) shall be used depending on the capacity and due to redundancy reasons in the preferred ring topology. The distance between two radio relay systems in a ring shall be > 5km and the angle between two SDH radio relay systems at one site for the same ring shall be >60. Two SDH radio relay systems at one site should be avoided. At one site in the Regional layer two media shall be used to get redundancy with media diversity.

nxE3 and nxSTM-1 Leased Lines can be used.2.1.4. Backbone Layer for both Suppliers

2.1.4.1. Topology

Figure 3.4.2.1.1. Traffic covered by the Core Backbone

MSCtoMSC

MSCtoDTAG*

MSCtoRNC

RNCtoRNC**

* Interconnection Traffic, could be an alternative Operator too

** not included in the early phase of the UMTS LaunchFigure 3.4.2.1.2 QTB Backbone Layer

The above drawing represents a typical Core Backbone, in this case the own QTB Network.

The blue ring is representing the high dense optical Network WDM

The red ring is representing the Transport Network Layer SDH.

The Backbone Network is the main feed for Capacity in the G3G Network and will carry most likely traffic between the MSC(s) as well as partly traffic from all of different regions or any other traffic attached on it. Therefore the Network shall be:

Based upon PDH/SDH only to avoid delays or echo

transparent to any used Protocol such as ATM or IP.

The Network shall be build either by :

Leased Capacity (LL) dark Fibre G3G own Infrastructure QTB

Note: Because of the expected high demand on Bandwidth in this section, the use of Microwave or Radio Links should not be taken in consideration.

In the early phase, the G3G Backbone shall have at least a ring structure. This might be suitable enough for the beginning, and the Network will consist of 2 MSCs only. Since the traffic is growing and more MSCs will be implemented in the future, the Network must be re-designed, must be completed. To gain highest availability as well as redundancy, a fully meshed Network Structure is required.

Figure 2: The Ring Structure

Figure 3: Meshed Network

Start up Phase

Final Solution

Additional use of the Backbone:

Since the Backbone Network does cover a number of major cities and Areas in Germany, we shall even take in consideration to use the Network partly as a collector - to add and drop traffic in the regions. This regional traffic then could be forwarded i.e. to the attached RNC(s) in the region.

Example:RNC to RNC Interconnection

TCP1 to RNC Interconnection

This could be done easily since the QTB consist of several PoPs across the country. To do so, the construction of Sub-Networks within the Backbone is mandatory. In Figure 4 you will see how a Sub Network could look like. In the drawing the red Ring represents the Core Backbone which will connect the MSC to each other. Where the yellow and the green little ring represents the regional sub Network. Those are collecting traffic in the regions from any outside Area to forward it to the RNC.

Figure 4: Sub Networks in the regions

Note: Whether this option can be used or not, does mainly depends on the availability of regional PoPs and Transmission Infrastructure, as well as the basic Network coverage.

2.1.4.2. Unavailability

The design Target of the Transmission Backbone should match the ITU-T G.826 recommendation for Transmission Systems and Media.

MSC MSCRNC - RNC

Treshold BER = 10 6

Pua = 210min per Year (0,04%)

99,96 % AvailabilityTreshold BER = 10 6

Pua = 210min per Year (0,04%)

99,96% Availability

2.1.4.3. Capacities

Because of the expected Traffic in the near future, Voice & Data, (Iu & IuB), and to protect the Network, the Capacity shall be not under 2,5Gbit. In addition, spare capacity is needed to be able to expand the Backbone quickly if needed and to avoid downtime or outages of the Network once it is in Operation.

To allow SDH protection, not the full capacity shall be used for traffic. Depending on the amount of protected Links, at least of the capacity shall be reserved for SHD protection.

2.1.5. Selection of Transmission Concentration Points 1st and 2nd order locations The future capacity needs for UMTS Transmission Networks will be about 2 to 4 times higher than operators are used to in GSM networks. The best suited frequencies will be occupied in the city centres very soon, if all operators continue present kind of microwave network planning.

Here are collected some requirements for new type of thinking for selection purposes of microwave concentration points.

Concentration points are used for Access and Regional Transmission Layers. Operator collects to these sites traffic from neighbour Node B connections through microwaves, (and leased line operators networks, if necessary) to one place.

On these sites may be located some kind of node or switch for collecting the traffic from sites towards RNC sites. This could be a multiplexing technology (ATM ; PDH or SDH) if technical and economical feasible.

The most important requirements for the selection of a suitable concentration point are:

The concentration point shall be on a place from where it is possible to see every site, but the concentration point should not be seen from other concentration points to avoid interferences.

Planning concentration points in heavy populated city areas, where the number of microwave radios operating on different frequency bands is high, it should be avoided to use (very) high buildings, masts or towers in order to decrease interference problems. The building heights should be optimal.

It should also be avoided very big hills or tops of mountains in the vicinity of big cities.

It should be payed attention to the terrain of the area and seek interference-free points, and choose them for concentration points.

The sites shall be selected so that there is direct shadowing from objects/obstacle as buildings, the wall of the own building, rocks, mountains, dense coniferous forests, etc to the most dense centres where the number of microwaves is highest.

There shall be enough place for the antennas (recommended 5-10 Antennas), additional Node B and transmission equipment.It is necessary that in a Concentration Point at least one Leased Line Service Provider, or Carrier, is present in the building with fibre connection. The optimum choice would be that there are more than one Carrier present, to be able to negotiate better rates for the leased lines. Especially this condition must be fulfilled in TCP of 1st Order, where city fibre rings will bring the concentrated traffic towards the RNC. 2.1.6. Dimensioning of Iu, Iub and Iur interfaces

All the transmission equipment, Suppliers designs and Group3G Transmission Planners designs must comply with standard interfaces defined by the ITU-T, ITU-R, 3gpp and ATM Forum.

ATM is specified by 3gpp as the protocol used in the layer 2 for the Iub, Iur and Iu-cs interfaces. Besides this, each Supplier has the freedom to decide the configuration of the VPC (Virtual Path Connection) and the VCC (Virtual Channel Connection) in their UMTS equipment.

Some general guidelines are given in the following section to be able to dimension the Iub, Iur and Iu interfaces, taking into account the up-to-now knowledge of Ericsson and Nortels technology. But Group3G should develop a close relationship to the Supplier in order to be able to optimize and adapt the characteristics of the Suppliers equipment configuration to Group3Gs own needs.

In their UMTS equipment, Ericsson offers only one type of traffic rate, CBR (Constant Bit Rate). No other type will be offered in future releases. On the other hand, Nortel offers from the first release several traffic rates, and especially advices to configure VBR (Variable Bit Rate) in the network.

There is usually one ATM VPC per physical interface, but if several traffic rates are supported at the same time (CBR, VBR, etc.), there shall be the possibility of one VPC per traffic rate (although this is not the case in the first releases of the Suppliers). The VPC configured should transport separated VCC, and the number of these depend on the interface and the Supplier.

Once the topology of the transmission network is fixed, the bandwidth and number of VPC and VCC per interface depends on the traffic figures.

2.1.6.1. Iub InterfaceThe Iub interface is connecting the Node B with the corresponding RNC. The user plane of the Iub interface carries three different types of services: voice, circuit switched data and packet switched data. ATM AAL2 is used as the standard transport layer for the Iub interface.

The control plane is based on ATM AAL5 protocol and it is used to carry the signalling, synchronization and O&M traffic. All the information exchanged between the RNC and the Node B uses NBAP and ALCAP protocols.

2.1.6.1.1. Provisioning of VPC

The ATM VPC configured in this interface shall transport separated VCC for the user and control planes, and the number of VCC depends on the Supplier.

In the case of Ericsson, one ATM VPC supporting CBR traffic is provisioned. The VPC should have the following distribution:

1 AAL5 VCC for O&M 1 AAL5 VCC for NBAP signaling support 1 AAL5 VCC for ALCAP signaling support n AAL2 VCC for user traffic. Ericsson advices that n should be at least 2. Generally, it is recommented that one VCC per QoS class is configured (Conversational, Streaming, Interactive and Background).Note: the maximum number of VPC per 8xE1 interface card is 8 and the maximum number of VCC per 8xE1 interface card is 240.

Nortel advises to provision one ATM VPC supporting VBR (real time-Variable Bit Rate) traffic for the user and the control planes. The VPC should have the following distribution:

1 (1) AAL5 VCC for OMC-B (with fixed VPI.VCI number 0.32) 1 (1) AAL5 VCC for NBAP-c signaling support (part of the Node B CP) 1 (6) AAL5 VCC per CEM (Channel Element Module) for NBAP-d signaling support (part of the common CP) 1 (1) AAL5 VCC for ALCAP traffic n (16) AAL2 VCC for user traffic. Nortel advices that n should be at least 2, one for delay sensitive traffic and the other one for non-delay sensitive traffic. Generally, it is recommented that one VCC per QoS class is configured (Conversational, Streaming, Interactive and Background).Note: in brackets the maximum number of VCC of each type possible to provision per Node B.

As noted before, these are only the first values and characteristics of UMTS equipment given by the Suppliers. Further information will be provided after the Suppliers answer to the Start-up document.2.1.6.1.2. Dimensioning of the interface

In order to calculate the needed bandwidth for the Iub, it is necessary to consider the following factors:

Estimated user peak traffic that the interface will have to support, including activity factor. It will be necessary to determine the transport characteristics (Constant Bit Rate, CBR, or Variable Bit Rate, VBR), data rate and QoS, together with the requirements in terms of blocking and delay.

30% of the connections will be assumed to be in soft handover situation, which means that the Iub interface should be dimensioned to be able to transport at least 30% to 50% more traffic.

Traffic must be increased with the overhead introduced due to the ATM headers. Additionally, AAL2 and AAL5 headers must be considered, depending on the type of traffic and number of VPC and VCC, as well as the headers of the Frame Protocol following the 3GPP specs (25.430). It is assumed that in the very beginning is necessary to add 23% capacity due to all headers

Signaling (NBAP, ALCAP) traffic must also be considered. 10% of traffic due to signaling must be considered (at least 64 kbit/s).

Operation & Maintenance traffic must also be considered. 2% of traffic due to O&M must be considered (a minimum of 64 kbit/s).

Finally and since the coded voice with Adaptative Multi Rate (AMR), is a bursty traffic with a certain burstingness factor, and since real time traffic is sensitive to delay, it will be necessary to dimension the links with a margin, in order to cope also with the bursts. In most cases, considering a burstingness factor of 80% is enough.

For this, the bandwidth needed on the physical link is,

where,

TIub is the total traffic to be transmitted through the Iub, considering user data, soft handover, headers, signaling and O&M traffic.BIub is the needed bandwidth on the physical link (1984 kbps in the case of E1 interfaces)

(Iub is the burstingness factor

Considering all these factors, it is seen that is necessary to have a bandwidth in the physical link (BIub) that supports 110% - 125% more traffic than user peak traffic.

Since transmission costs are high, it is proposed to start with low bandwidth and increase it, as the user data to be transmitted increases, that is, when the number of users in the network or its traffic profile implies more bandwidth needs. This should specially be applied in case of leased lines, where the aim is to pay only for the minimum required bandwidth.

The user peak traffic is an input from the Radio Planning team. They calculate this value based in the traffic estimation from Group3Gs business plan and the number of channel elements installed in the Node B. It is clear that no further capacity will be needed than the given by the calculation based in the channel elements of the Node B, since the Node B will not be able to process more traffic. The configuration of the Node B has been agreed together with E-Plus for the ISA contract.

For the starting phase, the Radio Planning team has used the most probable case according to Group3Gs business plan (same amount of voice users than 64 kbps data users) to calculate the starting bandwidth needed for the Iub interface.

Following these criteria and more assumptions taken in the UTRAN design, in the very beginning Group3G expects half an E1 of traffic for both Operators, that is, one E1 physical interface connected from each Node B to the next TCP (Traffic Concentration Point) to carry the Iub traffic. A TCP could be one transmission equipment, another Node B or the RNC.

2.1.6.1.3. Physical interfaces

Ericsson has available for the first phase release, the P2.1, the following transmission interfaces for the Iub interface, both in the Node B and in the RNC (since they are both based in the Cello platform and the cards are also exchangeable):

Transmission InterfaceStandardsConnector Type

8 x E1 ATM (ET-M1)ETS 300 420

ITU G.703/G.704120 Ohms symmetrical *)

8 x E1 IMA (ET-MC1)AF-PHY-0086.001

ETS 300 420

ITU G.703120 Ohms symmetrical *)

1 x STM-1 ATM (ET-M4)ITU I.432.2

ITU G.783, G.957Optical SC/PC singlemode. **)

1 x STM-1 Channelized (ET-MC4, 63xE1 ATM) ITU I.432.2


In the next release, P.3, Ericssons UMTS equipment will support also E3 ATM (ET-M3) interface.

On the other hand, Nortel has available for the first phase release, the V3, the following transmission interfaces for the Iub interface. In the Node B:


8 x E1 ATM, Fractional

(supporting IMA)AF-PHY-0086.001

ETS 300 420


In the RNC:


2 x STM-1 ATM ITU I.432.2


Therefore, additional ATM multiplexers are needed in front of the RNC in order to gather all the E1 ATM traffic coming from the Node Bs and multiplex it into the STM-1 interfaces of the RNC. Nortel proposes the Passport 7480, and because it is already existing in Group3Gs network, it will be implemented in the starting phase as interface adapter. This would leave the possibility of the following transmission interfaces for the Iur:



(MSA, supporting 11xE1 IMA)AF-PHY-0086.001

ETS 300 420


In future releases Nortels UMTS equipment will support also optical STM-1 Channelized interface and E3 interface, as well as grooming functions in the Node B. This information will be checked once again once the response to the Start-up Notice document is received.

*) In case a DDF (Digital Distribution Frame) is present in the site, the connectors offered by the Supplier should be LSA-Plus for Krone DDF and wire-wrap for ADC DDF. In case there is no DDF and the connection is made with a direct cable, the connector offered by the Supplier should depend on the transmission equipment available.

**) In case there is an ODF (Optical Distribution Frame) installed in the site, the connectors offered by the Supplier will be E2000. In the case that there is no ODF and the connection is made with a direct cable, the connector offered by the Supplier should depend on the transmission equipment available.

In the Start-up Notice document it has been requested that the Suppliers develop electrical STM-1 interfaces, and E2. The answer to this request will be published when the Start-up Notices answer is received.2.1.6.1.4. Infrasharing case

The above mentioned characteristics refer to the case of Area Sharing with E-Plus. Only few words can be said to the Infrasharing case, since the technology is not very well developed yet by the Suppliers and not much information is available. Ericsson will have his technology prepared for UTRAN sharing in the future release P3 (Group3G will install P2.1) and Nortel already in the release installed for Group3Gs network, the V3.

In the case of Ericsson and Nortel, also one Iub interface will support both Operators traffic, but it will be possible to logically split the user planes from each Operator into two different VCCs.

2.1.6.2. Iur InterfaceThe Iur interface is connecting the RNCs among themselves to exchange signalling information. In the Iur interface user data is transported for users that are in soft handover with more than one RNC involved. The user data is of the same type as on the Iub interface with the same type of overhead.

The control plane is based on ATM AAL5 protocol and it is used to carry the signalling, synchronization and O&M traffic. One user function of the control plane is the RNSAP protocol (Radio Network Subsystem Application Part).

The AAL2 is used as transport layer for DCH data streams on the Iur interface.

This interface is supporter by Ericsson from release P2 (Group3G will install P2.1 for the starting phase) and by Nortel since release V1 (Group3G will install V3 for the starting phase). But especial care must be taken when implementing this interface, because it is not very well developed among the UMTS Suppliers.2.1.6.2.1. Provisioning of VPC

The ATM VPC configured in this interface shall transport separated VCC for the user and control planes, and the number of VCC depends on the Supplier.

The VCC to be configured in this interface are under review by the Suppliers, the information Group3G has available for the moment is the following:

AAL5 VCCs for signalling supported by RNSAP and signalling supported by ALCAP (for transport network control).

AAL2 VCCs for user plane traffic. There should be at least one VCC per QoS class (Conversational, Streaming, Interactive and Background).Depending on the bandwidth needed several VCCs of one type could be needed. In the case of Ericsson, the only provisioning detail data that is available for the moment is the maximum number of VCC per interface (30 VCC in the case of E1 interface).

Once Group3G receives the answer to the Start-up notice document, further details will be published.

2.1.6.2.2. Dimensioning of the interface

The Iur should not be dimensioned calculating the Iur traffic from each carrier in the Node B and multiplying this by the number of carriers. These calculations would lead us to erroneous figures.

In the dimensioning of the Iur it is therefore necessary to consider the soft handover traffic between Node Bs. In a very dense network with several number of RNC in the same city, it is usually considered that 30% to 40% of the connections are in soft handover. To this 30% it is necessary to apply a factor minor than one in order to consider only the traffic between RNC due to soft handover between cells that belongs not to the same RNC.

The most common solution for this interface would be a point to point interface multiplexed physically in the Iu interface, since the amount of traffic in this interface is really small.The amount of Iur traffic load depends on how large the probability is that an UE is in Soft Handover with two or more RNC involved. If a RNC does not have any borders towards another RNC, there is no traffic on the Iur interface.

For the moment, no two RNC have common borders and therefore no Iur interface will be used or connected. The only exceptions are the two RNC from Nortel located in Hamburg in the same physical room or building. In this case a direct redundant n x STM-1 connection shall be established through in-house cabling.

2.1.6.2.3. Physical interfaces

The implementation of this interface is optional and in case of using it, it will free the Core Network from switching the cells assigned to two different RNC. The existence of this interface does not mean a physical implementation of the cabling between the two RNCs. That is, the Iur interface may be multiplexed together with the Iu interface towards the Core Network nodes.

If it is decided to implement the Iur interface with a physical interface, Ericsson has available for the first release the following transmission interfaces:


8 x E1 ATM (ET-M1)ETS 300 420



ETS 300 420




1 x STM-1 Channelized (ET-MC4, 63xE1 ATM, up to 8xE1 IMA) ITU I.432.2


In the next release, P3, Ericssons UMTS equipment will support also E3 ATM (ET-M3) interface.

In the case of Nortel,

additional ATM multiplexers are needed in front of the RNC in order to gather all the E1 ATM traffic coming from the Node Bs, this ATM multiplexer will also provide smaller physical interface possibilities for the Iur. . This would leave the possibility of the following transmission interfaces for the Iur:




ETS 300 420




In future releases Nortels UMTS equipment will support also optical STM-1 Channelized interface and E3 interface, as well as grooming functions in the Node B. This information will be checked once again once the response to the Start-up Notice document is received.



In the Start-up Notice document it has been requested that the Suppliers develop electrical STM-1 interfaces, and E2. The answer to this request will be published when the Start-up Notice is received

2.1.6.2.4. Infrasharing case

The above mentioned characteristics refer to the case of Area Sharing with E-Plus. Only few words can be said to the Infrasharing case, since the technology is not very well developed yet by the Suppliers and not much information is available. Ericsson will have his technology prepared for UTRAN sharing in the release P3 (Group3G will install P2.1) and Nortel already in the release installed, the V3.

In the case of Ericsson and Nortel, also one Iur interface will support both Operators traffic, but it will be possible to logically split the user planes from each Operator into two different VCCs.

2.1.6.3. Iu Interface

The Iu interface connects the UTRAN with the Core Network nodes. The Iu interface toward the circuit switched domain is called Iu-cs and the Iu interface towards the packet switched domain is called the Iu-ps.

Iu-cs:

In the radio control plane, RANAP (Radio Access Network Application Part) is used as the application protocol. For the transport control plane, ALCAP is again used.

AAL2 Signalling Protocol is used for establishing AAL2 connections towards the PSTN/ISDN domain, that is, as the standard transport layer for the user plane.Iu-ps:

AAL5 virtual circuits are used to transport the IP packets across the Iu interface toward the packet switched domain. There is usually a one-to-one relationship between the VCC configured in this interface and the IP address as required by Classical IP over ATM.

In the user plane, GTP-U and UDP are used one top of the AAL5 layer. The control plane uses RANAP protocol as well.

2.1.6.3.1. Provisioning of VPC

Iu-cs:

The ATM VPC configured in this interface should transport separated VCC for the following purposes:

AAL5 VCC for signalling supported by RANAP_CS.

AAL2 VCC for QoS class conversational.The number of VCC depends on the Supplier. Depending on the bandwidth needed several VCC of one type could be needed.

Iu-ps:

The ATM VPC configured in this interface should transport separated VCC for the following purposes:

AAL5 VCCs for signalling supported by RANAP_PS

at least one AAL5 VCC per QoS class (Streaming, Interactive and Background).

Depending on the bandwidth needed several VCC of one type could be needed.

In the case of Ericsson, not further details are known. In the case of Nortel, a table with the different possibilities in order to provision the VPC, in the case of one and two STM-1 interfaces, is shown below:

AALVPIVCINumber

STM-1VCC CS CPAAL52*Rncld32-4716

VCC ALCAPAAL5450-48132

VCC CS UPAAL248-7124

VCC PS CPAAL52*Rncld-132-4716

VCC PS UP IpCos1AAL548

VCC PS UP IpcCos2AAL549



VCC OAMAAL5/

Second STM-1VCC CS UPAAL22*Rncld48-7124

VCC PS UP IpCos1AAL52*Rncld-148




VCC OAMAAL5/

Further details will be added when the Suppliers respond to the Start-up Notice document.

2.1.6.3.2. Dimensioning of the interface

In order to calculate the needed bandwidth for the Iu, it is necessary to consider the following factors:

Estimated user traffic for voice and data coming from the Node Bs under each RNC and that should access the Core Network elements. To the final values, converted from Erlang to Kbps, a blocking probability of 0,005 could be considered.

Traffic must be increased with the overhead introduced due to the ATM headers. Additionally, AAL2 and AAL5 headers must be considered, depending on the type of traffic and number of VPC and VCC, as well as the headers of the Frame Protocol following the 3GPP specs (25.430). It is assumed that in the very beginning is necessary to add 23% capacity due to all headers

Signaling (NBAP, ALCAP) traffic must also be considered. 10% of traffic due to signaling must be considered (at least 64 kbit/s).

Finally a security margin of a 20% should be considered due to the burst characteristic of this traffic and the delay sensitivity.

Since transmission costs are high, it is proposed to start with low bandwidth and increase it, as the user data to be transmitted increases, that is, when the number of users in the network or its traffic profile implies more bandwidth needs. This should specially be applied in case of leased lines, where the aim is to pay only for the minimum required bandwidth.

The user traffic is an input from the Radio Planning team. They calculate this value based in the traffic estimation from Group3Gs business plan and the number of channel elements decided to be installed in the Node B and the RNC capabilities.

Because in the starting phase, there will not be a very big amount of traffic, the connection between the RNC and the Core Network Nodes must be based on redundant n x STM-1 interfaces. Since the circuit switch and the packet switched Core Network nodes will be located in the same physical location, the Iu-cs interface and the Iu-ps interface should be multiplexed in the same physical link. The Media Gateway, in the Core Network site, will be responsible for demultiplexing this traffic into the two different logical connections, the Iu-cs and the Iu-ps.2.1.6.3.3. Physical interfaces

Ericsson has the following transmission interfaces to support the Iu interface:


8 x E1 ATM (ET-M1)ETS 300 420



ETS 300 420




1 x STM-1 Channelized (ET-MC4, 63xE1 ATM, up to 8xE1 IMA) ITU I.432.2


In the next release, P3, Ericssons UMTS equipment will support also E3 ATM (ET-M3) interface.

In the case of Nortel,

additional ATM multiplexers are needed in front of the RNC. This would leave the possibility of the following transmission interfaces for the Iu interface in the case of Nortel:




ETS 300 420




In future releases Nortels UMTS equipment will support also optical STM-1 Channelized interface and E3 interface. This information will be checked once again once the response to the Start-up Notice document is received.



In the Start-up Notice document it has been requested that the Suppliers develop electrical STM-1 interfaces, and E2. The answer to this request will be published when the Start-up Notice is received

2.1.6.3.4. Infrasharing case

The above mentioned characteristics refer to the case of Area Sharing with E-Plus. Only few words can be said to the Infrasharing case, since the technology is not very well developed yet by the Suppliers and not much information is available. Ericsson will have his technology prepared for UTRAN sharing in the release P3 (Group3G will install P2.1) and Nortel already in the release installed, the V3.

In the case of Ericsson and Nortel, two Iu-cs and two Iu-ps interfaces will support both Operators traffic, one to each Infrasharing Partners Core Network nodes.

2.1.7. Selection of RNC locations

There are two steps to follow when choosing the location or geographical area of an RNC:

Definition of RNC areas

Definition of exact RNC location

Of course, the procedure of how to select an RNC location is very much related, not only to the transmission costs savings, but also to the inputs from the Radio Planning team. Therefore, the steps mentioned above should be developed together with the Radio Planning team.

2.1.7.1. Definition of RNC area

The need for having a new RNC should be identified by the Radio Planners in the Headquarters. The definition of the RNC area or city is responsibility of the Transmission Planners also in the Headquarters, in a very close relationship to the Radio Planners.

In the very beginning, Group3G will start the deployment of the UMTS network only in some big cities in Germany. Depending on the vendor characteristics of the RNC and the number of Node Bs deployed in each city, there might be a need for one or more RNC in that city. The decision of the number and the location of the RNC will have a big impact in the performance of the RNC.

The location of the RNC in the UTRAN is very flexible. RNC can be collocated with a MSC or following a decentralized concept. The final decision of the RNC area depends mainly on the following factors:

Transmission costs

Physical space available

RNC performance and characteristics (capacity, maximum number of physical interfaces, delays, etc.)

Minimum inter-RNC handover

The optimum location of RNC depends mainly in the transmission leased lines tariff structure. Generally, locating the RNC collocated with the MSC gives no possibility of multiplexing the traffic between the Node Bs and the MSC, and therefore the full capacity has to be transported even if the links are not totally used. This problem can be solved with additional investments in cross-connects, which would multiplex the physical layer (not 100% efficient), or in ATM switches, which would multiplex the ATM adaptation layers.

On the other hand, if the RNC is collocated remotely, it can be used as a multiplex (ATM switch in most of the cases), which will send over to the MSC only the capacity really used.

If Group3G owns a Transmission Backbone Network, it should be as much utilized as possible, since it means no additional transmission costs. If this is the case, the location of the RNC will not depend very much of the transmission costs.

On the other hand, if Group3G has to lease capacity from different Service Providers, a business case has to be very carefully studied. If the UMTS traffic is not very high and the agreements with the Carriers gives Group3G very competitive prices, maybe the transmission costs have no big impact in the final location of the RNC.

But if the traffic is estimated to grow very much in the near future, it could be a reason to decentralize the RNC. Also in general, if the transmission tariff correlates strongly with the distance of the transmission link (case of DTAG, for example), distributed RNC concept is favored.

The location of the RNC also depends very strongly in the space available within the area. Renting of new space is very difficult and expensive. In the case that Group3G already owns sites within the city with enough space available for the installation of the RNC, then the decentralized RNC concept is a better solution. In the case that there is a Core site within the city in question, then the RNC should be collocated with the MSC.

The capabilities of the RNC are also important factors to take into account when choosing the RNC location. For example, in the case of Nortel, the RNC is only able to handle 200 Node Bs and therefore, in the city of Hamburg more than one RNC would be needed (Hamburg 200 Node Bs). Since it is estimated that Bremen needs only 66 Node Bs to be covered, it makes sense to locate the RNC of Bremen in Hamburg and share the capacity with Hamburg Node Bs.

Other important factors, that affect very much the final decision, are the planning objectives that the transmission network must fulfill in order to comply with the UMTS quality parameters. Delay is one of the most important of these factors. The connection Node Bs RNC and RNC MSC must fulfill some delay requirements specified in the section Quality Planning Objectives. If, once chosen the RNC location, one of the delay requirements is not fulfilled, moving the RNC closer to the Node Bs or to the MSC could achieve the optimum solution.

Of course, the selection of the RNC location should be decided in order to minimize the inter-RNC handover. This takes place in the Iur interface between RNCs, and it is not very well developed and successfully tested by most of the Suppliers. The use of this interface should be avoided when possible, unless it is very much needed as in the case mentioned above of Bremen and Hamburg.

Taking into account all the criteria mentioned in the above paragraphs, the selection of the RNC locations for the Start-up phase has been already taken. The following is the list of cities and corresponding RNC locations, following the decentralized concept:

RNC AreaCity coveredNumber of Node BNumber of RNCSupplier

Frankfurt am MainFrankfurt am Main721Ericsson

DsseldorfEssen831Ericsson

Dortmund76

HamburgHamburg2002Nortel

Bremen66

MnchenMnchen1511Nortel

StuttgartStuttgart861Nortel

LeipzigLeipzig691Nortel

The main reason for this decision is the fact that Group3G is planning to own a Transmission Backbone Network that has physical space available for the installation of the RNC in each of the PoPs in the cities.

The only special cases are Hamburg and Bremen, explained before, and Essen and Dortmund. Due to the small number of Node Bs deployed in these two cities, it is wise to share the capacity of the same RNC, located in the Core Network site in Dsseldorf (because of available space and Operational advantages).

2.1.7.2. Definition of exact RNC location

The definition of the exact location of the RNC within one city is decided by the Transmission Planners in the Region responsible for that city, with the support of the Radio Planners.

The location of the RNC depends on many factors:

The transmission costs

The topology of the access and regional networks

The physical space available in the city

The suitability of using the RNC site as Transmission Concentration point (TCP)

The availability of one or several Carriers in the same location

etc.

The location of the RNC should be a Group3Gs Backbone site when possible. This will give the possibility to connect the RNC to the assigned Core Network nodes, also located in Backbone sites.

The chosen location should also be a suitable site for a TCP of 1st Order (see section Selection of Transmission Concentration Points 1st and 2nd order locations). This is due to the fact that all TCP of 2nd Order and several Node Bs will be connected to this site. Also, the building should have at least one Carrier available besides Group3Gs own Backbone, to be able to design the city rings connecting the TCPs of 1st Order, and to be able to connect the Node Bs directly to the RNC site.

In the hypothetical case that all Node Bs were connected directly to the RNC (star topology), a theoretical optimum location of the RNC, based on the transmission costs, can be calculated based on the locations of the Node Bs, with the following formula:

= ( ( ) / ()

= ( ( ) / ()

where (, ) and (, ) are the coordinates of the RNC and the Node B. This formula could be only applied when the transmission link costs depend in the distance and the capacity.

In case of different topologies, this model can be used as a guideline to find an optimum estimated RNC location.

2.1.8. RNC configuration regarding transmission

This section will be developed when the Suppliers answer to the Start-up Notice document.2.1.9. Node B configuration regarding transmission

This section will be developed when the Suppliers answer to the Start-up Notice document.

2.1.10. Redundancy concept

The concept to improve the reliability of a system with multiple construction of essential components of the system

This chapter presents a redundancy concept for access, regional and backbone networks of Group3G. The network shall be designed to support two-way-traffic-redundancy and two-media-traffic-redundancy, regarding to costs, reliability and availability.

2.1.10.1. 2.1.10.1.1.

2.1.10.1.2.

2.1.10.2. Path Redundancy

2.1.10.2.1. General definitions:

1+1 Switching: In 1+1 switching, traffic is transmitted over both, the worker and protection paths simultaneously, selection of the appropriate path is made at the receive end.

The protection path or channel card cannot be used for any other purpose other than protection because it is always carrying the same traffic as the worker path.

1:1 Switching: In 1:1 Switching, traffic is transmitted over either the worker or the protection path, a selection is made at the transmit and receive ends of the line to select the appropriate path.

The protection path or channel card can be used for a purpose other than protection, e.g. low priority traffic, (or as it is sometimes called, extra traffic or casual traffic). When switching takes place the traffic on the worker path and the protection path swap lines. In that case low priority traffic may be lost when it is switched to the faulty path.

Enhanced functions are 1:N or N+1 protection switching which work in the same manner as the two basic modes. In following figures the two possibilities are shown:

Fig.: Switching 1:1

Fig.: Switching 1+1

Typical switching criterias shall be:

There are fault conditions that would cause protection switching to occur.

Signal Fail (SF) (A high-priority fault is present)

Loss of Signal (LoS)

Loss of Frame (LoF)

Loss of Point (LoP)

Bit Error Rate (BER) > 10-,

Switching may also be operator controlled. This may be useful when diverting traffic from a active path or equipment so that maintenance work can be done.

For the beginning fully equipment redundancy shall only be used in backbone.

2.1.10.2.2. Access Network Protection

In the Access Layer (between TCP 1st and TCP 2nd and last Node B), no redundancy shall be used. There could be some exceptional cases of high density TCP 2nd Order, which has to be decided by case. Business case calculations are mandatory for choosing a second link for protection. Exceptional cases are hot spots like airports, fairs, etc.2.1.10.2.3. Regional Network Protection

Definition: Path protection is a dedicated protection mechanism that can be used on a physical structure, i.e. ring, meshed or mixed. It provides a means of protection against failure of the components of a path, including parts of the multiplexers as well as optical components, interconnecting fibres and intermediate regenerators and equipment. With a delay of 50ms the protection switch to a redundant link.

Definition: SNCP or Sub-network connection protection is based on path protection and can be used to protect a portion of a path, i.e. a portion where two separate paths segments are available, between two termination points. Sub-Network Connection is a linear protection scheme that can apply on a individual basis to VC-n signals

Path and trail protection shall be used in Regional network.

2.1.10.2.3.1. SDH Technology in Regional Networks: case of NortelEnd to end protection mechanism shall be used, that operate at VC-12, VC-3 or VC-4 level in combination with 1+1 SNCP functionality. The Virtual Containers (VC12, VC3, VC4) should be dropped out and in tributary units of the multiplexer. The aggregated VC4s (depending on used capacity) shall be transmitted via two ways. Especially signalling and O&M traffic shall be prioritised and totally protected.

Also media diversity (leased line and microwave) is the preferred in the Regional network. The distance between two microwave links inside one Regional network shall be more than 5 km. There shall not be two microwave links, part of the Regional ring, at one site unless they have an angle more than 60.

In the following figure, there is an overview of the regional network ring:

Fig.: Transmission network ring between TCPs and Backbone or RNC

For configuration of maximum 2 x STM1 for both aggregates, a small multiplexer and a single STM1 ring shall be chosen.

In case that the RNC is not located in the backbone site, the regional ring shall be directly connected to Group3Gs backbone. For terminate transmission regional network rings in backbone, both STM1 (east and west) shall be connected to backbone SNCP protected tributaries of backbone multiplexers, as shown in figure below:

Fig.: Configuration of connections between transmission network rings and backbone

2.1.10.2.3.2. ATM Technology in Regional Networks: case of EricssonDefinition: For dimensioning the needed capacity in E1, E3, or STM1 granularity and system performance, the traffic shall not be qualitative degraded in comparison to not using statistical multiplexing.

Redundancy shall be done by ATM load balancing. Each ATM switch shall have at least two line ports. Doubling of equipment for card protection purposes shall not be used. Protection switching shall be done by automatic ATM routing.

By dimensioning ATM Network rings, delay and traffic load are critical values. Because dimensioning depends on many parameters, there is an own guide to dimension it. Generally delay depends on connected ATM Switches in chain, adding delay to transmission. As Ericsson has stated, the maximum chain shall be 5 ATM switches between Node B and RNC.

In front of RNC a ATM switch shall be to concentrate traffic to RNC.

In case of Ericsson no additional ATM switches are needed. The ATM Switch in the Node B placed at TCP1 site shall be installed with 2 additional STM1 cards to connect to the regional ring.

In case of Nortel only a ATM switch in front of RNC is needed. At TCP1 site a Add drop multiplexer shall be placed to multiplex the traffic in the regional ring.2.1.10.2.4. Backbone Network

In SCNP redundancy the traffic between two backbone network terminations is transported through both direction to the destination, but the backup way is always reserved for the failure. Mostly the shorter way is used as primary (worker) way because it has at better quality (delay, jitters, wander, ...).All connection shall be fully SNCP 1+1 protected.

2.1.10.3. Equipment Redundancy

2.1.10.3.1. Card Protection

Definition: Types of card protection, Card protection can be 1+1, 1:1 or 1:N.

1+1 protection is where one worker card is backed up by one protection card. The protection card operates while waiting to be switched into operation.

1:1 protection is where one active card is backed up by one protection card. The protection card is idle until it is switched into operation.

1:N protection is wher a number of active cards are backed up by one protection card. If one of the active cards fails its function is transferred to the protection card. The remaining worker cards are unprotected until the fulty active card is replaced and placed back into operation. If all active cards are OK the protection card is idle.

Fig.: Configuration of card protection

In regional network, 1+1 card protection shall be used for power supply, matrix and control processor. No card protection shall be used in tributaries. Ports at two different cards shall be used to aggregate the traffic into the ring.

No card protection in the access (TCP1 TCP2 Node B) layer.

Fully Card protection is only used in backbone and only for important equipment. If a failure occurs on one of these cards, the function or traffic is transferred to the protection card. The faulty card can then be replaced without further interruption of traffic.

2.1.11. Quality planning objectives The design target for the Unavailability of the Transmission Network shall be according ITU R F-1493 with the following definitions: NodeB - TCP 2 - TCP 1TCP 1 RNC Core Network nodeRNC-RNC

Threshold BER=10E-6

Pua= 262min (0,05%) / yearThreshold BER =10E-6


Pua= 210min (0,04 %) / year

The unavailable time should be subdivided equally between the following events:


UA due to equipment


The design target for the Performance of the Transmission Network shall be ITU-T G.826 and ITU-R F 1491.

The Unavailability for Leased Lines can be different and depends on the contracts with

the Leased Line Carrier (e.g. SLA).The design target for the delay between the Node B and the RNC shall be:

SupplierNodeB - RNCRNC-MSC

EricssonOptimum: 5 ms

Maximum: 20 msOptimum: 2 ms

Maximum: 10 ms

NortelMaximum: 15 msMaximum: 10 ms

This delay shall be calculated only taking into account the transmission equipment (microwave, transmission media, ATM switches, etc). If two Node Bs are chained, the additional delay while multiplexing the traffic of both Node Bs should also be included in this calculation.2.2. INFRA SHARING CONTRACT WITH E-PLUS 2.2.1. General aspects

The Responsible Party (Group3G or E-Plus) is in charge to provide transmission capacity to the Node Bs of the basic grid in its (Sub-)Regions. The (Sub-) Regions are mentioned in the Main Body of the ISA contract.

Each Party is responsible to connect the Node Bs to the RNC and/or PoPC in his assigned (Sub-)Regions meeting the agreed QoS criteria as defined in Annex 7 (QoS), including the planning, construction, operation and maintenance activities according to the definitions in the Annex 5 (Transmission) and to the common planning principles as agreed in the respective supplements. It is the responsibility of each Party to find the optimum solution to meet the agreed transmission capacity and quality. For avoidance of doubt the network of the Responsible Party can consist of own links and leased lines; also the Other Party can supply capacity or links.

Group3G has to use E-Plus existing transmission capacity if E-Plus has got 1xE1 available from NodeB to RNC and the price and conditions justifies this.

Group3G uses to a reasonable extent the existing and commonly with E-Plus planned (future) Regional Network infrastructures. This means e.g. the topology, the capacity and the costs have to justify the use of the existing and commonly with E-Plus planned (future) Regional Network infrastructures.

The Annex 5 (Transmission) is completed by the following supplements:

1. Common Network Topology and Planning Principles. It describes the technical details related to the Access Layer and the Regional Layer of the Transmission Network.

2. Pricing Models and Procedures. It describes all cases, that require a commercial solution as well as the according procedures. The pricing details are defined in Annex 6 (Commercial).

3. Connection Points. It lists the geographical locations for UPoNRC, PoISC and PoPC.

4. Realization and Processes. It describes the technical realization, responsibilities, procedures, etc. for the physical connection of transmission capacity between both Parties.

The main subjects are included in this Planning recommendation. However the supplements can be requested by the regional planner if needed.

The demand of transport capacity to the last hop to one physical Node B is based on the NodeB dimensioning as defined in the Annex 1 (UTRAN) as amended from time to time.

The target is that the Responsible Party will provide the following Recommended Capacities for UMTS in the Transmission Network on the last hop to one physical Node B:Recommended Capacity for UMTS on

Existing link or

Leased line

(to be revised twice a year)new link or

upgraded link

(to be revised every four years or on demand)

link (last hop) to the last Node B1 E1

(0,5 E1 per Operator)8 E1

(4 E1 per Operator)

TrunkSee Supplement Common Network Topology and Planning PrinciplesSee Supplement Common Network Topology and Planning Principles

The Responsible Party has the right to provide less capacity than the Recommended capacity for UMTS as long as the minimum requirements for Quality of service are met as defined in Annex 7 (QoS) and mentioned in this document.

Quality of service of Annex 7:

KPI Unavailability in transmission network shall be according the defined values in Supplement Common Planning Princip

Documents

Transmission Planning Handbook v0.3