Technical Handbook. Wholesale NGN Ethernet Products

Transcription

1 Technical Handbook Wholesale NGN Ethernet Products

2 Document Control Revision History Version Date Revised by Revision details rd rd February 2010 February th March th October th December th February th March 2012 Initial version IBH and ISH solutions added Information on Planning Ratios added 1. Traffic based CoS option added 2. Correction made to ISH distances in Section WEIL max service limits added in Table 3: 4. Traffic based COS EtherType restriction in Section 8 5. Clarification on OAM PDU support added to Capability for Operators to optionally specify S-VLAN ID at E-NNI added to Section Additional network performance figures added to Section Information on request for configuration of loopbacks on WSEA connections added in Section WEIL Ethertype options added to Table 20, Table 21, and Table 22 in Section 8 3. Statement added to end of Section regarding Operator obligations associated with specifying S-VLAN IDs at E-NNI. Addition of the following: 1. WEIL ENH 2. Increased limits for EF and AF 3. Updates for APT 4. WSEA IBH 5. Revised technical performance figures. 6. Multi-service WSEA 7. circuit based class of service Option 6, 100% AF 8. WEIL ENH Added p-bits 6, nd Sept Added 10G WSEA 2. Added ALU 7210 SAS-M 10GigE NTU option 3. Added new section on NGN Capacity Management 05/05/2015 Version 1.8 Page 1

3 Version Date Revised by Revision details Technical Handbook Wholesale NGN Ethernet Services th May Updated to fix table numbering issues 2. Added NGN Bitstream Ethernet Connectivity Service (BECS) 05/05/2015 Version 1.8 Page 2

4 Table of Contents Technical Handbook Wholesale NGN Ethernet Services 1 Introduction Products Overview Symetric Ethernet Access Bitstream Ethernet Connectivity Service Network Solution Overview Symetric Ethernet Access WSEA (UNI) WEIL (E-NNI) Traffic Flow WEIL (E-NNI) WSEA (UNI) Traffic Flow WSEA WEIL Connection Scenarios VLAN Tagging Model Bitstream Ethernet Connectivity Service Layer 3 Redundant Connectivity IP Routing for the Redundant Solution Operator configuration for BGP Service Parameters Symmetric Ethernet Product WSEA and WEIL Bitstream Ethernet Connectivity Service Bandwidth Profiles WSEA Bandwidth Profile WEIL Bandwidth Profiles Bitstream Ethernet Connectivity Service WSEA CoS Models Circuit-based QoS Traffic-based QoS WEIL CoS Model Mapping of UNI - E-NNI CoS Markings Traffic Based QoS UNI to E-NNI Traffic Flows E-NNI to UNI Traffic Flows Network Performance NGN Capacity Management Planning Ratios WSEA Interface Specification WSEA Customer Sited Handover (CSH) WSEA In Building Handover (IBH) Multi-service WSEA Mapping of WSEA logical Services to Different WEILs WEIL Customer Sited Handover (CSH) WEIL In Span Handover (ISH) WEIL In Building Handover (IBH) WEIL Edge Node Handover (ENH) WEIL ENH 1+1 Protection /05/2015 Version 1.8 Page 3

5 List of Figures Figure 1: Products Overview Symetric Ethernet Access...78 Figure 2: Products Overview FTTC VDSL Ethernet AccessError! Bookmark not defined.9 Figure 3: UNI E-NNI Traffic Flow Figure 4: E-NNI UNI Traffic Flow Figure 5: WSEA/WEIL Connection Scenarios Figure 6: VLAN Tagging Model Figure 7: UNI -> E-NNI Traffic Flow... Error! Bookmark not defined.15 Figure 8: E-NNI -> UNI Traffic Flow... Error! Bookmark not defined.16 Figure 9: VLAN Tagging Model... Error! Bookmark not defined.17 Figure 10: WEIL Service Access Bandwidths Figure 11: Network Forwarding Classes / Queues Figure 12: WEIL CoS Model Figure 13: Mapping of UNI - E-NNI CoS Marking Figure 14: RAD ETX-202A Figure 15: ALU 7210 SAS-M 10GigE Figure 16: WSEA Customer Sited Handover (CSH) Figure 17: WSEA In Building Handover Figure 18: WSEA Multiple UNI Ports Figure 19: WSEA with Multiple Ports Figure 20: Mapping of WSEA logical Services to Different WEILs Figure 21: WEIL Customer Sited Handover (CSH) Figure 22: WEIL In Span Handover (ISH) Figure 23: Minimum Power Levels 1G In Span Interface - ZX SFP Figure 24: Minimum Power Levels 10G In Span Interface ER XFP Figure 25: Minimum Power Levels 10G In Span Interface - ZR XFP Figure 26: WEIL In Building Handover (IBH) Figure 27: Minimum Power Levels - 1Gbit/s In Building Interface Figure 28: Minimum Power Levels - 10Gbit/s In Building Interface Figure 29: WEIL Edge Node Handover Figure 30: 10G ENH Minimum Power Levels Figure 31: WEIL with 1+1 Protection Figure 32: Active-Standby WEIL Design... Error! Bookmark not defined. Figure 33: Active-Active WEIL Design... Error! Bookmark not defined. Figure 34: Direct Fibre Access to NGN Aggregation node Figure 35: Fibre Access via an Intermediate Exchange Figure 36: NGN Extended Reach using APT Figure 37: Aggregation node to PE connection Figure 38: APT connectivity to NGN /05/2015 Version 1.8 Page 4

6 List of Tables Table 1: WSEA and WEIL Service Parameters...16 Table 2: Layer 2 Control Protocol Processing...17 Table 3: WEIL Maximum Service Number Limits...17 Table 4: WSEA Bandwidth Options (Mbit/s)...19 Table 5: SAB Bandwidth Options (Mbit/s)...21 Table 6: NGN BECS Bandwidth Options (Mbit/s)...22 Table 7: Circuit-based QoS Options...25 Table 8: Traffic-based CoS...25 Table 9: Traffic-based CoS - CIR and PIR Values...26 Table 10: Traffic-based CoS Options...26 Table 11: Aggregate EF and AF Bandwidth Options...27 Table 12: Ingress Mapping of CoS Markings...29 Table 13: Egress Mapping of CoS Markings...29 Table 14: Ingress Mapping of CoS Markings...30 Table 15: Indicative Target Network Performance Figures...31 Table 16: Planning Ratios...34 Table 17: WSEA NTU Accommodation Requirements...36 Table 18: WSEA UNI Interface Specification...37 Table 19: WSEA NTU Accommodation Requirements...38 Table 20: UNI specification for WSEA...39 Table 21: NTU Port Configuration...41 Table 22: CSH NTU Accommodation Requirements...43 Table 23: CSH E-NNI Interface Specification...45 Table 24: 10G CSH NTU Accommodation Requirements...45 Table 25: 10G CSH E-NNI Interface Specification...46 Table 26: ISH E-NNI Interface Specification...49 Table 27: IBH E-NNI Interface Specification...51 Table 28: ENH E-NNI Interface Specification...53 Table 29: LAG Parameters: WEIL with 1+1 Protection /05/2015 Version 1.8 Page 5

7 Glossary AF APT CIR CoS C-VLAN DSLAM EF E-Line ENH E-NNI EXP FC FTTC IBH ISH L2CP LAG LAN LLF MPLS MTU NGN NTU ODF PIR PoH QoS SAB SFP STD S-VLAN TLI UNI VLAN VLL WEIL WSEA Technical Handbook Wholesale NGN Ethernet Services Assured Forwarding Access Packet Transport Committed Information Rate Class of Service Customer Virtual Local Area Network Digital Subscriber Line Access Multiplexer Expedited Forwarding Ethernet Line Edge Node Handover External Network to Network Interface Experimental Forwarding Class Fibre to the Cabinet In-building handover In-span handover Layer 2 Control Protocol Link Aggregation Group Local Ethernet Network Link Loss Forwarding Multiprotocol Label Switching Maximum Transmission Unit Next Generation Network Network Termination Unit Optical Distribution Frame Peak Information Rate Point of Handover Quality of Service Service Access Bandwidth Small Form-factor Pluggable Standard Service - Virtual Local Area Network Transparent LAN Interconnect User Network Interface Virtual Local Area Network Virtual Leased Line Wholesale Ethernet Interconnect Link Wholesale Symmetrical Ethernet Access 05/05/2015 Version 1.8 Page 6

8 1 Introduction The purpose of this document is to provide a technical description of the Wholesale Next Generation Network (NGN) based Ethernet products in order to assist Operators in the design and development of their own product offerings. Please note that this is a working document and therefore subject to regular updates as new products and product enhancements are introduced. 2 Products Overview The section provides a high level overview of the Wholesale Symmetrical Ethernet Access (WSEA) product, the Wholesale Ethernet Interconnect Link (WEIL) product, and the Bitstream Ethernet Connectivity Service (NGN-BECS). 2.1 Symetric Ethernet Access Customer Site NGN Network Operator Handover Site NTU NGN Node NGN Node NTU WSEA WEIL WSEA Logical (WES) Figure 1: Products Overview Symetric Ethernet Access The Wholesale NGN Symmetric Ethernet products consist of a number of components: 1. Wholesale Symmetrical Ethernet Access (WSEA) provides physical connectivity from a customer site to the NGN o Wholesale Ethernet Service (WES) is the WSEA logical connection providing an E-line service between an Operator s end user site and the Operator s handover site 2. Wholesale Ethernet Interconnect Link (WEIL) provides physical connectivity from an Operator s handover site to the NGN Multiple WSEA logical (WES) connections may be aggregated within the NGN network and handed-over to an Operator on a WEIL. Multiple WSEA logical connections may be supported on an individual physical WSEA with each WSEA logical connection terminated on a single WEIL or different WEILs. 05/05/2015 Version 1.8 Page 7

9 2.2 Bitstream Ethernet Connectivity Service NGN Operator Site IES 7750PE 7450ESS NTU WEIL (EIL) Figure 1: NGN BECS The Wholesale NGN BECS service consists of the following components: 1. Wholesale Ethernet Interconnect Link (EIL) provides (physical) connectivity from an Operator s handover site to the 7450ESS. 2. Logical connection providing handover of ADSL Bitstream L2TP traffic from ADSL DSLAMs to an Operator s IP network. Two options are available for the logical connection: NGN BECS Bitstream (NBB) NGN BECS Bitstream Managed Backhaul (NBM) a. NGN BECS Bitstream logical connection (NBB) b. NGN BECS Bitstream Managed Backhaul logical connection (NBM) For purposes of clarity, NGN BECS is a layer 3 service and will be available on all EIL types (e.g. ISH, IBH, CSH and ENH) and whether the EIL is configured as a port or a lag. 05/05/2015 Version 1.8 Page 8

10 3 Network Solution Overview Technical Handbook Wholesale NGN Ethernet Services This section provides a high-level technical overview of how the Wholesale NGN Ethernet products are supported on the NGN network and illustrates the interaction of the products. 3.1 Symetric Ethernet Access WSEA (UNI) WEIL (E-NNI) Traffic Flow Figure 2: UNI E-NNI Traffic Flow The following describes how customer traffic is treated in the UNI E-NNI direction: 1. End User traffic is presented to the network at a physical port (UNI) on an managed NTU which is installed at the end user site. The UNI port is configured as an 802.1Q trunk. The WSEA product is a Transparent LAN Interconnect (TLI) service and hence the end user traffic presented at the UNI may be tagged or untagged. The end user can present double tagged frames at the UNI. The C-VLAN tag(s) are carried transparently across the network. 2. Only one UNI port on the NTU may be used for end user traffic per WSEA logical connection. The 10GigE NTU is limited to one UNI port so Multiple WSEA connections are not supported. An S-VLAN tag is added to the end user traffic at the UNI on the NTU. The assigned S-VLAN ID is not visible to either the end user or the Operator. 3. A service policy (i.e. CoS profile/bandwidth) is applied to end user traffic associated with the S-VLAN and the end user traffic is mapped to the appropriate Forwarding Class (FC) within the core NGN network (Please refer to Section 5.1 for the WSEA bandwidth options and Section 6.1 for a description of the CoS solution). The S-VLAN tag is removed and the end user traffic is carried within a Virtual Leased Line (VLL) across the Core NGN network. 4. An S-VLAN tag is added to the end user traffic on egress of the Core NGN network and the appropriate service policy is applied to the end user traffic. The end user traffic is passed to the managed NTU located at the Operator handover site. 5. The S-VLAN on the network-facing port is mapped to the Operator facing port on the NTU (E-NNI port). The E-NNI port is configured as an 802.1ad port. The S- 05/05/2015 Version 1.8 Page 9

11 VLAN ID is assigned by or selected by the Operator and is used to identify the end user traffic associated with an individual WSEA logical connection. In the event that an Operator has more than one WEIL, an Operator must specify which WEIL is to be associated with each WSEA logical connection. The WEIL can only support services that originate in an end user site WEIL (E-NNI) WSEA (UNI) Traffic Flow Figure 3: E-NNI UNI Traffic Flow The following describes how Operator traffic is treated in the E-NNI UNI direction: 1. Operator traffic is presented to the network at a physical port (E-NNI) on an managed NTU located at the Operator handover site. The E-NNI port is configured as an 802.1ad port. The Operator must add an S-VLAN tag to their traffic prior to presentation at the E-NNI. The S-VLAN tag is assigned by or selected by the operator and is associated with the destination WSEA connection. 2. A service policy (i.e. CoS profile/bandwidth) is applied to the traffic associated with the S-VLAN and the Operator traffic is mapped to the appropriate Forwarding Class (FC) within the Core NGN network. The S-VLAN tag is removed and the Operator traffic is carried within a Virtual Leased Line (VLL) across the Core NGN network. 3. An S-VLAN is added to the Operator traffic and the appropriate service policy is applied. The assigned S-VLAN ID is not visible to either the end user or the Operator. 4. The S-VLAN tag is removed from the Operator traffic on the NTU. 5. Operator traffic is presented to the end user at the UNI port on the NTU WSEA WEIL Connection Scenarios The previous sections described at a high-level the network solution whereby the WSEA and WEIL connections are provided off different NGN Nodes and the WSEA logical connection is carried across the NGN core network. In certain cases the WSEA and its associated WEIL connection may be within the same serving exchange area and therefore may be connected to the same NGN Node. The technical description of the Wholesale NGN Ethernet products in this document is based on different NGN Nodes being used for the WSEA/WEIL connections (i.e. Scenario A below) in order to provide a comprehensive description of the products. 05/05/2015 Version 1.8 Page 10

12 Scenario A WSEA NGN Network WEIL NGN Node NGN Node Scenario B WSEA WEIL NGN Node Figure 4: WSEA/WEIL Connection Scenarios However, the technical functionality and parameters associated with the Wholesale NGN Ethernet products which are described in this document are unchanged if the WSEA and WEIL are both connected to the same NGN Node as per scenario B above. WSEA/WEIL connections served off the same NGN Node provide the capability to configure the WSEA logical connection as an uncontended connection. The NGN Node is a non-blocking device and therefore the Operator may inventory manage the WEIL to ensure that it is not overbooked, i.e. that the sum of the WSEA bandwidths do not exceed the Service Access Bandwidth on the WEIL (see Section 5.2 for description of Service Access Bandwidths). In this scenario QoS is not required on the WSEA as the connection is uncontended and is not carried across the core NGN network (Circuit based QoS Option 5 as described in Section should be specified for these connections). Conversely, the Operator may inventory manage the Wholesale Ethernet Interconnect Link which may result in the Wholesale Ethernet Interconnect Link being overbooked, i.e. that the sum of the Wholesale Symmetrical Ethernet Access bandwidths exceed the Service Access Bandwidth on the Wholesale Ethernet Interconnect Link (see Section 5.2 for description of Service Access Bandwidths). In this scenario QoS may be required on the Wholesale Symmetrical Ethernet Access and the required level of circuit based class of service should be specified by an Operator. It should be noted that more than one NGN Node may be installed in some serving exchange sites. These NGN Nodes within the same serving exchange site will not be connected locally VLAN Tagging Model This section summarises the VLAN tagging model used for the Wholesale NGN Symmetric Ethernet products: 05/05/2015 Version 1.8 Page 11

13 NGN Network Technical Handbook Wholesale NGN Ethernet Services CPE S-VLAN S-VLAN Operator Network Customer Site NTU NGN Node NGN Node NTU Operator Handover Site Frame frame S frame frame S frame S Untagged Customer Traffic OR Frame Frame C1 C2 Frame C1 Frame C1 Frame C2 S Frame C2 Frame C1 Frame C1 Frame C2 S Frame C2 S Tagged Customer Traffic Figure 5: VLAN Tagging Model The Operator can use the full C-VLAN ID range (i.e ) to tag their end user traffic. The C-VLAN IDs are carried transparently across the network. Default is for to assign the S-VLAN ID presented at the E-NNI. The assigned S- VLAN IDs will be in the range The Operator can optionally specify the S-VLAN ID presented at the E-NNI. If the Operator chooses to specify their own S-VLAN ID s on a specific E-NNI, then the Operator will be responsible for specifying all S-VLAN IDs within the range on that E-NNI. 3.2 Bitstream Ethernet Connectivity Service The Bitstream Ethernet Connectivity Service (BECS) is used for handing over standard (NGB and Legacy) Wholesale Bitstream traffic to an Operator using L2TP. BECS is not used for hand-over of the Bitstream Ethernet Access (BEA) service which is described in this document. For simplicity, the description is based on an NTU being deployed on the EIL connection. However, as described in the previous sections it is possible to have an EIL connection with no NTU. 05/05/2015 Version 1.8 Page 12

14 Bitstream Traffic Modem ADSL DSLAM BRAS Operator Handover Site Modem Modem ADSL ADSL DSLAM DSLAM Aggregation Network BRAS BRAS L2TP Tunnels IP Core NGN Node NTU E-NNI 1G/10G Optical/Electrical Port Operator Network (LNS) Figure 2: NGN BECS Traffic Flow The following describes how End User traffic is treated for the NGN BECS service: 1. The Operator s customers connect to an ADSL line with PPPoE credentials on the ADSL modem set according to the Operator s requirements (e.g. username@oao1.ie, password) 2. The Bitstream traffic is forwarded from the DSL Access Multiplexer (DSLAM) to the BRAS nodes. First stage End User authentication is performed on the BRAS based on the realm associated with the Operator 3. An L2TP tunnel is established between the BRAS (LAC) and the Operator BRAS (LNS) based on the RADIUS profile associated with the realm. The L2TP traffic is routed over the IP core to the BECS handover site. 4. At the BECS handoff an S-VLAN is added to the traffic and the appropriate service policy (e.g. CoS profile/bandwidth) is applied. All of the L2TP traffic for a particular NBB or NBM connection is presented on this S-VLAN. 5. The BECS traffic is passed to the managed NTU located at the Operator s handover site. The S-VLAN on the network-facing port is mapped to the Operator facing port on the NTU (E-NNI port). The E-NNI port is configured as an 802.1ad port. The S-VLAN ID is assigned by and is used to identify the end user traffic associated with an individual NBB or NBM connection. The Operator can optionally specify the S-VLAN ID presented at the E-NNI. 6. BECS traffic is handed off to the Operator s IP network where it is routed to the Operator s LNS to terminate the L2TP tunnels and perform second stage authentication for the end users PPP session. The BECS product supports the hand-off of legacy Bitstream (NBB) and Bitstream Managed Backhaul (NBM) traffic types on the same physical BECS connection. The handoff of both types of traffic is on separate S-VLANs. 05/05/2015 Version 1.8 Page 13

15 NGN Individual WES Logical Connections SAP SAP SAP E-NNI NGN BECS Connections Bitstream BMB SAP SAP NTU 7450ESS Figure 3: Support for BMB and Legacy Bitstream Traffic on EIL Individual NGN BECS connections are required to carry the Legacy Bitstream (NBB) and Bitstream Managed Backhaul traffic (NBM) on the same EIL. Usage based measurements are required for the BMB traffic and therefore this traffic cannot be carried in the same service as the Legacy Bitstream traffic. SABs Layer 3 Redundant Connectivity In order to offer increased levels of resilience to their End Users, Operators can add further physical connections to the IP network. In the event of one link failing, the remaining link(s) will carry the IP traffic. Bitstream traffic is only permitted to flow to and from the Operator over these specific links. In summary, the redundant solution will permit the BRAS s to build multiple L2TP tunnels to multiple Operator BRAS s (LNS s). Individual PPP sessions will be carried in these L2TP tunnels. The BRAS actually selects the LNS randomly from the list, but where there is a high number of PPP sessions, the distribution becomes evenly spread across the LNSs. BRAS IP Network L2TP Tunnels Operator IP Network ADSL Network NGN BECS NGN BECS NGN BECS Operator Edge Router 1 Operator Edge Router 2 Operator Edge Router 3 LNS 1 LNS 2 L2TP Tunnels Figure 4: Load-balancing Network Map When an Operator s End User enters their username and password, the PPP session causes the BRAS to lookup one of the DSL RADIUS servers. Based on the domain name entered by the End User, the RADIUS server will send back the IP addresses of the Operator s LNS devices. 05/05/2015 Version 1.8 Page 14

16 The first PPP session to each LNS will cause the BRAS s to build L2TP tunnels to each of these IP addresses. Subsequent PPP sessions to each LNS will travel within the existing L2TP tunnel. The BRAS s will load-share PPP sessions into the L2TP tunnels IP Routing for the Redundant Solution An ebgp peering is established across each link between and the Operator. The ebgp peering uses an MD5 password. Each Operator router advertises the IP addresses of every L2TP tunnel termination point to. Each edge router only advertises the IP addresses of each required BRAS tunnel initiation point to the Operator. Routing of traffic from the Operator to is under the control of the Operator. sets the MED to 100 for all BRAS prefixes. By setting the MED to the same value for all prefixes, no link appears closer than another. The use of the BGP MED (Multi-Exit Discriminator) parameter allows individual paths to be preferred for individual prefixes. For example, on the first link the Operator could advertise the IP addresses as follows: A.B.C.D/32 with MED of 10 A.B.C.E/32 with MED of 20 On the second link, the Operator would advertise the IP addresses as follows: A.B.C.D/32 with MED of 20 A.B.C.E/32 with MED of 10 In normal operation the network will see two possible paths to each individual prefix, and the path advertising the lower MED will be preferred. In the event of the preferred link failing, the path with the less preferred higher MED will automatically be selected. Hence, in normal operation the outbound L2TP tunnel to A.B.C.D will prefer the first link; the outbound L2TP tunnel to A.B.C.E will prefer the second link Operator configuration for BGP The Operator must use their own public AS number. Private AS numbers are not used. The Operator (Peer) announces the /32s only for their L2TP LNS (home gateways) announce /32s only for the BRAS loopbacks. Peer should not also announce their /32s at INEX as this could cause traffic to go the wrong way and be blackholed. will honor the MED values the Operator sends to unless there is an internal traffic engineering (capacity) requirement for to sent the traffic back a different way. will announce all of the BRAS /32s with MED 100 so that they appear equidistant. Peer has control of which peering they send traffic to on. If there are any capacity management issues this may have to be tweaked in co-operation with to ensure optimum performance. 05/05/2015 Version 1.8 Page 15

17 4 Service Parameters 4.1 Symmetric Ethernet Product WSEA and WEIL This section summarises the service parameters associated with the WSEA and WEIL products: Parameter Value MAC Address learning Off Max Frame Size 9000 bytes Max no. of S-VLANs per E-NNI 3990 Max number of C-VLANs per UNI 4096 C-VLAN ID Preservation Yes C-VLAN CoS Preservation Yes Multicast traffic limit No limit Broadcast traffic limit No limit Unknown Unicast traffic limit No limit Table 1: WSEA and WEIL Service Parameters Protocol Behaviour Spanning Tree Protocol (STP), Rapid Spanning Tree Protocol (RSTP), Multiple Spanning Tree Protocol (MSTP) Tunnelled PAUSE (802.3x) Link Aggregation Control Protocol (LACP) Discarded Tunnelled Marker Protocol Tunnelled 05/05/2015 Version 1.8 Page 16

18 Authentication (802.1x) All LANs Bridge Management Group Block of Protocols Tunnelled Tunnelled Generic Attribute Registration Protocol (GARP) Block of Protocols Tunnelled Cisco Discovery Protocol (CDP) Cisco VLAN Trunking Protocol (VTP) Tunnelled Tunnelled Table 2: Layer 2 Control Protocol Processing Due to the finite minimum buffer allocation assigned to a WEIL port, Table 3: details the theoretically maximum number of service that can be supported on a WEIL without service degradation being experienced. Queue PIR/CIR Default Buffer allocation No. of Services Supported 10M M M M 18 1G 9 Table 3: WEIL Maximum Service Number Limits 4.2 Bitstream Ethernet Connectivity Service This section summarises the service parameters associated with the Wholesale NGN BECS service: Parameter CoS Offering Value All NGN BECS traffic will be mapped to the Best Effort queue. Bandwidth Sharing SAB Bandwidth may be shared across S-VLANs or dedicated to each service. 05/05/2015 Version 1.8 Page 17

19 Service Bandwidth Dedicated bandwidth per S-VLAN or shared across both (i.e. bandwidth sharing). Routing Options AS Number BGP password LNS/HGW Addresses BGP Operator provides their own AS number Default: Operator provides BGP password Operator provides their LNS/HGW IP addresses (/32s) Routing Table Options No Routing table options supported. advertises all B-RAS loopbacks (loopback0 for Bitstream, loopback2 for BMB) MTU size bytes Layer 2 S-VLAN Tagging Layer 2 Access Model Tagged Tagged EPL type access Service Type Required S-VLAN allocation A single SAB may support either Bitstream or BMB services or both (2 S-VLANs) Default: Assigned by Eircom Option: Provided by Operator Table 1: NGN BECS Service Parameters 5 Bandwidth Profiles 5.1 WSEA Bandwidth Profile The WSEA bandwidth options for 1G and 10G physical access are shown in the following table: 1 Gbit/s SEA Bandwidth (Mbit/s) 10 Gbit/s SEA Bandwidth (Mbit/s) 10* 10 1 The 1620 byte setting permits a standard customer MTU of 1500 bytes and the associated L2TP header. 05/05/2015 Version 1.8 Page 18

20 1 Gbit/s SEA Bandwidth (Mbit/s) 10 Gbit/s SEA Bandwidth (Mbit/s) 20* 20 30* 30 40* 40 50* 50 75* * * * * Table 4: WSEA Bandwidth Options (Mbit/s) Bandwidth limits apply to WSEA connections when associated with the selected QoS option (please refer to Section 6.1 for a description of the QoS options): One of these values (10M to 300M) will have to be selected if the WSEA connection is delivered over an APT system. In all other cases the WSEA logical bandwidth will default to 1 Gbit/s for a 1G WSEA or 10 Gbit/s for a 10G WSEA. 05/05/2015 Version 1.8 Page 19

21 Technical Handbook Wholesale NGN Ethernet Services 300 Mbit/s upper limit will apply to the amount of EF traffic which can be ordered on an individual 1 Gbit/s WSEA connection. 600 Mbit/s upper limit will apply to the amount of AF traffic which can be ordered on an individual 1 Gbit/s WSEA connection. 600 Mbit/s upper limit will apply to the sum of EF and AF traffic which can be ordered on an individual 1 Gbit/s WSEA connection (300 Mbit/s upper limit for EF traffic still applies). 3 Gbit/s upper limit will apply to the amount of EF traffic which can be ordered on an individual 10 Gbit/s SEA connection. 6 Gbit/s upper limit will apply to the amount of AF traffic which can be ordered on an individual 10 Gbit/s SEA connection. 6 Gbit/s upper limit will apply to the sum of EF and AF traffic which can be ordered on an individual 10 Gbit/s SEA connection (3 Gbit/s upper limit for EF traffic still applies). The above limits apply to the aggregate of the WSEA Logicals supported on a WSEA physical. For Circuit Based CoS the WSEA logical bandwidth is used for the calculation of the upper limits. For Traffic Based CoS the % values of traffic mapped to the EF and AF queues is used for the calculation of the upper limits. The bandwidth values listed in Table 4:Table 6: above include the Ethernet frame overhead, preamble, and interframe gap. The bandwidth/throughput on the WSEA connection is not dependent on frame size. 5.2 WEIL Bandwidth Profiles The Wholesale NGN Ethernet products may use one or more WEIL Service Access Bandwidths (SABs) on a single physical WEIL bearer as shown in the flowing diagram: NGN Network WEIL Service Access Bandwidth 1 WEIL Service Access Bandwidth 2 Individual WSEA Logical Connections E-NNI NTU WEIL Service Access Bandwidth 3 NGN Node Figure 6: WEIL Service Access Bandwidths The Operator must specify which one of the WEIL Service Access Bandwidths is associated with a WSEA logical connection on a per-order basis. 05/05/2015 Version 1.8 Page 20

22 The Operator may specify a maximum of 10 EIL Service Access Bandwidths on a 1 Gbit/s EIL bearer or a maximum of 20 EIL Service Access Bandwidths on a 10 Gbit/s EIL bearer. The Operator is required to specify the WEIL Service Access Bandwidth(s) on the WEIL bearer. The WEIL Service Access Bandwidth options are shown in the following section. 1 Gb/s WEIL Service Access Bandwidths (Mbit/s) 10 Gb/s WEIL Service Access Bandwidths (Mbit/s) Table 5: SAB Bandwidth Options (Mbit/s) The sum of the WEIL Service Access Bandwidths which share the same physical WEIL bearer cannot exceed the physical speed of the connection (i.e. 1 Gbit/s or 10Gbit/s). 5.3 Bitstream Ethernet Connectivity Service The NGN BECS bandwidth options are shown in the following table: 1 Gbit/s EIL 10 Gbit/s EIL /05/2015 Version 1.8 Page 21

23 Table 6: NGN BECS Bandwidth Options (Mbit/s) As part of the BECS order, the Wholesale Operator will have to specify both the EIL and SAB reference IDs (and SAB bandwidth) to which the NGN BECS services are associated. NBM and NBB S-VLANs can be ordered against the same SAB or individual SABs. Refer to section 5.2 for a description of SABs NGN EIL Service Access Bandwidth 1 SAP Individual WSEA Logical Connections SAP SAP E-NNI NGN BECS Connections Bitstream BMB SAP SAP NTU EIL Service Access Bandwidth ESS EIL Service Access Bandwidth 3 Figure 5: Non-Bandwidth Sharing Model NGN BECS In non-bandwidth sharing mode the Legacy Bitstream and BMB NGN BECS services are each associated with their own dedicated SAB. 05/05/2015 Version 1.8 Page 22

24 NGN EIL Service Access Bandwidth 1 Individual WSEA Logical Connections SAP SAP SAP E-NNI NGN BECS Connections Bitstream BMB SAP SAP NTU EIL Service Access Bandwidth ESS Figure 6: Bandwidth Sharing Model NGN BECS In bandwidth sharing mode the Legacy Bitstream and BMB NGN BECS services are both associated with the same SAB. 05/05/2015 Version 1.8 Page 23

25 6 Class of Service (CoS) Technical Handbook Wholesale NGN Ethernet Services This section describes the basic Class of Service (CoS) design for the Wholesale NGN Ethernet products. It is broken down into the CoS design on the WSEA connections and the CoS design on the WEIL connections. There are three forwarding classes, or network queues, used within the Core NGN network for the Wholesale NGN Ethernet products: the Expedited Forwarding (EF) class, the Assured Forwarding (AF) class, and standard/best effort (STD) class. NGN Network WSEA EF AF STD WEIL Figure 7: Network Forwarding Classes / Queues The Expedited Forwarding class is serviced before the Assured Forwarding class and is intended to be used for real-time delay-sensitive traffic. There is a committed information rate (CIR) associated with the Expedited Forwarding class. End user traffic which exceeds the configured CIR will be dropped on ingress to the Core NGN network. The Assured Forwarding Class is serviced before the Standard Forwarding class and is intended for business applications which require priority access to available bandwidth over standard applications. There is a committed information rate (CIR) and peak information rate (PIR) associated with the Assured Forwarding class. The Assured Forwarding class provides the ability to classify ingress traffic as either in-profile or out-of-profile based upon the traffic arrival rate. A queue is considered in the in-profile state if the rate at which the queue is being serviced is less than its configured CIR. A queue is considered out-of-profile if the rate at which the queue is being serviced is greater than its CIR, but less than its PIR. After the profile state of the packet is determined at network ingress, the profile state of the packet influences the packets queuing priority and drop preference. The Standard forwarding class is used for carrying all remaining traffic. This remaining traffic generally uses protocols that are capable of maintaining some form of connectivity during congestion. For the Wholesale NGN Ethernet products the end user traffic is mapped to the appropriate forwarding class on ingress to the Core NGN network for both the WSEA and WEIL connections. 6.1 WSEA CoS Models There are two CoS options available for the WSEA product; Circuit-based QoS and Trafficbased QoS which are described in more detail in the following sections. Only one of these options can be selected for an individual WSEA logical connection Circuit-based QoS With circuit-based QoS will not inspect the end user 802.1p CoS markings (if present) on network ingress. All customer traffic is mapped to a single forwarding class on 05/05/2015 Version 1.8 Page 24

26 ingress to the Core NGN network. The following options are available for circuitbased QoS: CIR PIR Queuing Option 1 CIR = 100% PIR=100% All traffic mapped to EF queue Option 2 CIR = 50% PIR=100% All traffic mapped to AF queue Option 3 CIR = 10% PIR=100% All traffic mapped to AF queue Option 4 CIR = 5% PIR=100% All traffic mapped to AF queue Option 5 CIR = 0% PIR=100% All traffic mapped to STD queue Option 6 CIR = 100% PIR=100% All traffic mapped to AF queue Table 7: Circuit-based QoS Options The percentage values in Table 7 refer to the percentage of the WSEA bandwidth. The end user 802.1p CoS markings (if present) will be carried transparently across the network and will not be re-marked by Traffic-based QoS With traffic-based QoS the end user marks the 802.1p bits in their Ethernet frame headers on network ingress. End user traffic is mapped to a forwarding class on ingress to the Core NGN network based on the 802.1p markings. The following table shows the policy map associated with the traffic-based CoS option: Policy Map Name Policy Map (allocation and allowed values) 802.1p CoS 7/6/5/4 3/2 All traffic mapped to EF queue Flexible CoS A % B % All traffic mapped to AF queue Table 8: Traffic-based CoS Traffic with 802.1p marking of 7 or 6 or 5 or 4 will be mapped to the EF queue Traffic with 802.1p marking of 3 or 2 will be mapped to the AF queue Traffic with 802.1p marking of 1 will be mapped to the STD queue Traffic with other 802.1p markings not specified above will be mapped to the STD queue. The A % value will define the CIR associated with the Expedited Forwarding class. The B % value will define the CIR associated with the Assured Forwarding class. The Operator will specify the A % value and B % value on a per WSEA logical order basis. These values will be expressed as percentage values. The sum of A + B cannot exceed 95% of the WSEA physical bandwidth. The WSEA bandwidth will define the PIR associated with the assured forwarding class and standard forwarding class. 05/05/2015 Version 1.8 Page 25

27 CIR PIR EF Queue (<=300M)Note 1 A % A % AF Queue (<=600M)Note 1 B % WSEA Bandwidth STD Queue 0 % WSEA Bandwidth Table 9: Traffic-based CoS - CIR and PIR Values Note 1: Sum of EF and AF queues cannot exceed 600M (see section 5.1) The end user 802.1p CoS markings will be carried transparently across the network and will not be re-marked by. Percentage EF/AF Mix Increment 1-5% EF only 1% 5% EF and AF 5% 10-20% EF and AF 10% 25% Fixed option EF=50% AF=25% 30-50% EF and AF 10% 55% Fixed option EF=20% AF=55% 55% EF only 60-70% EF and AF 10% 75% Fixed option EF=0% AF=75% 80-90% EF and AF 10% 95% Fixed option EF=95% AF=0% Table 10: Traffic-based CoS Options 05/05/2015 Version 1.8 Page 26

28 WEIL CoS Model Technical Handbook Wholesale NGN Ethernet Services This section describes the CoS model for the WEIL associated with the Wholesale NGN Ethernet products. The WEIL CoS model is shown in the following diagram: WSEA-1 EF AF WSEA-2 WSEA-x STD EF AF STD EF AF Scheduler for for Aggregate EF EF Scheduler for for Aggregate AF AF Scheduler for for WEIL Service Access Bandwidth (SAB) WEIL STD NGN Node Figure 8: WEIL CoS Model The Operator selects a WEIL Service Access bandwidth (SAB) for a WEIL. The Operator can specify a maximum of 10 WEIL Service Access Bandwidths (SABs) on a WEIL bearer. The Operator specifies the aggregate EF and AF bandwidths for each WEIL SAB. Aggregate EF Bandwidth Aggregate AF Bandwidth Range 0-100% 0-100% Table 11: Aggregate EF and AF Bandwidth Options The percentage values in Table 11:Table 15: refer to the percentage of the WEIL Service Access Bandwidth (SAB) and the sum of the aggregate EF and AF bandwidths cannot exceed 100% of the SAB bandwidth. The CoS profiles on the WEIL associated with the individual WSEA logical connections is the same as the CoS profiles configured for the individual WSEA connections as outlined in Section 6.1. It is the responsibility of the Operator to ensure that the sum of the CIR values of the individual WSEA connections associated with a WEIL SAB do not exceed the WEIL SAB and aggregate EF/AF bandwidths specified for that WEIL SAB. 6.3 Mapping of UNI - E-NNI CoS Markings Traffic Based QoS The section describes the mapping of CoS markings across the network. It should be noted that the end user C-VLAN 802.1p markings will never be remarked by and will always be tunnelled across the network. 05/05/2015 Version 1.8 Page 27

29 NGN Network Technical Handbook Wholesale NGN Ethernet Services WSEA EF AF STD WEIL Network Ingress C-VLAN 802.1p Marking MPLS EXP Bits Marking Network Egress S-VLAN 802.1p Marking Network Egress C-VLAN 802.1p Marking MPLS EXP Bits Marking Network Ingress S-VLAN 802.1p Marking Figure 10: Mapping of UNI - E-NNI CoS Marking 05/05/2015 Version 1.8 Page 28

30 6.3.1 UNI to E-NNI Traffic Flows Classification on ingress to the Core NGN network is based on the end user C-VLAN 802.1p markings. Traffic is forwarded to the core Forwarding Class based upon ingress policy (i.e. the EF and AF bandwidths if specified). As per Table 12:Table 17: the mapping of the C-VLAN 802.1p markings to the core Forwarding Classes is as follows: C-VLAN 802.1p Marking Forwarding Class 7 EF 6 EF 5 EF 4 EF 3 AF 2 AF 1 STD Table 12: Ingress Mapping of CoS Markings On egress of the Core NGN network the classification is based on the MPLS EXP bits markings associated with the different core Forwarding Classes. The MPLS EXP bits markings are mapped to the S-VLAN 802.1p markings on the E-NNI as per the following table: Forwarding Class S-VLAN 802.1p Marking EF 5 EF 4 AF 3 AF 2 STD 1 Table 13: Egress Mapping of CoS Markings The original end user C-VLAN 802.1p markings are tunnelled across the network and are presented to the Operator at the E-NNI within the C-VLAN ID header. 05/05/2015 Version 1.8 Page 29

31 6.3.2 E-NNI to UNI Traffic Flows The Operator must ensure that the S-VLAN 802.1p markings are also mapped to the C-VLAN (or vice-versa) prior to presentation at the E-NNI. Classification on ingress to the Core NGN network is based on the Operator S-VLAN 802.1p markings. Traffic is forwarded to the core Forwarding Class based upon the ingress policy specified for the WSEA connection (i.e. the EF and AF bandwidths if specified). The mapping of the S-VLAN 802.1p markings to the core Forwarding Classes is as follows: S-VLAN 802.1p Marking Forwarding Class 7 EF 6 EF 5 EF 4 EF 3 AF 2 AF 1 STD Table 14: Ingress Mapping of CoS Markings On egress of the Core NGN network the classification is based on the MPLS EXP bits markings associated with the different core Forwarding Classes. The MPLS EXP bits markings are mapped to the S-VLAN 802.1p markings on egress of the Core NGN network. The S-VLAN tag is stripped off prior to presentation to the customer on the UNI and the original end user/operator C-VLAN 802.1p markings are presented on the UNI. 05/05/2015 Version 1.8 Page 30

32 6.4 Network Performance The following table outlines indicative target network performance figures for the NGN: Traffic Type Parameter Real-Time (EF) Business (AF) Standard Delay (One-way) 10ms (Note 1) 25ms (Notes 1, 2) 35ms Delay Variation (Oneway) 2.5ms (Note 1) 3.5ms(Note 2) 3.5ms Frame Loss 0.001% (Note 1) 0.01% (Notes 1,2) 0.1% Table 15: Indicative Target Network Performance Figures Note 1: The specified product performance will be met when measured over any 15 minute interval. A failure to meet the product performance over this interval twice in any 24-hour period will constitute a fault. A single failure to meet the performance in each of any two consecutive 24-hour periods shall constitute an intermittent fault Note 2: Only applies to in-profile AF traffic. Note 3: The above figures only apply under normal network conditions. Note 4: The above figure only apply if the Operator/End User has not caused congestion on their own access connection. The Operator must ensure that the actual EF and AF bandwidth levels associated with a SAB on the WEIL do not exceed the selected EF and AF bandwidth values (i.e. that the aggregate EF and AF schedulers do not become congested). In the event of congestion occurring at the aggregate EF and AF schedulers, the delay figures outlined in Table 15:Table 20: above may increase significantly. This is normal QoS behaviour as EF/AF traffic will start to be buffered once congestion arises at the EF/AF aggregate schedulers and the EF/AF traffic associated with individual WSEA connections will be served on a round-robin basis which will result in additional latency. It is recommended that the end user/operator shape their traffic to conform to the selected CoS profile prior to presentation on ingress to the UNI/E-NNI interfaces. Failure to do so may result in the above performance figures being impacted. 6.5 NGN Capacity Management The NGN Architecture, see Figure 1, consists of three layers: Core Network layer Provider Edge Layer Aggregation Layer The basic units of capacity in the NGN network are bundles of 10GigE or 1GigE links. Aggregation Nodes are dual-homed to a pair of PE Nodes using n * 1GigE links or n * 10GigE links. A maximum of 4 * 1GigE links can be used for each of the individual connections between the Aggregation and PE layers. 10GigE links are used on the Aggregation to PE connections when the aggregate uplink utilisation exceeds 3.2 Gbit/s. 05/05/2015 Version 1.8 Page 31

33 Every link between PE Layer to Core Layers has a diversely routed partner. Each PE Node is connected to 2 Core Nodes using n * 10GigE links. The inter-pe Node connections also use n * 10GigE links. The Core Layer consists of two planes of 4 fully meshed routers which are interconnected using n * 10GigE links. n * GE or n * 10GE n * GE or n * 10GE n * 10GE n * 10GE n * 10GE PE (Edge) Nodes PE (Edge) Nodes NGN Aggregation Nodes Core Nodes NGN Aggregation Nodes Figure 11: NGN Network Architecture As the NGN Nodes deployed by are non-blocking devices, the main capacity management activity is focused on the management of the aggregate network link capacity which is described in more detail in the following section. Aggregate Network Link Capacity Management pro-actively monitors the actual utilisation levels on NGN network links in order to avoid the possibility of any congestion occurring in the network. In the event that a certain threshold is exceeded on a link, an upgrade to the link in question will be triggered. The following threshold values have been defined for the NGN network: Aggregation to PE Layer: PE to Core Layer: Core to Core Layer: Greater than 80% link utilisation in failure conditions across aggregate resilient links triggers link upgrades. Greater than 80% link utilisation in failure conditions across aggregate resilient links triggers link upgrades. Greater than 80% link utilisation in failure conditions triggers link upgrades. Using the following diagrams for reference and Aggregation Node to PE Node links as an example: 05/05/2015 Version 1.8 Page 32

34 Link A PE Node Aggregation Node Link A + Link B 80% Link B PE Node Non-failure Condition Link A PE Node Aggregation Node Link A 80% Link B PE Node Failure Condition Aggregation Nodes are dual-homed to a pair of PE Nodes using n * 1GigE links or n * 10GigE links. The physical capacity of both uplinks are equal (i.e. physical capacity of Link A = physical capacity of Link B) In the event of a failure on one of the uplinks, the remaining uplink should have sufficient capacity to accommodate all traffic (i.e. the original traffic on the remaining uplink + the re-routed traffic from the failed uplink) The capacity threshold for the aggregate traffic utilisation on both uplinks is 80% In non-failure and real-world conditions the two uplinks are not evenly load balanced e.g. Link A could have 30% utilisation whereas Link B could have 45% utilisation. Even though Link B would be greater than 40%, this would not trigger an upgrade as the aggregate link utilisation is less than 80% so in the event of a failure the remaining link would still be below the 80% threshold. The 80% threshold values are defined as having been exceeded if the actual utilisation level exceeds these values for greater than 60 minutes in a minimum of two days over a rolling seven day period. Note: Capacity upgrades between NGN layers in the network happen in pairs of links. 6.6 Planning Ratios The following planning ratios are used for capacity management purposes in the Core NGN Network: Traffic Type Planning Ratio EF 1:1 05/05/2015 Version 1.8 Page 33

35 Traffic Type Planning Ratio AF In Profile 1:1 AF Out of Profile 5:1 STD 5:1 1 STD Table 16: Planning Ratios 05/05/2015 Version 1.8 Page 34

36 05/05/2015 Version 1.8 Page 35

37 7 WSEA Interface Specification Technical Handbook Wholesale NGN Ethernet Services The section details the interface specification for the WSEA product. A WSEA connection requires the installation of an NTU (RAD ETX 202A) or (ALU 7210 SAS-M 10GigE) at the end user s premises or in the Operator s rack located in an exchange. Figure 12: RAD ETX-202A NTU Model Power Supply Options Power Consumption Dimensions RAD ETX-202A Single 220V AC PSU (Default) Dual 220V AC PSU (Optional) 18.5W max NTU with single PSU: Height: 43.7 mm (1.7 in) Width: 215 mm (8.4 in) Depth: 300 mm (11.8 in) NTU with dual PSU: Height: 43.7 mm (1.7 in) Width: 440 mm (17.4 in) Depth: 240 mm (9.5 in) 19 Rack-mountable Yes Table 17: WSEA NTU Accommodation Requirements The following table details the UNI specification for the WSEA product: UNI Physical Interface options UNI Physical interface presentation Optical Wavelength Optical Power Budget 10/100/1000BaseT (Default) 1000BaseSX 1000BaseLX RJ45 for 10/100/1000 Base-T (Default) LC connector for 1000BaseSX (multimode fibre) LC connector for 1000BaseLX (single mode fibre) 1000BaseSX SFP 850 nm 1000BaseLX SFP 1310 nm 1000BaseSX SFP 05/05/2015 Version 1.8 Page 36

38 Technical Handbook Wholesale NGN Ethernet Services Input Power (dbm) Min: -17 Max: 0 Output Power (dbm) Min: -9.5 Max: BaseLX SFP Input Power (dbm) Min: -20 Max: -3 Output Power (dbm) Min: -9.5 Max: -3 Auto negotiation support Full Duplex Support Autosensing Enabled UNI Port Setting Options EtherType Support Link Loss Forwarding (LLF) Support Yes Yes Yes 1000BaseSX/LX Port Auto Negotiate (Default) 1000M Full Duplex 10/100/1000 Base-T Port C-VLAN Auto Negotiate (Default) 100M Full Duplex 10M Full Duplex (Not supported for Traffic based COS) 88A8 (Not supported for Traffic based COS) UNI port on the NTU will shut-down in the event of loss of service on NTU-NGN Node connection. The tunnelling of end user/operator LLF L2CP traffic is supported on the network. By default all L2CP traffic is marked as 802.1p 7 and will get mapped to the STD Forwarding Class on ingress to the network. Table 18: WSEA UNI Interface Specification 05/05/2015 Version 1.8 Page 37

39 Figure 13: ALU 7210 SAS-M 10GigE NTU Model Power Supply Options Power Consumption Dimensions Alcatel-Lucent 7210 SAS-M 10GigE Redundant, hot swappable power supplies -DC input: -36 V DC to -72 V DC, output: +12 V DC (Default) -AC input: 100~240v, 50~60 Hz output: +12 V DC 60W (typical), 205 BTUs per hour Height: 67 mm (2.64 in.) Width: 436 mm (17.17 in.) Depth: 253 mm (9.96 in.) 19 Rack-mountable Yes Table 19: WSEA NTU Accommodation Requirements The following table details the UNI specification for the WSEA product: UNI Physical Interface options UNI Physical interface presentation Optical Wavelength Optical Power Budget 10 Gbit/s Access 10000BaseSR 10000BaseLR 10000BaseER 10000BaseZR 10 Gbit/s Access LC connector for 10000BaseSR (multimode fibre) LC connector for 10000BaseLR (single mode fibre) LC connector for 10000BaseER (single mode fibre) LC connector for 10000BaseZR (single mode fibre) 10000BaseSR XFP 850 nm 10000BaseLR XFP 1310 nm 10000BaseER XFP 1550 nm 10000BaseZR XFP 1550 nm 10000BaseSR XFP Input Power (dbm) Min: -9.9 Max: -1 05/05/2015 Version 1.8 Page 38

40 Output Power (dbm) Min: -7.3 Max: BaseLR XFP Input Power (dbm) Min: Max: 0.5 Output Power (dbm) Min: -8.2 Max: BaseER XFP Input Power (dbm) Min: Max: -1 Output Power (dbm) Min: -4.7 Max: BaseZR XFP Input Power (dbm) Min: -24 Max: -9 Output Power (dbm) Min: -1 Max: 4 Auto negotiation support Full Duplex Support Autosensing Enabled UNI Port Setting Options EtherType Support Link Loss Forwarding (LLF) Support No Yes No 10000BaseSR/LR Port 10000M Full Duplex S-VLAN 0x8100 C-VLAN 0x8100 0x9100 (Not supported for Traffic based COS) 0x88A8 (Not supported for Traffic based COS) Not supported Table 20: UNI specification for WSEA WSEA Customer Sited Handover (CSH) The WSEA CSH option involves the installation of an NTU (RAD ETX 202A) or (ALU 7210 SAS-M 10GigE) at the end user site. 05/05/2015 Version 1.8 Page 39

41 Operator UNI NTU NTU NGN node NGN core End User Site Exchange Figure 14: WSEA Customer Sited Handover (CSH) 7.2 WSEA In Building Handover (IBH) An In Building Handover (IBH) option is supported for WSEA connections. An NTU will be installed in the Operator s rack in the same Exchange as the serving NGN node. Figure 15: WSEA In Building Handover 7.3 Multi-service WSEA The initial release of the Wholesale Symmetrical Ethernet Access product was limited to a single UNI port on the managed NTU installed at the End User site. This 05/05/2015 Version 1.8 Page 40

42 enhancement will enable the Operator to request the use of multiple UNI ports on the managed NTU. Note that the ALU 7210 SAS-M 10GigE NTU does not support Multiservice WSEA as the node only has one UNI port available. End User Equipment NTU Fibre NGN node NGN UNI Ports Figure 16: WSEA Multiple UNI Ports The 1G NTU has the capability to support up to 5 UNI ports. One WSEA logical connection per UNI is supported. The port configuration for the 1G NTU is shown below: ETX-202A/NULL/NULL/4UTP Port No. 1 Port No. 2 Port No. 3 Port No. 4 Port No. 5 Port No. 6 Table 21: Empty SFP slot (not available as UNI port) Empty SFP slot 10/100/1000BaseT port (RJ-45) 10/100/1000BaseT port (RJ-45) 10/100/1000BaseT port (RJ-45) 10/100/1000BaseT port (RJ-45) NTU Port Configuration UNI S-VLAN SAP End User Equipment Operator EIL(s) UNI S-VLAN SAP NTU 7450ESS Figure 17: WSEA with Multiple Ports 05/05/2015 Version 1.8 Page 41

43 The sum of all of the WSEA logical service bandwidths on an NTU has to be equal to, or less than, the WSEA bandwidth (see Table 4:Table 6: for WSEA physical bandwidth options) Mapping of WSEA logical Services to Different WEILs In the event that an Operator has more than one WEIL, an Operator must specify which WEIL is to be associated with each WSEA logical service. If multiple WSEA logical services are provided on a single physical WSEA connection, each WSEA logical service can be associated with a different WEIL. The WEIL can only support logical WSEA services that originate at an End User site. NGN node NGN node NGN WES 1 EIL x NGN node SEA WES 2 EIL y WES 3 EIL z NGN node Figure 18: Mapping of WSEA logical Services to Different WEILs 05/05/2015 Version 1.8 Page 42

44 8 WEIL Interface Specification A number of handover options are supported on the WEIL product: Technical Handbook Wholesale NGN Ethernet Services Customer Sited Handover (CSH) In Span Handover (ISH) In Building Handover (IBH) Edge Node Handover (ENH) The interface/e-nni specification differs depending on which of the above option an Operator selects for the WEIL. Master Plan Plus (MPP) SLA is not supported on no NTU variants (IBH and ISH) of the WEIL product including ENH, IBH and ISH. ( no-ntu variants are those where a physical NTU device acting as a demarcation point e.g. RAD ETX-202A is not installed and where the demarcation point is instead an ODF or Eircom patch panel.) 8.1 WEIL Customer Sited Handover (CSH) The WEIL CSH option involves the installation of an NTU (RAD ETX 202A) or (ALU 7210 SAS-M 10GigE) at the Operator site. E-NNI Operator NGN Network NGN Node NTU NTU Figure 19: Exchange WEIL Customer Sited Handover (CSH) Operator Site NTU Model Power Supply Options Power Consumption RAD ETX-202A Dual 220V AC PSU Dual -48V DC PSU 18.5W max Dimensions Height: 43.7 mm (1.7 in) Width: 440 mm (17.4 in) Depth: 240 mm (9.5 in) 19 Rack-mountable Yes Table 22: CSH NTU Accommodation Requirements 05/05/2015 Version 1.8 Page 43

45 E-NNI Physical Interface options E-NNI Physical interface presentation Optical Wavelength Optical Power Budget 10/100/1000BaseT (Default) 1000BaseSX 1000BaseLX RJ45 for 10/100/1000 Base-T (Default) LC connector for 1000BaseSX (multimode fibre) LC connector for 1000BaseLX (single mode fibre) 1000BaseSX SFP 850 nm 1000BaseLX SFP 1310 nm 1000BaseSX SFP Input Power (dbm) Min: -17 Max: 0 Output Power (dbm) Min: -9.5 Max: BaseLX SFP Input Power (dbm) Min: -20 Max: -3 Output Power (dbm) Min: -9.5 Max: -3 Auto negotiation support Full Duplex Support Autosensing Enabled E-NNI Port Setting Options Yes Yes Yes 1000BaseSX/LX Port Auto Negotiate (Default) 1000M Full Duplex 10/100/1000 Base-T Port Auto Negotiate (Default) 100M Full Duplex 10M Full Duplex EtherType Support Link Loss Forwarding (LLF) Support S-VLAN 0x88A8 (Default) 0x8100 C-VLAN 0x8100 0x9100 (Not supported for Traffic based COS) 0x88A8 (Not supported for Traffic based COS) E-NNI port on the NTU will shut-down in the event of loss of service on NTU-NGN Node connection. The tunnelling of end user/operator LLF L2CP traffic is supported on the network. By 05/05/2015 Version 1.8 Page 44

46 default all L2CP traffic is marked as 802.1p 7 and will get mapped to the STD Forwarding Class on ingress to the network. Table 23: CSH E-NNI Interface Specification NTU Model Alcatel-Lucent 7210 SAS-M 10GigE Power Supply Options Power Consumption Dimensions Redundant, hot swappable power supplies -DC input: -36 V DC to -72 V DC, output: +12 V DC (Default) -AC input: 100~240v, 50~60 Hz output: +12 V DC 60W (typical), 205 BTUs per hour Height: 67 mm (2.64 in.) Width: 436 mm (17.17 in.) Depth: 253 mm (9.96 in.) 19 Rack-mountable Yes Table 24: 10G CSH NTU Accommodation Requirements E-NNI Physical Interface options E-NNI Physical interface presentation Optical Wavelength Optical Power Budget 10 Gbit/s Access 10000BaseSR 10000BaseLR 10000BaseER 10000BaseZR 10 Gbit/s Access LC connector for 10000BaseSR (multimode fibre) LC connector for 10000BaseLR (single mode fibre) LC connector for 10000BaseER (single mode fibre) LC connector for 10000BaseZR (single mode fibre) 10000BaseSR XFP 850 nm 10000BaseLR XFP 1310 nm 10000BaseER XFP 1550 nm 10000BaseZR XFP 1550 nm 10000BaseSR XFP Input Power (dbm) Min: -9.9 Max: -1 Output Power (dbm) Min: -7.3 Max: BaseLR XFP Input Power (dbm) 05/05/2015 Version 1.8 Page 45

47 Min: Max: 0.5 Output Power (dbm) Min: -8.2 Max: BaseER XFP Input Power (dbm) Min: Max: -1 Output Power (dbm) Min: -4.7 Max: BaseZR XFP Input Power (dbm) Min: -24 Max: -9 Output Power (dbm) Min: -1 Max: 4 Auto negotiation support Full Duplex Support Autosensing Enabled E-NNI Port Setting Options EtherType Support Link Loss Forwarding (LLF) Support No Yes No 1000BaseSR/LR Port 10000M Full Duplex S-VLAN 0x8100 0x88a8 (Default) C-VLAN 0x8100 0x9100 (Not supported for Traffic based COS) 0x88A8 (Not supported for Traffic based COS) Not supported Table 25: 10G CSH E-NNI Interface Specification 8.2 WEIL In Span Handover (ISH) For the Wholesale Ethernet Interconnect Link (WEIL) ISH option the fibre will join the Operator fibre at a Point of Handover (PoH) outside the same exchange as the serving NGN node. 05/05/2015 Version 1.8 Page 46

48 E-NNI Operator NGN Network NGN Node 1000Base ZX SFP U/G Manhole FC/PC Connectors 12 fibre cable Operator Duct Operator ODF ODF Closure Operator fibre cable Exchange Metres Point of Handover (PoH) Site Figure 20: WEIL In Span Handover (ISH) The PoH for a WEIL ISH is the point at which the optical fibres from an Operator s cable are connected to the optical fibres of an cable. An Operator is responsible for providing a Raychem FOSC400 Xcon (ODF type closure) at the PoH. will terminate its optical fibres on one side of the optical distribution frame (ODF) and an Operator will terminate its fibres on the other side of the ODF. The PoH is housed in a U/G manhole located within 100 Metres outside the boundary of an exchange. The maximum distance between the terminal ends of the ISH shall be determined with reference to the available optical power budget. The required minimum Receive Level at the PoH shall be 19 dbm at start of life, to ensure a minimum of 22 dbm to end of life, following subsequent repairs and the natural ageing of systems measured at the optical distribution frame interface. -3 dbm - 19 dbm OAO Figure 21: Minimum Power Levels 1G In Span Interface - ZX SFP As shown in the diagram above, will provide, to an Operator, with a 1Gbit/s In Span Interface, an optical signal at -3dBm minimum at an Operator optical distribution frame interface. An Operator will provide, to, an optical signal at -19dBm minimum at the optical distribution frame interface. -7 dbm - 13 dbm Operator Figure 22: Minimum Power Levels 10G In Span Interface ER XFP 05/05/2015 Version 1.8 Page 47

49 As shown in the diagram above, will provide to an Operator with a 10Gbit/s In- Span Interface ER XFP an optical signal at -7 dbm minimum at an Operator Optical Distribution Frame interface. The Operator will provide, to, an optical signal at - 13dBm minimum at the Optical Distribution Frame (ODF) interface. -5 dbm - 23 dbm Operator Figure 23: Minimum Power Levels 10G In Span Interface - ZR XFP As shown in the diagram above, will provide to an Operator with a 10Gbit/s In- Span Interface - ZR XFP an optical signal at -5 dbm minimum at an Operator Optical Distribution Frame interface. The Operator will provide, to, an optical signal at - 23dBm minimum at the Optical Distribution Frame (ODF) interface. Note that a 3dB margin is used in the above figures to account for losses associated with connectors, ODFs, etc, between the 7450ESS port and the E-NNI. E-NNI Physical interface presentation Optical Wavelength Optical Power Budget (see Figure 21:Figure 25:, Figure 22:Figure 26:, Figure 23:Figure 27:) Single Mode Fibre FC/PC connector 1550 nm 1 Gbit/s Access ZX SFP Input Power (dbm) Min: -19 Max: -3 Output Power (dbm) Min: -3 Max: Gbit/s Access ER XFP Input Power (dbm) Min: -13 Max: -1 Output Power (dbm) Min: -7 Max: Gbit/s Access ZR XFP Input Power (dbm) Min: -23 Max: -9 Output Power (dbm) Min: -5 05/05/2015 Version 1.8 Page 48

50 Max: +2 Auto negotiation support Full Duplex Support Autosensing Enabled E-NNI Port Setting Options EtherType Support Link Loss Forwarding (LLF) Support Yes Yes Yes 1000BaseZX Port Auto Negotiate (Default) 1000M Full Duplex 10000BaseER/ZR Port S-VLAN C-VLAN Auto Negotiate (Default) 10000M Full Duplex 0x88A8 (Default) 0x8100 0x8100 0x9100 (Not supported for Traffic based COS) 0x88A8 (Not supported for Traffic based COS) The tunnelling of end user/operator LLF L2CP traffic is supported on the network. By default all L2CP traffic is marked as 802.1p 7 and will get mapped to the STD Forwarding Class on ingress to the network. Table 26: ISH E-NNI Interface Specification 05/05/2015 Version 1.8 Page 49

51 8.3 WEIL In Building Handover (IBH) Technical Handbook Wholesale NGN Ethernet Services For a Wholesale Ethernet Interconnect Link (WEIL) In Building handover a fibre cable is installed between the Operator collocation footprint and an optical distribution frame (ODF) in the same exchange as the serving NGN node. E-NNI Operator Operator Cable Tray Operator LLU Collocation Rack NGN Network NGN Node fibre 1000Base LX SFP ODF Operator 12 fibre cable SC Connectors Operator Patch Panel Operator Equipment Exchange Figure 24: WEIL In Building Handover (IBH) dbm - 17 dbm OAO Figure 25: Minimum Power Levels - 1Gbit/s In Building Interface As shown in the diagram above, for a 1Gbit/s IBH WEIL, will provide, to an Operator, an optical signal at dbm minimum at the ODF. An Operator will provide, to, an optical signal at -17dBm minimum at the ODF dbm dbm Operator Figure 26: Minimum Power Levels - 10Gbit/s In Building Interface As shown in the diagram above, for a 10Gbit/s IBH WEIL will provide, to an Operator, an optical signal at dbm minimum at an Operator optical patch panel. The Operator will provide to an optical signal at -12.4dBm minimum at an Operator optical patch panel. Note that a 2dB margin is used in the above figures to account for losses associated with connectors, ODF, etc, between the 7450ESS port and the E-NNI. 05/05/2015 Version 1.8 Page 50

52 E-NNI Physical interface presentation Optical Wavelength Optical Power Budget (see Figure 25:Figure 29:and Figure 26:Figure 30:) Auto negotiation support Full Duplex Support Autosensing Enabled E-NNI Port Setting Options EtherType Support Link Loss Forwarding (LLF) Support Single Mode Fibre SC connector 1310 nm 1 Gbit/s Access Input Power (dbm) Min: -17 Max: -3 Output Power (dbm) 10 Gbit/s Access Max: Yes Yes Yes Min: Max: -3 Input Power (dbm) Min: Max: Output Power (dbm) 1 Gbit/s Access Min: Auto Negotiate (Default) 1000M Full Duplex 10 Gbit/s Access S-VLAN C-VLAN Auto Negotiate (Default) 10000M Full Duplex 0x88A8 (Default) 0x8100 0x8100 0x9100 (Not support for Traffic based COS) 0x88A8 (Not support for Traffic based COS) The tunnelling of an Operator's LLF L2CP traffic is supported on the network. By default all L2CP traffic is marked as 802.1p 7 and will get mapped to the STD Forwarding Class on ingress to the network. Table 27: IBH E-NNI Interface Specification 05/05/2015 Version 1.8 Page 51

53 8.4 WEIL Edge Node Handover (ENH) Technical Handbook Wholesale NGN Ethernet Services The Operator facing port on the NGN node will be connected directly to an patch panel for the ENH handover. Figure 27: WEIL Edge Node Handover dbm dbm Operator Figure 28: 10G ENH Minimum Power Levels As shown in the diagram above, for a 10G WEIL will provide, to an Operator, an optical signal at dbm minimum at the optical patch panel. The Operator will provide to an optical signal at -12.4dBm minimum at the optical patch panel. Note that a 2dB margin is used in the above figures to account for losses associated with connectors, patch panel, etc, between the NGN node port and the E-NNI. 05/05/2015 Version 1.8 Page 52

54 The following table lists the UNI interface specification for ENH: Technical Handbook Wholesale NGN Ethernet Services E-NNI Physical interface presentation Optical Wavelength Optical Power Budget Auto negotiation support Full Duplex Support Autosensing Enabled E-NNI Port Setting Options EtherType Support Link Loss Forwarding (LLF) Support Single Mode Fibre SC connector 1310 nm 10 Gbit/s Access Yes Yes Yes Input Power (dbm) Min: Max: Output Power (dbm) 10 Gbit/s Access S-VLAN C-VLAN Min: Max: Auto Negotiate (Default) 10000M Full Duplex 0x88A8 (Default) 0x8100 (Optional) 0x8100 0x9100 (Not support for Traffic based COS) 0x88A8 (Not support for Traffic based COS) The tunnelling of an Operator's LLF L2CP traffic is supported on the network. By default all L2CP traffic is marked as 802.1p 7 and will get mapped to the STD Forwarding Class on ingress to the network. Table 28: ENH E-NNI Interface Specification 05/05/2015 Version 1.8 Page 53

55 8.4.1 WEIL ENH 1+1 Protection A 1+1 Protection option will be offered on WEIL connections to provide end users/operators with additional protection on the NNI/WEIL interfaces. This will be achieved by grouping multiple WEILs into Link Aggregation Groups (LAGs). The WEILs which are members of a LAG will be configured in active/active mode. End User/ Operator Equipment SAP NGN 1+1 Protection 7450ESS Figure 29: WEIL with 1+1 Protection The following 1+1 Protection options are offered: 10G WEIL 1+1 Protection (2 * 10G LAG) The WEIL bandwidth options as outlined in Section 0 will still apply for this option (i.e. the max bandwidth supported a 10G 1+1 Protected WEIL will be 10 Gbit/s). This will ensure that, in the event of a failure on one of the LAG ports, the remaining link will have sufficient capacity to carry all traffic. There will be minimal service interruption due to the switchover. For WEIL ENH with 1+1 protection the two 10Gb/s customer facing ports will be allocated from different IOM card modules on the NGN node. The following LAG parameters must be used by the Operator/end user equipment: Parameter Setting LACP Enabled Mode Active Active (1) Port threshold 0 (2) QoS Table 29: LAG Parameters: WEIL with 1+1 Protection PIR enforced across both LAG ports (1) Ensures that traffic is active on both links. (2) Indicates that the LAG is only taken down if both links are down 05/05/2015 Version 1.8 Page 54

56 9 Configuration of Loopbacks on WSEA Connections An Operator can request for a loopback to be configured on an Wholesale WSEA connection. The loopback will be internally configured on the end-user facing port (UNI) on the NTU deployed at the end-user site. The Operator should use one of the following methods to submit a request for a loopback to be configured: The Operator can request for a loopback to be configured via the Unified Gateway by selecting 'Configuration of Loopback on NGN WSEA Circuit' on the fault reporting screen and referencing the WSEA Circuit ID (WESxxxxxxxx). The Operator can log a request for a loopback to be configured by calling The Operator must request a 'Configuration of Loopback on NGN WSEA Circuit' and reference the WSEA Circuit ID (WESxxxxxxxx). The Operator must submit an additional request to have the loopback removed using the contact options outlined above (the WSEA Circuit ID must be referenced in all cases). The length of time that a loopback is applied should not exceed 24 hours. Note that this service is supported on the RAD NTU but it is not available when using the ALU 7210 SAS-M 10G NTU. 05/05/2015 Version 1.8 Page 55

57 10 NGN Node Types / Fibre Access Model There are three possible types of fibre access model used to connect the NTU at the end user site to the NGN (please refer to Figure 30:Figure 34:, Figure 31:Figure 35:, Figure 32:Figure 36:) 1. The default fibre access model is to connect the end user site to an NGN Aggregation Node by a direct fibre (Figure 30:Figure 34:). 2. In certain circumstances an end user may be connected to an exchange which is fibre enabled but does not have an NGN Aggregation Node (i.e. a Node Reach site). An inter-exchange (core) fibre pair is required between the NGN Aggregation Node enabled exchange and the end user s local exchange. This interexchange fibre pair is then patched to the customer s access fibre pair (Figure 31:Figure 35:). 3. The Access Packet Transport (APT) network is used to connect to the NGN at certain exchanges where an NGN Aggregation Node is not deployed. The APT network is a Carrier Ethernet transport network that uses MPLS-TE and extends the reach of the NGN to remote exchanges.the end user site is connected via fibre to an APT node (Figure 32:Figure 36:). The APT network appears as a transparent pipe to the services delivered over it, so the service parameters remain the same as for options 1 and 2 above, with the restriction that there is a 300 Mbit/s bandwidth limit for each physical access, whereas the limit is 1Gbit/s for options 1 and 2. A customer requirement for a service > 300 Mbit/s at an APT enabled site will be delivered over direct fibre (i.e. Node Reach solution) where practicable. UNI Operator Site NTU SFP/ XFP Fibre Exchange ODF NGN aggregation node SFP/ XFP Figure 30: Direct Fibre Access to NGN Aggregation node 05/05/2015 Version 1.8 Page 56

58 UNI Customer Site NTU SFP/ XFP Fibre Exchange ODF Fibre Exchange ODF NGN Aggregation Node SFP/ XFP Figure 31: Fibre Access via an Intermediate Exchange UNI Customer site Exchange Exchange Exchange NTU 1G Fibre APT Node APT Domain APT Node 1G Fibre NGN Aggregation Node NGN Figure 32: Max. 300Mb/s per WSEA physical NGN Extended Reach using APT For resilience, an NGN aggregation node is dual homed to 2 PE routers (Figure 33:Figure 37:). APT nodes are daisy-chained to a Head-End APT node that is co-located with the NGN Aggregation Node (Figure 34:Figure 38:). There is a single link between APT nodes, no protection is currently implemented. NGN PE NGN Aggregation Node NGN core NGN PE Figure 33: Aggregation node to PE connection 05/05/2015 Version 1.8 Page 57