WCDMA Transport Network Overview (ATM & IP)

Peng Chen GSDC China 2008-5-28

Agenda
Interface Protocol IMA Infrastructure Service Topology Link Redundancy and Protection

WCDMA RAN Logical Network Reference Model

Iu interface
The Iu interface is an open, standardized interface providing multivendor equipment to be supplied for both the Core Network and the WCDMA RAN. The Iu interface capabilities include the following:
Procedures to establish, maintain, and release Radio Access Bearers Procedures to perform Inter-Radio Access Technology handover Procedures for transferring of Non Access Stratum (NAS) signaling messages between UE and Core Network Location services by transferring requests from the Core Network to the WCDMA RAN and location information from the WCDMA RAN to the Core Network Simultaneous access to multiple Core Network domains for a single UE Mechanisms for resource reservation for packet data streams

Iur Interface
The Iur interface is a WCDMA internal interface for the communication between two RNC nodes (and between two RNSs). It is an open and standardized interface. The interface contains a control plane for radio signaling and AAL2 connection establishment and a user plane supporting guaranteed QoS on ATM or IP. Note that all ATM user plane traffic is carried on the same type of AAL2 connection, packet data, voice, and Unrestricted Digital Information (UDI).

Iub Interface
The information transferred over the Iub is categorized as follows:

Radio application-related signaling

The Iub interface allows the RNC and the RBS, for example, to negotiate about radio resources, to add and delete cells controlled by the RBS to support communication of a dedicated connection. The Iub interface provides the means for transport of uplink and downlink radio frames between the RNC and the RBS. The macro-diversity combining function of the RNC uses RBS quality estimations of the uplink radio frames. Accurate time synchronization between the soft handover branches is also required, which includes frame synchronization and node synchronization as a base.

Radio frames Quality estimation of uplink radio frames and synchronization data

Control Plane
The Radio Network protocol for the control plane is called Node B Application Protocol (NBAP) and transports Radio Network control plane messages between the RBS and the Controlling RNC (CRNC). The NBAP is carried on a signaling bearer based on SAAL-UNI on top of ATM. The control plane includes functionality for establishing AAL2 connections. When using IP transport, NBAP is carried over SCTP. The RNC and the RBS support a user plane over the Iub interface that is based on transport channels. The Radio Network User plane uses several frame protocols: FACH FP, RACH FP, PCH FP, DCH FP and HS-DSCH FP. The user plane carries data and control frames that are transferred between the RBS and the CRNC, containing radio interface (Uu) user data and user associated control data. Frame synchronization applies for the Iub interface. The Frame Synchronization function provides timing and supervision of traffic frames between the RNC and RBSs, in both the uplink and downlink directions. This is achieved by managing frame offset values that can be set different for downlink during operation. The frame offset values can be predefined in the system, but can also be refined, voice frame offsets are normally lower than for packet data, to be able to fulfil the delay requirements. The Node Synchronization function is a base for the Frame synchronization function, which is also a procedure applied for the Iub. The Node Synchronization functionality results in the knowledge of timing relationship between the CRNC and each RBS.

User Plane

Synchronization Plane

Protocols(1)
There are different protocols to build both Transport Network User Plane protocol stack and Transport Network Control Plane protocol stack on the Iu, Iur, and Iub interfaces.
e.g. SCCP, SCTP, MTP3b, M3UA, SAAL-NNI, SAAL-UNI, Q.2630, Q.2150.1, Q.2150.2 and GTP-U.

SCCP
The signaling Connection and Control Part (SCCP) is a protocol that provides additional functions to the Message Transfer Part level 3 (MTP3b and M3UA) and provides connectionless and connection-oriented network services, as well as address translation capabilities. The SCCP protocol provides the functionality to transport signaling messages between SCCP users and also through signaling Transfer Points (STPs) and/or Signaling Gateways (SGs). The SCCP is used to carry RANAP messages in the Iu interface and RNSAP messages in the Iur interface. The messages associated with a specific UE use SCCP connection-oriented mode; all other messages use SCCP connectionless mode. The SCCP uses for addressing, a Destination Point Code (DPC) and Subsystem Number (SSN), that allow direct routing by the SCCP and MTP3b or M3UA. The SSN is used by a terminating node (for example, the RNC) to identify different applications, for example, RANAP(142) or RNSAP(143). In connectionless mode, all information required to route the data to its destination is contained in each data packet and is analyzed in each node. Thus, no logical connection is established between the two end nodes.The connectionless mode is typically used to transfer small amounts of real-time-critical information between two users. The connection-oriented service is a way of exchanging signaling information between two network service users by establishing a logical connection between them. In connectionoriented mode, the data transfer messages only contain local reference numbers to identify the connection and they follow the previously established path.

Protocols(2)
SCTP
The Stream Controlled Transmission Protocol (SCTP) protocol layer provides a connection-oriented protocol between SCTP end nodes by the means of SCTP associations. An SCTP endpoint is defined by the SCTP transport address which consists of one or more IP address and an SCTP port. An SCTP association provides a reliable connection between two SCTP end nodes. SCTP provides the following:
Acknowledged error-free non-duplicated transfer of data Detection of data corruption, loss of data and duplication of data Selective retransmission mechanism to correct loss or corruption of data

SCTP provides a general-purpose transport protocol for message-oriented applications such as signaling operating on top of a connectionless packet service such as IP. The Message Transfer Part level 3 (MTP3b) protocol is responsible for transmission of signal units, with information provided by user parts, from one signaling Point (SP) to another. In WCDMA RAN Q.2630 and SCCP protocols are identified as users of MTP3b. The Message Transfer Part level 3 (MTP3b) protocol is responsible for transmission of signal units, with information provided by user parts, from one signaling Point (SP) to another. In WCDMA RAN Q.2630 (12) and SCCP (3) protocols are identified as users of MTP3b. The MTP3b allows transportation of signaling messages through SAAL-NNI links and extends the functionality offered by the SAAL-NNI to provide network layer functionality. MTP3b functions are responsible for distribution, discrimination, and routing of messages. The MTP is able to perform the following functions:
Ensure that incoming messages are either distributed to a user part in its own node or routed and forwarded to the next signaling point or signaling transfer point. Ensure that outgoing messages are directed to the appropriate signaling link. Control and manage all functions that are needed on each signaling link to provide reliable transfer.

MTP3b

M3UA
The MTP Level 3 User Adaptation (M3UA) protocol layer provides the same connectionless service to its user as MTP3b protocol provides. Internally M3UA provides and uses a connection-oriented service to support MTP3 signaling over IP. In WCDMA RAN Q.2630 and SCCP protocols are identified as users of M3UA.

Protocols(3)
SAAL-NNI
The signaling ATM Adaptation Layer-Network Node Interface (SAAL-NNI) offers link layer functionality ensuring that the two end points of a signaling link can exchange signaling messages reliably. It incorporates error checking, flow control, and sequence checking. SAAL-NNI is used as signaling link in both Iu and Iur interfaces. The following services are provided by the SAAL-NNI:
Assured data transfer over point-to-point SAAL-NNI connections.
Message delimitation, and alignment, error detection, and error correction are part of the assured data transfer service of this SAAL. The SAAL-NNI service relieves the user from loss, insertion, and corruption of data that may occur. In some cases, due to unrecoverable errors in the ATM adaptation layer, duplication or loss of service data units (SDUs) may occur.

Ability to retrieve SDUs already delivered to the SAAL-NNI. One signaling link error monitoring function.
It is employed when a link is in the proving state of the initial alignment procedure.

Re-establishment of assured data transfer mode when a protocol error has been detected by the SAAL layer (recovery).

Protocols(4)
SAAL-UNI
The SAAL-UNI offers link layer functionality ensuring that the two end points of a signaling link can exchange signaling messages reliably. It incorporates capabilities, such as error checking, flow control, and sequence checking. The use of SAAL-UNI allows an efficient transport of the NBAP in the Iub interface as well as to support Q.2630 transport in Iub. The following services are provided by the SAAL-UNI:
Unassured data transfer over point-to-point SAAL-UNI connections. The SAAL-UNI unassured data transfer service does not relieve the user from loss or insertion of data which may occur. Assured data transfer over point-to-point SAAL-UNI connections. The SAAL-UNI service relieves the user from loss, insertion, and corruption of data which may occur. In some cases, due to unrecoverable errors in the ATM adaptation layer, duplication or loss of service data units (SDUs) may occur. Establishment and release of assured data transfer mode on an established SAAL-UNI connection. Reestablishment of assured data transfer mode on request from the user (re-synchronization). Reestablishment of assured data transfer mode when a protocol error has been detected by the SAAL layer (recovery).

Q.2630
This AAL type 2 signaling protocol supports the dynamic establishment and release of individual AAL type 2 point-to-point connections. Q.2630 relies on the MTP3b or M3UA (in Iu and Iur interfaces) or SAAL-UNI (in Iub interface) for transport of messages between the relevant nodes. Q.2630 has the following characteristics:
On-demand establishment, maintenance, and release of end-to-end AAL2 connections over AAL2 network comprised AAL2 end-points and AAL2 switching points Support hop-by-hop routing Ability to control AAL2 connections on more than one underlying ATM VCC (ATM Virtual Channel Connection) Selection of AAL2 path type

Q.2150.1, in Iur and Iu-cs interfaces, defines the transport of Q.2630 for MTP3b and provides the necessary reliability for the signaling transport. Q.2150.2 defines the transport of Q.2630 directly over SAAL-UNI in the Iub interface.

Protocols(5)
GTP-U
The GPRS Tunneling Protocol User plane (GTP-U) is used between the RNC and the Serving GPRS Support Node (SGSN) in the Core Network. GTP-U is a tunnelling protocol that is used to transport user packet data over the Iu-PS interface. The main function of GTP-U is to provide separation between packets that belong to different user data flows, so there is a GTP-U tunnel for each PDP context in the SGSN or each RAB in the RNC. A GTP-U tunnel is defined between a pair of tunnel endpoints. A tunnel endpoint is identified by a Tunnel Endpoint Identifier (TEID), which is carried in the GTP-U header. For one tunnel, there is a different TEID for uplink and downlink traffic. These TEIDs (as well as IP addresses for RNC and Core Network) are exchanged over Iu-PS in RANAP signaling when the packet data RAB is established. A GTP-U path is defined between two GTP layers in different IP addresses, there is at least one GTP-U tunnel up between those two GTP layers. In the ATM case, typically one (but could be more) GTP-U path is carried over UDP/IP path and AAL5. When IP transport is used, the GTP-U layer is unchanged and is carried over UDP/IP and Ethernet. A UDP/IP path is a connection less path defined by two end points where an IP address and a UDP port number define an end point.

IMA(1)

The purpose of IMA is to provide an inverse multiplexing of an ATM cell stream over multiple physical links. Several physical links can thus be seen logically as one unique ATM link with a capacity equal to the sum of the capacities of the individual links. The feature supports:
Simplification of the ATM O&M procedures, i.e. monitoring as one single ATM link instead of several separate Higher trunking level and statistical multiplexing gain Redundancy and robustness

The ATM Inverse Multiplexing technique involves inverse multiplexing and de-multiplexing of ATM cells in a cyclical fashion among links grouped to form a higher bandwidth logical link whose rate is approximately the sum of the link rates. This is referred to as an IMA group. IMA groups terminate at each end of the IMA virtual link.
In the transmit direction, the ATM cell stream received from the ATM layer is distributed on a cell-by-cell basis, across the multiple links within the IMA group. At the far-end, the receiving IMA unit recombines the cells from each link, on a cell-by-cell basis, recreating the original ATM cell stream. The aggregate cell stream is then passed to the ATM layer. The IMA interface periodically transmits special cells that contain information that permit reconstruction of the ATM cell stream at the receiving end of the IMA virtual link. These cells, defined as IMA Control Protocol (ICP) cells, provide the definition of an IMA frame. The receiver end reconstructs the ATM cell stream after accounting for the link differential delays, smoothing Cell Delay Variation (CDV) introduced by the control cells, etc. The transmitter must align the transmission of IMA frames on all links; the receiver can detect the differential delays by measuring the arrival times of the IMA frames on each link. At the transmitting end, the cells are transmitted continuously. If there are no ATM layer cells to be sent between ICP cells within an IMA frame, then the IMA transmitter sends filler cells to maintain a continuous stream of cells at the physical layer. The insertion of filler cells provides cell rate decoupling at the IMA sublayer. The IMA receiver discards the filler cells.

IMA(2)
IMA achieves automatic reconfiguration if one, or more, physical links fail.
The ATM traffic is redistributed on the physical links still in operation. When the link fault is repaired, the automatic reconfiguration takes place again. The ATM traffic is redistributed on the physical links inclusive on the repaired ones

It is possible to use IMA on E1/T1/J1 (ET-MC1) and channelized 155Mbps boards (ET-MC41/ET-MC41s/ET-MF41).
IMA groups must be configured within the same ET-MC1 board and within the same circuit group on the ET-MC41 board. The ET-MF41 does not have the circuit group limitation so IMA groups can be configured freely within one port. All links in an IMA group must have the same bandwidth, normally a full E1/T1/J1.

A good way to protect the IMA group in case of physical link failure is to utilize the IMA bandwidth adaptation feature.
In the event of link failure, strict QoS services such as voice and video are undisturbed but best effort capacity is reduced. This can be achieved by the following configurations:
one UBR+ VP which adapts automatically at link failure. one CBR VP and one UBR VP. CBR and UBR+ VCs must be carried in the CBR VP.

Infrastructure(1)
TNINF is a set of logical links used to connect the WCDMA RAN transport nodes to each other and to the Core Network nodes. TNINF logical links are:
PDH Link SDH Link ATM Link AAL2 Link AAL5 Link Ethernet Link

All types of equipment and transport media in a transport network introduce delay and delay variance, which shall be minimized in order to maintain QoS and radio network performance and node synchronization and buffering performance.
The following requirements are placed on ATM-based transport network solutions: Delay at Iub Propagation delay, target: 5ms
Propagation delay, maximum: 30ms Delay variation, maximum: 2ms Propagation delay difference, if multiple links are used between RBS and RNC, target: 3ms, maximum: 10ms Propagation delay, maximum: 10ms Delay variation, maximum: 2ms Propagation delay, maximum: 12ms Delay variation, maximum: 2ms

Delay at Iu Propagation delay, target: 2ms

Delay at Iur Propagation delay, target: 4ms The proposed values are quite high and should easily be fulfilled in most cases. In some rare cases even higher values can be tolerated but performance will degrade.

Infrastructure(2)
The following delay values are suggested for IP-based transport network solutions: Delay for real-time traffic on Iub Propagation delay, target: 5ms
Propagation delay, maximum: 30ms Delay variation, maximum: 10ms Packet Loss, maximum: 10-6

Delay for HSPA traffic on Iub Propagation delay, target: 10ms

Propagation delay, maximum: 100ms Delay variation, maximum: Packet Loss, target: 10-6 Packet Loss, maximum: 10-4

Delay for non-HS best effort traffic on Iub Propagation delay, target: 10ms
Propagation delay, maximum: 50ms Delay variation, maximum: 12ms Packet Loss, target: 10-6 Packet Loss, maximum: 10-4

Delay at Iur Propagation delay, target: 4ms

Propagation delay, maximum: 12ms Delay variation, maximum: 5ms Packet Loss, target: 10-6

Infrastructure(3)
Exchange Terminal Board
The RBS, RXI, and RNC support physical interfaces in the range 1.5 Mbps to 155 Mbps.

ET-MC1:

The ET-MC1 is a 1.5 Mbps and 2 Mbps interface. Eight ports on each board n* 64 kbps fractional support on each port Maximum number of terminated VCs on each board: 240 Maximum number of terminated VPs on each board: 16 (two per port) Number of AAL2 muxes on each board : 16 Maximum number of terminated AAL2 connections on each board: 3968 Number of TDM channels on each board: 32 VP shapers on each board: 16 (2 on each port) Fractional ATM implemented Circuit emulation implemented IMA HW support with 1- 4 IMA groups with 2 - 8 ports each. Only full E1/T1/J1 ports can be part of an IMA group. Physical Media: Twisted pair

Infrastructure(4)
ET-MC41
The ET-MC41 is a channelized STM-1 line interface board. The board is configurable according to American or European standards STS-3c/OC3 (VT1.5), STM-1(VC-12). The board has:
One port per board 84 VT-1.5 or 63 VC-12 on each port, configurable at initiation Maximum number of terminated VCs on each board: 1890 if configured with 63 VC12 virtual containers or 2520 if configured with 84 VT-1.5 virtual tributaries Maximum number of terminated VPs: 2 on each virtual tributary/container (VT-1.5 or VC-12) Number of AAL2 muxes on each board: 126 if configured with 63 VC-12 or 168 if configured with 84 VT-1.5 Maximum number of terminated AAL2 connections on each board: 31248 (with VC12), 41664 (with VT-1.5) Number of TDM channels on each board: 252 (with VC-12), 336 (with VT-1.5) VP shapers: 2 on each virtual tributary/container (VT-1.5 or VC-12) One ATM on fractional on each virtual tributary/container Circuit emulation implemented, 1-30 TDM terminations using AAL1 per virtual container (VC-12) and 1-24 TDM termination using AAL1 per virtual container (VT1.5) Up to 30/42 IMA groups (2-8 VT/VC flows per group) for STM-1/OC3 respectively Physical Media: Optical S-1.1 (G.957) or IR-1 (ANSI) Physical contact for S-1.1 is SC, the range is at least 15 km Hardware support for MSP 1+1 link and equipment protection

Infrastructure(5)
ET-MFG
The ET-MFG is an exchange terminal board that enables direct IP connections to the RNC node with a one Gigabit Ethernet interface (1000Base-T). The board has:
1+1 Gigabit Ethernet ports per board (one port is active and the other one is standby) Each port has an RJ45 connector in the MDF Auto configuration. Only 1000 Mbps, full duplex mode with Symmetric Pause is supported VLAN support. It is possible to handle the following Ethernet frame variants: Untagged Tagged using VLAN identity and priority bit Priority tagged using priority bit only Protection switching is completed within 700 ms after a link failure

ET-MFX
The ET-MFX is an exchange terminal board used for Iub over IP/Ethernet. The board has:
Native IP Ethernet interface. Single slot (15mm) Used for Iub over IP/Ethernet, site LAN, and for IP aggregation Ethernet interfaces equipped in three board variants: 6 x RJ 45 10/100/1000Base-T + 1 x SFP for 1000 Base-X for RBS (ET-MFX11) 6 x RJ 45 10/100/1000Base-T + 1 x SFP for 1000 Base-X for RNC and RXI (ET-MFX12) 1 x RJ 45 10/100/1000Base-T + 6 x SFP for 1000 Base-X for RNC and RXI only (ET-MFX13) Auto negotiation for 10/100/1000 Mbps selection SFP is targeted for optical ethernet connectivity. SFP variants available through Ericsson Marconi SFP product family VLAN 802.1Q, L2 Ethernet priority 802.1p, L3 IP Diffserv priority Optional Network Synch Server for IP Transport Optional Ethernet Switch functionality

Services(1)

Services(2)
The following three blocks are identified within transport and signaling services:
Transport Service Control Plane
The following functionality is supported: Control plane functionality including addressing, routing, and connection admission control to support dynamically (on-demand) establishment and release of AAL2 connections. For each network connection, a CPP-based node can act as an originating node, a transit node, or as a destination node. Management plane functionality to establish and release AAL2 paths (ATM virtual Channel connection, ATM VCC) Management plane functionality to establish and release permanent cross-connected ATM VC links Management plane functionality to establish and release permanent ATM VP terminations Management plane functionality to establish and release Iub user plane transport bearers using the IPv4 transport option

Signaling Services
All the signaling services supported by CPP, such as SAAL-NNI, SAAL-UNI, MTP3b, M3UA, SCCP, SCTP, Q.2150.1 and Q.2150.2.

Transport Services User Plane

The transport service provides applications to send or receive data on established point-to-point AAL2 or AAL5 connections and cross-connected ATM Virtual Channel Links (VCLs) and transport of the Iub common transport channels and user data streams over UDP/IP.

The following Service Categories to support the ATM connectivity

CBR, UBR+, UBR

VP VC Aal2

Services(3)
The AAL2 Transport Service provides bandwidth-efficient transmission of low-rate, short, and variable length packets in delay-sensitive applications. AAL2 Connection Admission Control is the set of actions taken by the network during the AAL2 connection establishment phase in order to establish whether a connection request can be accepted or rejected. The connection admission control handles resources such as channel numbers, buffer space, and bandwidth.AAL2 Connection Admission Control uses functions such as: Routing, Link Admission Control, and Node Admission Control in the AAL2 network layer. AAL2 Addressing and Numbering Plan
An AAL2 connection that spans several nodes is established hop-by-hop AAL2 addresses must be unique in the Iub and Iur AAL2 networks An hierarchical numbering plan is required. All the RBSs under the same CRNC must have a common and unique address prefix assigned to the RNC. AAL2 switching consists of the multiplexing of AAL2 connections from different incoming AAL2 VCs onto one or several outgoing AAL2 VCs. The AAL2 packets are extracted from the incoming AAL2 VCs and then remultiplexed onto different outgoing AAL2 VCs Using AAL2 switching in hubs, it is possible to reduce the bandwidth required and therefore save transmission costs Using AAL2 switching in hubs, it is possible to offer redundancy through load-sharing when one or more physical links are used towards the RNC. Class A: Stringent QoS High priority "minimum delay", e.g. conversational speech Class B: Stringent QoS Low priority, e.g. high bitrate interactive/background packet data on dedicated channels Class C: Tolerant QoS - High priority "unspecified delay", e.g. HSDPA Class D: Tolerant QoS - Low priority "unspecified delay" The purpose of the Iub Flow Control between RNC and RBS is to keep the Priority Queues, in RBS, appropriately filled and to not overload the transport network. Reduced delays Increased data rates Increased capacity

AAL2 Switching

AAL2 QoS separation

AAL2 Path capacity utilization and Iub Flow Control

Enhanced Uplink

 AAL5
Movable CEP

Services(4)

Effectively hides the RNC module concept. RBSs are allocated to RNC modules automatically and evenly in each RNC subrack by configuring a preferred subrack for each RBS. Optimises control plane load distribution within the node. Reduces the need for standby processors in the node.

ATM over Fractional Transmission IP Network QoS

Differentiated Services Code Point (DSCP)

Ethernet Quality bits (Pbits) Configurable Transport Bearer QoS class Network Synchronization over IP (SoIP)
The SoIP Server and Client are communicating using SNTP over UDP, running on the same subnet as the traffic.

Topologies(1) Iub Topologies

Ring Tree Star and chain

Topologies(2)
The following capabilities apply for the WCDMA RAN at Iub using ATM transport:
Max ATM Cross Connect aggregation levels: 4 Max AAL2 Switching aggregation levels: 4 Max RBS/ RXI ring nodes in an ATM ring configuration: 4 plus RNC Max no of RBSs in an ATM ring configuration, including ring nodes: 64 Max no. of RBSs under a hub RBS acting as ATM ring node: 20 Max no of RBSs under a RXI acting as ATM ring node: 32 Max AAL2 switched RBSs in Hub-RBS: 8 Max AAL2 switched RBSs in RXI: 350 Max ATM cross connected RBSs in Hub-RBS: 50 Max ATM cross connected RBSs in RXI: 550

The following capabilities apply for the WCDMA RAN at Iub IP / Ethernet transport (note these figures are preliminary and subject to change):
Max Ethernet switching aggregation levels: 10 Max no of RBSs under a RBS hub node: 50 Max no of RBSs under a RXI node: 768 Max nodes in an Ethernet ring configuration: 5

Topologies(3)

Direct Connections over PDH/SDH networks

Connections over External ATM Network

One-level ATM Cross Connect aggregation using hub RBS or RXI

Two-level of AAL2 Switching using RXI and/or hub RBS

Cascading and Chaining of Hub RBS or RXI

RXI, Hub RBSs, and RNC Interconnected in an ATM Ring Topology

Topologies(4)
Iu and Iur Topologies:
For IuCS, an RNC can only be controlled by one MSC. One MSC can control several RNCs. The exception to this is when the Iu Flex functionality is activated which enables one RNC to be controlled by several MSCs. For IuPS, an RNC can only be controlled by one SGSN. One SGSN can control several RNCs. The exception to this is when the Iu Flex functionality is activated which enables one RNC to be controlled by several SGSNs.

Iu Topology with Single RNSGW

Iu Topology with Dual RNSGW Nodes

Iu Flex ATM Topology with Single RNSGW

Link Redundancy and Protection

Link redundancy is valid for Iu, Iur, and Iub as well as for the Mut, Mur, and Mub interfaces. Protection on Iub Interface
Protection on Iub interface is mainly used from a RXI or Hub RBS node to the RNC. The typical configuration is to have two different STM-1 physical links. Redundant VCCs (for each RBS) for management (Mub) communication, Iub signaling (NBAP-C, NBAP-D and Q.2630) and AAL2 paths over those two different physical links have to be configured in order to achieve Iub link redundancy. AAL2 switching is used for load sharing distribution over the STM-1 links between RXI/Hub RBS and the RNC. e.g. Rapid Spanning Tree Protocol (RSTP), MSP 1+1, Redundancy with IMA Parallel ATM Virtual Channel Connections (ATM VCC) for user plane communication over Iu and Iur are used for redundancy and are allocated to separate physical connections.
AAL2 Connections for user plane communication over Iu-CS and Iur are distributed in load sharing among parallel ATM VCCs. AAL5 Connections for user plane communication over Iu-PS are distributed in load sharing among parallel ATM VCCs.

Protection on Iu and Iur Interface

Parallel ATM Virtual Channel Connections (ATM VCC) for signaling over Iu are used for redundancy and are allocated to separate physical connections.
Parallel SAAL-NNI links, at separate VCC, are configured over Iu and Iur for redundancy reason. The SAAL-NNI links carry, over the MTP-3b layer, the Iu and Iur application protocol (RANAP and RNSAP) and transport network signaling protocol (Q.2630).

When IP is utilized for Iu-PS user plane or SIGTRAN, the Gigabit Ethernet interface board is used. Redundancy is achieved by allocating two parallel physical connections. Furthermore, there is a 1+1 port on the Gigabit Ethernet interface board which can complete protection switching within 700 ms after a link failure.