Fundamentos Bsicos de ATM

ATM
A fin de aprovechar al mximo la capacidad de los sistemas de transmisin, sean estos de cable o radioelctricos, la informacin no es transmitida y conmutada a travs de canales asignados en permanencia, sino en forma de cortos paquetes (celdas ATM) de longitud constante y que pueden ser enrutadas individualmente mediante el uso de los denominados canales virtuales y trayectos virtuales.

Asynchronous Transfer Mode

Voice

Video

Data

ATM cells

ATM 53-octet cells are switched in hardware

48-octet Payload

A compromise for voice, data, and video QoS defined/negotiated when the initial connection is made Asynchronous on layer 2 of the OSI reference model Compromise of STM and PTM

 ATM = Asynchronous Transfer Mode Fast packet switching and multiplexing technology (cell-based ) Support the universe of services
Provides quality of service guarantee and reliability Offers "bandwidth on demand" Connection-oriented, no error correction for user traffic
voice, video and data traffic

What is ATM?

Utilises statistical multiplexing

error correction for user traffic is handled by the end user the advantages are increased speed of switching and elimination of associated delay
less bandwidth can be reserved than if bandwidth reservation would be based on the peak rate of the connections. transmission cost saving is achieved

Synchronous and Asynchronous Multiplexing

A A

B
C

Synchronous Multiplexing

Asynchronous Multiplexing

Asynchronous Transfer Mode

Benefits:
+ Over-subscription in the service provider network possible + Guaranteed bandwidth, delay and jitter are possible + Proven technology

Drawbacks:
Most applications are based on IP instead of ATM A lot of overhead for IP over ATM Complicated protocol architecture

Implemented in voice and data networks Transport layer in 3G networks

Why is ATM used as transport network in 3G?

ATM provides efficient support for transmission of voice, data, and video ATM provides QoS guarantee and reliability ATM utilises statistical multiplexing, so
less bandwidth can be reserved transmission cost saving are considerable

ATM supports the soft handover functionality

ATM Cell
53 bytes Header 5 bytes Payload 48 bytes

Header contains routing and error control information Payload carries the actual user information, either voice, data or video

ATM Cell
8

Provides local functions, such as identifying multiple stations that share a single ATM interface

The 1st bit - indicates whether the cell contains user data or control data The 2nd bit - indicates congestion

4 VPI

3 2

8 VPI

3 2

GFC

VPI
Header (5 bytes) VCI VCI

VCI

VPI
VCI

VCI

PT

CLP

VCI

PT

CLP

HEC
Payload (48 bytes) Payload

HEC
Payload

User Network Interface (UNI)

GFC VPI VCI

Network Node Interface (NNI)

Indicates two levels of priority for ATM cells, CLP=1 should be discarded in preference to cells with the CLP=0

Generic Flow Control PT Payload Type Virtual Path Identifier CLP Cell Loss Priority Virtual Channel IdentifierHEC Header Error Control

ATM Cell Header

GFC provides local functions. VPI indicates the virtual path over which the cell should be routed. VCI identifies a virtual channel over which the cell is to travel.

PT discriminates between a cell carrying management information or one, which is carrying user information.
CLP indicates two levels of priority for ATM cells.

HEC checks for an error and corrects the contents of the header by using a CRC algorithm.

Relation Between a Transmission Path, VPs and VCs

11

2006 Nokia Introduction to the Transmission Architecture used in WCDMA / 2006-10-23 / TKo,RBe,FAs

VC and VP Switching
VC switch/cross-connect

Endpoint of VPC

VCI 21

VCI 22

VCI 23

VCI 24

VCI 21 VCI 22 VCI 21 VCI 22

VPI 1

VPI 2

VCI 24 VCI 23 VCI 21 VCI 22

VPI 3
VPI 4 VPI 5

VC switch/cross-connect

VP Switching
VCI 21 VCI 22
VCI 23 VCI 24 VCI 25 VCI 24 VPI 1 VPI 4 VCI 23 VCI 24 VCI 25 VCI 24 VCI 21 VCI 22

VPI 2

VPI 5

VPI 3

VPI 6

Vp switch/cross-connect

ATM Adaptation Layer at ATM end Systems

AAL2 Cell Packing

Introduction to transmission technology SDH/Sonet

European Standard: SDH (Synchronous Digital Hierarchy) North-American Standard: Sonet (Synchronous Optical NETwork) Granularities
SDH STM-0 STM-1 STM-4 STM-16 Sonet STS-1 / OC-1 STS-3 / OC-3 STS-12 / OC-12 STS-48 / OC-48 Bit rates 55.84 Mbit/s 155.52 Mbit/s 622.08 Mbit/s 2488.32 Mbit/s

Introduction to transmission technology PDH

Plesiochronus Digital Hierarchy
European (ETSI)
E1 balanced, Impedance 120 E1 un-balanced, impedance 75 bit rate 2048 kbits/s

North American (ANSI) standards

T1 balanced interface, Impedance 100 bit rate 1.544 Mbits/s

E1 Frame Structure
E1 Frame Structure
32 Timeslots
30 timeslots for user data 1 timeslot for synchronization (TS 0) 1 timeslot reserved for signalling (e.g. Nokia Q1 EOC) Each timeslot carries 8 bits, equivalent to 64kbit/s channel bandwidth

125 s

timeslots

0 1 2
bits

16

31

1 2 3 4 5 6 7 8

ATM bandwidth unit

53 bytes Header 5 bytes Payload 48 bytes

In transmission technologies PDH / SDH the bandwidth parameter is usually given in bits per time, e.g. 1.920 Mbits/s. In ATM the bandwidth definition uses the unit of ATM cells per time, e.g. 4528 cps (cells per second). The number of bits in an ATM cell is the factor to convert from ATM cells per second to bits per second. The number of bits in an ATM cell is calculated from: 53 Bytes/ATM cell x 8 bits/Byte = 424 bits per ATM cell

ATM bandwidth calculations: single E1

TS 0 1 2 Data Frame alignment, CRC, remote alarm Payload Payload

Total E1 : 32 TSs x 8 bits/TS 125 s

. . .
15 16 Payload reserved for signaling, e.g. Nokia Q1

15 TSs

= 2048 kbits/s Payload E1 : 30 TSs x 8 bits/TS 125 s = 1920 kbits/s

17

Payload

. . .
29 30 Payload Payload

15 TSs

ATM Cell Rate: 1920 kbits/s 424 bits/ATM cell = 4528 cps

31

Payload

IMA group Capacity

Inverse Multiplexing (licensed feature) allows to group several E1, T1 or JT1 interfaces in order to provide an ATM interface of higher bandwidth than a single interface Max number of IMA links is 8 per IMA group Bandwidth T1/JT1 IMA group
int ( n * 3622 cps * (m - 1 / m) * (2048 / 2049))

Bandwidth E1 IMA group

int ( n * 4528 cps * (m - 1 / m) * (2048 / 2049)) n = number of IMA links m = IMA frame length (currently fixed to 128 in Flexi WBTS)

ATM bandwidth calculations: STM-1 / STS-3 at VC4 level

RSOH

Total STM-1 / STS-3: 270 Bytes x 8 bits/Byte x 9 lines

PO H VC4 Payload

Pointer
MSOH

125 s = 155.52 Mbits/s VC 4 payload: 260 Bytes x 8 bits/Byte x 9 lines

9 lines
= 149.76 Mbits/s ATM Cell Rate:

125 s

149.76 Mbits/s 424 bits/ATM cell

= 353207 cps

260 Bytes

Objective of this Module

Objective of this module is to introduce essential ATM information for 3G networks using examples
Why ATM is used for UMTS VP/VC structure AAL2/AAL5 IMA, ATM over STM-1 VC-4 Cross connection at VP and VC level ATM service categories - CBR, UBR, UBR+ Traffic descriptor parameters Policing/shaping

Objective of this Module

Objective of this module is to introduce essential ATM information for 3G networks using examples
Why ATM is used for UMTS VP/VC structure AAL2/AAL5 IMA, ATM over STM-1 VC-4 Cross connection at VP and VC level ATM service categories - CBR, UBR, UBR+ Traffic descriptor parameters Policing/shaping

CIDs

Hierarchy
VP
Transmission path

VCs

AAL2 connection
Virtual channel connection

AAL2 link

Virtual channel link

VCCTP
VCLTP

VCLTP

VCLTP

VCLTP

VCCTP

Virtual path connection

VPCTP VPLTP

Virtual path link

VPLTP VPLTP VPLTP VPCTP

Transmission path

Physical link

The concept of Virtual Channels and Virtual Paths

Header Payload

Virtual path

Transmission path
Virtual path

ATM cells

Virtual connection (VC)

An ATM multiplexer continuously generates ATM cells an maps traffic in or fills with dummy content The ATM cell belongs to a specific Virtual Channel

In ATM any data is mapped into an ATM cell with the accordant ATM adaptation layer

The ATM Virtual Channel is mapped into a Virtual Path

The Virtual Channel is only terminated at application layer, in our case the WAM or FTM in BTS or the A2SU in RNC The Virtual Path is only terminated at any VC-Cross Connection Point

The ATM Virtual Path is mapped into a Physical Interface

The identifiers for VP and VC are part of the ATM cell header. At any switching point only the ATM cell header is read and the cell is processed accordingly. The payload remains untouched until VC is terminated

The physical media can be a PDH link, an IMA group or an SDH link

VP and VC Identifiers
Virtual channel
Virtual path

UNI Cell Header

7 6 5 4 3 2 1 0
Generic Flow Control

NNI Cell Header

7 6 5 4 3 2 1 0 VPI VPI VCI VCI 5 Bytes
Payload Type Identifier

VPI

VPI
VCI VCI

VCI

Transmission path
Virtual path

Payload Type Identifier

CLP

VCI

CLP

Header Error Check

Header Error Check

AAL2 connection

Payload
(48 bytes)

Payload
(48 bytes)

48 Bytes

Channel identifier (CID) is unique within an AAL2 VCC (only if AAL2 is used; handled later on) VCI is unique within VPC

VPI is unique within physical interface

Each VPC can contain numerous VCCs, identified by their VCI, depending on the allocated bits. ATM defines a maximum of 16 bits. The maximum of allocated bits is 13 in the RNC and 12 in the AXC (AXC Compact 7 bits fixed). Each physical interface connection can contain numerous VPCs, identified by their VPI, depending on the allocated bits. ATM defines a maximum of 8 (UNI) respectively 12 (NNI) bits for VPI numbering. Available bits depend on available maximum of network element and assigned bits for VCI

Note: AXC provides a maximum of 13 bits the RNC a maximum of 14 bits for VPI/VCI numbering

Identifiers have only local significance between two termination points

VP and VC Identifiers

GFC (Control de Flujo Genrico, Generic Flow Control, 4 bits): El estndar originariamente reserv el campo GFC para labores de gestin de trfico, pero en la prctica no es utilizado. Las celdas NNI lo emplean para extender el campo VPI a 12 bits. VPI (Identificador de Ruta Virtual, Virtual Path Identifier, 8 bits) y VCI (Identificador de Circuito Virtual, Virtual Channel Identifier, 16 bits): Se utilizan para indicar la ruta de destino o final de la clula. PT (Tipo de Informacin de Usuario, Payload type, 3 bits): identifica el tipo de datos de la celda (de datos del usuario o de control).Uno identifica el tipo de carga en el campo de usuario, otro indica si hay congestin en la red y el ltimo es el SDU. CLP (Prioridad, Cell Loss Priority, 1 bit): Indica el nivel de prioridad de la celda, si este bit est activo cuando la red ATM esta congestionada la celda puede ser descartada. HEC (Correccin de Error de Cabecera, Header Error Correction, 8 bits): contiene un cdigo de deteccin de error que slo cubre la cabecera (no la informacin de usuario), y que permite detectar un buen nmero de errores mltiples y corregir errores simples.

VP and VC Identifiers
PT (Tipo de Informacin de Usuario, Payload type, 3 bits): identifica el tipo de datos de la celda (de datos del usuario o de control).Uno identifica el tipo de carga en el campo de usuario, otro indica si hay congestin en la red y el ltimo es el SDU.

THE ATM ADAPTATION LAYER (AAL)

The AAL was conceived to provide a good interface between the different kinds of applications (voice, video and data) and the ATM network. Since the lower layers do not provide control functions, such as error and flow control, which are generally needed by applications, this layer is designed to bridge the gap. To define a suitable interface, the ITU-T considered that applications fell into three broad groupings as

AAL service protocols

THE ATM ADAPTATION LAYER (AAL)

Revised AAL protocols

ATM adaptation layers:

for circuit emulation for example "virtual E1 TDM trunks" through the ATM backbone Most recent AAL The most important AAL2 application is compressed VoATM

AAL 1 / 2 / 5
For variable length application packets Used for IPoATM (IP Over ATM) In R99 UTRAN for all Iu-PS user plane traffic + the Iu & Iub signaling

e.g. 2G traffic connections together with UTRAN

In R99 UTRAN for all user plane traffic in Iub, Iur & Iu-CS

Provides mapping of applications to ATM service of the same type Segments/reassembles into 48 payloads Hands 48 byte payloads to ATM layer

ATM Adaptation Layer 2 (AAL2)

Sector 1 flow Sector2 flow Sector 3 flow
Start field CPS-PH CPS-PP Zero padding

AAL2 ATM-cell flow

Cell header Cell payload Crossing cell boundary Offset to next CPS-PDU

One AAL2 cell flow consists of variable length "mini-cells" that are concatenated as a continuous stream into the 48-byte payload areas of ATM-cells One AAL2 flow can carry up to 248 channels, which are multiplexed by inserting channel data into mini-cell payloads The "mini-cells" are officially called CPS-packets (Common Part Sublayer) divided into packet header (PH) and packet payload (PP) parts Mini-cells have a "mini-header" (CPS-PH) to identify the channel and length of the mini-cell Mini-cell payload size can be anything from 1 to 45 bytes Channels can be multiplexed in any order desired, mini-cells can cross ATM-cell boundaries, cells can be padded All AAL2 ATM-cells begin with a start field which indicates the offset to 1st complete mini-cell within the ATM-cell

CID
8 bits

ATM Adaptation Layer 2 (AAL2) LI UUI HEC CPS-INFO

6 bits 5 bits 5 bits 1 to 45 / 64 octets

CPS Packet Header (CPS-PH) 3 Bytes CPS Packet

4 to 48 Bytes

CPS Packet Payload(CPS-PP)

Note: CID Channel Identifier LI Length Indicator UUI User-to-user information HEC Header Error Check

When AAL2 is used, 248 AAL2-connections can be mapped simultaneously in the same VCC Each of these AAL2-channels is identified by its Channel Identifier (CID). CID identifies user channel 8-255; CID 0-7 are reserved for Layer Management purposes.

CID is unique within an AAL2 VCC On Iub and Iur each active connection requires 2 CIDs (DTCH + DCCH) On Iu-CS each active connection requires 1 CID

The reserved CID and capacity within the VCC is a dynamic issue That means CID and capacity is available again after termination of specific connections Identifiers and capacities for VC and VP are static!

Packet #1

Packet #2

Zero padding

Zero padding

cell #1
Cell header Cell payload

cell #2
Crossing cell boundary

cell #3
AAL5 packet trailer

cell #4
AAL5 packet trailer

AAL5 ATM Adaptation layer is designed to carry fast streams of longer packets over ATM as simply as possible AAL5 doesn't provide any multiplexing itself, it is assumed that the packets carried identify themselves to higher protocol layers (like TCP/IP) AAL5 is also known as SEAL (Simple Efficient Adaptation Layer) Packets are simply segmented into 48-byte fragments of ATM-cell payloads, the last cell is padded and terminated with AAL5 trailer of 8 bytes AAL5 trailer indicates packet length & has 32-bit CRC Packets can be up to 64 Kbytes and are aligned to cell boundaries The last cell of a packet is indicated by setting the PT[0] bit in cell header

UU CPI length CRC-32

ATM Adaptation Layer 5 (AAL5)

Protocol Stacks of User Plane for CS/PS services

UE BTS RNC MGW / MSC / SGSN GGSN

UE
E.g., Vocoder RLC-U MAC
WCDMA L1

MGW RNC Node B

FP
WCDMA L1

MSC
A/mlaw PCM, UDI, etc. Link Layer PHY PHY

RLC-U MAC FP AAL2 ATM PHY

Iu-CS UP AAL2 ATM PHY

E.g., A/mlaw Vocoder PCM, UDI, Iu-CS UP etc. AAL2 ATM PHY Link Layer PHY

PSTN/ N-ISDN

AAL2 ATM PHY

UE
E.g., IPv4, IPv6

GGSN RNC
Node B
WCDMA L1

3G-SGSN
GTP-U UDP IP
LLC/SNAP

E.g., IPv4, IPv6

PDCP RLC-U MAC

WCDMA L1

FP AAL2 ATM PHY

PDCP RLC-U MAC FP AAL2 ATM PHY

GTP-U UDP IP
LLC/SNAP

GTP UDP IP
Link Layer PHY

GTP UDP IP
Link Layer PHY PHY

Also for NRT services AAL2 is used on ATM as Transport Layer

AAL5 ATM PHY

AAL5 ATM PHY

ATM over PDH ATM capacity on PDH

Most common transport for BTSs To transmit ATM traffic over an E1 link, the ATM cell is segmented Segments are simply mapped bit by bit into TS 01-15 and TS 17-31 and transmitted in the timeslot structure of an E1 frame Time slots 0 is used for frame alignment The transportation of ATM cells over an E1 link allows you to reuse the current PDH transmission network E1 links are also used to supply ATM-traffic to ATM network Total ATM capacity in an E1 frame:
30 bytes with ATM payload in a frame Frame repetition rate is 125 us 30 bytes/125 us = 240 000 bytes/s One ATM cell has 53 bytes/cell

The ATM traffic capacity in a 2 Mbps frame is:

240000 bytes/s/53 bytes = 4528 cps = 1.92 Mbps

IMA Inverse Multiplexing for ATM

b a
N 2 1

IMA

1 2

IMA
N

c
2 1

TX direction cells distributed across links in round robin sequence

RX direction cells recombined into single ATM stream in same sequence

It is not possible to split a logical interface among several physical interfaces Several physical E1 links are combined to one higher bandwidth logical interface known as IMA Group The purpose of Inverse Multiplexing for ATM (IMA) is to provide inverse multiplexing of an ATM cell stream over multiple physical links and to retrieve the original stream at the far-end from these links The ATM Inverse Multiplexing technique involves inverse multiplexing and demultiplexing of ATM cells on a cell-by-cell basis in a cyclic fashion among links grouped to form a higher bandwidth logical link.

IMA frame - ICP cell, Filler cell

Cell M-1 Cell 7 Cell 6 Cell 5
ICP

Cell 4

Cell 3

Cell 2

Cell 1

Cell 0

IMA frame no IMA 32 64

capacity 1920,000 1859,092 1889,078

loss in % 0,00 3,17 1,61

ATM

Filler ATM

ATM

ATM

ATM Filler ATM

Filler

ATM

ATM

ATM

ATM

ICP

Filler ATM

ATM 128 256 1904,070 1911,567 0,83 0,44

An IMA Frame

The ICP (IMA Control protocol) cell is sent once per IMA frame on each link with a different offset between different links, to adjust for differential link delays If there are no ATM layer cells to be sent the transmit IMA sends Filler cells. Note: IMA is not saving bandwidth but enables to utilize the available bandwidth more efficiently and protects from link and interface failure!
ICP fields contain e.g. link ID, IMA ID, IMA Frame Sequence Number, ICP Cell Offset

M consecutive cells transmitted on each link within the IMA group M can be 32, 64, 128 or 256. Nokia default: 128

IMA failure case

Upon the failure of one or more links or interfaces of an IMA group, the IMA Engine is able to recover and reassign the distribution of ATM cells to the remaining working PDH links within the IMA group.
The recovery time of the IMA engine upon PDH link failure is 2 s, all existing calls will be terminated. The minimum number of active IMA links that enable a group remaining active is specified by an IMA parameter and simultaneously the bandwidth available for that IMA group.
The parameter minimum number of links can be set in in AXC as well in RNC This parameter describes until what limit the IMA group continuous working in case one or more E1s drop out. Only the capacity of this IMA group is reduced but the IMA group remains operational.

Example:

IMA group with 4 E1 links An ATM interface with a VP of 17000 cps is assigned to that group Note: IMA group with 4 E1s provides a capacity of 17961 cps Minimum number of links is set to 2 In case 1 or 2 E1s drop out the IMA group continuous operating after restart with a limited bandwidth even if the logical interface cannot provide anymore the assigned VP capacity In case traffic increases available capacity cells are dropped, active connections might be released due to cell loss.

ATM over SDH (STM-1 / VC4)

When SDH is used as transport media the ATM cells are simply mapped Byte by Byte into the virtual container STM-1 has a transmission capacity of 155.52 Mbps. Each frame of STM-1 is organized into a 9 rows x 270 columns byte array 260 columns out of 270 carry the actual payload.

ATM Capacity in STM-1 VC-4 Frame

Payload of the STM-1 frame can accommodate 1 Virtual Container Level-4 (VC-4)
Total ATM capacity in a SDH frame: Payload per virtual container: 260x9 = 2340 bytes Frame repetition rate is 125 us. 2340 bytes/125 us = 18 720 000 bytes/s One ATM cell has 53 bytes/cell . The ATM traffic capacity in a SDH frame is:
18 720 000 bytes/s/53 bytes = 353 207 cps

In case STM-0 is used (3x VC3 within STM-1) 114113 cps are available per logical interface

ATM Cross Connection

From the origin to the termination of a Virtual Channel Trail there can be multiple intermediate Nodes where ATM Cross Connections might occur Traffic is collected from different origins to same destination
Statistical multiplexing gain can be achieved
BTS
E1

RNC
STM1

BTS
E1

BTS
E1

BTS
E1

BTS
E1

VP and VC Cross Connections

ATM cross connections can be performed at VP or VC level
VP level switching is preferred in some cases as it makes management and configurations easier

The physical link is terminated at any intermediate Node Virtual Paths can be cross connected to any Physical Interface Virtual Channels can be cross connected to any Virtual Path Which layer is terminated depends on the Cross Connection Level Cross Connections are carried out according to a Cross Connection Table

VC Cross Connection

VPI 1
VCI 21 VCI 22 VCI 21 VCI 22

VPI 2

VPI 3
VCI 24

VPI 1 VPI 4
VP Cross Connection

VCI 23
VCI 21 VCI 22

1. Physical layer Cross-Connection

Simply E1 mapping Recommended solution, if site already has cross-connection equipment (e.g. MW radio indoor units in Site Support cabinet). An ideal solution from network evolution viewpoint would be FIU 19 BTSs are separated in different physical lines so network management is easy Loop protection on physical layer is possible (not with FIU 19) No statistical multiplexing gain is available Also used if e.g. only channelized SDH is available. E1s are mapped into VC12. No effect on ATM layer, no changes in VPI/VCI numbering IMA group members can have separate routes

Physical link can be: SDH: STM-1(VC4 / VC3 / VC12) PDH: E1 or IMA group BTS
E1

BTS
E1 E1

RNC

Physical link

Physical link

E1

Can be e.g. FUI19 or other cross connect

2. VP level Cross-Connection
Physical link is terminated VPs can be mapped into any other physical interface when bandwith requirements are met In this example BTSs are separated in different virtual paths so the network management is easy When configuration is changed, only few settings are needed Only VP layer is affected, new VPI to be assigned Virtual Channel remains transparently within VP Statistical multiplexing gain in case of UBR overbooking Most common Cross Connection Level in intermediate Nodes in RAN

Physical link can be: STM1(VC4 / VC3 / VC12 E1 or IMA group

BTS

RNC

BTS
E1

E1

Physical link

STM1

Physical link

VP

3. VC level Cross-Connection
Physical and virtual paths are terminated Virtual Channels can be mapped into any other VP New created VPs can be mapped into any physical interface New VPIs and VCIs to be assigned All user traffic in one VP may not be best solution if something goes wrong on ATM level Complex configuration work and high risk of mess when changes have to be done Statistical multiplexing gain in case of UBR-VCCs The AAL2 connections are not affected on this level!

BTS
RNC

BTS
E1

E1

Physical link Physical link

STM1

VCs

VP

4. AAL2 Switching
Nodal function in Media Gateway (MGW) for Iur connections AAL2 switching also used
Iu-CS and Iur connections
UP VCC RNC

in Ultrasite BTS multiplexing traffic from different WAMs Iur

MGW

SIG VCC
RNC

AAL2 switching: Iur traffic can be transported within the Iu-CS UP-VCC. Due to the Nodal Function in the MGW, the Iur traffic can be switched on AAL2 level (Digit Analysis Tree) within the MGW to the other RNC.

Alternative is to use VCC switching. The Iur traffic can be transported in its own VCC together with the Iu-CS traffic. In this case the MGW acts like a Cross-Connect

VPI and VCI values

VPI/VCI is assigned at endpoints VPI/VCI is modified at connection points

VPI = 57 VCI = 65

VPI = 37 VCI = 41

VPI = 12 VCI = 41

VPI = 7 VCI = 65

ATM UNI

VC-level XCON

ATM NNI

VP-level XCON

ATM NNI

VC-level XCON

ATM UNI

VPI re-use One VPI value can be used once in the same ATM interface. Can be re-used in another ATM interface VCI re-use One VCI value can be used only once within the same VP, but can be re-used in another VP VP Cross Connection: VPIs will be re-assigned on the other interface, but could have previous value if available. VCIs remain the same VC Cross Connection The VP will be terminated The VCI will be re-assigned on the VP but could have previous value if available.

Virtual Channels and Paths in the Iub Interface

AAL2 channels remain with their CID in their dedicated VCC until VCT is terminated
VCT is terminated only in RNC (A2SU) and in BTS (WAM or FTM)

VCCs remain with their VCI in their dedicated VPC until VPT is terminated VPCs remain with their VPI in their dedicated physical interface until terminated On the Iub side VC numbering starts from VCI32 up to limitation regarding to assigned bits In case AAL2mux is used, VCT is also terminated in AXC

ATM terminology for AXC and RNC

Permanent Virtual Connection
VC XCON
VP XCON

BTS

RNC

AXC

AXC

VP TTP

VP Trail

VP TTP

VP Connection VP CTP VP Trail VC Connection VC Trail

VP Connection / VP Link

VPLtp
VPLtp

VP TTP VC TTP
VC TTP VC Connection VC CTP

VC Connection / VC Link

VCLtp VCLtp

To guarantee the required QoS, there is a method defined to describe connections behavior

Quality of Service

Service Classes

Service category classifies connections as CBR, UBR or UBR+

Traffic Parameters

Traffic parameters define mainly the bandwidth requirements e.g. PCR / MCR

QoS Parameters QoS Class

QoS parameters define finally the QoS of the Connection such as delay, cell loss etc.

The combination of the Traffic Class and the set of Traffic and QoS parameters constitute the Connection Traffic Descriptor of an ATM connection

ATM service categories - CBR

CBR Constant Bit Rate
Used for constant (maximum) bandwidth services For the connections that request a fixed (static) bandwidth, that is continuously available during the connection lifetime Is intended to support real-time applications requiring tightly constrained delay variation The applications are e.g. video, audio, circuit emulation, but use of CBR does not restricted to those applications Source may emit cells at or below the PCR at any time and for any duration (or may be silent) PCR is guaranteed Until RAS06 all user plane VCCs were CBR in Iub
Bandwidth

Time

ATM service categories - UBR

UBR Unspecified Bit Rate
Is intended for non-real-time applications, i.e., those not requiring tightly constrained delay and delay variation. Examples of applications are traditional computer communication applications, such as file transfer and email Service does not specify traffic related guarantees Sources are expected to transmit non-continuous bursts of cells PCR is not guaranteed and can send more use more capacity than PCR Traditionally DCN and Iu-PS connections have been UBR
Bandwidth

Time

ATM service categories UBR+

UBR+ Unspecified Bit Rate +
Is intended for non-real-time applications, i.e., those not requiring tightly constrained delay and delay variation. Examples of applications are traditional computer communication applications, such as file transfer and email Specified with PCR and MDCR MDCR is guaranteed and traffic can be transmitted up to PCR Picture is indicating that even though the MDCR is reserved for the traffic, the traffic can reach up to the PCR, if there is free capacity to be used e.g. in the shaped VPC or Bundle
Bandwidth

PCR

MDCR
Time

Note! UBR+ has also UBRshare proprietary parameter that is used to share excess capacity between different UBR+ VCC in the line card

Traffic and QoS Parameters

Traffic parameters describe traffic in terms of:
PCR - Peak Cell Rate MDCR - Minimum Desired Cell Rate
Maximum bandwidth in any situation Parameter defines the guaranteed cell rate

QoS Parameter:

CDVT - Cell Delay Variation Tolerance CLR - Cell Loss Ratio:

This parameter is set according to network element requirements (details follow!) Describes the ratio of lost cells to transmitted cells The CLR parameter is the value that the network agrees to offer as an objective over the lifetime of the connection If value will be exceeded an alarm will be raised or possibly further action will be triggered depending on parameter settings Usual values between 10-3 to 10-9

Traffic Management
Traffic management is needed in order to comply to the QoS requirements QoS is guaranteed as long as the connection is compliant with the Traffic contract The traffic offered to the network can be variable and therefore end-to end participation of network elements is required Traffic management functions with in ATM node are distributed among different elements

Interface

Interface

Switch fabric Ingress Cells in

Policing Buffering Congestion control Queuing Traffic priorities
Shaping Buffering

Egress Cells in

Control module

CAC

Traffic Contract
Traffic contract is negotiated during connection establishment Its an agreement between a user and a network, where the network guarantees a specific QoS if the user's cell flow conforms to a negotiated set of traffic parameters A traffic contract can be also a written contract between an Operator and e.g. a backbone / Leased Line provider. Traffic contract of includes:
! ! !
Definition of a Compliant Connection QoS class Traffic Descriptor ! Conformance Definition ! Cell Delay Tolerance ! Cell Loss Ratio ! Peak Cell Rate ! Service Category
A

Example on Traffic Management Parameters

In the transport network e.g. AXC polices the ingress traffic and shapes the egress traffic
No policing for ingress traffic in BTS
Shaping
Policing

Shaping CBR traffic in CBR VP in RNC

Policing
Shaping

BTS
E1

Shaping
Policing

Policing Shaping

RNC

Source level shaping in BTS

ATM switch

ATM switch

Policing for ingress CBR traffic in RNC

Example ATM application

Native ATM Switching Application in Native ATM Switching Application

Example ATM application

Native ATM Switching Application in Native ATM Switching Application The Tellabs 8600 system supports ATM switching of ATM Virtual Path Connections (VPCs), i.e. Virtual Path Circuits and Virtual Channel Connections (VCCs), i.e. Virtual Connection Circuits. The circuits to be switched are statically provisioned using Tellabs 8000 network manager or CLI. In a typical switching application a number of VPCs and VCCs are groomed from low speed ATM interfaces e.g. E1/T1 or P12s/DS1 in chSTM-1/chOC-3 to one high capacity uplink ATM interface such as STM-1/VC4/ATM or OC-3/STS-3c SPE/ATM. The switching application can also segregate the traffic from a number of low speed interfaces to two uplink ATM intrefaces going to a different service, e.g. mobile traffic to Radio Network Controller (RNC) and traffic from ATM DSLAMs to Internet Service Provider (ISP) Broadband Remote Access Server (BRAS) device. In the switching applications the Tellabs 8600 system is equipped with a standard ATM switch feature set such as ATM traffic management.

Example ATM application

ATM Pseudo Wire Application

Example ATM application

ATM Pseudo Wire Application The Tellabs 8600 system supports interconnection service of native ATM connections over an MPLS-based network. It is also possible to expand the connectivity over 3rd party MPLS core network between two Tellabs 8600 subnetworks.
The application takes advantage of the ATM pseudo wires which create circuitswitched MPLS tunnels over the MPLS network. Native ATM VCCs or VPCs can be inserted to the MPLS tunnels of an MPLS network at ingress when an ATM interface module is used. This takes place also vice versa, i.e. VCCs and VPCs are extracted from the tunnel at egress. One MPLS tunnel is created individually for each VPC and VCC. The Tellabs 8600 system acts as a native ATM interface towards an external device. The circuits to be tunnelled are statically provisioned using Tellabs 8000 network manager or CLI. The tunnels are provisioned statically or by using LDP and MPLS network may dynamically change the route of the MPLS tunnel based on its internal routing decisions. In a pseudo wire application the Tellabs 8600 system is equipped with a standard ATM switch feature set such as ATM traffic management towards external ATM devices. The ATM Quality of Service (QoS) is preserved over the MPLS network using Differentiated Services (DiffServ) queuing mechanism and MPLS traffic engineering.

Example ATM application

ATM Pseudo Wire Application in DSLAM and Mobile Application

Example ATM application

ATM Pseudo Wire Application in DSLAM and Mobile Application
The Tellabs 8600 system can setup an ATM pseudo wire also over Layer 3 connectivity. In the case of Layer 3, the packet-switched network (PSN) tunnels are emulated by IP connections. A typical application for this context is IP-routed DSL access between the Tellabs 8600 network elements. In DSL access the IP address is typically distributed either by Point-to-Point Protocol over Ethernet (PPPoE(oA) or by Point-to-Point Protocol over ATM (PPPoA) session with IP Control Protocol (IPCP). A fixed IP address can be provided, i.e. the given user always gets the same address. In the solution, a PPP session is terminated in the DSL router that terminates the ADSL line. Thus, the PPP session is between the L2TP network server (LNS) and the DSL router. Since it is necessary to have IP connectivity to the Tellabs 8600 system at cell site, the DSL router has to operate in half bridge mode where it relays the IP address which it gets from LNS to Tellabs 8620 access switch. Dynamic Host Configuration Protocol (DHCP) can be used between the DSL router and the Tellabs 8600 switch or the IP address can be manually preconfigured. This same IP address can be used to manage the Tellabs 8600 system. The static route for that IP address is needed in Tellabs 8660 edge switch. In addition, Tellabs 8660 edge switch needs to be configured to support MPLS over that IP connectivity. The figure below depicts the case where 3G/UMTS High-Speed Downlink Packet Access (HSDPA) data is carried using DSL access.

Example ATM application

3G and 2G Mobile Backhaul with DSLAM Aggregation

Perspectiva de la tecnologa ATM

El Modo de Transferencia Asncrona fue la apuesta de la industria tradicional de las telecomunicaciones por las comunicaciones de banda ancha. Se plante como herramienta para la construccin de redes de banda ancha (B-ISDN) basadas en conmutacin de paquetes en vez de la tradicional conmutacin de circuitos. El despliegue de la tecnologa ATM no ha sido el esperado por sus promotores. Las velocidades para las que estaba pensada (hasta 622 Mbps) han sido rpidamente superadas; no est claro que ATM sea la opcin ms adecuada para las redes actuales y futuras, de velocidades del orden del gigabit. ATM se ha encontrado con la competencia de las tecnologas provenientes de la industria de la Informtica, que con proyectos tales como la VoIP parece que ofrecen las mejores perspectivas de futuro. En la actualidad, ATM es ampliamente utilizado all donde se necesita dar soporte a velocidades moderadas, como es el caso de la ADSL, aunque la tendencia es sustituir esta tecnologa por otras como Ethernet que est basada en tramas de datos.

ATM resources in the RNC

To create the ATM resources on the Iub, we should use the RNC RNW Object browser. For other interfaces we use MML. 3. all RNC Nevertheless, the same parameters for ATM resources are used for the Create VPLtp interfaces:

Iub VP/VC to created with CoCo objects The needed resources are
ATM interface Access profile VPLtp for CBR traffic VPLtp for UBR traffic VCLtp for CBR traffic VCLtp for UBR traffic VCLtp for UBR+ traffic (RAS06!) 1. Create ATM IF & 2. Access Profile

Iub Iu-CS Iu-PS Iur Iu-BC All related SS7 signaling channels O&M

Physical layer SDH/ PDH / IMA

ATM VPC IF VPC

VCC VCC VCC

VCC

4. Create VCLtp

Note: VPLtp stands for Virtual Path Link Termination Point VCLtp stands for Virtual Channel Link Termination Point

3GTPL Session 7 RNC Detailed Parameter Planning JKl 05.12.2008