CHAPTER 5: Signaling in ATM Networks

CHAPTER 5:Signaling in ATM Networks

Objective: Users must have the capability of signaling connection across the network.

1. Switched Virtual Circuits (SVC) => by Signaling

2. Permanent Virtual Circuits (PVC) => by Network

Management

THREE CLASSES of SIGNALING PROTOCOLS

1. ITU Protocols (descendent from ISDN protocols)* Q.2931 and related “Recommendations=Standards”

e.g., Q.931 used for N-ISDN.

2. ATM FORUM PROTOCOLS* UNI 3.0 (Sept’93) & UNI 3.1 ( Q.2931) (Sept.’94)* UNI 4.0 (April’95)

3. Vendor Specific Protocols (Proprietary)* SPANS (Simple Protocol for ATM Network Signaling

from FORE)* Several protocols developed at Research Centers and

Universities.

THREE CLASSES of SIGNALING PROTOCOLS

Remarks: * Only Difference between 3.0 and 3.1 Data link protocol SSCOP used for reliable transport of ATM signaling packets.• Incompatible message formats between UNI 3.0 & Q.2931 but functionality is very similar.• UNI 3.1 attempts to reconcile UNI 3.0 & Q.2931 but not completely successful. => Three incompatible standards.• UNI 4.0 attempts supporting QoS. (Various aspects of an ATM network have been split into Signaling 4.0, Traffic Management 4.0, the PNNI, the ILMI, and the various physical interface documents.)

Need a universal standard to ensure interoperability.

UNI and NNI Signaling

UNI = “Private” User/Network InterfaceNNI = “Public” Network/Network Interface

PrivateATM

Switch“user”

PublicATM

Switch

PublicATM

Switch

PublicATM

Switch

PublicATM

Switch

UNI signaling UNI signaling

Private UNI Public UNI

UNI signaling

Public UNI

NNI signaling

• UNI and NNI signaling protocols are very similar in functionality.

• P-NNI is the routing protocol.

Basic Connection Setup Protocol

NETWORK

Setup

Call Proceeding

Setup

Call Proceeding

(Status Indication but not finished processing)

Connect

Connect ACK

Connect

Connect ACK

StartCall

SetupComplete

Call Received

Call Accepted

Local ACK, optional

SENDER RECEIVER

•Connection setup completed in one round-trip.•Signaling performed through dedicated Virtual

Channels. - UNI signaling VC: VPI=0, VCI =5

Basic Connection Setup Protocol (Cont.)

Remark: Signaling between the end-system and the ATMswitch usually takes place over VPI=0 and VCI=5,although this is not always the case.

If we implement a feature such as SVC tunneling, the signaling

channel will often be over a VPI =/ 0.

Also, some proprietary UNI and NNI protocols use VCI =/5 for signaling.

Note that the use of more than one VP at an interface does

not imply that we require multiple signaling connections,since a single signaling link may service multiple VPCs.

Signaling Channels

1. Reserved VPI/VCI• x/1 = Meta-signaling• x/2 = Broadcast signaling (not used initially)• 0/5 = ATM endpoint to local network

signaling both point-to-point and point-to-multipoint

signaling (NONASSOCIATED SIGNALING MODE: all VC connections are created, controlled,

released via this channel)• x/5 = point-to-point signaling with other

endpoints and other networks (ASSOCIATED SIGNALING MODE: only VC

connections within the VP x are created, controlled and released via this channel).

Meta-Signaling• Used to set up signaling channels

• All meta-signaling messages are one cell long and have

VPI/VCI = 0/1

• Sets up 3 types of signaling channels:

- Point-to-point- General broadcast- Selective broadcast

• Procedures to:- Set up new signaling channels- Release channels- Verify channels

Signaling Messages

ATM Signaling is a protocol used to set up, maintain, and clear SVCs between two ATM end users over private or public UNIs.

The protocol is, in fact, an exchange of messages that takes place between the ATM end user (caller or receiver) and adjacent ATM switch.

The messages contain information that is used to build, maintain, or clear the connection.

The messages themselves are segmented into cells at the signaling AAL and then transported over a standard signaling channels, VPI=0, VCI=5.

Signaling Messages

There are four types of messages:

• CALL ESTABLISHMENT

• CALL STATUS

• CALL CLEARING

• POINT-TO-MULTIPOINT OPERATIONS

SETUP. Sent by calling, or source ATM end user, to network (defined here as nearest ATM switch connected to ATM end user over UNI) and from network (defined here as nearest ATM switch connected to destination ATM end user over UNI) to called, or destination ATM end user. Used to initiate connection setup. Contains information, such as destination ATM address, traffic descriptors, AAL Info, and QoS.

1.CALL ESTABLISHMENT PROCESS

NNIUNI

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

UNI

DATA

CALL PROCEEDING

CONNECT

CONNECT ACK

SETUP

CALL PROCEEDING

CONNECT

CONNECT ACK

SETUP

CALL PROCEEDING

Accept

CALL ESTABLISHMENT PROCESS

• CALL PROCEEDING. Sent by destination ATM end user to network and by network to source ATM end user to indicate that call establishment has been initiated.

• CONNECT. Sent by destination ATM end user to network and by network to source ATM end user to indicate that destination ATM end user accepts connection request.

• CONNECT ACKNOWLEDGE. Sent by network to destination ATM end user to indicate call is accepted. May also flow from source ATM end user to network maintain symmetrical call-control procedures.

• ALERTING. Sent by the destination ATM end user to the network and by the network to the source ATM end user to indicate that the destination ATM end user alerting has been initiated. For human interface (e.g., voice).

• PROGRESS. Sent by the ATM end user or the network to indicate the progress of a call in the event of inter-working.

• STATUS. Sent by the ATM end user or network in response to a STATUS ENQUIRY message.

• STATUS ENQUIRY. Sent by the ATM end user or network to solicit STATUS message.

• NOTIFY. Sent by the ATM end user or network to indicate information pertaining to a call/connection.

2. CALL STATUS

• RELEASE. Sent by an ATM end user to request the network to clear the end-to-end connection or is sent by the network to indicate that the VCC is cleared and that the receiving ATM end user should release the VC and prepare to release the call reference after sending a RELEASE COMPLETE.

3. CALL CLEARING

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNIUNI

RELEASE COMPLETE

RELEASERELEASE

RELEASE COMPLETERELEASE COMPLETE

• RELEASE COMPLETE. Sent by an ATM end user or network to indicate that virtual channel and call reference have been released and that the entity receiving the message should release the call reference.

• RESTART. Sent by the ATM end user or network to request the recipient to restart the indicated virtual channel or all virtual channels controlled by the signaling channel.

• RESTART ACKNOWLEDGE. Acknowledges restart message and indicates restart is complete.

3. CALL CLEARING

Point-to-multipoint SVCs enable a single ATM end user to communicate with one or more ATM end users. Information flowing from the source ATM end user is replicated by the network, not at the source and received by all destination ATM end users attached to the point-to-multipoint connection. The calling or source ATM end user is called the ROOT, and the called or destination ATM end users are called LEAVES. Conceptually viewed, leaves are connected to the root in a tree structure.

4. POINT-TO-MULTIPOINT OPERATIONS

NNIATM

End User

ATM_1End UserATM

SwitchATM

Switch

ATMSwitch

ATMSwitch

UNI UNI

ADD PARTY

ADD PARTY ACKCALL PROCEEDING

CONNECT

CONNECT ACK

SETUP

ATM_2End User

ADD PARTY ACK

The root establishes a connection to the first leaf using standard call-establishment messages as shown in Figure. After that, additional leaves can be added or removed to the point-to-multipoint tree by the root. The leaves have the option of accepting the invitation and unilaterally removing themselves.

Point-to-multipoint messages consist of the following:

• ADD PARTY. Adds party (leaf) to an existing connection.

• ADD PARTY ACKNOWLEDGE. Acknowledges a successful ADD PARTY.

• ADD PARTY REJECT. Indicates that ADD PARTY request was unsuccessful.

• DROP PARTY. Drops or removes party (leaf) from an existing point-to-multipoint connection.

• DROP PARTY ACKNOWLEDGE. Acknowledges a successful DROP PARTY

Figure shows the messages required to be sent by the root to add a leaf (ATM_2) to an existing point-to-multipoint connection.

POINT-TO-MULTIPOINT OPERATIONS

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

ADD PARTY NAK

ADD PARTYSETUP

CALL PROCEEDING

RELEASE

(REJECT)RELEASE COMPLETE

ADD PARTY

ADD PARTY NAK

(REJECT)

CASE 1 ADD THE NEXT PARTY. THE PARTY REJECTS.

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

DROP PARTY ACK

DROP PARTYRELEASE

RELEASE COMPLETE

DROP PARTY

DROP PARTY ACK

CASE 2 ROOT DROPS A PARTY.

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

RELEASE COMPLETE

RELEASERELEASE

RELEASE COMPLETE

RELEASE

RELEASE

CASE 3: ROOT DROPS LAST PARTY.

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

DROP PARTY ACK

DROP PARTY

RELEASE

RELEASE COMPLETEDROP PARTY ACK

DROP PARTY

CASE 4: A PARTY DROPS OUT.

COMPLETE

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

RELEASE COMPLETE

RELEASE

RELEASE

RELEASE COMPLETE

CASE 5 NETWORK CLEARS THE CALL.

CONNECTION

TERMINATED

DROP PARTY

DROP PARTY ACK

Leaf Initiated Join (LIJ)

UNI 3.1 only allowed the root the option of adding leaves to an existing point-to-multipoint connection.

This was deemed restrictive and would not provide the flexibility for applications to take full advantage of this capability.

Therefore, UNI Signaling 4.0 added a capability for leaves to join a point-to-multipoint connection without intervention from the root.

This is called leaf initiated join (LIJ).

LIJ is supported in one of the FOLLOWING MODES.

Leaf Initiated Join (LIJ)

MODE 1. ROOT SETS UP A NETWORK LIJ CALL

(ROOT PROMPTED JOIN) (ROOT LIJ CONNECTION)

MODE 2. LEAF PROMPTED JOIN WITHOUT ROOT

NOTIFICATION

MODE 3: LEAF JOIN TO AN INACTIVE LIJ CALL (I.e., No (I.e., No

MulticastMulticast Group Exists and a leaf wants to initiate Group Exists and a leaf wants to initiate a a

multicast group)multicast group)

MODE 4: LEAF JOIN TO A NON-LIJ CALL (A multicast group (A multicast group

exists but notexists but not created by a LIJ procedure. A Leafcreated by a LIJ procedure. A Leaf

wants to join that existing multicast group.)wants to join that existing multicast group.)

LIJ EXTENSION IN UNI 4.0

Two new messages were added to support LIJ

in Signaling 4.0:

LEAF SETUP REQUEST. Sent by leaf to initiate leaf-joining procedures.

LEAF SETUP FAILURE. Sent to leaf by root or network to indicate failure

to join the point-to-multipoint connection.

Also known as Root-prompted Join. In this model the leaf generates and sends a request over the UNI to join a point-to-multipoint connection. This request, in turn, is forwarded up to the root which then invokes established procedures for adding a leaf to an existing connection. This is also called a root LIJ connection.

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

MODE 1: ROOT SETS UP A NETWORK LIJ CALL.

(The set up message contains LIJ parameters.)

ATMEnd User

LIJ SET UP REQUESTLIJ SET UP REQUESTLIJ SET UP REQUEST

ADD PARTY ADD PARTY SET UP

CALL PROCEEDINGCALL PROCEEDINGCALL PROCEEDING

CONNECT ADD PARTY ACKADD PARTY ACK

CONNECT ACK CONNECT ACK CONNECT ACK

ATMEnd User

ATMEnd User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

CALL PROCEEDING

CONNECT

MODE 2: LEAF JOINS TO AN ACTIVE LIJ CALL.

LIJ SETUP REQ

SETUP

Also known as Leaf-prompted Join without root notification. In this model a leaf generates and sends a request over the UNI to join a point-to-multipoint connection. The network handles the request and the leaf joins the connection without notifying the root. This is called a NETWORK LIJ CONNECTION.

CONNECT ACK

NO ROOTNOTIFICATION

ATMEnd User

MODE 3: LEAF JOIN TO AN INACTIVE LIJ CALL

ATMEnd User

ATM_1End User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

Leaf SetupLeaf SetupLeaf Setup

SetupSetup

SetupCall Proceeding Call Proceeding

Call Proceeding

ConnectConnect

Connect

Connect AckConnect Ack

Connect Ack

No Multicast Group Exists and a LEAF wants to initiate a MULTICAST GROUP.

MODE 4. LEAF JOIN TO A NON-LIJ CALL

ATMEnd User

ATM_1End User

ATMSwitch

ATMSwitch

ATMSwitch

ATMSwitch

NNIUNI UNI

ATM_2End User

Leaf SetupLeaf SetupLeaf Setup

Add PartyAdd Party

Setup

Call Proceeding

Add Party AckAdd Party Ack

Connect

Connect Ack

Suppose a MULTICAST GROUP exists but created by a NON-LIJset-up. Now a LEAF wants to join that existing MULTICAST GROUP by LIJ Set Up Request.

SIGNALING PROTOCOL STACK (S-AAL)

SAAL

SAP

UNI 4.0UNI 3.1Q.2931

SSCSSSCF

SSCOP

CPCS

SAR

SSCS: Service SpecificCommon Part Sub-layer

SSCF: Service SpecificCoordination Function

SSCOP: Service SpecificConnection-OrientedProtocol

CPCS: Common PartConvergence Sub-layer

SAR: Segmentation &Reassembly

UNI Signaling Protocol Stack

•Signaling messages transported over ATM network using Signaling AAL (SAAL)

•Based on AAL 5•SSCOP provides reliable transport

Physical Layer

ATM Layer

AAL 5CPCS

SAR

Service-Specific Connection-Oriented Protocol (SSCOP)

Signaling Protocol

AAL service-specific part (SSCS)

Signaling AAL (SAAL)

Service-Specific CoordinationFunction (SSCF)

Signaling AAL (SAAL)

Provides a structured & reliable means to transport signaling traffic between two ATM end users. As part of the C_PLANE (Control Plane), it serves as the interface between higher layer control & signaling functions such as UNI 3.1/Q.2931 and the ATM layer.

SSCF: responsible for mapping the higher layer application to SSCOP.

SSCOP: a powerful connection-oriented data link protocol that provides a reliable transport for signaling messages. It supports end-to-end error detection, correction, frame sequencing & selective frame recovery.

SSCOP Operation

• Flow control based on sliding windowWindow size can be dynamically controlled by receiver (buffer size)

• Error control by selective retransmission of frames (only lost frame is retransmitted; AAL5 packets which consist of multiple ATM cells).

• Separate frames for control and data

•Data frames up to 64 Kbytes

SSCOP can be used as general Transport Layer over ATM

SSCOP Control Frames

{POLL, STAT, USTAT} used for reliable deliveryTCP uses a timer, whereas SSCOP uses explicit delivery•POLL(Next): Periodically used by SOURCE to

request status of receiver (DESTINATION)

Contains sequence number (SN) of next frame (NEXT) to be transmitted and timestamp --- if frames #1, #2, and #3 sent, POLL will send number for frame #4.

Receiver responds with sequence number of next sequential frame expected and list of any outstanding frames.

SSCOP Control Frames

• STAT(Next;Missing): Status ACKing next and list of missing data. Response to POLL frame generated by receiver SOURCE uses STAT frame to:

1. Retransmit lost frames

2. Release ACKed frames from the retransmission buffer

3. Advance transmission window to last sequential frame received by receiver

• USTAT (Unsolicited Status): Sent by receiver upon detecting a “hole” in the received sequence of frames Enables fast retransmission in the presence of random

loss Used to improve performance. Can have in-frequent

POLLs, yet not send too many frames following a lost frame.

SSCOP Operation: Example

0 1 2 3 4 5 6 7 8 9 7 10 11 12

Transmitter

Receiver

timeSTAT(5)AcknowledgesFrames 0-4

USTATReceiver detectsLoss of frame 7

STAT(10;7)Again requestsRetransmissionOf frame 7

Transmitter does notRetransmit frame 7Because POLL transmittedBefore first retransmission

Transmitter buffer size determined by frame rate and round trip delay between POLL and Receiver STAT

POLL POLL

X

ATM AddressingPEER MODEL: Use existing IP or MAC Addresses. IP routing protocols (OSPF) could be used.Advantages: Simplifies end system address administration.Disadvantages: Increases the complexity of ATM switches since they must act like multiprotocol routers and support address tables for all current protocols.

OVERLAY MODEL: Decouple ATM from existing Infrastructure and have unique addressing mechanism.Accordingly new routing protocols were needed +Address resolution protocols from IP to ATM or from LANsto ATM address conversions.Advantages: Decoupling of ATM from higher layers allows independent development, applications and ATM technology.

OVERLAY MODEL CHOSEN!!!

ATM Host Addressing

Each ATM end user must have a UNIQUE ADDRESS. There is a STANDARDIZED ADDRESSING STRUCTURE for both PUBLIC and PRIVATE ATM NETWORKS!!!

Private Networks: (based on HIERARCHICAL ADDRESSING DOMAINS)

- 20-byte format based on syntax of OSI Network Service Access Point (NSAP) address

-Two different formats:

1. DCC (Data Country Code) 2. ICD (International Code Designator)

ATM Host Addressing

Public Networks:

- 8-byte (64 bits) E.164 format defined by ITU-T

(has 16 digits each code with Binary Coded Decimal

(BCD) using 4 bits.)

- Can be extended to a 20-byte private address format by

appending end-system address (e.g., MAC address)

MakeCompatible

WithPrivate

Network address

ATM Network Address Formats

Authority&Format Identifier(39)“BD”

DataCountryCode

High-order DomainSpecific Part

End-SystemIdentifier

Selector

Private Network Supplied

ATM Forum specifies 3 NSAP (Network Service AccessPoints) – like Address Formats

EndSystem Supplied

Not usedIn Routing

1 byte 2 bytes 10 bytes 6 bytes 1 byteIDP DSP

Multiple addresses assigned to the same ATM adapter.

DCC ATM Format

IDI

Authority&Format Identifier(47)“C5”

Int’lCodeDesignator



Selector

1 byte 2 bytes 10 bytes 6 bytes 1 byte

Private

IDP DSP

ETHERNET & TokenRing address field lengthfor IEEE assigned addresses

ICD ATM Format

ICD codes identify particular international organizations.

IDI

Authority&Format Identifier(45)“C3”

E.164 Address



Selector

1 byte 8 bytes 4 bytes 6 bytes 1 byte

Public

IDP DSP (Domain Specific Part) IDI

NSAP-Encoded E.164 Format

High order domain specific part addresses can be assigned by “hand”.

Addressing

The IDP specifies an administration authority which has the responsibility for allocating and assigning values for the DSP.

IDP has AFI (Authority and Format Identifier (AFI)) and IDI (Initial Domain Identifier (IDI).

AFI specifies the format of the IDI, and the abstract syntax of the DSP field.

IDI specifies the network addressing domain, from which DSPs are allocated and the network addressing authority responsible for allocating values of the DSP from that domain.

DCC (Data Country Code) Specifies the country in which the address is

registered. These addresses are administered by the ISO’s national member body in each country. The digits of data country code are encoded using BCD.

Addressing

ICD (International Code Designator) Identifies an authority which administers

a coding scheme. This authority is responsible for the

allocation of identifiers within this coding scheme to organizations.

The registration authority for the international code designator is maintained by the British Standards Institute.

The digits of ICD are encoded using BCD.

ATM Forum extended E.164 address to NSAP format. E.164 number is filled with leading zeroes to make 15 digits.

A F16 is padded to make 8 bytes.

High Order DSP (HO-DSP) field will be used to construct multi-level address hierarchies for routing.

Remark: In real NSAPs, DSP is subdivided into a hierarchy of Routing Domain (RD) and an Area Identifier (AREA) and an End System Identifier (ESI).

ATM Forum combined the RD and AREA fields into HO-DSP.

A range of addressing hierarchies will be supported -> increases the scalability.

End System Identifier (ESI): 48-bit IEEE MAC address (to identify a specific host within an ATM subnet) (Token Ring, Ethernet LAN MAC addresses)

SELECTOR is for use inside the host and is not used for routing (used for local multiplexing within end stations and has no network significance).

This is used to distinguish between different destinations reachable at the end device.

All ATM addresses are 20 bytes long.

Addressing

•Private networks must support all three formatsType of Number field = UnknownNumbering Plan Indication field = ISO NSAP

•Public networks must support native E.164 and may optionally support three NSAP-encoded formats.

For E.164:Type of Number field = International NumberNumbering Plan Indication field = Recommendation E.164

•If only native E.164 addresses, subaddress field in signaling messages used to convey private ATM address across.

•One Transit network selection possible using carrier identification code field

E.164 Addresses (ITU-T)

•North American Numbering Plan (NANP): 1(614)-555-1212

•E.163 numbering plan for telephony: 12 digits

•E.164 numbering plan for ISDN: 15 digits•Defined in ITU-T recommendation E.164 for

ISDN•ISDN numbers uniquely identify interfaces to

public networks•Several ISDN numbers can identify the same

interface•ISDN signaling allows ISDN number followed

by a sub-address (extension) of up to 40 digits

•Administrated by public networks (Therefore, are not easily available for private network use)

EXAMPLE

ATM Address for SVC connecting BWN Lab Testbed (GCATT) to OIT-OC-3 (Rich Building)

39.840F.8001.BC88.2280.4110.4002.4000.OC80.0020.00

ATM Signaling Message Format

(Q.2931, UNI 3.0, UNI 3.1)• Each message includes the following components:8 7 6 5 4 3 2 1Bits Octets

Protocol Discriminator

Length of call reference value(in octets)

0 0 0 0

Call reference valueFlagCall reference value (continued)Call reference value (continued)

Message typeMessage type (continued)

Message lengthMessage length (continued)

Information ElementsInformation Elements

1

2

3

4

5

6

7

8

9

10

11...

• Protocol Discriminator (1 byte): Distinguishes Q.2931 messages from other messages.

• 08 = Q.931

• 09 = Q.2931

• Call Reference (4 bytes): Identifies call to which this message is related to. One user may have many calls simultaneously.

• Flag = 1 : Message is from call reference originator

• Flag = 0 : Message is to call reference originator

• Message Type (2 bytes): Many types, e.g., connect, call proceeding, setup, release, etc.

• Message Length (2 bytes): Length of contents of this message

Message Format

Sample Message Types

Bits 876 Bits 54321 Type

000 Call Establishment Messages

00010 Call proceeding

00111 Connect

01111 Connect ACK

00101 Setup

01101 Setup ACK

010 Call Clearing Messages

01101 Release

11010 Release Complete

011 Information

10101 Status Inquiry

11101 Status

111 Reserved for Extension

Protocol discriminator: Distinguishes messages for ATM end user-to-user network call control from other messages.

Call reference value: Associates message with connection at UNI. Local significance only.

Message type: Identifies type of message as described in previous section .

Message length.

TLV information elements: Parameters associated with a particular message.

The presence of the first four components is mandatory in every message. A message will contain different information elements (IE) depending on the type of message. The IEs are Type/Length/Value (TLV) fields that contain information that is used by the ATM end user or network to process the connection

Information Elements

Table describes most of the relevant IEs that have been defined in UNI 3.1 and UNI Signaling 4.0, and may be present (mandatory or optional) in ATM signaling messages:

Table. UNI Signaling Information Elements

Information element Max. length (bytes)

Description

Cause 34 Why certain messages are generated and may provide diagnostic information.

Call state 5 Current status of call.

Endpoint reference 7 Identifies individual endpoint in point-to-multipoint connection.

Endpoint state 5 Indicates state of an endpoint (i.e., add parity, drop parity, etc) in point-to-multipoint connection.

AAL parameters 20 AAL specific parameters such as CPCS-SDU size, AAL type

ATM traffic descriptors

30 Forward and backward PCR, SCR, and burst sizes


Description

Alternative ATM traffic descriptors

30 Describes alternate ATM traffic-descriptor values and is used during negotiation of these values and is used during negotiation of these values in UNI Signaling 4.0.

Minimum acceptable traffic descriptors

Describes minimum acceptable ATM traffic-descriptor values and is used during negotiation of these values in UNI signaling 4.0.

Connection identifier 9 VPI and VCI values.

QoS parameters 6 QoS class (0-4).

Extended QoS parameters

25 Indicates individual QoS values acceptable on a per-call basis. These include acceptable and cumulative forward and backward cell loss ratio.

Broadband high-layer information

13 Validate compatibility of high-layer information such as ISO or vendor-specific protocols

Broadband bearer capability

7 Indicates request for connection-oriented service that will provide interworking (I.e., DS1 emulation), ATM only, or VP service (for switched VPs). Also specifies CBR or VBR.

Table. UNI Signaling Information Elements (Cont’d)


Description

Broadband low-layer information

17 Validates compatibility of layer-2 and layer-3 protocols.

Broadband locking shift

5 Indicates new active code set.

Broadband nonlocking shift

5 Indicates temporary shift to specified code set.

Broadband sending complete

5 Indicates completion of the called party number.

Broadband repeat indicator

5 Indicates if IE is repeated in message and how they should be interpreted.

Calling party number

26 ATM address of source ATM end user.

Calling party subaddress

25 Used to convey a private ATM address across a public E.164 network.

Called party number 26 ATM address of destination ATM end user.

Called party subaddress

25 Used to convey a private ATM address across a public E.164 network.


Information element Max. length

( bytes)

Description

Transit network selection

8 Identifies requested transit network

Restart indicator 5 Identifies class of facility to be restarted, such as indicated VC or all VCs.

Narrowband low-layer compatibility

20 Q.2931-based IE used to validate low-layer compatibility for N-ISDN interworking device.

Narrowband high-layer compatibility

7 Q.2931-based IE used to validate high-layer compatibility for N-ISDN interworking device.

Notification indicator 5 Q.2931-based IE used to indicate information pertaining to a call.

Progress indicator 6 Q.2931-based IE used to describe an event which has occurred during the life of a call.

Narrowband bearer indicator

Q.2931-based IE used to indicate a requested circuit-mode N-ISDN bearer service to be provided by the network.

LIJ call indentifier 9 Identifies point-to-multipoint LIJ call at root’s interface.

LIJ parameters LIJ parameters used by root to associate options with call at call setup.



Description

LIJ sequence number 8 Used by joining leaf to associate SETUP, ADD PARTY, or LEAF SETUP FAILURE response message with corresponding LEAF SETUP REQUEST.

End-to-end transit delay

12 Q.2931-based IE used to indicate the maximum end-to-end transit delay acceptance on a per-call basis, and to indicate the cumulative transit delay actually experienced by a virtual channel connection. Equal to forward maximum cell transfer delay per traffic management V4.0 specification.

Extended end-to-end transit delay

12 Indicates backward maximum cell transfer delay.

Generic identifier transport

30 Used to indicate session and resource identifier for video-on-demand virtual connections.

Connection scope selection

6 Enables calling user to indicate to the network that the connection should proceed within the selected routing range. Used to limit search for anycast services.

OAM traffic descriptor

56 Provides information relating to the presence and handling of the end-to-end F5 OAM information flow for performance management and user-originated fault management associated with the user connection involved in the call.


Information element

Max. length (bytes)

Description

ABR setup parameters

36 Specifies set of ABR parameters used during connection setup.

ABR additional parameters

14 Specifies additional ABR parameters.


Bandwidth Contract

User specifies 12 leaky bucket parameters

Forward Backward

CLP=0 Peak Cell Rate

Sustainable Cell Rate

Maximum Burst Size

Peak Cell Rate


Maximum Burst Size

CLP=0+1 Peak Cell Rate


Maximum Burst Size

Peak Cell Rate


Maximum Burst Size

AAL Parameters

AAL 1 Parameters1. CBR Rate2. Clock Recovery type3. Error correction type4. Structured Data Transfer5. Partially filled cells

AAL ¾ Parameters1. MID Size

AAL Parameters

AAL 3/4/5 Parameters1. Forward maximum SDU size2. Backward maximum SDU size3. Mode: message or streaming4. Service Specific Convergence Sublayer

(SSCS) Type Null, Assured SSCOP, non-assured SSCOP, or Frame relay

Sample Information Elements

Bits 87654321

Information Element

0111000001110001011110000110110001101101010110000101100101011010010111000100001001011110

Called party numberCalled party sub-addressTransit network selectionCalling party numberCalling party sub-addressAAL parameterATM Traffic DescriptorConnection IdentifierQuality of Service ParameterEnd-to-end transit delayBroadband bearer capability

UNI 3.0 Overview

•Two types of VC’s

• Unidirectional channel to single destination or multicast tree (e.g., distribute video or audio streams)

* Joins to multicast VC must be initiated by root (Set up VC then use “Add Party” according to instructions from source)

• Two channels in opposite directions with symmetric or asymmetric bandwidths (asymmetric: files/ACK)

•Defines interaction between end-points and the network

•Support for specifying traffic burstiness (CBR, VBR)

•No support for QoS (can define service classes; differentiate types of services but no support for parameterized service classes)

•Flexibility in ATM address formats

UNI 3.0 Overview

Three classes of services1. Class A: CBR with stringent cell-loss,

delay, and jitter requirements.2. Class C: VBR with no end-to-end timing

requirements no delay guarantees3. Class X:

– User-defined traffic type and timing requirements (setup message specifies only desired bandwidth and QoS)

(Not universal, only work within your private network)

UNI 3.0 Overview

Traffic parameters supported– Peak Cell Rate (PCR)– Sustained Cell Rate (SCR)– Maximum Burst Size– Can be specified separately for CLP = 0 and 1– No support for traffic parameter negotiation(If over-request on network resources (bandwidth),

request will be rejected at signaling, e.g., ask for 100 Mbps but network can only support 50 Mbps, then the request will receive a “Reject”!!!)

QoS Support– Allows five distinct QoS classes (0-4) to be specified in

SETUP message– Parameterized QoS not supported

UNI 3.1 Overviewvery similar to 3.0

Goal: Revision of UNI 3.0 to make it compatible with Q.2931

Several incompatibilities with UNI 3.0Many message formats differentRefinement of Domain-Specific Part (DSP) to

E.164 address

UNI 3.1 still incompatible with Q.2931Three incompatible protocols {UNI

3.0, UNI 3.1, Q.2931}

UNI Signaling 4.0 Enhancements

UNI 4.0 Signaling includes a number of significant functional enhancements over AND above what is provided in UNI 3.0/3.1. Some have already been discussed, but it is helpful to review here the primary functional enhancements:

Leaf-initiated join. Allows users to dynamically join existing point-to-multipoint connections.

Group addressing: Well-known functional addresses can be utilized to reduce signaling and configuration overhead.

UNI Signaling 4.0

Enhancements Anycast Addressing. Enables server (and or servers)

to be assigned a well-known group address. The ability to control access to anycast servers is included using a connection/membership scope.

Proxy Signaling. Used to support devices that do not support ATM signaling, such as residential broadband.

Switched Virtual Paths. ATM virtual paths can be dynamically provisioned. This will reduce administrative overhead in public and private networks.

Multiple Signaling Channels. This enables multiple ATM end users to share a single UNI interface or port on the switch.

Frame Discard Capability. Cells belonging to an entire frame can be discarded to prevent or relieve congestion, e.g., “early packet discard” and “partial packet discard” enabled/disabled on a per VC basis.

Available Bit Rate (ABR) Signaling for Point-to-Point Connections. Parameters for ATM end users requesting ABR service can be designed into the network.

UNI Signaling 4.0

Enhancements

Signaling of Individual QoS Parameters. Cell loss ratio, mean cell transfer delay, maximum cell transfer delay, and cell delay variation can be signaled into the network rather than just one QoS class.

Traffic Parameters Negotiation. ATM traffic parameters can be negotiated between ATM end users.

N-ISDN Inter-working. Enables inter-working between narrowband and broadband ISDN networks.

UNI Signaling 4.0

Enhancements

Proxy SignalingNetworkPSA

(Proxy Signaling Agent)PSA

CONNECT

CONNECT (to User A)

SETUP (to User B, from A,VPCI at UNI A) SETUP (to User B, from A,

VPCI at UNI B)

Network has established ATM connectionbetween users A and B

Suppose you have a low cost adapter, a “proxy” can do the signaling for that piece of equipment.

• Allows a user (proxy agent) to perform signaling on behalf of other users.

• Used to support devices that do not support ATM signaling, such as residential broadband.

• Enables multiple ATM interfaces on end-system to share same ATM address, e.g., high performance server with 3 adapters.

• Proxy agent can be remotely located on another switch

A well-known group address associated with a PARTICULAR SERVICE.

( like dialing 555-1212 in the telephone service)

Examples: Name server, time-of-day server, etc..; Specify service only and protocol will find server that provides desired service. Do not have to know location of service; just specify the desired service.

Allows a user to request a point-to-point connection to any host within a specified group

Useful for connecting to a server providing a specific service (e.g., LAN Emulation Server, ARP Servers, NHRP Servers)

• ATM group addresses obtained by setting most significant bit of first byte (AFI) in the address formats.

• Well-known groups for standardized service functions

• Also allows specification of “scope” of service provided by each member (within LAN, within organization, etc.). (Route the request to the server valid for the service “context”).

Anycasting

Overview of UNI 4.0 FunctionsThis table shows the new functions provided in the UNI Signaling 4.0, and whether the functions are mandatory (M) or optional (O) on both ATM end user (terminal equipment) and switch (switching system).Feature Terminal Equipment Switching System

Point-to-point callsPoint-to-multipointLeaf-initiated join capabilityNotification of end-to-end connection completionATM anycast capabilityMultiple signaling channelsSwitched virtual path (VP) serviceProxy signalingFrame discard capabilityABR signaling for point-to-point callsGeneric identifier transportTraffic parameters negotiationSignaling of individual QoS parametersSupplementary servicesDirect dialing in (DDI)Multiple subscriber number (MSN)Calling line identification presentation (CLIP)Calling line identification restriction (CLIR)Connected line identification presentation (COLP)Connected line identification restriction (COLR)Subaddressing (SUB)User-user signaling (UUS)

MOOOOOOOOOOOOO-OOOOOOO

MMMMM

*OOO†OOOO-OOOOO‡O

* This feature is optional for public networks/switching systems and is mandatory for private networks/switching systems.† Transport of the frame discard indication is mandatory.‡ This feature is mandatory for network/switching systems (public and private) that support only native E.164 address formats.

Documents

CHAPTER 5: Signaling in ATM Networks