2.

ATM and B-ISDN Concepts

This chapter introduces to the reader the following concepts used in ATM and B-ISDN:

ATM cell Virtual Path and Virtual Channel Protocol Reference Model Reference Configuration Models for B-ISDN

Chapter Objectives
After completing this chapter you will be able to: explain concepts and definitions for ATM and B-ISDN
Figure 2.1

Chapter objectives.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM Essentials

2.1 Introduction to B-ISDN and ATM

The concept of Broadband ISDN grew out of the work with ISDN in the early 1980s. One could foresee a demand for the support of services requiring bandwidths greater than those for ISDN. In the middle of the 1980s, three high-speed channels with fixed bandwidths were outlined as a way of providing greater bandwidths in the access. However, this solution was not seen as ideal, since these fixed rates would of course be inefficient for services requiring other bandwidths. In 1988, the ITU-T recommendation I.121 Broadband Aspects of ISDN, giving the basic outline of B-ISDN, was completed. ATM was now chosen as the technology for B-ISDN, as a technology capable of supporting arbitrary, variable bandwidths, uniting the best of circuit switching with the best of packet switching. ATM had not yet been specified, so discussions about, among other things, the cell size now followed. There were many alternatives presented such as the U.S.s of 5+64 (header + payload) and the European proposal of 4+32. The discussion was lively because the European view on the subject was from a telecom side, which meant that the delays through the network should be as small as possible and therefore requiring a cell as small as possible. The U.S. and Japan were more interested in datacom applications, which meant that they were more worried about header (or overhead) to payload ratio than about varying delays and therefore wanted a longer cell than the Europeans. Finally, a compromise of 5+48 was accepted.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM and B-ISDN Concepts

Figure 2.2

ATM cell flow

The cell is the basic building block of the ATM technology. Figure 2.2 shows how information flows, containing arbitrary and variable bit-rates, are organised in a uniform fashion into cells, containing 53 octets. All cells are then inserted into a continuous cell stream common to all cells. The ITU-T definition of Asynchronous Transfer Mode in Recommendation I.113 is as follows: A transfer mode in which the information is organised into cells, it is asynchronous in the sense that the recurrence of cells containing information from an individual user is not necessarily periodic. The word "asynchronous" means that cells from one user will appear at different distances in the cell stream, that is, the cells from one user are not sent synchronously in the cell stream. However, when viewing all the cells that are transmitted, ATM is exceedingly synchronous. Cells are always sent out regardless of whether or not there is something to be sent. The cell stream is filled out by idle cells when there is nothing to send.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM Essentials

2.2 The ATM cell

Payload

Header

HEC

P T I

C L P

VCI

VPI

GFC/ VPI

Figure 2.3

The ATM cell

Each cell consists of a header of 5 octets and a payload (user information) consisting of 48 octets. The same cell format is used both at the user interface and throughout the network. ATM technology allows the transparent forwarding of information packets similar to the workings of a circuit-switched connection. The cell type depends on the information carried in the payload. In a normal cell the payload contains user information but the cell can also be used for Operation and Maintenance or signalling information. Idle cells are cells with empty payload. The header is primarily an address field, identifying the connection and supplying an address for routing the cell to its final address. The address number identifies the connection. Figure 2.3 shows the cell header with the following fields: GFC - Generic Flow Control, VPI - Virtual Path Identifier, VCI - Virtual Channel Identifier, CLP - Cell Loss Priority, PTI - Payload Type Identifier and HEC Header Error Control.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM and B-ISDN Concepts

The following gives a short description of each field:

GFC - Generic Flow Control can be used for access control, which
can exist at the interface to the public network.

VPI - Virtual Path Identifier and VCI - Virtual Channel Identifier

contain address information and are used for routing. CLP - Cell Loss Priority indicates the relative priority of the cell.

PTI - Payload Type Identifier indicates whether the cell contains user
data, signalling data or maintenance information. HEC - Header Error Control is used for checksum control of the header. Checksum control provides the possibility of correcting singlebit errors and detecting multi-bit errors. The only bit-error checking in an ATM network is in the control of the cell header. The payload field is transmitted transparently through the ATM network. It is the task of the connecting terminals to check/correct any bit errors in the payload field.

2.3 Virtual Path and Virtual Channel

Virtual Channels
116 118 125

Virtual Path VP 5

Virtual Channels
115 125

Virtual Path VP 4

Physical link

Virtual Channels
116 118 120

Virtual Path VP 8

137

Figure 2.4

The relationship between VC and VP

The ATM cell flows on a physical link are logically organised in Virtual Channels and Virtual Paths. The user needs a Virtual Channel for sending its user data. Virtual Channels are grouped into Virtual Paths. There may
Ericsson Telecom AB 1996, Stockholm, Sweden

ATM Essentials

be one or several Virtual Channels in one Virtual Path and one or several Virtual Paths on a Physical link. Figure 2.4 shows a link containing three Virtual Paths with different numbers. Each Virtual Path contains Virtual Channels with different numbers. Each user is connected to one of the Virtual Channels. The fields VCI and VPI in the cell header indicate to which Virtual Channel and Virtual Path a ATM cell belongs. Cells in a cell flow from the same user have the same VCI and VPI value. The VCI values are unique within a Virtual Path and the VPI values are unique within a physical link. In the figure Virtual Channels 116, 118 and 125 belong to Virtual Path 5, Virtual Channels 115 and 125 to Virtual Path 4 and Virtual Channels 116, 118, 120 and 137 to Virtual Path 8.

VCI=117 VPI=7

4 I=1 3 V C I=3 VP
port=2

t=3 por
port=12

ATM

User A

VC 145

VP 5 VP 5

User B

VP V P 1717

VCI=117 VPI=7

VC 136

Figure 2.5

VCI and VPI values in a connection

The VPI and VCI values have only local meaning and apply only to a link between the user and the switching node or between switching nodes. Figure 2.5 shows two switching nodes and two connections set up between users connected to the nodes. For each incoming port in the switching nodes, the incoming VCI value is translated to a VCI value in the outgoing port and the incoming VPI value is translated to a VPI value in the outgoing port specified for this connec Ericsson Telecom AB 1996, Stockholm, Sweden

VP 5

ATM

t=5 por

port=5

VP 3

VP 5

VCI=145 VPI=5

VP 7

VC 117

VC 125

ATM and B-ISDN Concepts

tion. The VCI and VPI values set in this way then follow the connection to the next switching node. In figure 2.5, user A sends its data on VC 145 in VP 5 to switching node A. User A is connected to the link with port number 12. After switching node A, the user data should be switched to VC 134 in VP 3 in the outgoing link with port number 5. Then the VCI and VPI values in the cell header are changed from VCI=145 and VPI=5 to 134 and 3 respectively and sent out on the link. In the same way the VCI and VPI values are changed in switching node B to new values (VCI=117 and VPI=7) before the cell is sent to the user. The translation tables in each switching node are called routing tables and will be described later in this course. It is not necessary to send the address of the B-subscriber in the cell headers, because the ATM network is connection-oriented. When both VCI and VPI values are translated in a switching node, it is called a VC-switching node. It is also possible to have VP-switching nodes where only the VPI values are changed and the VCI values are always left unchanged.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM Essentials

2.4 Protocol reference model

Management Plane

Plane Management

Control Plane

User Plane

Layer Management

Higher Layer Protocols

Higher Layer Protocols

ATM Adaptation Layer

ATM Layer

Physical Layer

Figure 2.6

Protocol reference model

ITU-T has standardized a protocol reference model for B-ISDN which is very similar to the OSI model. Like the OSI model it is a layered model. The ATM protocols must be simple and reside at a low level, otherwise the system cannot support the bit-rates associated with broadband communications. The primary layers in the model are the Physical Layer, the ATM Layer and the ATM Adaptation Layer (AAL). AAL is a service-dependent layer and adapts various high-level protocols to the service-independent ATM layer. In the ATM layer the cells are created and handled. The ATM layer is service-independent. The Physical Layer takes care of the interface towards the physical media on which the cells are sent. The layers are subdivided into different planes. The User Plane is the part of the model that handles the transmission of data between users and the Control Plane supports signalling. Subdivision into User and Control planes is used by protocols above and including parts of the AAL. In the layers below there is no difference between the planes. A Management Plane is responsible for control of the ATM nodes and the network. The management plane is divided into Layer Management and
Ericsson Telecom AB 1996, Stockholm, Sweden

ATM and B-ISDN Concepts

Plane Management. Layer Management contains Operation and Maintenance functions related to each layer. Plane Management contains the management functions related to the whole system. In this course we will only describe the function related to the primary layers, that is the Physical Layer, the ATM Layer and the ATM Adaptation Layer (AAL).

Figure 2.7

Segmentation and multiplexing of different broadband services

Figure 2.7 illustrates the primary tasks of ATM. The information flows for various services with different bit-rates, are segmented and packed into different ATM cells. This process is performed in, for example, a terminal adapter. The different ATM cells are then multiplexed into a common cell flow. The cells from a certain user will always occur in the correct order in the cell flow but the cells from one user are not sent synchronously in the cell flow. The cells from one user will appear in the cell flow at different distances depending on the type of traffic and the traffic load. During heavy traffic some cells have to be buffered at different nodes, which will also affect the distance between the cells from one user in the cell flow. At the receiving end the cells will be demultiplexed and unpacked in the user adapter. The payload from the cells is then sent to the different receivers.
Ericsson Telecom AB 1996, Stockholm, Sweden

ATM Essentials

User information

User information

48 octet

48 octet

48 octet AAL

A TM Cell A TM layer

ATM Cell

A TM Cell

Physical layer

Figure 2.8

The Physical, ATM and AAL layers

Figure 2.8 illustrates the primary tasks divided among the ATM Adaptation, ATM and Physical layers. The data information from different sources is segmented in the ATM Adaptation Layer (AAL). AAL will add some data and send packets of 48 octets to the ATM Layer. The ATM Layer adds a header to the data packet received from the ATM Adaptation Layer and the ATM cell is formed. The cell is forwarded to the Physical Layer. In the Physical Layer the ATM cell is adapted to the physical media. This means that the ATM cells are placed into different kinds of frames depending on the transport technologies used in the physical media. In figure 2.8 the cells are placed into an SDH frame.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM and B-ISDN Concepts

2.5 Reference configuration models for B-ISDN

Private N etwork

Public Net work

S B-TE1

B B-NT2 TB B-NT1 U B

B-TE2 R B

B-TA S B

Figure 2.9

Reference configuration model

Figure 2.9 shows the general reference configuration model for B-ISDN. It is built on the same principles as the model for narrowband ISDN. The model consists of functional groups and reference points. The functional groups are units with a certain standardized function. The reference points are interfaces (defined in the standard) between the functional groups. The following functional groups are defined:

Broadband Network Termination 1 (B-NT1) Broadband Network Termination 2 (B-NT2) Broadband Terminal Equipment (B-TE1 and B-TE2) Broadband Terminal Adapter (B-TA)

B-NT1 is used in the public network and includes functions like termination of line transmission, handling of the transmission interface and related OAM functions. It is the B-ISDN access to the public network. B-NT2 is used in the private network and includes functions like adaptation functions for different interface media, multiplexing/demultiplexing of traffic, buffering of ATM cells, signalling protocol handling and switching of ATM cells. LAN and PABX are examples of B-NT2.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM Essentials

B-TE is the terminal that terminates all protocols used by the users. B-TE1 is a terminal with standard interface but the B-TE2 needs to be connected to a Terminal Adapter (B-TA) which adapts the terminal to the standard interface. The different functional groups can be implemented in different equipment or in the same equipment, for example B-NT1 and B-NT2 can be implemented in the same equipment. There are four defined types of reference point:

R is the reference point between a terminal and terminal adapter SB is the reference point between a terminal with standard interface and
the network TB is the reference point between B-NT1 and B-NT2

UB is the reference point between different nodes in the network

B-ICI ATM

ATM

NNI NNI

ATM ATM NNI

NT1

UNI public

Private N etwork

NT 2 UNI private B-TE1

Figure 2.10

UNI, NNI and B-ICI interfaces in public networks

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM and B-ISDN Concepts

In the ATM standards three interfaces are specified: User Network Interface (UNI), Network Node Interface (NNI) and Broadband Inter-Carrier Interface (B-ICI). Figure 2.10 shows where in an ATM network these interfaces are found. UNI is the interface between a user and the network. Figure 2.10 shows a private UNI and a public UNI. The private UNI appears between B-TE and B-NT2, which is equivalent to the reference point S B. The public UNI is the user interface towards the public network and appears between NT1 and NT2. It is equivalent to the reference point TB. NNI is the interface between different nodes in the network. In some specifications the interface between nodes is also called Intra-Network Interface (INI). B-ICI is the interface between nodes in different networks from different operators.

2.6 Summary
The basic building block of ATM technology is the ATM cell. The size of the cell is 53 octets, 5 octets of header and 48 octets of payload. The cell header consists of 6 fields and is primarily an address field, identifying the connection and supplying a link address for the routing of the cell to its final address. The header also contains fields to indicate the type of payload, cell loss priority, checksum on the header and a generic flow control for users. The user information is stored in the payload. The ATM cell flows on a physical link are logically organised in Virtual Channels and Virtual Paths. The user needs a Virtual Channel for sending its user data. Virtual Channels are grouped into Virtual Paths. There may be one or several Virtual Channels in one Virtual Path and one or several Virtual Paths on a physical link. The fields VCI and VPI in the cell header indicate to which Virtual Channel and Virtual Path an ATM cell belongs. Cells in a cell flow from the same user have the same VCI and VPI value. ITU-T has standardized a protocol reference model for B-ISDN. The primary layers in the model are the Physical Layer, the ATM Layer and the ATM Adaptation Layer (AAL). AAL is a service-dependent layer and adapts various high-level protocols to the service-independent ATM layer. In the ATM layer the cells are created and handled. The ATM layer is service-independent. The Physical Layer takes care of the interface towards the physical media on which the cells are transmitted. The general reference configuration model for B-ISDN is built on the principles of narrowband ISDN. The model consists of the functional groups B-NT1, B-NT2, B-TE1, B-TE2 and B-TA, and the reference points R, SB, TB and UB.

Ericsson Telecom AB 1996, Stockholm, Sweden

ATM Essentials

In an ATM network, three interfaces are specified in the ATM standard: User Network Interface (UNI), Network Node Interface (NNI) and Broadband Inter-Carrier Interface (B-ICI).

Ericsson Telecom AB 1996, Stockholm, Sweden