Every cell header contains a VPI field and a VCI field, which explicitly associate a cell with a given

virtual channel on a physical link. It is important to remember the following attributes of VPIs and

VCIs:

• VPIs and VCIs are not addresses, such as MAC addresses used in LAN switching.

• VPIs and VCIs are explicitly assigned at each segment of a connection and, as such, have only

local significance across a particular link. They are remapped, as appropriate, at each switching

point.

Using the VCI/VPI identifier, the ATM layer can multiplex (interleave), demultiplex, and switch

cells from multiple connections.

Point-to-Point and Point-to-Multipoint Connections

Point-to-point connections connect two ATM systems and can be unidirectional or bidirectional. By contrast, point-to-multipoint connections (see Figure 1-7) join a single source end system (known as the root node) to multiple destination end-systems (known as leaves). Such connections can be unidirectional only, in which only the root transmits to the leaves, or bidirectional, in which both root and leaves can transmit.

Note that there is no mechanism here analogous to the multicasting or broadcasting capability

common in many shared medium LAN technologies, such as Ethernet or Token Ring. In such

technologies, multicasting allows multiple end systems to both receive data from other multiple

systems, and to transmit data to these multiple systems. Such capabilities are easy to implement in

shared media technologies such as LANs, where all nodes on a single LAN segment must

necessarily process all packets sent on that segment. The obvious analog in ATMto a multicast LAN group would be a bidirectional multipoint-to-multipoint connection. Unfortunately, this obvious solution cannot be implemented when using AAL5, the most common ATM Adaptation Layer (AAL) used to transmit data across ATM networks.

AAL 5 does not have any provision within its cell format for the interleaving of cells from different

AAL5 packets on a single connection. This means that all AAL5 packets sent to a particular

destination across a particular connection must be received in sequence, with no interleaving

between the cells of different packets on the same connection, or the destination reassembly process would not be able to reconstruct the packets.

This is why ATM AAL 5 point-to-multipoint connections can only be unidirectional; if a leaf node

were to transmit an AAL 5 packet onto the connection, it would be received by both the root node

and all other leaf nodes. However, at these nodes, the packet sent by the leaf could well be interleaved with packets sent by the root, and possibly other leaf nodes; this would preclude the reassembly of any of the interleaved packets.

Solutions

For ATM to interoperate with LAN technology, it needs some form of multicast capability. Among

the methods that have been proposed or tried, two approaches are considered feasible (see

Figure 1-8).

• Multicast server. In this mechanism, all nodes wishing to transmit onto a multicast group set up

a point-to-point connection with an external device known as a multicast server. The multicast

server, in turn, is connected to all nodes wishing to receive the multicast packets through a

point-to-multipoint connection. The multicast server receives packets across the point-to-point

connections, serializes them (that is, ensures that one packet is fully transmitted prior to the next

being sent), and retransmits them across the point-to-multipoint connection. In this way, cell

interleaving is precluded.

• Overlaid point-to-multipoint connections. In this mechanism, all nodes in the multicast group

establish a point-to-multipoint connection with each other node in the group and, in turn, become

a leaf in the equivalent connections of all other nodes. Hence, all nodes can both transmit to and

receive from all other nodes. This solution requires each node to maintain a connection for each

transmitting member of the group, while the multicast server mechanism requires only two

connections. The overlaid connection model also requires a registration process for telling nodes

that join a group what the other nodes in the group are, so that it can form its own

point-to-multipoint connection. The other nodes also need to know about the new node so they

can add the new node to their own point-to-multipoint connections.

Of these two solutions, the multicast server mechanism is more scalable in terms of connection

resources, but has the problem of requiring a centralized resequencer, which is both a potential

bottleneck and a single point of failure.

The layers of the ATM reference model have the following functions:

• Physical layer—manages the medium-dependent transmission. The physical layer is divided into

two sublayers:

— Physical medium-dependent sublayer—synchronizes transmission and reception by sending

and receiving a continuous flow of bits with associated timing information, and specifies

format used by the physical medium.

— Transmission convergence (TC) sublayer—maintains ATM cell boundaries (cell

delineation), generates and checks the header error-control code (HEC), maintains

synchronization and inserts or suppresses idle ATM cells to provide a continuous flow of

cells (cell-rate decoupling), and packages ATMcells into frames acceptable to the particular

physical layer-implementation (transmission-frame adaptation).

• ATM layer—establishes connections and passes cells through the ATM network. The specific

tasks of the ATM layer include the following:

— Multiplexes and demultiplexes cells of different connections

— Translates VPI/VCI values at the switches and cross connections

— Extracts and inserts the header before or after the cell is delivered to the AAL

— Maintains flow control using the GFC bits of the header

• ATM adaptation layer (AAL)—isolates higher-layer protocols from the details of the ATM

processes by converting higher-layer information into ATM cells and vice versa. The AAL is

divided into two sublayers:

— Convergence sublayer (CS)—takes the common part convergence sublayer (CPCS) frame,

divides it into 53-byte cells, and sends these cells to the destination for reassembly.

— Segmentation and reassembly sublayer—segments data frames into ATM cells at the

transmitter and reassembles them into their original format at the receiver.

• Higher layers—accept user data, arrange it into packets, and hand it to the AAL.

ATM Addressing

If cells are switched through an ATMnetwork based on the VPI/VCI in the cell header, and not based

directly on an address, why are addresses needed at all? For permanent, statically configured virtual

connections there is in fact no need for addresses. But switched virtual connections, which are set

up through signaling, do require address information.

Switched virtual connections work much like a telephone call. When you place a telephone call you

must have the address (telephone number) of the called party. The calling party signals the called

party’s address and requests a connection. This is what happens with ATM switched virtual

connections; they are set up using signaling and therefore require address information.

The types and formats of ATM addresses, along with their uses, are described in the chapter “ATM Signaling and Addressing.”