Asynchronous Transfer Mode
Asynchronous Transfer Mode |
|
ATM connections between endpoints are not distinguished only by their various Quality of Service parameters and the formats of their addressing schemes. They also fall into one of two larger categories: point-to-point connections and point-to-multipoint connections. Which of these connection types any particular ATM connection uses depends on how ATM signaling builds its connection.
Signaling components exist at the endstation and at the ATM switch. The signaling layer of ATM software is responsible for creating, managing, and terminating switched virtual circuits (SVCs). The ATM standard wire protocol implemented by the signaling software is called the User Network Interface (UNI). The way one ATM switch signals another ATM switch comprises a second signaling standard, called the Network Network Interface (NNI).
Figure 14.10 ATM Signaling
When an ATM-aware process seeks to connect to another process elsewhere on the network, it asks the signaling software to establish an SVC. To do this, the signaling software sends an SVC creation request to the ATM switch using the ATM adapter and the reserved signaling VC. Each ATM switch forwards the request to another switch until the request reaches its destination. An ATM switch determines which switch to send the request to next based on the ATM address for the connection and the switch's internal network database (routing tables). Each switch also determines whether or not the request's service category and Quality of Service needs can be met. At any point in this process, a switch can refuse the request.
If all the switches along the path can support the virtual circuit as requested, the destination endstation receives a packet that contains the VC number. From that point on, the ATM-aware process can communicate with the destination process directly by sending packets to the VPI/VCI that identify the specified VC.
The ATM adapter shapes data traffic for each VC to match the contract made with the ATM network. If too much data is sent for any reason, the ATM switch can ignore — and lose — the data in favor of providing bandwidth to another contract or set of contracts. This is true for the entire breadth of the network; if bandwidth or speed exceeds the limits established by the contract, any device, including the ATM adapter, can simply drop the data. If this happens, the endstations concerned are not notified of the cell loss.
Unlike a standard LAN environment, ATM is a connection-oriented medium that has no inherent capabilities for broadcasting or multicasting packets. To provide this ability, the sending node can create a virtual circuit to all destinations and send a copy of the data on each virtual circuit. However, this is highly inefficient. A more efficient way to do this is through point-to-multipoint connections. Point-to-multipoint connects a single source endpoint, known as the root node, to multiple destination endpoints, known as leaves. Wherever the connection splits into two or more branches, the ATM switches copy cells to the multiple destinations.
Point-to-multipoint connections are unidirectional; the root can transmit to the leaves, but the leaves cannot transmit to the root or to each other on the same connection. Leaf-to-node and leaf-to-leaf transmission requires a separate connection. One reason for this limitation is the simplicity of AAL5 and the inability to interleave cells from multiple payloads on a single connection.
