• international
• Knowledge
  • Introduction
  • RFC´s
  • reserved IP
  • IP-Klassen
  • IP-Subnetze
  • OSI
  • Encoding
  • PCM
  • E1, E2, E3
  • X.25
  • G.703
  • G.704
  • Frame Relay
  • ATM
  • POS
  • FDDI
  • TokenRing
  • Fiber Technics
  • DTE and DCE
  • X.21 Kabel
  • X.21 Kabel 2
  • FME-Multiplexer
  • X.21 FAQs
  • G.703-G.704 FAQs
  • 2610 - X.21
  • dns-root
• Glossary
• Fragen?
• SSL/Schlüssel
• Reliability
• Suchen
• Inhaltsverzeichnis

The frame relay network provides similar types of connectivity as the more traditional X.25 public networks. A frame relay node performs almost no processing on the frames that pass through it. All flow control and error recovery are performed end-to-end by the user's terminals. For constantly high-bandwidth applications, leased lines will continue to provide a better solution, but esspecially for bursty applications (graphics on a LAN to LAN application), frame relay is often preferred.

There are two basic conditions that should exist for frame relay to be justifiable.

1. The transmission line has to be good. Frame relay will only function efficiently if the error rate of the underlying physical medium is low.

2. The nodes connecting to the frame relay are not dumb terminals, but will run their own protocols for flow control, error recovery, and acknowledgements.

ANSI standards T1.606 and T1.618 define the frame relay core procedures: These are the procedures used to handle users' data frames at a frame relay network node. ANSI standard T1.617 defines maintenance procedures for frame relay networks. These specify types of messages exchange between a user terminal and the node through which it connects to the network. Annex D of this standard defines procedures applicable to permanent virtual circuits (PVCs).

Before ANSI standard T1.617 Annex D, a consortium of companies defined a mechanism for frame relay PVC management, called LMI (Link Management Interface). The LMI defines a functionality similar to that defined (later) by ANSI standard and is currently a widely supported standard in existing frame relay networks.

There are three different classes of nodes within the frame relay network:

1. End user nodes

2. Network access nodes

3. Internal network nodes

End user nodes are the only terminals in the network that handle the application data. They communicate directly with network access nodes: these are the nodes through which the frame relay network is accessed by the user terminals.

Functionally, the network access nodes and internal network nodes are similar, performing switching on all data frames they receive. Yet they are not identical; two main differences exist. First, the network access nodes perform the network side of the maintenance procedures (ANSI T1.617 or LMI). These procedures do not involve internal network nodes. The second difference involves the checking of users' bandwidth consumption. This is usually done in the network access node (although is can be done in any node in the network).

The frame relay protocol is based on the core aspects of LAPD (Link Access Protocol D), the link layer protocol specified by CCITT. Frame relay uses the LAPD frame to enclose both user data and the address information used to route the frame. The address information is known as the Data Link Connection Identifier (DLCI). The frame relay header, which is a two-byte field within the frame, is specified for enclosing a DLCI representing the destination of the frame. A Frame Check Sequence (FCS) is used between the access device and the network to ensure bit integrity of the frame. Frames with errors are discarded. Frames are recovered by reinitiating transmission.