Previous Table of Contents Next


Frame Relay Network

The Frame Relay network (Figure 11.2) consists of specified network interfaces in the form of the User Network Interface (UNI) and Network-to-Network Interface (NNI). The specifics of the internal carrier network are based on ISDN, making use of Permanent Virtual Circuits (PVCs). While Switched Virtual Circuits (SVCs) are not yet implemented; many of the major carriers intend to make them available by mid-1997. Among the first to implement SVCs will be Ameritech, MCI, and CompuServe [11-5]. Frame Relay networks can be public, private, or hybrid.


Figure 11.2  Frame Relay network.

User Network Interface (UNI)
is the demarcation point between the user DTE and the network, and is in the form of a FRAD (Frame Relay Access Device) and a FRND (Frame Relay Network Device).
Network-to-Network Interface (NNI)
is defined as the interface between frame relay networks and is based on multinetwork PVCs. While some early implementation work has begun in this regard, most network-to-network connections are provided over digital trunks or ATM, where it is available. This current method of internetwork connection has implications relative to Implicit Congestion Notification and Network Management.
ISDN
is the basis of the internal Frame Relay network although the user links generally are not ISDN, largely due to lack of availability. Additionally, the ISDN LAP-D protocol governs the user links, although they generally are not ISDN for reasons of availability and additional cost.
Permanent Virtual Circuits (PVCs)
define fixed paths through the network for each source-destination pair, based on programmed logic. Such definitions are fixed in network routing tables, with alternative routes defined and invoked in the event of a network failure. Regardless of network traffic, the PVC will always be used to serve a given source-destination pair. Frame Relay networks currently are based on PVCs.
Switched Virtual Circuits (SVCs)
are set up call by call, based on programmed network routing options. At the end of the conversation, the SVCs are torn down; the next SVC provided for the same source-destination pair could be quite different, depending on network availability. While the standards envision SVC networks, they are not yet available.
Mesh Networking
is easily accomplished with Frame Relay, and on a scalable basis. That is to say that the cost of Frame Relay implementation is relatively proportionate to the task at hand. As more sites are added to the network, more FRADs, access links and FRNDs are added; as the bandwidth requirement increases, the capacity of the link is increased. The costs grow relatively gracefully as the network is expanded to provide access to additional sites, and as bandwidth is increased on a site-by-site basis.
Network Processing
is not performed in Frame Relay, at least not to the extent we experienced in X.25 networks. A Frame Relay network assumes that the frame is error free. As the link is digital (ideally SONET) and based on ISDN, there is no compelling requirement to check for errors. Once the frame is identified as being valid and once the address is recognized, the frame is relayed to the next node and, ultimately, to the end user. Frame Relay also assumes that all frames are passed through the network successfully, and are received in the order transmitted. This is the equivalent of removing OSI Layer 3 functions from the X.25 model. Further, Frame Relay assumes no responsibility for protocol conversion.
The upshot of this avoidance of network processing is that the cost of the network is reduced; latency is reduced, as well. Rather than the carrier’s assuming such responsibilities, and the associated costs, they are shifted to the user domain. While this may appear to be burdensome, the capabilities of such DCE has increased tremendously over the past few years, while the cost of such capability has decreased.


Previous Table of Contents Next