|
|
|
Fibre/Copper Distributed Data Interface (FDDI/CDDI)
| |
The Fibre Distributed Data Interface (FDDI) standard was produced by the ANSI X3T9.5 standards committee in the mid-1980s. During this period, high-speed workstations were beginning to tax the capabilities of existing LANs (primarily Ethernet and Token Ring). A new LAN was needed that could easily support these workstations and their new distributed applications. At the same time, network reliability was becoming an increasingly important issue as system managers began to migrate mission-critical applications from large computers to networks, and FDDI was developed to fill these needs.
Today, although FDDI implementations are not as common as Ethernet or Token Ring, FDDI has gained a substantial following that continues to increase as the cost of FDDI interfaces diminishes. FDDI is frequently used as a backbone technology as well as a means to connect high-speed computers in a local area. In fact, FDDI is used at Bradford University as the backbone of the Computer Centre LAN.
Technology Basics
FDDI [ROS87] [LIN96] specifies a 100 Mbps, token-passing, dual-ring LAN using a fibre-optic transmission medium. It defines the physical layer and medium access-control sublayer of the data link layer, and so is roughly analogous to IEEE 802.3 and IEEE 802.5 in its relationship to the OSI reference model (ISO 7498). The international standard for FDDI networks is ISO 9314.
Although it operates at faster speeds, FDDI is similar in many ways to Token Ring. The two networks share many features, including topology (ring), media-access technique (token passing), reliability features (redundant rings) and others. However, there are many significant differences between the two LANs.
One of the most important characteristics of FDDI is its use of optical fibre as a transmission medium. Fibre offers several advantages over traditional copper wiring, including security (fibre does not emit electrical signals that can be tapped), reliability (fibre is immune to electrical interference) and speed (optical fibre has much higher throughput potential than copper cable). FDDI defines use of two types of fibre: single mode and multimode. Modes can be thought of as bundles of light rays entering the fibre at a particular angle. Single-mode fibre allows only one mode of light to propagate through the fibre, while multimode fibre allows multiple modes of light to propagate through the fibre.
Single-mode fibre is capable of higher bandwidth and greater cable run distances than multimode fibre. This is because multiple modes of light propagating through the fibre may travel different distances (depending on the entry angles), causing them to arrive at the destination at different times (a phenomenon called modal dispersion.) Multimode fibre uses light-emitting diodes
(LEDs) as the light-generating devices, while single-mode fibre generally
uses lasers.
FDDI Specifications
FDDI is defined by four separate specifications:
FDDI specifies the use of dual rings. Traffic on these rings travels in opposite directions. Physically, the rings consist of two or more point-to-point connections between adjacent stations. One of the two FDDI rings is called the primary ring; the other is called the secondary ring. The primary ring is used for data transmission, while the secondary ring is generally used as a backup.
Class B or single-attachment stations (SAS) attach to one ring; Class A or dual-attachment stations (DAS) attach to both rings. SASs are attached to the primary ring through a concentrator, which provides connections for multiple SASs. The concentrator ensures that failure or power down of any given SAS does not interrupt the ring. This is particularly useful when PCs, or similar devices that frequently power on and off, connect to the ring.
Each dual-attachment station (DAS) on FDDI has two ports, designated A and B. These ports connect the station to the dual FDDI ring. Therefore, each port provides a connection for both the primary and the secondary ring.
Traffic Types
FDDI supports real-time allocation of network bandwidth, making it ideal for a variety of different application types. FDDI provides this support by defining two types of traffic: synchronous and asynchronous. Synchronous traffic can consume a portion of the 100-Mbps total bandwidth of an FDDI network, while asynchronous traffic can consume the rest. Synchronous bandwidth is allocated to those stations requiring continuous transmission capability. Such capability is useful for transmitting voice and video information, for example. Other stations use the remaining bandwidth asynchronously. The FDDI SMT specification defines a distributed bidding scheme to allocate FDDI bandwidth.
Asynchronous bandwidth is allocated using an eight-level priority scheme. Each station is assigned an asynchronous priority level. FDDI also permits extended dialogues, where stations may temporarily use all asynchronous bandwidth. The FDDI priority mechanism can essentially lock out stations that cannot use synchronous bandwidth and have too low an asynchronous priority.
Fault-Tolerant Features
FDDI provides a number of fault-tolerant features. The primary fault-tolerant feature is the dual ring. If a station on the dual ring fails or is powered down or if the cable is damaged, the dual ring is automatically "wrapped" (doubled back onto itself) into a single ring.
As an example, consider a network with 4 stations called Stations 1, 2, 3 and 4. If Station 3 fails, the dual ring is automatically wrapped in Stations 2 and 4, forming a single ring. Although Station 3 is no longer on the ring, network operation continues for the remaining stations. FDDI can also compensate for a wiring failure. If the wiring between Stations 3 and 4 fails, the stations wrap the ring within themselves.
As FDDI networks grow, the possibility of multiple ring failures grows. When two ring failures occur, the ring will be wrapped in both cases, effectively segmenting the ring into two separate rings that cannot communicate with each other. Subsequent failures cause additional ring segmentation. In this situation, optical bypass switches can be used to prevent ring segmentation by eliminating failed stations from the ring.
Critical devices such as routers or mainframe hosts can use another fault-tolerant technique called dual homing to provide additional redundancy and help guarantee operation. In dual-homing situations, the critical device is attached to two concentrators. One pair of concentrator links is declared the active link; the other pair is declared passive. The passive link stays in backup mode until the primary link (or the concentrator to which it is attached) is determined to have failed. When this occurs, the passive link is automatically activated.
Frame Format
The format of FDDI frames is similar to that of Token Ring frames.
The fields of an FDDI Information frame (I-frame) are as follows:
An FDDI Token contains the Preamble (PA), Start delimiter (SD), Frame control (FC) and Ending delimiter (ED) fields as above. A summary of the FDDI frame formats is shown in the following table. Control symbols are used instead of control bits as in token ring. The number of control symbols (as opposed to data symbols) in each field is given in parentheses (this is not the field length). Control symbols are 5-bits long.
| Token Format | |||
| Preamble (16 or more) | Starting Delimiter (2) | Frame Control (2) | Ending Delimiter (1/2) |
| I-Frame Format | ||||||||
Preamble (16 or more) | Starting Delimiter (2) | Frame Control (2) | Destination Address (4/12) | Source Address (4/12) | Data (0 or more) | Frame Check Sequence(8)
| Ending Delimiter (1/2) | Frame Status (3) |
Copper Data Distributed Interface (CDDI)
The high cost of fibre-optic cable has been a major impediment to the widespread deployment of FDDI to desktop computers. At the same time, shielded twisted-pair (STP) and unshielded twisted-pair (UTP) copper wire is relatively inexpensive and has been widely deployed. The implementation of FDDI over copper wire is known as Copper Distributed Data Interface (CDDI).
CDDI uses the same data link control mechanisms as FDDI, but with the Physical sublayer being replaced with one specific to copper wire. The Twisted Pair-Physical Medium Dependent (TP-PMD) working group was established to develop a specification for implementing FDDI protocols over UTP/STP wire. The standard was approved by ANSI in 1994, and approval in Europe is to follow. The figure below summarises the structure of the FDDI/CDDI standard showing the different physical-medium dependent (PMD) layers that are now available.
| FDDI Medium Access Control (MAC) sublayer | FDDI Station Management (SMT) | ||
| FDDI Physical (PHY) layer | |||
| Twisted-Pair Wire PMD | Single-Mode Fibre PMD | Multimode Fibre PMD | |
While it offers significant cost savings, Unshielded Twisted Pair (UTP) wire also poses two significant technical design challenges. First, its information-carrying capacity is low relative to fibre, which means information must be compressed into a narrower frequency range in order to carry the high data rate. Second, UTP radiates energy when carrying signals, and those radiation levels increase directly as the transmission frequency increases. The acceptable level of such radiation specified by the American FCC and the CE regulations in Europe is quite low, so moving data over UTP is essentially a balancing act - an attempt to maximize its data throughput capacity while minimizing signal radiation. These problems have been solved by inventing solutions that combine scrambling, encoding, and equalisation algorithms as follows.
| LANs Home | Introduction | Token Ring | Ethernet | Future LANs | References |
Adrian Catchpole - A.G.Catchpole@comp.brad.ac.uk