The fiber distributed data interface FDDI is a high-speed backbone network using optical fiber as the transmission medium. It can be used to interconnect a single computer and a local area network.
The FDDI standard adopts the IEEE802 architecture and LLC protocol, studies FDDI's own MAC protocol, and proposes a physical layer medium (Physical Layer Medium Dependent, PMD) sublayer and physical layer protocol (Physical Layer Protocol, PHY) at the physical layer. Sub-layer. In 1992, the research on the interface standard of FDDI and SONET interconnection was completed.
FDDI is specially designed for data transmission. In order to transmit voice, image and video services, the FDDI-II standard will extend the FDDI basic mode (Basic Mode) that supports packet switching to the hybrid mode (Hybrid Mode). The mixed mode can support packet switching and circuit switching at the same time. Currently, the next-generation FDDI standard under study is called FFOL (FDDI Follow-On LAN).
In essence, FDDI consists of four sub-parts, each with its own specific function. The various parts together enable FDDI to provide high-speed connections between upper-layer protocols (such as TCP/IP, IPX) and media (such as optical cables).
The four sub-standards of FDDI are media access control (MAC), physical layer protocol layer (PHY), physical medium related layer (PMD) and station management (SMT). MAC specifies how to access the medium, including frame format, addressing, token processing, cyclic redundancy check algorithm (CRC) and error recovery mechanism required by the protocol; PHY specifies transmission encoding and decoding procedures, clock requirements and other functions ; PMD specifies the characteristics that the transmission medium should have, including fiber-optic link, power level, bit-error rate, optical component, and connector ( Connector); SMT specifies features such as FDDI station configuration, ring configuration and ring control, including station insertion and deletion, startup, fault separation and recovery, mode arrangement and statistical collection.
The working principle of FDDI is mainly embodied in the three main processes of FDDI. The three main processes are the establishment of site physical connection, ring initialization and data transmission.
(1) Establishment of site physical connection
When the FDDI network is in normal operation, the station management SMT has been monitoring the operating conditions of the ring and managing the activities of all stations. The connection management module in the station management SMT is responsible for establishing an end-to-end physical connection on the bidirectional optical cable between each pair of PHY/PMD at the station.
The station exchanges information such as port types and connection rules with neighboring stations by sending and receiving certain line state sequences to test the quality of the physical connection. In the test process, once a fault is detected, the method of tracking diagnosis is used to determine the cause of the fault, isolate the fault, and reconfigure the network.
(2) Ring initialization
After the physical connection of the site is completed, the ring is then initialized. Before the initialization of the ring, the target token cycle time TTRT of the system must be determined first. Each station can use the announcement frame to propose their own TTRT value. The system determines the TTRT value according to the established competition rules, and completes the initialization of the ring through the station with the selected TTRT value. The process of determining the TTRT value is usually called the declaration process, and the declaration process is used to determine the TTRT value.
(3) Data transmission
FDDI data transmission includes three processes: data sending, receiving and deleting.
In the FDD1 ring network, the station that wants to send data must wait for the token to arrive at the station and capture the token before sending one or more data frames until all data transmission is completed or until THT times out. Finally, the site releases a new token.
Each station on the FDDI ring is monitoring the frames passing by the station at any time. The station compares the destination address in the frame with the station address to determine whether to receive the frame. If the destination address of the frame matches the station address, the station When receiving the frame, the "A" flag of the FS field of the frame is set to "1", indicating that the destination station exists; while receiving the frame, the station also performs error checking on the frame. If no error is found, the station will copy the data field in the frame and place “1” in the “C” flag position of the FS field of the frame, indicating that the frame has been received by the destination station; if a CRC error is found, the frame The "E" mark position of the FS field is "1".
In an FDDI ring network, when a station receives a frame, it also forwards the frame to the next station. The sending station will continue to monitor the frames passing by the station after sending the data. When the sending station detects that the source address of a data frame is the same as the address of this station, it immediately stops forwarding the frame and makes it an invalid frame, so that part of the frame information that has been forwarded to the ring will be lost when it reaches the next station. They are discarded as frame fragments to prevent these frame fragments from continuing to be transmitted on the ring. Then, the sending station is responsible for deleting the remaining part of the frame from the ring and detecting the FS field in the frame at the same time. If the "A" and "C" flag bits in the FS field are both "1", it means that the frame has been successfully received by the destination station; if only the "A" flag bit is "1", it means the destination station exists but not received This frame indicates that an error occurred during the transmission of the frame; if the "A" and "C" flag bits are both "0", it indicates that the destination address in the frame is wrong.
Under normal circumstances, only the primary ring works, and the backup ring acts as a backup. Once the network fails, whether it is a line failure or a site failure, the FDDI network will automatically rebuild the double loop into a single loop through winding to ensure that the network will not be interrupted. This is an important feature that distinguishes FDDI from other LANs.
The FDDI network defines two types of sites: one is a single connection site (SAS) or type B site; the other is a dual connection site (DAS) or type A site. The single connection site is connected to the main ring of the FDDI double ring through a concentrator. The concentrator can connect to multiple SAS sites at the same time, and it is guaranteed that any error or power failure of the SAS site will not affect the FDDI network. The dual connection site can be connected to two loops at the same time.
FDDI can be mainly used in the following environments.
(1) Computer room network, also known as back-end network, is used for the connection between large computers and high-speed peripherals in the computer room, as well as environments that require high reliability, transmission rate, and system fault tolerance.
(2) The backbone network of the office or building group, also known as the front-end network, is used to connect a large number of minicomputers, workstations, personal computers and various peripherals.
(3) The backbone network of campus network or enterprise network, used to connect minicomputers, servers, workstations, personal computers and local area networks distributed in various buildings in the campus or enterprise.
(4) Multiple campus networks or corporate backbone networks are used to connect multiple campus networks and corporate networks geographically separated by several kilometers to become a regional backbone network interconnecting multiple campus networks and corporate networks.
The main performance indicators of FDDI are as follows:
(1) Use 802.2LLC protocol, compatible with IEEE802LAN.
(2) Use the token passing MAC protocol based on the IEEE 802.5 token ring standard.
(3) Use dual-ring topology to ensure that the network has fault tolerance.
(4) Multi-mode fiber, single-mode fiber or twisted pair can be used as the transmission medium.
(5) The transmission rate is 100Mbps, and the transmission rate of optical signal symbols is 125Mbps.
(6) The number of connections is not more than 1000. If each node is dual-connected, it is not more than 500.
(7) The maximum station distance is 2km, and the maximum loop length is 100km.
(8) It has the ability to dynamically allocate bandwidth, and can support synchronous and asynchronous data services at the same time.
(9) The maximum packet length is 4500 bytes.