Data Communications Systems : Fiber Distributed Data Interface (FDDI)

Fiber distributed data interface (FDDI) is a computer network protocol that uses fiber optic cable as the transmission medium to provide high-speed data transmission service to LANs. FDDI is a token protocol. The basic transmission rate of FDDI is 100 Mbps. FDDI is commonly used as a backbone network that interconnects several LANs within a company.

The FDDI specification is IEEE 802.2 and FDDI data transmission speed range from 100 to 200 Mbps. 1000 Mbps and higher FDDI speeds are in development.

FDDI is a LAN architecture that is based on redundant fiber rings that transmit in opposite directions. One of the rings is the primary ring and the other ring is the secondary ring. When the primary ring ceases to be operational (such as a cut cable) the network reconfigures itself (called “self-healing”) and it reconfigures the secondary ring as the primary ring.

Both single mode fiber and multimode fiber cable systems can be used with FDDI. Multimode fibers have a wider optical bandwidth transmission capability. However, this introduces distortion and limits the maximum distance for multimode fiber systems to about 2 kilometers. Single mode fiber systems have maximum range of approximately 60 km.

FDDI is a token passing architecture differing from token ring in that while a station has a token it can transmit as many frames as possible before the token expires. Because of this, there can be multiple frames on the ring at any time.

The interconnection devices in a FDDI network include a dual attached concentrator (DAC) and dual attached station (DAS). These devices remove and insert data to the FDDI ring. Each of these devices has dual transmission capability. If the fiber ring is cut, they can automatically redirect data onto its other channel (the secondary ring).

The DAC is a concentrator the converts the optical data on the FDDI system into another format that can be used to connect to other data networks. This allows one FDDI network node to connect to many other data communication devices.

Figure 1 shows FDDI system that uses dual rings that transmit data in opposite directions. This diagram shows one dual attached station (DAS) and a dual attached concentrator (DAC). The DAS receives and forwards the token to the mainframe computer. The DAC receives and token and coordinates its distribution to multiple data devices that are connected to it.


Figure 1: Fiber Distributed Data Interface (FDDI)

Data Communications Systems: Token Ring

Token Ring
Token ring is a LAN system developed by IBM that passes a token to each computer connected to the network. Holding of the token permits the computer to transmit data. The token ring specification is IEEE 802.5 and token ring data transmission speed range from 4 Mbps or 16 Mbps. 100 Mbps and higher token ring speeds are in development.

Token ring networks are non-contention based systems, as each computer connected via the token ring network must have received and hold a token before it can transmit. This ensures computers will not transmit data at the same time. Token ring systems provide an efficient control system when many computers are interconnected with each other. This is the reason token ring systems will not see data traffic degradation when many new users are added compared to Ethernet systems. However, passing tokens does add overhead (additional control messages) that reduces the overall data transmission bandwidth of the system.

The token ring LAN architecture was invented by IBM and touted to be the standard for clients of IBM mainframes who sought to replace aging 3270 terminals with LAN’s. IBM also developed cabling standards along with hub-like devices called multi-station access units (MAU’s). The original MAU’s formed a star network with the client PC’s and simulated the ring internally. The PC’s were connected to the MAU via IBM category type 1, 2, or 3 cable.

Figure 1 shows a typical token ring LAN. This diagram shows that the network is logically setup in a ring and each computer in the token ring network must receive a token before it can transmit. Since the token is relatively small compared to the packets of data that are sent, the token can rapidly move from computer to computer. When a computer receives a token, it can transmit data for a limited amount of time before it is required to forward the token.


Figure 1: Token Ring

Data Communications Systems: Ethernet

Ethernet
Ethernet is a packet-switching transmission protocol that is primarily used in LANs. Ethernet is often characterized by its data transmission rate and type of transmission medium (e.g., twisted pair is T and fiber is F). Ethernet systems in 1972 operated at 1 Mbps. In 1992, Ethernet progressed to 10 Mbps data transfer speed (called 10BaseT). In 2001, Ethernet data transfer rates included 100 Mbps (100BaseT) and 1 Gbps (1000Base T). In the year 2000, 10 Gigabit fiber Ethernet prototypes had been demonstrated.

Ethernet can be provided on twisted pair, coaxial cable, wireless, or fiber cable. In 2001, the common wired connections for Ethernet was 10 Mbps or 100 Mbps. 100 Mbps Ethernet (100BaseT) systems are also called “Fast Ethernet.” Ethernet systems that can transmit at 1 Gbps (1 Gbps = 1 thousand Mbps) or more, are called “Gigabit Ethernet (GE).” Wireless Ethernet have data transmission rates that are usually limited from 2 Mbps to 11 Mbps.

Wired Ethernet conforms to IEEE 802.3 standards and wireless Ethernet conforms to 802.11. IEEE 802.3 standard and uses carrier sense multiple access with collision detection (CSMA/CD) media access control (MAC).

Ethernet is the older than token ring and is based on linear bus technology. Originally installed using RG-6/8 coaxial cable (called “thicknet”), it was used for high-speed bus applications to interconnect mainframes and mini-computers. With the growth of personal computer (PC) workstations in the 80’s and early 90’s, a new wiring strategy was implemented using thinner RG-58 coaxial cable (called “thinnet”). In the mid-90’s newer twisted pair standards were set and higher speeds were achieved. 10 Mbps (10BaseT) became achievable on Category 3 unshielded twisted pair (UTP) wire.

Because Ethernet systems can use different cabling systems (e.g., twisted pair and coax), network interface cards (NICs) must contain a connector that is compatible with the cabling systems. Some NIC cards come with multiple connectors. The different types of connectors include:

DB-15 AUI connector for thicknet, 10Base5

BNC coaxial connector for thinnet, 10Base2

RJ-45 for twisted pair, 10BaseT or 100BaseT.

The maximum distance between devices in an Ethernet network is determined by the type of cable selected and performance of the NIC. Figure 1 shows different types of Ethernet LAN systems and the approximate distances devices can be connected together in these networks. Thicknet Ethernet uses a low loss coaxial cable to provide up to 500 meters of interconnection without the need for repeaters. Thinnet systems use a relatively thin coaxial cable systems and the typical signal loss in this cable restricts the maximum distance to approximately 185 meters. 100 BaseT systems use category 5 UTP cable and the maximum distance is approximately 100 meters.


Figure 1: Ethernet

Technologies: Routers, Gateways, Firewall

Routers
A router is a device that directs (routes) data from one path to another in a network. Routers base their switching information on one or more information parameters of the data messages. These parameters may include availability of a transmission path or communications channel, destination address contained within a packet, maximum allowable amount of transmission delay a packet can accept, along with other key parameters. Routers that connect data paths between different types of networks are sometimes called gateways.

Routers provide some of the same functionality as network switches. Their primary function is to provide a path for each routable packet to its destination. When a router is initially installed into a network, it begins its life by requesting a data network address. Using this data network address, it sends messages to nearby routers and begins to store address connections of routers that are located around it. Routers regularly exchange their connection information (lists of devices it is connected to) with nearby routers to help them keep the latest packet routing information.

A router can make decisions on where to forward packets dependent on a variety of factors including the maximum distance or packet priority. Distance vector routing and link state routing allow the router to select paths that match the needs of the data that is being sent through it.

Routers may also have fixed routing tables that are manually programmed by the network administrator. These static routing tables may be inflexible, however the use of static routing ensures other router’s that may have corrupt routing tables does not change the table.

Figure 1 shows a how a router can dynamically forward packets toward their destination. This diagram shows that a router contains a routing table (database) that dynamically changes. This diagram shows a router with address 100 is connected to two other routers with addresses 800 and 900. Each of these routers periodically exchanges information allowing them to build routing tables that allow them to forward packets they receive. This diagram shows that when router 100 receives a packet for a device number 952, it will forward the packet to router 900. Router 900 will then receive that packet and forward it on to another router that will help that packet reach its destination.


Figure 9.11: Router

Gateways
Gateways are devices that enable information to be exchanged between two dissimilar computer systems or data networks. A gateway reformats data and protocols in such a way that the two systems or networks can communicate. Gateways can convert packets between dissimilar networks.

Figure 2 shows how a gateway can convert large packets from a FDDI into very small packets in an ATM network. Not only does the gateway have to divide the packets, it must also convert the addresses and control messages into formats that can be understood on both networks.


Figure 2: Gateway

Firewall
A firewall is a device or software program that runs on a computer that provides protection from external network intruders by inhibiting the transfer of unauthorized packets and by allowing through packets that meet safe criteria. There are various processes that can be used by firewalls to determine which packets are authorized and packets that should be rejected (not forwarded).

Because firewalls can use many different types of analysis to determine packets that will be rejected, they can be complicated to setup. If a firewall is not setup correctly, it can cause problems for users that are sending and expected return packets that may be blocked by the firewall. Because firewalls process and analyze information, this process requires additional time and this can slow down network data transfer and response time.

Figure 3 shows how a firewall works. This diagram shows that a user with address 201 is communicating through a firewall with address 301 to an external computer that is connected to the Internet with address 401. When user 201 sends a packet to the Internet requesting a communications session with computer 401, the packet first passes through the firewall and the firewall notes that computer 201 has requested a communication session, what the port number is, and sequence number of the packet. When packets are received back from computer 401, they are actually addressed to the firewall 301. Firewall 301 analyzes the address and other information in the data packet and determines that it is an expected response to the session computer 201 has initiated. Other packets that are received by the firewall that do not contain the correct session and sequence number will be rejected.


Figure 3: Firewall

Firewalls are also appropriate for small office and home office (SOHO) applications. There are low-cost software packages and hardware equipment that offer a moderate level of increased security. They cannot stop all hackers, but they will stop some of them.

Technologies: Data Modems, Hub, Bridge

Data Modems
Data modems are devices that convert signals between analog and digital formats for transfer to other lines. Data modems are used to transfer data signals over conventional analog telephone lines. The term modem also may refer to a device or circuit that converts analog signals from one frequency band to another.

A point-to-point analog data circuit requires a modem at each end to transfer digital signals. The type of modems used on each end must be compatible due to encoding and decoding processes. Analog communication lines are restricted to audio bandwidth of 300 Hz to 3300 HZ. To communicate digital data and control signals, the modems vary the frequency of the carrier in each direction based on an agreed to algorithm for encoding bits.

Figure 1 shows a modem with its functional responsibilities listed. From the DTE (serial interface RS 232-C) to the line the modem performs a digital-to-analog conversion and from the line to the DTE an analog-to-digital conversion.


Figure 1: Data Modem

Digital Service Unit (DSU)/Channel Service Unit (CSU)
DSU/CSU’s are the digital equivalent of the analog modem and are translation codecs (COde and DECode) coupled with a network termination interface (NTI). DSU/CSU’s operate only in a digital environment. DSUs/CSUs work together to reformat and channelize digital signals for transmission on multiple channel lines.

Hub
A hub is a communication device that distributes communication to several devices in a network through the re-broadcasting of data that it has received from one (or more) of the devices connected to it. A hub generally is a simple device that re-distributes data messages to multiple receivers. However, hubs can include switching functional and multi-point routing connection and other advanced system control functions. Hubs can be passive or active. Passive hubs simply re-direct (re-broadcast) data it receives. Active hubs both receive and regenerate the data it receives.

Figure 9.9 shows an Ethernet hub. This diagram shows that one of the computers has sent a data message to the hub on its transmit lines. The hub receives the data from the device and rebroadcasts the information on all of its transmit lines, including the line that the data was received on. The hub’s receiver and transmit lines are reversed from the computers. This allows the computers that are connected to the hub to hear the information on their receive lines. The sending computer uses the echo of its own information as confirmation the hub has successfully received and retransmitted its information. This indicates that no collision has occurred with other computers that may have transmitted information at the same time.


Figure 9.9: Hub

Bridge
A bridge is a data communication device that connects two or more segments of data communication networks by forwarding packets between them. The bridge operates as a physical connector and buffer between similar types of networks.

Bridges extend the reach of the LAN from one segment to another. Bridges have memory that allows them to store and forward packets. Bridges are protocol independent as the do not perform protocol adaptations.

Bridges contain a packet address-forwarding table (routing table) that they use to determine if the packets should be forwarded between networks. The packet-forwarding table contained within the bridge can be initially programmed or learned by the bridge. A self-learning bridge can monitor packet traffic in the network to continually update its packet-forwarding table

Bridges primarily operate at the physical layer and link layers of the OSI reference model. A bridge receives packets from one network, review the address of the packet to determine if it should be routed to the other network(s) it is connected to, and retransmits the packet following the standard protocol rules for the systems it is connected to.

Figure 3 shows the basic operation of a bridge that is connecting 3 segments of a LAN network. Segment 1 of the LAN has addresses 101 through 103, segment 2 of the LAN has addresses 201 through 203, and segment 3 of the LAN has addresses 301 through 303. The table contained in the bridge indicates the address ranges that should be forwarded to specific ports. This diagram shows a packet that is received from LAN segment 3 that contains the address 102 will be forwarded to LAN segment 1. When a data packet from computer 303 contains the address 301, the bridge will receive the packet but the bridge will ignore (not forward) the packet.


Figure 3: Bridge