Chapter 12 Local Area Networks

1. Chapter 12Local Area Networks Project 802 Ethernet Token Ring FDDI

2. LAN LAN is a data comm system that allows a number of independent devices to communicate directly with each other in a limited geographic area. Dominated by four architecture: ethernet, token bus, token ring and FDDI Data link control portion of the LAN protocols in use today are all based on HDLC. However, each protocol has adapted HDLC to fit the specific requirements of its own tech.

3. PROJECT 802 1985 IEEE started a project called project 802 to set standards to enable intercommunication between equipment from a variety of manufacturers. P802 is a way of specifying functions of the physical layer, data link layer and network layer for interconnectivity of major LAN protocols. Project 802 not intend to replace OSI.

4. See fig. 12.1 about the relationship of IEEE P802 and OSI. The IEEE has subdivided the data link layer into two sublayers: LLC Logical Link Control MAC Medium Access Control About the two sublayer, P802 contains a section governing internetworking. This section assures the compatibility of different LANs and MANs across protocols and allows data to be exchanged across otherwise incompatible networks.

5. The strength of P802 is modularity. Subdivide the function and numbering it (pls refer fig 12.2) This benefit to LAN designer who were able to standardize functions that can be generalized and to isolate functions that must remain specific.

8. IEEE 802.1 IEEE 802.1 is the section of Project 802 devoted to internetworking issue in LANs and MANs. Not yet complete. Although not yet complete, it seeks to resolve the incompatibilities between network architectures without requiring modifications in existing addressing, access, and error recovery mechanisms, among others. Used in networking/internetworking devices such as repeater, bridge, routers and gateway.

9. LLC Logical Link Control IEEE P802 model takes the structure of HDLC frame and divides it into 2 sets of functions. One set of function contains the end-user portions of the frame: the logical addresses, control information and data which handled by the 802.2 LLC protocol.

10. MAC Medium Access Control The second set of functions, (MAC) sublayer resolves the contention for the shared media. It contains the synchronization, flag, flow, and error control specification necessary to move info from one place to another, as well as the physical address of the next station to receive and route a packet. MAC protocols are specific to the LAN using them(Ethernet, Token Ring, Token Bus, etc.)

11. LLC: Protocol Data Unit (PDU) PDU? Data unit in the 1st sublayer LLC Contain 4 field familiar from HDLC: a destination service access point (DSAP), source service access point(SSAP), a control field, and an info field. (see fig 12.3) The DSAP and SSAP are addresses used by the LLC to identify the protocol stacks on the receiving and sending machines that are generating and using the data. See fig. 12.3 for description. The control field of the PDU is identical to the control field in HDLC. Refer fig 12.4 about I-frame, S-frames and U-frames. The PDU has no flag fields, no CRC, and no station address. These fields are added in the lower sublayer (the MAC layer)

14. 802.3 Ethernet Dev by Xerox, intel and digital equipment IEEE 802.3 defines 2 categories: baseband and broadband. Base specifies a digital signal. Broad specifies analog signal. IEEE divides the baseband category into 5 different standard see fig 12.5 IEEE defines only one specification for the broadband. See fig 12.5 Access method for ethernet is CSMA/CD

16. Ethernet access method: CSMA/CD Whenever multiple users have unregulated access to a single line, there is a danger of signals overlapping and destroying each other. Such overlaps, which turn the signals into unusable noise are called collisions. A LAN therefore needs a mechanism to coordinate traffic, minimize the number of collision that occur, and maximize the number of frames that are delivered successfully. So access mechanism should be used is CSMA/CD

17. How CSMA/CD work? Any workstation wishing to transmit must first listen for existing traffic on the line. A device listens by checking for a voltage. If no voltage is detected, the line is considered idle and the transmission is begin. During the data transmission, the station checks the line for the extremely high voltages that indicate a collision. If a collision is detected, the station quits the current transmission and waits a predetermined amount of time for the line to clear, then sends its data again.

18. Ethernet: addressing and electrical specification Addressing: each station got NIC which provide 6-byte physical address Signal: baseband system use manchester digital encoding digital/digital conversion For broadband (only one: 10broad36) it uses digital/analog conversion differential PSK. Data rate: support between 1 and 100Mbps

19. Ethernet frame format Consist of 7 fields se fig 12.7 Preamble : this 7 bytes field alert the receiving system to the coming frame and enable it to synchronize its input timing. Start Field Delimeter: tells the receiver that everything that follows is data, starting with the addresses. Destination Address: this field contain the physical address of the packet�s next destination

20. Source Address: contains the physical address of the last device to forward packet Length type of PDU: This two bytes field indicate the number of bytes in the coming PDU. 802.2 frame PDU. This field of the 802.2 frame contains the entire 802.2 frame as a modular, removable unit. The PDU can be anywhere from 46 to 1500bytes long. CRC. The last field in the 802.3 frame contains the error detection information.

22. Ethernet 802.3: implementation Although project 802 standard focus on the data link layer of the OSI model, the 802 model also defines some of the physical specifications for each of the protocol defined in the MAC layer. In the 802.3 standard, the IEEE defines the types of cable, connections, and signals that are to be used in each of five different Ethernet implementations. All Ethernet LANs are configured as logical buses, although they may be physically implemented in bus or star topologies. Each frame is transmitted to every station on the link but read only by the station to which it is addressed.

23. Thick Ethernet: 10BASE5 The first of the physical standards defined in the IEEE 802.3 model. Also known as thicknet. (See fig 12.9) This is a bus topology LAN that uses baseband signaling and has a maximum segment of 500m. In 10BASE5, networking devices (bridges, repeaters can be used to overcome the size limitation of local area networks. In thicknet LAN can be divided into segments by connecting devices. In this case, the length of each segment is limited to 500meters. However, to reduce collision, the total length of the bus should not exceed 2500 meters (5 segment). Also, the standard demands that each station be separated from each neighbor by 2.5meters (200 station per segment and 1000 stations in total.) see fig 12.8

26. 10Base5 topology Physical connectors and cables utilized by 10Base5 include coaxial cable, NIC, transceiver, and attachment unit interface AUI cables. See fig 12.9 for interaction of components RG-8 cable: thick coaxial cable (RG=Radio Government) Transceiver/MAU: Each station is attached by an AUI cable to an intermediary device called transceiver (transmitter/receiver). This device performs the CSMA/CD function (check voltage and collison) AUI cables: Each station is linked to its corresponding transceiver by an attachment unit interface (AUI), also called a transceiver cable. An AUI is a 15-wire cable with plugs that performs the physical layer interface functions between the station and the transceiver. 50m only. Each end of this cable is DB-15 connector see fig 12.10. Transceiver tap: used to connect transceiver into the line. The tap is a thick cable-sized well with a metal spike in the centre. See fig 12.10. Also known as vampire tap because it bites the cable to allow electrical connection between trancseiver and the cable.

31. Compared to 10Base5, 10Base2 topology is much simpler. In this technology, the transceiver circuitry has moved into the NIC and the transceiver tap has been replaced by a connector that splices the station directly into the cable, eliminating the need for AUI cables. Devices used in this topology are: NIC: provide same function as NIC in thicknet with transceiver capability to check for voltages on the link in order to perform CSMA/CD. Thin coaxial cable: these cable are relatively easy to install and move around (especially inside existing building where cabling must be pulled through the walls and ceilings.) BNC-T: The BNC-T connector is a T-shaped device with 3 port: one for the NIC and one each for the input and output ends of the cable.

34. 10Base-T: twisted pair ethernet Instead of individual transceiver, 10Base-T Ethernet places all of its networking operations in an intelligent hub with a port for each station. Station are linked into the hub by 4 pair RJ-45 cable (8 wire unshielded twisted-pair cable) terminating at each end in a male-type connector much like a telephone jack. The hub fans out any transmitted frame to all of its connected stations. Logic in the NIC assures that the only station to open and read a given frame is the station to which that frame is addressed.

35. As fig 12.12 shows, each station contains an NIC. A length of 4pair UTP of not more than 100 meters connects the NIC in the station to the appropriate port in the 10Base-T hub. The weight and flexibility of the cable and the convenience of the RJ-45 jack and plug make 10Base-T the easiest of the 802.3 LANs to install and reinstall. When a station needs to be replaced, a new station can simply be plugged in.

38. Other Ethernet Networks During the last decade there has been an evolution in Ethernet networks. Several new schemes have been devised to improve the performance and the speed of Ethernet LANs. This new improve LAN are: Switched Ethernet Fast Ethernet Gigabit Ethernet

39. Switched Ethernet Switched ethernet is an attempt to improve the performance of 10Base-T Ethernet. The 10Base-T Ethernet is a shared media network, which means that the entire media is involved in each tranmission. This is because the topology, though physically a star, is logically a bus. When a station sends a frame to a hub, the frame is sent out from all ports(interfaces) and every station will receive it. In this situation, only one station can send a frame at any given time. If more than one station try to send frame simultaneously, of course there is a collision.

42. Fast Ethernet With new applications such as CAD, image processing, and real-time audio and video being implemented on LANs, there is a need for a LAN with a data rate higher than 10Mbps. And�fast ethernet operates at 100 Mbps. Fast ethernet is a version of ethernet with a 100 Mbps data rate. There is no change in the access method. There is no change in the frame format.

43. The only two changes in the MAC layer are the data rate and the collision domain. The data rate is increased by a factor of 10; The collision domain is decreased by a factor of 10. This means that the collision domain must be decreased 10 times from 2500meters to 250meters. This decrease is not a problem because LANs today connect desktop computers that are not more than 50 to 100 meters away from the central hub. This means the collision domain is between 100 and 200 meters.

44. In the physical layer, the specification developed for fast ethernet is a star topology similar to 10base-T. However, to match the physical layer to different resources available, IEEE has designed two categories of Fast Ethernet: 100Base-X (use 2 cable between station and hub) 100Base-T4 (use 4 cable between station and hub) See fig 12.16. 100Base-X itself is divided into two types TX and FX.

46. 100Base-X: 100Base-TX Uses 2 category 5 (CAT5) UTP or 2 STP to connect a station to the hub/switch. One pair is used to carry frames from the station to the hub and the other to carry frames from the hub to the station. The signaling method (dig/dig) is NRZ-I. Encoding method is 4B/5B to handle the 100Mbps. The distance between the station and the hub/switch should be less than 100 meters. Refer fig 12.17

48. 100Base-X: 100Base-FX The 100Base-FX design uses 2 optical fibers, one to carry frames from the station to the hub. And the other from the hub to the station. The encoding is 4B/5B. Signaling method is NRZ-I. The distance between the station and the hub/switch should be less than 2000 meters. Refer fig 12.18

50. 100Base-T4 The 100Base-T4 scheme was designed in an effort to avoid rewiring. It requires four pairs of CAT3 UTP (voice grade) that are already available for telephone service inside most building. Two of the four pairs are bidirectional; the other two are unidirectional. This means that in each direction, three pairs are used at the same time to carry data. Because a 100-Mbps data rate cannot be handled by a voice-grade UTP, the specification splits the 100Mbps flow of data into three 33.66 Mbps flows. Refer fig 12.19

53. Gigabit Ethernet The migration from 10 Mbps to 100 Mbps encouraged the IEEE 802.3 committee to design Gigabit Ethernet, which has a data rate of 1000Mbps or 1 Gbps. MAC layer and the access method remain the same, but the collision domain is reduced. The physical layer- the transmission media and the encoding system however changes. To cater 1000mbps, optical fiber cable is used instead of twisted pair cable although the protocol doesn�t eliminate the use of twisted pair cables. Gigabit Ethernet usually serves as a backbone to connect Fast Ethernet networks.

54. Cont�d Four implementations have been designed for gigabit ethernet: 1000Base-LX 1000Base-SX 1000Base-CX 1000Base-T

56. 10Broad36 10BROAD36 - 10BROAD36 is a seldom used Ethernet specification which uses a physical medium similar to cable television, with CATV-type cables, taps, connectors, and amplifiers.