- Количество слайдов: 35
LAN Topologies and Media Access Control BUS 188 Chapter 7
The Lan System n n Generally purchase from vendors Three examples n n Novell IEEE 802. 3 network Banyan Vines Microsoft Windows NT All three examples are discussing different types of n n n Software MAC protocol Network topology
Media Access Control n Refers to a sublayer of the OSI reference model n n Resides in the datalink layer This protocol defines station access as well as data transmission
Topologies n Used to describe different models of creating networks n n n Mathematically topologies refers to points and surfaces in space A topology is a physical layout LAN’s have three basic types of topologies n n n BUS RING STAR
Bus Network n n n Linked by a twisted wire, coaxial cable, or fiber optic no host computer control subject to traffic problems
Bus Standards n Most common Ethernet n Originally proposed in 1972 by Xerox n Later DEC and Intel joined in n Became the foundation for the IEEE 802. 3 standard n n n IEEE 802. 3 standard- contention protocol IEEE 802. 4 standard – token bus BUS speeds are 1, 2. 5, 5, 100 and 1000 Mbps and uses twisted pair or coaxial cable
A Closer look at the IEEE 802. 3 Standard n n n n The objectives of 100 BASE-T 2 are as follows: a)To support the CSMA/CD MAC; b)To support the 100 BASE-T Media Independent Interface (MII), repeater, and Auto-Negotiation; c)To support full duplex operations (Clause 31); d)To provide 100 Mb/s data rate at the MII; e)To provide for operating over two pairs of Category 3, 4, or 5 balanced twisted-pair cabling systems installed in accordance with ISO/IEC 11801, as specified in 32. 7, at distances up to 100 m (328 ft); f)To support operation of other applications on adjacent pairs; g)To allow for a nominal network extent of 200 m including 1)Balanced cabling links of 100 m to support both half duplex and full duplex operation and 2)Two-repeater networks of approximately 200 m span; h)To provide a communication channel with a symbol error rate of less than one part in 10 10 at the PMA service interface.
Ring Network n n Same type of system like a bus except it is connectionoriented no host computer control
Ring Standards n n All nodes on the network receive each message— to a point A node may be either active or inactive n n n Active –means the node is able to send or receive messages Inactive – the node is not able to send or receive messages Most common is IBM’s token-passing ring Also referred to as IEEE 802. 5 standard Ring speeds are 4, 16, and 100 Mbps and uses twisted pair and fiber optic cable
Star Network n n n Good for centralized information processing files can be shared easily open to vulnerability
IEEE Project 802 Subcommitees n 802. 1 High Level Interface n n 802. 2 Logical Link Control n n n Defines overall network architecture, management, interconnection, and all OSI issues Responsible for dividing the datalink layer into two sublayers, the LLC and the MAC 802. 3 CSMA/CD – carrier sense with multiple access & collision detection 802. 4 Token Bus 802. 5 Token Ring 802. 6 Deals with FDDI and MANS
IEEE subcommittees cont n n n 802. 7 Broadband 802. 8 Fiber Optic 802. 9 Integrated Data and Voice Networks 802. 10 LAN Security 802. 11 Wireless LANs
The ANSI FDDI Standard n n n Was created to connect different LANs that were distributed in a metropolitan area. Backbone – a term that is used to describe as a network that connects these disparate networks FDDI specs allow for token-LANs to operate at speeds of over 100 Mbps, and max cable segment is 2 km. Up to 1000 nodes are allowed.
Token Ring n n n n The Token Ring technique is based on the use of a small frame called a token, that circulates when all stations are idle. Whenever a station wishes to send a frame it waits until it receives the token. As it seizes the token one bit is changed which then transforms the token into a start of frame sequence for a data frame. The station then transmits the remainder of the data fields necessary to complete a data frame. The data frame includes the destination station address at its head. The frame is repeated (received, checked and retransmitted), by each station on the network until it circulates back to the source station, where it is removed. In addition to repeating the frame, the destination station retains a copy of the frame and this is indicated by the setting of the response bits, at the end of the frame. The manner in which stations release the token after transmitting data depends on the ring data rate. With rings which operate at the conventional 4 Mbps, the token is released only after the transmitted frames response bits have been received back at the transmitting station. With modern rings which operate at the higher rate of 16 Mbps. , the token is released after transmitting the last bit of the frame. This is known as Early Token Release (ETR).
Data Link Protocols n Different protocols establishes the rules for gaining access to the medium and for exchanging messages. The six main areas that engineers focus on are: n n n Delineation of data Error control Addressing Transparency Code independence Media access
Delineation of data n Each type of protocol has specific message formats that allow for the following services n n n Line control information Station addresses Error detection information Committed information rate Actual data size In the case of tokens—priority status also included
The Token Format n n Tokens are 3 bytes (24 bits) in length and consist of a start delimiter, an access control byte, and an end delimiter. The start delimiter serves to alert each station to the arrival of a token (or data/command frame). This field includes signals that distinguish the byte from the rest of the frame by violating the encoding scheme used elsewhere in the frame. The access control byte contains the priority and reservation fields, as well as a token bit (used to differentiate a token from a data/command frame) and a monitor bit (used by the active monitor to determine whether a frame is circling the ring endlessly). Finally, the end delimiter signals the end of the token or data/command frame. It also contains bits to indicate a damaged frame and a frame that is the last in a logical sequence.
The Data Frame Format n n Data/command frames vary in size, depending on the size of the information field. Data frames carry information for upper-layer protocols; command frames contain control information and have no data for upper-layer protocols. In data/command frames, a frame control byte follows the access control byte. The frame control byte indicates whether the frame contains data or control information. In control frames, this byte specifies the type of control information. Following the frame control byte are the two address fields, the destination address field and the source address fields, which identify the destination and source stations. As with IEEE 802. 3, addresses are 6 bytes in length. The data field follows the address fields. The length of this field is limited by the ring token holding time, which defines the maximum time a station may hold the token.
The Data Format Continued n n 1. 2. 3. n Following the data field is the frame check sequence (FCS) or checksum field. This field is filled by the source station with a calculated value dependent on the frame contents. The destination station recalculates the value to determine whether the frame may have been damaged in transit. If so, the frame is discarded. As with the token format, the end delimiter indicates the end of the data/command frame. It also contains an E bit which is set if any interface detects an error. The frame status byte is only present in Token Ring frames. It contains the A and C bits. When a frame arrives at the interface of a station with the destination address, the interface sets the A bit (=1), as it passes through. If the interface copies the frame to the station, it also sets the C bit (=1). A station might fail to copy a frame due to lack of buffer space or other reasons. When the station which sent the frame strips it from the ring, it examines the A and C bits. The three possible combinations are; A=0 and C=0; Destination not present or powered up. A=1 and C=0; Destination present but frame not accepted. A=1 and C=1; Destination present and frame copied. This arrangement provides an automatic acknowledgment of the delivery status of each frame.
Error Control n Two main techniques are n n Parity Cyclic redundancy checks These techniques have been around for about the last 30 years They work reasonably well, but they come at a price
Addressing n Networks use a hierarchical scheme similar to the postal system n n n Application Network node Network addresses must be unique In ethernet and IBM token ring each address is 48 bits long and set by the manufacturer In ARCnet (star) a node address is 8 bits long and set by the LAN administrator
Transparency n n n The ability to send any bit string as data in a message This allows us to be able to send binary data (such as object programs) as well as text data. Includes start-of-text and end-of-text framing characters (see figure 7 -13).
Code Independence n Allows computers using different data codes to talk to each other. n Ex. Some computers use a 7 -bit others use an 8 -bit code.
Media Access n n Perhaps the most important part of data link protocols. This part determines the way a particular device gains the right to transmit data on the medium There are two main standards that are adhered to—contention and token passing.
Contention n Each station has equal access to the medium. n n Each node monitors the medium to see if a message is being transmitted. If no message is detected then any node can transmit.
Collision n n One of the challenges with contention is the possibility that two stations will transmit at the same time. Collision window example n n Take a 1000 m bus network with two nodes at opposite ends The propagation delay is approx 5 ns per meter which equates to approx 5 microseconds
CSMA/CD n n An alternative format to straight contention Known as a fair protocol n n n All nodes listen to the medium to see whether a message is being transmitted If medium is quiet, transmit message. If medium is busy, wait for the signal to clear than transmit. If a collision occurs, wait for the signal to clear, then wait for a random interval, and then retransmit
Token Passing Media Access Protocol n This is a round-robin process (as opposed to the two previous ones—polling) n n n Nodes wait for transmit token If transmit token is received and there is no message to send, send the token to the next mode. If transmit token is received and there is a message to send, then: n n n Transmit message Wait for acknowledgement When acknowledgement is received, pass the token to the next mode.
Token Ring versus Ethernet (CSMA/CD) n n n Token Ring networks are deterministic in nature -nodes may only transmit at certain well defined times. Result is high bandwidth efficiency. Up to 90% in Token Ring, 40% in Ethernet. Guaranteed sequential access to network eliminates fluctuating response times experienced on other network topologies. Token Ring performance does not deteriorate to the same extent as Ethernet when network traffic increases. This means that at high loads, the presence of collisions of data frames on Ethernet networks becomes a major problem and can seriously affect the throughput. By its nature Token Ring has a higher reliability, the ring can continue normal operation in most cases despite any single fault. Ethernet has an advantage over Token Ring in that the cost of network equipment is lower for Ethernet. Token Ring networks tend to be more expensive to set up and maintain than Ethernet, although hardware costs for Token Ring are decreasing. Advances in Ethernet technology have tended to be much more rapid than Token Ring. Gigabit Ethernet being an example of this. Token Ring technologies are being developed however to allow data transfer rates of 100 Mbps using technologies such as Token Ring Switching.
Token Ring versus Token Bus n n n These standards although similar in some respects were developed with different applications in mind. Token Bus (IEEE 802. 4), networks use a Bus technology and do not have a centralized active monitor. Token Bus was designed with large factories in mind where machines and equipment would be moving around under computer control. Network failures could have serious consequences and so had to be avoided at all costs. On the other hand the Token Ring network standard was designed with office automation in mind, where a failure once in a rare while could be tolerated as the price for simpler system. Token Bus is not as deterministic as Token Ring but like Token Ring it has excellent throughput and efficiency at high load. Token Bus uses broadband transmission and cabling which obviously gives it a bandwidth advantage over Token Ring but the downside is the equipment costs (modems, wideband amplifiers, etc). The IEEE 802. 4 protocol is also extremely complex compared to IEEE 802. 5 and it has a substantial delay at low loads.
Other comments on differences n n n Token Rings offer a steady reliable network that works well under mission critical situations. Developments in the field of token rings promise a high-speed and scaleable network that competes well with the more popular Ethernet. The reliability of token-rings stems from the sequential manner in which it serves computers. Each station on a token ring is given equal access to the network guaranteeing that no one station dominates the bandwidth. This also means that, unlike collision-based Ethernet, performance does not degrade with increased traffic on the network. The efficiency of token-rings use of a networks may be as high as 90% compared with 40% for Ethernet. Despite rapid developments in Ethernet token ring vendors have remained competitive and willing to invest in further developing token rings. The introduction of token ring switches which allow rings to be segmented has vastly improved the performance of token rings. These switches allow stations to be given dedicated access to a ring which in turn allows stations to run in full-duplex mode, effectively doubling its bandwidth.
Other comments continued n n n Token Ring vendors have been eager to compete with fast Ethernet by standardizing high-speed token ring, a hybrid of existing technologies which will allow token rings to reach speeds of 100 MB/sec. Standards are also being developed which will allow this 100 MB/sec to be scaled up to 1 GB/sec. Despite these developments token rings cannot compete with the advances in Ethernet bandwidth technology. Token rings also share the cabling problems of Ethernet - STP cables are expensive and difficult to install while UTP cables have very limited distances and require routers and bridges to overcome this. Overall, token rings are a viable networking solution that compares favorably with its competitors and will continue to be the preferred network for some applications in many organizations.