Local Area Networks Ch. 4.pptx

Local Area Networks Ch. 4 • All data communications were originally point-topoint, much as airplane travel was before the advent of the airport as a hub • Data communication in the 1970s was a costly venture, as bandwidth, disk storage, and printers were expensive • Data communications followed a...

Local Area Networks Ch. 4 • All data communications were originally point-topoint, much as airplane travel was before the advent of the airport as a hub • Data communication in the 1970s was a costly venture, as bandwidth, disk storage, and printers were expensive • Data communications followed a similar pattern by utilizing wide area data networks to increase efficiency and reduce costs • The aforementioned fork in the road comes at this point: do we want to network locally or do we want to go through the public network? Definitions • For the past 25 years, the process industries have had large numbers of “smart” devices that communicate electrically to monitor or control a process or processes • The combination of such devices and the electrical connectivity that enables them to communicate is, by definition, a network • However, industrial networks are referred to by many different names : data highways, fieldbuses, distributed control systems , and so on • Industrial LANs seem to be everywhere and every instrument manufacturer has a proprietary variant • In addition to the vendor-proprietary LANs, there are now several standards for industrial LANs, including: FOUNDATION Fieldbus , PROFIBUS and PROFINET, plus Modbus/TCP, and EtherCat Programmable controllers can be used to form a distributed control system when they are connected by a local area network with a third-party human-machine interface to provide integration and a graphics interface Definitions We prefer the following definition: If a company owns the medium of transmission, the infrastructure, and three or more devices that are in communication, it has a LAN An alternate definition of local area network is this: a system with more than three nodes that uses a protocol with defined rules and performs a specified set of functions IEEE 802 LAN Model THE IEEE 802 MODEL IS A LAYER 1 AND 2 MODEL CONCERNING ITSELF, FROM INCEPTION TO THIS DAY, WITH VARIABLE SIZE PACKET NETWORKS ONE SHOULD FIRST LOOK AT THE OPEN SYSTEMS INTERCONNECT ISO 7498 MODEL AND THE IEEE 802 MODEL, TO DISCOVER HOW THESE MODELS COMPLEMENT ONE ANOTHER THE 802 STANDARD DESCRIBES NETWORK FUNCTIONS THE CURRENT 802.2 STANDARD ALSO PROVIDES FOR THE SHARING OF A LAN BY MULTIPLE PROTOCOLS AND SERVICES, EACH HAVING DIFFERENT UPPER-LAYER PROTOCOL “STACKS”— ISO MODEL LAYERS 3 THROUGH 7 OR INTERNET PROTOCOL VERSION 4 AND INTERNET PROTOCOL VERSION 6 ), WITH THE ABILITY TO ROUTE MESSAGES TO THE RIGHT PROTOCOL STACK Layer 1, The Physical Layer • How data is placed onto and taken off of the medium determines the type and method for sharing the network • In general, if modulation-demodulation is used—that is, if it uses a carrier signal and data modulation/demodulation—it is a broadband or carrier-band system • If data is placed digitally at a bit rate that approaches the bandwidth capacity of the channel, it is a baseband system Broadband • Broadband refers to separating signals by frequency or analog broadband) or time or digital broadband) • Broadband services can include network technology for multiple LANs on one trunk or they can consist of carrier-supplied services, such as Integrated Services Digital Network and variations of Digital Subscriber Line , which are supplied to networks from a WAN provider • The bulk of the technology found in an analog frequency division, multiplexing LAN evolved directly from cable TV systems, right down to the connectors and the 75-ohm cable Layer 1, The Physical Layer • How data is placed onto and taken off of the medium determines the type and method for sharing the network • In general, if modulation-demodulation is used—that is, if it uses a carrier signal and data modulation/demodulation—it is a broadband or carrier-band system • If data is placed digitally at a bit rate that approaches the bandwidth capacity of the channel, it is a baseband system Broadband • Broadband refers to separating signals by frequency or analog broadband) or time or digital broadband) • The main bus connects the devices in the network and it will have a large number of different signals on it • Broadband services can include network technology for multiple LANs on one trunk or they can consist of carriersupplied services, such as Integrated Services Digital Network and variations of Digital Subscriber Line , which are supplied to networks from a WAN provider • The bulk of the technology found in an analog frequency division, multiplexing LAN evolved directly from cable TV systems, right down to the connectors and the 75-ohm cable Broadband • Analog or digital broadband trunks offer a multitude of services—data, voice, and television—all on the one cable Baseband • In baseband transmission, there is only one digital signal on the bus or shared medium • Switched media has gained acceptance as a lower cost, high-performance alternative to statistical-based media access schemes • Switching is accomplished by using a fast-switching matrix, which is a set of electronic switches arranged so there can be a path from any port to another port depending upon a signal-contained address • Another method of sharing is a contention method, wherein everyone is allowed to grab the channel for a short period of time if it is currently idle • This approach can lead to “collisions,” when two devices attempt to initiate communications concurrently Baseband • This kind of sharing method is like getting to an intersection and going right across if there is no cross traffic or dashing across as soon as there is a break in traffic Carrier Band • Used in industrial environments, carrier band has only a single carrier on the bus, which is used to translate the digital signal away from a DC reference, i.e., to modulate the digital signal to the frequencies available with the medium • Carrier band is a broadband technology but it has some of the restrictions of baseband; notably, it only has a single channel • Carrier band uses coaxial cable and for the most part has been replaced by other twisted pair schemes; it does not, at this time , have a significant market share in instrumentation and control Topologies • Whether a system is baseband or broadband has an effect on the system’s topology, or how the system is laid out • In practice, most broadband systems use the bus topology , while most baseband systems now use star topology • That is not to say that they are always that way, particularly if the medium is fiber optic, where the topology may be some form of point-to-point or star Topology Definitions • Network topology is the arrangement of the various links and nodes of a network, as well as the schematic of the various components within a network • While topology refers to the physical layout, a network systems’ architecture refers to the logical connections between intermediary and end devices • The term logical connection usually refers to the way in which the network operating software views the dataflow of a device Star Topology • The star topology has been a fundamental computer network topology since the computer joined networking • In one of the first instances of the star topology , a central computer, usually a mainframe, time-shared all inputs and outputs • In many instances of star topology, all linked devices input to the computer and the computer outputs to all linked devices • The star topology is the way computer networks were arranged until the decreasing cost and size of electronics allowed processing power to be spread out among users • The star topology is identical to the topology of direct digital control systems Ring Topology • A physical ring is just what it sounds like: it is comprised of point-to-point connections and each device in the network is in the path of data travel • Each device must be active on the ring or there must be some provision for bypassing the device in the event of failure or inactivity • Token rings simplify protocol and the maintenance of synchrony throughout the ring • Due to the fact that fiber optic cable is not open to easy line taps , terminations are difficult , and most fiber connections are considered a point-topoint connection, the ring topology is ideal for fiber optic-based networks Ring Topology • In most ring topologies there is always a message being transmitted, even if it is just the special “token” message being passed to the next node in the ring • A node’s failure to “hear” a token message within a defined time window will cause some form of token regeneration to be activated Bus Topology • The bus topology is used extensively in industry because of the installation flexibility it affords and because the failure of a node to bypass when it’s inactive is not necessarily devastating to network operations • Most industrial bus systems used to be token passers and were used in a logical ring; that is, they were physically connected as a bus • However, the system software had them using circular token passing to control access, as if in a physical ring • Coaxial cable-based Ethernet was eventually replaced by wiring hubs, where the drop to each node was a UTP cable with an 8p8c plug and connector at each end Bus Topology • A hub is a dumb device that electrically works much like a bus ; the main difference is that a hub failure causes total communication failure because the hub is an electronic device • Ethernet switches are essentially powerful computers with large numbers of Ethernet network interface cards , whose function is to take incoming Ethernet frames from connected computers and pass them back out to their intended destinations based on the MAC addresses in the frames Transmission Media • There are basically three types of transmission media available: copper, fiber optic, and wireless • In the next section we will discuss Ethernet media, as all three types of media are currently recommended for Ethernet systems Ethernet Media • The traditional twisted-pair telephone wiring was investigated carefully by early network designers because of the tremendous savings that could be realized if the already installed media could be used instead of special cabling • However, even the first networks ran at speeds at 10 MBps, 1980) that required either specific adjustments to use with telephone media or required the use of different media altogether • At current network speed and interface requirements, the medium will almost always be a dedicated twisted-pair cable or higher), wireless, or fiber optic, rather than the installed telephone wiring Ethernet Media • Categories are, for our purposes, defined in IEC 11801 and assign the maximum data rate a cable is capable of carrying over a stated distance • As Ethernet assumes a larger share of the industrial interconnect market, its Layer 1 specification, listed in table 4-2 for types 10, 100, 1,000, and 10,000 ), will likely dominate • Having successfully conquered the office and business environment, Ethernet has continued on to assume a large share in the industrial arena • The IEEE 802 standard now encompasses more than local area networks, which are used generally as Layers 1 and 2 in many networking applications • This is by no means a complete list, just a selection of the main committees 802.11 • The original 802.11 standard is now a legacy; there is no 802.11 standard anymore, just the amendments, due to the marketplace availability of much higher data speeds • The original 802.11 standard, approved in 1997, provided for 1 or 2 Mbps transmission in the 2.4 GHz band using either frequency hopping spread spectrum or direct sequence spread spectrum • 802.11a Standard 802.11a, approved in 1999, used the 5 GHz band, which provided up to 54 Mbps and used an orthogonal frequency division multiplexing encoding scheme rather than FHSS or DSSS 802.11b • The first of the 802.11 amendments for which commercial product became available was 802.11b • It extended the original 802.11 to provide 11 Mbps transmission in the 2.4 GHz band using DSSS • The revised standard, 802.11b, was accepted in 802.11g • It works in the 2.4 GHz band but it has a maximum data rate of 54 Mbps and is backward compatible with 802.11b • 802.11n At the beginning of 2004 the IEEE formed a new 802.11n task group to develop a new wireless modulation method capable of achieving a data rate of up to 540 Mbps • Proposed standard 802.11n was built on previous 802.11 standards by adding multiple antennas technology) to make possible spatial multiplexing and other coding schemes 802.11ac • IEEE 802.11ac-2013, published in December 2013, is an amendment to IEEE 802.11 that builds on 802.11n • Changes, compared to 802.11n, include wider channels in the 5 GHz band, more spatial streams , higher order modulation the number describing the available states in one quadrant of a cycle vs. 64QAM), and the addition of multi-user, multiple-input multiple-output technology • As of October 2013, high-end implementations support 80 MHz channels, three spatial streams and 256-QAM, yielding a data rate of up to 433.3 Mbps per spatial stream, 1300 Mbps total, in 80 MHz channels in the 5 GHz band 802.11ad • IEEE 802.11ad is a published standard that defines a new physical layer for 802.11 networks to operate in the 60 GHz millimeter wave spectrum • This frequency band has significantly different propagation characteristics than the 2.4 GHz and 5 GHz bands where Wi-Fi networks operate • The certification program is now being developed by the Wi-Fi Alliance, instead of the now defunct WiGig Alliance • 802.11af IEEE 802.11af, also referred to as “White-Fi” and “Super Wi-Fi,” is a standard, approved in February 2014, which allows WLAN operation in the TV white space spectrum in the VHF and UHF bands between 54 and 790 MHz 802.11ah • IEEE 802.11ah is an emerging standard that defines a WLAN system operating at sub-GHz license-exempt bands , with final approval slated for March • Due to the favorable propagation characteristics of the low frequency spectra, 802.11ah will provide improved transmission range, compared with the conventional 802.11 WLANs operating in the 2.4 GHz and 5 GHz bands • 802.11ah can be used for various purposes, including large scale sensor networks, extended range hotspots, and outdoor Wi-Fi for cellular traffic offloading, although the available bandwidth is relatively narrow due to the lower frequency carrier used 802.11ah • All the preceding WLAN technologies are essentially used to bridge one or more wireless clients to a wired LAN, or possibly to each other, via a wireless access point • They are designed to operate over moderate distances and provide a bit rate comparable to wired LANs Figure 4-7 shows how APs can be integrated into a wired LAN • The 802.11af and 802.11ah standards are still too new to be included in this table 802.16 WiMAX • WiMAX is a metropolitan area network standard meant to cover much larger areas than the 802.11 standards • Standard IEEE 802.16 was approved in 2002 for operation over a broad frequency range of 10 to 66 GHz • WiMAX is intended for two distinctly different markets: rural distribution of broadband services and as a contender for broadband delivery of data to and from cell phones Wireless Mesh Networks • A wireless mesh network is a communications network made up of radio nodes organized into a mesh topology • The mesh clients are often laptops, cell phones, and other wireless devices; the mesh routers forward traffic to and from the gateways which may, but need not, connect to the Internet • Wireless mesh networks can be implemented with various wireless technologies including 802.11, 802.15, 802.16, cellular technologies, or combinations of more than one type LAN Infrastructure • A LAN consists of devices other than the servers and clients • Aside from the physical media, there are also active components that form a LAN and these components generally share a common definition, although sometimes it gets stretched by vendors’ or users’ usage • Some of these components are merely dumb electronic devices, while others contain impressive computing power and associated software, requiring extensive configuration Repeater • A repeater is a Layer 1 device that regenerates the input signal, thus restoring its amplitude and clock sequence • It is used to extend a particular segment of a network • There are two types of repeaters: a simple amplifier that is designed to boost a degraded signal , and a receiver/transmitter device that sits between LAN segments listening for signals/messages on either segment and then totally regenerating a new, duplicate signal/message on the other segment • Hub Basically obsolete now, hubs first became popular with the advent of 10BASE-T Ethernet, which featured a twistedpair hub consisting of 4 to 24 twisted-pair connectors in commercial use Repeater • Hubs can also take the form of optical fiber hubs but, in either case, hubs may be called wiring concentrators Layer 1 and Layer 2 Devices • The limitations of a repeater arise from the fact that it only knows 1s and 0s and attaches no meaning to either state, it just regenerates the signal condition; that is, it outputs a signal with the correct rise and fall time and voltage state achieved • One way to extract better performance is to have a bridge between the two network nodes • A bridge consists of a repeater with intelligence that reads the Layer 2 information: the destination and source adapter addresses; the “type or length” field; and the frame check sequence , which is a cyclic redundancy check Bridge • Containing two sets of Layers 1 and 2, a bridge is a device that connects two network segments of the same type, perhaps to connect different segments or to split a single segment • A bridge provides physical connections, namely, MAC addressing and signaling • Modern bridges are “learning bridges,” that is, they look at all message traffic on both segments and their Layer 2 addresses Ethernet Switch • A switch is essentially a bridge with a large number of Ethernet NICs • Like a bridge, the switch learns the MAC addresses of the devices on each of its ports and, just like a bridge, a switch examines the Layer 2 addresses to decide where the destination node is located • Because logic, memory, and arbitration procedures are built into the switch, each node thinks it has total use of the network • Switch technology has come down in price enough to make bridges and hubs obsolete Router • A router contains logic that examines Layers 1, 2, and 3 information and will connect two or more networks, providing network addressing, error correction, physical signal conversion, and conversion to compensate for differences in signal frame size • Routers often perform a gateway function in converting messages from the format used on one network to the format used on another, in addition to dealing with signaling and media differences • A router may be a separate computer-based device designed specifically for routing functions or a set of software and network interfaces added to a generalpurpose computer or server Brouter • This is also called a brouter , a network device that works as a bridge and as a router • A Layer 3 switch is a device capable of reading Layer 2 and Layer 3 addresses, so it knows whether the message is to stay on the same physical segment of a network, to be bridged to another physical segment , or to leave for another network • Brouters operate at both the network layer for routable protocols and at the data link layer for non-routable protocols Gateways • The organization of 1s and 0s on System A may vary considerably with that of System B. The protocols and addressing may be quite different • In order to accommodate the differences, it takes a dual transceiver, that is, it takes all seven layers for each network to produce the native data, which can then be sent back down the other system • In Internet terminology, a gateway is the computer/device that makes the connection between the local autonomous network and the Internet • An important and common use of gateways in industrial settings is to connect incompatible networks to allow for data exchange using only three layers of the seven-layer OSI Model Gateways • In the SCADA world, gateway devices can be used to perform protocol conversion so that, for example, a SCADA system that uses asynchronous, serial DNP3.0 protocol to poll its remote terminal units can communicate with an RTU that only supports asynchronous, serial Modbus protocol Layer 2 Functions • We briefly mentioned some of the Layer 2 functions when discussing the bridge • At this point, we will take a more detailed look at Layer 2 operation MEDIA ACCESS MEANS HOW A DEVICE ON A NETWORK, SUCH AS A LAN, GAINS CONTROL OF THE NETWORK MEDIA IN ORDER TO TRANSMIT ITS INFORMATION—IN OTHER WORDS, WHO TALKS AND WHEN MANY DIFFERENT METHODS ARE AVAILABLE FOR GAINING ACCESS AND MOST DEPEND ON NETWORK PHILOSOPHY , NOT NECESSARILY ON NETWORK TOPOLOGY BEFORE THE RISE OF DISTRIBUTED PROCESSING, MULTI-DROP SYSTEMS AND WHAT THERE WAS OF NETWORKS USED THE STAR TOPOLOGY Media Access ALTHOUGH POLLING MAY BE ONE OF THE OLDEST ACCESS METHODS, IT IS STILL USED AND, IN CERTAIN SITUATIONS, QUITE EFFECTIVELY IN POLLING, A SCHEDULER, WHETHER A SOFTWARE OR HARDWARE ENTITY , DETERMINES WHO SHOULD SPEAK, WHO SHOULD LISTEN, AND UNDER WHAT CONDITIONS POLLING CONSISTS OF A PRIMARY NODE ASKING ONE OR MORE SECONDARY NODES , EACH IN TURN, IF THEY HAVE ANY TRAFFIC TO SEND Polling AN EXAMPLE OF A MODIFIED METHOD IS INTERRUPT POLLING, ALSO KNOWN AS EVENT-DRIVEN POLLING OR “HUBBING.” IN THIS SCHEME, THE MASTER STATION DOES NOT QUERY STATIONS BUT WAITS UNTIL ONE OF THE STATIONS ANNOUNCES ITS INTENT TO TRANSMIT BY RAISING AN INTERRUPT OR BY OTHERWISE SIGNALING ITS INTENT IT IS MUCH LIKE THE CLASSROOM, WHERE THE INSTRUCTOR MAY OR MAY NOT INITIATE ANY COMMUNICATION, BUT IN ORDER FOR YOU, THE STUDENT, TO INITIATE COMMUNICATION, YOU MUST RAISE YOUR HAND Event-Driven Polling Token Passing • In the token-passing access method, a short message with a unique digital pattern is transferred from peer-to-peer among all the participating peers • Token passing was briefly discussed in an earlier description of the token-passing ring and that material is still applicable to token passing in general • Token passing is deterministic; every station knows when the token will arrive, so it knows within a specified period of time when it will receive a transmission initiated by another station Carrier Sense Multiple Access with Collision Detection • In order to ensure that no receiving station treats such messages as valid, and so that the transmitting stations know that their transmissions have failed, there is a need to detect such collisions • To prevent this from happening, the Ethernet frame has a minimum length • At 10 Mbps, this minimum length requirement is accounted for in the minimum payload size of 46 octets • With 100 Mbps Ethernet, the NIC will append the necessary “frame extension symbols” to achieve a minimum of 512 octets for the frame • IEEE 802.3/Ethernet: A Layer 1 and 2 Standard Although IEEE 802.3 is not specifically an industrial protocol as such, it is used throughout industry today With its ado pti on as the Use f IEEE 802. 3 aperformance, s an industria l D ueoto its proven underpinnings o f va ri o us netwcost, ork proand to col wa stransmission ini tia l ly highly low instrumenta ti o nhigh LANs, such as deba ted, wihas th arguments presented speeds, become the deFOUNDATIitON Fieldbus H SE abo ut its a bil ity to pro v i de and P ROFI NET1, and it rea2ches a l l theinwa y facto Layer standard determini sti c perfo rmance due to do wn to the end devi ces o n the pla nt many networks co ll isioindustrial ns flo o r Carrier Sense Multiple Access with Collision Detection 10BASE5 • This is interpreted as meaning “10 Mbps, Baseband, 500 m end-to-end ,” using Belden 89880 coaxial cable, also known as ThickNet or, if you will, “the frozen orange garden hose”—a reference to the difficulty in handling that this particular cable presents • The cable impedance is 50 ohms and it must be terminated at each end in a 50-ohm non-inductive precision resistor 10BASE2 • This is interpreted as “10 Mbps, Baseband, 185 m end-toend ,” using RG58A/U coaxial cable, also called ThinNet • The coaxial cable is terminated in 50-ohm resistors at each end node, the node being connected by a BNC “T” connector, which is attached to the NIC • ThinNet/10BASE2 is usually much easier to work with than ThickNet; however, it does have two cables coming into the back of each node, aside from the end ones • The node NIC is connected to the hub or switch with a 100ohm, UTP Category 3 cable, using only two pair of the fourpair cable • Due to the prevalence and low cost of Ethernet NICs capable of “fast” Ethernet, this standard is also considered obsolete but most NICs support 10/100BASE-T and can drop down to 10 Mbps if needed 10BASE-FL • This is “10 Mbps Ethernet over fiber.” • It requires one transmit and one receive fiber and uses an optical switching hub • Depending on the fiber cable used, 10BASE-FL may have a maximum distance of nearly 2 km 100BASE-T • This is “100 Mbps, Baseband, Twisted Pair.” • The 100BASE-T specifications are generally identical to 10BASE-T, with the exception that the connecting cable must be Category 5 , as opposed to Category 3 for 10 Mbps • Figure 4-19 represents 100BASE-T as well 100BASE-FX • This is “100 Mbps Ethernet over fiber.” • When run in full duplex mode, 100BASE-FX can achieve a distance of 2 km • Half-duplex is limited to 400 m 1000BASE-T • This means “1 gigabit per second Ethernet.” • Originally specified for Category 5 cable, 1000BASE-T requires a tighter specification cable, such as Category 5e or 6a, to operate properly over copper at 100 m , its maximum distance • Note that all four pairs of the cable are used 1000BASE-CX • This is a shielded twisted-pair cable with a 25 m maximum distance • It is no longer considered an active specification 1000BASE-SX • This is a short-wavelength fiber cable with a maximum distance of up to 550 m 1000BASE-LX • This is a long-wavelength fiber cable with a maximum distance of up to 5 km • 10GbE – 10 Gigabit Ethernet Over Fiber This is a standard meant for trunks , operated originally only over fiber optic • This standard includes several variations and designations: 10GbBASE-SR, 10GbBASE-LR, 10GbBASE-LRM, 10GbBASE-ER, 10GbBASE-LX4 , and 10GbBASE-PR • The following are the fiber versions of 10GbE media: SR was developed to run on multi-mode fiber for a distance of 26 m • 10 GbE – 10 Gigabit Ethernet Over Copper Due to both monetary and certain installation requirements, a copper version of the 10 GbE was developed for unshielded twisted pair • The standard was known as 10GBASE-T and was published in September IEEE 802 Media Access Control • The media access control sublayer interfaces the Logical Link Control sublayer with the Physical layer • The MAC logic provides the service of sending and receiving 1s and 0s by arranging them into frames • In transmit mode, the MAC: Initiates transmission , Assembles the frame, Calculates the Frame Check Sequence, Sends the frame, and Ceases transmitting • In both transmit and receive modes, the Media Access Control logic also monitors for collisions and executes the applicable recovery logic based on the mode • It is the MAC layer that is primarily responsible for the media access method—token bus or CSMA/CD—hence its name: Media Access Control • The server software had algorithms to determine priority, level of access, and so on IEEE 802 Media Access Control • This could be rectified by using multiple servers or by reducing the server software so the station that was acting as the server could also be used as a station and, therefore, the responsibility for being server could be moved around • In practice, a single station or set of stations must have the management responsibility for LAN additions, deletions, security and so forth Industrial Token Passing • One of the other methods of access used in industrial networking is token passing • It is deterministic and operates efficiently even at heavy loads • However, it is a much slower protocol, even though it is deterministic, because all nodes are offered a chance to transmit, even if they have no need, and that wastes bandwidth Token-Passing Bus • The token-passing bus actually does what its name implies • The token will be passed to the next rider when the present token holder gets off, so that next person may ride the bus • Although most of the legacy IEEE 802.4 standard described a “medium” as a broadband coaxial cable with repeaters, it should be understood that token-passing buses use other media • The scheme described next is a generalized concept that a token-passing bus might use • One of the important parameters of a token-passing system is the token rotation time—the time it takes the token to make a round trip of the bus Token-Passing Bus • Some networks have a four-level priority scheme in effect; that is, starting at the highest and going to the lowest, a station transmits its high-priority packets first • If there is time left on the token-hold timer that corresponds to that priority, then the station can transmit its lower-priority packets • Without the priority scheme, all messages have the same high priority and all use the high-priority token timer • At intervals, a “solicit successor” frame is sent and allows one response frame • Stations whose address falls into this range can then respond with a set successor frame Logical Link Control • Together, the MAC and the Logical Link Control are responsible for placing and retrieving information without errors on the Physical layer • The specifications in IEEE 802.2 outline the LLC’s responsibilities, regardless of which media access/topology is employed, for any • The functional definition of the Data Link layer includes framing data blocks or packets , determining the check character, and determining network addresses Introduction to Types of Service • “Connection-oriented” service means service that is pointto-point oriented • On the other hand, “connectionless” service places addressed packets on the media without establishing a persistent data channel • Type 1: Connectionless service allows two LLCs to exchange data without establishing a persistent channel connection • This type of service essentially hands off end-to-end reliability to another layer, typically Layer 4 for TCP/IP or Layer 7 for many industrial applications • The single-frame basis means that the transmit station will keep its last transmitted frame until the called station acknowledges this frame Introduction to Types of Service • If the frame is not acknowledged the transmit station will retransmit the frame a set number of times and then assume that something is awry • In such an event, automatic or manual recovery methods are needed to determine the reason for the lack of acknowledgment • The software could also determine that—due to the lack of response—the called station was out of service and apply maintenance procedures • X frame, the 802.2 information is identical at the LLC interface Gigabit Ethernet Jumbo Frames • As the speed of Ethernet reached gigabit rates, many of the equipment manufacturers of switches wanted to see a change to allow an Ethernet frame to carry a much larger data payload • Although several variations have been proposed , there is general support for a “jumbo” frame that supports a roughly 9,000 byte data area • There is an EtherType code to designate a jumbo frame and all such frames are usually of fixed size , so no length value is actually required 802.1q Tagged Ethernet Frames • Another variation in the Ethernet frame was generated to add more functionality to Ethernet switches • Vendors wanted the ability to provide different classes of service/priority to some messages in order to minimize latency and delay through the Ethernet switch • Vendors also wanted to be able to provide a means for logically isolating message traffic to create a virtual local area network across interconnected switches • These changes required the addition of two data fields to the Ethernet frame • This data, a four-octet long IEEE 802.1q header, is inserted into the frame between the source MAC address and the Length/EtherType code • The first two octets are an EtherType code of 0x8100 that tells the receiving switch that this is a “tagged” frame 802.1q Tagged Ethernet Frames • In concept, IEEE 802.5 has a similar frame, although the number of octets and the use of a poll octet instead of a type/length octet are both different, and by the time the information reaches the LLC, it is the same interface • In the 16-Mbps version of 802.5, the frame length can be up to 17.1 K octets although, as mentioned, there are “jumbo” frames for 802.3 of 9 K octets for use with 1/10 Gb switches Non-Routing IP Addresses • By the mid-1990s, most corporations/organizations had implemented TCP/IP within their own networks, as their Layer 3 and 4 architecture • The IETF had to provide a means for allowing those networks to have IP addresses without using up the rapidly-dwindling supply of available addresses • The IETF designated several IPv4 address ranges as being non-routable and only for use on internal networks • The question then arises: If the Internet will not recognize or deliver messages to or from computers with these IP addresses, how come you can send and receive email from people all around the Internet? Network Address Translation • As was mentioned earlier, most private networks have one point of connectivity to the Internet at the default gateway device that routes internal traffic out to the Internet • The IETF proposed the idea of allowing internal users with non-routable IP addresses to “borrow” the valid IP address of the gateway device for sending and receiving messages over the Internet • In this example, the default gateway device has two network interfaces, one on the internal network with a non-routable IPv4 address and one on the Internet with a “real” IPv4 address Network Address Translation IPv6 • IPv6 brings several changes to the inter-networking protocol • The first change that we will address is in the datagram header • Of the 40 octets in the header, IPv6 uses 16 octets each for the source and the destination IP address • During the transition from IPv4 to IPv6—which may take several more years—devices will need to know which IP/header version they are to process • This allows IPv6 to offer enhanced delivery performance for traffic, such as streaming audio and video • Hop Limit , in practice, serves the same purpose as the Time to Live field does in IPv • Although TTL could have been measured in seconds, it was almost always measured in hops THE OSI LAYER 3 PROTOCOL IS CONSIDERABLY MORE DETAILED THAN THE PROTOCOLS DESCRIBED THUS FAR Another Layer 3 Scheme: OSIIP THE U.S. GOVERNMENT MADE A CONSIDERABLE PUSH TO ADOPT THE OSI PROTOCOLS , 1990) OSI ROUTING PROTOCOLS ARE WELL THOUGHT OUT AND CERTAINLY MUCH MORE ROBUST THAN EITHER IPV4 OR IPX STANDARDS; IS IS A ROUTER) LAN Layer 4 Overview • Layer 4 is concerned only with end users • In most industrial networks of the threelayer variety, protocols are either connection oriented or the application layer handles the packet sequencing and accountability • With the encroachment of public standards upon the proprietary industrial networks, the most used standards are the TCP/IP protocols AS WE’VE SEEN, TCP/IP, AN ACRONYM FOR TRANSMISSION CONTROL PROTOCOL/INTERNET PROTOCOL, WAS DEVELOPED BY THE DEPARTMENT OF DEFENSE’S ADVANCED RESEARCH PROJECTS AGENCY TO WORK ACROSS DISSIMILAR COMPUTER PLATFORMS, RANGING FROM MAINFRAMES TO MICROS IP INTERFACED TO HIGHER LAYERS USING TCP , TO THE LOWER LAYERS , AND TO PEER, GATEWAYTO-GATEWAY TRANSACTIONS TCP WAS DEVELOPED TO PROVIDE THIS CLASS OF SERVICE AND TO SUPPORT CONNECTION-ORIENTED COMMUNICATIONS TCP/IP Connection or Connectionless • The difference between a connection-oriented and connectionless service can seem confusing, however, there are good examples in real life • The postal service will do its best, but delivery is not guaranteed • TCP, a Transport Layer 4 protocol, supports several reliability mechanisms for ensuring successful message delivery, along with other higher layer functions • TCP/IP has found wide use and is widely adapted to LANs, particularly those with mixed operating system platforms • TCP/IP is built into all versions of UNIX, Linux, Windows, Mac OSX, Solaris, and other operating systems, including some embedded operating system versions • On LANs, TCP/IP is mostly used layered above Ethernet -type networks, although it is usable on any network Connection or Connectionless • The primary difference between OSI-compliant standards and TCP/IP types is that the OSI model encompasses and attempts to standardize all possible elements, so a manufacturer knows precisely what to put into a data communications product • The TCP/IP concept has made it possible for a range of new Internet applications to be developed • The Transport Control Protocol assumes that the lower layers only offer unreliable datagram service Port Numbers • This diagram represents how a TCP/IP frame information is acquired by Decapsulation—the opposite of the process that encapsulated the data within the layer information on the transmit side • An Ethernet Frame is received at the correct MAC address as long as there are no bit errors • In an IP-based network architecture, programs wanting to communicate will request a port number and be assigned one, either permanently or temporarily Port numbers range from 0 to 65,534 but fall into three groupings 49,152–65,534: Temporary, reusable port numbers • Note that most programs and services will accept messages on the same port, regardless of whether delivery is by TCP or UDP • Some applications and services require either TCP message delivery or UDP message delivery but not both • There is the OSI TP and Novell’s SPX • The OSI Transport layer provides five classes of transport, four of which require connection-mode network service • The congestion avoidance algorithm allows the Network layer to notify the Transport layer when congestion is detected, for which the Transport layer can reduce its outstanding packet balance • TCP has similar mechanisms for flow control and congestion relief 49,152–65,534: Temporary, reusable port numbers • TCP also has a mechanism called slow start, whereby messages are initially sent using the minimum segment size and then each successive message is doubled in size until the MSS is reached or the sender is notified of delivery failure • In this second case, the segment size is dropped back to the prior size and all further message segments will use that size unless subsequent delivery problems arise due to traffic congestion and the segment size has to be reduced again • As has been mentioned, TCP, being a connection-oriented protocol, establishes a persistent connection between the two end-point applications that are communicating THIS FUNCTION IS USUALLY INTEGRATED INTO THE OPERATING SYSTEM BUT MAY BE ACCESSIBLE FOR ENCRYPTION OR FOR OTHER FORMS OF SYNTAX CONVERSION LAN LAYER 7: APPLICATION LAN LAYER 7 IS WHERE THE DATA IS INTERFACED TO THE APPLICATION SOFTWARE ON ONE SIDE OF THE COMMERCIAL LAN WORLD IS THE UNIX CAMP, WHICH USES THE NETWORK FILE SYSTEM , AND ON THE OTHER SIDE IS THE WINDOWS CAMP, WHICH USES THE SERVER MESSAGING BLOCK LAN Layer 6: Presentation Summary • In this chapter, local area networks were discussed in concept from the vantage point of Layers 1 through • The reader should be aware that—beyond Layer 1—you need detailed knowledge of a particular protocol to achieve more than a conceptual understanding • In addition to 802.3, this chapter also discussed a number of other protocols, including the newer wireless ones under the 802 umbrella