1. www.ciscopress.com Networking Basics CCNA 1 Chapter 6. Ethernet Fundamentals

2. www.ciscopress.com Ethernet Fundamentals The History of Ethernet Created by Robert Metcalfe and others at Xerox in early-to-mid 1970s They later teamed with Digital Equipment Corporation (DEC) and Intel; published the first Ethernet standard known as the DIX (DEC, Intel, Xerox) standard. Version 2 of the standards, DIXv2, was published in 1980 and became 10BASE5. And DIX Ethernet is now known as Ethernet II. The first products developed using the Ethernet standard were sold during the early 1980s, transmitting at 10Mbps over thick coaxial cable up to 2km.

3. www.ciscopress.com Ethernet Fundamentals The History of Ethernet IEEE took over the development of these standards in the early 1980s The first committee (committee 802) was given the task of developing IEEE standards for LANs The 802.3 subcommittee worked on Ethernet standards (802.3 is now the official IEEE Ethernet standard). The 802.2 subcommittee worked on several types of LANS, including Ethernet and Token Ring

4. www.ciscopress.com Ethernet Fundamentals The History of Ethernet Ethernet standards include: –10BASE5 (1980) –10BASE2 (1985) –10BASE-T (1990) –Fast Ethernet – 100 Mbps (1995) –Gigabit Ethernet – 1 Gbps (1998) –10 Gigabit Ethernet – 10 Gbps (2002)

5. www.ciscopress.com Ethernet Fundamentals The Names of Different IEEE Ethernet Types Ethernet standards differ in two main respects: –Speed –Type of cabling supported

6. www.ciscopress.com Ethernet Fundamentals The Names of Different IEEE Ethernet Types Commonly used Ethernet names list the basic differentiating features: –Speed – the speed is listed in Mbps before the word “BASE” –Baseband transmission – all current Ethernet standards use baseband transmission –Cabling – the text after the word “BASE” indicates the type of cabling; e.g. “T” means “twisted-pair”

7. www.ciscopress.com Ethernet Fundamentals The Names of Different IEEE Ethernet Types Commonly used Ethernet names list the basic differentiating features: –Example: 10BASE-T 10 Mbps Baseband transmission Twisted-pair cabling

8. www.ciscopress.com Ethernet Fundamentals The Names of Different IEEE Ethernet Types Baseband versus broadband –Baseband means that a single frequency is used to encode bits –Some earlier (now obsolete) Ethernet technologies used broadband to send signals 10BROAD36 Broadband is a range of frequencies

9. www.ciscopress.com Ethernet Fundamentals The History of Ethernet The underlying Ethernet technology remain unchanged. Each new type of Ethernet supports the same basic Ethernet frame and address scheme. The 802.3 family of Ethernet is all compatible. By using common framing, Ethernet has remained simple and successful – it is very scalable.

10. www.ciscopress.com Ethernet Fundamentals

11. www.ciscopress.com Ethernet Fundamentals Sample Campus LAN with Different Types of Ethernet The Scalability of the Internet

12. www.ciscopress.com Ethernet Fundamentals The History of Ethernet Reasons for Ethernet’s continued success: –Relatively simple –Able to easily incorporate new technologies (same frame structure) –Ethernet is reliable, with well-tested components and protocols –Ethernet is inexpensive; new types generally experience rapid price reductions within a few years of introduction

13. www.ciscopress.com Ethernet Fundamentals IEEE Ethernet Standards and the OSI Model IEEE 802.3 Standards concern several details: –Physical transmission details (cables, connectors, encoding, speeds) –Media access issues (carrier sense multiple access with collision detection – CSMA/CD) –Errors during transmission –MAC address (location and format) –Ability of NICs to synchronize to the incoming signal by using a preamble and Start Frame Delimiter (SFD)

14. www.ciscopress.com Ethernet Fundamentals IEEE Ethernet Standards and the OSI Model Since Ethernet is comprised of standards at the lower layers (Data Link layer and Physical layer), generalizing, it may best be understood in reference to the OSI model. In the IEEE Ethernet standard, the Data Link layer is further divided into two sublayers - Logical Link Control (LLC) sublayer and the Media Access Control (MAC) sublayer. The use of these sublayers contributes significantly to compatibility between diverse end devices as LLC sublayer is media independent.

15. www.ciscopress.com Ethernet Fundamentals The Logical Link Control Sublayer (802.2) Makes the connection with the upper layers Frames the Network layer packet Identifies the Network layer protocol Remains relatively independent of the physical equipment

16. www.ciscopress.com Ethernet Fundamentals The Media Access Control Sublayer (802.3) Data Encapsulation –Frame delimiting –Addressing –Error detection Media Access Control –Control of frame placement on and off the media –Recovery transmission failure due to collisions

17. www.ciscopress.com Ethernet Fundamentals Comparing Ethernet Standards to the OSI Model

18. www.ciscopress.com Ethernet Fundamentals Comparing Ethernet Standards to the OSI Model The IEEE 802.3 standard defines many physical details, as well as the lower half of the data link layer - Media Access Control (MAC) LCC remains relatively independent of the physical equipment and allows multiple Layer 3 protocols.

19. www.ciscopress.com Ethernet Fundamentals

20. www.ciscopress.com Ethernet Fundamentals Ethernet Framing Framing refers to two things: –The process of encapsulating data inside a header and possible a trailer –The meaning given to the bits inside those headers and trailers Framing defines the meaning of transmitted bits Mailing a letter –Write the letter –Put in envelope –Write address on envelope in a particular place –Put the stamp in a particular place

21. www.ciscopress.com Ethernet Fundamentals Ethernet Framing Framing is similar to mailing a letter –The NIC adds a header and trailer around the Layer 3 protocol data unit (PDU) to encapsulate the data –The header and trailer must have properly formatted information to be delivered correctly

22. www.ciscopress.com Ethernet Fundamentals Encapsulating Packets Inside Ethernet Frames The process of adding a header and trailer around the Layer 3 PDU is called “encapsulation” Sometimes called “framing” The term “frame” refers to the resulting bits that include the Ethernet header and trailer

23. www.ciscopress.com Ethernet Fundamentals Ethernet Encapsulation and the IEEE 802.3 Frame

24. www.ciscopress.com Ethernet Fundamentals The Encapsulation Process 1.The IP software gives the IP packet to the Ethernet software 2.The Ethernet software encapsulates the IP packet between and Ethernet header and trailer 3.The Ethernet NIC physically transmits the bits that comprise the frame over an Ethernet LAN

25. www.ciscopress.com Ethernet Fundamentals Three Styles of Ethernet Framing

26. www.ciscopress.com Ethernet Fundamentals The Fields in the IEEE 802.3 Frame Header and trailer contain a defined number of bytes Bytes are grouped together to form fields Different sets of bytes have different meanings

27. www.ciscopress.com Ethernet Fundamentals Three Styles of Ethernet Framing The differences between framing styles are minimal. The most significant difference between the two standards are the addition of a Start Frame Delimiter (SFD) and the change of the Type field to a Length field in the 802.3.

28. www.ciscopress.com Ethernet Fundamentals IEEE 802.3 Header

29. www.ciscopress.com Ethernet Fundamentals IEEE 802.3 Header

30. www.ciscopress.com Ethernet Fundamentals The Type/Length Field in the IEEE 802.3 Header If the two-octet value is equal to or greater than 600 hex, the frame is interpreted according to the DIX type code. Otherwise, the frame is interpreted as a 802.3 frame and the length of the frame is indicated. The early DIX Ethernet (Ethernet II) frame format is the most commonly used frame type with TCP/IP- based Ethernet LANs.

31. www.ciscopress.com Ethernet Fundamentals The Length of an Ethernet Frame Both the Ethernet II and IEEE 802.3 standards define the minimum frame size as 64 bytes and the maximum as 1518 bytes. The Preamble and Start Frame Delimiter fields are not included when describing the size of a frame. If the size of a transmitted frame is less than the minimum or greater than the maximum, the receiving device drops the frame.

32. www.ciscopress.com Ethernet Fundamentals The Preamble and SFD Fields Ethernet with throughput speeds of 10 Mbps and slower are asynchronous. Asynchronous communication uses the 8 bytes of preamble (timing information) to synchronize the receiver. 100 Mbps and higher are synchronous, which means timing information is not required. However, for compatibility reasons, the Preamble and Start Frame Delimiter (SFD) fields are still present.

33. www.ciscopress.com Ethernet Fundamentals The Format of MAC Addresses Media Access Control address characteristics: –6-byte hexadecimal (hex) numbers –Each NIC has a burned-in (permanent) MAC address –First 3 bytes (6 hex digits) are called an Organizationally Unique Identifier (OUI) and refer to the manufacturer –Second 3 bytes are unique numbers to the manufacturer (a serial number)

34. www.ciscopress.com Ethernet Fundamentals Structure of a MAC Address

35. www.ciscopress.com Ethernet Fundamentals Media Access Control MAC refers to protocols that determine which computer on a shared medium environment (collision domain) is allowed to transmit the data. Two categories of MAC protocols: –Deterministic (taking turns): Token Passing –Nondeterministic (first come, first serve): CSMA/CD

36. www.ciscopress.com Ethernet Operation Token Passing in Token Ring

37. www.ciscopress.com Ethernet Fundamentals Media Access Control for the Three Topological Implementations Ethernet – Logical bus topology (information flow on a linear bus) and physical star or extended star (wired as a star) Token Ring – Logical ring topology (information flow controlled in a ring using a token) and physical star or extended star (wired as a star) FDDI (Fiber Distributed Data Interchange) – Logical ring topology and physical dual-ring topology

38. www.ciscopress.com Ethernet Operation CSMA/CD (Carrier Sense multiple Access / Collision Detect) CSMA/CD algorithm defines rules that NICs must follow In the original Ethernet physical bus topology, only one device can send at a time If multiple devices try to send at the same time, their electrical signals overlap and are added together The resulting signal does not represent the original signal of either device Such an occurrence is called a “collision” How to detect collision?

39. www.ciscopress.com Ethernet Operation CSMA/CD – the Algorithm The original MAC standard suggests: –Wait until the LAN is unused, then send the frame (listen before transmit – carrier sense) –Listen to detect if the frame collided with another frame (collision detect while transmitting) –If no collision occurred, the frame must have made it across the LAN –If a collision did occur, wait and then try to send the original frame again

40. www.ciscopress.com Ethernet Operation CSMA/CD – the Algorithm The CSMA/CD algorithm in detail: –Wait until the LAN is silent –Send a frame –Transmitting device listens for collisions while transmitting –If no collisions occur, process is complete –All devices whose transmitted frames collided send a 32- bit jamming signal to invoke a back-off algorithm –All devices stop transmitting for a random amount of time to allow the collision signals to subside. –Once the timer expires, a device can begin the process again

41. www.ciscopress.com Ethernet Operation Origins of the Name CSMA/CD

42. www.ciscopress.com Ethernet Operation Collision Domains The connected devices that access a common media via a hub or series of directly connected hubs make up what is known as a collision domain. A collision domain is also referred to as a network segment. Hubs and repeaters therefore have the effect of increasing the size of the collision domain.

43. www.ciscopress.com Ethernet Operation

44. www.ciscopress.com Ethernet Operation Full Duplex, Half Duplex, and Collision Domains CSMA/CD causes throughput over a LAN to degrade as the LAN gets busier Collisions cause devices to wait to retransmit, further slowing the network Originally, Ethernet LAN performance was restricted because Ethernet NICs used half duplex logic – they could send or transmit (duplex), but not do both simultaneously To improve LAN performance, standards evolved where collisions cannot occur: full duplex (receive and transmit simultaneously)

45. www.ciscopress.com Ethernet Operation Preventing Collisions with Switch Buffering Switches prevent collisions by buffering frames If several PCs send frames to the same address at the same time, the switch holds the frames in memory - a process called buffering The switch then forwards the frames one at a time

46. www.ciscopress.com Ethernet Operation Switch Buffering Example

47. www.ciscopress.com Ethernet Operation Collision Domains with a Switch

48. www.ciscopress.com Ethernet Operation Collision Domains and Full Duplex Each port of a switch creates its own collision domain A frame in one collision domain does not cause collisions in another collision domain The switch’s buffering process prevents the collisions When full duplex is enabled: –NICs/Interfaces can send and receive at same time –CSMA/CD is no longer needed as collisions can’t occur –Because CSMA/CD is disabled, the NIC disables its loopback logic –Full duplex cannot be used if a hub is plugged into the port

49. www.ciscopress.com Ethernet Operation Switches and Hubs: Case That Does Not Allow Full Duplex

50. www.ciscopress.com Ethernet Operation Autonegotiation of Duplex and Speed Ethernet supports a wide variety of speeds and two duplex options An IEEE standard sets the process by which Ethernet NICs and switch ports can automatically negotiate the speed and duplex setting – called “autonegotiation” NICs and switch ports exchange information about their capabilities, using Fast Link Pulse (FLP) bursts [a series of Network Link Pulses (NLPs) that allow the devices to send each other a set of bits] These bits do not look like normal frames, so they are not processed like normal frames

51. www.ciscopress.com Ethernet Operation Autonegotiation of Duplex and Speed The preferred choices of autonegotiation speed from fastest to slowest –1000 Mbps, full duplex –1000 Mbps, half duplex –100 Mbps, full duplex –100 Mbps, half duplex –10 Mbps, full duplex –10 Mbps, half duplex

52. www.ciscopress.com Ethernet Operation Autonegotiation of Duplex and Speed Speed and duplex may be manually configured or automatically negotiated If one end disables autonegotiation and the other end uses a different speed the link will not work If one end disables autonegotiation and the other end uses a different duplex, the connection may work, but with excessive collisions Autonegotiation works well on copper cabling, but is not supported on fiber

53. www.ciscopress.com Ethernet Operation Ethernet – Physical Layer 10 Mbps Ethernet - 10BASE-T using  Cheap and easy-to-install Cat3/Cat5 unshielded twisted-pair (UTP) cable (2 pairs, full-duplex allowed).  Manchester-encoding  physical star topology  up to 100 meters before requiring a hub or repeater.  8-pin RJ-45 connector (pin 1/2 for Tx, 3/6 for Rx, other pins unused)

54. www.ciscopress.com Ethernet Operation Ethernet – Physical Layer 100 Mbps (Fast) Ethernet – 100BASE-TX using  Cat5 or later UTP (2 pairs of wires)  4B/5B encoding  physical star topology  Switch  up to 100 meters

55. www.ciscopress.com Ethernet Operation Ethernet – Physical Layer 100 Mbps (Fast) Ethernet – 100BASE-FX using  fiber-optic cable  physical point-to-point topology  4B/5B encoding  Up to 400 meters  May use switch

56. www.ciscopress.com Ethernet Operation Ethernet – Physical Layer Advantages of optical fibre over UTP small physical size noise immunity increased unrepeated distances But more expensive and harder to install

57. www.ciscopress.com Ethernet Operation Ethernet – Physical Layer 1000 Mbps (Gigabit) Ethernet – 1000BASE-T using –all four pairs in Category 5 or later UTP cable. –4D-PAM5 line encoding to obtain 1 Gbps data throughput. –Up to 100 meters

58. www.ciscopress.com Ethernet Operation Ethernet – Physical Layer 1000 Mbps (Gigabit) Ethernet – 1000BASE-SX and 1000BASE-LX using –two strands of optical fiber. –8B/10B encoding. –-SX: short laser wavelength for up to 550 meters –-LX: long wavelength for up to 10,000 meters.

59. www.ciscopress.com Summary The Ethernet family includes 10 Mbps, Fast Ethernet and Gigabit Ethernet Any new upgrades to Ethernet supplement the IEEE 802.3 standard Ethernet uses baseband signaling IEEE 802.3 operates at the bottom two layers of the OSI model –Lower half of the data link layer: the MAC sublayer –The physical layer

60. www.ciscopress.com Summary Ethernet at Layer 1 involves interfacing with media, signals, bit streams that travel on the media, components and various physical topologies All types of IEEE-standard Ethernet use the same frame structure –Preamble –Start Frame Delimiter (SFD) –Destination Address –Source Address –Length or Type –Data and Pad –Frame Check Sequence

61. www.ciscopress.com Summary Address fields of the Ethernet frame contain Layer 2, or MAC addresses, for source and destination All frames are susceptible to errors – the FCS field contains a number that is calculated by the source – the destination performs the same calculation to determine if the frame is good Ethernet uses the CSMA/CD algorithm to manage access to shared media – it is a nondeterministic protocol

62. www.ciscopress.com Summary Other MAC algorithms, such as token-passing methods used by Token Ring and FDDI, are deterministic With CSMA/CD, a NIC waits for the absence of a signal on the media and the begins transmitting If two or more nodes transmit at the same time, a collision occurs If nodes detect a collision, they wait a random time and then retransmit Autonegotiation detects speed and duplex mode (half or full duplex) of the device on the other end of the wire and adjusts to match the settings