Transcript Chapter 1 - Local Area Network Technologies

Chapter 1 - Local Area Network
Technologies
How IP Datagrams are
Encapsulated
• IP datagrams are found at the OSI Network
layer
• IP datagrams are sent to the DataLink Layer
prior to being sent out the physical medium
• See Diagram on handout
Items Included in Data Link
Layer Header and Trailer
• Delimitation – used to distinguish frames in the
Data Link Layer from each other. Both a start and
end of frame is used.
• Protocol Identification – identifies what particular
protocol is being used (TCP/IP, IPX, etc.)
• Addressing – source node and destination node
• Bit-level integrity check – to detect bit-level errors
in the frame. A checksum is used. The checksum is
computed by source node and included in frame
header or trailer. The destination recalculates
checksum and compares result to included
checksum.
LAN Encapsulation
• The way a network encapsulates data to be
transmitted is called the “frame format”
• The frame format corresponds to information
placed on the frame by the Data Link Layer
• The frame format consists of a header and trailer
• All nodes on the same network segment bounded
by routers must use the same frame format
• Each network type (Ethernet, Token Ring, FDDI)
uses a different frame format
Ethernet
• Began as a radio transmission system developed at
the University of Hawaii called ALOHA
• All transmissions share a common channel
• Stations contend for access to the channel in order
to transmit
• Uses CSMA/CD- transmitter listens to channel to
determine when to send
• 1979 – DEC, Intel, and Xerox created industry
standard 10Mbps Ethernet (known as Ethernet II)
• 1981 – IEEE formed the 802.3 standard to make
10 Mbps an international standard
Ethernet – Continued
• 1995 – IEEE approved a 100 Mbps version of
Ethernet – referred to as Fast Ethernet – uses CAT
5 twisted pair cable
• Ethernet existed before the 802.3 standard was
developed. This has resulted in multiple Ethernet
standards.
• There are thus multiple ways of encapsulating data
to be transmitted on an Ethernet network.
• This situation can result in two hosts on an
Ethernet network segment unable to communicate
with each other
Ethernet – Continued
• 10/100 Ethernet – dual-speed Ethernet –
especially helpful for transitions from 10
Mbps to 100 Mbps technology
• Gigabit Ethernet – operates at 1000 Mbps
Ethernet Frame Formats
• IP datagrams use either Ethernet II (10
Mbps) or IEEE 802.3 SubNetwork Access
Protocol (SNAP) encapsulation
Ethernet II
• Created by DEC, Intel, and Xerox before 802.3
specification
• Also known as Digital Intel Xerox (DIX) frame
format
• Frame format contains:
–
–
–
–
–
–
Preamble
Destination Address
Source Address
EtherType
Payload
Frame Check sequence
Ethernet II Preamble
• 8 bytes long
• 7 bytes of alternating 1s and 0s (10101010)
to synchronize a receiving station
• 1 byte – 10101011 to indicate the start of
the frame
• The Preamble field is not visible with
Network Monitor
Ethernet II Destination and
Source Address
• Destination address is 6 bytes long
• Destination can be unicast, multicast, or broadcast
• The unicast address is the physical MAC (Media
Access Control) address
• The broadcast address is all 1s (FF-FF-FF-FF-FFFF)
• Source address is 6 bytes long and indicates
senders unicast address
Ethernet II EtherType
• 2 bytes long
• Indicates the upper layer protocol contained
within the Ethernet frame.
• Acts as an identifier for the Ethernet II
frame format.
• For IP datagram, this field is set to 0x0800
Ethernet II Payload
• Consists of a protocol data unit (PDU) of an
upper layer protocol
• Maximum size of 1500 bytes
• Minimum size of 46 bytes – if the PDU is
smaller than this, it is padded so that it is at
least this size
Ethernet II Frame Check
Sequence
•
•
•
•
4 bytes long
Provides bit-level error detection
Also called CRC – Cyclic Redundancy Check
Source node calculates Frame Check Sequence
bits and places result in the appropriate field
• Destination re-calculates FCS bits and checks
these with the bits sent
• If the two values match, the destination node
processes the frame
• If the two values don’t match, the frame is silently
discarded
Frame Check Sequence – Cont.
• FCS is obtained by dividing by a 33-bit prime
number
• This prime number is divided into the number
consisting of the bits in the frame (except for the
Preamble and the FCS fields)
• The remainder of this division is placed in the
FCS field (32 bits)
• This process can detect 100 percent of all singlebit errors
• It is highly improbable that random noise would
damage the frame and not be detected by this
Frame Check Sequence – Cont.
• FCS does not provide security
• An intermediate node could have intercepted the
frame, altered the contents, and calculated a new
FCS
• Use IP Security to detect the above (Chapter 20)
• This process does not indicate where the error is
located or how to correct it
• Other types of CRC calculations provide
additional information
• The Checksum field in the ATM header provides
error detection and limited error recovery
See P. 7 for Example of Ethernet
II Frame Format from Network
Monitor
Ethernet Interframe Gap
• The end of the Ethernet frame is not explicitily
indicated
• An implied postamble is used that consists of a
gap between each frame
• This gap is called the Ethernet interframe gap
• This gap is a specified measure of time – that
needed to send 96 bits of data
• It the wire goes silent for 96 bit times, the last bit
in the received frame occurred 96 bit times ago.
Ethernet Minimum Frame Size
• All frames must be at least 46 bytes in length
• This is because of how the collision detection
scheme works
• To detect a collision, the transmission must be
long enough for the signal indicating the collision
to be propagated back to the sending node
• See Figure 1-2 of a 10Base5 Ethernet network that
obeys the 5-4-3 rule:
– Maximum of 5 physical segments
– Maximum of four repeaters between any two nodes
– Maximum of at most three segments populated
Ethernet Minimum Frame Size
• Propagation time – time for signal to propagate
from one end of network to the other
• Propagation delay for this maximum extent
Ethernet is 28.8 microsec.
• Slot time – time for signal to make a round trip =
28.8 * 2 = 57.6 microsec.
• In 57.6 microsec, at a transmission rate of
10Mbps, you will transmit 57.6 * 10 = 576 bits
• Thus the entire frame, including the Preamble
filed must be a minimum of 576 bits (72 bytes)
• Subtract 8+6+6+2+4 = 26 yields 46 bytes
IEEE 802.3 Ethernet Frame
Format
• Result of the IEEE 802.2 and 802.3
specifications
• Contains an 802.3 header and trailer as well
as an 802.2 header
• See Figure 1-3 for specific fields
IEEE 802.3 Header
• Preamble – 7 bytes long consisting of alternating
1s and 0s. Used to synchronize receiving station.
• Start Delimiter – 10101011 – Indicates start of
frame
• Destination & Source Address – same
• Length – 2 bytes that indicate the number of bytes
from the LLC header’s first byte to the payload’s
last byte – minimum of 46 and maximum of 1500
IEEE 802.2 LLC Header
• Destination Service Access Point (DSAP)indicated the destination upper layer
protocol for the frame
• Source Service Access Point (SSAP)indicates the source upper layer protocol
Similarities between Ethernet II
and IEEE 802.3 Frame Formats
• II Preamble is identical to IEEE Preamble
and Start Delimiter fields
• Source Address and Destination Address are
very similar
• FCS is identical
How are the Two Frame Formats
Differentiated?
• To differentiate the two, examine the first 2
bytes past the Source Address field.
• If this field is > 1500, then this is an
Ethernet II frame format
• If this field is <= 1500, then it is a length
field and an 802.3 frame
IEEE 802.3 SNAP
• Was created as an extension of IEEE 802.3
to allow protocols that were designed for
Ethernet II to be used in an IEEE 802.3
compliant environment
What Frame Format does
Windows 2000 Use?
• The default frame format for Windows 2000
is the Ethernet II format
• Windows 2000 will receive both types of
frame format
• You may instruct windows 2000 to send
IEEE 802.3 SNAP-encapsulated frames
Details of the MAC Address
• Individual/Group Bit – indicates whether
the address is individual (0) or group (1)
• Universal/Locally Administered Bit –
indicates whether address is allocated by
IEEE (0) or locally administered (1)
Homework
• Obtain a listing of the defined EtherType
fields allowed from the web site listed in
our text
• Complete reading Chapter 1 of text