Packet Transmission (Part III)

Local Area Networks

(LANs)

Identifying A Destination

• All stations on shared-media LAN receive all transmissions

• To allow sender to specify destination– Each station assigned unique number– Known as station’s address– Each frame contains address of identical

recipient

Ethernet Addressing

• Standardized by IEEE

• Each station assigned unique 48-bit address

• Address assigned when network interface card (NIC) manufactured

NIC Cards

Internal NIC

Ethernet Addressing

• Ethernet addresses are MAC addresses

• They are station (NIC) address at the MAC layer

• Usually expressed for humans in hexadecimal– Base 16: Each Symbol Represents 4 bits (for example,

0101=5 hex)

– Convert each four bits to a hex symbol between 0 and F

– Put dashes between each pair

– Example: A1-34-CD-7B-DF-89

Ethernet Address Recognition

• Each frame contains destination address

• All stations receive a transmission

• Station discards any frame addressed to another station

• Important: interface hardware, not software, checks address

Possible Destinations

• Packet can be sent to:– Single destination (unicast)– All stations on network (broadcast)– Subset of stations (multicast)

• Address used to distinguish

Advantages Of Address Alternatives

• Unicast– Efficient for interaction between two computers

• Broadcast– Efficient for transmitting to all computers

• Multicast– Efficient for transmitting to a subnet of

computers

Broadcast On Ethernet• All 1s address specifies broadcast

• Sender– Places broadcast address in frame– Transmits one copy on shared network– All stations receive copy

• Receiver– Station’s unicast address– The broadcast address

Multicast On Ethernet

• Half of address reserved for multicast

• Network interface card– Always accepts unicast and broadcast– Can accept zero or more multicast address

• Software– Determine multicast address to accept– Informs network interface card

Identifying Frame Contents

• Integer type field tells recipient the type of data being carried

• Two possibilities– Self-identifying or explicit type (hardware

records type)– Implicit type (application sending data must

handle type).

Conceptual Frame Format

• Header – Contains address and type information– layout fixed

• Payload– Contains data being sent

Illustration Of Ethernet Frame

• Sender places– Sender’s address in source– Recipient’s address in destination– Type of data in frame type– Cyclic redundancy check in CRC

Octet

• A collection of 8 bits

• Called a byte in computer science

• The term “Octet” is widely used in networking

Octet = group of 8 bits

Ethernet Frame

Preamble (7 Octets)10101010 …

Begins synchronization

Start of Frame Delimiter (1 Octet)10101011

Ends synchronizationSignals start of

content of frame

Field Description

1010101010101010101010101010101010…Synchronizes the receiver’s clock with the sender’s clock

Ethernet Frame

Destination Address (6 Octets)

48 BitsExpressed for humans in

HexadecimalExample: A1-34-CD-7B-DF-34

Unique value set at factory

Field Description

Source Address (6 Octets)

48 BitsExpressed for humans in

HexadecimalExample: C7-66-3B-AC-89-00

Unique value set at factory

Ethernet Frame

Field Description

Length (2 Octets)Length of Data Field in octets

1,500 (decimal) maximum

Data (variable) LLC Frame

PAD

(Needed if Data Field is lessthan 46 octet minimum; Data

Field plus Pad will be 46 Octets)

Ethernet Frame

Field Description

CRC(4 Octets)

Error detection field;If error is detected, frame is

discarded

The CRC field is for error checking

Both sender and receiver compute it from the other bits in the frame.

If the two computed values are different, there has been an error.

In case of error, the receiver discards the frame; there is no request for retransmission.

Example Ethernet Types

When Network HardwareDoes Not Include Types

• Sending and receiving computers must agree– To only send one type of data– To put type information in first few octets of

payload

• Most system need type information

Illustration Of Type Information Added to Data

• In practice – Type information small compared to data

carried– Format of type information standardized

A Standard For Type Information

• Defined by IEEE

• Used when hardware does not include type field

• Called LLC/SNAP header

Demultiplexing On Type

• Network interface hardware– Received copy of each transmitted frame– Examines address and either discards or accepts– Passes accepted frames to system software

• Network device software– Examines frame type– Passes frames to correct software module

Ring Topology

• Second most popular LAN topology

• Bits flow in single direction

• Several technologies exist

• Used with ring topology

• Guarantees fair access: IEEE 802.5 standards

• Token– Special (reserved) message– Small (a few bits)

• Frame formats:

Token Passing

token: SD AC FC

packet: SD AC FC dest addr src addr data checksum ED FS

Frame Formats

• SD, ED: mark start, end of frame• AC: Access Control Byte:

token bit: value 0 means token can be seized, value 1 means data packet to follow access control bytepriority bits: priority level of this packetreservation bits: station can write these bits to prevent stations with lower priority packet from seizing the token next time it becomes free

• FC: Frame Control: used for monitoring and maintenance, distinguish control frames and information frames

Frame Formats (cont’d)

• source and destination address: as in Ethernet

data: packet from network layer

checksum: error detection

FS: Frame Status: set by destination, read by receiver

set to indicate that destination is up, packet copied OK from ring

Token Passing Paradigm• Station

– Waits for token to arrive– Transmits one packet around ring– Transmits token around ring

• When no station has data to send– Token circulates continuously

Token Passing Ring Transmission

• Station waits for token before sending

• Signal travels around entire ring

• Sender receives its own transmission

Token Release Mechanisms

• Release After Reception (RAR): – Each station reissues the free token only after it

receives the transmitted frame.

– Used on lower speed token rings ( <= 4Mbps).

• Release After Transmission (RAT):– Each station attaches a free token at the end of its frame

– Possible multiple frames propagate in a ring

– Used on higher speed token rings (>= 16Mbps)

Strengths Of Token Ring Approach

• Easy detection of– Broken ring– Hardware failures– Interference

Weaknesses Of Token Ring Approach

• Broken wire disables entire ring

• Point-to-point wiring– Awkward in office environment– Difficult to add/remove stations

Token Passing Ring Technologies

• IBM Token Ring– Originally operated at 4 Mbps– Later version operated at 16 Mbps

• Fiber Distributed Data Interface (FDDI)– Operates at 100 Mbps

FDDI• Fiber Distributed Data Interface

• A high-speed (100 Mbps) LAN (token ring) technology

• Suitable also for MAN

• It uses fiber as its transmission medium

• Can attach 1000 stations, can be up to 200 km

• Based on two counter-rotating (independent) fiber rings. Only one is used

• Maximum frame size = 4500 bytes

FDDI (cont’d)

• Using RAT mechanisms• Modes of data transfer

– Support synchronous (periodic) traffic: • voice and video• Highest priority• Useless if it is late• Allocated a maximum amount of time to transmit

sync data at each station

– Support asynchronous data: normal data

FDDI Terminology

• FDDI– Uses optical fibers

– High reliability

– Immune to interference

• CDDI– FDDI over copper

– Same frame format

– Same data rate

– Less noise immunity

FDDI Hub Technology

• Part of FDDI standard

• Stations attach to hub

• Same frame format and data rate as FDDI

• Called star-shaped ring

FDDI Failure Recovery• Uses two rings. Automatic failure recovery• Terminology

– DAS: Dual Attachment Stations– Concentrator: allow for multiple stations on the ring– SAS: Single Attachment Stations

• The two counter-rotating provide fault-tolerance, i.e., self-healing ring

FDDI Failure Recovery (cont’d)

FDDI Timers and Data Transmission

• THT (Token Holding Time): upper bound on how long a station can hold the token

• TRT (Token Rotation Time): interval between two successive arrivals of the token– TRT = ActiveNodes THT + RingLatency

• TTRT (Target Token Rotation Time): the maximum limit that a given station can hold the token (could differ between stations)

Timed Token Protocol

• Each station measures its TRT

• Upon receiving the Token, transmits any waiting sync data (up to some limit)

• If measured TRT > TTRT, token is too late, station can not transmit

• If TRT < TTRT, token is early, station can send data up to THT (THT < TTRT-TRT).

Timed Token Protocol

• See example in class