20_otherlink

Download Report

Transcript 20_otherlink

20:
Other Technologies used at the
Link Layer
Last Modified:
4/6/2016 8:18:16 AM
5: DataLink Layer
5a-1
Token Passing: IEEE802.5 standard
 4 Mbps
 max token holding time: 10 ms, limiting frame length
 SD, ED mark start, end of packet
 AC: access control byte:
 token bit: value 0 means token can be seized, value 1 means
data follows FC
 priority bits: priority of packet
 reservation bits: station can write these bits to prevent
stations with lower priority packet from seizing token
after token becomes free
5: DataLink Layer
5a-2
Token Passing: IEEE802.5 standard
 FC: frame control used for monitoring and
maintenance
 source, destination address: 48 bit physical
address, as in Ethernet
 data: packet from network layer; checksum: CRC
 FS: frame status: set by dest., read by sender

set to indicate destination up, frame copied OK from ring
 limited number of stations: 802.5 have token
passing delays at each station
5: DataLink Layer
5a-3
Point to Point Data Link Control
 one sender, one receiver, one link: easier
than broadcast link:
 no need for explicit MAC addressing
 full-duplex simultaneous bi-directional
operation = no need for media access
control
 e.g., dialup link, ISDN line
 popular point-to-point protocols:
 PPP (point-to-point protocol)
 HDLC: High level data link control
5: DataLink Layer
5a-4
PPP Design/Features
 packet framing: encapsulation of network-layer




datagram in data link frame
 carry network layer data of any network layer
protocol (not just IP) at same time
 ability to demultiplex upwards
bit transparency: must carry any bit pattern in the
data field
error detection (no correction)
connection liveness: detect, signal link failure to
network layer
network layer address negotiation: endpoint can
learn/configure each other’s network address
5: DataLink Layer
5a-5
PPP non-requirements
 no error correction/recovery
 no flow control
 no need to support multipoint links (e.g.,
polling)
Error recovery, flow control, data re-ordering
all relegated to higher layers!|
5: DataLink Layer
5a-6
PPP Data Frame
 Flag: delimiter (framing)
 Address: does nothing (only one option)
 Control: does nothing; in the future
possible multiple control fields
 Protocol: upper layer protocol to which
frame delivered (eg. IP, PPP-LCP, IPCP, etc)
5: DataLink Layer
5a-7
PPP Data Frame
 info: upper layer data being carried
 check: cyclic redundancy check for error
detection
5: DataLink Layer
5a-8
Byte Stuffing
 “data transparency” requirement: data field must
be allowed to include flag pattern <01111110>
 Q: is received <01111110> data or flag?
 Sender: adds (“stuffs”) extra < 01111110> byte
after each < 01111110> data byte
 Receiver:
 two 01111110 bytes in a row: discard first byte,
continue data reception
 single 01111110: flag byte
5: DataLink Layer
5a-9
Byte Stuffing
flag byte
pattern
in data
to send
flag byte pattern plus
stuffed byte in
transmitted data
5: DataLink Layer 5a-10
PPP Data Control Protocol
Before exchanging networklayer data, data link peers
must
 configure PPP link (max.
frame length,
authentication)
 learn/configure network
layer information
 for IP: carry IP Control
Protocol (IPCP) msgs
(protocol field: 8021) to
configure/learn IP
address
5: DataLink Layer 5a-11
IP over Other Wide Area
Network Technologies
 ATM
 Frame Relay
 X-25
5: DataLink Layer 5a-12
ATM architecture
 Adaptation layer (AAL): only at edge of ATM
network
 roughly analogous to Internet transport layer
 ATM layer: “network” layer
 Virutal circuits, routing, cell switching
 physical layer
5: DataLink Layer 5a-13
ATM Layer: ATM cell
 5-byte ATM cell header
 48-byte payload
Why?: small payload -> short cell-creation delay
for digitized voice
 halfway between 32 and 64 (compromise!)

Cell header
Cell format
5: DataLink Layer 5a-14
ATM cell header
 VCI: virtual channel ID
 will
change from link to link thru net
 PT: Payload type (e.g. RM cell versus data
cell)
 CLP: Cell Loss Priority bit
 CLP = 1 implies low priority cell, can be
discarded if congestion
 HEC: Header Error Checksum
 cyclic redundancy check
5: DataLink Layer 5a-15
ATM: network or link layer?
Vision: end-to-end
transport: “ATM from
desktop to desktop”
 ATM is a network
technology
Reality: used to connect
IP backbone routers
 “IP over ATM”
 ATM as switched
link layer,
connecting IP
routers
5: DataLink Layer 5a-16
Datagram Journey in IP-overATM Network
 at Source Host:
IP layer finds mapping between IP, ATM dest address
 passes datagram to AAL5
 AAL5 encapsulates data, segments to cells, passes to
ATM layer
 ATM network: moves cell along VC to destination (uses
existing one or establishes another)
 at Destination Host:
 AAL5 reassembles cells into original datagram
 if CRC OK, datgram is passed to IP

5: DataLink Layer 5a-17
X.25 and Frame Relay
Like ATM:
 wide area network technologies
 virtual circuit oriented
 origins in telephony world
 can be used to carry IP datagrams and can
thus be viewed as Link Layers by IP
protocol just like ATM
5: DataLink Layer 5a-18
X.25
 X.25 builds VC between source and
destination for each user connection
 Per-hop control along path
 error control (with retransmissions) on
each hop
 per-hop flow control using credits
• congestion arising at intermediate
node propagates to previous node on
path
• back to source via back pressure
5: DataLink Layer 5a-19
IP versus X.25
 X.25: reliable in-sequence end-end
delivery from end-to-end
 “intelligence
in the network”
 IP: unreliable, out-of-sequence end-
end delivery
 “intelligence
in the endpoints”
 2000: IP wins
 gigabit routers: limited processing
possible
5: DataLink Layer 5a-20
Frame Relay
 Designed in late ‘80s, widely deployed in
the ‘90s
 Frame relay service:
 no error control
 end-to-end congestion control
5: DataLink Layer 5a-21
Frame Relay (more)
 Designed to interconnect corporate customer LANs
typically permanent VC’s: “pipe” carrying aggregate
traffic between two routers
 switched VC’s: as in ATM
 corporate customer leases FR service from public
Frame Relay network (eg, Sprint, ATT)

5: DataLink Layer 5a-22
Frame Relay (more)
flags address
data
CRC
flags
 Flag bits, 01111110, delimit frame
 Address = address and congestion control
10 bit VC ID field
 3 congestion control bits
• FECN: forward explicit congestion
notification (frame experienced congestion
on path)
• BECN: congestion on reverse path
• DE: discard eligibility

5: DataLink Layer 5a-23
Frame Relay -VC Rate Control
 Committed Information Rate (CIR)
defined, “guaranteed” for each VC
 negotiated at VC set up time
 customer pays based on CIR

 DE bit: Discard Eligibility bit
Edge FR switch measures traffic rate for each
VC; marks DE bit
 DE = 0: high priority, rate compliant frame;
deliver at “all costs”
 DE = 1: low priority, eligible for discard when
congestion

5: DataLink Layer 5a-24
Summary
 principles behind data link layer services:
error detection, correction
 sharing a broadcast channel: multiple access
 link layer addressing, ARP
 various link layer technologies
 Ethernet hubs, bridges, switches
 IEEE 802.11 LANs
 PPP
 ATM, X.25, Frame Relay
 journey down the protocol stack now OVER!

5: DataLink Layer 5a-25
A bit about physical
connections
5: DataLink Layer 5a-26
Rating wide area internet
connections
 T0, DS0 – 1 voice channel, 65 Kbps

What homes get for 1 telephone line
 T1 (Level 1 transmission line) or DS1
 1.544 Mbps, 24 voice channels at 64 Kbps
 T3 or DS3 = 28 T1 lines, 44.746 Mbps
 OC3 = 3 DS3s
 OC12 = 12 DS3s
 OC48 = 48 DS3s, 2488 Mbps
 OC192 = 192 DS3s
5: DataLink Layer 5a-27
SONET and SDH
 Higher data rates often achieved using
synchronous optical networking (SONET)
and Synchronous Digital Hierarchy (SDH)

SONET in the US and Canada and SDH in the
rest of the world
 Transport over optical fiber using lasers/
LEDs
 Transporting large amounts of telephone
calls and data traffic over the same fiber
without synchronization problems
5: DataLink Layer 5a-28
 T0 = typical phoneline connection
 DS3 delivered native on a copper trunk or
converted to an optical fiber run when
needing longer distances between
termination points
DS3 transported over SONET is encapsulated
in a STS-1 SONET channel
 Still analog when delivered over fiber

 When delivering data over an OC3 or
greater SONET is used.
OC-3 SONET link contains three STS-1s, and
therefore may carry three DS3s.
 Likewise, OC-12, OC-48, and OC-192 may carry
12, 48, and 192 DS3s respectively.
5: DataLink Layer

5a-29
More on SONET
 Designed to carry multiple real-time,
uncompressed, circuit-switched voice lines
encoded in Pulse-Code Modulation (PCM)
format
 Also multiple digital bit streams of
differing origin within single framing
protocol
Multiplex circuit mode communications (T1, T3,
DS1, DS3,etc.) from a variety of different
sources over same fiber
 Emphasis is on merging many different flow into
5: DataLink Layer
one quickly

5a-30
 STM-1 (Synchronous Transport Module,
level 1) frame is the basic transmission
format for SDH.
 STM-1 frame is transmitted in exactly
125 µs, therefore, there are 8,000 frames
per second on a 155.52 Mbit/s OC-3 fiberoptic circuit
5: DataLink Layer 5a-31
 Protocol neutral
 Not communications protocols in and of
themselves

Generic, all-purpose transport containers for
moving both voice and data.
 Used to carry ATM, Ethernet, TCP/IP etc.
5: DataLink Layer 5a-32
 SONET standard defined by Telcordia and
American National Standards Institute
(ANSI) standard T1.105 and T1.119
 SDH standard specified in International
Telecommunication Union (ITU) standards
G.707, G.783, G.784, and G.803

SDH originally defined by the European
Telecommunications Standards Institute
(ETSI)
5: DataLink Layer 5a-33
Carrier Pricing
 Two simple components: local loop and port
 Local loop = cost to transport the signal
from the end user's central office (CO) to
the point of presence (POP) of the carrier
 Local
loop cost based on geography/distance
from CO to POP
 Port = cost to access the network through
the carrier's network

Port cost based on access speed and yearly
commitment level
5: DataLink Layer 5a-34
Fiber cable runs
 One example
from the North
Country
5: DataLink Layer 5a-35
Undersea cables
5: DataLink Layer 5a-36
 multiple SONET signals can be transported
over multiple wavelengths on a single fiber
pair by means of wave length-division
multiplexing, including dense wavelengthdivision multiplexing (DWDM) and coarse
wavelength-division multiplexing (CWDM).
 DWDM circuits are the basis for all
modern submarine communications cable
systems and other long-haul circuits.
5: DataLink Layer 5a-37
Other
 Satellite Links
 Pros and Cons
5: DataLink Layer 5a-38
Outtakes
5: DataLink Layer 5a-39
IEEE 802.11 MAC Protocol
802.11 CSMA Protocol:
others
 NAV: Network
Allocation
Vector
 802.11 frame has
transmission time field
 others (hearing data)
defer access for NAV
time units
5: DataLink Layer 5a-40
IEEE 802.11 MAC Protocol:
CSMA/CA
802.11 CSMA: sender
- if sense channel idle for
DISF sec.
then transmit entire frame
(no collision detection)
-if sense channel busy
then binary backoff
802.11 CSMA receiver:
if received OK
return ACK after SIFS
5: DataLink Layer 5a-41
IP-Over-ATM
Classic IP only
 3 “networks” (e.g., LAN
segments)
 MAC (802.3) and IP
addresses
Ethernet
LANs
IP over ATM
 replace “network” (e.g.,
LAN segment) with ATM
network
 IP addresses -> ATM
addresses just like IP
addresses to 802.3 MAC
addresses!
Ethernet
LANs
ATM
network
5: DataLink Layer 5a-42
ARP in ATM Nets
 ATM network needs destination ATM address
just like Ethernet needs destination Ethernet
address
 IP/ATM address translation done by ATM ARP
(Address Resolution Protocol)
 ARP server in ATM network performs
broadcast of ATM ARP translation request to
all connected ATM devices
 hosts can register their ATM addresses with
server to avoid lookup

5: DataLink Layer 5a-43
Access Control
 802.11 working group considered 2
proposals for a MAC algorithm
Distributed access protocols
 Centralized access protocols

5: DataLink Layer 5a-44
Distributed Access Protocols
 Distribute the decision to transmit over all
the notes
 Like Carrier-sense mechanisms in Ethernet
 Makes sense especially for an ad hoc
network of peer workstations
 Can also be good for busty traffic
5: DataLink Layer 5a-45
Centralized Access Protocols
 Regulation of transmission by a centralized
decision maker
 Natural for networks with a base station
 Especially good if network is highly utilized
( avoid fighting it out among peers)
 Also good if some data is time
sensitive/high priority
5: DataLink Layer 5a-46
Distributed Foundation
Wireless MAC
 Compromise was Distributed Foundation
Wireless MAC (DFWMAC)
 Distributed Access control mechanism with
an optional centralized control layer on top
of that
Distributed Coordination Function (DCF) on top
of physical layer
 On top of that is optional Point Coordination
Function (PCF) that provides contention free
service

5: DataLink Layer 5a-47
Access Control
CSMA
 DCF uses Carrier Sense Multiple Access (CSMA)
 CSMA means listen before you send to make sure the
medium is idle
 No Collision Detection - Not CSMA/CD like Ethernet



CD based on listening while you send to make sure you hear only
your signal
Wireless HW not made to send and listen at same time
Large dynamic range of possible signals – cannot effectively
distinguish incoming weak signals from noise and the effects
of its own transmission
5: DataLink Layer 5a-49
IFS = interframe space
Medium Access Control Logic
Each time fail
increase time to
wait before send
Interframe Space (IFS) Values
 Actually three different IFS values
 Short IFS (SIFS)
 Shortest IFS
 Used for immediate response actions
 Point coordination function IFS (PIFS)
 Midlength IFS
 Used by centralized controller in PCF scheme when using polls
 Distributed coordination function IFS (DIFS)
 Longest IFS
 Used as minimum delay of asynchronous frames contending for
access
5: DataLink Layer 5a-51
Priority
 Stations using SIFS have “priority” over
others because they will test for idle
faster find and then start transmitting
 Others that wait longer will find the
channel busy when they listen after PIFS
or DIFSs
5: DataLink Layer 5a-52
IFS Usage
 SIFS
 Acknowledgment (ACK)
 Clear to send (CTS)
 Poll response( for PCF)
 PIFS
 Used by centralized controller in issuing polls
(for PCF)
 Takes precedence over normal contention
traffic
 DIFS
 Used for all ordinary asynchronous traffic
5: DataLink Layer 5a-53
Contention Periods/
Contention-Free Periods
 The DCF and PCF respectively operate in
Contention Periods (CPs) and Contention Free
Periods (CFPs)
 In CPs, stations compete with each other to win
channel access
 In CFPs, an Access Point (AP) grants the
opportunity of transmission to stations by polling
5: DataLink Layer 5a-54
Polling
 Since PIFS smaller than DIFS, coordinator
can seize coordinator and lock all traffic (
at least traffic that obeys the rules) while
it polls and receives responses
 When polling coordinator sends a poll to a
station, it can respond using SIFS ( beating
the next PIFS and any DIFS)
5: DataLink Layer 5a-55
Polling
 In a CFP, a PC polls the first station in its polling list,
and it may also piggyback some data to the polling frame.
 The polled station responds either with an ACK or a data
frame piggybacked to the ACK frame.
 An SIFS separates the polling and responding frames.
 Once the frame exchange sequence with the first
station is done, the PC waits for a PIFS and then polls
another station in its polling list.
5: DataLink Layer 5a-56
Superframes
 CPs and CFPs alternate in a superframe
 A superframe is an interval between two beacon
frame transmissions.
 A beacon frame is broadcasted by APs in BSSs or
random stations in IBSSs.

It carries management information to the stations.
5: DataLink Layer 5a-57
IEEE 802.11 MAC Timing
PCF Superframe Construction
5: DataLink Layer 5a-58
Superframe
 Point coordinator would lock out asynchronous traffic by
issuing polls
 Superframe interval defined



During first part of superframe interval, point coordinator polls
round-robin to all stations configured for polling
Point coordinator then idles for remainder of superframe
Allowing contention period for asynchronous access
 At beginning of superframe, point coordinator may seize
control and issue polls for given period

Time varies because of variable frame size issued by responding
stations
 Rest of superframe available for contention-based access
 At end of superframe interval, point coordinator contends
for access using PIFS
 If idle, point coordinator gains immediate access


Full superframe period follows
5: access
DataLink Layer
If busy, point coordinator must wait for idle to gain
5a-59
Acknowledgements
 When station received frame addressed directly
to it ( not broadcast or multicast) it replies with
an ACK after waiting SIFS
 ACKs allow for recovery from collision since no
collision detection
 Use of SIFS allows for efficient delivery of an
LLC data unit that requires multiple MAC frames


Just get SIFS between ACK and then next frame
No one else will gain control of the channel until the
entire LLC if over
5: DataLink Layer 5a-60
802.11 Physical Layer
Standards
Op. Freq.
Data Rate
Typical/Max
(Mbit/sec)
Range
Indoor/Outdoor
(meters)
Legacy
802.11-1997
2.4 GHz
½
?
802.11a (1999)
5 GHz
25/54
15-30
802.11b (1999)
2.4 GHz
5.5/11
45-90
802.11g(2003)
2.4 GHz
25/54
45-90
802.11n(2009)
5 and 2.4 GHz
144/600
91/182
5: DataLink Layer 5a-61
 802.11b was the first, followed by 802.11a (
higher BW, less popular)
 802.11g higher BW, directly compatible
with b
 802.11n – even higher BW, backwards
compatible with b and g
5: DataLink Layer 5a-62
RC4
 WEP uses RC4 a stream cipher
Stream ciphers are vulnerable to attack if the same key is used
twice (depth of two) or more.
 Say we send messages A and B of the same length, both encrypted
using same key, K. The stream cipher produces a string of bits C(K)
the same length as the messages. The encrypted versions of the
messages then are:
 E(A) = A xor C
 E(B) = B xor C
 where xor is performed bit by bit.





Say an adversary has intercepted E(A) and E(B). He can easily
compute:
E(A) xor E(B)
However xor is commutative and has the property that X xor X
= 0 (self-inverse) so:
E(A) xor E(B) = (A xor C) xor (B xor C) = A xor B xor C xor C =
A xor B
5: DataLink Layer 5a-63