2019 term 3 tele3118: network technologies · network technologies 1c-15 n single shared broadcast...
TRANSCRIPT
2019 term 3TELE3118: Network Technologies
Week 1: Data Link Layer Framing, Error Control, Multiple Access
Some slides have been taken from:qComputer Networking: A Top Down Approach Featuring the Internet, 7th edition. Jim Kurose, Keith Ross. Addison-Wesley, 2016. All material copyright 1996-2016. J.F Kurose and K.W. Ross, All Rights Reserved.qComputer Networks, 5th edition. Andrew S. Tanenbaum. Prentice-Hall, 2010.
Link layer: introduction
1c-2
terminology:n hosts and routers: nodesn communication channels
that connect adjacent nodes along communication path: linksq wired linksq wireless linksq LANs
n layer-2 packet: frame, encapsulates datagram
Network Technologies
data-link layer has responsibility of transferring datagram from one node to physically adjacent node over a link
Where is the link layer implemented?
1c-3Network Technologies
n in each and every hostn link layer implemented in “adaptor” (aka network interface card NIC) or on a chipq Ethernet card, 802.11
card; Ethernet chipsetq implements link, physical
layern attaches into host’s system
busesn combination of hardware,
software, firmware
controller
physicaltransmission
cpu memory
host bus (e.g., PCI)
network adaptercard
applicationtransportnetwork
link
linkphysical
Link Layer Functions
1c-4Network Technologies
n Framingn Error controln Addressing and Multiple-access controln Flow controln Reliable transfer
Framing
1c-5Network Technologies
A character stream. (a) Without errors. (b) With one error.
Framing (2)
1c-6Network Technologies
(a) A frame delimited by flag bytes.(b) Four examples of byte sequences before and after stuffing.
Framing (3)
1c-7Network Technologies
Bit stuffing(a) The original data.(b) The data as they appear on the line.(c) The data as they are stored in receiver’s memory after destuffing.
Error detection
1c-8Network Technologies
EDC= Error Detection and Correction bits (redundancy)D = Data protected by error checking, may include header fields
n Error detection not 100% reliable!n protocol may miss some errors, but rarelyn larger EDC field yields better detection and correction
Parity checking
1c-9Network Technologies
single bit parity:§ detect single bit
errors
two-dimensional bit parity:§ detect and correct single bit errors
Theory: Hamming Distances
1c-10Network Technologies
n Hamming distance of two codewords:q The number of bit positions in which they differq Example: 10001001 and 10110001 have Hamming distance 3
n Hamming distance of a code:q Minimum distance between any two valid codewordsq E.g. the code: 0000000000, 0000011111, 1111100000,
1111111111 has Hamming distance 5n message size n = data bits m + redundant bits r
q Of 2n codewords, only 2m are validn How many redundant bits needed for 1-bit error
correction?q Should satisfy: (n+1)2m ≤2n : why?
Cyclic redundancy check
1c-11Network Technologies
n more powerful error-detection codingn view data bits, D, as a binary numbern choose r+1 bit pattern (generator), Gn goal: choose r CRC bits, R, such that
q <D,R> exactly divisible by G (modulo 2) q receiver knows G, divides <D,R> by G. If non-zero remainder: error
detected!q can detect all burst errors less than r+1 bits
n widely used in practice (Ethernet, 802.11 WiFi, ATM)
CRC example
1c-12Network Technologies
want:D.2r XOR R = nG
equivalently:D.2r = nG XOR R
equivalently:if we divide D.2r by G, want remainder R to satisfy:
R = remainder[ ]D.2r
G
CRC properties
1c-13Network Technologies
n Can catch all errors with odd number of inverted bitsn Odd number of bits in error
q Represented by a polynomial with odd number of terms (e.g. 𝑥" + 𝑥$ + 1 )n No polynomial with odd number of terms has (x+1) as a factor in
modulo 2 n Add (x+1) as a factor in G(x) Jn Can catch all burst errors length ≤ r n Burst error of length k is represented 𝑥' 𝑥()* + ⋯+ 1
q there are k consecutive 1’s in your error
n If G(x) contains an 𝑥, term, it will never have 𝑥' as a factorq G(x) ends with +1 at the end! J
n G(x) = 𝑥-$ + 𝑥$. + 𝑥$- + 𝑥$$ + 𝑥*. + 𝑥*$ + 𝑥** + 𝑥*, + 𝑥/ + 𝑥0 +𝑥" + 𝑥1 + 𝑥$ + 𝑥* + 1
Multiple access links, protocols
1c-14Network Technologies
two types of “links”:n point-to-point
q PPP for dial-up accessq point-to-point link between Ethernet switch, host
n broadcast (shared wire or medium)q old-fashioned Ethernetq upstream HFCq 802.11 wireless LAN
shared wire (e.g., cabled Ethernet)
shared RF(e.g., 802.11 WiFi)
shared RF(satellite)
humans at acocktail party
(shared air, acoustical)
Multiple access protocols
1c-15Network Technologies
n single shared broadcast channel n two or more simultaneous transmissions by nodes:
interference q collision if node receives two or more signals at the same time
multiple access protocoln distributed algorithm that determines how nodes
share channel, i.e., determine when node can transmitn communication about channel sharing must use channel
itself! q no out-of-band channel for coordination
An ideal multiple access protocol
1c-16Network Technologies
given: broadcast channel of rate R bpsdesiderata:
1. when one node wants to transmit, it can send at rate R.2. when M nodes want to transmit, each can send at average rate
R/M3. fully decentralized:
n no special node to coordinate transmissionsn no synchronization of clocks, slots
4. simple
MAC protocols: taxonomy
1c-17Network Technologies
three broad classes:n channel partitioning
q divide channel into smaller “pieces” (time slots, frequency, code)
q allocate piece to node for exclusive use
n random accessq channel not divided, allow collisionsq “recover” from collisions
n “taking turns”q nodes take turns, but nodes with more to send can take
longer turns
Channel partitioning MAC protocols: TDMA
1c-18Network Technologies
TDMA: time division multiple access
n access to channel in "rounds" n each station gets fixed length slot (length = packet
transmission time) in each round n unused slots go idle n example: 6-station LAN, 1,3,4 have packets to send,
slots 2,5,6 idle
1 3 4 1 3 4
6-slotframe
6-slotframe
Channel partitioning MAC protocols: FDMA
1c-19Network Technologies
FDMA: frequency division multiple access n channel spectrum divided into frequency bandsn each station assigned fixed frequency bandn unused transmission time in frequency bands go idle n example: 6-station LAN, 1,3,4 have packet to send,
frequency bands 2,5,6 idle
freq
uenc
y ba
nds
time
FDM cable
Random access protocols
1c-20Network Technologies
n when node has packet to sendq transmit at full channel data rate R.q no a priori coordination among nodes
n two or more transmitting nodes ➜ “collision”,n random access MAC protocol specifies:
q how to detect collisionsq how to recover from collisions (e.g., via delayed
retransmissions)n examples of random access MAC protocols:
q slotted ALOHAq ALOHAq CSMA, CSMA/CD, CSMA/CA
Slotted ALOHA
1c-21Network Technologies
assumptions:n all frames same sizen time divided into equal
size slots (time to transmit 1 frame)
n nodes start to transmit only slot beginning
n nodes are synchronizedn if 2 or more nodes
transmit in slot, all nodes detect collision
operation:n when node obtains fresh
frame, transmits in next slotq if no collision: node can
send new frame in next slot
q if collision: node retransmits frame in each subsequent slot with prob. p until success
Slotted ALOHA
1c-22Network Technologies
Pros:n single active node can
continuously transmit at full rate of channel
n highly decentralized: only slots in nodes need to be in sync
n simple
Cons:n collisions, wasting slotsn idle slotsn nodes may be able to
detect collision in less than time to transmit packet
n clock synchronization
1 1 1 1
2
3
2 2
3 3
node 1
node 2
node 3
C C CS S SE E E
Slotted ALOHA: efficiency
1c-23Network Technologies
n max efficiency: find p* that maximizes Np(1-p)N-1
n for many nodes, take limit of Np*(1-p*)N-1 as N goes to infinity, gives:max efficiency = 1/e = 0.37
efficiency: long-run fraction of successful slots (many nodes, all with many frames to send)
at best: channelused for useful transmissions 37%of time! !
n suppose: N nodes with many frames to send, each transmits in slot with probability p
n prob that given node has success in a slot = p(1-p)N-1
n prob that any node has a success = Np(1-p)N-1
Pure (unslotted) ALOHA
1c-24Network Technologies
n unslotted Aloha: simpler, no synchronizationn when frame first arrives
q transmit immediately n collision probability increases:
q frame sent at t0 collides with other frames sent in [t0-1,t0+1]
Pure ALOHA efficiency
1c-25Network Technologies
P(success by given node) = P(node transmits) .P(no other node transmits in [t0-1,t0] .P(no other node transmits in [t0-1,t0]
= p . (1-p)N-1 . (1-p)N-1
= p . (1-p)2(N-1)
… choosing optimum p and then letting n
= 1/(2e) = 0.18
even worse than slotted Aloha!
CSMA (carrier sense multiple access)
1c-26Network Technologies
CSMA: listen before transmit:if channel sensed idle: transmit entire framen if channel sensed busy, defer transmission
n human analogy: don’t interrupt others!
CSMA collisions
1c-27Network Technologies
n collisions can still occur: propagation delay means two nodes may not hear each other’s transmission
n collision: entire packet transmission time wastedq distance & propagation
delay play role in determining collision probability
spatial layout of nodes
CSMA/CD (collision detection)
1c-28Network Technologies
CSMA/CD: carrier sensing, deferral as in CSMAq collisions detected within short timeq colliding transmissions aborted, reducing channel wastage
n collision detection: q easy in wired LANs: measure signal strengths, compare
transmitted, received signalsq difficult in wireless LANs: received signal strength
overwhelmed by local transmission strength
n human analogy: the polite conversationalist
CSMA/CD (collision detection)
1c-29Network Technologies
spatial layout of nodes
“Taking turns” MAC protocols
1c-30Network Technologies
channel partitioning MAC protocols:§ share channel efficiently and fairly at high load§ inefficient at low load: delay in channel access,
1/N bandwidth allocated even if only 1 active node!
random access MAC protocols§ efficient at low load: single node can fully utilize
channel§ high load: collision overhead“taking turns” protocols
look for best of both worlds!
“Taking turns” MAC protocols
1c-31Network Technologies
polling:n master node “invites”
slave nodes to transmit in turn
n typically used with “dumb” slave devices
n concerns:q polling overhead q latencyq single point of failure
(master)
master
slaves
poll
data
data
“Taking turns” MAC protocols
1c-32Network Technologies
token passing:§ control token passed from
one node to next sequentially.
n token messagen concerns:
q token overhead q latencyq single point of failure
(token)
T
data
(nothingto send)
T
Summary of MAC protocols
1c-33Network Technologies
n channel partitioning, by time, frequency or codeq Time Division, Frequency Division
n random access (dynamic), q ALOHA, S-ALOHA, CSMA, CSMA/CDq carrier sensing: easy in some technologies (wire), hard in
others (wireless)q CSMA/CD used in Ethernetq CSMA/CA used in 802.11
n taking turnsq polling from central site, token passingq Bluetooth, FDDI, token ring