15-441 Lecture 4
Physical Layer &
Link Layer Basics
Copyright © CMU, 2008-10
Based on slides from previous 441 lectures
Lecture 4 15-441 © 2008-10 1
Last Time
• Application Layer
Application
• Example Protocols Application
Presentation
– ftp Presentation
Session
Session
– http
Transport
Transport
• Performance Network
Network
Datalink
Datalink
Physical
Physical
Today (& Tomorrow (& Tmrw))
1. Physical layer. Application
Application
2. Datalink layer Presentation
Presentation
introduction, Session
Session
framing, error Transport
Transport
coding, switched Network
Network
networks. Datalink
Datalink
3. Broadcast- Physical
Physical
networks, home
networking.
Transferring Information
• Information transfer is a physical process
• In this class, we generally care about
– Electrical signals (on a wire)
– Optical signals (in a fiber)
– More broadly, EM waves
• Information carrier can also be ?
Lecture 4 15-441 © 2008-10 4
Transferring Information
• Information transfer is a physical process
• In this class, we generally care about
– Electrical signals (on a wire)
– Optical signals (in a fiber)
– More broadly, EM waves
• Information carriers can also be
– Sound waves
– Quantum states
– Proteins
– Ink & paper, etc.
Lecture 4 15-441 © 2008-10 5
From Signals to Packets
Packet
Transmission Sender Receiver
Application
Application
0100010101011100101010101011101110000001111010101110101010101101011010
Packets
Presentation
Presentation
Header/Body Header/Body Header/Body
Session
Session
Transport
Transport
Bit Stream 0 0 1 0 1 1 1 0
Network
Network
Datalink
Datalink
“Digital” Signal
Physical
Physical
Analog Signal
Lecture 4 15-441 © 2008-10 6
From Signals to Packets
Packet
Transmission Sender Receiver
0100010101011100101010101011101110000001111010101110101010101101011010
Packets
Header/Body Header/Body Header/Body
Bit Stream 0 0 1 0 1 1 1 0
“Digital” Signal
Analog Signal
Lecture 4 15-441 © 2008-10 7
Today’s Lecture
• Modulation.
• Bandwidth limitations.
• Frequency spectrum and its use.
• Multiplexing.
• Media: Copper, Fiber, Optical, Wireless.
• Coding.
• Framing.
Lecture 4 15-441 © 2008-10 8
Why Do We Care?
• I am not an electrical engineer?
• Physical layer places constraints on what
the network infrastructure can deliver
– Reality check
– Impact on system performance
– Impact on the higher protocol layers
– Some examples:
• Fiber or copper?
• Do we need wires?
• Error characteristic and failure modes
• Effects of distance
Lecture 4 15-441 © 2008-10 9
Modulation
• Changing a signal to convey information
• From Music:
– Volume
– Pitch
– Timing
Lecture 4 15-441 © 2008-10 10
Modulation
• Changing a signal to convey information
• Ways to modulate a sinusoidal wave
– Volume: Amplitude Modulation (AM)
– Pitch: Frequency Modulation (FM)
– Timing: Phase Modulation (PM)
• In our case, modulate signal to encode a 0 or a
1.
(multi-valued signals sometimes)
Lecture 4 15-441 © 2008-10 11
Amplitude Modulation
• AM: change the strength of the signal.
• Example: High voltage for a 1, low voltage for
a0
0 0 1 1 0 0 1 1 0 0 0 1 1 1 0 0 0 1 1 0 0 0 1 1 1 0
1 0 1 0 1
Lecture 4 15-441 © 2008-10 12
Frequency Modulation
• FM: change the frequency
0 1 1 0 1 1 0 0 0 1
Lecture 4 15-441 © 2008-10 13
Phase Modulation
• PM: Change the phase of the signal
1 0 1 0
Lecture 4 15-441 © 2008-10 14
Baseband vs Carrier Modulation
• Baseband modulation: send the “bare”
signal.
• Carrier modulation: use the signal to
modulate a higher frequency signal
(carrier).
– Can be viewed as the product of the two signals
– Corresponds to a shift in the frequency domain
Lecture 4 15-441 © 2008-10 15
Amplitude Amplitude Carrier Modulation
Amplitude
Signal Carrier Modulated
Frequency Carrier
Lecture 4 15-441 © 2008-10 16
Why Different Modulation
Methods?
Lecture 4 15-441 © 2008-10 17
Why Different Modulation
Methods?
• Transmitter/Receiver complexity
• Power requirements
• Bandwidth
• Medium (air, copper, fiber, …)
• Noise immunity
• Range
• Multiplexing
Lecture 4 15-441 © 2008-10 18
What Do We Care About?
• How much bandwidth can I get out of a
specific wire (transmission medium)?
• What limits the physical size of the network?
• How can multiple hosts communicate over
the same wire at the same time?
• How can I manage bandwidth on a
transmission medium?
• How do the properties of copper, fiber, and
wireless compare?
Lecture 4 15-441 © 2008-10 19
Bandwidth
• Bandwidth is width of the frequency range in
which the fourier transform of the signal is
non-zero.
• Sometimes referred to as the channel width
• Or, where it is above some threshold value
(Usually, the half power threshold, e.g., -3dB)
• dB
– Short for decibel
– Defined as 10 * log10(P1/P2)
– When used for signal to noise: 10 *
log10(S/N)
Lecture 4 15-441 © 2008-10 20
Signal = Sum of Waves
+ 1.3 X
+ 0.56 X
+ 1.15 X
Lecture 4 15-441 © 2008-10 21
The Frequency Domain
• A (periodic) signal can be viewed as a sum
of sine waves of different strengths.
– Corresponds to energy at a certain frequency
• Every signal has an equivalent
representation in the frequency domain.
– What frequencies are present and what is their strength
(energy)
• E.g., radio and TV signals.
Lecture 4 15-441 © 2008-10 22
The Nyquist Limit
• A noiseless channel of width H can at most
transmit a binary signal at a rate 2 x H.
– Assumes binary amplitude encoding
Lecture 4 15-441 © 2008-10 23
The Nyquist Limit
• A noiseless channel of width H can at most
transmit a binary signal at a rate 2 x H.
– Assumes binary amplitude encoding
– E.g. a 3000 Hz channel can transmit data at a
rate of at most 6000 bits/second
Hmm, I once bought a modem that did 54K????
Lecture 4 15-441 © 2008-10 24
How to Get Past the Nyquist
Limit
Lecture 4 15-441 © 2008-10 25
How to Get Past the Nyquist
Limit
• Instead of 0/1, use lots of different values.
• (Remember, the channel is noiseless.)
• Can we really send an infinite amount of
info/sec?
Lecture 4 15-441 © 2008-10 26
Past the Nyquist Limit
• More aggressive encoding can increase the
channel bandwidth.
– Example: modems
• Same frequency - number of symbols per second
• Symbols have more possible values
psk
Psk+
AM
• Every transmission medium supports
transmission in a certain frequency range.
– The channel bandwidth is determined by the transmission
medium and the quality of the transmitter and receivers
– Channel capacity increases over time
Lecture 4 15-441 © 2008-10 27
Capacity of a Noisy Channel
• Can’t add infinite symbols
– you have to be able to tell them apart.
– This is where noise comes in.
Lecture 4 15-441 © 2008-10 28
Capacity of a Noisy Channel
• Can’t add infinite symbols
– you have to be able to tell them apart.
– This is where noise comes in.
• Shannon’s theorem:
C = B x log2(1 + S/N)
– C: maximum capacity (bps)
– B: channel bandwidth (Hz)
– S/N: signal to noise ratio of the channel
Often expressed in decibels (db) ::= 10 log(S/N)
.
Lecture 4 15-441 © 2008-10 29
Capacity of a Noisy Channel
• Can’t add infinite symbols
– you have to be able to tell them apart.
– This is where noise comes in.
• Shannon’s theorem:
C = B x log2(1 + S/N)
– C: maximum capacity (bps)
– B: channel bandwidth (Hz)
– S/N: signal to noise ratio of the channel
Often expressed in decibels (db) ::= 10 log(S/N)
• Example:
– Local loop bandwidth: 3200 Hz
– Typical S/N: 1000 (30db)
– What is the upper limit on capacity?
• Modems: Teleco internally converts to 56kbit/s digital signal,
which sets a limit on B and the S/N.
Lecture 4 15-441 © 2008-10 30
Example: Modem Rates
100000
Modem rate
10000
1000
100
1975 1980 1985 1990 1995 2000
Year
Lecture 4 15-441 © 2008-10 31
Transmission Channel
Considerations
• Every medium supports
transmission in a certain Good Bad
frequency range.
– Outside this range, effects such
as attenuation, .. degrade the
signal too much
• Transmission and receive
hardware will try to
maximize the useful
bandwidth in this frequency
band. Frequency
– Tradeoffs between cost, distance,
bit rate
• As technology improves,
these parameters change,
even for the same wire.
Signal
Attenuation & Dispersion
• Real signal may be a combination of many
waves at different frequencies
• Why do we care?
Good Bad
+ On board
Frequency
Lecture 4 15-441 © 2008-10 33
Limits to Speed and
Distance
•Noise: “random” energy is
added to the signal.
•Attenuation: some of the
energy in the signal leaks
away.
•Dispersion: attenuation and
propagation speed are
frequency dependent.
(Changes the shape of the signal)
Effects limit the data rate that a channel can
sustain.
» But affects different technologies in different ways
Effects become worse with distance.
» Tradeoff between data rate and distance
Today’s Lecture
• Modulation.
• Bandwidth limitations.
• Frequency spectrum and its use.
• Multiplexing.
• Media: Copper, Fiber, Optical, Wireless.
• Coding.
• Framing.
Lecture 4 15-441 © 2008-10 35
Today’s Lecture
• Modulation.
• Bandwidth limitations.
• Frequency spectrum and its use.
• Multiplexing.
• Media: Copper, Fiber, Optical, Wireless.
• Coding.
• Framing.
Lecture 4 15-441 © 2008-10 36
Supporting Multiple Channels
• Multiple channels can coexist if they transmit at
a different frequency, or at a different time, or
in a different part of the space.
– Three dimensional space: frequency, space, time
• Space can be limited using wires or using
transmit power of wireless transmitters.
• Frequency multiplexing means that different
users use a different part of the spectrum.
– Similar to radio: 95.5 versus 102.5 station
• Controlling time (for us) is a datalink protocol
issue.
– Media Access Control (MAC): who gets to send when?
Lecture 4 15-441 © 2008-10 37
Time Division Multiplexing
• Different users use the wire at different points in
time.
• Aggregate bandwidth also requires more spectrum.
Frequency
Frequency
Lecture 4 15-441 © 2008-10 38
FDM: Multiple Channels
Determines Bandwidth of Link
Amplitude
Determines
Bandwidth
of Channel
Different Carrier
Frequencies
Lecture 4 15-441 © 2008-10 39
Frequency versus
Time-division Multiplexing
•With FDM different
users use different
Frequency
parts of the frequency
spectrum. Frequenc
– I.e. each user can send all the Bands
time at reduced rate
– Example: roommates
•With TDM different
users send at different
times. Slot Frame
– I.e. each user can sent at full
speed some of the time
– Example: a time-share condo
•The two solutions can Time
be combined.
Today’s Lecture
• Modulation.
• Bandwidth limitations.
• Frequency spectrum and its use.
• Multiplexing.
• Media: Copper, Fiber, Optical, Wireless.
• Coding.
• Framing.
Lecture 4 15-441 © 2008-10 41
Copper Wire
• Unshielded twisted pair (UTP)
– Two copper wires twisted - avoid antenna effect
– Grouped into cables: multiple pairs with common
sheath
– Category 3 (voice grade) versus category 5
– 100 Mbit/s up to 100 m, 1 Mbit/s up to a few km
– Cost: ~ 10cents/foot
• Coax cables.
– One connector is placed inside the other
connector
– Holds the signal in place and keeps out noise
– Gigabit up to a km
• Signaling processing research pushes the
capabilities of a specific technology.
– E.g. modems, use of cat 5
Lecture 4 15-441 © 2008-10 42
UTP
• Why twist wires?
Lecture 4 15-441 © 2008-10 43
UTP
• Why twist wires?
– Provide noise immunity
• Combine with Differential Signaling
Lecture 4 15-441 © 2008-10 44
Light Transmission in Fiber
1.0 LEDs Lasers
tens of THz
loss
(dB/km) 0.5
1.3
1.55
0.0
1000 1500 nm
(~200 Thz)
wavelength (nm)
Lecture 4 15-441 © 2008-10 45
Ray Propagation
cladding
core
lower index
of refraction
(note: minimum bend radius of a few cm)
Lecture 4 15-441 © 2008-10 46
Fiber Types
• Multimode fiber.
– 62.5 or 50 micron core carries multiple “modes”
– used at 1.3 microns, usually LED source
– subject to mode dispersion: different propagation modes
travel at different speeds
– typical limit: 1 Gbps at 100m
• Single mode
– 8 micron core carries a single mode
– used at 1.3 or 1.55 microns, usually laser diode source
– typical limit: 10 Gbps at 60 km or more
– still subject to chromatic dispersion
Lecture 4 15-441 © 2008-10 47
Fiber Types
Multimode
Single mode
Lecture 4 15-441 © 2008-10 48
Gigabit Ethernet:
Physical Layer Comparison
Medium Transmit/ Distance Comment
receive
Copper 1000BASE-CX 25 m machine room use
Twisted pair 1000BASE-T 100 m not yet defined; cost?
Goal:4 pairs of UTP5
MM fiber 62 mm 1000BASE-SX 260 m
1000BASE-LX 500 m
MM fiber 50 mm 1000BASE-SX 525 m
1000BASE-LX 550 m
SM fiber 1000BASE-LX 5000 m
Twisted pair 100BASE-T 100 m 2p of UTP5/2-4p of UTP3
MM fiber 100BASE-SX 2000m
Lecture 4 15-441 © 2008-10 49
How to increase distance?
• Even with single mode, there is a distance
limit.
• I.e.: How do you get it across the ocean?
Lecture 4 15-441 © 2008-10 50
How to increase distance?
• Even with single mode, there is a distance
limit.
• I.e.: How do you get it across the ocean?
pump
laser
source
Lecture 4 15-441 © 2008-10 51
Regeneration and Amplification
• At end of span, either regenerate
electronically or amplify.
• Electronic repeaters are potentially slow, but
can eliminate noise.
• Amplification over long distances made
practical by erbium doped fiber amplifiers
offering up to 40 dB gain, linear response
over a broad spectrum. Ex: 40 Gbps at 500
km. pump
laser
source
Lecture 4 15-441 © 2008-10 52
Wavelength Division Multiplexing
•Send multiple wavelengths through the same
fiber.
– Multiplex and demultiplex the optical signal on the fiber
•Each wavelength represents an optical carrier
that can carry a separate signal.
– E.g., 16 colors of 2.4 Gbit/second
•Like radio, but optical and much faster
Optical
Splitter
Frequency
Lecture 4 15-441 © 2008-10 53
Wireless Technologies
• Great technology: no wires to install,
convenient mobility, …
• High attenuation limits distances.
– Wave propagates out as a sphere
– Signal strength attenuates quickly 1/d3
• High noise due to interference from other
transmitters.
– Use MAC and other rules to limit interference
– Aggressive encoding techniques to make signal
less sensitive to noise
• Other effects: multipath fading, security, ..
• Ether has limited bandwidth.
– Try to maximize its use
– Government oversight to control use
Lecture 4 15-441 © 2008-10 54
Things to Remember
•Bandwidth and distance of networks is limited
by physical properties of media.
– Attenuation, noise, dispersion, …
•Network properties are determined by
transmission medium and transmit/receive
hardware.
– Nyquist gives a rough idea of idealized throughput
– Can do much better with better encoding
• Low b/w channels: Sophisticated encoding, multiple bits per
wavelength.
• High b/w channels: Simpler encoding (FM, PCM, etc.), many
wavelengths per bit.
– Shannon: C = B x log2(1 + S/N)
•Multiple users can be supported using space,
time, or frequency division multiplexing.
•Properties of different transmission media.
Lecture 4 15-441 © 2008-10 55
