Crosslayer design

1
Lecture 2
2
Cross-layer design

• Recall
• Motivation for cross-layer design
• direct coupling between the physical layer and
  higher layers is ubiquitous and unavoidable in
  the wireless environment
• Wireless networks do not come with links
• Need QoS provisioning in a highly dynamic
  environment
• Interference management algorithms at various
  layers of the protocol stack strongly coupled
• Objective QoS support at all layers using
  adaptation protocols
• Common misconception
• Layered approach must be completely eliminated
  and all layers must be integrated and jointly
  optimized!
• Solution holistic view of wireless networking
• - maintains the layered approach,
  while accounting for interactions between various
  protocols at different layers.
• Questions to be answered
• Which layers should respond to channel
  variations?
• What layers should be jointly optimized?
• What information should be exchanged between
  layers?
• How should this information be used by adaptation
  protocols at each particular layer?

3
Cross-layer design framework

• We focus on the first three layers of the
  protocol stack
• Determine abstract models for the layers and QoS
  metrics for each layer
• Determine adaptation protocols at each layer
• Determine required information exchange between
  layers (minimal)
• What is the coupling element between layers?
• Joint optimization design optimal or heuristic?
• Determine performance tradeoffs performance,
  complexity and scalability

Design steps
4
Cross-layer design framework - continuation
?
Adaptation protocol
?
?
?
Adaptation protocol
Adaptation protocol
5
Design goal ?

• Deliver QoS
• QoS measures
• Physical layer
• BER (Bit error rate)
• MAC layer
• Access delay, throughput
• Network layer
• Delay, throughput, blocking probability, dropping
  probability
• Other important performance measures
• Energy (power consumption, network lifetime)
• User capacity

Impact all layers
6
QoS requirements

• Application dependent
• Voice
• Delay intolerant, tolerant to some errors
• Needs access priority
• BER target 10-2
• Data
• Delay tolerant average delay may be recommended
• BER target 10-6
• Packets may be retransmitted until correctly
  received (ARQ protocols in data link layers)

7
Introduction to wireless communication

• Two Wireless Network Paradigms
• Cellular
• Terminals communicate to each other through the
  base station fixed infrastructure, allows
  centralized management
• Ad hoc networks
• Self organizing, no infrastructure, no
  centralized management
• Deployed in unlicensed bands
• no infrastructure costs
• Peer-to-peer communication
• Multi-hop communication from source to destination

8
Physical layer abstraction

• Physical layer ? link

If physical layer QoS are met BER (bit error
rate)
?
A
B
A
B
A
B
A
B
We will first focus on characterizing a wireless
link !
9
Network layer abstraction for wireless systems

• Cellular

Thickness of link lines indicates signal
strength (BER)
10
Network layer abstraction for wireless systems

• Ad hoc networks

• Again, links may exist or not between any two
  nodes and can be
• characterized by some QoS measures
• - e.g. BER (related to signal strength),
  transmission power,
• transmission rate, delay, etc
• - QoS metrics for links may be translated
  into costs for transmissions on
• a particular link ? routing for cost optimization

11
Physical layer Characterization of wireless links
12
Simple model for wireless transmission
Information Source
Modulator
Channel
Information Destination
Demodulator
Assume information source is digital generates a
string of bits that must be transmitted using
electromagnetic waves (no wires) - modulates a
carrier - sinusoidal signals suitable
carriers, characterized
by - amplitude amplitude modulation -
frequency frequency modulation - phase phase
modulation
Example BPSK
13
Modulation examples Cont.

• QPSK modulate both the sine and cosine (the
  quadrature) carrier
• Better spectral efficiency
• Can you improve further the spectral efficiency?
• M-ary modulation
• Example M - QAM
• bits encoded into one symbol
• Large M higher rates!!!
• Question can we get unlimited high rates for a
  given bandwidth by increasing M ?
• NO we have to be able to distinguish between the
  received symbols

14
Signal Constellation and Detection

• 4-QAM
• 16 QAM
• Higher constellation, less room for errors l
• Problem channel introduces noise, fading,
  distortsions

bases functions
15
Physical Channel

• Higher-order (M-ary) ? increased spectral
  efficiency
• Rate of reliable data transmission
• ? limited by impairments due to physical
  properties of the channel - noise (receiver
  background)
• - path losses (spatial diffusion shadowing)
• - multipath (fading dispersion)
• - interference (multiple-access co-channel)
• - dynamism (mobility, random-access bursty
  traffic)
• - limited transmitter power (FCC regulations,
  battery constraint)

16
Noise

• Noise present in all communication systems.
• White Gaussian noise
• spectrum constant for all frequencies
• pdf is Gaussian
• Key parameter of noise
• zero mean
• spectral height No/2 (variance of the
  noise)
• Key performance parameter when no interference is
  present
• SNR Eb/No (signal-to-noise ratio) (Eb
  received energy per bit)
• Determines BER (bit error rate)
• Different BER for different modulation types

17
Propagation effects

• Two basic types of propagation effects
• Large-scale (spatial diffusion shadow fading)
• Small-scale (multipath fading)

Characterization of radio waves
Radio wave ? oscillating in time at its
frequency. ? traveling through the
air at the speed of light
c
300,000,000 meters per sec ?
characterized by wavelength, ?
? wavelength the distance the wave travels
as it goes through one period (or cycle) of
oscillation
18
Frequency bands and wavelengths

• Different frequency for the carrier ? different
  physical properties
• propagation beyond the horizon
• energy absorption by the air
• propagation through rain, walls, etc.
• attenuation with distance
• sources of noise
• These properties can be better understood in
  terms of the wavelengths of the radiation.
• Next slides frequency bands allocation and
  corresponding wavelengths

19
Frequency Band Allocations
RADIO
IR
VISIBLE
UV
X-RAYS
GAMMA RAYS
RADIO
VLF
LF
MF
HF
VHF
UHF
SHF
EHF
300k
3k
30k
3M
30M
300M
3G
30G
300GHz
VLF Very Low Frequency LF Low Frequency MF
Medium Frequency HF High Frequency VHF Very
High Frequency UHF Ultra High Frequency SHF
Super High Frequency EHF Extremely High
Frequency
20
Wavelengths of Frequency Bands

• VLF, LF ? long waves
• MF ? medium waves
• HF, VHF ? short waves
• UHF, SHF ? microwaves ?
• EHF ? millimeter waves
• Above microwave region, only certain windows of
  frequencies propagate freely through air, rain,
  etc.
• Infrared and visible light will not penetrate
  walls
• X-rays and gamma rays interact with matter

Propagate well beyond line of sight
The distance the signal travels decreases as
the frequency increases
21
Some US Frequency Allocations
AM Radio
540 1,600 kHz
(medium wave) Cordless Phones
46 - 49 MHz
(FM) or 902-928 MHz
2.4 - 2.4835 GHz
(Spread Spectrum) FM Radio
88
108 MHz TV
54 216 MHz (VHF)
420 890 MHz (UHF)

not contiguous Cellular
824 - 894 MHz
(UHF) not contiguous PCS

1.85- 1.99 GHz (UHF) not contiguous Satellite
Comms
SHF Wireless LANs
the upper ISM bands
and IR (not regulated). ISM Industry, Science
Medicine - transmit power of 1 watt or
less. ISM Bands 902 - 928 MHz 2.4 - 2.4835
GHz 5.725 - 5.850 GHz
22
Licensed spectrum and FCC policy

• FCC Federal Communications Commission
• regulates the use of spectrum
• Spectrum auctioned for use by commercial
  companies
• e.g. Verizon, ATT, TV broadcast companies, etc
• Recent paradigm shift in FCC policy
• More efficient spectral re-use
• Unused TV broadcast may go to wireless
  Advantages In TV band the signals may travel
  farther and penetrate buildings easier than the
  signals in the current bands
• Second market users for lightly used spectrum
• Implications smart radio features to control
  interference
• Cognitive radio technology the radios sense the
  environment and adapt their transmission
  parameters according to the current level of
  interference and local policy rules
• Invited lecture??

23
Propagation Effects

• Large scale propagation
• Several phenomena occur when a wave propagates
  close to the earth surface
• Reflection a wave encounters objects larger
  than its wavelength
• Diffraction when a radio path is obstructed by
  an irregular surface. Secondary waves are
  generated, resulting in bending of waves around
  and behind the obstacle.
• Scattering when radio wave travels through a
  medium containing lots of small (compared to the
  wavelength) objects

24

• Propagation in free space ignores any
  interactions
• Antenna radiates a sine wave with the carrier
  frequency
• 
  speed of light
• Friis free space equation
• Propagation along the earths surface 2-ray
  model
• Flat earth assumption
• Ground wave reflected
• Delay
• Phase shift
• Attenuation

Pt transmit power Pr received power gt, gr
transmit/receive antenna gains r distance
between the antennas
25

• Propagation close to the earth surface
  continuation
• When antenna heights are small compared with the
  distance between antennas

The received powers can also be expressed in dB
or dBm
? Path loss coefficient
Free space
26

• 2-ray model
• In practice measurements

Fit a a straight line from measurements -the
slope gives the propagation exponent
Received power (dB)
Free space
2-ray model
Distance (meters)
27

• Find an exponent for the path loss model

n path loss n 2 6 n 2 waveguide
effect r distance transmitter-receiver X
Gaussian random variable
Lognormal r.v.
Lognormal random variable with parameters ?, ?
28
Gaussian distribution ? mean, ? std. deviation
Interference and channel reuse

• How should we design a cellular system
• One base station (BS), high power and large
  coverage?
• Split into cells to accommodate a larger density
  of users?

Example 2 BS use the same channels and are
situated at distance D Question how should we
choose D/R (R is the coverage radius for one
cell, Such that all users meet their target SIR
(signal-to-interference ratio) - SIR maps
the bit error rate performance (BER)
29
Assume noise is 0 Channel impairment is
determined by other users using the same
channel Duplex communication Equal transmission
powers

• Example continuation

BS2
BS1
D x1x2 T target SIR in dB
R
x2
x1 R
30

• Example continuation

? want small or big number?
From cellular efficiency point of view -gt want
small numbers
The effective number of channels/cell total
number of channels/ cell reuse factor (N)
How to determine N for given SIR requirement
For base stations situated in a straight line, it
can be shown by induction that
R
R
31

• For hexagonal cells

D
R
For CDMA systems N1
We can define the cellular efficiency
BS spectrum allocated to the cellular system BC
bandwidth/channel N channel reuse
32

• To compare different systems, we also have to
  account for their target SIR requirement

Target SIR requirement St (dB)
33

• Example that accounts for the shadow fading in
  the cell
• The received power (dBm) T Gaussian r.v.

R
At the cell boundary, what is the probability
that Tgt -100dBm ?
34

• Lecture reading assignment (due next week in
  class)
• J. Goldsmith and S. B. Wicker Design challenges
  for energy-constrained
• ad hoc wireless networks, IEEE Wireless
  Communications Magazine, vol 9,
• No.4., pp 8-27, August 2002.

Homework 1 due 2 weeks from now