EE/CpE440: Current Topics10/29/02
EE/CpE 440
Current Topics in EE/CpE
Beyond 3G - The Potentialfor LAN-like Data Rates
with Cellular-like Coverage
Bruce McNairBurchard 206201-216-5549
[email protected]://koala.ece.stevens-tech.edu/~bmcnair
October 29, 2002
EE/CpE440: Current Topics10/29/02
Outline
• Background/Motivation for 4G
• Theory & Practical Considerations
• Prototype Implementation
• Experimental Results
• Future Directions
EE/CpE440: Current Topics10/29/02
Outline
• Background/Motivation for 4G
• Theory & Practical Considerations
• Prototype Implementation
• Experimental Results
• Future Directions
EE/CpE440: Current Topics10/29/02
Fourth Generation (4G) Wireless Access
Home LANsHome LANsCable networksCable networks
intranetsintranets
1G/2G1G/2G
POTSPOTS
voice/analog data
sophisticated wired data networking demand ↑demand for mobility ↑
reliance on mobile computing/PDAs ↑
TheInternet
TheInternet
3G 4G
Need for sophisticated,high-speed wireless data⇒
EE/CpE440: Current Topics10/29/02
Design Considerations for a 4G System
Low start-up bandwidth,flexible frequency usage,coordinated parameters
Crowded spectrum,Interference potential
Co-existence with present services
Packet mode operationMulti-user operation, Spectral efficiency
Internet access
High peak data rate,Bandwidth requirementsSpectrum allocation
Multimedia applications,low latency
Asymmetrical capacity,smart antennas,channel coding,
Power consumption, size, transmit power, link budget
Portable devices, reasonable battery life
TradeoffsRestrictionsDesirable Characteristics
EE/CpE440: Current Topics10/29/02
Design Considerations for a 4G System
Low start-up bandwidth,flexible frequency usage,coordinated parameters
Crowded spectrum,Interference potential
Co-existence with present services
Packet mode operationMulti-user operation, Spectral efficiency
Internet access
High peak data rate,Bandwidth requirementsSpectrum allocation
Multimedia applications,low latency
Asymmetrical capacity,smart antennas,channel coding,
Power consumption, size, transmit power, link budget
Portable devices, reasonable battery life
TradeoffsRestrictionsDesirable Characteristics
EE/CpE440: Current Topics10/29/02
Design Considerations for a 4G System
Low start-up bandwidth,flexible frequency usage,coordinated parameters
Crowded spectrum,Interference potential
Co-existence with present services
Packet mode operationMulti-user operation, Spectral efficiency
Internet access
High peak data rate,flexible assignment
Bandwidth requirementsSpectrum allocation
Multimedia applications,low latency
Asymmetrical capacity,smart antennas,channel coding,
Power consumption, size, transmit power, link budget
Portable devices, reasonable battery life
TradeoffsRestrictionsDesirable Characteristics
EE/CpE440: Current Topics10/29/02
Design Considerations for a 4G System
Low start-up bandwidth,flexible frequency usage,coordinated parameters
Crowded spectrum,Interference potential
Co-existence with present services
Packet mode operationMulti-user operation, Spectral efficiency
Internet access
High peak data rate,flexible assignment
Bandwidth requirementsSpectrum allocation
Multimedia applications,low latency
Asymmetrical capacity,smart antennas,channel coding,
Power consumption, size, transmit power, link budget
Portable devices, reasonable battery life
TradeoffsRestrictionsDesirable Characteristics
• User perceived data rate capacity of a shared packet channel is the unused capacity of the channel• Larger peak-capacity channels look like larger capacity dedicated channels for the same average level of utilization
EE/CpE440: Current Topics10/29/02
4G WirelessA Direct migration path from 3G:
An existing 3G network provides: • signalling• control• 384 kb/s uplink• downlink for lower bandwidth traffic• ongoing compatibility with 3G terminals
4G WOFDM provides untetheredcable modem-like user experience:
• 10+ Mb/s peak user rate downlink• 384 kb/s EDGE uplink (asymmetric traffic rates)• power efficient design for portable terminals
EE/CpE440: Current Topics10/29/02
Outline
• Background/Motivation for 4G
• Theory & Practical Considerations
• Prototype Implementation
• Experimental Results
• Future Directions
EE/CpE440: Current Topics10/29/02
Limiting Factors in Mobile Wireless Communications
• Noise, SNR
• Multipath fading
• Interference
• Multipath dispersion
• Frequency selective fading
• Doppler shift
EE/CpE440: Current Topics10/29/02
Noise, SNR
Transmitter
Receiver
kTB noise
• Thermal noise is proportional to receiver bandwidth• -174 dBm/Hz • Increased by noise figure of receiver
• Transmitter signal attenuated by distance, obstacles• Square law attenuation in free space, ~3.5 power in terrestrial environment
• Link budget limits high-speed operation
Excess receivernoise
EE/CpE440: Current Topics10/29/02
Multipath fading
• Reflected signals take longer path than direct signal• Delayed reflections create constructive/destructive interference
• With no line-of-sight path, fading characterized by Rayleigh distribution– 10% of the time, signal is in 10 dB fade– 1% of the time, signal is in 20 dB fade
• It is impossible to design sufficient margin to operate through deepest fades
Transmitter
Receiver
EE/CpE440: Current Topics10/29/02
Interference
• Restricted spectrum availability creates greater interference potential
• Interference sources:– Co-channel operations– adjacent channel operations– Intermodulation products
Transmitter
Receiver
Interferer
EE/CpE440: Current Topics10/29/02
Multipath Dispersion
• Delayed versions of signal interfere with each other– Equivalent to intersymbol interference on a baseband wireline system
• Multipath is modeled as a delay profile - signal delay and average amplitude with randomly varying instanteous signal level
• For comparison purposes, typical indoor and outdoor delay profiles have been standardized:
– Typical Urban, Hilly Terrain, Mountainous Terrain profiles– Exponential delay profile
Transmitter
Receiver
t
EE/CpE440: Current Topics10/29/02
Multipath Dispersion/ Frequency Selective Fading
• Delayed versions of signal interfere with each other– Equivalent to intersymbol interference on a baseband wireline system
• Multipath is modeled as a delay profile - signal delay and average amplitude with randomly varying instanteous signal level
• For comparison purposes, typical indoor and outdoor delay profiles have been standardized:
– Typical Urban, Hilly Terrain, Mountainous Terrain profiles– Exponential delay profile
• Through the Fourier Transform, delay profile can be studied in frequency domain
Transmitter
Receiver
t
EE/CpE440: Current Topics10/29/02
Doppler Shift
• Relative motion of transmitter/receiver creates an apparent frequency offset due to Doppler shift
• Offset is proportional to speed and frequency
• At a 2 GHz carrier frequency, 60 mph creates a 200 Hz Doppler shift
EE/CpE440: Current Topics10/29/02
4G Wireless:The Challenge:
Doppler frequency shift
Noise, multipath fading,interference impair signal
time
-40 dB
0 dB
Signal power
Multipath dispersionimpairs high symbol rates
time
Channel response
path loss
multipath fading, dispersion
Frequency selective fadingimpairs single carrier systems
frequency
Channel response
0 dB
-40 dB
shadow fading
interference
“kTB” noise
EE/CpE440: Current Topics10/29/02
Modulation Considerations• Parameters that can be modified to transmit information:
– Frequency– Amplitude– Phase
• FSK has a constant envelop, allowing use of power efficient non-linear amplifiers, but it isn’t spectrally efficient
• ASK requires linear amplification
• PSK is the modulation method of choice for most wireless communications systems
EE/CpE440: Current Topics10/29/02
Modulation parameter tradeoffs
• Data rate = (symbol rate)*(bits per symbol)
• To attain high data rates:– Transmit high symbol rates, or– Transmit many bits per symbol
EE/CpE440: Current Topics10/29/02
Modulation parameter tradeoffs
• Data rate = (symbol rate)*(bits per symbol)
• To attain high data rates:– Transmit high symbol rates, or– Transmit many bits per symbol
• High symbol rates are limited by time dispersion, creating intersymbol interference
• Large number of bits per symbol -> dense constellations -> susceptibility to noise and interference
• What is the solution?
EE/CpE440: Current Topics10/29/02
Modulation parameter tradeoffs
• Data rate = (symbol rate)*(bits per symbol)
• To attain high data rates:– Transmit high symbol rates, or– Transmit many bits per symbol
• High symbol rates are limited by time dispersion, creating intersymbol interference
• Large number of bits per symbol -> dense constellations -> susceptibility to noise and interference
• What is the solution?
• Keep symbol rate low to avoid dispersion, keep constellation simple to deal with noise. Just transmit multiple carriers.
-> OFDM (Orthogonal Frequency Division Multiplexing)
EE/CpE440: Current Topics10/29/02
OFDM Basics
tone spacingftNt tones
Operatingbandwidth, fB
f
t
CyclicsuffixNC/2
CyclicprefixNC/2
Rampdown
NR
RampupNR
2B R C G FN N N N N= + + +
2F F
B F C R G
N NN N N N N
η = =+ + +
Block efficiency
Tolerance todelay spread C Ct N≈ ∝
Tone spacing vsactive block time
1t
F
ft
=
Total bandwidth B t tf N f=
Raw capacityfor M-ary tone
modulationtN M
GuardNG
NF FFT samples
OFDM block, NBsample time t
block length tB
tFtC/2 tC/2
EE/CpE440: Current Topics10/29/02
coder/intlvr
mod IFFTsignal/ctrl
pilots
PAPRwindow
cyclic ext
filter
dataintf
D/A filtercontroller
transmitter
RFRFRF
OFDM Transmitter
EE/CpE440: Current Topics10/29/02
filter rotate FFT
ch est demod decode/deintlv
A/D
∆t est
∆f est
filter
dataintfcontroller
receiver
RFRFRF
OFDM Receiver
EE/CpE440: Current Topics10/29/02
Data rate: 6, 9*, 12, 18*, 24, 36*, 48*, 54* MbpsModulation: BPSK, QPSK, 16QAM, 64QAM*
Coding rate: 1/2, 2/3, 3/4*Subcarriers: 52
Pilot subcarriers: 4
G
3.2 µs
4 µs
FFT
52=48+4 tones64 point FFT
Key 802.11a Physical Layer Parameters:
Symbol duration: 4 µsGuard interval: 800 ns
Subcarrier spacing: 312.5 kHzBandwidth: 16.56 MHz
Channel Spacing: 20 MHz20 ppm
FFT size: 64
Carrier accuracy:Carrier accuracy @5.8GHz: 114 kHz
BPSK QPSK QAM16 QAM64
6 12 24R=1/2
48R=2/3
9 18 36 54R=3/4
User data rates (Mbps):
* optional
EE/CpE440: Current Topics10/29/02
OFDM Basics 2: tradeoffs
802.11aDVB-T
2k mode4G
Datarate
6, 9, 12, 18, 24, 36, 48, 54 Mb/s
Tonemodulation
BPSK, QPSK,16QAM, 64QAM
Codingrate 1/2, 2/3, 3/4
Nt 52
tB 4 µs
tB-tF 800 ns
ft 312.5 kHz
fB 16.56 MHz
fop 5 GHz
4.98-31.67 Mb/s
QPSK, “16QAM,” “64QAM”
1/2, 2/3, 3/4, 5/6, 7/8
1705
231-280 µs
7-56 µs
4.464 kHz
7.6 MHz
~700 MHz
5-10 Mb/s
QPSK,16QAM
1/2, 2/3, 3/4
640
200 µs
40 µs
6.25 kHz
4 MHz
2 GHz
EE/CpE440: Current Topics10/29/02
DD/P
DD/P
D
f
t
WOFDM frame = 20 msec
one to four
1 MHz slots
burst = 1 msec
OFDM/TDMA Options for 4G
• Full peak data rates are achievable
• Dynamic Packet Assignment to base stations, mobiles is an option
• Portable terminals can process only relevant traffic for power savings
EE/CpE440: Current Topics10/29/02
Fourth Generation WirelessAn OFDM-based Solution:
• Current OFDM applications - high-speed wired and wireless digital systems:
- European DVB-T- Digital Audio Broadcast (DAB)- Digital Subscriber Loop (DSL)- IEEE 802.11 Wireless LAN
• Rapid ongoing advances in general purpose DSP technology directly benefit OFDM’s straightforward signal processing
DD/P
DD/P
D
f
t
WOFDM frame = 20 msec
one to four
1 MHz slots 10 codewords (transport units)/burst
burst = 1 msec
Tone-by-tone flexibility
OFDMOFDM
Adaptivemodulation
Smartantennas Per user
tone allocation
Bandwidthcontrol
Resistanceto interference
Time/frequencycoding Inherent resistance
to dispersionRobustness
Simple underlyingmodulation
System flexibility
Per base stationtone allocation
Dynamic packetassignment
6.25 kHzN*160 tones
N*1 MHz
f
EE/CpE440: Current Topics10/29/02
Outline
• Background/Motivation for 4G
• Theory & Practical Considerations
• Prototype Implementation
• Experimental Results
• Future Directions
EE/CpE440: Current Topics10/29/02
Base station
4G Wireless Research Prototype
Channel simulator
Mobile station
• prototype designed with general purpose DSPs for flexibility
• two-branch receiver diversity implemented at 1900 MHz
• performance measured on typical mobile outdoor channels
• robust performance demonstrated
EE/CpE440: Current Topics10/29/02
OFDM Prototype
Pentek A/D Pentek Node Controller
Custom Clock Board
Custom D/A
RF
• PCS band operation• 384 kb/s end user capacity in 800 kHz • with OFDM• DSPs programmed (mostly) in C • Experiment with 2-way diversity,
synchronization
RF interface
Pentek ‘C40 DSPs
Sundevelopment
platform
bmcnair6/17/99
EE/CpE440: Current Topics10/29/02
E
E
D
C
B
A
RF 2
A/D
A/D
Sampleclock
FFT
FFT
RSdecode
BERcount
AGC
RF 1Eras.det
monitor control
DSP-> RFinterface
Synth,RF init
D/AMonitorscope
demod
demod
∆testimate
∆ωestimate
Differential Equal Gain Diversity OFDM Receiver
G
G
G
G
framing
G
G
G
G
G
G
1
2
5
5
2
V
window
window
Hardware Architecture
Static gain adjust
Out of frame
A/D initialization
2.166 MHz Ts
512 FFT, 189 tones Differential in time demod, equal gain combining
RS(63,31) with 16 erasures
625 samples/288.46 µsec block
∆t estimation:differential in frequency
∆f estimation:differential in time acrosscyclic extension
EE/CpE440: Current Topics10/29/02
FBA C
rotate
FFT
WaitISR
WriteGRAM
rotate
FFT
StartDMA
updatecount
StartDMA
updatecount
INT INT
D
signalB
WaitA
writeGRAM
WaitA| B
ReadGRAM
demod
signalC
SignalD
RSdecode
dropsampl
dropsampl
frame
BER
∆test
∆ωest
Wait C
AGC
sum
checkctl
writeRF
E
Dropsamples
Sumout
FFT A
FFT B
signalC
startDMA
INT wait forinput
get r_ptrdecodeinput
displayoptions
Set r_ptrwrite ctlstatus
controlstatus
outputvectors
InitializeA/D
writeclock
∆ω
InitializeRF
Process flow
CampISR
Markblock
WaitB
signalA
Alone?Y
N
SignalE
WaitD
erasedetect
descramb
ReadGRAM
OFDM Receiver - DSP Software Architecture
EE/CpE440: Current Topics10/29/02
Datainterface
Coder ModulatorInverse
FFTD/A RF
OFDM transmitter
RF A/D FFT
DemodulatorErasure
detection DecoderDataIntf
RF A/D FFTOFDM receiver
• Fast Fourier Transform (FFT) forms basis of OFDM complexity advantage
• Multiple individual carriers retain antenna diversity advantage
• Good performance is maintained with high fading rates (i.e., high vehicle speeds)and channels with large delay spread (e.g., Hilly Terrain)
Internet/intranet
Internet/intranet
EE/CpE440: Current Topics10/29/02
800 kHz
RF A/D FFT
DemodulatorErasure
detection DecoderDataIntf
RF A/D FFTOFDM receiver
Two equal rays 5 µsec delay
EE/CpE440: Current Topics10/29/02
RF A/D FFT
DemodulatorErasure
detection DecoderDataIntf
RF A/D FFTOFDM receiver
“Typical Urban” channel
800 kHz
EE/CpE440: Current Topics10/29/02
Outline
• Background/Motivation for 4G
• Theory & Practical Considerations
• Prototype Implementation
• Experimental Results
• Future Directions
EE/CpE440: Current Topics10/29/02
+atten
+atten
noise
atten
splitter
atten
noise
LPFLPF
TXBaseband
TXRF
LPF
LPF
TwoBranchFader
LPF
LPF
TwoBranch
RXRF
RXBaseband
RXBaseband
Experimental setup for OFDMreceiver performance measurements
• Simulated channels: AWGN, flat fading, two-ray, GSM models• Tested one- and two-branch receiver• Excellent repeatability and agreement with simulation and theory
EE/CpE440: Current Topics10/29/02
Theoretical versus measured performance: one-branch receiver in AWGN
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
7 8 9 10 11 12
SNR
FE
R
measured
(47,31) ν=8
(63,31) ν=16
EE/CpE440: Current Topics10/29/02
Simulated vs. Measured Results
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
5.00 7.00 9.00 11.00 13.00 15.00 17.00 19.00 21.00
SNR (dB)
FER
measured: 2rx, 200Hz, 20usecmeasured: 2rx, 200Hz, 5usecmeasured: 2rx, 40Hz, 20usecmeasured: 2rx, 40Hz, 5usecsimulation: 2rx 200Hz 20usecsimulation: 2rx 200Hz 5usecsimulation: 2rx 40Hz 20usecsimulation: 2rx 40Hz 5usec
EE/CpE440: Current Topics10/29/02
FER vs SNRRS(63,31) with erasure detection (ρ=16)
OFDM receiver, 2 ray
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
5.00 7.00 9.00 11.00 13.00 15.00 17.00 19.00 21.00
SNR
FE
R
1RX 40Hz 2usec
2RX 40Hz 2usec
1RX 40Hz 5usec
2RX 40Hz 5usec
1RX 40Hz 10usec
2RX 40Hz 10usec
1RX 40Hz 20usec
2RX 40Hz 20usec
EE/CpE440: Current Topics10/29/02
FER vs SNRRS(63,31) with erasure detection (ρ=16)
one- and two-branch receiver, GSM delay profiles
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
5.00 7.00 9.00 11.00 13.00 15.00 17.00 19.00 21.00
SNR
FE
R 1RX HT 40Hz
1RX MT 40Hz
1RX TU 40Hz
1RX BU 40Hz
2RX HT 40Hz
2RX MT 40Hz
2RX TU 40Hz
2RX BU 40Hz
EE/CpE440: Current Topics10/29/02
FER vs SNRRS(63,31) with erasure detection (ρ=16)two-branch receiver, GSM delay profiles
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
5.00 7.00 9.00 11.00 13.00 15.00 17.00 19.00 21.00
SNR
FE
R
2RX HT 200Hz
2RX HT 100Hz
2RX HT 40Hz
2RX BU 200Hz
2RX BU 100Hz
2RX BU 40Hz
2RX BU 1Hz
EE/CpE440: Current Topics10/29/02
FER vs SNRRS(63,31) with erasure detection (ρ=16)
OFDM receiver, 2 ray
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
5 7 9 11 13 15 17 19 21
SNR
FE
R
1RX 1Hz 5usec1RX 5Hz 5usec1RX 100Hz 5usec1RX 200Hz 5usec2RX 1Hz 5usec2RX 5Hz 5usec2RX 40Hz 5usec2Rx 100Hz 5usec2RX 200Hz 5usec
.
EE/CpE440: Current Topics10/29/02
SNR required for 10% FER,two-branch receiver
SNR required for 10% FER,one-branch receiver
Fading rate: 1 Hz 5 Hz 40 Hz 100 Hz 200 Hz
AWGN 9.4 dB
Bad Urban 13.4 dB 13.5 dB 14.4 dB 18.3 dB
2 µsec 13.5 dB ~21 dB
5 µsec 13.1 dB 13.1 dB 13.1 dB 14.0 dB ~21 dB
10 µsec 13.2 dB
20 µsec 14.5 dB 14.0 dB >21 dB
Typical Urban 14.4 dB 14.8 dB 15.5 dB >21 dB
Hilly Terrain 15.8 dB 17.3 dB >>21 dB
Mountainous Terrain 16.0 dB >>21 dB
Channel:
Flat 17.1 dB 17.1 dB 19.1 dB >21 dB
Performance Summary
Fading rate: 1 Hz 5 Hz 40 Hz 100 Hz 200 Hz
AWGN 7.0 dB
Bad Urban 9.3 dB 9.4 dB 9.7 dB 10.8 dB
2 µsec 8.9 dB 11.9 dB
5 µsec 8.1 dB 8.6 dB 8.9 dB 9.2 dB 10.4 dB
10µsec 9.1 dB
20µsec 9.0 dB 9.5 dB 11.6 dB
Typical Urban 10.4 dB 10.5 dB 10.6 dB 12.0 dB
Hilly Terrain 11.2 dB 11.6 dB 14.2 dB
Mountainous Terrain 10.4 dB 13.1 dB
Channel:
Flat 11.2 dB 10.9 dB 11.2 dB 12.7 dB
EE/CpE440: Current Topics10/29/02
Outline
• Background/Motivation for 4G
• Theory & Practical Considerations
• Prototype Implementation
• Experimental Results
• Future Directions
EE/CpE440: Current Topics10/29/02
What is the Future of Wireless Communications?
1. Will there be an evolution to 4G as there has been from 1G-2G-3G?• Who can afford the nation-wide investment of a new network?• Will the top-down evolution continue?
2. Will there a grass roots deployment of Wireless LANs with extensions to outdoors, high-speed mobility?
• Will next generation wireless grow like the Internet did, bottom-up?• Who will deploy the infrastructure to make this happen?• Will 4G wireless be ubiquitous like cellular or localized like cable modem/DSL
service?
EE/CpE440: Current Topics10/29/02
OFDM tradeoffs
802.11aDVB-T
2k mode4G
Datarate
6, 9, 12, 18, 24, 36, 48, 54 Mb/s
Tonemodulation
BPSK, QPSK,16QAM, 64QAM
Codingrate 1/2, 2/3, 3/4
Nt 52
tB 4 µs
tB-tF 800 ns
ft 312.5 kHz
fB 16.56 MHz
fop ~5 GHz
4.98-31.67 Mb/s
QPSK, “16QAM,” “64QAM”
[1/2, 2/3, 3/4, 5/6,7/8] + RS(204,88)
1705
231-280 µs
7-56 µs
4.464 kHz
7.6 MHz
~500 MHz
2.56-8.96 Mb/s
QPSK,16QAM
1/2, 2/3, 3/4, 7/8
640
200 µs
40 µs
6.25 kHz
4 MHz
~2 GHz
EE/CpE440: Current Topics10/29/02
Extending the WLAN Standards
802.11aDVB-T
2k mode‘4G’
Datarate
6, 9, 12, 18, 24, 36, 48, 54 Mb/s
Tonemodulation
BPSK, QPSK,16QAM, 64QAM
Codingrate 1/2, 2/3, 3/4
Nt 52
tB 4 µs
tB-tF 800 ns
ft 312.5 kHz
fB 16.56 MHz
fop ~5 GHz
4.98-31.67 Mb/s
QPSK, “16QAM,” “64QAM”
[1/2, 2/3, 3/4, 5/6,7/8] + RS(204,88)
1705
231-280 µs
7-56 µs
4.464 kHz
7.6 MHz
~500 MHz
2.56-8.96 Mb/s
QPSK,16QAM
1/2, 2/3, 3/4, 7/8
640
200 µs
40 µs
6.25 kHz
4 MHz
~2 GHz
Investigations show that a hybrid 802.11a format with modifications for outdoor operation is practical
Limitingparameters
EE/CpE440: Current Topics10/29/02
Coding rate: 1/2, 2/3, 3/4Subcarriers: 52 - insufficient for high data rates in wide area
Pilots subcarriers: 4 - insufficient if number of subcarriers increased
Symbol duration: 4 µs - too short for efficient wide area operationGuard interval: 800 ns - too short for wide area operation
Subcarrier spacing: 312.5 kHz - too large for narrow channelsBandwidth: 16.56 MHz - too large for spectrum available
Channel Spacing: 20 MHz
G
3.2 µs
4 µs
FFT
Carrier accuracy: 20 ppm - leads to too much carrier errorCarrier error @5.8GHz: 114 kHz - too much for narrower channel spacing,
even at 1.9 GHz
Issues:
Data rate: 6, 9, 12, 18, 24, 36, 48, 54 MbpsModulation: BPSK, QPSK, 16QAM, 64QAM
52=48+4 tones64 point FFT
FFT size: 64 - too small for number of carriers in crowed spectrum
EE/CpE440: Current Topics10/29/02
Coding rate: 1/2, 2/3, 3/4subcarriers:
G
230.4 µs
FFT
Data rate: 1.66, 2.5, 3.33, 5, 6.66, 10, 13.33, 15 MbpsModulation: BPSK, QPSK, 16QAM, 64QAM
832 = 52*16
204.8 µs
832=768+64 tones2048 point FFT
Changes for high-mobility operation:
FFT size: 2048 = 64*32Symbol duration: 230.4 µs = 3.2*64 + .8*32
Guard interval: 25.6 µs = .8*32Subcarrier spacing: 4.833 kHz = 312.5/64
Bandwidth: ~5 MHz ª 16.56/4Channel Spacing: 5 MHz ª 20/4Carrier accuracy: .5 ppm for 5 GHz, 1 ppm for 2.4 GHz
Carrier error @5.8GHz: 2.9 kHz, 1.9 kHz @ 1.9 GHz
Pilot subcarriers: 64 = 4*16BPSK QPSK QAM16 QAM64
1.66 3.33 6.66R=1/2
13.33R=2/3
2.5 5 10 15R=3/4
User data rates (Mbps):
EE/CpE440: Current Topics10/29/02
Coding rate: 1/2, 2/3, 3/4subcarriers:
G
230.4 µs
FFT
Data rate: 1.66, 2.5, 3.33, 5, 6.66, 10, 13.33, 15 MbpsModulation: BPSK, QPSK, 16QAM, 64QAM
832
204.8 µs
832=768+64 tones2048 point FFT
802.11a
FFT size: 2048Symbol duration: 230.4 µsGuard interval: 25.6 µs
Subcarrier spacing: 4.833 kHzBandwidth: ~5 MHz
Channel Spacing: 5 MHzCarrier accuracy: .5 ppm for 5 GHz, 1 ppm for 2.4 GHz
Carrier error @5.8GHz: 2.9 kHz, 1.9 kHz @ 1.9 GHz
Pilot subcarriers: 64
BPSK QPSK QAM16 QAM64
1.66 3.33 6.66R=1/2
13.33R=2/3
2.5 5 10 15R=3/4
User data rates (Mbps):
802.11a+6, 9*, 12, 18*, 24, 36*, 48*, 54* Mbps
BPSK, QPSK, 16QAM, 64QAM*1/2, 2/3, 3/4*
524
4 µs800 ns
312.5 kHz16.56 MHz
20 MHz20 ppm
64
114 kHz
G
3.2 µs
4 µs
FFT
52=48+4 tones64 point FFT
BPSK QPSK QAM16 QAM64
6 12 24R=1/2
48R=2/3
9 18 36 54R=3/4
User data rates (Mbps):
EE/CpE440: Current Topics10/29/02
Coexistence of Simulcast/Individualized Traffic
Pure simulcast has marginal differentiation from broadcast
Pure individualized transmissionhas marginal differentiation
from cellular
Opportunity space is forcustomizable
simulcast - but existing systemsdon’t efficiently address this option
Interactive & location-basedservices are a natural extension
of customization
EE/CpE440: Current Topics10/29/02
Simulcast Dispersion
Typical UrbanRMS delay spread
is about 1 µsec
Simulcast dispersion for 5 km radius cells is about 15 µsec at the 90th percentile
Most radio link schemes have been optimized to deal with intracell delay spread only
Top Related