Sie sind auf Seite 1von 137

Orthogonal Frequency Division Multiplexing OFDM

fred harris Cubic Signal Processing Chair San Diego State University fred.harris@sdsu.edu
Vehicular Technology Conference - 2004

Textbooks and References


Wireless OFDM Systems: How to Make Them Work Marc Engels, Editor OFDM Wireless LANs: A Theoretical and Practical Guide Juha Heiskala and John Terry OFDM for Wireless Multimedia Communications Richard Van Nee and Ramjee Prasad Single and Multi-Carrier Quadrature Amplitude Modulation Lajos Hanzo, William Webb, and Thomas Keller ADSL, VDSL, and Multicarrier Modulation John Bingham Implementing ADSL David Ginsburg DSL Advances Massimo Sorbara, John Cioffi, and Peter Silverman

OFDM
OFDM also known as Multi-Carrier or Multi-Tone Modulation DAB-OFDM Digital Audio Broadcasting DVD-OFDM Digital Video Broadcasting ADSL-OFDM Asynchronous Digital Subscriber Line Wireless Local Area Network IEEE-802.11a, IEEE-802.11g ETSI BRAN (Hyperlan/2)

OFDM Systems
System Transform Size Number Carriers Channel Spacing kHz Bandwidth MHz Sample Rate MHz Symbol Duration sec Data Rate Mbits/s HyperLAN/2 802.11a 64 64 52 4 52 4 312.5 312.5 16.25 16.56 20 20 3.2 0.8 3.2 0.8 6-54 6-54

DVB-T DAB ADSL

2048 1024 2048 8192 256 (down) 64 (up)

1712 842 1536 36-127 7-28

4.464 1.00 4.3125

7.643 1.536 1.104

9.174 2.048 1.104

224 24/48/96 msec 231.9

0.68-14.92 3.072 0.64-8.192

OFDM Advantages
Efficiently Deals With Multi-path Fading Efficiently Deals With Channel Delay Spread Enhanced Channel Capacity Adaptively Modifies Modulation Density Robustness to Narrowband Interference

OFDM Disadvantages
OFDM Sensitive to Small Carrier Frequency Offsets OFDM Exhibits High Peak to Average Power Ratio OFDM Sensitive to High Frequency Phase Noise OFDM Sensitive to Sampling Clock Offsets

Single Carrier System

Sequential Transmission of Waveforms Waveforms are Short Duration T Waveforms Occupy Full System Bandwidth 1/T

Multi-Carrier System

Parallel Transmission of Waveforms Waveforms are Long Duration MT Waveforms Occupy 1/M th Of System Bandwidth 1/T

OFDM: Dense Multichannel System


Conventional Multichannel System Non Overlapping Adjacent Channels. Channels separated by More Than Their Two Sided bandwidth

OFDM Multichannel System 50% Overlap of Adjacent Channels Available bandwidth is Used Twice Channels separated by Half Their Two Sided bandwidth

Standard Digital Communication System

Bandlimited Channel
Nyquist Spectrum Nyquist Spectrum With Cosine Taper

Infinite Duration Nyquist Pulse

Finite Duration Nyquist Pulse

Translation in Time Domain Phase Slope in Frequency Domain

Translated Signals Are Orthogonal When Peak is Translated to Zero Crossings of Original

Channel Distortion
Modifies Received Wave Shape
Matched Filter Output For SQRT Nyquist Pulse 1 0.8 0.6 0.4 0.2 0 -0.2 -6 -4 -2 0 2 4 6

Matched Filter Output For Channel Distorted SQRT Nyquist Pulse 1 0.8 0.6 0.4 0.2 0 -0.2 -6 -4 -2 0 2 4 6

Inter Symbol Interference (ISI) Due to Channel Distorted Signal


Eye Diagram No Channel Distortion 2 1 0

-1

-2 -1

-0.8

-0.6

-0.4

-0.2

0.2

0.4

0.6

0.8

Eye Diagram With Channel Distortion 2

-1

-2 -1

-0.8

-0.6

-0.4

-0.2

0.2

0.4

0.6

0.8

Steady State Response of a Filter to a Sine Wave is a Sine Wave

Rectangle Pulse: DC Centered Spectrum with Equally Spaced Zeros

H0 (f ) = ATP

sin(2 f (2 f

TP ) 2

TP ) 2

Shift Spectrum with Linear Phase on DC Pulse: Move Spectrum to First Spectral Zero

sin(2 ( f k H k ( f ) = ATP

1 TP ) ) Tp 2 1 T (2 ( f k ) P ) Tp 2

Real Part of Complex Exponential Time Series: Integer Number of Cycles per Interval
1 0 -1 0 1 0 -1 0 1 0 -1 0 1 0 -1 0 0.5 1 0.5 1 1 0 -1 0 0.5 1 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1

Imaginary Part of Complex Exponential Time Series: Integer Number of Cycles per Interval
1 0 -1 0 1 0 -1 0 1 0 -1 0 1 0 -1 0 0.5 1 0.5 1 1 0 -1 0 0.5 1 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1 1 0 -1 0 0.5 1

Spectra Of Complex Exponential Time Series: Integer Number of Cycles per Interval
1 0.5 0 0 1 0.5 0 0 1 0.5 0 0 1 0.5 0 0 0.5 0.5 1 0.5 0 0 0.5 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5 1 0.5 0 0 0.5

Continuous Time: Orthogonal Time Signal Set

k (t ) :

k = , 2, 1, 0 , 1, 2, , 0t <T k = , 2, 1, 0 , 1, 2, , 0t <T 0 if n m T if n = m

2 k (t ) = exp( j k t ) : T
T

(t ) (t ) dt =
m

Discrete Time: Orthogonal Time Signal Set


k (n ) :
k = ..., 2, 1, 0 , 1, 2, ...., 0 n N 1 k = 0 ,1 , 2, ...., N 1 0 nT < NT k = 0 , 1, 2, ...., N 1 0n<N = 0 if n m N if n = m

2 k (n ) = exp( j k nT ) : NT 2 = exp( j k n) : N
(n ) n (n ) m n =0 N 1

NOTE : k (n ) = N + k (n ) k (n ) = N k (n )

OFDM Modulator

OFDM Demodulator

OFDM is a Block Process

Adjacent Symbol Interference (ASI) Symbol Smearing Due to Channel

Guard Interval Inserted Between Adjacent Symbols to Suppress ASI

Cyclic Prefix Inserted in Guard Interval to Suppress Adjacent Channel Interference (ACI)

Data Length Defines Sinc Width: Spectral Spacing Matches Width

Extended Data Length Reduces Sinc Width: Spectral Spacing Preserved

OFDM Symbol: Time and Spectra Channel Input and Output


Real Part of Time Series, Input to Channel 0.6 0.4 0.2 10 5 0 -5 -10 -15 -20 -0.4 -25 40 60 80 100 120 140 160 180 -30 -0.5 0 0.5 Spectrum

0 -0.2

20

Real Part of Time Series, Output of Channel 0.6 0.4 0.2 0 -0.2 -0.4 10 5 0 -5 -10 -15 -20 -25 -30 -0.5

Spectrum

20

40

60

80

100

120

140

160

180

0.5

OFDM Spectra Without and with Cyclic Prefix


OFDM spectral lines With Channel Without Cyclic Prefix 1.5 1 0.5 0 -0.5 -1 -1.5 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

OFDM spectral lines With Channel With Cyclic Prefix 1.5 1 0.5 0 -0.5 -1 -1.5 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

Overlaid Constellations , All Frequencies, Without and With Cyclic Prefix


OFDM Constellations With Channel Without Cyclic Prefix 1.5

OFDM Constellations With Channel With Cyclic Prefix 1.5

0.5

0.5

-0.5

-0.5

-1

-1

-1.5 -1.5

-1

-0.5

0.5

1.5

-1.5 -1.5

-1

-0.5

0.5

1.5

Constellations: Different OFDM Bins Without Cyclic Prefix


1 1 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

Constellations: Different OFDM Bins With Cyclic Prefix


1 1 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

-1 -1 0 1

Channel Estimate with Pilots


Channel, B andwidth, and S am ples 1 0.5 0 -1 1 0.5 0 -1 1 0.5 0 -1 0.01

-0.8

-0.6

-0.4

-0.2

0.2

0.4

0.6

0.8

Zero P ack ed S pec tral S am ples and E xtended Reflection -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

Interpolated S pectral P oints

-0.8

-0.6

-0.4

-0.2

0.2

0.4

0.6

0.8

M agnitude of Interpolation E rror For In-B and Frequencies 0.005

0 -1

-0.8

-0.6

-0.4

-0.2 0 0.2 Norm alized Frequency

0.4

0.6

0.8

DFT (FFT) as Signal Generator for Complex Sinusoids

DFT (FFT) As Signal Analyzer for Complex Sinusoids

H ( k ) = h ( n) e
n =0

N 1

2 nk N

: k = 0, 1, 2, ... , N 1

Radix-2 FFT Flow Diagrams

Input Vector FFT Mapped to Output Time Series, Up-Sampled, Converted Via DAC to Waveform, and I-Q Up-Converted

The FFT as Signal Generator and Interpolator

OFDM Modulation With IFFT and Interpolator

OFDM Demodulation With FFT

OFDM Transceiver

Time and Spectra of Sparse OFDM Symbol


Real Part OFDM Time Series 1 0.5 Amplitude

-0.5

-1

10

20

30

40

50 60 Normalized Time Spectrum

70

80

90

100

10 0 Log Magnitude (dB) -10 -20 -30 -40 -50 -60 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 Normalized Frequency 0.2 0.3 0.4 0.5

Time and Spectra With Frequency Offset = 0.1 Bin


Real Part OFDM Time Series with Offset Frequency = 0.1 Bin Width 1

0.5 Amplitude

-0.5

-1

10

20

30

40

50 Normalized Time

60

70

80

90

100

Spectrum With Frequency Offset = 0.1 Bin Width 10 0 -10 -20 -30 -40 -50 -60 -0.5

Log Magnitude (dB)

-0.4

-0.3

-0.2

-0.1

0 0.1 Normalized Frequency

0.2

0.3

0.4

0.5

Time and Spectra With Sample Clock Offset = 1.02 fs


Real Part OFDM Time Series with Sampling Clock = 1.02 fs 1

0.5 Amplitude

-0.5

-1

10

20

30

40

50 Normalized Time

60

70

80

90

100

Spectrum With Sampling Clock = 1.02 fs 10 0 -10 -20 -30 -40 -50 -60 -0.5

Log Magnitude (dB)

-0.4

-0.3

-0.2

-0.1

0 0.1 Normalized Frequency

0.2

0.3

0.4

0.5

Time and Spectra With Sample Clock Offset = 0.98 fs


Real Part OFDM Time Series with Sampling Clock = 0.98 fs 1

0.5 Amplitude

-0.5

-1

10

20

30

40

50 Normalized Time

60

70

80

90

100

Spectrum With Sampling Clock = 0.98 fs 10 0 -10 -20 -30 -40 -50 -60 -0.5

Log Magnitude (dB)

-0.4

-0.3

-0.2

-0.1

0 0.1 Normalized Frequency

0.2

0.3

0.4

0.5

Ideal I-Q Up and Down Conversion


cos(0t) I(t) S hape CHANNEL Q(t) S hape -sin(0t) n(t) -sin(0t) Match ^ Q(t) cos(0t) ^ Match I(t)

Spectral and Time Description of Real Sinusoids

Complex Sinusoids-I

Complex Sinusoids-II

Complex Baseband and Complex Band-Centered Spectra

Complex Baseband and Real Band-Centered Spectra

Complex Down Conversion

Gain and Phase Imbalance in I-Q Mixers

Spectral Image Due to Gain Imbalance

Spectral Image Due to Phase Imbalance

Line Spectral Images Due to I-Q Mismatch

Coupling Between Positive and Negative FFT Indices Due to I-Q Imbalance and First Order Correction Mechanism

(1 j ) ( j ) G( k ) 2 2 2 H (k ) G( k ) = H ( k ) ( j ) (1 j ) 2 2 2 (1 + j ) ( + + j ) H (k ) 2 2 2 G( k ) H ( k ) G ( k ) ( + + j ) (1 + j ) 2 2 2

Test Bench: Demonstration of Receiver I-Q Imbalances, Carrier Offset, and Timing Offset

Carrier Offset: 4% of FFT Bin Width

Timing Offset: 10% of Sampling Time Period

Timing Clock Offset: 5% of Sampling Time Period per Frame

Gain Imbalance: 10% Error

Phase Imbalance: 0.1 Radian Error

I-Q Mixer Imbalance; 20% Gain, 0.2 Radians

Differential Delay to I/Q Mixers, 10% of Sample Interval

Periodic Time Segments in OFDM Frame Obtained by Zero Packing Spectrum

Probe Mismatch During Short Repeated Preamble

Power Amplifier Non-Linearity


Nonlinear Transfer Function of Amplifier 4 3.5 3 2.5 2 1.5 1-dB Compression Point 1 0.5 0 0 1 2 3 4 -1 -1.5 -2 0 2 4 6 8 10 2 1.5 1 0.5 0 -0.5 Input and Output of Non-Linear Amplifier

Spectrum of Two Input Sinusoids 10 0 -10 -20 -30 -40 -50 -60 -0.5 10 0 -10 -20 -30 -40 -50 -60 -0.5

Spectrum of Two Output Sinusoids

0 Normalized Frequency

0.5

0 Normalized Frequency

0.5

16-QAM Input and Output Envelopes. Saturation and 1-dB Compression Circles
Saturation at 2-Times RMS Signal Level

Envelope at Input to Amplifier 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2 -21-dB -1 Compression 0 1

Saturation 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2 2

Envelope at Output of Amplifier Saturation

-21-dB -1 Compression

Limiting Amplifier Effect on Received QAM Constellation

Matched Filter Applied to Input of Amplifier 1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 -1 -0.5 0 0.5 1 1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 -1

Matched Filter Applied to Output of Amplifier

-0.5

0.5

Limiting Amplifier Effect on Signal Spectra


Spectrum at Input to Amplifier 10 0 -10 Log Magnitude (dB) -20 -30 -40 -50 -60 -4 -3 -2 -1 0 1 Normalized Frequency (f/fsym) 2 3 4

Spectrum at Output of Amplifier 10 0 -10 Log Magnitude (dB) -20 SPECTRAL REGROWTH -30 -40 -50 -60 -4 -3 -2 -1 0 1 Normalized Frequency (f/fsym) 2 3 4

16-QAM (=0.2) Envelope Statistics


16-QAM Histogram at Amplifier Input 0.016 0.014 0.012 0.01 0.008 0.006 0.004 0.002 0 0 0.5 1 1.5 Normalized Amplitude (x/ )
x
-1

10

Probabilty of Level Crossing

10

10

-2

2.5 10
-3

16-QAM Histogram at Amplifier Output 0.016 0.014 0.012 0.01 0.008 0.006 0.004 0.002 0 0 0.5 1 1.5 Normalized Amplitude (x/ x ) 2 2.5 10
-6

std dev =1.03

clip level 10
-4

10

-5

0.5

1 1.5 2 Normalized Amplitude (x/ x )

2.5

OFDM Input and Output Envelopes: Saturation and 1-dB Compression Circles
Saturation at 2-Times RMS Signal Level

Envelope at Input to Amplifier 3 Saturation 2 2 3

Envelope at Output of Amplifier Saturation

-1

-1

-2 1-dB Compression -3 -2 -1

-2 1-dB Compression -2 -1

-3

-3 0 1 2 3 -3

Limiting Amplifier Effect on OFDM Constellation

OFDM Constellation at Input to Amplifier 1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 -1 -0.5 0 0.5 1 1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 -1

OFDM Constellation at Output of Amplifier

-0.5

0.5

OFDM Envelope Statistics


OFDM Histogram at Amplifier Input 0.015 10
0

Probabilty of Level Crossing

0.01

10

-1

0.005 10
-2

1 2 3 Normalized Amplitude (x/ )


x

4 10
-3

OFDM Histogram at Amplifier Output 0.015 std dev =1.06 clip level 10 0.01
-4

10 0.005

-5

1 2 3 Normalized Amplitude (x/ )


x

10

-6

1 2 3 Normalized Amplitude (x/ )


x

OFDM Envelope Statistics with Selected Alternate Mapping


OFDM Histogram: One FFT 0.02 0.015 0.01 10 0.005 0
-1

10

Probabilty of Level Crossing

4 10
-2

OFDM Histogram: Two FFTs 0.02 0.015 0.01 0.005 0

10

-3

Two FFTs

One FFT

Four FFTs 10
-4

OFDM Histogram: Four FFTs 0.02 0.015 10 0.01 0.005 0 10


-6 -5

1 2 3 Normalized Amplitude (x/ x )

1 2 3 Normalized Amplitude (x/ x )

Clipping

Smart Clipping

Reserve Frequency Bins Form Clipping Pulses

Selecting Reserve Frequency Bins


Spectrum 11-Adjacent Frequencies 1 0.8 0.6 0.4 0.2 0 -60 -40 -20 0 20 40 60 1 0.8 0.6 0.4 0.2 0 -0.5 0 Time Series for 11-Equally Spaced Frequencies 1 0.8 0.6 0.4 0.2 -40 -20 0 20 40 60 0 -0.5 0 Time Series for 11-Randomly Spaced Frequencies 1 0.8 0.6 0.4 0.2 -40 -20 0 20 40 60 0 -0.5 0 0.5 0.5 0.5 Time Series for 11-Adjacent Frequencies

Spectrum 11-Equally Spaced Frequencies 1 0.8 0.6 0.4 0.2 0 -60

Spectrum 11-Randomly Spaced Frequencies 1 0.8 0.6 0.4 0.2 0 -60

Reserve Bin Canceller Clipping at 2.5 (8 dB)


Peak envelope, input to PAR control 5 5 data clip level data std dev average peak Peak envelope, output of first pass PAR control

50

100

150

200

250

50

100

150

200

250

Peak envelope, output of second pass PAR control 5 data clip level data std dev average peak 5

Peak envelope, output of third pass PAR control data clip level data std dev average peak

50

100

150

200

250

50

100

150

200

250

Statistics for Clip at 2.5 (8 dB)


input histogram 0.025 10
0

prob of level crossing input pass-1 pass-2 pass-3

0.02

0.015

10

-1

0.01 average =-0.648 dB 0.005 10


-2

4 10
-3

output histogram 0.025 std dev =0.928

0.02

10

-4

0.015

0.01 clip level 0.005

10

-5

10

-6

-5

5 PAR (dB)

10

Reserve Bin Canceller Clipping at 2.2 (6.9 dB)


Peak envelope, input to PAR control 5 5 data clip level data std dev average peak Peak envelope, output of first pass PAR control 4 4

50

100

150

200

250

50

100

150

200

250

Peak envelope, output of second pass PAR control 5 data clip level data std dev average peak 5

Peak envelope, output of third pass PAR control data clip level data std dev average peak

50

100

150

200

250

50

100

150

200

250

Statistics for Clip at 2.2 (6.9 dB)


input histogram 0.025 10
0

prob of level crossing input pass-1 pass-2 pass-3

0.02

0.015

10

-1

0.01 average =-0.653 dB 0.005 10


-2

4 10
-3

output histogram 0.025 std dev =0.928 0.02

10

-4

0.015

0.01 clip level 0.005

10

-5

10

-6

-5

5 PAR (dB)

10

Reserve Bin Canceller Clipping at 2.0 (6 dB)


Peak envelope, input to PAR control 5 5 data clip level data std dev average peak Peak envelope, output of first pass PAR control 4 4

50

100

150

200

250

50

100

150

200

250

Peak envelope, output of second pass PAR control 5 data clip level data std dev average peak 5

Peak envelope, output of third pass PAR control data clip level data std dev average peak

50

100

150

200

250

50

100

150

200

250

Statistics for Clip at 2.0 (6 dB)


input histogram 0.025 10
0

prob of level crossing input pass-1 pass-2 pass-3

0.02

0.015

10

-1

0.01 average =-0.659 dB 0.005 10


-2

4 10
-3

output histogram 0.025

0.02

std dev =0.927

10

-4

0.015

0.01 clip level 0.005

10

-5

10

-6

-5

5 PAR (dB)

10

OFDM 802.11a

Time-Frequency Profile of 802.11a Tones Pilot Tones Shown in Yellow

Preamble and Pilot Structure


Short Symbols
Start of Frame Detection Signal Strength Indication Frequency Offset Resolution

Long Symbols
Channel Estimate Fine Time Resolution

Distributed Pilots
Carrier Tracking Sample Clock Tracking

Preamble Time Structure


Magnitude IEEE 802.11a Preamble 0.8 0.6 0.4 0.2 0

0.5

1.5

2.5 Real Part

3.5

4.5

1 0.5 0 -0.5 -1

0.5

1.5

2.5 Imaginary Part

3.5

4.5

0.5

-0.5

0.5

1.5

2.5

3.5

4.5

Detecting Frame Start with Repeated Short Symbols

Signals in Preamble Detector


Envelope of Input Signal 0.8 0.6 0.4 0.2 0 0.8 0.6 0.4 0.2 0 0.8 0.6 Cross Correlation 0.4 0.2 0 0 0.5 1 1.5 2 2.5 Ratio of Cross to Auto Correlation 3 3.5 4 0 0.5 1 1.5 2 2.5 cross and auto correlations of Input Signal 3 3.5 4 0 0.5 1 1.5 2 2.5 Delayed Envelope of Input Signal 3 3.5 4

0.5

Detection Threshold

0.5

1.5

2.5

3.5

Detail of Signal in Preamble Detector


Envelope of Input Signal 0.8 0.6 0.4 0.2 0 0.6 0.8 0.6 0.4 0.2 0 0.6 0.8 0.6 0.4 0.2 Cross Correlation 0 0.6 1 0.7 0.8 0.9 1 1.1 1.2 Ratio of Cross to Auto Correlation 1.3 1.4 1.5 1.6 Auto Correlation 0.7 0.8 0.9 1 1.1 1.2 cross and auto correlations of Input Signal 1.3 1.4 1.5 1.6 0.7 0.8 0.9 1 1.1 1.2 Delayed Envelope of Input Signal 1.3 1.4 1.5 1.6

0.5

Detection Threshold

0 0.6

0.7

0.8

0.9

1.1

1.2

1.3

1.4

1.5

1.6

Maximum Likelihood Estimator for Frequency Offset

Frequency and Signal Strength Estimates


Estimate of Frequency Offset 10 Frequency Offset in FFT Bins 5

Known Offset 3.3 FFT Bins


0

-5

-10

0.5

1.5

2 Sample Time

2.5

3.5

Estimate of Signal Strength 0.1

0.08

Mag Square

0.06

0.04

0.02

0.5

1.5

2 Sample Time

2.5

3.5

Cross Correlation of Long Preamble


Cross Correlation of Input Signal With Long Preamble Section 14 12 10 Amplitude 8 6 4 2 0

3 Time Samples

Zoom to First Correlation Peak 15 15

Zoom to Second Correlation Peak

10 Amplitude Amplitude 5 Expected Peak Position

10

Expected Peak Position

0 4.3

4.4

4.5 4.6 Time Samples

4.7

4.8

0 5.3

5.4

5.5 5.6 Time Samples

5.7

5.8

Clipped Cross Correlation of Long Preamble


Clipped Cross Correlation of Input Signal With Long Preamble Section 50

40

Amplitude

30

Clipped Correlator Replica Signal Clipped Version of Template Signal Sign[Real(Template)]+j*sign[Imag(Template)]

20

10

100

200

300

400 500 Time Samples

600

700

800

900

Zoom to First Correlation Peak 50 50

Zoom to Second Correlation Peak

40

40

Amplitude

20

Amplitude 4.4 4.5 4.6 Time Samples 4.7 4.8

30

30

20

10

10

0 4.3

0 5.3

5.4

5.5 5.6 Time Samples

5.7

5.8

Channel Probe With Long Preamble


Channel Probe, Long Segment of Preamble 7 6 5 Amplitude 4 3 2 1 0 -0.5 Amplitude 7 6 5 4 3 2 1 0 -0.5 Response of Channel Probe, One Look

0 Normalized Frequency

0.5

0 Normalized Frequency

0.5

Response of Channel Probe, No Noise 7 6 5 Amplitude 4 3 2 1 0 -0.5 Amplitude 0 Normalized Frequency 0.5 7 6 5 4 3 2 1 0 -0.5

Response of Channel Probe, Average of Two Looks

0 Normalized Frequency

0.5

Constellation with Residual Carrier Offset


5000 Constellations, Zero Carrier Frequency Offset 1.5 1 0.5 0 -0.5 -1 -1.5 -1.5 400 pilot, Zero Carrier Frequency Offset 1.5 1 0.5 0 -0.5 -1 -1.5 -1.5

-1

-0.5

0.5

1.5

-1

-0.5

0.5

1.5

5000 Constellations, 5 ppm Carrier Frequency Offset 1.5 1 0.5 0 -0.5 -1

400 pilot, 5 ppm Carrier Frequency Offset 1.5 1 0.5 0 -0.5 -1

-1.5 -1.5

-1

-0.5

0.5

1.5

-1.5 -1.5

-1

-0.5

0.5

1.5

Frequency Domain Residual Carrier Offset


60 Frames Real Part FFT: Zero Carrier Frequency Offset 1.5 1 0.5 Amplitude 0 -0.5 -1 -1.5 -30

-20

-10

0 Frequency Index

10

20

30

60 Frames Real Part FFT: 5 ppm Carrier Frequency Offset 1.5 1 0.5 Amplitude 0 -0.5 -1

-1.5 -30

-20

-10

0 Frequency Index

10

20

30

Constellations with Sample Clock Offset


5000 Constellations, Zero Clock Frequency Offset 1.5 1 0.5 0 -0.5 -1 -1.5 -1.5 400 pilot, Zero Clock Frequency Offset 1.5 1 0.5 0 -0.5 -1 -1.5 -1.5

-1

-0.5

0.5

1.5

-1

-0.5

0.5

1.5

5000 Constellations, 300 ppm Clock Frequency Offset 1.5 1 0.5 0 -0.5 -1

400 pilot, 300 ppm Clock Frequency Offset 1.5 1 0.5 0 -0.5 -1

-1.5 -1.5

-1

-0.5

0.5

1.5

-1.5 -1.5

-1

-0.5

0.5

1.5

Frequency Domain With Sample Clock Offset


80 Frames Real Part FFT: Zero Clock Frequency Offset 1.5 1 0.5 Amplitude 0 -0.5 -1 -1.5 -30

-20

-10

0 Frequency Index

10

20

30

80 Frames Real Part FFT: 300 ppm Clock Frequency Offset 1.5 1 0.5 Amplitude 0 -0.5 -1

-1.5 -30

-20

-10

0 Frequency Index

10

20

30

Other Variants of OFDM


Amplitude and Phase Overlays

Shaped OFDM OQAM OFDM Coded OFDM CI OFDM

Shape to Control Spectral Side Lobes

Overlapped OFDM Frames

Polyphase Filter For Shaped OFDM Shaping and Matched Filter

Impulse Response of Shaped OFDM Modulator and Demodulator


1 0.5 0 0 1 Impulse Response at Output of IFFT 0.5 0 0 1 0.5 0 0 1.5 1 0.5 0 0 1 Impulse Response at Output of FFT 0.5 0 0 1 2 3 4 5 6 7 8 1 2 3 4 5 1 2 3 4 5 6 7 8 Impulse Response at Output of Polyphase Shaping Filter 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 Impulse at Input to IFFT (DC-bin)

Impulse Response at Output of Polyphase Matched Filter

Orthogonal: Adjacent Time Slots Non Adjacent Frequency Bins


Impulse Response Shaping Filter 1 0.8 0.6 0.4 0.2 0 -0.2 0 1 2 3 4 5 1 0.8 0.6 0.4 0.2 0 -0.2 0 2 4 6 8 10 Auto Correlation Response

Spectrum: Shaping Filter Centered on IFFT Spectral Bins 0 Adjacent Spectral Bins Correlated Alternate Spectral Bins Not Correlated -20

-40

-60 -4 -3 -2 -1 0 1 2 3 4

Impulse Response Time-Frequency Profile

0.8

0.6

0.4

0.2

0 8 6 OFD M 10
Fram

4 eN u mb er

5 0 2 0 -10 -5
cy frequ en bin

Orthogonality Between Real and Imaginary Part of Shaped OFDM Frequency Bins

Even and Odd Symmetric Wave Shapes from Adjacent Bins are Orthogonal in Shaped OFDM

Symmetry Considerations in Real and Imaginary Components of Offset Shaped OFDM Frames

Offset OFDM

Compare Spectra
Spectrum of Standard OFDM With Cyclic Prefix 0 Log Magnitude (dB)

-20

-40

-60 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 Normalized Frequency 0.2 0.3 0.4 0.5

Spectrum of OFDM/OQAM Without Cyclic Prefix 0 Log Magnitude (dB)

-20

-40

-60 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 Normalized Frequency 0.2 0.3 0.4 0.5

OFDM and Shaped OFDM PAR


Histogram: Standard OFDM 0.016 0.014 0.012 0.01 0.008 0.006 0.004 0.002 0 0 1 2 3 Normalized Amplitude (x/ x ) 4 Shaped OFDM/OQAM 10
-2

10

Probabilty of Level Crossing

10

-1

Histogram: Shaped OFDM/OQAM 10 0.016 0.014 0.012 0.01 0.008 0.006 0.004 0.002 0 0 1 2 3 Normalized Amplitude (x/ x ) 4 0 10
-4 -3

Standard OFDM

1 2 3 Normalized Amplitude (x/ x )

Complementary Codes

Canceling Correlation Side Lobes

Inserting CC in OFDM
AN (n ) AN (n ) + AN (n ) AN (n ) = 2N (n ) AN ( ) + BN ( ) = 2N (A Constant Power Level)
2 2

Since Sample Values AN (n ) are equal to 1 Average Power in AN ( )


2

=N 2N =2 N

Thus Peak to Average Power Ratio

Now Reverse Domains Use Complementary Code Sequence as amplitude of Carriers in Frequency Domain Then time series has Peak Squared Magnitude = 2N Average Magnitude = N for Peak to Average Power Ratio = 2

CC and Digital Filters

Equivalent Phase Coding

PAR in CCK OFDM and Standard OFDM


Mag Square of CC Rate 1/2 OFDM 2.5 2 1.5 Mean = 1 1 0.5 0 Peak = 2

500

1000

1500

2000

2500

3000

3500

Mag Square of Coded Rate 1/2 OFDM 8 Peak = 6.26

Mean = 1

500

1000

1500

2000

2500

3000

3500

OFDM and CC-OFDM PAR

C-I OFDM Carrier Interferometry


OFDM with Phase Overlay In Conventional OFDM Rectangle Envelope in Time Dirichlet Kernel in Frequency In CI-OFDM Rectangle Envelope in Frequency Dirichlet Kernel in Time Sin(x)/x in Time Domain Without Excess Bandwidth, No Square-Root Nyquist Shaping Filter

Frequency Domain Phase Slope in Continuous and in Sampled Data Domains

Circularly Shifted Time Domain Dirichlet Kernels


1 0.5 0 1 0.5 0 1 0.5 0 1 0.5 0 1 0.5 0

exp(-j 2 k 1/N) exp(-j 2 k 0/N) exp(-j 2 k 2/N) exp(-j 2 k 4/N) exp(-j 2 k 3/N) -0.5 -0.5 -0.5 -0.5 -0.5 -10 0 10 -10 0 10 -10 0 10 -10 0 10 -10 0 10

1 0.5 0

1 0.5 0

1 0.5 0

1 0.5 0

1 0.5 0

exp(-j 2 k 9/N) exp(-j 2 k 7/N) exp(-j 2 k 8/N) exp(-j 2 k 5/N) exp(-j 2 k 6/N) -0.5 -0.5 -0.5 -0.5 -0.5 -10 0 10 -10 0 10 -10 0 10 -10 0 10 -10 0 10

1 0.5 0

1 0.5 0

1 0.5 0

1 0.5 0

1 0.5 0

exp(-j 2 k 10/N) exp(-j 2 k 11/N) exp(-j 2 k 12/N) exp(-j 2 k 13/N) exp(-j 2 k 14/N) -0.5 -0.5 -0.5 -0.5 -0.5 -10 0 10 -10 0 10 -10 0 10 -10 0 10 -10 0 10

Linear Versus Circular Convolution

Fast Circular Convolution with the FFT

Single Symbol in CI-OFDM

M-Symbols in CI-OFDM

1-to-2 Interpolated Time Domain Data Points


CI-OFDM Real Time Series and 1-Modulation Sample 1 0.8 Amplitude 0.6 0.4 0.2 0 -0.2 -20 -15 -10 -5 0 Time 5 10 15 20

CI-OFDM Real Time Series and 4-Modulation Samples 1 0.8 Amplitude 0.6 0.4 0.2 0 -0.2 -20 -15 -10 -5 0 Time 5 10 15 20

CI-OFDM Data Frame


CI-OFDM Real Time Series and 32-Modulation Samples 2 1.5 1 Amplitude 0.5 0 -0.5 -1 -1.5 -2 -25 -20 -15 -10 -5 Time 0 5 10 15 20 Cyclic Prefix

Spectrum 10 0 -10 -20 -30 -40 -50 -60 -0.5

Log Magnitude (dB)

-0.4

-0.3

-0.2

-0.1

0 0.1 Normalized Frequency

0.2

0.3

0.4

0.5

CI-OFDM Statistics
Histogram: Standard OFDM (QPSK) 10 0.015
0

Probabilty of Level Crossing

0.01

10

-1

0.005
-2

10 0 1 2 3 Normalized Amplitude (x/ x ) Histogram: CI-OFDM (QPSK) 10 4

-3

0.1 10 0.05
-4

1 2 3 Normalized Amplitude (x/ x )

1 2 3 Normalized Amplitude (x/ x )

CI-OFDM Statistics
Histogram: Standard OFDM 10 0.015
0

Probabilty of Level Crossing

0.01

10

-1

0.005
-2

10 0 1 2 3 Normalized Amplitude (x/ x ) Histogram: CI-OFDM (16-QAM) 4

0.08

10

-3

0.06

0.04

10

-4

0.02

1 2 3 Normalized Amplitude (x/ x )

1 2 3 Normalized Amplitude (x/ x )

Thats all Folks

Professor harris, may I be excused? My brain is full.