Outline

Introduction
Fundamentals of OFDM transmissions
Possible research problems in OFDM transmissions

Chapter 3: Physical-layer transmission techniques

Section 3.4: Orthogonal Frequency Division Multiplexing

(OFDM) Fundamentals

Faculty of Electronics & Telecommunications Engineering

University of Science and Technology, The University of Danang

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 1

 Outline
Introduction
Fundamentals of OFDM transmissions
Possible research problems in OFDM transmissions

1 Introduction
Development of mobile communications systems - Revisit
Mobile broadband technology evolution - Revisit

2 Fundamentals of OFDM transmissions

Functional blocks in OFDM transmitter and receiver
Transmitted OFDM signals
Time-variant multipath channels
Received OFDM signals
MQAM symbol detection in the frequency domain

3 Possible research problems in OFDM transmissions

Carrier frequency offset (CFO) estimation
Doubly selective channel estimation

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 2

 Outline
Introduction Development of mobile communications systems - Revisit
Fundamentals of OFDM transmissions Mobile broadband technology evolution - Revisit
Possible research problems in OFDM transmissions

Development of mobile communications systems - Revisit

~ 1 Gbps (stationary),
~ 100 Mbps (mobile)
~ 14 Mbps (downlink),
~ 5.8 Mbps (uplink)
~ 50 Kbps code

code
space

frequency
frequency frequency
frequency
time

time
time

OFDM, SDMA (4G)

FDMA (1G) TDMA (2G) CDMA (3G) e.g., WiMAX, LTE
e.g., AMPS ~ 1980s e.g., GSM ~ 1990s e.g., W-CDMA ~ 2000s
2010s

Orthogonal multiple access is still exploited in 5G!

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 3

 Outline
Introduction Development of mobile communications systems - Revisit
Fundamentals of OFDM transmissions Mobile broadband technology evolution - Revisit
Possible research problems in OFDM transmissions

Development of mobile communications systems - Revisit

~ 1 Gbps (stationary),
~ 100 Mbps (mobile)
~ 14 Mbps (downlink),
~ 5.8 Mbps (uplink)
~ 50 Kbps code

code
space

frequency
frequency frequency
frequency
time

time
time

OFDM, SDMA (4G)

FDMA (1G) TDMA (2G) CDMA (3G) e.g., WiMAX, LTE
e.g., AMPS ~ 1980s e.g., GSM ~ 1990s e.g., W-CDMA ~ 2000s
2010s

Orthogonal multiple access is still exploited in 5G!

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 3

 Outline
Introduction Development of mobile communications systems - Revisit
Fundamentals of OFDM transmissions Mobile broadband technology evolution - Revisit
Possible research problems in OFDM transmissions

Mobile broadband technology evolution - Revisit

2005 2006 2007 2008 2009 2010 2011 2012

WCDMA LTE
R99 HSDPA HSPA LTE
3GPP Advanced

“4G”
CDMA 2000 EVDO EVDO IMT-Advanced
3GPP2 Rev A Rev B

Mobile 802.16e 802.16e 802.16m

WiMAX R 1.0 R 1.5 R 2.0

CDMA based OFDMA based

The last decade has witnessed numerous intensive studies in employing

orthogonal frequency division multiplexing (OFDM) for the emerging
broadband communications systems (e.g., WiFi, WiMAX, LTE) to exploit
its high spectral efficiency and robustness against multipath (frequency
selective) fading channels.
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 4
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Functional blocks in OFDM transmitter and receiver

Clk RF
Pilot insertion Osc LO

ci
Conv. MQAM S/P Insert CP P/S
mapping
IFFT DAC RF
Encoder

Burst structure
Information
bits, ui

Data OFDM Pilot OFDM

symbols symbols

RF Clk
LO Osc

ADC MQAM
RF CP S/P FFT Demapper
-1 decoder Information
bits, ui

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 5

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Transmitted OFDM signals

Consider a coded OFDM system using M -ary quadrature amplitude
modulation (M-QAM).
The encoded bit stream is bit-interleaved. Then, the resulting
sequence of interleaved bits is organized as a sequence of Q-bit
tuples {dk,m } where Q = log2 M and
T
dk,m = [dk,m,0 , ..., dk,m,Q−1 ] .
The sequence is further mapped to a complex-valued symbol
Xk,m ∈ A where A is the M -ary modulation signaling alphabet, and
m and k denote the indices of OFDM symbol and subcarrier,
respectively.
The m-th OFDM symbol
X 0,m X k ,m XN 1, m X 0, m 1 X k ,m 1 X N !1, m 1

d k , m ! [d k , m,0 ,..., d k , m,Q 1 ]T The k-th subcarrier

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 6

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Transmitted OFDM signals

Consider a coded OFDM system using M -ary quadrature amplitude
modulation (M-QAM).
The encoded bit stream is bit-interleaved. Then, the resulting
sequence of interleaved bits is organized as a sequence of Q-bit
tuples {dk,m } where Q = log2 M and
T
dk,m = [dk,m,0 , ..., dk,m,Q−1 ] .
The sequence is further mapped to a complex-valued symbol
Xk,m ∈ A where A is the M -ary modulation signaling alphabet, and
m and k denote the indices of OFDM symbol and subcarrier,
respectively.
The m-th OFDM symbol
X 0,m X k ,m XN 1, m X 0, m 1 X k ,m 1 X N !1, m 1

d k , m ! [d k , m,0 ,..., d k , m,Q 1 ]T The k-th subcarrier

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 6

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Transmitted OFDM signals (cont.)

Each OFDM symbol consists of N information bearing subcarriers
Xk,m , k = 0, ..., N − 1, where N is the size of the fast Fourier transform
(FFT) and inverse-FFT (IFFT) used in the multicarrier transmission.

CP samples

P/S DAC RF
IFFT
QAM symbols
X k ,m
xn , m

Clk RF
Pilot insertion Osc LO

ci
Conv. MQAM S/P Insert CP P/S
mapping
IFFT DAC RF
Encoder

Burst structure
Information
bits, ui
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 7
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Transmitted OFDM signals (cont.)

CP portion one OFDM symbol of samples

xN N g , m …. xN 1, m x0, m x1, m x2, m …. xN N g 1, m xN N g , m …. xN 1, m

Copy and paste the last Ng samples

After inverse FFT (IFFT) and cyclic prefix (CP) insertion, the
transmitted baseband signal of the mth OFDM symbol can be written as
N −1  
1 X j2πkn
xn,m = √ Xk,m exp , (1)
N k=0 N

where n ∈ {−Ng , ..., 0, ..., N − 1}, Ng denotes the CP length.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 8

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Transmitted OFDM signals (cont.)

CP portion one OFDM symbol of samples

xN N g , m …. xN 1, m x0, m x1, m x2, m …. xN N g 1, m xN N g , m …. xN 1, m

Copy and paste the last Ng samples

After inverse FFT (IFFT) and cyclic prefix (CP) insertion, the
transmitted baseband signal of the mth OFDM symbol can be written as
N −1  
1 X j2πkn
xn,m = √ Xk,m exp , (1)
N k=0 N

where n ∈ {−Ng , ..., 0, ..., N − 1}, Ng denotes the CP length.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 8

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Time-variant multipath channels

Local
Scattering
Two Main
Multipaths

P
ym = l xm−l hl,m + wb (m/W ) (2)

The lth (time-variant) channel tap gain that includes the effect of
transmit-receive filters and time-variant channel propagation is denoted
by hl,n,m where n and m stand for the indices of the time-domain sample
and OFDM symbol, respectively.
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 9
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Time-variant multipath channels (cont.)

In the time-variant multipath channels, a number of basis expansion
models (BEMs) can be employed for reducing the number of channel
parameters while capturing the time-variation of the channels.
Using BEMs, hl,n,m (after CP removal) can be approximated by
Q
X
hl,n,m = bn+Ng +mNs ,q cq,l , l ∈ {0, ..., L − 1}, (3)
q=1

where Ns = N + Ng denotes the OFDM symbol length after CP

insertion, n = 0, ..., N − 1, m = 0, ..., M − 1 and M is the number
of both data and pilot OFDM symbols in a burst. bn+Ng +mNs ,q
stand for the qth basis function values of the used BEM.
CP portion one OFDM symbol of samples

xN N g , m …. xN 1, m x0, m x1, m x2, m …. xN N g 1, m xN N g , m …. xN 1, m

Copy and paste the last Ng samples

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 10

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Time-variant multipath channels (cont.)

In the time-variant multipath channels, a number of basis expansion
models (BEMs) can be employed for reducing the number of channel
parameters while capturing the time-variation of the channels.
Using BEMs, hl,n,m (after CP removal) can be approximated by
Q
X
hl,n,m = bn+Ng +mNs ,q cq,l , l ∈ {0, ..., L − 1}, (3)
q=1

where Ns = N + Ng denotes the OFDM symbol length after CP

insertion, n = 0, ..., N − 1, m = 0, ..., M − 1 and M is the number
of both data and pilot OFDM symbols in a burst. bn+Ng +mNs ,q
stand for the qth basis function values of the used BEM.
CP portion one OFDM symbol of samples

xN N g , m …. xN 1, m x0, m x1, m x2, m …. xN N g 1, m xN N g , m …. xN 1, m

Copy and paste the last Ng samples

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 10

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Time-variant multipath channels (cont.)

It is noted that the use of BEMs offers a significant dimension
reduction in the time-variant channel representation, i.e., Q ≪ N M .
For instance, in the current LTE system settings, three LTE time
slots contain 21 OFDM symbols with 128-FFT (the smallest FFT
size used in the LTE settings) and the resulting number of the
time-variant channel parameters corresponding to one channel tap
gain hl,n,m in (3) will be N M = 128 × 21 = 2688.
By using the discrete prolate spheroidal (DPS)-BEM as shown in (3),
the number of the basis functions Q can vary from 3 (for low user
speeds, e.g., about 10km/h) to 5 (for moderate user speeds, e.g.,
about 100km/h) under the required MSE of the DPS-BEM-based
channel approximation below 10−10 as shown in Fig. 1.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 11

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Time-variant multipath channels (cont.)

It is noted that the use of BEMs offers a significant dimension
reduction in the time-variant channel representation, i.e., Q ≪ N M .
For instance, in the current LTE system settings, three LTE time
slots contain 21 OFDM symbols with 128-FFT (the smallest FFT
size used in the LTE settings) and the resulting number of the
time-variant channel parameters corresponding to one channel tap
gain hl,n,m in (3) will be N M = 128 × 21 = 2688.
By using the discrete prolate spheroidal (DPS)-BEM as shown in (3),
the number of the basis functions Q can vary from 3 (for low user
speeds, e.g., about 10km/h) to 5 (for moderate user speeds, e.g.,
about 100km/h) under the required MSE of the DPS-BEM-based
channel approximation below 10−10 as shown in Fig. 1.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 11

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Time-variant multipath channels (cont.)

It is noted that the use of BEMs offers a significant dimension
reduction in the time-variant channel representation, i.e., Q ≪ N M .
For instance, in the current LTE system settings, three LTE time
slots contain 21 OFDM symbols with 128-FFT (the smallest FFT
size used in the LTE settings) and the resulting number of the
time-variant channel parameters corresponding to one channel tap
gain hl,n,m in (3) will be N M = 128 × 21 = 2688.
By using the discrete prolate spheroidal (DPS)-BEM as shown in (3),
the number of the basis functions Q can vary from 3 (for low user
speeds, e.g., about 10km/h) to 5 (for moderate user speeds, e.g.,
about 100km/h) under the required MSE of the DPS-BEM-based
channel approximation below 10−10 as shown in Fig. 1.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 11

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Numerical results of BEM-based channel fitting

0
10
1 km/h
5 km/h
MSE of DPS−BEM−based channel fitting

10 km/h
−5
10 20 km/h
30 km/h
40 km/h
−10 50 km/h
10 60 km/h
70 km/h
80 km/h
−15 90 km/h
10 100 km/h
120 km/h
140 km/h
−20 160 km/h
10 180 km/h
200 km/h

−25
10

−30
10
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Number of used basis functions

Figure 1: Normalized MSE of DPS-BEM-based approximation (fitting) of

time-varying channels generated by Jakes model under different mobile speeds.
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 12
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Received OFDM signals

Over the above doubly-selective channels, after CP removal, the nth
received sample in the mth OFDM symbol can be represented by
L−1
X
yn,m = hl,n,m xn−l,m + zn,m , (4)
l=0
where n = 0, ..., N − 1 and zn,m is the additive white Gaussian noise
(AWGN) with variance No .
Channel impulse response

hN g ,0, m …. h1,0,m h0,0,m

xN N g , m …. xN 1, m x0, m x1, m x2, m …. xN N g 1, m xN N g , m …. xN 1, m

one OFDM symbol of samples

CP portion

y0,m ! hN g ,0,m xN " N g ,m h1,0, m xN "1, m h0,0, m x0,m

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 13
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Received OFDM signals in the frequency domain

At the receiver, after performing CP removal and FFT, the received
samples in the frequency domain can be determined by
N −1
1 X
Yk,m = √ yn,m e−j2πkn/N = Hk,m Xk,m + ρk,m + Zk,m , (5)
N n=0
PN −1 PL−1
where Hk,m = N1 n=0 l=0 hl,n,m e−j2πkl/N ,
Zk,m = √1N n=0 zn,m e(−j N ) and ρk,m =
PN −1 2πkn

PN −1 
PN −1 PL−1 
1 −j2πil/N
N i=0,i̸=k Xi,m n=0 l=0 hl,n,m e ej2πn(i−k)/N .
It is noted that the received samples in the time domain are:
L−1
X
yn,m = hl,n,m xn−l,m + zn,m .
l=0

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 14

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Received OFDM signals in the frequency domain

At the receiver, after performing CP removal and FFT, the received
samples in the frequency domain can be determined by
N −1
1 X
Yk,m = √ yn,m e−j2πkn/N = Hk,m Xk,m + ρk,m + Zk,m , (5)
N n=0
PN −1 PL−1
where Hk,m = N1 n=0 l=0 hl,n,m e−j2πkl/N ,
Zk,m = √1N n=0 zn,m e(−j N ) and ρk,m =
PN −1 2πkn

PN −1 
PN −1 PL−1 
1 −j2πil/N
N i=0,i̸=k Xi,m n=0 l=0 hl,n,m e ej2πn(i−k)/N .
It is noted that the received samples in the time domain are:
L−1
X
yn,m = hl,n,m xn−l,m + zn,m .
l=0

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 14

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Received OFDM signals in the frequency domain

At the receiver, after performing CP removal and FFT, the received
samples in the frequency domain can be determined by
N −1
1 X
Yk,m = √ yn,m e−j2πkn/N = Hk,m Xk,m + ρk,m + Zk,m , (5)
N n=0
PN −1 PL−1
where Hk,m = N1 n=0 l=0 hl,n,m e−j2πkl/N ,
Zk,m = √1N n=0 zn,m e(−j N ) and ρk,m =
PN −1 2πkn

PN −1 
PN −1 PL−1 
1 −j2πil/N
N i=0,i̸=k Xi,m n=0 l=0 hl,n,m e ej2πn(i−k)/N .
It is noted that the received samples in the time domain are:
L−1
X
yn,m = hl,n,m xn−l,m + zn,m .
l=0

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 14

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

An example of channel frequency response in time- and

frequency-domains

20

0
Amplitude (dB)

−20

−40

−60
120
100 140
80 120
60 100
80
40 60
20 40
20
Subcarrier index OFDM symbol index

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 15

 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Received OFDM signals over block-fading channel

At the receiver, after performing CP removal and FFT, the received
samples in the frequency domain can be determined by
N −1
1 X
Yk,m = √ yn,m e−j2πkn/N = Hk,m Xk,m + 0 + Zk,m , (6)
N n=0
PL−1
where Hk,m = l=0 hl e−j2πkl/N .
It is noted that the received samples in the time domain are:
L−1
X
yn,m = hl xn−l,m + zn,m .
l=0

Clk RF
Pilot insertion Osc LO

RF Clk
LO Osc
ci
Conv. MQAM S/P Insert CP P/S
mapping
IFFT DAC RF
Encoder
ADC MQAM
RF CP S/P FFT Demapper
-1 decoder Information
bits, ui
Burst structure
Information
bits, ui
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 16
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Received OFDM signals over block-fading channel

At the receiver, after performing CP removal and FFT, the received
samples in the frequency domain can be determined by
N −1
1 X
Yk,m = √ yn,m e−j2πkn/N = Hk,m Xk,m + 0 + Zk,m , (6)
N n=0
PL−1
where Hk,m = l=0 hl e−j2πkl/N .
It is noted that the received samples in the time domain are:
L−1
X
yn,m = hl xn−l,m + zn,m .
l=0

Clk RF
Pilot insertion Osc LO

RF Clk
LO Osc
ci
Conv. MQAM S/P Insert CP P/S
mapping
IFFT DAC RF
Encoder
ADC MQAM
RF CP S/P FFT Demapper
-1 decoder Information
bits, ui
Burst structure
Information
bits, ui
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 16
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Received OFDM signals over block-fading channel

At the receiver, after performing CP removal and FFT, the received
samples in the frequency domain can be determined by
N −1
1 X
Yk,m = √ yn,m e−j2πkn/N = Hk,m Xk,m + 0 + Zk,m , (6)
N n=0
PL−1
where Hk,m = l=0 hl e−j2πkl/N .
It is noted that the received samples in the time domain are:
L−1
X
yn,m = hl xn−l,m + zn,m .
l=0

Clk RF
Pilot insertion Osc LO

RF Clk
LO Osc
ci
Conv. MQAM S/P Insert CP P/S
mapping
IFFT DAC RF
Encoder
ADC MQAM
RF CP S/P FFT Demapper
-1 decoder Information
bits, ui
Burst structure
Information
bits, ui
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 16
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

MQAM symbol detection in the frequency domain

Under maximum likelihood (ML) detection, the transmitted QAM
symbols can be recovered in the frequency domain by
bk,m = arg min |Yk,m − Hk,m Xk,m |2 .
X (7)
Xk,m ∈A

Based on the detected QAM symbols X bk,m , the transmitted data

bits can be recovered accordingly by channel decoders.
It is noted that the ML symbol detection process (7) needs to know
channel frequency response (CFR) Hk,m before providing a detected
version of the transmitted symbol Xbk,m

Possible research problem

In OFDM-based transmissions (e.g., WiFi, WiMAX, LTE systems) over
time-variant multipath channels, the problem of estimating channel
responses (i.e., Hk,m ) is of importance in research and industry as well.
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 17
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

MQAM symbol detection in the frequency domain

Under maximum likelihood (ML) detection, the transmitted QAM
symbols can be recovered in the frequency domain by
bk,m = arg min |Yk,m − Hk,m Xk,m |2 .
X (7)
Xk,m ∈A

Based on the detected QAM symbols X bk,m , the transmitted data

bits can be recovered accordingly by channel decoders.
It is noted that the ML symbol detection process (7) needs to know
channel frequency response (CFR) Hk,m before providing a detected
version of the transmitted symbol Xbk,m

Possible research problem

In OFDM-based transmissions (e.g., WiFi, WiMAX, LTE systems) over
time-variant multipath channels, the problem of estimating channel
responses (i.e., Hk,m ) is of importance in research and industry as well.
Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 17
 Functional blocks in OFDM transmitter and receiver
Outline
Transmitted OFDM signals
Introduction
Time-variant multipath channels
Fundamentals of OFDM transmissions
Received OFDM signals
Possible research problems in OFDM transmissions
MQAM symbol detection in the frequency domain

Downlink multiuser MIMO-OFDM transmissions in

WiMAX systems (IEEE 802.16e)

User 1

FFT Decoder

User 2
MU-MIMO encoder

FFT Decoder
IFFT

User 3

FFT Decoder

Base station User 4

FFT Decoder

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 18

 Outline
Introduction Carrier frequency offset (CFO) estimation
Fundamentals of OFDM transmissions Doubly selective channel estimation
Possible research problems in OFDM transmissions

Carrier frequency offset (CFO) estimation

In the presence of CFO, after CP removal, the nth received sample
in the mth OFDM symbol can be represented by
L−1
2πε
X
yn,m = ej N (n+Ng +mNs ) hl,n,m xn−l,m + zn,m , (8)
l=0

where:
n = 0, ..., N − 1
∆f and ε = ∆f N T denote the absolute and normalized CFOs,
respectively.
T is the sampling period of the system.
As observed in (8), the presence of CFO introduces a time-domain
phase rotation that will translate into ICI in the frequency domain.
In addition, the time-variation of the multipath channels also induces
ICI in the frequency domain.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 19

 Outline
Introduction Carrier frequency offset (CFO) estimation
Fundamentals of OFDM transmissions Doubly selective channel estimation
Possible research problems in OFDM transmissions

Carrier frequency offset (CFO) estimation (cont.)

Consequently, the presence of both CFO and doubly selective

channels would incur a significant ICI power at OFDM receivers,
giving rise to a considerable irreducible error floor in the receiver
performance.
For CFO compensation and reliable coherent data
detection/decoding, the CFO and CIR estimates are indispensible at
OFDM receivers.
To obtain the estimates, CFO and channel estimates can be
obtained either jointly or separately by existing estimation
techniques (e.g., RLS, ML, MAP,...).

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 20

 Outline
Introduction Carrier frequency offset (CFO) estimation
Fundamentals of OFDM transmissions Doubly selective channel estimation
Possible research problems in OFDM transmissions

Doubly selective channel estimation: Introduction

The last decade has witnessed numerous intensive studies in

employing OFDM for broadband communication systems to exploit
its high spectral efficiency and robustness against multipath
(frequency-selective) fading channels.
In the current literature, most of these studies have assumed
frequency-selective channels to be time-invariant (i.e., quasi-static or
block-fading) within a transmission burst.
This channel assumption can be used in a wireless system with
stationary and/or low-speed users.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 21

 Outline
Introduction Carrier frequency offset (CFO) estimation
Fundamentals of OFDM transmissions Doubly selective channel estimation
Possible research problems in OFDM transmissions

Doubly selective channel estimation: Introduction (cont.)

In a wireless network with rapidly moving nodes (e.g., users in cars

and trains in 4G-LTE systems), the resultant time-selectivity of the
channel impulse response (CIR) introduces a large number of
channel parameters (much greater than that of
quasi-static/block-fading channels).
In addition, the time-variation of the channel leads to a loss of
subcarrier orthogonality, resulting in inter-carrier interference (ICI) in
OFDM receivers.
Under such a scenario, the assumption of quasi-static fading
channels becomes inappropriate.
As a result, time- and frequency-selective (doubly selective) channels
should be considered in the wireless system investigation and
analysis.

Wireless Communications - Chapter 3: Physical-layer transmissions Section 3.4: OFDM fundamentals 22