Progress In Electromagnetics Research, PIER 83, 119–131, 2008

PILOT-LESS SYNCHRONIZATION RECEIVER FOR

UWB-BASED WIRELESS APPLICATION

J.-H. Kim and Y.-H. You

uT Communication Research Center
Sejong University
98 Kunja-Dong, Kwangjin-Ku, Seoul 143-747, Korea

K.-I. Lee, K.-S. Jeong, and J.-H. Yi

YuJeong Systems Co. Ltd.
503 Dong-Il B/D, 826 Kuro-3Dong, Kuro-Ku, Seoul 152-879, Korea

Abstract—This paper suggests a pilot-less residual carrier frequency

offset (CFO) estimator for ultra-wideband orthogonal frequency
division multiplexing (UWB-OFDM) systems. The basic idea of our
approach is based on the fact that two adjacent OFDM symbols
carry the identical information in the UWB-OFDM system, thus
removing the need of pilot symbols. To demonstrate the efficiency
of the proposed pilot-less CFO estimator, analytical expression of the
mean square error (MSE) is reported and comparisons are made with
the conventional pilot-aided CFO estimator in terms of MSE and
throughput.

1. INTRODUCTION

The ultra-wideband (UWB) technology is one of the most innovative

and promising wireless means to enable cable-less transmission of high-
speed data between close-by nodes [1–4]. The WiMedia Alliance has
defined the inter-operability standards for UWB communication [5].
The work of the WiMedia Alliance has focused its efforts on providing
increased capability to enable new high-rate wireless applications.
There has been explosive adoption of UWB technology in the various
wireless applications [6–10]. A particularly interesting potential
WiMedia application is short-range wireless video transmission.
Uniquely, the availability of broad spectrum in UWB communication
120 kim et al.

provides an unequalled advantage for high-performance and high-

capacity wireless video networks [6]. Current wireless video solutions
such as Wi-Fi do not have the bandwidth to accommodate high-
definition (HD) or AVC/H.264 video streams, which require a
minimum of 20 Mbps each. UWB can offer data-rates exceeding those
that are required to convey high quality broadcast video and new
features such as watching broadcast quality video over a broadcast-
type wireless link will enable multimedia presentation in education,
libraries, advertising, or have a wireless connection at home [11, 12].
High-rate UWB is an orthogonal frequency division multiplexing
(OFDM) system designed to deliver data rates between 53.3 and
480 Mbps in pico-environments. In the WiMedia UWB system,
frequency hopping (FH) and OFDM are used for high data-rate
transmission, however this combination of FH and OFDM makes
carrier frequency offset (CFO) a considerable problem [13, 14].
Even though most systems compensate the frequency offset with
initial training sequences, there might still be a residual CFO. The
uncompensated residual CFO can cause inter-carrier interference (ICI)
and large signal constellation rotation. So it must be compensated,
otherwise it would lead to degradation in system performance [15, 16].
A number of algorithms to remove the effect of CFO are presented in
the literature using pilot subcarriers [13–20]. In the current WiMedia
UWB system, only a few pilots are embedded in one OFDM symbol
for frequency tracking. As a result, the major challenge in the UWB
systems is to increase frequency estimation accuracy with the limited
pilots.
In this paper, we present a fine synchronization receiver with algo-
rithm for tracking of residual CFO for UWB-based video transmission
applications. In order to design an improved synchronization receiver
without the use of pilot symbols, some modifications on the current
WiMedia UWB system are introduced. It is found that a slight modi-
fication leads to improved estimation performance in comparison with
the traditional pilot-aided synchronization scheme. Since we save the
pilots for synchronization, the throughput of the system is also in-
creased. In order to show the usefulness of the proposed scheme, it is
applied to H.264 based video communication system.
This paper is organized as follows: Section 2 describes the signal
model for the UWB-OFDM system. In Sections 3, a pilot-less fine
frequency estimation algorithm is suggested for UWB-OFDM system
and its performance is reported in terms of mean square error (MSE).
The simulation environment and detailed results are discussed in
Section 4. Finally, conclusion are drawn in Section 5.
Progress In Electromagnetics Research, PIER 83, 2008 121

2. SYSTEM MODEL

Consider the UWB-OFDM system using N points of IFFT and Nzp

zero-padded suffix. The transmitted baseband signal for the n-th
sample of the l-th OFDM symbol xl (n) can be simply expressed as
−Nn /2
1  1 
N/2−1
xl (n) = √ Xl (k)ej2πnk/N + √ Xl (k)ej2πnk/N
N k=−N/2+1
N k=Nn /2
(1)

where Xl (k) is the non-zero symbol transmitted on the k-th subcarrier

and Nn is the number of null subcarriers.
At the receiver, we assume that the symbol timing error is
perfectly compensated. Taking into account the small CFO, the
received OFDM symbol after FFT demodulation becomes [19][21]

sin(π∆)
Rl (k) = Gl Xl (k)ej2π∆l(N +Nzp )/N Hl (k) + Il (k) + Wl (k)
N sin(π∆/N )
(2)

where Gl is the lognormal shadowing term, Hl (k) is the channel’s

frequency response with zero-mean and variance σH 2 incorporating the
time-invariant phase term, Wl (k) is a zero-mean complex Gaussian
noise term with variance σW2 during the l-th symbol period, ∆ is the
CFO relative to the inter carrier spacing, and Il (k) is the ICI term. In
Eq. (2), the ICI caused by small CFO can be omitted, since its power
is very small compared with the additive noise power [21]. So, Eq. (2)
can be simplified into

Rl (k) ≈ Cl (k)Xl (k)ej2π∆l(N +Nzp )/N + Wl (k) (3)

with denoting Cl (k) = Gl Hl (k).

3. PILOT-LESS FINE FREQUENCY

SYNCHRONIZATION RECEIVER

In order to estimate the frequency error without the use of pilot

symbols, a simple way of implementing a fine frequency estimator is
suggested in this section and its estimation performance is analytically
derived.
122 kim et al.

3.1. Synchronization Algorithm

The current UWB-OFDM system provide time domain diversity by
time-domain spreading (TDS) and frequency domain diversity by
frequency domain spreading (FDS). Both FDS and TDS techniques
shall be used when the data unit is encoded at a data rate of 53.3
or 80 Mb/s. When the data unit is encoded at data rate of 106.7,
160 or 200 Mb/s, only TDS technique is adopted. In this paper, we
consider the UWB-OFDM system using only TDS for high-rate data
transmission.
When the UWB-OFDM system is equipped with only TDS, in
order to derive a pilot-less fine frequency estimator, the (l + 1)-th
OFDM symbol is modified into
Xl+1 (k) = Im{Xl (−k)} − jRe{Xl (−k)} (4)
where Re{x} and Im{x} denote the real and imaginary parts of x,
respectively. It is noted from Eq. (4) that xl+1 (k) can be effortlessly
obtained from xl (k) without additional IFFT operation. At the
receiver, the equalized signal is given by

Rl (k) = Rl (k)Ĉl∗ (k) (5)

where Ĉl (k) is the estimate of Cl (k). Using the TDS property which is
provided in the UWB-OFDM system, the fine frequency estimator is
based on post-FFT temporal correlation by using Nu = N − Nn − Ng
non-zero data symbols, where Ng is the number of guard subcarriers.
By using the relation between two consecutive symbols Xl (k) and
Xl+1 (k) in Eq. (4), one can find that Xl∗ (k)Xl+1 (−k) = (−j)|Xl (k)|2 .
When we consider the non-zero signal samples excluding the guard
subcarriers in the UWB-OFDM system, a temporal correlation is
devised to has the form
Tl (k) = jR∗l (k)Rl+1 (−k), k ∈ S0 ∪ S1 (6)
where S0 denotes the set of subcarrier indices in the left half k ∈
[−(N − Ng )/2 + 1, −Nn /2] and S1 denotes the set of indices in the
right half k ∈ [Nn /2, (N − Ng )/2 − 1]. The temporal correlation is
further derived by

Tl (k) = |Gl |4 |Hl (k)Hl+1 (−k)|2 |Xl (k)|2 ej2π∆(N +Nzp )/N + Ŵl (k)
(7)
+ α1 Il 1 (k) + α2 Il 2 (k) + α1 α2 Il 3 (k)

where Ŵl (k) denotes the additive noise component, αi = Ĉl+i−1 (k) −
Cl+i−1 (k), αi Il i (k) is the noise term due to the imperfect channel
Progress In Electromagnetics Research, PIER 83, 2008 123

estimation (CE), and the product of two estimation error terms

α1 α2 Il3 (k) are negligible at the high SNR. In Eq. (7), the noise
contribution Ŵl (k) reads
∗
Ŵl (k) = j|Gl |3 |Hl (k)|2 Hl+1 (−k)Xl∗ (k)Wl+1 (−k)
+ j|Gl |3 |Hl+1 (−k)|2 Hl (k)Xl+1 (−k)Wl∗ (k) (8)
∗
+ j|Gl |2
Hl (k)Hl+1 (−k)Wl∗ (k)Wl+1 (−k),

and CE-induced noise terms {Il i (k)} are given by

Il 1 (k) = |Gl |3 Hl∗ (k)|Hl+1 (−k)|2 |Xl (k)|2

+ j|Gl |2 Hl∗ (k)Hl+1
∗
(−k)Xl∗ (k)Wl+1 (−k)
(9)
+ j|Gl |2 |Hl+1 (−k)|2 Xl+1 (−k)Wl∗ (k)
+ jGl Wl∗ (k)Wl+1 (−k)Hl+1 ∗
(−k),

and
Il 2 (k) = |Gl |3 |Hl (k)|2 Hl+1 (−k)|Xl (k)|2
+ j|Gl |2 Hl (k)Hl+1 (−k)Xl+1 (−k)Wl∗ (k)
+ j|Gl |2 |Hl (k)|2 Xl∗ (k)Wl+1 (−k) + jGl Wl (k)Wl+1
∗
(−k)Hl (k)
(10)

where we omit the phase-rotated term due to ∆ because the statistical

property of Ŵl (k) and Il i (k) is untouched.
Since E{Ŵl (k)} = 0 and E{αi Il i (k)} = 0, it follows that

{E [Tl (k)]} = 2π∆(N + Nzp )/N (11)

where
is the angle operation. For relatively large Nu , the proposed

fine frequency estimator is based on the cumulative phases of Tl (k) on
negative and positive subcarriers given by
 
 
φl,i =
Tl (k) , i = 0, 1. (12)
 
k∈Si

Similar to [18, 19], by using Eq. (12), the fine frequency estimator takes
expression

ˆ = N φl,0 + φl,1
∆ . (13)
2π(N + Nzp ) 2
124 kim et al.

3.2. Performance Analysis

To demonstrate accuracy of the proposed frequency estimator, we
derive the error variance. In an analogy to [18][19], denoting ρ =
ej2π∆(N +Nzp )/N and making use of tan(a + b) ≈ tan(a) + tan(b), we
obtain
   
   
Im Tl (k)ρ Im Tl (k)ρ
   
k∈S0 k∈S1
ˆ
4π(∆ − ∆)(N + Nzp )/N ≈   +  
   
Re Tl (k)ρ Re Tl (k)ρ
   
k∈S0 k∈S1
(14)

with
 
  
Re Tl (k)ρ ≈ |Gl |4 |Xl (k)|2 |Hl (k)Hl+1 (−k)|2
 
k∈Si k∈Si (15)
Nu
= E{|Gl |4 }E{|Xl (k)|2 }E{|Hl (k)|2 }E{|Hl+1 (−k)|2 }.
2
So, the error of the estimate has the form:
 
  
Im Ŵl (k) + α1 Il 1 (k) + α2 Il 2 (k) ρ
 
k∈S0 ∪S1
∆ˆ −∆= 1
2πλ Nu E{|Gl |4 }E{|Xl (k)|2 }E{|Hl (k)|2 }E{|Hl+1 (−k)|2 }
(16)

where λ = (N + Nzp )/N . Since E{Ŵl (k)} = 0 and E{αi Il i (k)} = 0,

the variance of the noise contribution in Eq. (16) is given by
  2
2 1 2
σN = E Im Ŵl (k) + α1 Il (k) + α2 Il (k) ρ
k∈S0 ∪S1
 2
(17)
1 2
= Nu E Im Ŵl (k) + α1 Il (k) + α2 Il (k) ρ .
Progress In Electromagnetics Research, PIER 83, 2008 125

The MSE of fine frequency estimator can be obtained after some

straight forward calculations
 
ˆ 2 1 2B3 1
E ∆−∆ = 2 2 +
4π λ Nu SNR · B2
2
2SNR2 · B2
   (18)
B3 1 1 B1
+ E{|αi | }
2
+ + +
B22 2SNR2 · B2 SNR · B2 2SNR2 · B22

where SNR = σH 2 E /σ 2 , E = E{|X (k)|2 }, B = E{|G |2 }, B =

s W s l 1 l 2
E{|Gl | }, and B3 = E{|Gl |6 }. When the least square (LS) channel
4
estimation is used, E{|αi |2 } = 1/SNR.

4. SIMULATION RESULTS

In this section, extensive simulations are performed to verify the

accuracy of the MSE analysis. In our simulations, 106.7 Mbps UWB-
OFDM system with N = 128, Nn = 6, Ng = 10, and Nzp = 37
is considered. Here, the UWB channel model (CM) that has been
contributed in IEEE 802.15.SG3a is used in [22] and 104 realizations
of each UWB channel model are used, one realization being applied
along the frame duration. In this paper, we consider a pilot-aided
estimator developed in [21] as reference, which only uses Np = 12 pilot
symbols put in subcarriers k ∈ {−55, −45, · · · , −5, 5, · · · , 45, 55} [5].

1E-002
Perfect CE (analysis)
LS CE (analysis)
Perfect CE (simulation)
1E-003 LS CE (simulation)

1E-004
MSE

Pilot

1E-005

Pilot-less

1E-006

1E-007
5 10 15 20 25 30
SNR [dB]

Figure 1. MSE performance of fine frequency estimators.

126 kim et al.

Figure 1 shows the MSE performance of fine frequency estimators

in CM1. Here, we assume that the residual frequency error ∆ is
assumed to be 25% for ±20ppm frequency tolerance [5] and Nn = 112
non-zero data symbols are used for the pilot-less CFO estimation.
When compared to the conventional method, the proposed method
gives an improved estimation performance. The SNR difference
between LS and perfect channel estimations is 3 dB. Therefore, one
can expect that a sophisticated channel estimation scheme additionally
improves the performance [23, 24]. More importantly, it is apparent
that the theoretical results of fine frequency estimator according to
Eq. (18) match with the simulation results very well with the increase
of SNR.
120
Throughput [Mbps]

80

40

Perfect
Estimation
Pilot
Pilot-less
0
5 10 15 20 25 30
SNR [dB]

Figure 2. Throughput performance of fine frequency estimators.

In Fig. 2, the throughput performance of fine frequency estimators

in CM1 is depicted. In this example, the results were based on a
packet size of 1024 bytes and one-tap equalization is used. We can
find that the pilot-less estimator gives very accurate performance,
and it gives the almost same performance to perfect estimation
case. When compared to the pilot-aided estimation method, the
proposed UWB-OFDM receiver with pilot-less estimation provides
approximately 12% throughput enhancement because we save the
pilots for synchronization. However, the computational burden is
increased by using all non-zero data information.
Figure 3 presents the average peak signal-to-noise ratio (PSNR)
Progress In Electromagnetics Research, PIER 83, 2008 127

46
Perfect
Pilot-less
42 Pilot

38
Average PSNR [dB]

34

30

26

22

18
12 16 20 24 28
SNR [dB]

Figure 3. Average PSNR video quality of the UWB-OFDM

synchronization receivers.

performance of the UWB-OFDM receiver with the pilot-less fine

frequency estimator. In this example, LS channel estimation and one-
tap equalization are used. The test video data is ballroom sequence
whose resolution is 640 × 480. In this simulation, the structure of GOP
is “IPPPP· · · PIPP· · · ”, where the size of GOP is 11, i.e., 10 P pictures
are encoded between I pictures. With the help of 12% throughput
enhancement in the pilot-less synchronization receiver, each I picture
is transmitted repeatedly two times. The video codec used in this
simulation is JM13.2 version. In order to check the usefulness of
the proposed scheme, the average PSNR’s of the encoded images are
evaluated with respect to the original sequences at various SNRs,
where the PSNRs are averaged values after simulations are repeatedly
performed more than 100 times. As shown in Fig. 3, the PSNR’s of
the images resulted from the proposed pilotless scheme is much higher
than those from the conventional system. It implies that the proposed
algorithm increases the efficiency of the video communication system.
Figures 4–6 show the MSE and BER performances of fine
frequency estimators when CM2∼CM4 are used, respectively. The
MSE and BER performances in CM2 are approximately parallel to
these in CM1. When the system is encountered with high dispersive
UWB channels CM3 and CM4, the error floor for all estimators can
128 kim et al.
1E-001 1E+000
Pilot
Pilot-less
1E-002 1E-001

1E-003 1E-002
MSE

BER
1E-004 1E-003

1E-005 1E-004

1E-006 1E-005 Perfect

Pilot
Pilot-less
1E-007 1E-006
0 5 10 15 20 25 30 -5 0 5 10 15 20
SNR[dB] SNR[dB]

Figure 4. MSE and BER performances of fine frequency estimators

in CM2: (1) solid lines - perfect CE (2) dashed lines - LS CE.

1E-001 1E+000
Pilot
Pilot-less
1E-001
1E-002

1E-002
1E-003
MSE

BER

1E-003

1E-004
1E-004

1E-005
1E-005 Perfect
Pilot
Pilot-less
1E-006 1E-006
0 5 10 15 20 25 30 -5 0 5 10 15 20
SNR[dB] SNR[dB]

Figure 5. MSE and BER performances of fine frequency estimators

in CM3: (1) solid lines - perfect CE (2) dashed lines - LS CE.
Progress In Electromagnetics Research, PIER 83, 2008 129

1E-001 1E+000
Perfect
Pilot
Pilot
Pilot-less
Pilot-less
1E-002 1E-001

1E-003 1E-002

BER
MSE

1E-004 1E-003

1E-005 1E-004

1E-006 1E-005
0 5 10 15 20 25 30 -5 0 5 10 15 20
SNR[dB] SNR[dB]

Figure 6. MSE and BER performances of fine frequency estimators

in CM4: (1) solid lines - perfect CE (2) dashed lines - LS CE.

be observed because the delay spread of the channel is relatively

longer than CM1 and CM2, which causes the inter-symbol interference.
However, the performances of the proposed pilot-less estimator in CM3
and CM4 are also enhanced and come close to perfect estimation
case, with the comparison with these of the pilot-aided conventional
estimator.

5. CONCLUSION

In OFDM-based wireless UWB systems, a residual frequency offset

introduces time and subcarrier varying phase rotations, which seriously
degrade the performance of the systems. In this paper, we addressed
the problem of estimating the residual frequency error for UWB-based
wireless transmission systems without the aid of pilot information.
Analytical expression of the MSE of the fine frequency synchronization
scheme was derived and simulation results verified the accuracy of our
MSE analysis.

ACKNOWLEDGMENT

The authors wish to acknowledge the assistance and support of

the Center for Advanced Transceiver Systems and the Ministry of
Commerce, Industry and Energy of Korea, and this research is
supported by Seoul R&BD Program.
130 kim et al.

REFERENCES

1. Fan, Z. G., L. X. Ran, and J. A. Kong, “Source pulse optimizations

for UWB radio systems,” Journal of Electromagnetic Waves and
Applications, Vol. 20, No. 11, 1535–1550, 2006.
2. Soliman, M. S., T. Morimoto, and Z. I. Kawasaki, “Three-
dimensional localization system for impulsive noise sources using
ultra-wideband digital interferometer technique,” Journal of
Electromagnetic Waves and Applications, Vol. 20, No. 2, 515–530,
2006.
3. Chen, C.-H., C.-H. Liu, C.-C. Chiu, and T.-M. Hu, “Ultrawide
band channel caculation by SBR/IMAG techniques for indoor
communication,” Journal of Electromagnetic Waves and Appli-
cations, Vol. 20, No. 1, 41–51, 2006.
4. Liu, Y. J., Y. R. Zhang, and W. Cao, “A novel approach to the
refraction propagation characteristics of UWB signal waveforms,”
Journal of Electromagnetic Waves and Applications, Vol. 21,
No. 14, 1939–1950, 2007.
5. ECMA International, Standard ECMA-368, “High rate ultra
wideband PHY and MAC standard,” December 2007.
6. WiMedia Alliance, “Presentation to DLNA,” January 2007.
7. Xiao, S., J. Chen, X.-F. Liu, and B.-Z. Wang, “Spatial focusing
characteristics of time reversal UWB pulse transmission with
different antenna arrays,” Progress In Electromagnetics Research
B , Vol. 2, 223–232, 2008.
8. El-Fishawy, N., M. Shokair, and W. Saad, “Proposed MAC
protocol versus IEEE802.15.3a for multimedia transmission over
UWB networks,” Progress In Electromagnetics Research B , Vol. 2,
189–206, 2008.
9. Khani, H. and P. Azmi, “Performance analysis of a high data
rate UWB-DTR system in dense multipath channels,” Progress
In Electromagnetics Research B , Vol. 5, 119–131, 2008.
10. Yin, X.-C., C. Ruan, C.-Y. Ding, and J.-H. Chu, “A planar
U type monopole antenna for UWB applications,” Progress In
Electromagnetics Research Letters, Vol. 2, 1–10, 2008.
11. Cui, W., P. Ranta, T. Brown, and C. Reed, “Wireless
video streaming over UWB,” Proc. of ICUWB 2007, 933–936,
September 2007.
12. Kim, J., S. Lee, Y. Jeon, and S. Choi, “Residential HDTV
distribution system using UWB and IEEE 1394,” IEEE Trans.
Consumer Electronics, Vol. 52, No. 1, 116–122, February 2006.
Progress In Electromagnetics Research, PIER 83, 2008 131

13. Yak, C., Z. Lei, S. Chattong, and T. Tjhung, “Timing

synchronization and frequency offset estimation for Ultra-
Wideband (UWB) multi-band OFDM systems,” Proc. of
PIMRC 2005, 471–475, September 2005.
14. Yinghui, L., T. Jacobs, and H. Minn, “Frequency offset estimation
for MB-OFDM-based UWB systems,” Proc. of ICC 2006, 4729–
4734, June 2006.
15. Wang, X., T. T. Tjhung, Y. Wu, and B. Caron, “SER performance
evaluation and optimization of OFDM system with residual
frequency and timing offsets from imperfect synchronization,”
IEEE Trans. Broadcasting, Vol. 49, No. 2, 170–177, June 2003.
16. Ai, B., Z.-X. Yang, C.-Y. Pan, J.-H. Ge, Y. Wang, and Z. Lu,
“On the synchronization techniques for wireless OFDM systems,”
IEEE Trans. Broadcasting, Vol. 52, No. 2, 236–244, June 2006.
17. Shim, E. S. and Y. H. You, “Parameter estimation and error
reduction in multicarrier systems by time-domain spreading,”
Progress In Electromagnetics Research B , Vol. 7, 1–12, 2008.
18. Speth, M., S. A. Fechtel, G. Fock, and H. Meyr, “Optimum
receiver design for OFDM based broadband transmission —
Part II,” IEEE Trans. Commun., Vol. 49, 571–578, April 2001.
19. Shi, K., E. Serpedin, and P. Ciblat, “Decision-directed fine
synchronization for OFDM systems,” IEEE Trans. Commun.,
Vol. 53, 408–412, March 2005.
20. Chih-Peng, L., C. Po-Lin, and L. Tsui-Tsai, “Residual carrier
frequency offset tracking for OFDM-based systems,” Proc. of
APCCAS 2004, 989–992, December 2004.
21. Speth, M., S. A. Fechtel, G. Fock, and H. Meyr, “Optimum
receiver design for wireless broad-band systems using OFDM —
Part I,” IEEE Trans. Commun., Vol. 47, 1668–1677, November
1999.
22. Foerster, J., Channel Modeling Sub-committee Report, IEEE,
802.15-02/490r1-SG3a, Febuary 2003.
23. Baek, M. S., M. J. Kim, Y. H. You, and H. K. Song, “Semi-blind
channel estimation and PAR reduction for MIMO-OFDM system
with multiple antennas,” IEEE Trans. on Broadcasting, Vol. 50,
No. 4, 414–424, December 2004.
24. Koo, B. W., M. S. Baek, and H. K. Song, “Multiple
antenna transmission technique for UWB system,” Progress In
Electromagnetics Research Letters, Vol. 2, 177–185, 2008.