A BEM Method of Channel Estimation for OFDM
Systems in High-Speed Train Environment
Zhichao Sheng, Yong Fang, Chen Wang
School of Communication and Information Engineering
Shanghai University, China
E-mail: {kebon22, yfang, cwangsh} @ shu.edu.cn
Abstract—This paper investigates the channel estimation
problem of Orthogonal Frequency Division Multiplexing
(OFDM) systems in high-speed train environment. The OFDM
channel in the high-speed train environment is modeled by using
basis expansion model (BEM). The estimation problem of the
channel impulse response and Doppler frequency offset are
studied. In the simulation, the modeling performances of
different bases are compared. The estimation results of the
channel impulse response and Doppler frequency offset show the
effectiveness of this method in the high-speed train environment.
Keywords—Channel estimation; OFDM; basis expansion
model (BEM); high-speed train.
I. INTRODUCTION
Orthogonal Frequency Division Multiplexing (OFDM) has
become one of the key technologies featured in LTE (Long
Term Evolution), and efficient broadband access to OFDM is
a current research focus. On the other hand, high-speed train
has become one of the most important transportation means.
The high mobility of high-speed train brings large Doppler
shift. Moreover, the Doppler shift causes frequency offset of
line-of-sight (LOS) component [1]. Therefore accurate
channel estimation of OFDM systems in the high-speed train
environment is an integral part [2].
Recently, basis expansion model (BEM), which is a
parsimonious model, has become popular for channel
estimation. BEM is used to model rapidly time-varying
wireless channels and then a finite number of BEM
coefficients will be estimated to reconstruct the whole time
domain channel [3-6]. Because the number of estimated BEM
coefficients is much smaller than the unknowns of the channel,
it reduces the complexity of channel estimation. As a result,
BEM can reconstruct the LOS path in the high-speed train
environment, which is useful to estimate the Doppler
frequency offset.
In this paper, the BEM is adopted to model the channel in
the high-speed train environment. Based on the BEM, the LOS
path and scattered paths are estimated. The Doppler frequency
offset of the LOS path is also studied. In the simulation, the
modeling performances of different bases are compared. The
estimation results of the channel impulse response and
Doppler frequency offset show that this method based on
BEM can obtain accurate estimation performance.
II. SYSTEM MODELS
A. Channel Model in High-Speed Train Environment
In the high-speed train environment, due to the deployed
base stations (BS), there are a dominant LOS path and
scattered paths. The channel is often modeled as the Rician
fading channel [1].
Let us consider the OFDM systems with K subcarriers. In
order to avoid intersymbol interference (ISI), we assume that
the length of cyclic prefix
is larger than the number of
propagation paths
.
At the transmitter side, the
th transmitted symbols in the
frequency domain is
( ,0) ( , 1)
T
X X n X n K
.
After the IFFT processing,
and
is the unitary
FFT matrix with entries
. The
th received
symbols in the time domain is
(1)
where
( ,0) ( , 1)
T
y y n y n K
,
( ,0) ( , 1)
T
x x n x n K
and
is the column
vector of noise.
In the frequency domain, (1) can be rewritten as
(2)
where
( ,0), , ( , 1)
T
Y Y n Y n K
is the column vector
of received symbols,
is the channel matrix in
the frequency domain and
is the column vector of noise.
In (2), the time domain channel matrix
is
This work is supported by NSF of China #61271213.