Channel Estimate with PDP Assumption and
Interference RS knowledge in LTE System
Liang Lin
1,2
, Niu Kai
1
, Xu Wenjun
1
, Tian Baoyu
1
, Gong Ping
1
, Sun, Shaohui
2
1
Key Lab of Universal Wireless Communications, Ministry of Education, BUPT
2
State Key Laboratory of Wireless Mobile Communications, CATT
Beijing, P.R. China
Email: lianglin@bupt.edu.cn
Abstract—In this article, we discuss the LMMSE(linear
minimum mean square error) channel estimate methods on
OFDM(Orthogonal Frequency Division Multiplexing) in LTE sys-
tem. To reduce the complexity of LMMSE, a constant assumption
PDP(power delay profile) is used with little performance backoff.
The influence of different PDP assumption and channel condition
is simulated that the proposed algorithm is robust. Then we come
to SRS(sounding reference signal) which different users send
different sequence in the same time-frequency resources in LTE
system. Simulation results show that if user number is not more
than 4, the MSE performance nearly equals to single user, that
is LMMSE channel estimate can cancel orthogonal RS sequence
interference implicitly. DMRS(demodulation RS) estimate in
CoMP scenario has also been discussed whose performance can
be increased substantially with the knowledge of interference RS
sequence.
Index Terms—channel estimate, assumed PDP, LMMSE,
known RS sequence
I. INTRODUCTION
Driven by the need of high data rate in multimedia based
applications, novel techniques such OFDM and MIMO stand
as promising choices for future high data rate systems such as
LTE and WiMAX[1].
OFDM divides the available spectrum into a number of
overlapping but orthogonal narrowband subchannel, and hence
converts a frequency selective channel into a non-frequency
selective channel[2]. Moreover, ISI is avoid by the use of CP
which can also simplify the complexity of synchronization
and channel estimate. CP can make line convolution that
commonly model channel when digital signal processing is
considered in time domain into circular convolution by ignor-
ing the change of channel in one OFDM symbol which may
cause ICI.
In coherent wireless receivers, channel estimate is integral
part of the receive designs because unless the receiver ac-
curately estimate how the channel modifies the transmitted
signal, it can not recover the transmitted signal in coherent
system[3]. Channel estimate is a challenging problem in wire-
less systems for the radio channel is highly dynamic. Multipath
propagation, mobility, and local scattering cause the signal
to be spread in frequency, time, and angle which are related
to selectivity of the channel. Channel estimation performance
is directly related to these statistics, however sometimes it is
difficult to acquire the knowledge of channel characteristic.
Different techniques are proposed to exploit channel statistic
for better channel estimates[4][5].
In typical communication system, the CIR in time domain
is modeled as an unknown time-vary FIR filter, and the
estimation of these coefficients is channel estimate itself. In
OFDM based system which is a multi-carrier system can learn
from the method in single carrier system, however the unique
properties of multi-carrier transmission may bring about addi-
tional perspectives for the new approaches to estimate channel.
With the help of CP, we can transform the time domain CIR
which modeled as a FIR filter and computed in convolution
into the frequency domain CFR which result in multiplication
relations. Besides the estimation of CIR, radio channel can also
be estimated in frequency domain directly using the known
data on frequency domain.
Channel estimation can be grouped into two main cate-
gories: blind and non-blind. The non-blind estimation gains
more compared with blind case. The non-blind channel esti-
mation transmit data which is known by the receiver so that the
receiver can estimate the radio channel by demodulating the
received samples. We call these data the reference signal(RS).
The design of RS pattern should meet Nyquist Theorem
which satisfy the minimum sampling demands of coherence
bandwidth and coherence time.
There are numerous approaches to estimate the channels,
the direct estimation of the channel for subcarriers which
treat each subcarriers as independent. However, in practice,
the CFR is often oversampled, hence the frequency domain
channel coefficients are correlated. On the other hand, the
noise in these subcarriers can be independent. By utilizing
the correlation in frequency domain, the noise can be reduced
significantly. Therefor, the channel estimate accuracy can be
improved[6]. Similarly, the correlation in time and spatial
domain can be exploited to suppress noise since the noise
is considered to be independent in time and spatial domain as
well.
For fast fading channels, constant channel assumption leads
to ICI[7] which treated as part of noise in most systems. Of
course, better channel estimation performance can be obtained
by more accurate modeling[8].
Commonly, RS position could not hold all the resource
element, for data demodulation, the interpolation is invited to
estimate subcarriers channel response whose transmitting data.