Channel estimation for 3D MIMO system based on LOS/NLOS identification

Abstract

Channel estimation is one of the most important parts in three-dimensional multiple-input multiple-output (3D MIMO) systems. The characteristics of non-line-of-sight (NLOS) channel and line-of-sight (LOS) channel are different in 3D MIMO systems. If the same channel estimation scheme is used in LOS case as NLOS, the performance of estimation will be bad. In outdoor propagation environment, 3D MIMO channels between closely located antennas share the same delay support in temporal domain. With those prior knowledge, in this study, a new channel estimation scheme is proposed. The proposed scheme can be divided into two processes. First, it is needed to identify the received sounding reference signal whether is LOS or NLOS propagation. Then, different enhanced DFT-based channel estimation schemes are proposed separately according to the identification results. Simulation results verify the proposed algorithm outperforms traditional discrete Fourier transform (DFT)-based channel estimation. At signal-to-noise ratio of 20 dB, the proposed algorithm has 17.7 and 35.7% improvement in NLOS case and LOS case separately in terms of normalised mean squared error compared with traditional DFT-based channel estimation scheme, and is achieved with additional liner complexity.

Full text

IET Communications
Research Article
Channel estimation for 3D MIMO system
based on LOS/NLOS identification
ISSN 1751-8628
Received on 15th April 2018
Revised 14th December 2018
Accepted on 30th January 2019
E-First on 28th February 2019
doi: 10.1049/iet-com.2018.5211
www.ietdl.org
Minshan Xiang1,2, Yongyu Chang1,2 , Tianyi Zeng1,2
1School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, People's Republic
of China
2Wireless Theories and Technologies Laboratory, Beijing University of Posts and Telecommunications, Beijing 100876, People's Republic of
China
E-mail: yychang@bupt.edu.cn
Abstract: Channel estimation is one of the most important parts in three-dimensional multiple-input multiple-output (3D MIMO)
systems. The characteristics of non-line-of-sight (NLOS) channel and line-of-sight (LOS) channel are different in 3D MIMO
systems. If the same channel estimation scheme is used in LOS case as NLOS, the performance of estimation will be bad. In
outdoor propagation environment, 3D MIMO channels between closely located antennas share the same delay support in
temporal domain. With those prior knowledge, in this study, a new channel estimation scheme is proposed. The proposed
scheme can be divided into two processes. First, it is needed to identify the received sounding reference signal whether is LOS
or NLOS propagation. Then, different enhanced DFT-based channel estimation schemes are proposed separately according to
the identification results. Simulation results verify the proposed algorithm outperforms traditional discrete Fourier transform
(DFT)-based channel estimation. At signal-to-noise ratio of 20 dB, the proposed algorithm has 17.7 and 35.7% improvement in
NLOS case and LOS case separately in terms of normalised mean squared error compared with traditional DFT-based channel
estimation scheme, and is achieved with additional liner complexity.
1 Introduction
Three-dimensional multiple-input multiple-output (3D MIMO)
technology has become one of the possible technologies for
implementing massive MIMO in practical use. Channel estimation
is one of the most important parts in 3D MIMO system. The
accuracy of channel estimation will directly affect the system
performance.
Due to the low complexity and the acceptable performance in
practical systems, the discrete Fourier transform (DFT)-based
channel estimation has attracted lots of interests in channel
estimation studies. In [1], the DFT-based channel estimation was
first proposed for orthogonal frequency-division multiplexing
(OFDM) communication systems under time-variant radio channel
conditions. It employs low-pass filter in a transform domain which
can reduce inter-carrier interference and additive white Gaussian
noise (AWGN) components in the received pilot signals
significantly. An enhanced DFT-based channel estimation method
is proposed in [2]. The residual noise existing within the ‘energy
concentration’ region in time domain is reduced by reasonable
selection of a threshold value. Inaccurate noise power estimation
may cause worse estimation performance. Yang et al. [3] use
different window functions which can reduce the aliasing error of
the direct DFT-based channel estimation method. Huang et al. [4]
proposes an improved method to estimate noise power and use it
for an approximated MMSE-based channel estimation. However,
this method has a high complexity. In [5], an enhanced DFT-based
channel estimation for long-term evolution (LTE) uplink is
proposed. A sinc-null based noise power estimation method in
conjunction with a dynamic noise removal technique is proposed to
suppress the noise in the time domain and achieve better
performance while keeping at a low complexity. In [6], another
improved DFT-based channel estimation scheme is proposed. This
scheme tries to clean the reference signals in the time domain
before being used for interpolation. It uses the estimated noise
variance from reference signals on multiple OFDM symbols via the
property of DFT and does not need any priori channel information.
In [7], the DFT-based channel estimator by using the frequency-
domain weighting scheme for LTE-Advanced uplink is
investigated. The proposed scheme can improve the degraded
performance due to the non-orthogonal channels of multiple
antennas of the current reference signal pattern at the time domain.
Zhang et al. [8] proposes a cascaded MMSE channel estimation
with low computational complexity by utilising the channel spatial
correlation between antennas at the transmitter side. In [9], the
problem of 3D massive MIMO sparse channel estimation is
transformed into compressive sensing (CS) sparse signal
reconstruction problem and then adaptive filtering algorithm in CS
reconstruction problem is employed. Shafin et al. [10] studies the
angle and delay estimation for 3D massive MIMO systems under a
parametric channel modelling.
However, the schemes mentioned above do not take the line-of-
sight/non-line-of-sight (LOS/NLOS) statistical discrimination into
account. The channel characteristics for users in LOS condition
and NLOS condition are different. It will affect the system
performance in many aspect [11]. Tt will affect the accuracy of
channel estimation. Thus, in this paper, we propose an enhanced
DFT-based channel estimation scheme which takes account of
LOS/NLOS condition of every user. We utilise the characteristic of
LOS/NLOS channels and the sparse common support (SCS) in 3D
MIMO systems to make an identification of LOS users and NLOS
users. Then, in NLOS condition, we propose an adaptive de-
noising channel estimation scheme. In LOS condition, we propose
an united-decision scheme to estimate the delay spread by using
the prior knowledge of channels.
The remainder of this paper is organised as follows. Section 2
introduces the characteristics of 3D MIMO channel and the DFT-
based channel estimation. In Section 3, we present the proposed
channel estimation scheme. Section 4 shows the simulation results.
Section 5 concludes the paper.
Notations: AT, A
¯ denotes the transposed matrix of A and the
conjugate matrix of A, respectively. diag(a1, …, aN) is the diagonal
matrix with a1, …, aN as its diagonal elements. ∥a∥1, ∥a∥2,
∥a∥∞ denote the 1-norm, 2-norm, infinite-norm of a, respectively.
The absolute value of a is denoted by a. I is the identity matrix
IET Commun., 2019, Vol. 13 Iss. 7, pp. 898-904
© The Institution of Engineering and Technology 2019
898