D2D Asisted MU-MIMO Precoding to Enhance
LTE Network
Yanqiu Zhang, Shu Fang, Yuanchao Han, Yu Zeng
Department of National key Laboratory on Communications,
University of Electronic Science and Technology of China, Chengdu, China
Email: fangshu@uestc.edu.cn
Abstract—Multi-user multiple-input multiple-output (MU-
MIMO) has been adopted as a key enabling technique to
achieve the capacity demand in 4G wireless networks. However,
there are still some problems existing in the frequency division
duplex (FDD) MU-MIMO system, such as the channel quality
indicator (CQI) mismatch problem, large uplink feedback, low
pairing probability of MU communication. In this paper, we
exploit a MU-MIMO scheme in FDD assisted by device-to-device
(D2D) technique to enhance the conventional LTE network.
The proposed scheme could not only solve the above problems
in conventional MU-MIMO systems, but also achieve signifi-
cant performance improvement than conventional MU-MIMO
schemes. Since the user pairing process is done at the user side,
the feedback of the proposed scheme is reduced dramatically
compared with the conventional scheme. Simulation results show
that the proposed D2D assisted MU-MIMO scheme has clear
superiority in terms of throughput over conventional MU-MIMO
scheme in LTE networks. Because with D2D assisted MU-
MIMO, the proposed scheme tackles the CQI mismatch problem
and increases the multi-user pairing probability to enhance the
overall throughput.
Index Terms—Device-to-device, multiple-input multiple-
output, feedback, channel quality indicator, precoding
I. INTRODUCTION
Multi-user multiple-input multiple-output (MU-MIMO)
technique has been adopted as one of the core technologies in
LTE-A system. The basic feature of MU-MIMO is to empoly
the space division mutiple access (SDMA) technology to trans-
mit multiple data streams from multiple users simultaneously
[1]. Therefore, the inter-channel interference is existed in MU-
MIMO system and precoding technique is adopted for MU-
MIMO system so as to combat the interference.
In the time division duplex (TDD) mode, the evolved Node
B (eNB) can easily obtain the channel state information
(CSI) via channel reciprocity of uplink channel. However,
in frequency division duplex (FDD) mode, where channal
reciprocity is not existed, it is necessary to feed CSI back
to the eNB for each user in MU-MIMO system [2]. The
conventional MU-MIMO scheme in FDD mode will feed back
the Rank-1 channel quality indicator (CQI) and precoding
matrix information (PMI) without multi-user interference to
eNB. After eNB collects all of the feedback information, it
will do multi-user scheduling to pair some of the users at the
same time-frequency resource block to estimate the CQI with
multi-user interference. Therefore, the modulation and coding
scheme (MCS) is finally confirmed at eNB according to the
estimated CQI at eNB. Therefore, the CQI mismatch problem
is always existed in FDD MU-MIMO system. At the same
time, the pairing probability of FDD MU-MIMO is low.
Recently, device-to-device (D2D) communication in cellular
networks has received much attention thanks to its potential
to provide higher date rate [3]. D2D communication can
be defined as a direct communication between two UEs.
Specifically, unlike the conventional cellular communication
where all traffic is routed via eNB, the adjacent UEs can
establish direct links in the control of eNB to transmit data
to each other under cellular nerwork resources. It has the
advatages of improving spectrum utilization, shorting packet
delay, and reducing energy consumption, while enabling new
peer-to-peer and location-based applications and services [4].
There are some recent research work and literature con-
cerned with tackling the above issues existing in MU-MIMO
system. In [5], [6], the authors refers to the idea of joint
spatial division and multiplxing (JSDM) to realize savings in
CSI for FDD mode. It lies in clustering the UEs into groups
with approximately similar channel covariance eigenvectors,
while users across groups have near orthogonal covariances.
To mitigate or avoid mutual interference between cellular and
D2D transmissions, a new strategy of setting a restricted area
was proposed in [7], in which the cellular users can not coexist
with D2D users, thus the interference caused by the uplink
cellular users can be controlled within a certain range. In
[8], the authors found that interference can be completely
overcome if the number of canceled D2D interfering signals
is scaled with the number of base station (BS) antennas at an
arbitrarily slow rate. In [9], the authors proposed to suppress
co-channel interference by maximizing the signal-to-leakage-
noise ratio (SLNR) for all users simultaneously.
For the resource allocation of D2D communication in the
MU-MIMO system, an enhanced D2D resource allocation
scheme under the TDD mode is proposed [10], where eNB
acts as a relay. It receives data from the D2D users and
forward the data stream when the D2D link is interrupted,
thus guaranting the quality of D2D communication. Nash
bargaining solution and nash competitive games frameworks
are proposed to address the power allocation problem for D2D
underlay cellular network in [11]. In [12], the author used
game theory to study the resource allocation in the networks,
and an efficient resource allocation algorithm based on better
response dynamic is developed for a feasible suboptimal solu-