A Dynamic Scheduling Scheme for CoMP in Downlink FD-MIMO Transmission
Chengyu Lu, Binyan Lu, Yong Wang, Shu Fang
National Key Laboratory of Science and Technology on Communications,
University of Electronic Science and Technology of China,
Chengdu, China
e-mail: fangshu@uestc.edu.cn
Abstract—Full dimension MIMO (FD-MIMO) technique has
attracted considerable attention in both academia and industry
as a potential candidate technology in 5G. It is able to enhance
the performance of wireless communication system such as
throughput, spectral efficiency and energy efficiency due to the
exploitation of additional spatial freedom degrees. Meanwhile,
coordinated multiple points (CoMP) technique is also a
promising technology in Long Term Evolution (LTE).
Through coordinated processing, the throughput of cell-edge
users can be improved significantly. In this paper, we propose
a dynamic scheduling scheme with signal leakage to noise ratio
(SLNR) precoding for CoMP in downlink FD-MIMO
transmission. First, the possible CoMP UE is searched
dynamically with low signal to interference plus noise ratio
(SINR) that requires coordination with other base-stations
(BS). Second, the possible CoMP BS is searched semi-
dynamically to provide enough gain for users. Simulation
results show that the proposed scheme achieves almost the
same performance as the full joint processing (JP) scheme, but
with a significant reduction of computational complexity.
Keywords-CoMP; scheduling; FD-MIMO; precoding;
computational complexity
I. INTRODUCTION
With the explosive development of mobile internet,
people are eager for a higher data rate transmission and a
better quality of communication [1], [2]. But the current
mainstream system (long term evolution (LTE) and long
term evolution-advanced (LTE-A)) can not satisfy new and
unprecedented demands that highlighted the 5G
communication system [3]. Massive multi-input-multi-output
(M-MIMO) is one of the promising candidate technologies
currently studied in the third generation partnership project
(3GPP) for the next generation wireless system 5G.
Researches reveal that massive MIMO with linear precoding
processing can reach the performance bound and improve
the spectral efficiency, energy efficiency and robustness
greatly due to exploit additional degrees of freedom [4], [5].
In most studies, massive MIMO is implemented in a
horizontal way equipped with uniformly-spaced linear array
(ULA) [6], [7]. But in practical deployment, limited by the
space of base station (BS), this is not suitable to implement
in this approach [8]. FD-MIMO is capable to overcome this
space limitation of BS by exploiting 2D array structure, since
vertical dimension is also exploited as well as horizontal
domain. By utilizing additional multiplexing gain provided
by elevation domain, the performance of FD-MIMO will
improve greatly. Conventional 2D beamforming evolves into
3D beamforming, consequently, user equipments (UE)
distributed in different altitude can be distinguished by BS
[9]. Based on the channel state information (CSI) at the
transmitter (CSIT), the spatial multiplexing gain can be
obtained by optimizing each UEs specific precoding
coefficients, thus improves the overall system performance
[10].
Furthermore, in recent years, coordinated multiple point
(CoMP) is also considered as one of the promising
technologies in LTE-A, because the quality of service (QoS)
for cell-edge UEs can be improved significantly due to the
alleviation of inter-cell interference [11]. 3GPP therefore
pushes on the study of potential benefits utilizing advanced
interference mitigation techniques relying on CoMP [12].
Network clustering which turns a potentially interference
free system can reduce the inter cluster interference, plays a
vital rule as an essential part of each CoMP scheme [13].
Generally, CoMP technique is divided into two categories by
3GPP, joint processing CoMP (JP-CoMP) and coordinated
beamforming/coordinated scheduling CoMP (CB/CS-CoMP),
both of them share CSI among coordinated sectors while
only the former one utilizes all coordinated sectors to
transmit data [14]. In this paper, we mainly focus on JP-
CoMP, the information data will be available at cooperating
BS and UE receives the processed data from all BS.
Now comes the question, what will happen if FD-MIMO
meets CoMP, there are few papers focusing on this issue. [13]
proposed a novel precoding algorithm for massive MIMO
with FD-MIMO configuration in CoMP scenario, [15]
explored the benefits of combing cloud radio access
networks (C-RAN) with FD-MIMO. In this paper, we will
further discuss the CoMP scheme in FD-MIMO
configuration. We propose a novel dynamic scheduling
scheme with SLNR precoding for CoMP in downlink FD-
MIMO that is able to significantly reduce the complexity of
the conventional full JP-CoMP scheme. Furthermore, the
proposed scheme almost achieves the same performance
with the full JP-CoMP scheme. For the proposed scheme, we
first search the coordination UEs dynamically with low
SINR. Second, we search for the base-stations semi-
dynamically that is suitable for coordination and can provide
enough gain to the UEs. We evaluate the scheme in FD-
MIMO environment with SLNR precoding and the
simulation results demonstrate its efficiency.