Hindawi Publishing Corporation

International Journal of Antennas and Propagation

Volume 2014, Article ID 848071, 10 pages
http://dx.doi.org/10.1155/2014/848071

Research Article
Massive MIMO Channel Models: A Survey

Kan Zheng,1 Suling Ou,1 and Xuefeng Yin2

1
Wireless Signal Processing and Network Lab, Key Laboratory of Universal Wireless Communication, Ministry of Education,
Beijing University of Posts and Telecommunications (BUPT), P.O. Box 93, No. 10, Xi Tu Cheng Road, Haidian District,
Beijing 100876, China
2
Department of Electronics Science and Technology, Tongji University, Shanghai, China

Correspondence should be addressed to Kan Zheng; [email protected]

Received 25 January 2014; Accepted 28 March 2014; Published 16 June 2014

Academic Editor: Periklis Chatzimisios

Copyright © 2014 Kan Zheng et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The exponential traffic growth of wireless communication networks gives rise to both the insufficient network capacity and excessive
carbon emissions. Massive multiple-input multiple-output (MIMO) can improve the spectrum efficiency (SE) together with the
energy efficiency (EE) and has been regarded as a promising technique for the next generation wireless communication networks.
Channel model reflects the propagation characteristics of signals in radio environments and is very essential for evaluating the
performances of wireless communication systems. The purpose of this paper is to investigate the state of the art in channel models
of massive MIMO. First, the antenna array configurations are presented and classified, which directly affect the channel models
and system performance. Then, measurement results are given in order to reflect the main properties of massive MIMO channels.
Based on these properties, the channel models of massive MIMO are studied with different antenna array configurations, which
can be used for both theoretical analysis and practical evaluation.

1. Introduction in data throughput and link reliability without extra band-

width or boosting transmission power. Nowadays, MIMO
The requirements for high rate of wireless communication together with orthogonal frequency division multiplexing
networks grow exponentially with the applications of smart (OFDM) has been accepted as key techniques in the third
terminals. Thus, the capacity of the networks has to be generation (3G) long-term evolution (LTE) cellular networks
increased in order to guarantee the quality of service (QoS) and its advancement (Adv.). The evolved Node B (eNB)
of mobile applications. Improving the spectrum efficiency equipped with multiple antennas communicates with several
(SE) is one of the feasible ways to achieve the better network types of user equipment (UE); at the same time frequency
capacity. Besides it, with the excessive power consumption resources, named as multiuser MIMO (MU MIMO), can
of wireless communication networks, both the carbon emis- improve the spectrum efficiency, link reliability, and system
sions and operator expenditure increase every year [1, 2]. energy efficiency [6, 7]. In order to scale up these benefits,
Therefore, green communication has gained more atten- massive MIMO, also named as large-scale antenna system,
tion in the academic and industrial fields, and the energy very large MIMO, or hyper MIMO, has been first put
efficiency (EE) has been regarded as another vital metric forward in 2010 [8]. In massive MIMO system, hundreds
for evaluating the new technique as well as the spectrum of antennas are used to serve tens of UEs simultaneously.
efficiency [3–5]. Theoretical and measurement results indicate that massive
Multiple-input multiple-output (MIMO) technique has MIMO can significantly improve the spectrum efficiency and
attracted much attention in wireless communications for simultaneously reduce the radiated power [6, 9]. All the
more than ten years because it can offer significant increases features of massive MIMO are shortly summarized in Table 1.
2 International Journal of Antennas and Propagation

Table 1: Features of massive MIMO.

Feature Main reason
High spectrum efficiency [6, 9] Large multiplexing gain and array gain
High energy efficiency [6, 9] Radiated energy can be concentrated on UE
High reliability [9] Large diversity gain
Favorable propagation condition for i.i.d. Rayleigh
Efficient linear precoder/detector [9]
Advantages channel
Weak interuser interference and
Orthogonal UE channels and extremely narrow beam
enhanced physical security [6]
Channel harden phenomenon averages out the fast
Simple scheduling scheme [16]
fading
Robust to individual element failure [6] Large number of antenna array elements
Limited orthogonal pilots as of bounded coherent
Pilot contamination [8]
interval and bandwidth
High signal processing complexity [9] Large number of antennas and multiplexing UE
Disadvantages
Extremely narrow beam is sensitive to UE moving or
Sensitive to beam alignment [6]
antenna array swaying
Poor broadcast channel [6] Be blind to UE positions

In Release 11 (R11) specified by the 3rd generation partner- consideration of both the theoretical analysis and realistic
ship project (3GPP), the MIMO technique can only radiate system evaluation.
the beam in horizontal dimension as of the fixed down tilt of The remainder of this paper is organized as follows.
antenna array. In order to well exploit the vertical angular res- Section 2 discusses antenna array deployments of massive
olution of signal propagation, different kinds of the antenna MIMO. In Section 3, measurement results of massive MIMO
array configurations, such as the rectangular, spherical, and channel are mainly investigated. In Section 4, both the
cylindrical antenna array deployments, have been studied in CBSMs and GBSMs of massive MIMO are studied. Finally,
3GPP [13–15]. With these MIMO antenna array deployments, we conclude the paper in Section 5.
the eNB can adaptively adjust both azimuth and elevation
angles of signal propagation. In other words, the radiated
MIMO signals can be controllable in the three-dimension 2. Antenna Array Configuration
(3D) space and thus called as 3D MIMO. In order to
enhance the system capacity, 3D MIMO should employ more The antenna array structures roughly undergo three phases
antennas. And, massive MIMO should adopt the rectangular, with the development of technique in manufacturing. In
spherical, or cylindrical antenna array configurations for the the traditional passive antenna array, the radiofrequency
practical system considering the space of antennas array. circuit is usually connected to the physical antennas through
Therefore, 3D MIMO with a large number of antennas can the radiofrequency cable. Subsequently, in order to reduce
be regarded as one of practical forms of massive MIMO. the loss induced by the radiofrequency cable and save the
Up to now, massive MIMO has been paid much attention cost of installation and maintenance, the remote radio unit
to by both academic and industry organizations. There are (RRU) separated from the baseband unit (BBU) has been
lots of works done for massive MIMO, such as performance widely adopted [21]. The digital baseband signals generated
analysis, remedies for alleviating pilot contamination, and from the BBU are sent to the RRUs through the optical
channel estimation. The channel model is fundamental for fiber. The radiofrequency circuit is as close as possible to
theoretical analysis as well as for performance evaluation the physical antennas. Furthermore, active antenna array
of massive MIMO system. Therefore, this paper mainly system without radiofrequency cable may be made, which is
investigates the state of the art of the channel models for the helpful to the realization of the massive antenna array [22]. In
massive MIMO system. The configuration of antenna array active antenna array, both the radiofrequency circuits and the
as the critical factor deciding the characteristics of massive antenna elements are integrated into one circuit board, which
MIMO channels is first presented briefly. Based on the given is an important milestone for the development of antenna
antenna array configurations, the current measured activities array.
of massive MIMO channels are studied for working out the As shown in Figure 1, there are several typical antenna
main properties of massive MIMO channels, which are the array configurations for massive MIMO systems, that is,
basis for the precise channel models. With the new properties, linear antenna array, spherical antenna array, cylindrical
two kinds of channel models, namely, correlation-based antenna array, rectangular antenna array, and distributed
stochastic models (CBSMs) and geometry-based stochastic antenna array [6]. The linear array is an example of one-
models (GBSMs), are presented for the massive MIMO in dimension (1D) antenna array, which generally propagates
International Journal of Antennas and Propagation 3

Cylindrical

Distributed
Spherical

Linear

Rectangular

Figure 1: Antenna array configuration [6].

signals on the two-dimension (2D) plane. It is usually as the main properties for the realistic channel model [17, 23].
assumed for theoretical analysis and realistic measurements. However, further measurements should be implemented to
On the other hand, the spherical antenna array, cylindrical validate the two properties for the spherical, cylindrical, and
antenna array, and rectangular antenna array are some kind rectangular antenna array configurations. We will discuss the
of 2D antenna array, which can radiate the signals into any properties derived from channel measurements in detail as
directions in the 3D space. Considering the size of antenna follows.
array for the eNBs and UEs, the spherical, cylindrical, and The channel characteristics, such as channel gain, K-
rectangular antenna array configurations are more likely to factor, and angular power spectrum (APS), are measured
be used in practical systems. Moreover, distributed antenna with a bandwidth of 50 MHz, when the eNB employs a
array is mainly used for either indoor coverage enhancement linear 128-antenna array with the antenna space of half a
or outdoor cooperation. wavelength [17]. As shown in Figure 2, measurement results
The configurations of antenna array directly affect the indicate that the antenna array is beyond the Rayleigh
channel properties, furthering the performance of massive distance as of the large length of linear antenna array. Thus,
MIMO system. Taking the antenna space as an example, both the far-field and plane wavefront assumptions in the
it decides the mutual coupling and correlation matrix, fur- traditional channel models are inappropriate for the massive
thering the capacity of massive MIMO. However, the spaces MIMO channel. Moreover, different antennas at the eNB may
between adjacent antennas in current antenna array are observe different sets of clusters in different time slots, which
usually set as equal. Therefore, the study of the antenna array results in the shadow fading over the linear antenna array,
configuration on a constraint or unconstraint area is valuable namely, nonstationary phenomenon.
for the development of massive MIMO, such as the design of When the eNB employed a 32-antenna rectangular array
sparse antenna array. with half a wavelength between the adjacent antennas, the
indoor-to-outdoor channels were measured with a band-
3. Channel Measurement width of 50 MHz [18]. The correlation between different
UE channels in measurements is higher than that in inde-
Lots of channel measurements have been carried out in pendent identical distributed (i.i.d.) Rayleigh channel, since
order to discriminate the main properties of massive MIMO positions of two UEs are too close to undergo the similar
channel. Typical measurements are summarized in Table 2 for channels. Moreover, the average channel correlations for both
different antenna array configurations under a given scenario. measurement and i.i.d. Rayleigh channel decrease with the
These measurements mainly focus on the impacts of antenna increasing number of antennas; namely, the UE channels can
numbers on the small-scale propagation characteristics. be decorrelated by the very large antenna array.
When the linear antenna array is employed at the eNB, the Channel measurements for both the cylindrical and linear
nonstationary phenomenon and near-field effect are studied antenna array with 128-antenna were implemented, when
4 International Journal of Antennas and Propagation

Table 2: Channel measurement.

Scenario Antenna configuration Antenna space Bandwidth/carrier Measured metric
Channel gain, 𝐾-factor, APS, and eigenvalue
Outdoor [17] 128-antenna linear array Half wavelength 50 MHz/2.6 GHz
distribution correlation of antennas and users
Indoor-to-
32-antenna rectangular to
outdoor Half wavelength 50 MHz/2.6 GHz Correlation
128-antenna cylindrical array
[18]
Cylindrical and linear array with
Outdoor [19] Half wavelength 50 MHz/2.6 GHz Large-scale fading and angular resolution
128-antenna
Outdoor [20] 112-antenna virtual array — 20 MHz/2.6 GHz Correlation, inverse condition number

Cluster 1

Cluster 2 MS3

eNB Cluster 3

MS1

MS2
Cluster 4

Cluster 5

Figure 2: Illustration of nonstationary phenomenon and near-field effect [10].

the space of adjacent antennas is half a wavelength and the have been concluded in 3GPP [11]. The channel model of
bandwidth is 50 MHz [19]. Similar to [17], the nonstationary rectangle antenna array comprising these characteristics is
phenomenon can be observed over the linear antenna array. also established, which will be studied in the next section.
Meanwhile, large power variation can also be experienced
over the cylindrical antenna array as of both the polarization
and directional pattern. Moreover, the linear antenna array 4. Channel Model
has superior angular resolution but only in azimuth. The
Two kinds of channel models, namely, correlation-based
cylindrical antenna array has lower angular resolution, yet it
stochastic models (CBSMs) and geometry-based stochastic
can resolve in both azimuth and elevation, which may be an
models (GBSMs), are widely used to evaluate the perfor-
advantage for some scenarios, such as building coverage.
mances of the wireless communication systems. The com-
A scalable virtual antenna array with 112-antenna was
plexity of the former is lower and is mainly used for analyzing
employed at the eNB in order to investigate the channel
the theoretical performance of MIMO systems. However,
characteristics for massive MIMO in [20]. Both the line of
the accuracy is limited for the realistic MIMO system, and
sight (LoS) and non-line of sight (NLoS) scenarios were
it is difficult to model wireless channels considering the
studied with a bandwidth of 20 MHz. The correlation and
nonstationary phenomenon and spherical wave effects. In
inverse condition number were measured to evaluate the
contrast, the latter can accurately reflect the realistic channel
orthogonality of two-channel vectors and multiple channel
properties and is more suitable for massive MIMO channel
vectors, respectively. The difference of inverse condition
even with the higher computation complexity. The current
numbers for measured channel and i.i.d. Rayleigh channel
channel models have been summarized in Table 3. This
becomes larger with more multiple UE and/or deploying
section discusses the CBSMs for theoretical analysis and
more antennas. Deploying more antennas will make the
GBSMs for the realistic performance evaluation according to
channels of multiple UE more orthogonal. However, there
the categories in Table 3.
is less improvement when the number of antennas exceeds
a given number, which mainly relies on the propagation
environment and UE positions. 4.1. CBSM. The CBSMs are mainly used as the theoretical
The measurements for the characteristics of elevation model to evaluate the performance of massive MIMO system.
angles are extensive for 3D MIMO system recently [11, 24]. The current analysis generally assumes that the UEs employ
The main measurement results of rectangle antenna array single antenna in consideration of the complexity and spaces
International Journal of Antennas and Propagation 5

Table 3: Channel models of massive MIMO.

Modeling method Category Property
i.i.d. Rayleigh channel model Elements of fast fading are i.i.d. complex Gaussian variables
Contain correlation between transmit antennas or/and receive
CBSM Correlation channel model
antennas
Consider antenna impedance, load impedance, and mutual
Mutual coupling channel model
impedance
2D channel model Propagate beam on 2D plane, such as linear array
GBSM Propagate beam on 3D plane, such as spherical, rectangular, and
3D channel model
cylindrical array

for the terminal equipment with low carrier frequency. According to the large number of laws, we can find one char-
Therefore, the following CBSMs are discussed when the UE acteristic of the favorable propagation is the orthogonality of
employs single antenna, which are easy to be extended to different UE channels:
the cases that the UE with multiple antennas works at the
1 𝐻 0, 𝑖 ≠ 𝑗;
millimeter wave. Moreover, massive MIMO is more likely to h h ≈{ (3)
be used in time division duplex (TDD) system at present, 𝑁 𝑖 𝑗 1, 𝑖 = 𝑗.
since the downlink channel state information (CSI) can be Another is the channel harden phenomenon; namely, the
acquired from the uplink due to TDD reciprocity feature. Euclidean norm of each UE channel approximates to the
Therefore, only the uplink channel is taken as an example for large-scale fading factor [9]:
study in this section.
Let us consider an uplink MIMO system that 𝐾 single- 1 󵄩󵄩 󵄩󵄩2
󵄩g 󵄩 ≈ 𝛽𝑘 , 𝑘 = 1, 2, . . . , 𝐾. (4)
antenna UE transmits signals to the eNB equipped with 𝑁 󵄩 𝑘󵄩
𝑁 antennas simultaneously. When the NLoS channel is The favorable propagation can not only better the per-
assumed, the channel model can be generalized by formance but also simplify the algorithm design of mas-
sive MIMO system. The orthogonality can alleviate the
interuser/intercell interference, which is helpful to improve
G = HD1/2 , (1) the system capacity. Channel harden phenomenon can miti-
gate the impact of fast fading on the scheduling gain, which
simplifies the complexity of scheduling scheme [16].
where D = diag{𝛽1 , 𝛽2 , . . . , 𝛽𝐾 } is the large-scale propagation
matrix and 𝛽𝑘 = 𝜙𝑑𝑘−𝛼 𝜉𝑘 . 𝜙 is a constant related to the antenna
gain and carrier frequency, 𝑑𝑘 is the distance between the 4.1.2. Correlation Channel Model. In order to reflect the
antenna correlation caused by the insufficient antenna space
eNB and the 𝑘th UE, 𝛼 is the path loss exponent, and 𝜉𝑘 is
2 and the scattering environments, the correlation channel
the log-normal shadow fading with 10 log10 𝜉𝑘 ∼ N(0, 𝜎sh ).
𝑁×𝐾 model is established to evaluate the performance of the
H∈C is the fast fading matrix. massive MIMO system [25]. The fast fading channel vector
According to the fast fading matrix, the CBSMs can of each one of UE can be formed by the correlation matrix
be further simplified into three kinds of channels, namely, multiplied by standard complex Gaussian vector; namely,
i.i.d. Rayleigh fading model, correlation channel model, and
mutual coupling channel model. We discuss these three kinds h𝑘 = R𝑘 k𝑘 , 𝑘 = 1, 2, . . . , 𝐾, (5)
of models as follows in detail.
where the steering matrix R𝑘 ∈ C𝑁×𝐷𝑘 contains 𝐷𝑘 steering
vectors with different angles of arrival (AoAs) for the 𝑘th UE
4.1.1. i.i.d. Rayleigh Channel Model. The i.i.d. Rayleigh fading and k𝑘 ∼ CN(0, I𝐷𝑘 ).
channel model is widely adopted for the theoretical analysis When the linear antenna array is assumed at the eNB, the
of the massive MIMO system, which assumes there is no steering matrix R𝑘 can be written as
correlation and mutual coupling between transmit antennas
1
or receiver antennas. The elements of the fast fading matrix R𝑘 = [a (𝜃𝑘,1 ) , a (𝜃𝑘,2 ) , . . . , a (𝜃𝑘,𝐷𝑘 )] , (6)
H = [h1 , h2 , . . . , h𝐾 ] are i.i.d. Gaussian random variables; 𝐷𝑘
namely, ℎ𝑛,𝑘 ∼ CN(0, 1) (𝑛 = 1, 2, . . . , 𝑁; 𝑘 = 1, 2, . . . , 𝐾). where 𝜃𝑘,𝑖 is the 𝑖th AoA of 𝑘th UE and the steering vector
The favorable propagation is the most important property
a(𝜃𝑘,𝑖 ) ∈ C𝑁×1 is given as
of the i.i.d. Rayleigh fading channel in the massive MIMO
system [8]; namely, 𝑇
a (𝜃𝑘,𝑖 ) = [1, 𝑒(𝑗2𝜋𝑑/𝜆) sin 𝜃𝑘,𝑖 , . . . , 𝑒(𝑗2𝜋(𝑁−1)𝑑/𝜆) sin 𝜃𝑘,𝑖 ] , (7)

1 𝐻 where 𝑑 is the distance between the adjacent antennas and 𝜆

G G ≈ D, 𝑁 ≫ 𝐾. (2) is the carrier wavelength.
𝑁
6 International Journal of Antennas and Propagation

For the rectangle antenna array, the steering vector can be where R = [R1 , R2 , . . . , R𝐾 ]. As the UE is geometrically
attained by the steering matrix; namely, distributed, the received correlation matrix can be assumed
as I𝐾 . Then, we can get H ∼ CN(0, Σ𝑁 ⊗ I𝐾 ).
a (𝜃𝑘,𝑖 , 𝜙𝑘,𝑖 ) Mutual coupling channel model is more practical for
𝑇 massive MIMO. It is helpful to analyze the effects of antenna
= vec {[1, 𝑒(𝑗2𝜋𝑑/𝜆) sin 𝜃𝑘,𝑖 , . . . , 𝑒(𝑗2𝜋(𝑁−1)𝑑/𝜆) sin 𝜃𝑘,𝑖 ] space on the performance of massive MIMO, which is
significant for designing the antenna array configuration,
⊗ [1, 𝑒(𝑗2𝜋𝑑/𝜆) sin 𝜙𝑘,𝑖 , . . . , 𝑒(𝑗2𝜋(𝑁−1)𝑑/𝜆) sin 𝜙𝑘,𝑖 ] } , especially the sparse antenna array.

(8)
4.2. GBSM. GBSMs are mainly used for evaluating the
where 𝜃𝑘,𝑖 and 𝜙𝑘,𝑖 represent azimuth of arrival (AoA) and performance of practical wireless communication systems,
elevation of arrival (EoA), respectively. vec{⋅} denotes vector- which accurately comprise the channel properties. Consider-
ization of matrix. Note that AoA denotes the abbreviation of ing the elevation or not, the GBSMs of massive MIMO can be
angle of arrival or azimuth of arrival for linear antenna array classified into two kinds, namely, 2D and 3D channel models.
or rectangle antenna array, respectively. When the linear antenna array is employed at the eNB, the tilt
This correlation channel model introduces the AoAs, angle is fixed and the 2D channel model is precisely enough
which can be utilized to distinguish the UE and improve for evaluating the performance. However, if the spherical,
the accuracy of channel estimation [26]. When the UE is cylindrical, or rectangular antenna array configurations are
located at different orientations, the UE channels can almost adopted at the eNB, the 3D channel model in consideration
be separated by angle information, thereby alleviating the of both elevation and azimuth should be established for
pilot contamination. It is also helpful to analyze the interuser evaluation.
or intercell interference and develop UEs scheduling or cell
cooperation to alleviate the interference. 4.2.1. 2𝐷 Channel Model. Employing the linear antenna
array, the nonstationary phenomenon and spherical wave-
4.1.3. Mutual Coupling Channel Model. Since the number of front have been discriminated as the main properties for
antennas increases sharply in the massive MIMO system, the the massive MIMO channel through channel measurements
mutual impedance has to be considered as of the limited space [17, 19]. The properties are helpful to decorrelate the channels
for antenna array. Moreover, the load impedance and antenna for different type of UE, thereby providing a favorable channel
impedance should also be characterized in order to reflect the [17]. However, they raise the difficulty to establish the channel
realistic channel model. model for the massive MIMO system. Therefore, the realistic
The channel vector of the 𝑘th UE can be written as follows channel model comprising the properties is still on the way at
in consideration of both the impedances and correlation [27]: present [10].
Similar to [17, 19], the nonstationary phenomenon is
h𝑘 = ZR𝑘 k𝑘 , 𝑘 = 1, 2, . . . , 𝐾, (9) also observed through the measurements for the linear 128-
antenna array in a semiurban area [31]. In order to test the
where Z ∈ C𝑁×𝑁 represents the mutual coupling matrix, R𝑘 ∈
performance of proposed algorithms for the more practical
C𝑁×𝐷𝑘 denotes the steering matrix containing 𝐷𝑘 steering
massive MIMO system, much work has been done on the
vectors of the receiver antenna array, and k𝑘 ∼ CN(0, I𝐷𝑘 ).
channel modeling of massive MIMO reflecting this non-
According to [27, 28], the mutual coupling matrix can be
stationary phenomenon. Therefore, a cluster-based channel
expressed as
model comprising the nonstationary phenomenon has been
−1 established, which can be seen as the extension of the COST
Z = (𝑍𝐴 + 𝑍𝐿 ) (Γ + 𝑍𝐿 I) , (10)
2100 channel model.
with In consideration of both the nonstationary phenomenon
𝑍𝐴 𝑍𝑀 0 ⋅⋅⋅ 0 and spherical waveform effect, an elliptical GBSM was given
[𝑍𝑀 𝑍𝐴 𝑍𝑀 ⋅ ⋅ ⋅ 0 ] for massive MIMO with linear antenna array [23, 32, 33].
[ ] First, clusters on lots of confocal ellipses with different major
[ ]
Γ = [ 0 𝑍𝑀 𝑍𝐴 ⋅ ⋅ ⋅ 0 ], (11) axis represent different resolvable delays, which make the
[ .. .. .. ]
[ . . d d . ] GBSM possesses the spherical wavefront property. Then, the
[ 0 0 ⋅ ⋅ ⋅ 𝑍𝑀 𝑍𝐴 ] birth-death process of clusters is introduced to describe the
nonstationary property. Numerical results of the established
where 𝑍𝐴 , 𝑍𝐿 , and 𝑍𝑀 represent the antenna impedance, channel demonstrate that the phases responses of the linear
load impedance, and mutual impedance, respectively. Here, antenna array are no longer linear and the AoAs of the
the mutual impedances are only considered between adjacent antenna array shift gradually, which coincides with the
antennas, which can be seen from the nonzero 𝑍𝑀. The measurement results in the realistic environments.
further discussion about the mutual coupling based on the
electromagnetic analysis can be found in [29, 30]. Therefore,
the transmit correlation matrix can be written as follows [27]: 4.2.2. 3D Channel Model. The 3D channel models have been
investigated in the academic for several years already, such as
Σ𝑁 = 𝐸 [HH𝐻] = 𝐾ZRR𝐻Z𝐻, (12) [34, 35]. Meanwhile, several projects have been established
International Journal of Antennas and Propagation 7

especially for the 3D channel model, such as WINNER + Generally, the propagation channel contains several clus-
project [36]. However, most of these models mainly depend ters (𝑛 = 1, 2, . . .) with different delay 𝜏𝑛 and power factors.
on the literature survey rather than on realistic measurements Each cluster contains some rays (𝑚 = 1, 2, . . . , 𝑀) with
[24]. Thus, they are considered insufficient or obviously the same delay. However, in the 3D channel model, the two
questionable for the realistic channel properties, such as the strongest clusters are spread into three subclusters, with fixed
cross-correlation matrix of large-scale parameters. Therefore, delay offset 0, 5, and 10 ns. The other related delays, angles,
the existing 3D channel models need to be improved through and power can be generated by the given distributions and
realistic measurements and further used to evaluate the some measured parameters in [11].
realistic scenarios.
In WINNER + projects, the main procedures of 3D When the double polarization array is adopted at the eNB,
MIMO channel model have been described in detail [11, 12, as shown in Figure 5, the 𝑛th cluster of the fast fading channel
37], as shown in Figure 3. The main parameters of 3D channel from the 𝑠th antenna of the eNB to the 𝑢th antenna of the UE
model contain shadow fading (SF), delay spread (DS), K- in the NLoS scenario can be expressed as follows [11, 12, 24,
factor, AoA, azimuth of departure (AoD), EoA, and elevation 36, 37]:
of departure (EoD), especially the EoA and EoD. The cross-
correlation matrix has been expanded to 7-dimension from
5-dimension in the 2D channel model. In the stage of
𝑀 𝐹 𝑇
3GPP Release 12 (R12), three scenarios have been defined, 𝑟𝑥,𝑢,𝜃 (𝜃𝑛,𝑚,EoA , 𝜑𝑛,𝑚,AoA )
namely, urban microcell with high UE density (3D-UMi), 𝐻𝑢,𝑠,𝑛 (𝑡) = √𝑃𝑛 ∑ [ ]
urban macrocell with high UE density (3D-UMa), and urban 𝑚=1 𝐹𝑟𝑥,𝑢,𝜑 (𝜃𝑛,𝑚,EoA , 𝜑𝑛,𝑚,AoA )

macrocell with one high-rise per sector and 300 m intersite 𝜃𝜃 𝜃𝜑

𝑒𝑗Φ𝑛,𝑚 −1 𝑒𝑗Φ𝑛,𝑚
√𝜅𝑛,𝑚
distance (ISD) (3D-UMa-H). Based on the procedure in [ ]
Figure 3, the 3D channel model of rectangle antenna array has ×[ 𝜑𝜃
]
−1 𝑒𝑗Φ𝑛,𝑚 𝑗Φ𝜑𝜑
√𝜅𝑛,𝑚 𝑒 𝑛,𝑚 (16)
been given based on measurements. The related parameters, [ ]
their distributions, and the cross-correlation matrix for the
3D channel model have been concluded in the technical spec- 𝐹𝑡𝑥,𝑠,𝜃 (𝜃𝑛,𝑚,EoD , 𝜑𝑛,𝑚,AoD )
×[ ]
ification [11]. Next, we will discuss the modeling procedure 𝐹𝑡𝑥,𝑠,𝜑 (𝜃𝑛,𝑚,EoD , 𝜑𝑛,𝑚,AoD )
briefly according to [11, 12, 36, 37].
𝑇 𝑇
The large-scale propagation of the 3D MIMO channel × 𝑒𝑗2𝜋(̂𝑟𝑟𝑥,𝑛,𝑚 ⋅𝑑𝑟𝑥,𝑢 /𝜆 0 +̂𝑟𝑡𝑥,𝑛,𝑚 ⋅𝑑𝑡𝑥,𝑠 /𝜆 0 +V𝑛,𝑚 𝑡) ,
consists of both the path loss and shadow fading. The path
loss model has to be classified into outdoor and outdoor-
to-indoor (O2I) models in consideration of the different
propagation environments [11]. The illustration of path loss where 𝑃𝑛 is the power normalized factor such that the power
for both outdoor and O2I can be simply depicted in Figure 4. of all clusters equals one. {Φ𝜃𝜃 𝜃𝜑 𝜑𝜃 𝜑𝜑
𝑛,𝑚 , Φ𝑛,𝑚 , Φ𝑛,𝑚 , Φ𝑛,𝑚 } are four
In the outdoor scenarios, the path loss is the function of the random initial phases of different polarisation combinations
carrier frequency 𝑓0 , distance 𝑑3D between the eNB and the for the 𝑚th ray of the 𝑛th cluster, and 𝜅𝑛,𝑚 is the cross
UE, break point distance 𝑑bp , the height of UE ℎUE , and the polarisation power ratio. 𝐹𝑟𝑥,𝑢,𝜃 and 𝐹𝑟𝑥,𝑢,𝜑 are the patterns
height of eNB ℎeNB ; namely, of the 𝑢th receiver antenna in the direction of the spherical
PLOD = 𝑓OD (𝑓0 , 𝑑3𝐷, 𝑑bp , ℎUE , ℎeNB ) . (13) basis vectors 𝜃̂ and 𝜑,
̂ respectively. Similarly, 𝐹𝑡𝑥,𝑠,𝜃 and 𝐹𝑡𝑥,𝑠,𝜑
are the same for the 𝑠th transmit antenna. The spherical unit
As for the O2I scenarios, three items constitute the path loss, vector 𝑟̂𝑥,𝑛,𝑚 can be given as
namely, basic path loss PLOD related to 𝑑3𝐷-out and 𝑑3𝐷-in
which is similar to the outdoor scenario, the loss through wall
PLTW , and the indoor loss PLIN = 0.5𝑑2𝐷-in ; namely,
cos 𝜃𝑛,𝑚,Eoy cos 𝜑𝑛,𝑚,Aoy
PLO2I = PLOD + PLTW + PLIN . (14) [ ]
̂𝑟x,𝑛,𝑚 = [ cos 𝜃𝑛,𝑚,Eoy sin 𝜑𝑛,𝑚,Aoy ] , (17)
The shadow fading 𝜉𝑛 is dependent on the cluster 𝑛, which sin 𝜃𝑛,𝑚,Eoy
follows log-normal distribution; namely, [ ]

𝜗 1 (𝜗 ln 𝑥)2 10
𝑓𝜉𝑛 (𝑥) = exp {− 2
}, 𝜗= .
𝑥 √2𝜋𝜎2 2𝜎sh,𝑛 ln 10
sh,𝑛 where “𝑥 = 𝑡𝑥, 𝑦 = 𝐴” and “𝑥 = 𝑟𝑥, 𝑦 = 𝐷” represent
(15) the spherical unit vectors for receiver and transmitter, respec-
tively. 𝑑𝑟𝑥,𝑢 is the location vector of the 𝑢th receiver antenna,
The correlation between adjacent shadow fading values can
be characterized as 𝑅(Δ𝑥) = exp{−|Δ𝑥|/𝑑cor }, where 𝑑cor is 𝑑𝑡𝑥,𝑠 is the location vector of the 𝑠th transmit antenna, and 𝜆 0
𝑇
the correlation distance related to the propagation environ- is the wavelength. Doppler frequency 𝑓𝑛,𝑚 = 𝑟̂𝑟𝑥,𝑛,𝑚 ⋅V/𝜆 0 with
𝑇
ments. V = V ⋅ [cos 𝜃V cos 𝜑V , cos 𝜃V sin 𝜑V , sin 𝜃V ] .
8 International Journal of Antennas and Propagation

Set scenario, network, and antenna parameters

Assign propagation condition (NLOS/LOS)

Calculate path loss

Azimuth of departure (AoD) Azimuth spread at departure (ASD)

Subpath offset of AoD Elevation spread at departure (ESD)
Elevation of departure (EoD)
Subpath offset of EoD Lognormal shadowing (SF)

Path delays Delay spread (DS)

Average path powers
Azimuth of arrival (AoA) K factor
Subpath offset of AoA
Elevation of arrival (EoA) Azimuth spread at arrival (ASA)
Subpath offset of EoA Elevation spread at arrival (ESA)

Draw random initial phases

Generate channel coefficient

Apply path loss and shadowing

Figure 3: The procedure of 3D channel model [11, 12].

d3D
-out
Indoor
d3D
-in

d 3D
Outdoor
heNB

hUE
dbp
hUE

d2D d2D-out d2D-in

Figure 4: Illustration of path loss [11].

As for the LoS scenario, the fast fading with a main path where 𝛿(⋅) and 𝐾𝑅 are the indicative function and Rice factor,
can be written as follows [11, 12, 24, 36, 37]: respectively.
If only the vertical polarization is adopted in LoS or NLoS
1 𝐾𝑅 scenarios, the 2 × 2 polarisation matrix in (16) and (18) can
̃ 𝑢,𝑠,𝑛 (𝑡) = √
H H (𝑡) + 𝛿 (𝑛 − 1) √
𝐾𝑅 + 1 𝑢,𝑠,𝑛 𝐾𝑅 + 1 be replaced by exp(𝑗Φn,𝑚 ), and only vertically polarization
pattern is applied [11, 12].
𝑀 𝑇 According to the above procedures of channel modeling,
𝐹𝑟𝑥,𝑢,𝜃 (𝜃LoS,EoA , 𝜑LoS,AoA ) 𝑗ΦLoS
× ∑[ ] 𝑒 we can generate the realistic channel coefficients to evaluate
𝐹𝑟𝑥,𝑢,𝜑 (𝜃LoS,EoA , 𝜑LoS,AoA )
𝑚=1 the three aforementioned scenarios. However, the channel
model is not precise for simulations, since the nonstationary
1 0 𝐹 (𝜃 ,𝜑 )
×[ ] [ 𝑡𝑥,V,𝜃 LoS,EoD LoS,AoD ] phenomenon has not been reflected [19]. Moreover, the
0 −1 𝐹𝑡𝑥,V,𝜑 (𝜃LoS,EoD , 𝜑LoS,AoD )
channel models of other scenarios and antenna array config-
𝑇 𝑇 urations have to be given in order to extend the applications
× 𝑒𝑗2𝜋(̂𝑟𝑟𝑥,LoS ⋅𝑑𝑟𝑥,𝑢 /𝜆 0 +̂𝑟𝑡𝑥,LoS ⋅𝑑𝑡𝑥,𝑠 /𝜆 0 +VLoS 𝑡) , of massive MIMO. Therefore, the properties and modeling of
(18) 3D MIMO channel still need to be further investigated.
International Journal of Antennas and Propagation 9

z Conflict of Interests
The authors declare that there is no conflict of interests
eNB
𝜃 regarding the publication of this paper.
̂
𝜙
n̂
Acknowledgments
𝜃̂ y
𝜑 The work was supported by the China Natural Science
Funding (61271183), Program for New Century Excellent
x z UE Talents in University (NCET-11-0600), National Key Technol-
𝜃
ogy R&D Program of China under Grant 2013ZX03003005,
 and National High Technology Research and Development
Program of China (2014AA01A705).
x
𝜑 y

Figure 5: Illustration of coordinate [11, 12]. References

[1] Z. Hasan, H. Boostanimehr, and V. K. Bhargava, “Green cellular
networks: a survey, some research issues and challenges,” IEEE
Communications Surveys & Tutorials, vol. 13, no. 4, pp. 524–540,
2011.
[2] D. Feng, C. Jiang, G. Lim, L. J. Cimini Jr., G. Feng, and G. Y. Li,
“A survey of energy-efficient wireless communications,” IEEE
Communications Surveys & Tutorials, vol. 15, no. 1, pp. 167–178,
5. Conclusion 2013.
[3] Y. Chen, S. Zhang, S. Xu, and G. Y. Li, “Fundamental trade-offs
Massive MIMO has been regarded as one of efficient ways on green wireless networks,” IEEE Communications Magazine,
to improve both spectrum efficiency and energy efficiency vol. 49, no. 6, pp. 30–37, 2011.
for the broadband wireless communication systems. As the [4] G. Y. Li, Z. Xu, C. Xiong et al., “Energy-efficient wireless com-
first step to evaluate the performance of any communication munications: tutorial, survey, and open issues,” IEEE Wireless
systems, the channel models for massive MIMO are necessary Communications, vol. 18, no. 6, pp. 28–35, 2011.
to be investigated, which is the main concern of this paper. [5] G. Auer, V. Giannini, C. Desset et al., “How much energy is
According to the ability of array to radiate the signals, the needed to run a wireless network?” IEEE Wireless Communi-
antenna array configurations are classified into 2D and 3D cations, vol. 18, no. 5, pp. 40–49, 2011.
antenna arrays. Generally, the linear antenna array is 2D [6] E. G. Larsson, O. Edfors, F. Tufvesson, and T. L. Marzetta,
antenna array, and the rectangle, spherical, and cylindrical “Massive MIMO for next generation wireless systems,” IEEE
antenna array belong to 3D antenna array. From studying Communications Magazine, vol. 52, no. 2, pp. 186–195, 2013.
the current measurement results of different configurations, [7] A. Ghrayeb and T. M. Duman, Coding for MIMO Communica-
it can be found that the linear antenna array gives rise to tion System, John Wiley & Sons, New York, NY, USA, 2007.
the nonstationary phenomenon and near-field effect. The 3D
[8] T. Marzetta, “Noncooperative cellular wireless with unlimited
antenna array has lower angular resolution, yet it can resolve numbers of base station antennas,” IEEE Wireless Communica-
in both azimuth and elevation. Increasing the number of tions, vol. 9, no. 11, pp. 3590–3600, 2010.
antennas at eNB can decorrelate different user channels.
[9] F. Rusek, D. Persson, B. K. Lau et al., “Scaling up MIMO:
At present, CBSMs are mainly used to analyze the theoret-
opportunities and challenges with very large arrays,” IEEE
ical performance of massive MIMO due to its simplification. Signal Processing Magazine, vol. 30, no. 1, pp. 40–60, 2013.
There are three kinds of simplified CBSMs widely used for
[10] F. Tufvesson, “Very large MIMO systems ICASSP, tutorial—part
different objectives, namely, i.i.d. Rayleigh channel model,
II: propagation aspects of very large MIMO,” in Proceedings of
correlation channel model, and mutual coupling channel
the 37th IEEE International Conference on Acoustics, Speech, and
model. The channel model reflecting both the nonstationary Signal Processing (ICASSP ’12), 2012.
phenomenon and spherical wave effect has been established
[11] “Technical specification group radio access networkstudy on 3D
for the linear antenna array based on the cluster model. The
channel model for LTE, (Rel 12),” 3GPP TR- 36.873 V1.1.1, 2013.
improved 3D channel model of rectangle antenna array has
been proposed in 3GPP in order to evaluate three kinds of [12] Y.-H. Nam, B. Ng, K. Sayana et al., “Full-dimension MIMO
scenarios. However, the channel models of other antenna (FD-MIMO) for next generation cellular technology,” IEEE
Communications Magazine, vol. 51, no. 6, pp. 172–179, 2013.
array configurations and scenarios have to be supplemented,
and the nonstationary property for spherical, cylindrical, [13] “Requirements, candidate solutions and technology roadmap
and rectangular antenna array needs further measurements. for lte rel-12 onward,” 3GPP RWS-120010, 2012.
Therefore, the channel model of massive MIMO is still an [14] “Technologies for Rel-12 and onward,” 3GPP RWS- 120021, 2013.
open topic, which needs much more investigation for its [15] “Views on Rel-12 and onwards for LTE and UMTS,” 3GPP RWS-
improvements in the next steps. 120006, 2013.
10 International Journal of Antennas and Propagation

[16] B. M. Hochwald, T. L. Marzetta, and V. Tarokh, “Multiple- [33] A. Ghazal, C.-X. Wang, H. Haas et al., “A non-stationary MIMO
antenna channel hardening and its implications for rate feed- channel model for high-speed train communication systems,”
back and scheduling,” IEEE Transactions on Information Theory, in Proceedings of the 75th IEEE Vehicular Technology Conference
vol. 50, no. 9, pp. 1893–1909, 2004. (VTC ’12), pp. 1–5, Yokohama, Japan, June 2012.
[17] S. Payami and F. Tufvesson, “Channel measurements and [34] T. Aulin, “A modified model for the fading signal at a mobile
analysis for very large array systems at 2.6 GHz,” in Proceedings radio channel,” IEEE Transactions on Vehicular Technology, vol.
of the 6th European Conference on Antennas and Propagation 28, no. 3, pp. 182–203, 1979.
(EuCAP ’12), pp. 433–437, Prague, Czech Republic, March 2012. [35] K. Kalliola, H. Laitinen, P. Vainikainen, M. Toeltsch, J. Laurila,
[18] X. Gao, O. Edfors, F. Rusek, and F. Tufvesson, “Linear pre- and E. Bonek, “3-D double-directional radio channel character-
coding performance in measured very-large MIMO channels,” ization for urban macrocellular applications,” IEEE Transactions
in Proceedings of the 74th IEEE Vehicular Technology Conference on Antennas and Propagation, vol. 51, no. 11, pp. 3122–3133, 2003.
(VTC ’11), pp. 1–5, Budapest, Hungary, September 2011. [36] J. Meinilä, P. Kyösti, L. Hentilä et al., “D5.3: WINNER+
[19] X. Gao, F. Tufvesson, O. Edfors, and F. Rusek, “Measured final channel models,” Wireless World Initiative New Radio
propagation characteristics for very-large MIMO at 2.6 GHz,” WINNER, 2010.
in Proceedings of the 46th IEEE Asilomar Conference on Signals, [37] “WINNER II interim channel models,” IST-WINNER, 2006.
Systems and Computers (ASILOMAR ’12), pp. 295–299, Pacific
Grove, Calif, USA, November 2012.
[20] J. Hoydis, C. Hoek, T. Wild, and S. ten Brink, “Channel
measurements for large antenna arrays,” in Proceedings of the
9th IEEE International Symposium on Wireless Communication
Systems (ISWCS ’12), pp. 811–815, Paris, France, August 2012.
[21] “C-RAN: the road towards green radio access network,” Tech.
Rep., China Mobile Research Institute, 2012.
[22] “Active antenna system: utilizing the full potential of radio
sources in the spatial domain,” Tech. Rep., Huawei, 2012.
[23] C.-X. Wang and S. Wu, “Massive MIMOchannel measurements
and modeling: advances and challenges,” submitted to IEEE
Wireless Communion Magazine.
[24] Z. Zhong, X. Yin, X. Li, and X. Li, “Extension of ITU IMT-
advanced channel models for elevation domains and line-of-
sight scenarios,” in Proceedings of the 78th IEEE Vehicular
Technology Conference (VTC ’13), pp. 1–5, Las Vegas, Nev, USA,
September 2013.
[25] J. Hoydis, S. ten Brink, and M. Debbah, “Massive MIMO in the
UL/DL of cellular networks: how many antennas do we need?”
IEEE Journal on Selected Areas in Communications, vol. 31, no.
2, pp. 160–171, 2013.
[26] H. Yin, D. Gesbert, M. Filippou, and Y. Liu, “A coordinated
approach to channel estimation in large-scale multiple-antenna
systems,” IEEE Journal on Selected Areas in Communications,
vol. 31, no. 2, pp. 264–273, 2013.
[27] C. Masouros, M. Sellathurai, and T. Ratnarajah, “Large-scale
MIMO transmitters in fixed physical spaces: the effect of
transmit correlation and mutual coupling,” IEEE Transactions
on Communications, vol. 61, no. 7, pp. 2794–2804, 2013.
[28] B. Clerckx, C. Craeye, D. Vanhoenacker-Janvier, and C. Oestges,
“Impact of antenna coupling on 2 × 2 MIMO communications,”
IEEE Transactions on Vehicular Technology, vol. 56, no. 3, pp.
1009–1018, 2007.
[29] C. Masouros, J. Chen, K. Tong, M. Sellathurai, and T. Rat-
narajah, “Towards massive-MIMO transmitters: on the effects
of deploying increasing antennas in fixed physical space,” in
Proceedings of the Future Network and Mobile Summit, pp. 1–10,
2013.
[30] C. A. Balanis, Antenna Theory: Analysis and Design, John Wiley
& Sons, Hoboken, NJ, USA, 2012.
[31] X. Gao, F. Tufvesson, O. Edfors, and F. Rusek, “Channel behav-
ior for very-large MIMO systems-initial characterization,” in
Proceedings of the Joint Workshop on Small Cell Cooperative
Communications (COST IC1004 ’12), 2012.
[32] E. H. Aggoune, “A non-stationary wideband channel model for
large MIMO communication systems”.
 International Journal of

Rotating
Machinery

International Journal of
The Scientific
Engineering Distributed
Journal of
Journal of

Hindawi Publishing Corporation

World Journal
Hindawi Publishing Corporation Hindawi Publishing Corporation
Sensors
Hindawi Publishing Corporation
Sensor Networks
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of

Control Science
and Engineering

Advances in
Civil Engineering
Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Submit your manuscripts at

http://www.hindawi.com

Journal of
Journal of Electrical and Computer
Robotics
Hindawi Publishing Corporation
Engineering
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

VLSI Design
Advances in
OptoElectronics
International Journal of

International Journal of
Modelling &
Simulation
Aerospace
Hindawi Publishing Corporation Volume 2014
Navigation and
Observation
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
in Engineering
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Engineering
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
http://www.hindawi.com Volume 2014

International Journal of
International Journal of Antennas and Active and Passive Advances in
Chemical Engineering Propagation Electronic Components Shock and Vibration Acoustics and Vibration
Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014