Massive MIMO Design for 5G Networks: An Overview on Alternative Antenna Configurations and Channel Model Challenges

Full text

Massive MIMO Design for 5G Networks: An
Overview on Alternative Antenna Configurations
and Channel Model Challenges
H. M. El Misilmani and A. M. El-Hajj
ECE Department, Beirut Arab University, Lebanon
hilal.elmisilmani@ieee.org, a.elhajj@bau.edu.lb
Abstract—With the growth of mobile data application and
the ultimate expectations of 5G technology, the need to expand
the capacity of the wireless networks is inevitable. Massive
MIMO technique is currently taking a major part of the ongoing
research, and expected to be the key player in the new cellular
technologies. This papers presents an overview of the major
aspects related to massive MIMO design including, antenna
array general design, configuration, and challenges, in addition
to advanced beamforming techniques and channel modeling and
estimation issues affecting the implementation of such systems.
Keywords—Massive MIMO, 5G, Antenna arrays, Channel
estimation, Beamforming.
I. INTRODUCTION
The fifth generation of wireless communication standards
is the next evolution that is expected to hit the markets by
2020. In addition to improved data rates of up to 10 Gbps
and reduced latencies below 1 ms, this evolution promises
to enable a network of connected machines, devices that
operate in conjunction with regular subscribers [1]. This will
introduce new communication mechanisms such as device to
device communication (D2D). The road to 5G deployment
is essentially facilitated by the introduction of new concepts
that will help 5G systems reach the projected theoretical
specifications. Among others, ultra-network densification will
transform the traditional cell architecture from a collection
of macrocells covering large areas to multitude of small cells
providing higher capacity and better services to the users while
using a lower transmit power. The jump to the millimeter wave
band is another novelty which will allow to benefit from very
large bandwidths and achieving very high data rates. However,
these high frequencies impose additional constraints to the
system design in terms of signal blockage and attenuation.
This is why multi-antenna approaches such as Massive MIMO
become a necessity in future communication standards since
they enable an efficient adaptation of the parameters of the
transmitted signal to counteract the millimeter wave channel
effects.
The interference is the main limitation of wireless networks.
Communications engineers have strived to exploit the prop-
erties of multipath wireless channel in order to improve the
performance of communications standards through an increase
of the radio link capacity. Several interference reduction
techniques have been studied, such as: multiuser MIMO [2],
multicell processing [3], and interference alignment [4]. How-
ever, these techniques cannot be used to reach the high data
rates expected from future technologies. Network densification
is taking a big interest in research as a candidate solution.
On way of applying this technique is by cell-size shrinking.
This could be done by installing femto or small cells [5]
, but this increases interference and adds cost of additional
equipment. Another option that is taking huge interest in
wireless communication is the use of Very Large MIMO arrays
or Large-Scale Antenna Systems, known as Massive MIMO.
This technique, similarly called Full Dimension MIMO, Hyper
MIMO, and ARGOS, use a great number of elements, fully
operating in a coherent and adaptive way. Massive MIMO
takes the original concept of multiple-input multiple-output
to another level going from tens to hundreds or thousands
of antennas with the aim of increasing the spectral efficiency
in the system, and providing a uniform quality of service in
different environments, notably in urban and suburban areas
complementing or even replacing the process ultra-network
densification.
Usually, a single element antenna possesses a poor di-
rectivity with a relatively wide and wide radiation pattern.
For 5G technology, high directivity is strongly required. This
can be achieved through constructing antenna arrays, in a
suitable electrical and geometrical configuration, without the
need for optimizing the size of antenna elements, which is the
motivation behind the use of Massive MIMO. Since it was first
introduced [6], the use of massive antennas has been receiving
important interest in wireless technology [6]–[14]. Some of the
work focused on the number of antenna elements needed to
achieve optimized performance [10], other investigated detec-
tion methods that can further enhance the gain [11]. In [12] the
motivation of massive MIMO and a comparison between using
more antennas or more base stations are presented. Moreover,
a TDD cellular system employing nocooperative BSs equipped
with Massive MIMO is presented in [6]. A recitation on
MIMO progress, importance, and challenges facing Massive
MIMO from a detection perspective is presented in [13].
Massive MIMO are considered to be adopted in 5G network,
at the Base Station. These large-sized antenna arrays can adapt

(a) (b)
Figure 1. (a) Single element structure, (b) Array configuration using 4×4
antenna elements [15]
flexibly to complex environment, and by scaling up the order
of MIMO system and applying beamforming techniques, the
signal transmitted from the BSs can be highly focused into
small regions of interest, towards each user, resulting in greatly
reduced interference. As a result, the spatial multiplexing
in each time-frequency resource block, along with multi-
antenna diversity and beamforming, is expected to improve
the transmission rate, the multiplexing capability, the spectrum
efficiency, and maximize the signal-to-noise-plus-interference
ratio (SNIR or SINR). Even, a nearly interference-free com-
munication link would be established between the user and its
BS, if highly directly beams with low sidelobe levels are used.
The performance can be further enhanced if more antennas
are at the BS, and eventually higher data rates required in
5G can be achieved. Further enhancement can be realized
by installing more antennas in the users’ mobile devices. In
addition, as a result of the channels orthogonality of different
users, increasing the number of antennas can actually result
in simpler transmit/receive processing techniques, even in the
presence of interference. Nevertheless, through averaging out
many random impairments, Massive MIMO systems are able
to enhance the energy efficiency with potential power savings,
while providing robust and secure communicating link.
In the following section a brief overview on the antennas
used in Massive MIMO is presented, along with the main
configurations and challenges found. Next, 5G channel mod-
eling and estimation in the presence of Massive MIMO is
investigated, and finally the beamforming concepts in Massive
MIMO are examined.
II. MA SS IV E MIMO ANT EN NAS
A. General Description and Review
Usually, for a simpler and practical antenna system, the
design of different array elements constituting the antenna
array is identical, however this is not necessary. Having such
an array results in more freedom in controlling the array
pattern of an array without changing its physical dimensions.
This is done through adopting proper geometrical antenna
(a)
(b)
Figure 2. (a) Prospective view the proposed antenna unit, (b) Subarray with
four antenna units [18]
array configuration. This antenna configuration, along with the
pattern of a single element, the separation between different
elements and mutual coupling, exhibit a significant effect on
performance of the system.
Theoretically speaking, and neglecting the coupling between
the different elements, the fields radiated by the individual
elements can be added using vector addition to determine the
total field of the array. Since every element has its own pattern,
a constructive interference of the different fields is essential
in the desired direction, whereas a destructive interference is
required to cancel the radiation in other directions, resulting in
a maximum intensity at a specific desired location [16]. More
advanced antenna array design consists of the use of phased
arrays. In a phased array, the feeding mechanism is designed in
such a way that different relative phases will be used with the
different antenna elements, reinforcing the maximum radiation
pattern in a desired direction [17].
An essential objective behind the use of Massive MIMO in
5G technology is to control the overall pattern of the antenna
for interference reduction and long distance communication
over high frequency. This pattern is highly affected by the ar-
ray configuration, the separating distance between the antenna
elements, the phase and amplitude of the excitation of different
elements, and the corresponding pattern of each.
In the design process, the chosen configuration should
be studied in terms of the total number of antenna ele-
ments, the resulting radiation characteristics: radiation pattern,
beamwidth and gain. In addition, a care should be taken in
studying the mutual coupling between the elements and how
they affect the power of the received signal, the coverage,
and the overall channel capacity. A proper study of the above
should be done on one or more of the operating frequencies
of the 5G technology, the 6 GHz, 27 to 28 GHz band, and 60
to 70 GHz bands.
Different anttennas can be utilized in such arrays, such as

Figure 3. The configuration of Turning Torso antenna array [24]
printed or microstrip antennas, horn, and dipoles antennas.
The choice of the element type depends on the application in
hand, the performance needed, the overall size of the system.
A4×4array configuration of a compact millimeter wave
BS, having 36 sub-sectors consisiting each of a 16 planar
ALTSA elements is presented in [15]. The antenna, shown
in Fig. 1, has a half-power beamwidths of 10.7◦and 5.3◦, in
the H- and E- planes respectively, with a realized gain of 25.6
dBi. An antenna array of 144 ports having dual polarization,
horizontal and vertical, and operating at 3.7 GHz is presented
in [18]. The array has 18 low profile sub-arrays, each sub-
array consists of 4 elements connected to power splitters. The
antenna array is shown in Fig. 2. A three stack levels Massive
MIMO system of orthohexagonal rings consisting each of six
sub-arrays is presented in [19] to establish a compact dual-
polarized antenna. The antenna, shown in Fig. 3, has a gain of
16.6 dBi, with a HPBW of 12.5◦in the azimuth plane. Using
the steerable feature, 18 beams can be generated covering a
whole circumference. Massive MIMO using active antennas
of 32 ports is presented in [20], providing an increased cell
average throughput gain compared to conventional system. A
massive MIMO system based on multi-mode antennas design
to operate as an UWB system in the 6−8.5GHz band
is presented in [21]. Different number of antenna elements
consisting each of a miniaturized circular patch microstrip
antenna, operating at 1.8 GHz, is presented in [22]. A method
to synthesize the array patterns with a desired sidelobe level
with uniform circular arrays using Dolph-Chebyshev approach
has been presented in [23].
B. Antenna Configuration
Massive MIMO antennas could be collocated at the base
station site in a uniform linear, square, circular, or cylindrical
arrays. They can be distributed geographically or installed on
the face of a building.
1) Planar Array: In linear arrays, the elements are located
along a straight line, vertically or horizontally, as shown in
Figs. 4, (a) and (b). This topology is called uniform linear
arrays (ULAs) when the inter-element spacing is uniform.
ULAs with equal sidelobes, which lead to the narrowest
beamwidth, are presented in [25] and are denoted the Cheby-
shev arrays. Taylor arrays, on the other hand, are famous for
their decaying sidelobes. In the synthesis of Chebyshev and
(a) (b) (c)
Figure 4. (a) Horizontal array, (b) Vertical array, (c) Planar or Rectangular
array
Taylor ULAs, it is possible to control the sidelobe level but
not the beamwidth. More recently, a method for the design
of ULAs with independently controllable beamwidth and SLL
has been proposed in [26]. To provide additional variables,
planar arrays, shown in Fig. 4 (c), are adopted to control the
radiation of the antenna. In planar arrays, the antenna elements
are arranged over some planar surface, called UPA, in a planar
or rectangular form. The use of such array configuration results
in reduced lower sidelobes, while directing the maximum
radiation in a desired location, one of the main objectives of
5G transmission.
For Mand Nelements arranged along the x- and y-
axes respectively, and assuming different excitation of each
element, the array factor can be given by [16]:
AF =SxmSyn (1)
where:
Sxm =
M
X
m=1
Im1ej(m−1)(kdxsin θcos φ+βx)
Sym =
N
X
n=1
I1nej(n−1)(kdysin θsin φ+βy)
(2)
with dx,dyare the separating distance between the elements,
and βx,βy, are the progressive phase shift between them,
along the x−and y−axes respectively. Assuming a uniform
excitation, the normalized array factor can be given by [16]:
AFn(θ, φ) = 

1
M
sin M
2ψx
sin ψx
2


1
N
sin N
2ψy
sin ψy
2
(3)
with ψx=kdxsin θcos φ+βxand ψy=kdysin θsin φ+
βy.
A planar version of Chebyshev ULAs is presented in [27].
The planar version of Taylor arrays is given in [28]. The design
of planar arrays with independently adjustable beamwidth and
SLL is introduced in [29].
2) Circular and Cylindrical Arrays: In circular arrays,
shown in Fig. 5 (a), the antenna elements are arranged in a
circular ring, called UCA. This configuration usually provides

wider angle radiation direction. Assuming the peak of the main
beam is directed in the (θ0, φ0)direction, the array factor in
this case can be given by [16]:
AF (θ, φ) =
N
X
n=1
Inejkρ0cos(θn−ξ)(4)
with Inbeing the amplitude excitation of the n,ρ0and ξ
given by [16]:
ρ0=a(sin θcos φ−sin θ0cos φ0)2
+ (sin θsin φ−sin θ0sin φ0)21/2(5)
ξ= tan−1sin θsin φ−sin θ0sin φ0
sin θcos φ−sin θ0cos φ0(6)
(a) (b)
Figure 5. (a) Circular array, (b) Cylindrical array
Through stacking circular arrays of equal radius one above
the other, separated by the same displacement, cylindrical
arrays can be formed. Cylindrical array, shown in Fig. 5
(b), can be actually seen as a linear array along the vertical
line on the cylindrical surface, and a circular array along the
transversal plane cutting the cylindrical surface. Both Circular
and cylindrical arrays possess the advantage of symmetry
in azimuth, which makes them ideally suited for full 360◦
coverage. One of the most important properties of cylindrical
arrays is that the multiplication of the array factors of both,
linear and circular array, results in the array factor of the whole
cylindrical array [30], [31], i.e.:
AF (θ, φ) = AFlinear(θ, φ)×AFcircular(θ, φ)(7)
Hence, while the circular arrays provides 360◦coverage,
stacking them in such cylindrical way provides increased gain
and directivity.
C. Design Challenges
Although Massive MIMO systems provide significant per-
formance enhancement with essential advantages, the use of a
large antenna of antenna arrays bring several challenges that
must be considered.
For instance, the use of Massive MIMO at the BSs could
result in increased hardware complexity and signal processing
costs. One solution for this is to decrease the size of the
antenna element. This can be done through working with
millimeter wave antennas operating at high frequency ranges,
for which 5G technology is expected to operate at. Another
technique to reduce this cost is to use electromagnetic lens
antenna (ELA). This lens, shown in Fig. 6, focuses the power
of any indicent plane wave to a small subset of the antenna
array [32], [33]. This system requires less number of required
RF chains, and as a result the implementation cost is reduced.
Figure 6. Proposed design with the EM-lens embedded antenna array [32]
Another issue is related to the different antenna array
configurations, for which each configuration results in different
channel characteristics and hence will have an impact on the
performance of the overall system. Assuming a linear array
system in adopted and for a fixed total number of elements, a
higher spectrum efficiency can be achieved by placing more
antennas in the horizontal direction, whereas, for the same
total number of antenna elements, the performance is degraded
in the horizontal direction if more antennas are used in the
vertical direction.
The separation between the antenna elements is another
important factor that affects the array radiation. Although
reducing the antenna spacing could meet the installation
requirements, if the elements displacement is less than
λ/2of the antenna, mutual coupling and fading correlation
become increasingly dominant, resulting in degrading the
capability of the Massive MIMO array to distinguish the
users, and hence degrading the system performance. Hence,
an important attention has to be made while choosing the
separation between the antenna elements. If the physical
space is limited, the separation between the antenna elements
must be reduced in order to increase the total number
of antenna elements. An investigation of this effect for a
single- or multi-users has been presented in [34]–[36]. It
was shown that that due to these effects, a practical limit on
the maximum number of BS antennas could be found that
results in a maximum spectral efficiency. To optimize the
system performance, matching networks can be integrated
with the design of compact antenna arrays. If strong mutual
coupling is found, applying an optimal matching improves
the performance, but reduces the system bandwidth [37]–[44].
The impact of such coupling and its effect on the bandwidth
of circular arrays has been investigated in [19].

Inspecting the general configuration of an array system at
the BS, a conventional 2D beamforming exhibits an important
deficiency. The transmitted beam is only adjusted in the
horizontal dimension, i.e. the beam angle is only adjusted
with the azimuth angle. This could be done by controlling
the signal distribution through altering the phase and ampli-
tude of the excitation on each antenna port. In the vertical
dimension, the transmission between the BS and cell users is
with a fixed angle. Hence, an optimal throughput could not
be achieved, and eventually, with equal azimuth angles of the
users, an inevitable interference will occur. In order to solve
this deficiency, 3D MIMO system is recommended, for which
Active Antenna Systems (AAS) are used. These AASs consist
of RF modules integrated with the design, used to control each
element separately. This results in an increased efficiency, and
flexible beam control. Hence, a Massive MIMO should be
designed with a 3D spacial channel models through adjusting
the beam angle with both azimuth and elevation angles.
III. 5G CHANNEL MODELING AND MASSIVE MIMO
The move from current sub-6 GHz transmission to the
millimeter range has a significant impact on the performance
of the large-scale antenna systems. While MIMO systems
have been considered as additional feature for current wire-
less communication technologies, massive MIMO systems
are a necessity for the operation of millimeter-wave based
systems [45]. The main reason for that is that at very high
frequencies, the pathloss of each of the links become rather
significant. The frequency bands projected for 5G operation
(e.g., 28 GHz) suffer from new forms of losses related to rain,
atmospheric gas absorption, foliage, etc. Several properties of
massive MIMO systems, such high array gains, are needed to
make communication viable, even for small distances.
Currently employed channel models assume random scat-
tering model for each link in addition to the presence of in-
dependent scatterers [46]. Moreover, the scattering considered
is of diffuse nature, ignoring the specular propagation where
the latter become notably dominant at high frequencies [47].
Hence, to correctly characterize the performance of the MIMO
system, a more realistic channel model is needed. This model
relies on spatial consistency where geometric locations of the
scatterers, in particular near the transmitter and receiver, needs
to be specified.
Furthermore, massive MIMO systems are characterized by
a high directly and pencil-shaped beamforms. Current channel
models based on plane wave propagation fail to provide the
necessary angular resolution in the analysis of the propagation
of each of the links originating from the MIMO array. Thus,
it has been recommended to use new channel models that
provide higher angular resolution and a better representation
of the amplitude distribution of each of the rays. The channel
models are also expected to rely on spherical wave propaga-
tion [46].
IV. CHANNEL ESTIMATION
Channel estimation is at the core of the operation of any
MIMO system. The knowledge of the channel state informa-
tion is necessary in order to perform adequate precoding in
the MIMO system. Time division duplexing (TDD) has been
the technique of choice to get channel information, mainly to
make use of the channel reciprocity principle [48]. According
to this principle, electromagnetic waves transmitted over the
same frequency band in the uplink and downlink, experience
the same propagation conditions. Using, this technology, the
need for feedback of channel estimates diminishes while
having channel state estimates at the transmitter. Recently,
frequency division duplexing was also investigated for channel
information acquisition in massive MIMO systems. The main
idea is to either implement precoding technique with partial
channel state information or use compressed sensing to reduce
the feedback overhead [49]. From an antenna array point of
view, there is a certain level of correlation among antennas.
Therefore, it is not always necessary to get the channel
estimates for all the antennas.
V. BE AM FO RM IN G CO NC EP TS I N MA SS IV E MIMO
The conventional MIMO concept was built around the idea
of utilizing advanced signal processing techniques to generate
beams with high directivity that is pointed to a target user,
and in the optimal case, having the weakest sidelobes in the
direction of the non-served user (thus causing minimal inter-
ference). The implementation of the intended beamforming
technique at the transmitter (transmit beamforming) or at the
receiver (receiver beamforming) offer the network designer
several axes of freedom to optimize the network performance.
The advent of millimeter wave technologies adds new con-
siderations to the design of beamforming systems. First and
foremost, the switch to higher frequency bands originated from
the need to use large bandwidths. A direct implication would
be a degradation in the SNR. As discussed in [50], this will
result in problems during the beam-formation especially that
the array gain is small in this phase since wide beams are used
for user localization.
Another issue is the large bandwidth employed which is
usually greater than the coherence bandwidth of the chan-
nel [51]. Normally, this indicates a frequency selective chan-
nel and a large possibility for intersymbol interference. So,
advanced equalization needs to be added, complementing
the selected beamforming technique. The deep changes in
the propagation properties of the channel also lead to two
emerging topics related to beamforming in Massive MIMO,
namely, elevation beamforming [52] and efficient codebook
design [53]. The main premise of elevation or 3D beam-
forming is to exploit the channels degrees of freedom in
the elevation direction [52], [54] paving the way to what is
known as full dimension MIMO systems. In simple terms, the
beams are adjusted in the horizontal and vertical direction.
A necessity for the implementation of such systems involves

the definition of double directional channel models [55]. In
addition to 3D beamforming, these channel model will allow
us to design beam-tracking which are necessary to circumvent
the shortcomings of millimeter wave transmission in terms of
signal blockage and small range (e.g., [56]).
Efficient codebook design is another green area of research.
The main premise of the approach is to reduce the hardware
complexity in the implementation of beamforming techniques
by having fixed beam patterns which are generated by pre-
determined antenna weight vectors at the transmitter and
receiver. One advantage of codebook design is that it can be
easily designed for any antenna array geometry and specifi-
cations [50]. In [53], an efficient codebook-design algorithm
for millimeter wave-based massive MIMO system is presented.
The approach is based on cross-entropy optimization to jointly
identify the optimal analog precoder and analog combiner pair.
VI. CONCLUSION
This paper presented an overview on 5G technology require-
ments that are expected to be facilitated by Massive MIMO
technology. An overview of this promising systems has been
presented, with a major focus on its antenna part, general
design and antenna configurations, along with major design
challenges. Then, the modeling of channels in 5G with the
presence of Massive MIMO has been discussed, along with
the channel estimation and beamforming concepts.
REFERENCES
[1] E. Dahlman, S. Parkval, and J. Skold, 4G, LTE-Advanced Pro and The
Road to 5G, 3rd ed. Elsevier-Academic Press, 2016.
[2] D. Gesbert, M. Kountouris, R. W. Heath, C.-B. Chae, and T. Salzer,
“Shifting the MIMO paradigm,” IEEE Signal Processing Magazine,
vol. 24, no. 5, pp. 36–46, September 2007.
[3] D. Gesbert, S. V. Hanly, H. Huan, S. Shamai, O. Simeone, and W. Yu,
“Multi-cell MIMO Cooperative Networks: A New Look at Interference,”
IEEE Journal on Selected Areas in Communications, vol. 28, no. 9,
p. 13801408, December 2010.
[4] M. Maddah-Ali, A. S. Motahari, and A. Khandani, “Communication
over MIMO X Channels: Interference Alignment, Decomposition, and
Performance Analysis,” IEEE Transactions on Information Theory,
vol. 54, no. 8, p. 34573470, August 2008.
[5] J. Hoydis, M. Kobayashi, and M. Debbah, “Green small-cell networkse,”
IEEE Vehicular Technology Magazine, vol. 6, no. 1, pp. 37–43, March
2011.
[6] T. L. Marzetta, “Noncooperative Cellular Wireless with Unlimited
Numbers of Base Station Antennas,” IEEE Transactions On Wireless
Communications, vol. 9, no. 11, pp. 3590–3600, November 2010.
[7] J. Jose, A. Ashikhmin, P. Whiting, and S. Vishwanath, “Channel
Estimation and Linear Precoding in Multiuser Multiple-Antenna TDD
Systems,” IEEE Transactions on Vehicular Technology, vol. 60, no. 5,
p. 21022116, June 2011.
[8] H. Q. Ngo, E. G. Larsson, and T. L. Marzetta, “Analysis of the
Pilot Contamination Effect in Very Large Multicell Multiuser MIMO
Systems for Physical Channel Models,” IEEE International Conference
on Acoustics, Speech and Signal Processing (ICASSP), May 2011.
[9] B. Gopalakrishnan and N. Jindal, “An Analysis of Pilot Contamination
on Multi-user MIMO Cellular Systems with Many Antennas,” IEEE
International Workshop in Signal Processing Advances in Wireless
Communications, June 2011.
[10] H. Huh, G. Caire, H. C. Papadopoulos, and S. A. Ramprashad, “Achiev-
ing Massive MIMO Spectral Efficiency with a Not-so-Large Number of
Antennas,” IEEE Transactions on Wireless Communications, vol. 11,
no. 9, pp. 3226–3239, September 2011.
[11] J. Hoydis, S. T. Brink, and M. Debbah, “Massive MIMO: How Many
Antennas do we Need?,” 49th Annual Allerton Conference on Commu-
nication, Control, and Computing, December 2011.
[12] J. Hoydis and M. Debbah, “David versus Goliath: Small Cells versus
Massive MIMO,” tech. rep., Alcatel-Lucent.
[13] S. Yang and L. Hanzo, “Fifty Years of MIMO Detection: The Road
to Large-Scale MIMOs,” IEEE Communications Surveys and Tutorials,
vol. 17, no. 4, pp. 1941–1988, September 2015.
[14] E. Yaacoub and M. Al-Husseini, “Achieving Physical Layer Security
with Massive MIMO Beamforming,” The 11th European Conference on
Antennas and Propagation (EuCAP 2017), pp. 1762–1766, March 2017.
[15] R. Ma, Y. Gao, L. Cuthbert, and Q. Zeng, “Antipodal inearly tapered
slot antenna array for millimeter-wave base station in massive MIMO
systems,” IEEE Antennas and Propagation Society International Sym-
posium (APSURSI), July 2014.
[16] C. A. Balanis, Antenna Theory Analysis and Design, Third Edition. John
Wiley & Sons, Inc., 2005.
[17] “Definition of Phased Arrays.” Federal Standard 1037C. [Online],
Available: http://www.its.bldrdoc.gov/fs-1037/fs-1037c.html.
[18] Y. Gao, R. Ma, Y. Wang, Q. Zhang, and C. Parini, “Stacked Patch
Antenna With Dual-Polarization and Low Mutual Coupling for Massive
MIMO,” IEEE Transactions on Antennas and Propagation, vol. 64,
no. 10, pp. 4544–4549, October 2016.
[19] P. S. Taluja and B. L. Hughes, “Diversity Limits of Compact Broadband
Multi-Antenna Systems,” IEEE Journal on Selected Areas in Commu-
nications, vol. 31, no. 2, pp. 326–337, January 2013.
[20] Y. H. Nam, B. L. Ng, K. Sayana, Y. Li, J. Zhang, Y. Kim, and J. Lee,
“Full-Dimension MIMO (FD-MIMO) for Next Generation Cellular
Technology,” IEEE Communications Magazine, vol. 51, no. 6, pp. 172–
179, 2013.
[21] N. Doose and P. A. Hoeher, “Massive MIMO Ultra-Wideband Commu-
nications Using Multi-Mode Antennas,” International ITG Conference
on Systems, Communications and Coding, February 2015.
[22] Q. Zhang, Z. Chen, Y. Gao, C. Parini, and Z. Ying, “Miniaturized An-
tenna Array with Co2Z Hexaferrite Substrate for Massive MIMO,” IEEE
Antennas and Propagation Society International Symposium (APSURSI),
July 2014.
[23] B. K. Lau and Y. H. Leung, “A Dolph-Chebyshev Approach to the
Synthesis of Array Patterns for Uniform Circular Arrays,” IEEE Inter-
national Symposium on Circuits and Systems, May 2000.
[24] C. P. Domizioli, “Noise Analysis and Low-noise Design for Compact
Multi-Antenna Receivers: A Communication Theory Perspective,” Ph.D.
dissertation, North Carolina State University, Raleigh, 2009.
[25] C. L. Dolph, “A current Distribution for Broadside Arrays Which
Optimizes the Relationship Between Beam Width and Sidelobe Level,”
Proc. IRE, vol. 34, no. 6, pp. 335–348, 1946.
[26] M. Al-Husseini, E. Yaacoub, M. Baydoun, and H. Ghaziri, “Independent
Control of the Beamwidth and Sidelobe Level of Taylor One-parameter
Arrays,” The 38th Progress in Electromagnetics Symposium (PIERS
2017), St. Petersburg, Russia, May 2017.
[27] F. I. Tseng and D. K. Cheng, “Optimum Scannable Planar Arrays with
an Invariant Sidelobe Level,” Proc. IEEE, vol. 56, no. 11, pp. 1771–
1778, 1968.
[28] K. Y. Kabalan, A. El-Hajj, and M. Al-Husseini, “Bessel Planar Arrays,”
Radio Science, vol. 39, no. 1, pp. 1–9, February 2004.
[29] M. Al-Husseini, H. Ghaziri, E. Yaacoub, and K. Kabalan, “Rectangular
and Circular Arrays with Independently Controlled Beamwidth and
Sidelobe Level,” The 2017 IEEE International Symposium on Antennas
and Propagation (APS/URSI 2017), San Diego, CA, USA, July 2017.
[30] R. J. Mailloux, Phased Array Antenna Handbook,2nd edition. Artech
House, 2005.
[31] E. Yaacoub, M. Al Husseini, A. Chehab, A. El Hajj and K. Y. Kabalan,
“Hybrid Linear and Circular Antenna Arrays,” Iranian Journal of
Electrical and Computer Engineering, vol. 6, no. 1, pp. 48–54, 2007.
[32] Y. Zeng, R. Zhang, and Z. N. Chen, “Electromagnetic Lens-Focusing
Antenna Enabled Massive MIMO: Performance Improvement and Cost
Reduction,” IEEE Journal on Selected Areas in Communications,
vol. 32, no. 6, pp. 1194–1206, June 2014.
[33] T. Kwon, Y. G. Lim, B. W. Min, and C. B. Chae, “RF Lens-Embedded
Massive MIMO Systems: Fabrication Issues and Codebook Design,”
IEEE Transactions on Microwave Theory and Techniques, vol. 64, no. 7,
pp. 2256–2271, July 2016.
[34] C. Masouros and a. T. R. M. Sellathurai, “Large-Scale MIMO Trans-
mitters in Fixed Physical Spaces: The Effect of Transmit Correlation

and Mutual Coupling,” IEEE Transactions on Communications, vol. 61,
no. 7, pp. 2794–2804, July 2013.
[35] S. Shen, M. R. McKay, and R. D. Murch, “MIMO Systems with Mutual
Coupling: How Many Antennas to Pack into Fixed-Length Arrays?,” In-
ternational Symposium On Information Theory Its Applications, October
2010.
[36] X. Artiga, B. Devillers, and J. Perruisseau-Carrier, “Mutual Coupling Ef-
fects in Multi-User Massive MIMO Base Stations,” IEEE International
Symposium on Antennas and Propagation, July 2012.
[37] J. W. Wallace and M. A. Jensen, “Termination-Dependent Diversity
Performance of Coupled Antennas: Network Theory Analysis,” IEEE
Transactions on Antennas and Propagation, vol. 52, no. 1, pp. 98–105,
January 2004.
[38] J. W. Wallace and M. A. Jensen, “Mutual Coupling in MIMO Wireless
Systems: A Rigorous Network Theory Analysis,” IEEE Transactions on
Wireless Communications, vol. 3, no. 1, pp. 1317–1325, July 2004.
[39] M. J. Gans, “Channel Capacity Between Antenna Arrays Part I: Sky
Noise Dominates,” IEEE Transactions on Communications, vol. 54,
no. 9, pp. 1587–1592, September 2006.
[40] M. J. Gans, “Channel Capacity Between Antenna Arrays Part II:
Amplifier Noise Dominates,” IEEE Transactions on Communications,
vol. 54, no. 11, pp. 1983–1992, November 2006.
[41] B. K. Lau, J. B. Andersen, G. Kristensson, and A. F. Molisch, “Impact
of Matching Network on Bandwidth of Compact Antenna Arrays,” IEEE
Transactions on Antennas and Propagation, vol. 54, no. 11, pp. 3225–
3238, November 2006.
[42] C. P. Domizioli, B. L. Hughes, K. G. Gard, and G. Lazzi, “Receive
Diversity Revisited: Correlation, Coupling, and Noise,” IEEE Global
Communications Conference, December 2007.
[43] C. P. Domizioli, B. L. Hughes, K. G. Gard, and G. Lazzi, “Optimal
Front-End Design for MIMO Receivers,” IEEE Global Communications
Conference, December 2008.
[44] C. P. Domizioli, “Noise analysis and low-noise design for compact
multi-antenna receivers: A communication theory perspective,” Ph.D.
dissertation, North Carolina State University, Raleigh, NC, USA, 2009.
[45] Y. Liu, M. Tao, B. Lin, and H. Shen, “Hybrid Beamforming for Massive
MIMO A Survey,” arXiv preprint arXiv:1609.05078, 2016.
[46] V. Nurmela et al., “Deliverable D1.4: METIS Channel Models ,” tech.
rep., Metis, 2015.
[47] J. Medbo, H. Asplund, J.-E. Berg, and N. Jalden, “Directional Channel
Characteristics in Levation and Azimuth at an Urban Macrocell Base
Station,” EuCap, March 2012.
[48] H. Haas and S. McLaughlin, Next Generation Mobile Access Technolo-
gies: Implementing TDD. Cambridge University Press, 2007.
[49] L. Lu, G. Li, A. Swindlehurst, A. Ashikhmin, and R. Zhang, “An
Overview of Massive MIMO: Benefits and Challenges,” IEEE Journal
of Selected Topics in Signal Processing, vol. 8, no. 5, pp. 742–758,
October 2014.
[50] S. Kutty and D. Sen, “Beamforming for Millimeter Wave Communica-
tions: An Inclusive Survey,” IEEE Communications Surveys & Tutorials,
vol. 18, no. 2, pp. 949–973, February 2016.
[51] T. Rappaport, R. Heath, and R. Daniels, Millimeter Wave Wireless
Communications. Prentice-Hall, 2014.
[52] A. Kammoun, H. Khanfir, Z. Altman, M. Debbah, and M. Kamoun,
“Preliminary Results on 3D Channel Modeling: From Theory to Stan-
dardization,” IEEE Journal on Selected Areas in Communications,
vol. 32, no. 6, pp. 1219–1229, June 2014.
[53] J.-C. Chen, “Efficient Codebook-Based Beamforming Algorithm for
Millimeter-Wave Massive MIMO Systems,” IEEE Transactions on Ve-
hicular Technology, pp. 1–9, 2017.
[54] F. Hu, Opportunities in 5G Networks: A Research and Development
Perspective. CRC Press, 2016.
[55] T. Thomas, H. Nguyen, G. MacCartney, and T. Rappaport, “3D mmWave
Channel Model Proposal ,” IEEE Vehicular Technology Conference
(VTC-Fall), September 2014.
[56] J. He, T. Kim, H. Ghauch, K. Liu, and G. Wang, “Millimeter Wave
MIMO Channel Tracking Systems,” IEEE Globecom Workshops, De-
cember 2014.