MIMO: State of the Art, and the Future in Focus

Mboli Sechang Julius

Abstract-Antennas of transmitters and receivers have In between SISO and MIMO lies smart antennas
been manipulated to increase the capacity of transmission (Antenna diversity), which is either transmit diversity,
and reception of signals. Using many elements in antennas
to shape beams and direct nulls in a particular point for
multiple-Input, Single-Output (MISO), or receive
optimum signal transmission and reception has over diversity, Single-Input, multiple-Output (SIMO) [1, 3, 4].
decades, had tremendous positive influence in received
power and signal to noise ratio (SNR). However, since the MIMO systems can either be single-user or multiuser
antenna elements manipulation can be done both at base channels. If a transmitter uses n antennas while a receiver
station and device terminal, it gives rise to an important
method of using several antennas to put and obtain signals uses m antennas, there are then ways of
to and from space with increased capacity. This principle is conFigureuring the system [5]. When the emitters and
termed Multiple-input and Multiple-output (MIMO). sensors are respectively collocated on one transmitter and
This paper discusses application of MIMO in the state of
one receiver which is coordinated at both ends, the
the art and next generation of wireless systems (5G). It also
discusses four models of MIMO; SISO, SIMO, MISO and system is termed “Single User MIMO (SU MIMO)” or
MIMO, considering three method of combing the signals “point to point” [6]. In this system, the receiver tries to
from multipath propagations; Selection combining (SC), pick up the “n” input signal(s) simultaneously. A multi
Equal gain combing (EGC) and maximum ratio combining
(MRC). Spatial diversity and spatial multiplexing are also user MIMO (MU MIMO) consists of multiple access and
discussed as form of MIMO. Finally, Massive or Hyper broadcast channel.
MIMO which is a new method of increasing transmission
capacity by very large scale for fifth generation of wireless A multiple access channel is a multipoint-to-point
system is discussed with its challenges and opportunities. channel where multiple users or transmitters from
different geographical locations access the same source
Key terms-Diversity combining techniques, spatial
multiplexing, channel state information (CSI). Massive with very little or no coordination. A good example of
MIMO this system is the uplink channel in cellular networks. On
the other hand, MIMO broadcast channel is typically a
1. INTRODUCTION point-to-multipoint system where users or receivers
In radio communications, antennas play a critical role which are geographically separated receive distinct
and their capabilities have been significantly improved messages that are sent simultaneously by a lone
within the past decades [1]. A typical antenna for today’s transmitter which may have one or more coordinated
wireless communication is technically a lone radome that emitters.
usually contained multiple elements which can be
manipulated to steer nulls and beams of the antenna to
desired directions[1]. When an antenna is used to send or
input signals into space, it is technically referred to as
input device for the space and if it is used to receive
signal from space, it is referred to as output device. It is
this impression that gives rise to the term Multiple Input,
Multiple Output (MIMO) which is when multiple
antennas are used at both systems involved in wireless
communications.

Bell Labs patented MIMO in 1984 [2] and since then,

its conFigureuration has been extensively worked upon to
improve the reliability of transmission and reception.
MIMO is used on both downlink and uplink channels. For
instance, when base station (BS) and user equipment Figureure 1: SU MIMO
(UE) are both equipped with single antenna each, it is
conventionally referred to as single input, single output
(SISO). In this system, the well-known Shannon formula
gives the capacity as;

(1)
 Figure 1. SU MIMO if the MIMO system must be effective. There are mainly
three techniques of combing the signals; selection
combining (SC), equal gain combining (EGC) and
maximum ratio combing (MRC) [1, 3, 7].

3.1 Selection Combining (SC)

In SC, all the signals received are first sampled and the
largest is sent to demodulator for further process. This
method is very easy to implement but unfortunately is not
effective as it does not make simultaneous use of all the
signals. This is done by simply comparing the various
SNRs of the signals and eventually selecting the signal
Figureure 2: MU MIMO with highest SNR. SC is the cheapest method and does
not require any additional RF receiver chain though in
The downlink channel in cellular network is a good real life situation, the system only pick the strongest
example. The Shannon formula for calculating the signal as it may be difficult to measure the SNRs
channel capacity in MIMO system is individually [1, 7].

(2) 3.2 Maximum Ratio combining (MRC)

In MRC, the various individual signals from the

2. Spatial Diversity and Spatial Multiplexing different branches are co-phased and weighted with
Spatial diversity basically makes use of multiple respect to their individual SNRs and then summed to
antennas in wireless transmission where the antennas has obtain the output. MRC gives an average SNR output
similar characteristics and physically separated [1]. This which is the same as summing the average SNRs of the
is usually done when the incoming signal incidence is several branches if it is assumed that all the branches
considered, so that the effect of co-channel interference is have equal average SNR. The output SNR will be
mitigated. The use of cells and sectors in cellular acceptable at receiver since it probably gives the largest
networks like UMTS takes advantage of spatial diversity SNR. A typical result or sum from MRC will be like the
which helps in efficient use of the spectrum among black curve in Figure.6
multiple users. MISO and SIMO are essential transmit
and receive diversity respectively. Diversity only makes
the transmission more robust but does not increase the
data rate of the transmitted signals [1, 4].

Spatial multiplexing is a method that transmits data

streams or combined data for separate antennas to take
advantage of the available space dimension so that it is
reused. Spatial multiplexing MIMO systems have great
advantage over conventional MIMO systems in that
instead of single data transmission, data streams are Figure 3: Selection Combining (SC)
transmitted to make the transmission more robust as
opposed to ordinary spatial diversity, it is intended to
increase the data rate [1]. The concepts of spatial transmit
diversity and spatial multiplexing raises an interesting
question; how are these signals combined and
demodulated?

3. Combining Techniques

The existence of transmit diversity needs receive

diversity as well, it is even required that the number of
receive antennas equals or exceeds the transmit antennas
Figure 4: Maximum Ratio combining (MRC)
 obtain an SNR that is acceptable from the various SNRs
which may have been unacceptable. In comparison to
MRC, EGC is less complex and its performance is
marginally less effective than MRC.

4. STATE OF THE ART

It is now established that MIMO increases data rate,
makes transmission more robust, lowers the bit error rate,
increases coverage area and improves position estimation.
The downside of MIMO is mainly its complexity and the
limitation of meeting the ever increasing and challenging
demand of data transmission [2].

In the past, second generation technology like GSM

Figure 5: Maximum Ratio combining (MRC), Signals Addition used shared transmission channel that users were
allocated by dividing time and the available frequency,
that is frequency division multiple access and time
division multiple access (FDMA/TDMA) and third
generation system like the UMTS used code with time
and frequency, that is, code division multiple access
(CDMA) in two modes; either frequency division duplex
(FDD) or time division duplex (TDD) [8]. But now in
LTE-A (long term evolution advance), MIMO is used in
both uplink and downlink channels in different
conFigureurations [2, 9]. When 4 antennas on the
eNodeB transmit to 2 antennas in the UE, it is a 4×2
MIMO downlink channel, and if 4 antennas on the UE
transmit to 4 antennas on the eNodeB, then it is a 4×4
MIMO uplink channel.

For downlink, release 8 and 9 of LTE allows up to 4×4

MIMO while 2×2 MIMO is supported in the uplink [8].
Figure 6: Maximum Ratio combining (MRC) output sample.
This conFigureuration is modified to 8×8 MIMO for
downlink channel while uplink channel is increased to
4×4 MIMO. Downlink with 8 antennas for both eNodeB
and UE (needs more power and large size) is very
sophisticated and costly design. This implies that this
technology may be dropped in the near future [8].

If an 8×8 MIMO system implements 20 MHz carrier

as described above, its throughput will be about 600
Mbps [8]. With this throughput, if it is assumed that an
8×8 MIMO system deploys the aggregation of 5×20
MHz carriers in LTE-A, then maximum throughput will
be 600 Mbps × 5 carriers = 3 Gbps, but it’s only a
theoretical value though. MIMO system is used in two
methods in 3GPP (Third Generation Partnership Project);
Figure 7: Equal Gain combining (EGC)
SU-MIMO and MU-MIMO. Any cell can be in any of
3.3 Equal Gain combining (EGC) these modes dynamically switching but not in both
methods simultaneously [2, 8].
EGC works in similar manner as MRC but the
difference is that the weights of the individual SNRs from
the branches are all made unity. It is still possible to
 reports are required for a 4×4 MIMO system and
formations of 64 channels are needed for 8×8 MIMO
system that is supported in LTE-A. The use of codebook
which is simply a set of predefined matrices with each
standing for a kind of transmission parameters for a
particular channel, is limited in size. CSI problem is
much simpler to solve in TDD mode since it employs
same channel for both uplink and downlink channel [2, 8,
10].

Generally, it is a good idea to employ spatial

multiplexing, when good channel conditions exist, but
when the channel conditions are poor, then an alternative
method should be transmit diversity. This is true because
for successful performance, spatial multiplexing will
definitely need high quality radio channel but if the
channel is weak, then multiple antennas can be employed
Figure 8: Illustration of 4×2 downlink MIMO. [10] to proffer solution for high throughput [8].

An 8×8 SU-MIMO system is capable of delivering The uplink channel in LTE-A uses MIMO that is a bit
peak data rate for just one user but as mentioned earlier different from the downlink MIMO. For uplink channel,
this technology will likely not be used because it is first thing to consider is size of UE which usually should
required that the several layers are not correlated spatially be very small and the desire to make it even smaller and
and that both the eNodeB and UE make use of 8 flexible is ever on the increase. Release 11 of LTE
transmitting and receiving antennas respectively [8]. This support maximum number of 4 antennas due to their
technology will also mean less effectiveness since most small size in comparison to eNodeB and since they are
applications will not actually need high data rates or hand hell devices, techniques like beamforming in uplink
deliver high data rates. Therefore, MU-MIMO technology channels is impracticable [8].
will actually be a better option in this case. All LTE-A
enabled-UEs can be used to exploit MU-MIMO since 5. NEXT GENERATION MIMO
they are likely to be in different physical locations and so
As everyone is already talking of fifth generation
the MIMO layers are spatially separated, promising
technology (5G), there is also need for evolution in
higher throughput in a cell or sector than that of SU-
MIMO since it still suffers some challenges at this
MIMO [2, 8].
present generation. The new research area already in
The LTE downlink channel uses spatial multiplexing in progress for this 5G technology is termed as massive
two loops; either closed loop or open loop depending on MIMO. Simply put as the technology that will furnish
whether there is feedback from the UE or not [2, 8, 10]. It base stations with very high number of antennas
is a closed loop when the channel state information (CSI) (hundreds or even thousands) in comparison to
is fed back to the eNodeB and it is an open loop where conventional MIMO [8, 11, 12]. Massive MIMO may
there is no feedback. The transmitter would normally use also be called very large MIMO, ARGOS, full dimension
the CSI to recompense loses in the channel in such a way MIMO or hyper MIMO [11, 13].
that the receiver gets the best quality of service as much
5G networks are expected to support a very dense
as possible. However, this leads to two issues; the
system with large numbers of devices that will support
feedback loop can take up high bandwidth thereby
machine-to-machine and machine-to-person
affecting its efficiency and CSI delivery delay is also a
communications [5, 14]. Therefore, the frequencies range
major problem when CSI changes rapidly.
for massive MIMO is expected to go from
The challenge of CSI delivery delay is solved by using centimetre/millimetre range, from 6 GHz to about 100
codebooks [8] in LTE-A, but the problem of bandwidth GHz to support these ever craving data consumers. The
efficiency remains a major challenge since radio resource traffic will likely increase tremendously. One billion new
is limited. For instance, in 2×2 MIMO system, up to 4 mobile subscribers expected to join the existing 3.6
reports channels are expected to be delivered while 16 billion to make it 4.6 billions by 2020 [15], and to cope
with this increase, Massive MIMO might need to proffer
up to 100 Mbit/s data speed and peak data speed of about Since the performance of wireless communication
10 Gbit/s [14]. systems largely depends and in fact, is governed by the
wireless channel environment, new and hot research
areas is springing up in in MIMO or topics connected to
it. All new technologies being developed daily from the
several applications of Internet of Things (IoT) and Big
Data depend either directly or indirectly on high and
reliable data rate which can be possible via Massive
MIMO as already discussed. The analysis of wireless
communication systems is always difficult due to the fact
that the wireless channels employed is always
unpredictable and dynamic. This and many challenges
are opening up research opportunities in wireless
communications.
Figure 9: Sample Base station with many antennas [16].

Major advantages include:

6. CONCLUSION
i. Increase in capacity by a factor of 10 or more.
ii. Very high increase in data rate due to large This paper gave a general overview of MIMO as a
number of antennas leading to high transmits system that started with just SISO and went all the way to
diversity. MIMO. It related how transmission all started with just a
iii. Low and cheap power components can also be single line and progresses onto multiple channels. The
used to build Massive MIMO so that its output desire for more efficiency and quality of service in the
power lies in the mili-watt range. communication industries led us to MIMO and now
iv. Efficiency will largely be increased due to the Massive MIMO is on the way to even make the system
fact that beams and nulls from base station will more suitable for 5G networks. This is possible because
be steered to where they are needed, thereby generations of networks has been evolving every decade,
avoiding wastage and interference [17]. so it is normal to expect 5G in 2020.
v. Latency can be greatly reduced in the open air
interface because fading dips can be avoided REFERENCES
with beamforming.
[1] "A quick introduction to MIMO technology," in Microwaves,
The main aim of the coding techniques in massive Communications, Antennas and Electronics Systems, 2009.
MIMO is to be able to purely separate the distinct users’ COMCAS 2009. IEEE International Conference on, 2009, pp. 1-23.
[2] C. Cox, 2012. An introduction to LTE : LTE, LTE-advanced, SAE,
data streams at the receiver by employing optimal and 4G mobile communications / Christopher Cox. [Online
maximum likelihood detector, decision feedback eBook]. Available: [Accessed:
detectors and linear detectors[16]. [3] S. Alamouti, "A simple transmit diversity technique for wireless
communications," Selected Areas in Communications, IEEE
Journal on, vol. 16, no. 8, pp. 1451-1458, 1998.
However, with so much anticipated benefits of massive [4] A. Ismail, S. Sezginer, J. Fiorina, and H. Sari, "A Simple and Robust
MIMO, there are limitations and research challenges. Equal-Power Transmit Diversity Scheme," Communications
The limitations include pilot contamination, channel Letters, IEEE, vol. 15, no. 1, pp. 37-39, 2011.
[5] T. L. Narasimhan, P. Raviteja, and A. Chockalingam, "Large-scale
reciprocity, non-channel state information (CSI) at the multiuser SM-MIMO versus massive MIMO," in Information
transmitter operation and physical separation between Theory and Applications Workshop (ITA), 2014, 2014, pp. 1-9.
base stations and users. If UEs with line of sight (LOS) [6] A. J. Paulraj, D. A. Gore, R. U. Nabar, and H. Bolcskei, "An
overview of MIMO communications - a key to gigabit wireless,"
are close to the base station, interference will occur due to Proceedings of the IEEE, vol. 92, no. 2, pp. 198-218, 2004.
high correlation. The number of antennas will also be [7] M. L. Pablo-Gonzalez, M. Sanchez-Fernandez, and E. Rajo-
Iglesias, "Combination of the Three Types of Diversity to Design
limited at certain level where additional antennas will not
High-Capacity Compact MIMO Terminals," Antennas and
make any change again [5, 11]. Wireless Propagation Letters, IEEE, vol. 13, no. pp. 1309-1312,
2014.
[8] J. Korhonen, Introduction to 4G mobile communications. Boston: [14] B. ALASTAIR. 5G mobile network features [Online]. Available:
Artech House, 2014. http://www.unwiredinsight.com/2014/5g-mobile-network-
[9] S. Sesia, I. Toufik, and M. Baker, LTE - the UMTS long term features [Accessed: 15/12/2014 ].
evolution : from theory to practice, 2nd ed. Chichester: Wiley, [15] GSMA. GSMA Press Release. [Online]. Available:
2011. http://www.gsma.com/newsroom/press-release/one-billion-
[10] B. ALASTAIR. LTE MIMO theory and practice [Online]. Available: new-unique-mobile-subscribers-by-2020-gsma/ [Accessed:
http://www.unwiredinsight.com/2013/lte-mimo [Accessed: 18/3/2015].
21/03/2015]. [16] Practical many-Antennas Base Station [Online]. Available:
[11] E. Larsson, O. Edfors, F. Tufvesson, and T. Marzetta, "Massive http://argos.rice.edu/ [Accessed: 17/3/2015].
MIMO for next generation wireless systems," Communications [17] K. T. Truong and R. W. Heath, "Effects of channel aging in massive
Magazine, IEEE, vol. 52, no. 2, pp. 186-195, 2014. MIMO systems," Communications and Networks, Journal of, vol.
[12] X. Gao, O. Edfors, F. Rusek, and F. Tufvesson, "Massive MIMO in 15, no. 4, pp. 338-351, 2013.
real propagation environments." arXiv preprint arXiv:1403.3376,
2014.
[13] A. L. Anderson and M. A. Jensen, "A Generalized Sum-Rate
Optimizer for Cooperative Multiuser Massive MIMO Link
Topologies," Access, IEEE, vol. 2, no. pp. 1040-1050, 2014.