ENERGY EFFICIENCY OF MIMO SYSTEM

A PROJECT REPORT

Submitted to

in the partial fulfillment of the requirements for

the award of the degree of
in the Department of
Electronics and Communication Engineering [ECE]
by
B Prasanth Lara-16311A04C4
N Tarun-16311A04F1
N Surya-16311A04F2
Project co-ordinator Under the Guidance of
DR .D Ajitha DR SPV Subba Rao
Associate Proffessor Assistant Professor
Department of ECE Department of ECE

DEPARTMENT OF ELECTRONICS & COMMUNICATION ENGINEERING

SREENIDHI INSTITUTE OF SCIENCE & TECHNOLOGY

(AUTONOMOUS)
Yamnampet, Ghatkesar, R.R District, Hyderabad – 501301

1
 Date: 10/03/2020

CERTIFICATE

This is to certify that the Project Work entitled

“ENERGY EFFICIENCY OF MIMO SYSTEM” being submitted by

B Prasanth Lara-16311A04C4

N Tarun-16311A04F1

N Surya-16311A04F2

in fulfillment of B.Tech 4th year 2nd nd semester in Department of Electronics and

Communication Engineering [ECE], Sreenidhi Institute of Science and Technology, an
autonomous institute under Jawaharlal Nehru Technological University, Telangana, is a record
of bonafide work carried out by them under our guidance and supervision.

Project Coordinator- Project Guide-

Dr D Ajitha Dr. S.P.V. SUBBA RAO
Assistant Professor Head & Professor
DEPARTMENT OF ECE DEPARTMENT OF ECE

Head of the department

Dr SPV SUBBA RAO
Professor and HOD
ECE Dept

ACKNOWLEDGEMENT

2
We would like to thank my supervisor Dr D Ajitha, Associate professor, Department of ECE for
giving us constant guidance, support and motivation throughout the period this course work was
carried out. Their readiness for consultation at all times, their educative comments and assistance
even with practical things have been invaluable. We are thankful that she had gave the freedom
to do the work with our ideas.

We express my sincere gratitude to our project guide DR. S.P.V.SUBBA RAO, Head of
Department, ECE for helping us in carrying out this project giving support throughout the period
of our study in SNIST.

We are also thankful to our principle DR. CH. SIVA REDDY, giving us this guidance and
support, motivation throughout the period of our B.Tech course work carried out.

We convey our special thanks to honorable Executive Director DR. P.N.REDDY SNIST for his
continuous support in maintaining pleasant and friendly atmosphere in campus for studies and
course work.

We are also thankful to all teaching and non-teaching staff of our department who has rendered
their co-operation in completion of this project report.

We also thank our parents and friends and well-wishers who aided us in completion of the
project report.

DECLARATION

3
We hereby declare that the work described in this report, entitled “ENERGY EFFICIENCY
OF MIMO SYSTEM”, which is being submitted by us in fulfillment of B.Tech 4th year, 2nd
semester in Department of Electronics and Communication Engineering [ECE], Sreenidhi
Institute Of Science & Technology affiliated to Jawaharlal Nehru Technological University
Hyderabad, Kukatpally, Hyderabad (Telangana) -500 085 is the result of investigations carried
out by us under the Guidance of Dr SPV SUBBA RAO Professor, Department of ECE,
Sreenidhi Institute Of Science And Technology, Hyderabad. The work is original and has not
been submitted for any Degree/Diploma of this or any other university.

Place: Hyderabad
Date:11/03/2020

Signature

B Prasanth Lara-16311A04C4
N Tarun-16311A04F1
N Surya-16311A04F2

ABSTRACT

4
5G provides users with a higher system capacity, low latency, and low system complexity. The
core technology, massive multiple-input and multiple-output (MIMO), improves spectral
efficiency and degrees of freedom and lower system complexity. In my project, we will analyze
the relationship between signal-to-noise ratio (SNR) and capacity. In the second part, we will
discuss energy efficiency. The consumed power of the system includes transmit power, circuit
power consumption, and idle power consumption. The circuit power consumption dominates the
system power consumption when the transmitter is equipped with massive number of antennas.
Hence, to analyze this problem, we propose a power consumption model. To solve this problem,
it is important to find the trade-off between system capacity and energy efficiency.
Keyword: Massive MIMO, Spectral efficiency, Energy Efficiency

CONTENTS- Page Nos

Acknowledgement

5
Abstract

Chapter 1

Introduction

Chapter 2

History of MIMO systems

Chapter 3

MIMO System Model

Chapter 4

Energy efficiency of MIMO system

Chapter 5

Advantages and Disadvantages

Chapter 6

Applications

Chapter 7

Results

Chapter 8

Matlab Code

6
Chapter 9

Conclusions

Chapter 10

Future Prospects

Chapter 11

References

Abbreviations:

3GPP: 3 rd Generation Partnership Project

5G: 5 th Generation of Wireless technology

7
C-RAN: Cloud Radio Access Network

eMBB: Enhanced Mobile Broadband

GSM: Global System for Mobile Communications

GSMA: GSM Association

HLF: High Level Forum

IMT: International Mobile Telecommunications

IoT: Internet of Things

LSA: Licensed Shared Access

LMLC: Low Mobility Large Cell

LTE: Long Term Evolution

M2M: Machine to Machine

MIMO: Multiple Input Multiple Output

mMTC: Massive Machine Type Communication

NDCP: National Digital Communication Policy

NFV: Network Function Virtualization

NR: New Radio

NSA: Non-Stand Alone OFDMA: Orthogonal Frequency Division Multiple

QoS: Quality of Services

RAN: Radio Access Network

SA: Stand Alone

SDN: Software Defined Networking

TRAI: Telecom Regulatory Authority of India

URLLC: Ultra Reliable and Low Latency Communication

8
 Chapter 1

Introduction-

5G is the fifth generation of wireless communication technology that supports cellular data
networks. Big-scale adoption began in 2019 and today almost all telecommunications service
providers in developed countries are upgrading their infrastructure to provide 5G performance.

The frequency spectrum of 5G is divided into millimeters, mid-band, and low-band. Low-band
uses the same frequency range as its predecessor, 4G. The 5G millimeter wave is the fastest, with
a real speed that is usually down to 1-2 Gbit / s. Frequency above 24 GHz reaches 72 GHz which
is above the threshold of the highest speed band. Access is shorter, so more cells are needed. The
millimeter waves have difficulty in pulling many walls and windows, so the surface coverage is
limited.

The 5G mid-band is the largest, with more than 30 networks. The band speed at 100 MHz band
is usually 100-400 Mbit / s down. In the lab and occasionally in the field, the speed can exceed
gigabit per second. The frequencies used range from 2.4 GHz to 4.2 GHz. Sprint and China
China use 2.5 GHz, while others are between 3.3 and 4.2 GHz. Many areas can be covered by
simply upgrading existing towers, which reduces costs. Mid-band networks are more efficient,
bringing costs closer to 4G costs.

The low-band 5G offers the same capacity for the advanced 4G. In the United States, T-Mobile
and AT&T launched low-cost services in the first week of December 2019. T-Mobile CTO
Neville Ray warns that the speed at their 600 MHz 5G is below 25 Mbit / s down. AT&T, using
850 MHz, will usually deliver less than 100 Mbit / s by 2019. Performance will improve, but
there can be no greater than 4G power in the same spectrum.

Verizon, AT&T, and almost all 5G providers in 2019 have a range between 25- 35 milliseconds.
The "air latency" (between the phone and the tower) in 2019 actions is 8 - 12 ms. The latency on
the server, back in the network, raises the rate to ~ 30 ms, 25- 40% lower than normal 4G.
Adding "Edge Servers" near the towers can reduce latency in 10 - 20 ms. The low latency, which
is similar to the usual 1 tout, is old and does not take time off the server.

9
Speed

5G speed will start from ~ 50Mbit / s to over 2Gbit / s initially. [14] The fastest 5G, known as
mmWave, delivers speeds up to 2Gbit / s. As of July 3, 2019, mmWave had a maximum speed of
1.8Gbit / s on AT & T's 5G network.

Sub-6 GHz 5G (mid-band 5G), on average, will deliver between 100 and 400 Mbit / s, but will
reach far beyond mmWave, especially outside.

The low-band display provides remote placement but is slower than others.

The speed of 5G NR on sub-6 GHz bands can increase slightly more than 4G with the same
amount of spectrum and antennas, even though other 3GPP 5G networks will be slower than
other developed 4G networks, such as T-Mobile's LTE network / LAA, reaches 500+ Mbit / s in
Manhattan [18] and Chicago. [19] The 5G specification permits LAA (Licensed Access Use),
but the LAA on 5G has not yet been shown. Adding LAA to an existing 4G plan can add
hundreds of megabits per second with speed, but this is a 4G expansion, not a new 5G segment.
[18]

The similarity in terms of the similarities between 4G and 5G in existing bands is because 4G is
already close to the Shannon limit with data connection rates. The average 5G apparent
millimeter wave speed, with its very wide bandwidth and short distance, and consequently
frequency reuse, can be very high.

Latency

For 5G, the "air latency" on mission deployment in 2019 is 8-12 milliseconds. The latency on the
server must be added to "air latency". Verizon reports that latency in its original 5G usage by 30
ms Edge Servers near the towers can reduce the latency to 10 - 20 ms. 1 - 4 ms will rarely exceed
years outside the lab.

What is the difference between 4G and 5G?

A: There are several differences between 4G vs 5G:

5G may be a unified platform that's more capable than 4G

5G uses spectrum better than 4G

10
5G is faster than 4G

5G has more capacity than 4G

5G has lower latency than 4G

5G may be a unified platform that's more capable than 4G

While 4G LTE focused on delivering much faster mobile broadband services than 3G, 5G is
meant to be a unified, more capable platform which will not only elevate mobile broadband
experiences, but also support new services like mission-critical communications and therefore
the massive IoT. 5G also will natively support all spectrum types (licensed, shared, unlicensed)
and bands (low, mid, high), a good range of deployment models (from traditional macro-cells to
hotspots), as well as new ways to interconnect (such as device-to-device and multi-hop mesh).

Chapter 2

History of MIMO Systems

On the radio, multi-input and multi-output, or MIMO is a way to increase the volume of the
radio connection using multiple transmitting and receiving horns to exploit the frequency of
transmission. MIMO has become an integral part of wireless communication standards including
IEEE 802.11n (Wi-Fi), IEEE 802.11ac (Wi-Fi), HSPA + (3G), WMAMAX (4G), and Long term
Evolution (4G LTE). Recently, MIMO has been used for 3-wire electrical wiring connections as
part of the standard ITU G.hn specification with HomePlug AV2.

At one point, in wireless the word "MIMO" was referring to the use of multiple horns in the
transmitter and receiver. In today's application, "MIMO" specifically refers to the process used to
send and receive more than one data ad at the same radio station by exploiting multipath

11
propagation. MIMO is distinctly based on smart antenna techniques developed for the
performance of a single data signal, such as brightness and diversity.

MIMES often follow back to the 1970s research papers on transmission systems that use
multiple digital channels and interventions (crosstalk) at midnight with a wire bundle: AR Kaye
and DA George (1970), Branderburg and Wyner (1974) , [ and W. van Etten (1975, 1976).
Although these are not examples of exploiting multipath propagation to send multiple sources of
information, other mathematical techniques for dealing with visual disturbances are not useful
for the development of MIMO. In the mid-1980s Jack Salz at Bell Laboratories re-conducted the
study, investigating a number of services that worked through "integrated networks with
additional audio sources" such as frequency communication and radio programs designed for
them.

Mechanisms were developed to improve the effectiveness of mobile radio networks and to
enable more aggressive frequency use in the early 1990s. Multiple Space-division multiple
access (SDMA) uses discrete or intelligent antennas to communicate at the same frequency as
users at different locations within the same channel range. It was proposed by the SDMA
program by Richard Roy and Björn Ottersten, researchers at ArrayComm, in 19911. Their US
Patent (No. 5515378 issued in 1996 ) describes how to increase capacity through the "horn
detection list at the base station" overflowing of users from afar. ”Types of antennas

Multi-antenna MIMO technology (or single user MIMO) has been produced and operates at
certain levels, e.g., 802.11n products.

• SISO / SIMO / MISO for special MIMO cases

o Multiple Input and Single Output (MISO) is a special case where the receiver has one cable.

o Input and dual output (SIMO) is a special case when a transmitter has one antenna.

o Single-input single output (SISO) is a standard radio program where no transmitter or receiver
has multiple antennae.

• MIMO techniques do not use a single user

o Bell Laboratories layered Space-Time (BLAST), Gerard. J. Foschini (1996)

o Per Antenna Rate Control (PARC), Varanasi, Guess (1998), Chung, Huang, Lozano (2001)

12
o Choosing Per Antenna Rate Control (SPARC), Nokia (2004)

Some limitations

o Body antenna space is chosen to be large; multiple waveforms at the base station. The antenna
division of the finder has limited space in the offices, although the antenna discussion and
algorithm are still being developed.

Multiple user types

Recently, research results on MIMO technology for many users have been emerging. While
many MIMO users (or the MIMO network) may have high potential, in reality, the (small)
research of most MIMO users (or most users with the MIMO antenna) is very effective.

• MASSIVE Multi-MIMO User (MU-MIMO)

o At the latest 3GPP and WiMAX standards, MU-MIMO is treated as one of the widely accepted
candidate candidate technologies set by many companies, including Samsung, Intel, Qualcomm,
Nokia, TI, Huawei, Philips, Nokia

13
 Chapter 3

MIMO SYSTEM MODEL-

SISO systems are generally more complex than most multi-input (MIMO) systems. It's often,
and it's very easy to order a larger or more predictable future "flight" or "back in an envelope".
MIMO systems have so many connections that most of us quickly, completely, and efficiently
get into our heads.

Typical domain strategies for the analysis and design of control structure theory for the SISO
control system. Bode structure, Nyquist stiffness, Nichols ice, and the locus are common tools
for SISO system analysis. Controls can be programmed using a polynomial, a root method to
name just two popular ones. Usually the SISO controllers will be either PI, PID, or lead-lag.

MIMO can be divided into three main categories: precoding, spatial multiplexing (SM), and
coding variants.

Precoding involves multiple distributions, in small details. In general terms, it is assumed that all
processing takes place at the destination. In the (single-stream) variations, the same signal is
released from each transmitter of the antenna at the appropriate frequency and gain weight in
such a way that the signal strength is increased at the receiver input. The benefits of
beamforming are to maximize the signal gain - by making the signals emitted from various
antennas adding constructively - as well as reducing the effect of fainting. In the line-of-sight
distribution, the synchronization of results shows a well-defined descriptive mechanism.

14
However, standard beams are not a good comparison for mobile networks, which are
characterized mainly by multipath propagation. When the receiver has multiple antennas, the
transmission of the transmitter cannot simultaneously increase the signal intensity with each
receiving signal, and placing an email with multiple streams is often advantageous. Note that
preoding requires channel channel information (CSI) information for the transmitter and
receiver.

Air multiplication requires MIMO antenna configuration. In spatial frequency, a high-frequency

signal is separated by very low currents and each stream is transmitted from another unique
antenna of a normal channel. When these signals come to the receiver of an antenna that
combines enough local signals and the receiver with an accurate CSI, they can split these signals
into parallel (almost) channels. Spatial multiplication is a very powerful way to increase channel
capacity with high signal-to-noise ratios (SNR). The maximum number of local sources is
limited below the number of antimicrobials in the transmitter or receiver. Local replication can
be used without CSI in the transmitter, but can be combined with precoding when CSI is
available. Spatial multiplication can also be used for simultaneous transmission to multiple
recipients, known as multiple space-division or multiple-user, where CSI is required for the
transmitter. The arrangement of receivers with different spatial coordinates allows for better
separation.

Various encoding techniques are used when there is no channel information on the transmitter.
By means of variations, a single stream (unlike many streams in spatial extension) is transmitted,
but the signal is composed of so-called space-time coding techniques. The signal is extracted
from each transmission beam by full or near orthogonal coding. Coding variants exploit the
independence of multiple antenna interfaces to improve signal variability. Because there is no
channel information, no variables or collections are received from the various codes. Differential
encoding can be combined with adding a geographical space where specific channel information
is available to the recipient.

15
This thesis will study energy efficiency and spatial efficiency in massive MIMO based on three
pre-coding algorithms including MRC, MMSE and ZF in mathematical and simulation
approches. A simplified massive MIMO system is built in this system which equips with
adjustable number of transmitted antenna and 3 users with signal received antenna. In particular,
a computational efficient resource allocation will be designed for the maximization of energy
efficiency of the massive MIMO system. Besides, the impact of the amount of transmitted

16
antenna on capacity will be assessed under different pre-coding algorithms. Based on this,
energy efficiency will be then discussed. Finally, optimal and suboptimal transmission scheme
will be presented in order to maximize the usage of materials and resources.

For wireless communication, the regional CSI specification (CSI) refers to known areas of the
communication channel. This information describes how the signal is transmitted from one
translator to the receiver and represents the combined effect of, for example, scattering,
dissipation and energy decay with distance. The method is called channel measurement. CSI
makes it possible to synchronize transmissions in current channel conditions, which is essential
for achieving reliable communication with high data rates in multantenna systems.
CSI needs to be limited to the receiver and the frequency is calculated along with the response to
the transmitter (although recalculation rates may occur in TDD systems). Therefore, the trans
person and the recipient may have different CSI. The CSI in the transmission and the CSI in the
receiving area are sometimes referred to as CSIT and CSIR, respectively.
There are basically two levels of CSI
1. Quick CSI
2. Statistics CSI.
Accelerated CSI (or transient CSI) means that current channel conditions are known, which can
be viewed as being able to respond to negative digital filters. This provides an opportunity to
synchronize the transmitted signal in response to an impact and thus increase the received signal
for spatial repetition or achieve very low values.

17
Statistical CSI (or long-term CSI) means that channel observations are known. This description
may include, for example, the nature of the end-distribution, the gain of a normal channel, the
component of the peripheral, and the relation of the surface. As with CSI immediately, this
information can be used for transfer purposes.
CSI adoption is limited to how quickly channel conditions change. In fast-ending systems where
channel conditions vary rapidly under the transmission of a single information signal, only the
statistical CSI makes sense. On the other hand, in programs of gradual termination the CSI can
be estimated with valid accuracy and applied to the synchronization distribution for some time
before it expires.
In multi-user MIMO, a transmitter of multiple antennas communicates simultaneously with
multiple receivers (each with multiple or multiple antennas). This is known as SDMA multi-
space access. From an implementation standpoint, the mododing algorithms for SDMA systems
can be divided into nonlinear and nonlinear types. Linear access algorithms are not linear, [7] but
straight line input methods meet reasonable performance with very low complexity. Linear input
strategies include high frequency transmission (MRT), zero-forcing (ZF) transfer mode, and
Wiener specification There are also data input techniques that are designed for low frequency
status information response, for example random placement. The nonlinear interpretation is
designed based on the concept of paper coding (DPC), which indicates that any known
interference in the transmission can be eliminated without the benefit of radio sources if a high
quality preoding system can be used in the transmission signal.

18
While performance enhancements have a clear meaning in point-to-point MIMO, the multi-user
system can't maximize their performance at the same time. This can be considered a problem of
accessibility for different purposes where each objective is related to increasing the capacity of
one of the users. [3] The most common way to simplify this problem is to select a program
function; for example, the amount of weight at which the weight corresponds to the raw material
used first by the users. In addition, there may be more users than the data stream, requiring a
programming algorithm to determine which users will be active at a given time.

19
Linear precoding with full channel state information

This sub-method cannot reach the weight limit, but it can increase the weighted average
performance (or some other metric of the achievable metric below the solid line). High-line
queue has no display form closed, but it takes the MMSE signal-resolution status of a single
antenna. Therefore, foreoding can be interpreted as finding the perfect balance between gaining a
strong signal gain and reducing the inconvenience of some users.

Finding single-weighted MMSEs is easy, leading to precise methods where the instruments are
selected optimally. The most common method is to focus on the number or number of the
specified rate; that is, the average transmission rate (MRT) and zero-forcing (ZF) precoding.
MRT only increases the signal gain for the target user. MRT is very close to limited systems,
where user interference is ignored when compared to noise. ZF encryption aims to reduce user
interference, with the aim of losing some signal gain. The ZF specification can achieve
performance near a threshold value when the number of users is large or the system is interrupted
(e.g., the noise is not strong compared to the interrupt). The balance between MRT and ZF is
achieved by the so-called zero-termination [ (also known as the sign-to-leakage-and-interference
ratio (SLNR). ]

And in the multuser MIMO system setup, another method has been used to convert the average
weight access problem into a weighted MSE value problem by adding additional MSE weight
per symbol to. However, this function cannot solve this problem completely (i.e., its solution is
more). On the other hand, the method of presentation of the mind is also considered and found a
more equitable solution to use the weight ratio.

Note that linear linear regression can be calculated using monotonic optimization algorithms, but
the magnitude of the complexity of the algorithms comes quickly with the number of users.
These algorithms are therefore only useful for benchmarking in small programs.

In practice, channel status information is limited to the transmitter due to measurement and
amplification errors. Incorrect channel information can result in significant loss of system
coverage, as interference between extended streams can be completely controlled. In closed
systems, response capabilities determine which input strategies are feasible. Each recipient may

20
report any version of the total channel information or focus on specific indicators of significant
activity (e.g. channel profit).

When complete channel information is returned with good accuracy, one can use techniques
designed to have complete channel information with minimal performance degradation. Zero
synchronization can achieve even the maximum gain of a multiplier, but only if the accuracy of
the channel response increases directly with the signal-to-noise (in dB) range. Quantization and
feedback of channel position information is based on vector number, and codebooks based on
Grassmanni line placement have shown good performance.

Other clarification methods have been developed for that case with very low channel response
rates. Randomized factor analysis was proposed as an easy way to achieve optimum performance
on a scale such as a numerical model where the number of acceptors is large. In this sub-strategy,
a set of decorative directions is chosen at random and users consume a few pieces to tell the
passenger which beam offers the best performance and what level they can expect to use.

Linear Precoding Schemes –

In massive MIMO systems, when the amount of transmitted antennas approaches infinity, the
system can be simplified as a Single-input-to-Single-output (SISO) systems Therefore, to
optimize spectral resources in massive MIMO systems, pre-coding is used at the transmit side in
order to reduce the complexity of system, diminish noise effect and optimize stream data
transmission based on channel state information (CSI) There are three common linear pre-coding
schemes, including MRC, ZF and MMSE.

(1) Maximum-ratio Transmission

This scheme is to maximize SNR by seeking to maximize the power at the receiver combiner.
MRC is considered as a viable linear reception scheme for massive MIMO systems since it can
be applied in a distributed manner. The mathematical model for MRC is shown as below

𝐴 = 𝒈𝒌‖𝒈𝒌‖

21
Moreover, MRC has a satisfactory performance in the low-power regime, even approaching to
optimal performance as the amount of antennas grows infinitely. However, as the power
increase, systems based on MRC scheme suffer from serious inter-user interference [1].
(2) Zero-forcing

The ZF scheme is to eliminate inter-user interference by projecting the received signals into the
orthogonal elements. It can be written as
𝑨= 𝑮(𝑮𝐻𝑮)−1
Since ZF scheme does not take noise into consideration, system based on ZF precoding scheme
has a poor performance in low power regime. The performance in high-power regime approaches
to optimal

(3) MMSE

MMSE scheme seeks to eliminate inter-user interference as well as noise. Compared to MRC
and ZF, system complexity of MMSE is relatively higher.
𝑨 =𝑮(𝑮𝐻𝑮 + 1𝑃𝑢𝐼𝑘)−1

From mathematical perspective, MRC has the lowest complexity among these three precoding
schemes. MMSE requires perfect channel state information.

22
 Chapter 4

Energy Efficiency of MIMO System-

The energy efficiency (EE) is defined as the ratio of capacity versus transmitted power
consumption, presenting the number of bits transmitted per Joule In massive MIMO system,
when allocating transmitted power equally to antennas, the spatial efficiency will be improved by
M and √𝑀 for perfect CSI and imperfect CSI system, respectively (where M represents the
number of transmitted antennas).
Moreover, with the increase amount of antennas, noise and small scale fading will decrease
correspondingly; correlation among channels will be reduced by expending distance among
antennas. According to statistics, channels between antennas andusers will approach to
orthogonal when the number of antennas is overwhelmingly large than that of users Traditional
MIMO systems pay more attention on transmitted power consumption instead of EE

Power Consumption Model

Since a large number of antennas are involved in, circuit power consumption which is mainly
yielded by radio frequency (RF) chain cannot be neglected . In communication system, each
antenna is equipped with one RF chain. RF chain in downlink consists of mixer, filter, digital-to-
analog converter (DAC) as well as synchronizers

Received signal 𝑥 is firstly fed into band-selective filter, power amplifier, synchronization, low-
pass filer and auto gain control module (AGC). Then, an A/D converter is adapted to converts
analogue signal to digital signal by quantization.

In this case, more antennas contribute to more circuit power consumption. As a result, massive
MIMO comprises energy efficiency for spatial efficiency. Additionally, system loss associated to

23
hardware for RF seriously threats on system performance. Hardware loss includes quantization
error, phase error, phase shift of carrier frequency and sampling frequency, nonlinear power
amplifier.

Based on analysis above, we can get the mathematical model for total power consumption,
𝑃𝑘,𝑡𝑜𝑡𝑎𝑙=𝑃𝑡𝜂(1−𝜎𝑓𝑒𝑒𝑑)+𝑃𝑐𝑖𝑟+𝑃𝑠𝑡𝑎(1−𝜎𝐷𝐶)(1−𝜎𝑀𝑠)(1−𝜎𝑐𝑜𝑜𝑙) (
Where
𝑃𝑘,𝑡𝑜𝑡𝑎𝑙 is kth use’s total power consumption.
𝑃𝑐𝑖𝑟 is circuit power consumption which can be calculated by 𝑃𝑐𝑖𝑟=𝑁(𝑃𝑑𝑎𝑐+𝑃𝑚𝑖𝑥+𝑃𝑓𝑖𝑙𝑡)
+𝑃𝑠𝑦𝑛,
𝑃𝑑𝑎𝑐 is DAC power consumption.
𝑃𝑚𝑖𝑥 is mixer power consumption. 𝑃𝑓𝑖𝑙𝑡 is filter power consumption.
N is activated transmitted antennas.
𝑃𝑠𝑡𝑎 is idle power consumption.
𝜎𝐷𝐶,𝜎𝑓𝑒𝑒𝑑,𝜎𝑐𝑜𝑜𝑙,𝜎𝑀𝑠 are the loss factors of antenna DC-DC power supply, antenna feeder,
active cooling system and main power supply respectively
In this project, we ignore loss factors
Therefore, EE can be given in mathematical way as below

𝐸𝐸= Σ𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦Σ𝑃𝑖,𝑡𝑜𝑡𝑎𝑙

In conclusion, a large amount of antennas in massive MIMO systems does improve spatial
efficiency and capacity but lead to extra power consumption based on analysis. Therefore, it is
important to find a balanced trade-off among different aspects of system performance.

Energy efficiency is determined by precoding scheme, the number of antennas as well as

transmitted power consumption. We mainly focus on how the transmitted power influences on
energy efficiency

24
 CHAPTER 5

Advantages and Disadvantages-

Following are the benefits or advantages of MIMO:

1. The higher data rate can be achieved with the help of multiple antennas and SM (Spatial
Multiplexing) technique. This helps in achieving higher downlink and uplink throughput.
It helps in achieving reduction in BER (Bit Error Rate) due to application of advanced
signal processing algorithms on the received data symbols by multiple antennas.

2. The techniques such as STBC (Space Time Block Coding) and BF (Beamforming) when
employed in MIMO system helps in achieving extension of cell coverage.

3. MIMO based system minimize fading effects seen by the information traveling from
transmit to receive end. This is due to various diversity techniques such as time,
frequency and space.
There is lower susceptibility of tapping by unauthorized persons due to multiple antennas
and algorithms.

4. The systems with MIMO offers high QoS (Quality of Service) with increased spectral
efficiency and data rates.

5. The wide coverage supported by MIMO system helps in supporting large number of
subscribers per cell.

6. The MIMO based system is widely adopted in latest wireless standards viz. WLAN

(802.11n, 802.11ac etc.), WiMAX (IEEE 802.16e), LTE, LTE-Advanced etc .

25
Disadvantages-

1. Following are the disadvantages of MIMO:

The resource requirements and hardware complexity is higher compare to single antenna
antenna based system. Each antenna requires individual RF units for radio signal
processing. Moreover advanced DSP chip is needed to run advanced mathematical signal
processing algorithms.

2. The hardware resources increase power requirements. Battery gets drain faster due to
processing of complex and computationally intensive signal processing algorithms. This
reduces battery lifetime of MIMO based devices.

3. MIMO based systems cost higher compare to single antenna based system due to
increased hardware and advanced software requirements.

26
 Chapter 6

Applications-

Automobiles
5G Automotive Association have been promoting the C-V2X communication technology
that will first be deployed in 4G. It provides for communication between vehicles and
communication between vehicles and infrastructures.

Public safety
Mission-critical push-to-talk (MCPTT) and mission-critical video and data are expected to
be furthered in 5G.

Fixed wireless
Fixed wireless connections will offer an alternate to fixed line broadband (ADSL, VDSL,
Fiber optic, and DOCSIS connections) in some locations.
Wireless video transmission for broadcast applications
Sony has tested the possibility of using local 5G networks to replace the SDI cables
currently used in broadcast camcorders

• 5G will make unified global standard for all.

• Network availability will be everywhere and will facilitate people to use their computer
and such kind of mobile devices anywhere anytime.
• Because of the IPv6 technology, visiting care of mobile IP address will be assigned as per
the connected network and geographical position.
27
• Its application will make world real Wi Fi zone.
• Its cognitive radio technology will facilitate different version of radio technologies to
share the same spectrum efficiently.
• Its application will facilitate people to avail radio signal at higher altitude as well

Chapter 7
28
Results-

29
30
31
32
Combining theoretical knowledge, Dinkelbach Method and numerical results, if we do not
consider system complexity, MMSE has the best performance in terms of spectral efficiency and
energy efficiency (perfect CSI is assumed). Compared to ZF, MRC performs better in low
power regime (with SNR ranging from 0dB to 15dB). On the contrary, the performance of ZF in
high power regime (with SNR larger 10dB) is better and even approaches to that of MMSE.
Moreover, the simulation results demonstrate that massive MIMO systems cannot enhance
spectral efficiency and energy efficiency infinitely due to the existence of inference among
users.

33
 Chapter 8
Matlab Code-
clc; 
clear all;
close all;
%..............................................................
% Initiation
%..............................................................
no_of_data_bits = 64%Number of bits per channel extended to 128
M =4 %Number of subcarrier channel
n=256;%Total number of bits to be transmitted at the transmitter
block_size = 16; %Size of each OFDM block to add cyclic prefix
cp_len = floor(0.1 * block_size); %Length of the cyclic prefix
%............................................................
% Transmitter
%.........................................................
%.........................................................
% Source generation and modulation
%........................................................
% Generate random data source to be transmitted of length 64
data = randsrc(1, no_of_data_bits, 0:M-1);
figure(1),stem(data); grid on; xlabel('Data Points'); ylabel('Amplitude')
title('Original Data ')
% Perform QPSK modulation on the input source data
qpsk_modulated_data = pskmod(data, M);
figure(2),stem(qpsk_modulated_data);title('QPSK Modulation ')
%............................................................
%.............................................................
% Converting the series data stream into four parallel data stream to form
% four sub carriers
S2P = reshape(qpsk_modulated_data, no_of_data_bits/M,M)
Sub_carrier1 = S2P(:,1)
Sub_carrier2 = S2P(:,2)
Sub_carrier3 = S2P(:,3)
Sub_carrier4 = S2P(:,4)
figure(3), subplot(4,1,1),stem(Sub_carrier1),title('Subcarrier1'),grid on;
subplot(4,1,2),stem(Sub_carrier2),title('Subcarrier2'),grid on;
subplot(4,1,3),stem(Sub_carrier3),title('Subcarrier3'),grid on;
subplot(4,1,4),stem(Sub_carrier4),title('Subcarrier4'),grid on;
%..................................................................
%..................................................................
% IFFT OF FOUR SUB_CARRIERS
%.................................................................
%..............................................................
number_of_subcarriers=4;
cp_start=block_size-cp_len;
34
ifft_Subcarrier1 = ifft(Sub_carrier1)
ifft_Subcarrier2 = ifft(Sub_carrier2)
ifft_Subcarrier3 = ifft(Sub_carrier3)
ifft_Subcarrier4 = ifft(Sub_carrier4)
figure(4), subplot(4,1,1),plot(real(ifft_Subcarrier1),'r'),
title('IFFT on all the sub-carriers')
subplot(4,1,2),plot(real(ifft_Subcarrier2),'c')
subplot(4,1,3),plot(real(ifft_Subcarrier3),'b')
subplot(4,1,4),plot(real(ifft_Subcarrier4),'g')
%...........................................................
%...........................................................
% ADD-CYCLIC PREFIX %..........................................................
%............................................................
for i=1:number_of_subcarriers,
ifft_Subcarrier(:,i) = ifft((S2P(:,i)),16)% 16 is the ifft point
for j=1:cp_len,
cyclic_prefix(j,i) = ifft_Subcarrier(j+cp_start,i)
end
Append_prefix(:,i) = vertcat( cyclic_prefix(:,i), ifft_Subcarrier(:,i))
% Appends prefix to each subcarriers
end
A1=Append_prefix(:,1);
A2=Append_prefix(:,2);
A3=Append_prefix(:,3);
A4=Append_prefix(:,4);
figure(5), subplot(4,1,1),plot(real(A1),'r'),title('Cyclic prefix added to all the sub-carriers')
subplot(4,1,2),plot(real(A2),'c')
subplot(4,1,3),plot(real(A3),'b')
subplot(4,1,4),plot(real(A4),'g')
figure(11),plot((real(A1)),'r'),title('Orthogonality'),hold on ,plot((real(A2)),'c'),hold on ,
plot((real(A3)),'b'),hold on ,plot((real(A4)),'g'),hold on ,grid on 
%Convert to serial stream for transmission
[rows_Append_prefix cols_Append_prefix]=size(Append_prefix)
len_ofdm_data = rows_Append_prefix*cols_Append_prefix
% OFDM signal to be transmitted
ofdm_signal = reshape(Append_prefix, 1, len_ofdm_data);
figure(6),plot(real(ofdm_signal)); xlabel('Time'); ylabel('Amplitude');
title('OFDM Signal');grid on;
%...............................................................
Passing time domain data through channel and AWGN

channel = randn(1,2) + sqrt(-1)*randn(1,2);

after_channel = filter(channel, 1, ofdm_signal);
awgn_noise = awgn(zeros(1,length(after_channel)),0);
recvd_signal = awgn_noise+after_channel; % With AWGN noise

35
figure(7),plot(real(recvd_signal)),xlabel('Time'); ylabel('Amplitude');
title('OFDM Signal after passing through channel');grid on;
recvd_signal_paralleled = reshape(recvd_signal,rows_Append_prefix, cols_Append_prefix);
%........................................................
%........................................................
% Remove cyclic Prefix
%.......................................................
%......................................................
recvd_signal_paralleled(1:cp_len,:)=[];
R1=recvd_signal_paralleled(:,1);
R2=recvd_signal_paralleled(:,2);
R3=recvd_signal_paralleled(:,3);
R4=recvd_signal_paralleled(:,4);
figure(8),plot((imag(R1)),'r'),subplot(4,1,1),plot(real(R1),'r'),
title('Cyclic prefix removed from the four sub-carriers')
subplot(4,1,2),plot(real(R2),'c')
subplot(4,1,3),plot(real(R3),'b')
subplot(4,1,4),plot(real(R4),'g')
%...................................................
%...................................................
% FFT Of recievied signal
for i=1:number_of_subcarriers,
% FFT
fft_data(:,i) = fft(recvd_signal_paralleled(:,i),16);
end
F1=fft_data(:,1);
F2=fft_data(:,2);
F3=fft_data(:,3);
F4=fft_data(:,4);
figure(9), subplot(4,1,1),plot(real(F1),'r'),title('FFT of all the four sub-carriers')
subplot(4,1,2),plot(real(F2),'c')
subplot(4,1,3),plot(real(F3),'b')
subplot(4,1,4),plot(real(F4),'g')
%................................
%..............................
% Signal Reconstructed
%..................................
%..................................
% Conversion to serial and demodulationa
recvd_serial_data = reshape(fft_data, 1,(16*4));
qpsk_demodulated_data = pskdemod(recvd_serial_data,4);
figure(10)
stem(data)
hold on
stem(qpsk_demodulated_data,'rx');

36
grid on;xlabel('Data Points');ylabel('Amplitude');
title('Recieved Signal with error')
clear
N = 10^6; % number of bits or symbols
Eb_N0_dB = [0:25]; % multiple Eb/N0 values
nTx = 2;
nRx = 2;
for ii = 1:length(Eb_N0_dB)

% Transmitter
ip = rand(1,N)>0.5; % generating 0,1 with equal probability
s = 2*ip-1; % BPSK modulation 0 -> -1; 1 -> 0

sMod = kron(s,ones(nRx,1)); %
sMod = reshape(sMod,[nRx,nTx,N/nTx]); % grouping in [nRx,nTx,N/NTx ] matrix

h = 1/sqrt(2)*[randn(nRx,nTx,N/nTx) + j*randn(nRx,nTx,N/nTx)]; % Rayleigh channel

n = 1/sqrt(2)*[randn(nRx,N/nTx) + j*randn(nRx,N/nTx)]; % white gaussian noise, 0dB
variance

% Channel and noise Noise addition

y = squeeze(sum(h.*sMod,2)) + 10^(-Eb_N0_dB(ii)/20)*n;

% Receiver

% Forming the MMSE equalization matrix W = inv(H^H*H+sigma^2*I)*H^H

% H^H*H is of dimension [nTx x nTx]. In this case [2 x 2]
% Inverse of a [2x2] matrix [a b; c d] = 1/(ad-bc)[d -b;-c a]
hCof = zeros(2,2,N/nTx) ;
hCof(1,1,:) = sum(h(:,2,:).*conj(h(:,2,:)),1) + 10^(-Eb_N0_dB(ii)/10); % d term
hCof(2,2,:) = sum(h(:,1,:).*conj(h(:,1,:)),1) + 10^(-Eb_N0_dB(ii)/10); % a term
hCof(2,1,:) = -sum(h(:,2,:).*conj(h(:,1,:)),1); % c term
hCof(1,2,:) = -sum(h(:,1,:).*conj(h(:,2,:)),1); % b term
hDen = ((hCof(1,1,:).*hCof(2,2,:)) - (hCof(1,2,:).*hCof(2,1,:))); % ad-bc term
hDen = reshape(kron(reshape(hDen,1,N/nTx),ones(2,2)),2,2,N/nTx); % formatting for
division
hInv = hCof./hDen; % inv(H^H*H)

hMod = reshape(conj(h),nRx,N); % H^H operation

yMod = kron(y,ones(1,2)); % formatting the received symbol for equalization

yMod = sum(hMod.*yMod,1); % H^H * y
yMod = kron(reshape(yMod,2,N/nTx),ones(1,2)); % formatting
yHat = sum(reshape(hInv,2,N).*yMod,1); % inv(H^H*H)*H^H*y

% receiver - hard decision decoding

37
 ipHat = real(yHat)>0;

% counting the errors

nErr(ii) = size(find([ip- ipHat]),2);

end

simBer = nErr/N; % simulated ber

EbN0Lin = 10.^(Eb_N0_dB/10);
theoryBer_nRx1 = 0.5.*(1-1*(1+1./EbN0Lin).^(-0.5));
p = 1/2 - 1/2*(1+1./EbN0Lin).^(-1/2);
theoryBerMRC_nRx2 = p.^2.*(1+2*(1-p));

close all
figure
semilogy(Eb_N0_dB,theoryBer_nRx1,'bp-','LineWidth',2);
hold on
semilogy(Eb_N0_dB,theoryBerMRC_nRx2,'kd-','LineWidth',2);
semilogy(Eb_N0_dB,simBer,'mo-','LineWidth',2);
axis([0 25 10^-5 0.5])
grid on
legend('theory (nTx,nRx, ZF)', 'theory (nTx,nRx, MRC)', 'sim (nTx, nRx, MMSE)');
xlabel('Average Eb/No,dB');
ylabel('Bit Error Rate');
title(“Energy Efficiency for MIMO system with MMSE equalizer (Rayleigh channel)');

clear
N = 10^6; % number of bits or symbols
Eb_N0_dB = [0:25]; % multiple Eb/N0 values
nTx = 2;
nRx = 2;
for ii = 1:length(Eb_N0_dB)

% Transmitter
ip = rand(1,N)>0.5; % generating 0,1 with equal probability
s = 2*ip-1; % BPSK modulation 0 -> -1; 1 -> 0

sMod = kron(s,ones(nRx,1)); %
sMod = reshape(sMod,[nRx,nTx,N/nTx]); % grouping in [nRx,nTx,N/NTx ] matrix

h = 1/sqrt(2)*[randn(nRx,nTx,N/nTx) + j*randn(nRx,nTx,N/nTx)]; % Rayleigh channel

38
 n = 1/sqrt(2)*[randn(nRx,N/nTx) + j*randn(nRx,N/nTx)]; % white gaussian noise, 0dB
variance

% Channel and noise Noise addition

y = squeeze(sum(h.*sMod,2)) + 10^(-Eb_N0_dB(ii)/20)*n;

% Receiver

% Forming the ZF equalization matrix W = inv(H^H*H)*H^H

% H^H*H is of dimension [nTx x nTx]. In this case [2 x 2]
% Inverse of a [2x2] matrix [a b; c d] = 1/(ad-bc)[d -b;-c a]
hCof = zeros(2,2,N/nTx) ;
hCof(1,1,:) = sum(h(:,2,:).*conj(h(:,2,:)),1) ; % d term
hCof(2,2,:) = sum(h(:,1,:).*conj(h(:,1,:)),1) ; % a term
hCof(2,1,:) = -sum(h(:,2,:).*conj(h(:,1,:)),1); % c term
hCof(1,2,:) = -sum(h(:,1,:).*conj(h(:,2,:)),1); % b term
hDen = ((hCof(1,1,:).*hCof(2,2,:)) - (hCof(1,2,:).*hCof(2,1,:))); % ad-bc term
hDen = reshape(kron(reshape(hDen,1,N/nTx),ones(2,2)),2,2,N/nTx); % formatting for
division
hInv = hCof./hDen; % inv(H^H*H)

hMod = reshape(conj(h),nRx,N); % H^H operation

yMod = kron(y,ones(1,2)); % formatting the received symbol for equalization

yMod = sum(hMod.*yMod,1); % H^H * y
yMod = kron(reshape(yMod,2,N/nTx),ones(1,2)); % formatting
yHat = sum(reshape(hInv,2,N).*yMod,1); % inv(H^H*H)*H^H*y

% receiver - hard decision decoding on second spatial dimension

ipHat2SS = real(yHat(2:2:end))>0;
ipHatMod2SS = 2*ipHat2SS-1;
ipHatMod2SS = kron(ipHatMod2SS,ones(nRx,1));
ipHatMod2SS = reshape(ipHatMod2SS,[nRx,1,N/nTx]);

% new received symbol - removing the effect from second spatial dimension
h2SS = h(:,2,:); % channel in the second spatial dimension
r = y - squeeze(h2SS.*ipHatMod2SS);

% maximal ratio combining - for symbol in the first spatial dimension

h1SS = squeeze(h(:,1,:));
yHat1SS = sum(conj(h1SS).*r,1)./sum(h1SS.*conj(h1SS),1);
yHat(1:2:end) = yHat1SS;

% receiver - hard decision decoding

ipHat = real(yHat)>0;

% counting the errors

39
 nErr(ii) = size(find([ip- ipHat]),2);

end

simBer = nErr/N; % simulated ber

EbN0Lin = 10.^(Eb_N0_dB/10);
theoryBer_nRx1 = 0.5.*(1-1*(1+1./EbN0Lin).^(-0.5));
p = 1/2 - 1/2*(1+1./EbN0Lin).^(-1/2);
theoryBerMRC_nRx2 = p.^2.*(1+2*(1-p));

close all
figure
semilogy(Eb_N0_dB,theoryBer_nRx1,'bp-','LineWidth',2);
hold on
semilogy(Eb_N0_dB,theoryBerMRC_nRx2,'kd-','LineWidth',2);
semilogy(Eb_N0_dB,simBer,'mo-','LineWidth',2);
axis([0 25 10^-5 0.5])
grid on
legend('theory (nTx=1,nRx=2, MRC)', 'sim (nTx=2, nRx=2, ZF-SIC)');
xlabel('Average Eb/No,dB');
ylabel('Bit Error Rate');
title('Energy Efficiency for MIMO and ZF-SIC equalizer (Rayleigh channel)');

40
 Chapter 9

Conclusions-

To optimize system performance and maximize economic benefit, it is significant to find out the
optimal power allocation.

Since the appearance of Massive MIMO, the system performance of wireless communication
system has been improved significantly in terms of capacity, latency, reliability and e.t.
In this thesis,
(1) we firstly analyze the history of wireless communication system.

(2) Introduce the importance of massive MIMO, system model, mathematic model and problem
formulation.

(3) And to make the problem formulation solvable, we apply Dinkelbach Method to address this
optimization problem with CVX.

(4) Finally, combining numerical results, we present that massive MIMO remarkably improves
the energy efficiency and spectral efficiency.

We mainly focus on energy efficiency in massive MIMO systems. Although we have studied
power allocation in uplink transmission, there are still some aspects that should be done in the
future.

41
 Chapter 10

Future Prospects-

 It is observed that multiple antenna systems with huge amount of antenna elements at base
station are competent to increase data rate by many folds, without requirement of any extra
bandwidth, as compared to other existing technologies. Massive-MIMO combined with multiple
carrier systems (Massive-MIMO-OFDM) followed by suitable signal detection schemes, like
beam forming, gives overwhelming results. Application: With possibilities of further research
and continuous improvements, Massive-MIMO system is one of the best suitable choices, among
various technologies, for next generation wireless communication systems, like 5G.

The energy efficiency model in this thesis only suits to single cell massive MIMO system. In the
future, it can be developed to multi-cells Massive MIMO system. We consider that both transmit
and receive sides have perfect CSI. It is expected to develop to imperfect CSI. Distances between
users and BS is not be taken into account in this thesis. This should be considered in the future
for more accurate simulation results.

42
 Chapter 11

References-
1. Raleigh, Gregory; Cioffi, John M. (1996). Spatio-temporal coding for wireless
communications (PDF). Global Telecommunications Conference, 1996. London, UK
November 18–22, 1996.
2.  Rakhesh Singh Kshetrimayum  (2017). Fundamentals of MIMO Wireless
Communications. Cambridge University Press.
3. Telatar, Emre (1999). "Capacity of Multi-antenna Gaussian Channels". European
Transactions on Telecommunications. 10 (6): 585–95. 
4. Rosas, F. & Oberli, C. (April 16, 2013). "Nakagami-m approximations for multiple-input
multiple-output singular value decomposition transmissions". IET
Communications. 7 (6): 554–561. doi:10.1049/iet-com.2012.0400.
5. David Tse; Pramod Viswanath (2005). Fundamentals of Wireless
Communication. Cambridge.
6. Claude Oestges; Bruno Clerckx (2007). MIMO Wireless Communications: From Real-
world Propagation to Space-time Code Design. Academic Press.
7. Ezio Biglieri; Robert Calderbank; Anthony Constantinides; Andrea Goldsmith;
Arogyaswami Paulraj; H. Vincent Poor (2010). MIMO Wireless
Communications. Cambridge University Press.
8. S. Cui; A. J. Goldsmith & A. Bahai (August 2004). "Energy-efficiency of MIMO and
Cooperative MIMO in Sensor Networks". IEEE Journal on Selected Areas in
Communications. 22 (6): 1089–1098. 
9. B. Kumbhani, R S Kshetrimayum (2017). MIMO Wireless Communications over
Generalized Fading Channels. CRC Press.
10.  T. E. Bogale and L. B. Le, Pilot optimization and channel estimation for multiuser
massive MIMO systems in Proc. IEEE Conference on Information Sciences and Systems
(CISS), Princeton, USA, Mar. 2014.

43
44