International Journal of Advanced Research in Engineering and Technology (IJARET)

Volume 11, Issue 11, November 2020, pp. 1469-1483, Article ID: IJARET_11_11_134
Available online at http://www.iaeme.com/IJARET/issues.asp?JType=IJARET&VType=11&IType=11
ISSN Print: 0976-6480 and ISSN Online: 0976-6499
DOI: 10.34218/IJARET.11.11.2020.134

© IAEME Publication Scopus Indexed

EVALUATION OF PAPR REDUCTION ABILITY

OF 5G INSPIRED WAVEFORM CANDIDATES
Pathuri Lavanya
Research Scholar, Department of ECE, Koneru Lakshmaiah Education Foundation,
Vijayawada, Andhra Pradesh, India.

Dr. Penke Satyanarayana

Department of Electronics and Computer, Koneru Lakshmaiah Education Foundation,
Vijayawada, Andhra Pradesh.

Dr. Mohammad Mohatram

Department of Electrical and Electronics, Global College of Engineering and Technology,
Muscat, Oman.

ABSTRACT
The multifold increase in multimedia technology triggers an insatiable thirst for
high data rates and increased demand for Orthogonal Frequency Division
Multiplexing based wireless systems, a superior technology for the high-speed
wireless communication systems. OFDM technique possesses multiple benefits
superior to its predecessor, namely, its high capacity and resistance to multipath
fading channels. Nonetheless, few hurdles remain yet unresolved in OFDM
systems. The foremost issue stands out to be the high Peak to Average Power Ratio,
which results in non-linearity in the High-Power Amplifiers and drives In-Band
Distortion and Out-Of-Band Emission. Research plans are currently under
development to standardize, given the economic 5G boom for market
commercialization. The rise of superior waveform competitors possessing higher
spectrum throughput, lower Peak to Average Power Ratio having a more
straightforward system design than the celebrated OFDM. Potential candidates
include numerous single carrier and multi-carrier waveforms. This paper undertakes
an assessment, highlights the strengths and weaknesses of various 5G waveform
candidates under the Key Performance Indicators, Simulation is carried out to study
the PAPR reduction capabilities of these waveforms.

http://www.iaeme.com/IJARET/index.asp 1469 [email protected]

 Pathuri Lavanya, Dr. Penke Satyanarayana, Dr. Mohammad Mohatram

Keywords: 5G, FBMC, GFDM, KPI, UFMC, F-OFDM Multi-carrier, PAPR

reduction.
Cite this Article: Pathuri Lavanya , Dr. Penke Satyanarayana, Dr. Mohammad
Mohatram, Evaluation of Papr Reduction Ability of 5G Inspired Waveform
Candidates, International Journal of Advanced Research in Engineering and
Technology 11(11), 2020, pp. 1469-1483.
http://www.iaeme.com/IJARET/issues.asp?JType=IJARET&VType=11&IType=11

1.INTRODUCTION
Orthogonal frequency division multiplexing (OFDM) is the best-adapted technology far and
wide for fourth generation (4G) communications [1]. Current standards, namely Long-Term
Evolution (LTE) and its next-generation LTE-Advanced (LTE-A) adopt OFDM [2]. Even
though exciting research is underway to enhance channel performance of OFDM with
minimal inter-carrier interference (ICI), better spectrum utilization, and minimal inter symbol
interference (ISI) using a cyclic prefix (CP), it exhibits some intrinsic drawbacks. Firstly,
leakage in frequency occurs due to the rectangular nature of pulse. Secondly, loss in the
spectrum occurs due to CP add-on to eradicate fading in a multipath environment. Thirdly,
stringent frequency and time synchronization fail to preserve the orthogonality in sub-carriers
even though It helps preserve interferences within and amongst cells. Fourthly, very high peak
to average power ratio (PAPR). Fifthly, high out-of-band radiations (OOB), as OFDM
deploys square pulses for the baseband. Lastly, low-grade performance and high bit error rate
(BER) increases in proportion to modulation levels. 5th Generation (5G) mobile
communication standard is on its way to being commercially operational in a few years and is
expected to bring about a substantial revolution in various market realms such as scientific,
business, industry, public, to mention a few. It would provide enhanced energy efficiency,
throughput, data coverage, and latency compared with the present 4G structure. The
prevailing 4G standard is devoid of capable infrastructure that would enable it to be forward
compatible with the forthcoming 5G network. One of the most critical concerns for
implementing 5G communication is selecting the best waveform that is best suited for that
particular application. In order to improvise OOB radiation and PAPR reduction further,
academicians and researchers have proposed and developed several single carrier and multi-
carrier waveforms in recent years to suit specific physical layers requirements.
This survey paper's prime objective is to exhibit the brief principle of operation,
advantages, and limitations of these waveforms. Further, the waveforms are compared
amongst themselves on the criteria of Key Performance Indicators (KPI) factors such as
PAPR and BER. Research is presented in a structured approach in this paper. The Section 2
exhibits how single carrier and multicarrier waveforms candidates are categorized following
an organized explanation of the waveforms. Section 3 describes research summary of the
studies conducted over the years. Section 4 describes experimental setup. Section 5 is solely
devoted to the comparative PAPR reduction performance in simulation. Finally, section 6
presents brief conclusion and also brings light into possible future directions

2. WAVEFORM CANDIDATES
Waveforms can be categorized based on how the spectrum gets appropriated utilized, namely
single-carrier (SC) waveforms and multi-carrier (MCM) waveforms. The comprehensive
research conducted provides insight into how these waveforms have been developed and
improvised over ten years to suit 5G standards. Figure 1 illustrates a generalized framework
that is stationed to generate the waveforms. It follows a sequential procedure. The desired
waveforms are generated by selecting the desired design parameters and activating the

http://www.iaeme.com/IJARET/index.asp 1470 [email protected]

 Evaluation of Papr Reduction Ability of 5G Inspired Waveform Candidates

necessary stages. Table 1 explain the necessary parameters and the operational block to be
initiated to generate the desired waveform candidates. The method at which waveforms
candidates is appropriately classified is illustrated in Figure 2.

Figure 1. Generalized waveform synthesis framework to generate any 5G waveform candidate.

Desired waveforms may be generated by activating necessary operations.

Table 1 Generalized Design Factors and Design Parameters to Generate 5G Waveform Candidates
Waveform Design Factor Design Process
Single Carrier Waveforms
DFT-s-OFDM Subcarrier Spacing, DFT blocks, Size of FFT, I, III, VI. VII
Length of CP
ZT-DFT-s-OFDM Sub-carrier Spacing, DFT blocks, Size of FFT and I, II, III, IV,V
CP, Zero-Tail samples
UW-DFT-s-OFDM CPM, Subcarrier Spacing, DFT blocks, Size of FFT I, (II), III, IV, V,
CP Length VI
Multi Carrier Waveforms
CP-OFDM Subcarrier Spacing , Size of FFT, Length of CP I, IV, V
W-OFDM Subcarrier Spacing, Length of CP, Size of FFT, I, IV,V, VI
Length of Window
UW-OFDM Subcarrier Spacing, Unique Word II, IV, V
UF-OFDM Subcarrier Spacing, Sub-band filter, Guard band, I, II, IV
Size of FFT CP, zero-tailing
FBMC-OQAM Subcarrier Spacing, prototype filters with I,IV
percentage of overlap
GFDM Subcarrier Spacing, prototype filters with I, IV
percentage of overlap, Size of FFT

2.1 Multicarrier Waveforms

The waveforms discussed in this article are classified into two categories, those that are
supported by the previous OFDM transceivers and the waveforms that require new
transceivers to be installed for operation. The former category is further grouped based on the
signal processing methods involved such as Sub-band filtering and full-band filtering. The
latter category falls under Sub-carrier filtering.

http://www.iaeme.com/IJARET/index.asp 1471 [email protected]

 Pathuri Lavanya, Dr. Penke Satyanarayana, Dr. Mohammad Mohatram

Figure 2 . Generalized 5G waveform classification of single carrier waveforms and multi-carrier

waveforms.
2.1.1 Sub-Carrier Filtering
These techniques are intended to lessen OOB radiation in MCM waveforms. Pair of
transceiver filters are deployed for every single sub-carrier. Inter Symbol Interface (ISI) is
eradicated by introducing cyclic-prefix (CP) before the signal that is selected for transmission.
However, due to the presence of long filter lengths, waveforms in this category are not
capable to support services requiring low latency. E.g. Filter-bank multi-carrier (FBMC) and
Generalized frequency division multiplexing (GFDM).
(a) Filter bank multi-carrier-OQAM: FBMC waveform was proposed in the year 1960s [3].
Using this waveform, individually each of the designated subcarrier are filtered with very
narrow band filter and long filter length, that helps makes it immune to OOB emissions. The
following steps I and IV are enabled from the Figure 1. In FBMC, data symbols, and
subcarriers for each of the contained symbols are considered. represents a real data
symbol that consists of both real and imaginary components symbolized by and
respectively.

Real part of the subcarrier modules, is shifted in phase of by the upper IFFT,
and imaginary parts is phase rotated by a value of by the lower IFFT. A real
symmetric filter prototype of response is applied to each subcarrier resulting from
times iterative lower and upper IFFT vector output copied and repeated. Principly in FBMC,
first a prototype filter is incorporated by frequency shifting. Second, modulated filters called
filter bank as a group, helps transmit the set of data symbols in parallel. Such filters lowers
out-of-band (OOB) radiation and inter symbol interference (ISI).On the other side, it results in
frequency spreading due to absence of CP and complex computational structure.
(b) Generalized frequency division multiplexing: Principle of generation of GFDM
waveforms [4] is defined as follows, complex data symbols are modulated over
subcarriers, resulting in complex natured baseband data vector. It is followed by serial to

http://www.iaeme.com/IJARET/index.asp 1472 [email protected]

 Evaluation of Papr Reduction Ability of 5G Inspired Waveform Candidates

parallel (S/P) conversion of the data symbols and division into independent streams, each
stream consisting of sub-symbols. sub symbol in each stream, is
oversampled be a factor . Then processed by a pulse shaping filter circularly
convoluted in time of length that shifts samples of sub symbol. Finally, modulated
with subcarriers, summed and affixed with cyclic prefix, giving rise to the resultant
output of GFDM waveform

Here denotes the base-band data information, supported by the sub-carrier

. is prototype filter of vector that circularly
shifts in time domain.
2.1.2 Sub-Carrier Filtering
These waveform categories are designed for reduced filter length. Here, filters are developed
considering only a sub-band wherein each sub-band consists of many subcarriers [12]. Sub-
band filter length can be shortened since it is quite longer and designed to work for multiple
sub-carriers. The sub-band filtering candidates are designed to overcome the glitches of
subcarrier filtering. They find themselves being deployed to serve applications in 5G network
for low latency. E.g. UFMC, UF-OFDM and F-OFDM.

(a) Filtered-OFDM (F-OFDM): The working principle of F-OFDM is that it divides the data
bandwidth into sub-bands with permitted filter length extensions beyond the length of CP [5].
Filter is supposed to be deliver the most uniform waveform output while operating in the
pass-band with bands of tight transition. Most popular filter used is the windowed-sinc filter
having ideal impulse response

The term represents the chosen window. F-OFDM’s filter length is very sensitive and
vital because filter length that are larger than fundamental length results in degraded
frequency spectrum while shorter filter length growths both spectrum utilization, OOB
radiation and computational complexity. Overlapping in time domain is vital for preserving
bandwidth utilization. Benefits include less computations, flexibility to cohabitate with other
waveforms in asynchronous fashion.
(b) Universal Filtered Multi-Carrier: In UFMC, the system is developed as follows. The
input data gets split by allocated vector of sub-carriers into sub-bands [6]. The data
symbols in every single sub-band is IDFT processed and serially filtered. Data vector is
multiplied and modulated with a pre-coding matrix signified as . Data vector
is split into sub-bands. Each sub band comprises of sub
carriers such that . -point (IDFTs) are processed for data in each sub-band. The
filter length is an central factor in determining the characteristics of the UFMC waveform,

http://www.iaeme.com/IJARET/index.asp 1473 [email protected]

 Pathuri Lavanya, Dr. Penke Satyanarayana, Dr. Mohammad Mohatram

such as PAPR, OoB emissions and inclusive performance. One such filter is the Dolph
Chebyshev filter with length . Filters with long length results in huge overhead reducing
the system capacity. Shorter filter lengths effects performance, owing to multipath nature of
fading channel. Cumulative of the filtered signal output of all sub-bands are summed that
results in the generated UFMC signal to be transmitted. The UFMC waveform can be
represented

(c) Universal Filtered OFDM (UF-OFDM): UF-OFDM efficiently suppresses the OOB
radiation by its principle of separating all subcarriers in a number of subbands and filtering
each sub-band independently. Because of which UF-OFDM thrives high efficiency in
spectrum utilization, are immune to OOB radiation. The input data using the subcarriers
gets slotted into subbands [7], every subband having subcarriers. UF-OFDM
transmission signal regarding the data symbol is

Here and . signifies output data symbol of the inverse discrete

Fourier transform (IDFT) and signifies time response of filter length. represents a
complex part of symbol pertaining to the subcarrier. denotes the sub-band
matrix output of IDFT. denotes the linear filter convolution Toeplitz matrix, in the
subband. Length of effective guard interval (GI) of the UF-OFDM is expressed as
and represents half of the filter transition time. However, performance
quality declines exponentially in long delayed multipath fading environment due to longer
time response longer and narrow filter’s main lobe.

2.1.3 Full-band Filtering

The full-band filters are designed to be applicable for the entire signal bandwidth [7]. One
transceiver filter pair is sufficient for the entire stipulated bandwidth Eg. F-OFDM.
(a) Windowed OFDM (W-OFDM): Windowed OFDM [8] is specifically waveform greatly
researched to produce smooth waveform transitions amongst in line OFDM symbols in time
domain. Accurate designing of W-OFDM results in optimal time and frequency localization,
improved throughput in link even in noise affected channel path. Implanted at the
beginning of the w-OFDM symbol as CP. The samples is used as “prefix/suffix”
parameter for the sub-band. w-OFDM symbol denoted as contains
+2 samples. Prefix and suffix will be smoothed with a windowing operation, suffix

http://www.iaeme.com/IJARET/index.asp 1474 [email protected]

 Evaluation of Papr Reduction Ability of 5G Inspired Waveform Candidates

of the w-OFDM symbol will be overlapped with the first samples of the
w-OFDM symbol. The windowed symbol is obtained from the extended symbol
via

2.2 Single-carrier Waveforms

Single carrier (SC) waveforms are a promising prospect for the 5G transmission standard.
They are remarkably crucial in mm-Wave transmission [6]. Firstly, they display a much lower
peak-to-average-power ratio (PAPR) and the flexible system configuration incorporation than
OFDM; hence Uplink (UL) of 3GPP LTE deploys SC waveforms. Secondly, multiple
modulations and implementation structures can be inflicted on the data information. Thirdly,
they contribute higher flexibility in scheduling, permits equalization in the frequency domain
with lower OOB emissions. Fourthly, power consumption significantly reduces for high
power amplifiers (PA) with a much excellent coverage range. Lastly, it maintains superior
MIMO congeniality than pure single-carrier waveforms but not OFDM. These specific
characteristics mark this technology as a viable candidate with low PAPR for high-frequency
communication. This family is further classified into waveforms based on internal and
external guard intervals [8].

2.2.1 External Guard Band

External guard period is appended as Cyclic prefix to immune the transmitted data in
multipath channel environment from ISI and timing synchronization errors. CP length is
predefined, so is only suitable for the worst-case scenario. Users experience reduced system
performance at shorter delay spreads. CP-based waveforms introduce additional overhead
suffer from limited flexibility and poor spectral containment between different numerologies
(a)DFT-s-OFDM: LTE uplink transmission widely incorporates the usage of Discrete
Fourier Transform-spread-OFDM waveform. Its principle works as follows. Firstly, the DFT
block spreads the input data symbols, then proceeded for the IDFT stage. CP is prepended to
the symbol's start to avoid interference within adjacent symbol (ISI) that gets induced due to
the channel's multipath. It allows one-tap frequency domain equalization (FDE). Even though
it possesses simple design and precise implementation, DFT-S-OFDM manifests high out-of-
band (OOB) emissions, similar to OFDM. The broadcasted DFT-s-OFDM signal in discrete
time is denoted as

Here is the nth input from the input vector denoted the sum
of subcarriers of data carriers and non-data carriers .

2.2.2 Internal Guard Band

The same effect of the CP can be achieved using an internal guard period. The total length of
the guard band and data band remains predefined. However, flexibility can arise in the length
of the ratio of the lengths of data to the guard band. It helps preserve Orthogonality between
the signals on individual sub-bands is preserved. Internal guard band helps smooth transition
between symbols unlike that of the CP based waveforms significantly lessening the adjacent
channel interference (ACI) with better symbol energy usage. Unlike CP-based waveforms,
data bits in form of CP need not be stripped off at the receiver. Such waveform candidates

http://www.iaeme.com/IJARET/index.asp 1475 [email protected]

 Pathuri Lavanya, Dr. Penke Satyanarayana, Dr. Mohammad Mohatram

include Zero Tail DFT-s-OFDM (ZT DFT-s-OFDM) and Unique Word DFT-s-OFDM (UW
DFT-s-OFDM) [9]. Let us consider a generic data vector having length which is
mapped over the following column vector Here denotes the transpose
operator, and represents the corresponding time domain zero-head and the zero-tail
sequences, and represents the respective lengths. is a
represents column vector of data symbols. Vector has a length of
is fed to the DFT blocks whose output is then mapped over the frequency subcarriers and
IFFT-processed. Resultant time signal column vector can be expressed as:

Here denotes the unnormalized FFT matrix, for

and is the matrix which maps the data on the
frequency subcarriers filtering. The latter category falls under Sub-carrier filtering. The length
of the significant energy parts of the signal output is represented by

The identical sequences and , in practice the head and tail of the signal are only
quasi-identical since they contain the energy leakage from the data vector which varies at
each symbol.
(a) Zero Tail DFT-s-OFDM: One such adapted kind of SC – FDMA is the Zero Tail (ZT)
DFT spread OFDM technique [10]. There is almost no change compared to the previous
waveform, except stage II that attaches zero head and tail samples before the DFT stage.
Greater frequency and time localization is achieved than the SC-FDMA or OFDM. In large
delay spread multipath channel environments, ZT DFT-S-OFDM usually suffers from ISI as
the tail is not precisely zero which may be diverse from one symbol to the next.
(b) Unique Word DFT spread OFDM: Unique Word-spread-OFDM works on similar lines
to that of ZT-DFT-s-OFDM, the only difference is that a distinctive signal process is
conducted to generate Unique Word (UW) bits that gets appended to the trailing end of each
OFDM symbol [11]. On the contrary, several complexity considerations needs to be
considered wrt to the transmitter and receiver design. Firstly the unique word bits ought to be
well defined within the predefined bandwidth. Secondly, interference that occurs due to the
presence of word bits in the signal requires proper handling at the receiver
Principle of unique word DFT spread OFDM (UW-DFT-S-OFDM) and ZT-DFT-s-
OFDM are analogous, the only difference is that a distinctive signal process is conducted to
generate a Unique Word (UW) bits that gets appended to the trailing end of each OFDM
symbol. Hence they are specially deployed for uplink applications as they possess exciting
qualities like sequencing of symbols in time-domain, less peak-to-average power ratio, higher
power transmission efficiency at increased data rates [1].

3. RESEARCH ANALYSIS
Comprehensive analysis [12-19] has been conducted to estimate the performance of the MCM
waveforms using the Key Performance Indicators (KPI) factors and is presented in Table 2.
F-OFDM and UFMC are more flexible in reducing the complexity, compared to FBMC
and GFDM. The study found that PAPR rates for all candidates are lower than OFDM. This
may be a result for using filter and windowing techniques. The review found that a better

http://www.iaeme.com/IJARET/index.asp 1476 [email protected]

 Evaluation of Papr Reduction Ability of 5G Inspired Waveform Candidates

performance to reduce the PAPR was obtained when incorporating FBMC using orthogonal
quadrature amplitude modulation and GFDM more than F-OFDM and UFMC. This review
investigated the effects of the filter length on 5G. waveform candidates. This part affected the
performance of the waveforms and its usefulness in the overall complexity, latency and
spectral efficiency. It approved that all candidates acquired spectral efficiency more than
OFDM waveforms. However, a better spectral efficiency attained by FBMC waveform
obtained in high spectral localization. This criterion will result in a suitable coexistence with
CP-OFDM to enhance the power spectral density (PSD) of CP-OFDM.

Table.2. Pros and Cons of Multicarrier and Single Carrier Waveforms

Multi-Carrier Waveforms
Waveform Pros Cons
FBMC-OQAM 1.High Spectral Efficiency 1. High Adjacent Channel Leakage
2.Very low OOB 2. Not MIMO compatible,
3.Best frequency localization 3. Delay for short bursts
3.Robust to asynchronous access 4. No immunity to ISI due to lack of CP
4.Suitable for high-mobility applications 5. High implementation complexity with sync.
6. Increased power consumption due to
OQAM
W-OFDM 1.Robust to spectral & temporal offsets 1.Degradation due to time localization
2.Advantages same as CP-OFDM 2.Either poor spectral efficiency or BER
2.Lower OOBE than CP-OFDM
UFMC 1. Better OOB leakage suppression than 1. ISI from multipath fading
CP-OFDM 2. CP not used
2.Good frequency localization 2. Complex Tx/Rx than CP-OFDM
3.Shorter filter length than FBMC and
GFDM
4.Compatible with MIMO
GFDM 1. Flexible bandwidth with reduced sub- 1. Receiver cannot reduce ISI/ICI
carriers 3. Requires large guard band
2. Flexible design 4. Higher latency due to block processing
3. Good frequency localization 5. Bad MIMO integration and pilot design
4. Reduced PAPR 6. High implementation complexity
F-OFDM 1.Compatibility to other waveforms 1.Vulnerabe to filter length changes
2. Flexible filtering granularity 2.Very high implementation complexity
3.Better frequency localization
4.Shorter filter length than FBMC and
GFDM
Single-Carrier Waveforms
ZT-DFT-s- 1. Flexible bandwidth 1. Control signaling
OFDM 2. Inter-symbol guard 2. Low power ZT.
3. Better OOB 3. Strict synchronization
4. Low PAPR 4. Reduced performance for high order
5. Better spectral efficiency modulation
DFT-s-OFDM 1.Flexible bandwidth 1. Higher PAPR
2.Allows FDE 2. Bad ACI
3 Features as OFDM 3. High OOBE
4.Low PAPR 4. Strict synchronization
5. Hard-coded CP
UW-DFT-s- 1.Flexible guard interval 1. Strict synchronization
OFDM 2.Best spectral efficiency 2.Extra signaling
3.Lowest OOBE, PAPR 3. Complex Structure
Research studies clearly prove that though discrete Fourier transforms have been implemented
in FBMC, GFDM and UFMC for lower PAPR uplink communication in 5th gen. They have

http://www.iaeme.com/IJARET/index.asp 1477 [email protected]

 Pathuri Lavanya, Dr. Penke Satyanarayana, Dr. Mohammad Mohatram

been mostly in the scientific research stage and not have been released for commercial
purposes due to its high computational complexity and power consumption. Therefore, our
research bases its grounds with generalized discrete Fourier transform waveforms without
cyclic prefix.

4. EXPERIMENTAL SETUP
The experimental work is executed on a (HP) Hewlett-Packard personal workstation,
operating system Windows 7 Ultimate built with an Intel (R) i5 – 4200 U core processor,
processing at frequency speed of 2.30 GHz. In this simulation as tabulated in Table.3.

Table.3.Simulation Parameters for 5G Waveforms Candidates

Structure Parameters
CP-OFDM
Carrier frequency 2GHz
No. of sub-blocks 16
System Bandwidth 20MHz
No. of Frames 10000
Size of block data , 1200, 2048
8 ; 84
Symbols used per sub-frame 14 in DFT-OFDM
15 in ZT- DFT-s-OFDM
Number of bits per symbol 4
Sampling frequency 30.72 MHz
Cyclic prefix length 72
GFDM
GFDM block length 15
Number of guard symbols 2
RRC filter roll-off 0.2
Number of IC iterations 4
UFMC
Guard interval length 72
Sub-band width 12
Filter length 80
Side-lobe attenuation 40dB.
FBMC
Overlapping factor 4
Phase Rotation Vectors 4
F-OFDM
GI Length 128
Number of sub-bands 25
Filter scheme Dolph-Chebyshev
Filter length 129

http://www.iaeme.com/IJARET/index.asp 1478 [email protected]

 Evaluation of Papr Reduction Ability of 5G Inspired Waveform Candidates

Side-lobe attenuation 40
W-OFDM
Physical Resource Block 15

Roll-off factor 0.6

10000 uniformly distributed input OFDM signals are dealt using 16-QAM modulation. First,
simulations are performed to understand the effect of PAPR reductions deploying different
modulations types, overlapping factors and number of subcarriers. Complementary
cumulative distribution function (CCDF) and bit error rate (BER) have been the two most
important benchmarks to assess the PAPR. Firstly, PAPR performance of OFDM waveform
having 10000 input data signals for 16-QAM incorporating various sub-carriers N is
simulated in Matlab 2019a.
For CP-OFDM of length of 72. For GFDM, block length of
=15. 2 guard bands, RRC filter roll-off of 0.2 and uses 4 iterations. For UFMC, a
guard interval 72, using filter length 80 and side lobe attenuation of 40dB. For FBMC, 4
phase rotation vectors. For F-OFDM, GI Length of 128, number of subcarriers per sub-band
of 12, with 25 sub-bands . For W-OFDM, 15 physical resource blocks, and 12
subcarriers are deployed. To implement ZT-DFT-s-OFDM, 15 zero padding symbols used per
sub-frame, UW-DFT-s-OFDM, 15 unique word symbols per frame. The ratio of head ratio to
tail ratio of 8 : 84. However, 14 symbols per frame is deployed in DFT-OFDM. The
factors are deployed to provide same computational complexity similar to OFDM and brings
about the total of same length of data with CP/GI padded irrespective of single carrier and
multiple carrier waveforms considered for simulation. Total priority would be to consider the
PAPR reduction performance, processing time and bit error rate reduction.

5. PERFORMANCE OUTCOMES
Firstly, we compare the CCDF vs. PAPR graph outcomes of the multicarrier waveforms at 10 -
3
dB as illustrated in Figure 3. DFT-s-OFDM has the maximum PAPR reduction of 1.5dB at
10-4dB of CCDF to 10dB from 11.5 dB where conventional OFDM stands. UFMC results in
the minimum PAPR reduction of 0.2 dB to 11.4dB from the original OFDM performance of
around 11.6dB. The PAPR values of F-OFDM, W-OFDM, FBMC, and GFDM lowers to 11.8
dB, 11.6dB, 11.4dB, 11.1dB respectively.
0
10
OFDM
UFMC
f-OFDM
W-OFDM
FBMC
-1
10 GFDM
DFT-s-OFDM
CCDF (dB)

-2
10

-3
10
6 7 8 9 10 11 12 13
PAPR (dB)

Figure 3 CCDF vs. PAPR of various multicarrier 5G waveform candidates

http://www.iaeme.com/IJARET/index.asp 1479 [email protected]

 Pathuri Lavanya, Dr. Penke Satyanarayana, Dr. Mohammad Mohatram

Similarly, Figure 4 pictorized the PAPR values of single carrier waveforms without cyclic
prefix or internal guard band are studied at 10-3dB CCDF. PAPR using DFT-s-OFDM, ZT-
DFT-s-OFDM, UW-DFT-s-OFDM lowers to 9.5 dB, 9.7dB and 9.3 dB respectively. In this
simulation, complexity of the techniques including the multiplication units and addition units
are considered to the same as that of OFDM. Processing time is not considered. That would
be the future scope of the research.
0
10
OFDM
ZT-DFT-s-OFDM
DFT-s-OFDM
-1 UW-DFT-s-OFDM
10
CCDF (dB)

-2
10

-3
10

-4
10
2 4 6 8 10 12 14
PAPR (dB)

Figure 4 CCDF vs. PAPR for various single carrier candidates

Now considering the Signal to Noise Ratio (SNR) variation to Bit error rate (BER) in
multicarrier waveforms at a CCDF of 10-3 dB as in Figure 5. BER value of UFMC is almost
similar or higher than OFDM at about 25.2 dB when OFDM stands at 25 dB. F-OFDM has
the minimum reduction in signal to noise ratio is almost same value as original OFDM of 24.2
dB at 10-3 dB (BER).
0
10
UFMC
OFDM
f-OFDM
W-OFDM
GFDM
-1 FBMC
10
DFT-s-OFDM
Bit Error Rate

-2
10

-3
10
17 18 19 20 21 22 23 24
SNR [dB]

Figure 5 BER vs. SNR of various multicarrier 5G waveform candidates using 16QAM modulation,
=4, =128 subcarriers.
Figure 6 illustrates the BER results of DFT spread single carrier waveforms. Discrete Fourier
Transform-spread-OFDM results in the maximum signal to noise ratio reduction of 1dB to
24.2dB below the conventional OFDM system nearing to 25.2dB. The signal to noise ratio of
the Single carrier waveforms 10-3 dB BER using Zero Tail-DFT-s-OFDM and Unique Word-

http://www.iaeme.com/IJARET/index.asp 1480 [email protected]

 Evaluation of Papr Reduction Ability of 5G Inspired Waveform Candidates

DFT-s-OFDM lowers to 18.3 dB and 18.1dB respectively. Future research projects in this
domain may rely on further considering other metrics to further enhance the peak to average
power ratio performance measurements
0
10
OFDM
ZT-DFT-s-OFDM
DFT-s-OFDM
UW-DFT-s-OFDM

-1
10
Bit Error Rate

-2
10

-3
10
0 2 4 6 8 10 12 14
SNR [dB]

Figure 6 BER vs. SNR for various single carrier candidates using 16QAM modulation, =4, =128
subcarriers.

6. CONCLUSIONS
The extensive investigation on 5G waveform candidates developed to overcome OFDM's
limitations is presented in this review. Firstly, an in-detail description, advantages, and
limitations of individual waveform candidates are presented. The KPI factors are considered
to assess the relative usefulness of the respective waveforms 5G implementation. High PAPR
reduction and OOB, radiation reduction capability, are given primary preference to review the
candidates.
The following are the resulting summary obtained from the simulation study of CCDF
variation with PAPR. Firstly, no waveform technology delivers superior KPI features. Every
waveform has its shortcomings. Secondly, the multicarrier waveforms possess complicated
structures than Single Carrier Waveforms. Thirdly, Single carrier waveforms exhibit lower
PAPR than OFDM. Lastly, all waveforms either appends guard-band (GB), zero-padding
(ZP), or cyclic prefix (CP) between symbols to avoid ISI and ICI. Implementation of the
waveform technology in specific scenarios and channel conditions must consider the KPI
factors. Future research in this domain calls for collaborative research to develop additional
benchmarks to integratively bring about collective KPI optimizations for the 5G waveform
candidates. The critical analysis and implementation should depend upon the appropriate
application.

REFERENCES
[1] A.K. Bashir, R. Arul, S. Basheer,G. Raja,R. Jayaraman, N,M,F Qureshi, An optimal multitier
resource allocation of cloud RAN in 5G using machine learning, Transactions On Emerging
Telecommunications Technologies, 30(8), 2019, e3627.

http://www.iaeme.com/IJARET/index.asp 1481 [email protected]

 Pathuri Lavanya, Dr. Penke Satyanarayana, Dr. Mohammad Mohatram

[2] M. V.Rao, B.T.P. Madhav, J.K.Y. Usha Devi, T. Anilkumar, B. P. Nadh, CSRR-loaded T-
shaped MIMO antenna for 5G cellular networks and vehicular communications,29(8), 2019,
e21799

[3] M. Laabidi, and R. Bouallegue, Three implementations of the tone reservation PAPR
reduction scheme for the FBMC/OQAM system, IET Comm., 13(7), 2019, pp. 918 - 925.

[4] K. Liu, W. Deng, and W. Liu, Theoretical Analysis of the Peak-to-Average Power Ratio and
Optimal Pulse Shaping Filter Design for GFDM Systems, IEEE Transactions on Signal
Processing, 67(13) 2019, pp. 3455 – 3470.

[5] L. Zhang, A. Ijaz, P. Xiao, M.M. Molu, R. Tafazolli, and L. A. Dasilva, Filtered OFDM
systems, algorithms, and performance analysis for 5G and beyond, IEEE Transactions of
Communications, 66(3), 2018, pp. 1205–1218,

[6] A.F. Almutairi, M.L Gharabally, and A. Krishna, Performance Analysis of Hybrid Peak to
Average Power Ratio Reduction Techniques in 5G UFMC Systems, IEEE Access, 2017(1),
2017.

[7] M.B. Mabrouk, M. Chafii, Y. Louet, and F. Bader, A Precoding-based PAPR Reduction
Technique for UF-OFDM and Filtered-OFDM Modulations in 5G Systems, 23th European
Wireless Conference, May 2017.

[8] H.G. Ryu, and A. Changyoung, Design and Performance Comparison of W-OFDM Under the
Nonlinear HPA Environment, Wireless Personal Communications vol. 98, 2018, pp.983–999

[9] M. Özden, C. Ülkü, A. Özen, An investigation of the effects of different UW sequences on the
PAPR performance of UW-OFDM systems, 26th Signal Processing and Communications
Applications Conference (SIU), May 2018.

[10] A. Şahin, R. Yang, E. Bala, C. Beluri, R.L. Olesen, Flexible DFT-S-OFDM: Solutions and
Challenges, IEEE Comm. Magazine, 54(11), 2016, pp. 106 – 112.

[11] G. Berardinelli, K. Pedersen, F. Frederiksen, T.B. Sorensen, P. Mogensen, Unique Word DFT-
Spread-OFDM for Fast Time-Varying Channels, 85th IEEE Vehicular Technology
Conference, June 2017

[12] S.Y Fathima, M.Z.U Rahman, K.M.Krishna, S. Bhanu, M.S. Shahsavar, Side Lobe
Suppression in NC-OFDM Systems Using Variable Cancellation Basis Function, IEEE
Access, 5, 2017, pp. 9415-9421.

[13] K. Sreelakshmi.P. Bora, M. Mudaliar,Y.B. Dhanade, B.T.P Madhav, Linear array Yagi-Uda
5G antenna for vehicular application”, (UAE), 7 (1.1), 2018, pp. 513- 517.

[14] S. Umar, N. Priya, P. Gayathri, T.S. Subba, A.M. Abdul , Design of jitter spectral shaping as
robust with various oversampling techniques in OFDM, Smart Innovation, Systems and
Technologies, 77, May 2017 ,pp. 641- 647.

[15] M.M.R. Nelluri, H. Khan, Power efficient 5G wireless networks based on self-organized
utility function, International Journal of Innovative Technology and Exploring Engineering,
8(7), 2019, pp.1857-1862.

[16] C. Vanaja,N. Pavithra, N. Sravya, M Manoj, Y.B Dhanade, Enhanced franklin antenna for the
future 5G communication applications, International Journal of Innovative Technology and
Exploring Engineering, 8(6), 2019, pp.1168-1172.

http://www.iaeme.com/IJARET/index.asp 1482 [email protected]

 Evaluation of Papr Reduction Ability of 5G Inspired Waveform Candidates

[17] Mahender K., Ramesh K.S., Kumar T.A., An efficient OFDM system with reduced PAPR for
combating multipath fading, Journal of Advanced Research in Dynamical and Control
Systems, 9(14), 2017, pp.1939-1948.

[18] Krishna A.R., Chakravarthy A.S.N., Sastry A.S.C.S., Utilization of SC-FDMA and OFDMA
based uplink and downlink resources in LTE-a network-assisted device to device
communication for effective spectrum management, ARPN Journal of Engineering and
Applied Sciences, 12(5), 2017, pp.1429-1439.

[19] Mahender K., Kumar T.A., Ramesh K.S., Performance study of OFDM over multipath fading
channels for next wireless communications, International Journal of Applied Engineering
Research, 12(20), 2017, pp.10205-10210.

http://www.iaeme.com/IJARET/index.asp 1483 [email protected]