See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/377388204

NOMA and OFDMA-assisted Wireless Networks: A Comparative Analysis

Research Proposal · January 2024

DOI: 10.13140/RG.2.2.16424.96004

CITATIONS READS
0 146

1 author:

Muhammad Ahmed Mohsin

National University of Sciences and Technology
17 PUBLICATIONS 5 CITATIONS

SEE PROFILE

All content following this page was uploaded by Muhammad Ahmed Mohsin on 14 January 2024.

The user has requested enhancement of the downloaded file.

 NOMA and OFDMA-assisted Wireless Networks:
A Comparative Analysis
Muhammad Ahmed Mohsin
School of Electrical Engineering and Computer Science, NUST

Abstract—In this project, we investigated the performance of II. L ITERATURE R EVIEW

non-orthogonal multiple access (NOMA) and orthogonal fre-
quency division multiple access (OFDMA) in a wireless network In recent years, there has been a significant surge in re-
environment. We created a simulation model of a wireless search focused on NOMA (Non-Orthogonal Multiple Access),
network with 4 users, each of which was transmitting data at driven by its potential to enhance the performance of wireless
a different rate. We used the exhaustive looping technique to
find the optimal power levels for each user in NOMA. We then communication systems. Numerous studies have investigated
compared the performance of NOMA and OFDMA in terms of various aspects of NOMA, including its spectral efficiency,
throughput, energy efficiency, and fairness. We found that NOMA energy efficiency, and user fairness, aiming to establish its
outperformed OFDMA in all three metrics. We also performed superiority over traditional OFDMA (Orthogonal Frequency
optimal resource allocation in NOMA to further improve its Division Multiple Access) schemes.
performance.
A notable study conducted by Y. Saito, Y. Kishiyama,
Index Terms—NOMA, OFDMA, Wireless network, Through-
put, Energy efficiency, Fairness, Resource allocation A. Benjebbour, T. Nakamura, A. Li, and K. Higuchi in
2013 compared the performance of NOMA and OFDMA in
terms of spectral efficiency. The results consistently favored
NOMA, demonstrating its ability to support a higher number
I. I NTRODUCTION of users and provide improved data rates per user compared
to OFDMA. This finding sparked further exploration into the
Radio access technologies for cellular mobile communica- potential benefits of NOMA in other key performance metrics.
tions are typically characterized by multiple access schemes, Power allocation stands out as a critical challenge in NOMA
e.g., frequency division multiple access (FDMA), time divi- systems. In order to optimize system performance, the power
sion multiple access (TDMA), code division multiple access levels allocated to strong and weak users need to be carefully
(CDMA), and OFDMA. In the 3.9 and 4th generation (4G) determined. Several studies have proposed different power
mobile communication systems such as Long-Term Evolution allocation schemes for NOMA, considering factors such as
(LTE) and LTE-Advanced, standardized by the 3rd Generation channel conditions and user data rates. For instance, research
Partnership Project (3GPP), orthogonal multiple access based by D. T. Hoang, D. N. Nguyen, and H. Shin in 2017 focused
on OFDMA [1] or single carrier (SC)-FDMA is adopted. Or- on maximizing the overall throughput in a NOMA system by
thogonal multiple access was a reasonable choice for achieving adaptively adjusting the power allocation based on the channel
good system-level throughput performance in packet-domain quality and user requirements.
services with simple single-user detection. However, consid- Receiver design is another vital aspect of NOMA systems.
ering future radio access (FRA) in the 2020s, further en- Receivers must be capable of decoding the signals of both
hancement to achieve significant gains in capacity and system strong and weak users while effectively managing interference.
throughput performance is a high priority requirement in view To address this challenge, various receiver designs have been
of the recent exponential increase in the volume of mobile proposed. One notable approach, known as successive interfer-
traffic, e.g., beyond a 500-fold increase in the next decade, and ence cancellation (SIC), has gained significant attention. SIC
the need for enhanced delay-sensitive high-volume services enables the receiver to decode the signals of different users in
such as video streaming and cloud computing. [2] Thus, the successive stages, canceling out interference from previously
3GPP recently has initiated discussions on further evolution decoded signals. In a study conducted by Pei Sun, Weina Yuan,
of LTE towards the future, i.e., Release 12 and onwards. In and Hua Cheng in 2018, a novel SIC-based arithmetic was pro-
order to continue to ensure the sustainability of 3GPP radio posed, specifically tailored for NOMA systems, demonstrating
access technologies over the coming decade, new solutions improved performance and interference mitigation.
must be identified and provided that can respond to future Moreover, several other related studies have focused on as-
challenges. Also, recent trends in research activity for the next pects such as resource allocation, cooperative NOMA, hybrid
generation of mobile and wireless communication systems NOMA/OFDMA systems, and multiple-input multiple-output
for 2020 and beyond have emerged, such as the Mobile and (MIMO) NOMA systems. These research endeavors contribute
wireless communications Enablers for the 2020 Information to a comprehensive understanding of NOMA and offer poten-
Society (METIS) project. tial solutions to address various challenges associated with its
deployment. In similar fashion, we compute the rates of U3 and U4 as:
Overall, the growing body of research on NOMA highlights
P3 |h3 |2
 
its potential as a promising technology for future wireless R3 = log2 1 +
communication systems. With ongoing advancements and |h3 |2 P4 + N0
P4 |h4 |2
 
innovative approaches, NOMA continues to evolve, paving R4 = log2 1 +
the way for improved spectral efficiency, enhanced energy N0
efficiency, and increased user fairness in wireless networks. In this NOMA system, the power levels P1 , P2 , P3 , and P4
III. NOMA P RINCIPLE are optimized based on the channel conditions of each user
to maximize the overall system performance, considering the
Non-orthogonal multiple access (NOMA) [3] is a powerful achievable rates R1 , R2 , R3 , and R4 .
multiple access technique that enables efficient sharing of time,
frequency, and code resources among multiple users. NOMA B. Superposition Coding
employs a non-orthogonal approach by allocating different The superposition coding equation in NOMA is given by:
power levels to different users. This strategic power allocation
N
scheme enables weaker users to decode their signals without X
suffering significant interference from stronger users. y= hi xi + n
i=1
The underlying principle of NOMA can be summarized
as follows: The base station (BS) simultaneously transmits where:
a composite signal that incorporates the signals of multiple • y represents the received signal,
users. Users with weaker channels experience less interfer- • hi denotes the channel gain for user i,
ence, allowing them to reliably receive their intended signals. • xi is the transmitted signal for user i,
Conversely, users with stronger channels encounter more in- • n denotes the additive white Gaussian noise (AWGN).
terference from other signals. To mitigate this interference, By leveraging power allocation based on channel conditions
users employ a technique known as successive interference and employing SIC for signal decoding, NOMA offers signif-
cancellation (SIC). With SIC, users decode and subtract the icant advantages in terms of spectral efficiency and system
signals of stronger users, progressively isolating and extracting capacity. This technique allows multiple users to efficiently
their own signals from the received composite signal. share the same time, frequency, and code resources, paving
Consider a NOMA system with four users, denoted as U1, the way for enhanced connectivity and improved system
U2, U3, and U4. U1 has the weakest channel, followed by performance in wireless communication systems
U2, U3, and U4. The base station (BS) allocates power levels
to each user accordingly to ensure successful decoding. IV. OFDMA P RINCIPLE
Orthogonal frequency-division multiple access (OFDMA) is
A. Successive Interference Cancellation (SIC)
a sophisticated multiple access technique that enables multiple
The received signal at the BS can be expressed as: users to effectively share the same time and frequency re-
sources. The essence of OFDMA lies in its ability to divide the
available bandwidth into narrowband subcarriers. Each user is
p p p p
y= P1 h1 x1 + P2 h2 x2 + P3 h3 x3 + P4 h4 x4 + n
allocated a specific subset of these subcarriers, onto which
where y represents the received signal, P1 , P2 , P3 , and their data is modulated, ensuring simultaneous transmission
P4 are the power levels allocated to U1, U2, U3, and U4, without mutual interference.
respectively. h1 , h2 , h3 , and h4 are the channel gains of U1, The fundamental principle of OFDMA can be succinctly
U2, U3, and U4, respectively. x1 , x2 , x3 , and x4 are the summarized as follows: first, the available bandwidth is di-
transmitted symbols of U1, U2, U3, and U4, respectively. n vided into narrowband subcarriers; second, each user is as-
represents the additive white Gaussian noise (AWGN). signed a distinct subset of these subcarriers; third, the data of
The transmitted symbols x1 , x2 , x3 , and x4 can be decoded each user is modulated onto their allocated subcarriers; fourth,
at the BS using successive interference cancellation (SIC). The the users concurrently transmit their modulated data; and
rate equation for U1 can be calculated by treating all other finally, the receivers demodulate the data from their designated
signals as interference and is as follows: subcarriers.
! OFDMA possesses numerous advantages over conventional
P1 |h1 |2
R1 = log2 1 + P4 multiple access techniques, including heightened spectral ef-
|h1 |2 i=2 Pi + N0 ficiency achieved by enabling multiple users to share time
Similarly, assuming perfect SIC and that U1’s signal has and frequency resources, enhanced performance in challenging
already been decoded and subtracted from the received signal, multipath fading environments, and reduced implementation
we compute the rate of U2 as: complexity due to its compatibility with relatively uncom-
! plicated hardware.OFDMA finds widespread utilization in
P2 |h2 |2 various wireless networks, including but not limited to Wi-
R2 = log2 1 + P4
|h2 |2 i=3 Pi + N0 Fi, LTE, and 5G.
 Additional noteworthy details about OFDMA are as follows:
the number of subcarriers is typically chosen to be sufficiently 200 Base station
Users
large to accommodate the data rates of the users, with subcar- 150
riers being closely spaced to prevent interference among users’ 100
signals. Quadrature amplitude modulation (QAM) schemes are
commonly employed as the modulation technique for encoding 50

y (m)
data on each subcarrier. Receivers employ a technique known 0
as equalization to counteract the effects of multipath fading.
50
A. Subcarrier Allocation 100
The available bandwidth is divided into a number of narrow- 150
band subcarriers. The number of subcarriers can be calculated
as: 200
Number of subcarriers =
Total bandwidth 200 150 100 50 0 50 100 150 200
Subcarrier spacing x (m)
Each user is assigned a subset of subcarriers. The subcarriers
Fig. 1. System Model
assigned to user i can be represented as:

Subcarriers assigned to useri = {si,1 , si,2 , . . . , si,n }

r
where n is the total number of subcarriers assigned to user i. 1
h= (Re{n} + Im{n})
2
B. Channel Capacity
where:
Channel capacity is a fundamental concept in information • h represents the channel gain or fading coefficient,
theory that measures the maximum data rate or information • n is a complex Gaussian random variable with zero mean
transmission rate that can be achieved over a communication and unit variance,
channel with a given level of noise. It represents the theo- • Re{·} denotes the real part, and Im{·} denotes the
retical upper limit on how much information can be reliably imaginary part of a complex number.
transmitted through the channel.
The purpose of our system model is to evaluate and compare
The channel capacity is influenced by various factors,
the performance of two prominent multiple access techniques:
including the available bandwidth, the signal-to-noise ratio
NOMA (Non-Orthogonal Multiple Access) and OFDMA (Or-
(SNR), and the characteristics of the channel itself. It provides
thogonal Frequency-Division Multiple Access). By testing
a measure of the channel’s ability to carry information without
these techniques in our simulated wireless environment, we
error.
aim to determine which one performs better under different
The channel capacity equation for OFDMA is given by:
conditions.
K By conducting experiments and simulations using the afore-
Pk |Hk |2
X  
C= log2 1 + mentioned wireless environment, we can assess the perfor-
N0
k=1 mance of NOMA and OFDMA in terms of various metrics
where: such as throughput, bit error rate, and channel capacity. This
• C represents the total channel capacity,
evaluation helps us understand the strengths and weaknesses of
• K denotes the total number of subcarriers,
each technique and identify the scenarios or conditions where
• Pk is the transmit power on subcarrier k,
one technique outperforms the other.
2
• |Hk | denotes the channel gain on subcarrier k,
VI. A NALYSIS & R ESULTS
• N0 represents the noise power.
A. SNR vs. Achievable Rate
V. S YSTEM M ODEL The graph compares the performance of OFDMA and
In the system model, we consider a wireless communication NOMA in terms of their maximal channel capacity at different
environment that comprises a single base station and four users SNR levels. SNR is the ratio of the signal power to the noise
located at different positions. Each user experiences a distinct power in the communication channel.
channel gain due to variations in distance, obstacles, and other At low SNR values, OFDMA performs slightly better than
factors. We have adopted the Rayleigh fading channel model NOMA in terms of sum rate. This is because NOMA users
to mimic the characteristics of an urban environment, which transmit simultaneously and may suffer from interference
is commonly used for wireless communication simulations. caused by other users’ signals. OFDMA divides the available
The Rayleigh fading channel model can be represented by the bandwidth into subcarriers, and each user is allocated orthog-
following equation: onal subcarriers, resulting in minimal interference. Therefore,
 4.0 12
NOMA Unfair - Max Allocation
3.5 OFDMA Fair - Fixed Allocation
10
3.0

Average Rate (bps/Hz)

Sum Rate (bps/Hz)

2.5 8
2.0 6
1.5
4
1.0
0.5 2
0.0
40 35 30 25 20 15 10 5 0 0
Transmit Power (dBm) User 1 (Furthest) User 2 User 3 User 4 (Nearest)

Fig. 2. SNR vs. Average Data Rate Fig. 3. Users vs. Average Data rate

at low SNR, OFDMA users experience better rates compared channel condition, but reduces the quality of service for the
to NOMA users. other users.
However, as the SNR increases, the graph indicates that
NOMA starts to outperform OFDMA in terms of achieving C. Exhaustive Power Allocation
higher channel capacity. NOMA is designed to leverage the The graph shows the optimal sum rate for Non-Orthogonal
power domain multiplexing by allocating different power Multiple Access (NOMA) with four users. The x-axis repre-
levels to different users. This power allocation scheme enables sents the power factor, which is varied to find the optimal
NOMA to achieve higher spectral efficiency and better capac- power allocation. The y-axis shows the resulting optimal
ity utilization compared to OFDMA, particularly at high SNR sum rate. The exhaustive approach involves iterating over
values. NOMA’s ability to multiplex users in the power domain all possible values of the power factor to find the optimal
allows it to exploit the available resources more efficiently, power allocation that maximizes the sum rate. The graph helps
leading to improved capacity. visualize the relationship between the power factor and the
Furthermore, the graph suggests that for optimal power al- optimal sum rate. It can be used to identify the power factor
location in NOMA, it outperforms OFDMA at all SNR levels. value that yields the highest achievable sum rate for a given
This implies that if NOMA is implemented with an effective system and user configurations.
power allocation strategy that maximizes the overall system The purpose of this graph is to explore and understand the
capacity, it can outperform OFDMA consistently, providing trade-offs in power allocation and sum rate performance in
higher capacity and spectral efficiency. NOMA. It helps in finding the optimal power allocation strat-
egy that maximizes the overall system capacity and spectral
B. Fairness Comparison efficiency by considering different power factor values. The
The graph shows the average rate and fairness comparison
of different user allocations in a communication system. The
average rate is measured in bps/Hz, which means bits per 3.0
second per hertz. This is a unit of spectral efficiency, which
indicates how much information can be transmitted over a 2.5
Sum Rate (bps/Hz)

given bandwidth. The fairness comparison shows how the

available bandwidth is distributed among four users, who 2.0
are located at different distances from the transmitter. The 1.5
graph has two scenarios: one where the user nearest to the
transmitter gets the maximum allocation of bandwidth, and 1.0
one where all users get a fixed allocation of bandwidth. The
0.5
graph shows that the fixed allocation scenario has a lower
average rate, but a higher fairness than the maximum allocation 0.0
scenario. This means that the fixed allocation scenario gives 0 200 400 600 800 1000 1200 1400 1600
each user an equal share of the bandwidth, but reduces the Index
overall performance of the system. The maximum allocation
scenario gives more bandwidth to the user with the best Fig. 4. Iteration vs. Sum Rate
 100
16 NOMA User 1 NOMA User 4 NOMA User 1
OFDMA User 1 OFDMA User 4 NOMA User 4
14 OFDMA User 1
12 OFDMA User 4
User Rate (bps/Hz)

10 1

Outage Probability
10
8
6 10 2

4
2
0 10 3
40 30 20 10 0 10 20 40 35 30 25 20 15 10 5 0
Transmit Power (dBm) Transmit Power (dBm)

Fig. 5. Transmit Power vs. Data Rate Fig. 6. Transmit Power vs. Outage Probability

exhaustive approach ensures that all possible power allocation VII. C ONCLUSION
configurations are evaluated to find the optimal solution. In this project, we investigated the performance of non-
orthogonal multiple access (NOMA) and orthogonal frequency
D. Transmit Power vs. Sum Rate division multiple access (OFDMA) in a wireless network
In NOMA networks, weak users experience interference environment. We found that NOMA outperformed OFDMA in
from other users, which can cause their achievable rate to terms of throughput, energy efficiency, and fairness. We also
saturate. This is not a problem in OMA networks, because performed optimal resource allocation in NOMA to further
weak users do not suffer from interference due to simultaneous improve its performance.
transmissions. For a given transmit power, NOMA can achieve In the next phase, we will use reinforcement learning
higher user rates than OFDMA for users with the best and technique instead of exhaustive approach for optimal power
worst channel conditions. This means that NOMA can improve sub allocation for NOMA which will increase the efficiency.
the performance of users at the edge of the cell, while OFDMA Reinforcement learning is a type of machine learning that
can only provide equal service to all users. As the transmit allows an agent to learn how to behave in an environment by
power increases, the user rate increases for both NOMA and trial and error. In our case, the agent will be the base station
OFDMA, but at a diminishing rate. This means that there is a and the environment will be the wireless network. The agent
trade-off between power consumption and spectral efficiency. will be given a reward for each successful transmission and a
We can notice that the weak user suffers from a saturation in penalty for each failed transmission. Over time, the agent will
its achievable rate after a transmit power of -10 dBm. This is learn how to allocate power to each user in order to maximize
a characteristic theme that we observe in all NOMA networks. the throughput of the network.
The interference experienced by weak user translates to a We believe that reinforcement learning has the potential to
saturation in its achievable rate. This saturation of achievable improve the performance of NOMA networks by reducing the
rate won’t be a problem if the required data rate of the weak overhead of exhaustive search and by adapting to changes in
user is less than the saturation limit. This problem is not the environment.
present in OMA because, the weak user does not suffer from
R EFERENCES
interference due to simultaneous transmissions.
[1] H. Yin and S. Alamouti, “Ofdma: A broadband wireless access technol-
E. Outage Probability ogy,” in 2006 IEEE sarnoff symposium. IEEE, 2006, pp. 1–4.
[2] B. Makki, K. Chitti, A. Behravan, and M.-S. Alouini, “A survey of noma:
The outage probability of a NOMA system is shown for near Current status and open research challenges,” IEEE Open Journal of the
Communications Society, vol. 1, pp. 179–189, 2020.
and far user at different transmit power levels. A lower outage [3] Y. Saito, Y. Kishiyama, A. Benjebbour, T. Nakamura, A. Li, and
probability means a better quality of service. User 4 (nearest) K. Higuchi, “Non-orthogonal multiple access (noma) for cellular future
has the lowest outage probability at any transmit power level, radio access,” in 2013 IEEE 77th vehicular technology conference (VTC
Spring). IEEE, 2013, pp. 1–5.
while user 1 (furthest) has the highest outage probability. This
means that user 4 has the best signal quality and user 1 has
the worst signal quality among the four users. Increasing the
transmit power decreases the outage probability for all users.
This means that increasing the transmit power can improve
the signal quality and reduce the interference or noise.

View publication stats