Received 26 May 2024, accepted 15 June 2024, date of publication 26 June 2024, date of current version 3 July 2024.

Digital Object Identifier 10.1109/ACCESS.2024.3419072

Car-to-Car Communication Using RF Cognitive

Radio With VLC Common Control Channel
NERMEEN M. OKASHA 1,2 AND FATMA A. NEWAGY 1, (Member, IEEE)
1 Department of Electronics and Communications Engineering, Faculty of Engineering, Ain Shams University, Cairo 11517, Egypt
2 Department of Electronics and Communications Engineering, Modern Academy for Engineering and Technology, Cairo 11585, Egypt

Corresponding author: Nermeen M. Okasha ([email protected])

ABSTRACT This paper combines the advantages of both cognitive radio (CR) and visible light communica-
tion (VLC) for car-to-car applications to achieve a high data rate with minimum (delay, outage probability,
bit error rate (BER), and cost). CR technology hops among the existing radio frequency (RF) available
channels to increase the RF spectrum usage efficiency and dodge the scarcity limitation. Moreover, using
CR as a license-free application will reduce car-to-car communication running costs. However, CRs require a
common control channel (CCC) to communicate the spectrum availability map within the CR network and to
inform the receiver end about the change in the transmitter-end channel. Therefore, the CCC is the bottleneck
in the car-to-car CR network. Then, we explore the types of CCCs and discuss using each of them to solve
this bottleneck issue. In the proposed scheme, we adopt using VLC as CCC. A MATLAB simulation for a
car-to-car framework is built to demonstrate the capabilities of VLC through the chosen metrics (i.e., data
rate, delay, outage probability, cost, and bit error rate). Our results show that VLC achieves up to 90% of the
licensed data rate with a small outage probability of 21.2% and moderate BER and delay. In addition, VLC
presents the minimum cost, placing second after the licensed type with a score of 84.2% in the combined
metric. In conclusion, with the VLC’s bright future of expansion and growth in the car-to-car application,
we have proven that VLC is worthy of implementation practically in modern cars.

INDEX TERMS Car to car, CCC, CR, fog effect, rain effect, VLC.

I. INTRODUCTION A. BACKGROUND
Due to the increased usage of social media platforms, people There are many communication methods to transmit data
nowadays get distracted more often even while driving. More- between cars. One of the V2V communication system meth-
over, the excessive dopamine dose received while watching ods is radio frequency (RF) communication [4]. However, due
social media platforms makes people quickly feel bored while to the popularity of RF, the RF channels are mostly booked
driving. Consequently, car accidents have increased dramat- and crowded. Using the standard channel assigning technique
ically in the last decade [1]. Therefore, many researchers to cars is not the best solution for transmitting data between
proposed adding the car-to-car communication function to all vehicles as there is no available backup spectrum for all
newly manufactured cars to increase road safety [2], [3]. That cars on the road. Some researchers propose using the current
communication means providing vital data about the road, infrastructure of wireless telecommunication networks [5].
ahead cars, and the following cars. The car-to-car commu- At the same time, the others propose the establishment of a
nication transfers data between cars to give each one a global stand-alone network that is designed spatially for car-to-car
view of the road. It aims to achieve accident-free traffic and communication [6]. These two assumptions are too costly to
smooth road flow with a small congestion probability. be implemented in real situations. Moreover, the RF spectrum
is too crowded to accept the new car-to-car communication
bandwidth either within the mobile communication network
or using the establishment of a stand-alone network [7].
The associate editor coordinating the review of this manuscript and To avoid this overcrowded RF problem, Academia has
approving it for publication was Hayder Al-Hraishawi . recently introduced the idea of communication systems in
2024 The Authors. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
VOLUME 12, 2024 For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/ 89339
 N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

the millimeter-wave (mmWave) bands to overcome the prob- nodes are in destructive interference, the remaining nodes
lem of over-occupied low-frequency bands (i.e., systems that correct their decision through cooperation. Such cooperation
operate in the RF band from 3 kHz to 300 GHz) [8]. In the requires a separate communication channel between SUs to
meantime, mmWave communication systems provide higher transmit control messages other than the data channels.
data rates. In contrast, the challenges of using the mmWave Consequently, cognitive radios require a common control
band in radio frequency (RF) systems are not limited to the channel to communicate the spectrum availability map within
maximum produced transmission power from the electronic the network and to tell the receiver end about the change in
devices in that band and the complexity of channel modeling the transmitting channel. Hence, the common control chan-
due to the rain attenuation and atmospheric varieties. More- nel is the bottleneck for this method [17]. There are many
over, the mmWave communication band faces practical issues types of common control channels and each of them tries
such as antenna alignments and beamforming because of its to solve this bottleneck issue from a philosophical point of
challenging multiple propagation characteristics [9]. On the view. Generally, the common control channel types fall into
other hand, communication systems in mmWave bands are two categories (i.e., in-band and out-of-band) [18], [19]. The
predicted to be the strongest candidate for 6G and beyond in-band common control channel is either an underlay chan-
and these systems might introduce some new IEEE standards nel using an ultra-wideband spread spectrum (e.g., CDMA)
(e.g., at the 60 GHz band, the IEEE standard 802.11ad is or uses one of the channels with the least existence of pri-
introduced) [10]. mary users (PUs) [20]. The underlay approach suffers from a
In contrast, many researchers proposed using cognitive complex transmitter and receiver design and a very small data
radio (CR) technology to hop among the existing RF avail- rate while using one of the CR channels does not guarantee
able channels [11]. This assumption will increase the RF the channel availability for important applications such as
spectrum usage efficiency and will dodge the aforementioned car-to-car communications. On the other hand, the out-of-
limitation. Moreover, using cognitive radio is a license-free band category contains 1) buying a license for the common
application that reduces the car-to-car communication cost control channel from mobile network service providers,
compared to the other RF-licensed techniques. In addition, 2) using the industrial, scientific, and medical (ISM) radio
using cognitive radio improves the quality of service (QoS) bands such as the free 2.4 GHz band which coexists with
by sensing environmental and inadvertent man-made radio wi-fi and other interferences, and 3) using a non-RF approach
interference [12]. Therefore, cognitive radios can select fre- such as wired communication or visible light communication
quency channels with higher signal-to-noise ratios (SNR) to (VLC) [21]. In contrast, buying a license for a common
improve the utilization efficiency of the radio wave spectrum control channel adds a subscription cost for each car, which
and postpone the spectrum scarcity problem [13]. In CR, increases the overall system cost. Meanwhile, the ISM band
the primary users (PUs) are the ones who have the license is overcrowded and suffers from a very high noise level [22].
to use that frequency band, meanwhile, the secondary users
(SUs) are the ones who do not have the license to use that B. RESEARCH GAP, MOTIVATION, AND PROBLEM
frequency band [14]. However, SUs are allowed access to STATEMENT
the PUs licensed spectrum bands opportunistically on the Therefore, aiming to achieve more spectral efficiency, opti-
condition of not causing interference to PUs [15]. To min- cal wireless networks can be used in incorporation with
imize the interference, the network of cognitive radio SUs CR [23]. Consequently, the network throughput of the hybrid
access the spectrum through overlay, underlay, or interweave cooperative CR RF/VLC is way more enhanced. To fur-
based on the sensed PUs network information and the type of ther increase the throughput, the hybrid system may adopt
regulatory available constraints [16]. using the multiple-input multiple-output (MIMO) in the net-
The main two factors that decide good CR design are the work [24]. Wireless communication in the band of visible
probability of miss-detection (PMD) for the primary user sig- light (400 THz to 800 THz) has been proposed to compensate
nal and the probability of false alarm (PFA) for free channels. for the spectrum scarcity problem in RF band communication
The effect of PMD is more severe as it harms PUs and the systems that operate at (3 kHz to 300 GHz) [25]. In 1880,
system might not be approved to work if this factor is above Bell and Tainter proposed the photo phone (i.e., the idea of
a certain threshold. The PFA affects degradation in the CR transmitting a signal using light) [26]. However, the proposed
system’s overall throughput and delay. These two factors are photo phone idea used a fluorescent lamp as a light source
heavily affected by the fading effect [17]. in the system implementation to transmit data [27]. After
In fading the signals from multipath aggregate to form that, the fluorescent lamp was replaced by light-emitting
constructive or distractive interferences. The distractive inter- diodes (LEDs) for fast switching and to produce a higher
ference causes degradation in the PUs signal which increases data rate [28]. Keio University’s researchers were able to use
the PMD. While the constructive interference amplifies the the same LEDs for both communication and illumination.
background noise sparks which increases the PFA [18]. The However, this experiment was in an indoor visible light com-
academia proposed using cooperation between nodes to over- munication (VLC) environment using multiple light sources
come these fading effects by taking advantage of the fading and detectors to form small-cell access points (APs) within
uniqueness effect at each node [19]. Therefore, if 10% of the wireless network. VLC is easier to prove operable in an
89340 VOLUME 12, 2024
N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

indoor environment as there are fewer interference sources (FPS) achieves data rate in the range from several bits per
(e.g., sun, moon, traffic lights, etc.) and fewer weather condi- second (bps) to several kilobits per second (kbps) [38]. This
tions (e.g., rain, fog, etc.) [29]. problem might be overcome by using high-speed sampling
Furthermore, the 2011 IEEE 802.15 working group for rate cameras of more than 1000 FPS [32], [39]. On the
Wireless Personal Area Networks published the first IEEE other hand, OCC has more advantages such as high SNR,
standard about VLC (802.15.7-2011). That IEEE standard low interference, high stability, and high security concerning
discussed the framework of VLC communication regarding non-fixed communication link distances, these characteristics
the dimming mechanisms, modulation schemes, and data make it one of the popular solutions for long-distance LOS
rates [30]. After that, the IEEE 802.15 working group pro- communication links.
posed STD 802.15.7-2018 to broaden the previous STD Our previous research used the VLC as the main car-to-
802.15.7-2011. The 2018 version includes categorizing the car communication technique either, in urban areas using
PHY layers into six categories to enable many more tech- mobile networks (e.g., 4G or 5G) as a backup communication
nologies in Optical wireless communications (OWC) [31]. method or on highway roads using low earth orbit satellite
Moreover, the 2018 version describes Optical Camera Com- constellations (e.g., Starlink, or OneWeb) as a backup while
munication (OCC) as the IEEE’s first standard to use OWC there is no mobile coverage [26], [40]. In both scenarios, VLC
in personal area networks (PANs) as the suffix well-known communications have great theoretical results which promise
as (OWPANs) [32]. Additionally, IoT applications tend to a great future for VLC communications.
take advantage of OCC technology benefits based on the However, we have conducted a practical experiment to
VLC approach beyond 5 G key technologies [33]. It also implement these laboratory simulations. By comparing the
takes advantage of the available infrastructures (i.e., LED results from both the lab experiment and implementation
lights in traffic lights, car headlights and taillights, parking using practical car parts (i.e. camera and LED), our practical
cameras, and road surveillance cameras) which are consid- experiment shows a huge gap between what we expect to have
ered the best compatible with several applications /scenarios. and what we have achieved [40]. The maximum achievable
Consequently, safety-related traffic information (e.g., direc- data rate in the practical experiment was 200 kbit/s. While in
tion and speed-related information, accident notifications, the lab the data rate could reach 355 Mbit/s in the pure VLC
and braking performance) can be exchanged using vehicle channel communication which is about three times the pure
communication through VLC [34]. The camera detector used RF channel communication (i.e., 108 Mbit/s). The difference
in OCC has a low frame rate that needs special modulation is that the lab uses some state-of-the-art cameras and LED
schemes to avoid flickering illumination [35]. OCC technol- sources. Consequently, we propose a plan for the VLC devel-
ogy is described within the IEEE 2018 standard of 802.15.7 opment map that starts with the practical setup at hand. Then
[36]. The IEEE 2018 standard of 802.15.7 physical layer after a decade of research and development, the car manu-
(PHY) for OCC contains three categories within it such as: facturer shall be convinced to implement these lab-advanced
parts in new cars for high-speed VLC communication.
• Category 4: For applications in outdoor areas and
Therefore, this paper proposes using the VLC as a common
support high mobility. Consequently, car-to-car and
control channel for the cognitive radio network in car-to-car
car-to-infrastructure communication applications can be
applications, because choosing the common control channel
used under this category.
in the RF makes the problem worse. Although VLC has
• Category 5: Unlike Category 4, this category is used
great potential to be the main method for transmitting data
for indoor applications with small distances. Commer-
between vehicles with huge bandwidth and higher data rates,
cial rolling shutter cameras used in smartphones might
the development of VLC as a means of transferring data
use this category for Li-Fi and similar communication
between cars is still in the early stages. The VLC communi-
applications.
cation transmission data rate depends on the properties of the
• Category 6: This category concerns communication
installed transmitting LED and the receiving camera. With the
through screens. It can be applied to smartphones,
currently installed equipment, cameras and LED lights could
tablets, and TV screens.
reach up to approximately 200 Kbit/sec. On the other hand,
Smart devices especially mobile phones have dramatically road safety cannot wait until the VLC technology reaches its
increased the usage of advanced cameras. OCC is one system full capability.
that uses smart device cameras rather than photodetectors to Consequently, motivated to achieve the highest data trans-
receive data [37]. The usage of OCC is open to making sev- fer rate between cars, we will use an RF cognitive radio
eral challenges in different applications (e.g., digital signage, with a backup common control channel using VLC. Our
mobile robot (MR) communications, vehicle-to-everything metrics include data rate, delay, outage probability, cost,
(V2X), smartphone positioning, augmented reality (AR), and scalability, adaptability, and bit error rate. The comparison
localized advertising). The OCC has a limited data rate analyses shall compare VLC versus the other common control
because the available commercial cameras have a low sam- channel types. We focus on comparing practical systems, not
pling rate. Particularly, using cameras of 30 frames per second theoretical or laboratory-achievable systems.

VOLUME 12, 2024 89341

 N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

C. KEY CONTRIBUTIONS II. RELATED WORK

• We propose building a car-to-car communication system Many surveys and published studies discussed the VLC sys-
that utilizes CR transmission with VLC as the CR’s com- tems’ limitations, advantages, and applications. The authors
mon control channel. In this proposal, we take advantage of [41] have introduced VLC as a 5G candidate commu-
of their license-free nature to reduce the car-to-car com- nication system and demonstrated the VLC advantages and
munication system running cost. limitations. The main advantage they found is the VLC can
• We target using the current car LED lights and sen- achieve a high data rate and avoid the spectrum sacristy
sor’s camera as the VLC communication hardware to problem of RF bands. However, the main limitations are
reduce the communication system cost. However, as our the needed complex pieces of equipment to be added to
previous practical experiment mentioned, the practically mobile phones and vehicles such as high frame per second
achievable data rate is below expectations and still needs cameras and MIMO LED lamps or screens. The authors
improvement to reach the theoretical levels. Mainly, due of [42] have been introduced using the automotive industry
to a lack of hardware specs. Therefore, we aim to use the LED headlamps in VLC and the effect of that on the VLC
car-to-car VLC communication technology to encour- data rate achieved in V2V applications and the utilization
age the industry to invest in upgrading VLC hardware limitations. They found that the possible ways to increase the
components. data rate using the automotive industry LED headlamps fall
• Choosing the VLC communication as the CR’s common into two categories 1) increasing the number of independent
control channel solves the CR bottleneck dilemma of ON/OFF LEDs in each headlamp and 2) changing the illu-
choosing a licensed CCC or unreliable CCC for cogni- mination level of each independent LED to form ASK levels.
tive radio common control channel communications. Therefore, if 2 headlamps where each headlamp containing
• We proposed dividing each of the car LED lights into 16 independent ON/OFF LEDs with 8 illumination levels,
segments. Each segment transmits its data using OOK then each frame can receive 96 bits. For 100 FPS cameras,
modulation separately and independently to form multi- they prove that the system can communicate at the data rate
inputs. On the other hand, camera pixels form multi of 9600 bit/s.
photodetector (i.e., multi-outputs). Therefore, the pro- In [2], the authors present a comprehensive survey of
posed system forms a MIMO to increase the data rate, hybrid RF/VLC communication systems. The author pro-
improve the bit error rate, and decrease the outage posed a comparison between RF and VLC health safety,
probability. spectrum availability, energy consumption, coverage and
• To validate the proposed car-to-car communication sys- mobility, and interference. They find that Visible light offers
tem’s robustness, we present the study of the proposed higher health safety than RF. Moreover, the spectrum avail-
system under different weather conditions. The sim- ability of visible light is wider than RF. In addition, from
ulation checks the resultant bit error rate and outage a security point of view, visible light has more immunity
probability under each weather condition. The detailed to interception and jamming compared to RF. Additionally,
results are in the results section. the energy consumption of RF is much greater than visible
light. While RF has higher EMI interference than visible
D. KEY LIMITATIONS light. On the other hand, the main advantage of RF is that
• Concerning cost, the system might face increased costs RF has more coverage and offers greater mobility than visible
if implemented with high lab specs. Such as the cost of light.
the complex pieces of equipment to be added to vehicles In [4], the authors propose using a CR-based ad-hoc
for example high FPS cameras and MIMO LED lamps. vehicle network (VANET) or CR-VANET. The CR-VANET
• Concerning communication distance, VLC has a limited is a highly technologically sophisticated network that aims
distance (point-to-point communication) of approxi- to compensate for the deficiency in VANET spectra.
mately 100m which makes it limited in most applica- CR-VANET has some problems in sensing relative to other
tions. However, the distance can be improved through CR-based networks, including the effects of high-speed
multiple hops relaying. In contrast, by increasing the mobility and vehicle routes, the heterogeneous quality of
number of hops, the noise accumulation might lead to service (QoS) specifications, security issues, shadowing, and
data loss. multi-path fading problems. The authors proposed using
• Concerning delay, increasing the communication dis- machine learning, CSS, new PU activity modeling, the con-
tance through multiple hops leads to a total huge delay struction of faster convergent algorithms, and beamforming
due to the needed processing time by each repeater in the antennas as solutions to these problems. However, there is
chain. Therefore, the number of hops must be limited still a great deal of potential for research and study into spec-
to decrease delay and prevent loss of data.to minimize trum sensing challenges in CR-VANET. In [11], the authors
the processing time in each node, we proposed using an introduced simultaneous light wave information and power
amplifier forward to decrease the delay in a tradeoff with transfer (SLIPT). The authors proposed using both the VLC
the received accumulated noise. and the CR simultaneously to mitigate the outage probability

89342 VOLUME 12, 2024

N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

TABLE 1. MATLAB model simulation parameters.

in both systems. Therefore, if the VLC channel is blocked III. SYSTEM MODEL
due to weather conditions or non-line-of-sight scenario con- We propose a framework of a two-way road with 60 cars
straints then the CR channel prevents the outage. On the other within six lanes (i.e., three lanes in each direction) As shown
hand, if all RF channels are occupied with PUs, then the in Figure 1. A blue arrow represents a single-hop or multi-hop
VLC channel prevents the outage. Moreover, a power transfer VLC, while purple lightning represents overlay, green arrows
is proposed through the VLC channel. Then, whenever the represent underlay, and yellow arrows represent licensed RF
device passes through a VLC access point such as street that communicates with the NSP access point mounted on
lights or commercials, it charges its battery and receives data traffic lights. When the VLC distance is greater than 100m
simultaneously. However, the paper only assumes communi- the communication is considered blocked as shown with a
cation to stationary devices with no mobility probability is red cross in Figure 1. Meanwhile, if there is a car in between
added. Therefore, their assumptions are not valid for car-to- it communicates through multi-hop VLC with the middle car
car communication applications. considered as a repeater as shown with the two consecutive
The authors of [43] have introduced a survey on systems blue arrows. In addition, overlay cognitive radio mostly hap-
that use wireless optical communication, their applications, pened away from the access point as shown with the purple
and network architectures. Wireless optical communication is lighting. The green arrow that represents the underlay is not
either used in an outdoor environment or an indoor environ- restricted to certain conditions. A MATLAB simulation of the
ment. The indoor applications are much easier to implement car-to-car communication system under this framework will
and operate and have a simpler network architecture. For the be used and modeled according to the following parameters.
downlink, the lighting system of an indoor home or mall is The road simulated distance is chosen as 10 km. Variations
used to transmit the needed data while the selfie cameras in the speed of each car are assumed to be from 40 to 80 km.
of the mobile phones receive and decode the transmitted Mount Carlo simulations are used to minimize the random-
messages. The uplink uses LED illuminance levels of mobile ness in the results and to increase the result’s robustness.
phone screens to transmit the needed data to the surveil- Each car is assumed to transmit a data set of 64 kbit to
lance cameras which decode the reply messages. On the its surrounding cars through the proposed communication
other hand, outdoor applications use commercial screens system to measure the algorithm’s metrics. Our MATLAB
and traffic lights for the downlink which are received by simulation parameters are summarized in Table 1.
vehicles parking cameras and mobile phone selfie cameras. In our simulation, the road has access points along its
The uplink uses LED illuminance levels of mobile phone way either mounted on lights, road signs, or commercial
screens or the vehicle’s headlamps and taillamps to trans- boards. The access points communicate with the on-road cars
mit the needed data to the road surveillance cameras which to deliver vital road instructions (e.g., speed limits, closed
decode the reply messages. The authors of [44] have intro- lanes ahead, congestion, etc.). However, these instructions
duced the communication system details based on VLC. only reach a portion of cars that are nearest to the access
They propose using LED lamps or LASER sources as VLC point with minimum transmitted power to be able to reuse
communication transmitters and photodetectors or cameras the frequency spectrum and save energy. All other cars (i.e.,
as VLC communication receivers. Moreover, they introduced out of the limited coverage range of the access points) on
possible applications such as indoor localization, vehicle-to- the road depend on hopping this information among them
vehicle (V2V) communication, and VLC sensing. In each through the car-to-car communication means. In addition,
application, they provide the advantages, possible challenges, each car transmits additional information to the following
future research trends, and optimization techniques used to cars regarding its intentions and directions on the road to
maximize the data rate. They concluded that most of the avoid crashes.
current research concerns the theoretical approach for VLC As aforementioned, we assume the main car-to-car com-
and that the practical experiment results are still in the munication uses RF cognitive radio communications. The RF
early stages of achieving the required data rate and outage frequency band of 70 GHz to 90 GHz is proposed because
probability. of its currently limited applications and huge bandwidth.

VOLUME 12, 2024 89343

 N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

FIGURE 1. The proposed framework of 60 cars communicates through VLC/RF within 6 lanes road (3 lanes in each direction).

However, this frequency band suffers from high attenua- the use of licensed bands guarantees a better performance
tion that limits its communication distance to a few tens of in terms of outage probability, SNR, and BER as the ser-
meters. This is not a problem in car-to-car communication vice provider’s role is to ensure these performance metrics.
because the required communication distance between cars However, the licensed bands cannot be used by all cars
is within that range. While the CR common control channel due to the spectrum scarcity problem and a portion of cars
is compared between the VLC, overlay-paid, overlay-ISM, must use other communication techniques. That portion is
and underlay. The traffic types on the common control chan- going to increase as more and more cars adopt car-to-car
nel contain the following instructions 1) spectrum sensing communication.
results, 2) cooperation packets among secondary users (i.e.,
cars), 3) request to transfer from the sender to the destination, B. USING UNLICENSED OVERLAY AS CCC OF CR
4) broadcast to allow the pool of secondary users to know that CAR-TO-CAR COMMUNICATION
this channel is reserved, 5) the acknowledgment packet from In RF communication in the ISM band, an omnidirectional
the destination to the sender, and 6) create and update the rout- antenna is used to transmit an electromagnetic wave in
ing map among the secondary users. These traffic types can the frequency band from 70 to 90 GHz with a bandwidth
be summarized into broadcast and unicast (i.e., point-to-point of 2 GHz. Consequently, that huge bandwidth might have suf-
full duplex). For comparison purposes, we assume alterna- ficient channels for car-to-car communication in the near and
tive common control channels such as 1) licensed overlay, medium future. However, this band has higher attenuation
2) unlicensed overlay in the ISM band, 3) underlay, and compared to other bands because the attenuation in free space
4) using VLC. In what follows, we will discuss each CCC is proportional to the carrier frequency and that frequency
alternative for exploring their advantages and drawbacks is more likely to suffer from gas attenuation peaks such as
theoretically. oxygen O2 and water vapor H2 O. This high level of attenua-
tion in such a band decreases the communication distance and
A. USING LICENSED OVERLAY AS CCC OF CR CAR-TO-CAR limits the number of applications (car-to-car communication
COMMUNICATION is one of them). The channel in RF communication is less
In RF communication in the licensed bands, each car buys affected by the air’s attenuation due to weather conditions
a license to use that band in car-to-car communication. That in comparison to the VLC channel. In contrast, attenua-
license might be a subscription (i.e., the amount paid every tion in the RF channel is affected by concretes or metals.
period even if not used by that band) or data-centric (i.e., On the other hand, The RF omnidirectional antenna receives
the amount paid per packet sent through that band). In this electromagnetic communication waves in both non-line-of-
paper and for comparison purposes, we adopt the data-centric sight and line-of-sight. This paper proposes a modulation
license approach from the service providers. Therefore, the of 1024 QAM because if the modulation increases beyond
overall system cost is the initial telecommunication sys- that limit the SNR will decrease dramatically which leads to
tem cost and the data transmission cost compared to other decreased communication distance. The proposed transmis-
approaches with no running costs at all. On the other hand, sion range of RF is less than 200 m to avoid high attenuation

89344 VOLUME 12, 2024

N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

levels because transmission through RF from all cars with the 1) INTERFERENCE AND POWER CONSTRAINTS IN OVERLAY
chosen communication distance is considered interference AND UNDERLAY COGNITIVE RADIO
concerning other cars. That’s why we need to minimize the In underlay cognitive radio, the secondary users must operate
number of cars using RF communication within a given area under strict power constraints to avoid causing interference
and make them more dependent on other communication to primary users. This can affect the transmission rate R and
techniques. thus the transmission delay Ttrans . The effective transmission
rate Reff considering the power constraint can be expressed
C. USING UNDERLAY AS CCC OF CR CAR-TO-CAR as:
COMMUNICATION  
Ps
In RF communication in the underlay, the CR SUs spread Reff = B log2 1 + (4)
N 0B + I p
the transmitted data using the spread spectrum (SS) key
techniques. In such techniques, the transmitted data occu- where
pies more bandwidth but with low power (i.e., lower than
B is the bandwidth.
noise level) to not cause any harm to PUs. The SU receiver
Ps is the transmission power of the secondary user.
reconstructs that data message perfectly from the surrounding
N0 is the noise power spectral density.
noise by knowing the transmitted SS key and multiplying
Ip is the interference power from primary users.
the received message with that key. However, there is a
finite number of SS keys that can be used orthogonally
without interfering with each other. Beyond that limit more E. GENERAL DELAY EQUATION IN OVERLAY AND
and more interference levels are present and all communi- UNDERLAY COGNITIVE RADIO
cations might be dropped. Therefore, CR using underlay in The delay general equation for both overlay and underlay
CCC communications limits the maximum number of CR cognitive radio systems is discussed in this section several
networks operating in the same area. However, underlay in parameters affect the delay influenced on secondary users.
CCC ensures no damage to PUs and its outage probability These parameters include the dynamic nature of spectrum
is not an opportunistic manner but rather is a function of the availability, transmission power constraints, and interference
number of current cars using the CR network. from primary users. Therefore, we calculate the following
parameters.
D. GENERAL BER EQUATION IN OVERLAY AND
UNDERLAY COGNITIVE RADIO 1) TRANSMISSION DELAY
The bit error rate general equation for both overlay and The time needed to send a packet from the transmitter to the
underlay cognitive radio systems is discussed in this section receiver is called the transmission delay. It can be expressed
This equation calculates the effective SNR that is affected by as:
both noise and interference. Combining the effective SNR L
with the modulation-specific BER, we get the general BER T trans = (5)
R
for a cognitive radio system with M-QAM, for instance, is:
s  where:
L is the packet length in bits.
 
1 SNReff
BERCR ≈ 2 1 − √ Q 3 ·  (1) R is the transmission rate in bits per second (bps).
M M −1

and: 2) QUEUEING DELAY

Ps The time a packet spends waiting in the queue before it is
SNROVL_eff = (2) transmitted is called Queueing delay. It is affected by the
N 0B + I
service rate and the traffic loads of the system. For a Poisson
Ps
SNRUL_eff = (3) arrival process system with service times exponentially dis-
N 0B + I + I P tributed (M/M/1 queue), the average queueing delay Tqueue is
where: calculated by:
SNRovl_eff is the overlay signal to noise ratio. ρ
T queue = (6)
SNRul_eff is the underlay signal to noise ratio. µ (1 − ρ)
B is the bandwidth.
Ps is the transmission power of the secondary where:
user. ρ is the traffic intensity, defined as ρ = µλ .
N0 is the noise power spectral density. λ is the average packet arrival rate.
Ip is the interference power from primary users. µ is the service rate, which is the reciprocal of the
I is the interference from other secondary users. average service time.

VOLUME 12, 2024 89345

 N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

3) PROPAGATION DELAY 2) THROUGHPUT CALCULATION

The time it takes for a signal to propagate from the transmitter As the data is successfully transmitted and received it is called
to the receiver is called Propagation delay and It is calculated Throughput T . It can be expressed as the product of the
by: effective transmission rate and the packet success probability:
d T = Reff × P success (12)
T prop = (7)
v
Substituting the effective transmission rate and packet suc-
where:
cess probability:
d is the distance between the transmitter and receiver.
Ps
v is the propagation speed of the signal (approximately T = Blog2 (1 + ) × (1−BER)L (13)
the speed of light for wireless communication). N 0B + I p

3) GENERAL THROUGHPUT EQUATION

4) PROCESSING DELAY The general throughput equation for an overlay and underlay
The time it takes for a node to make routing decisions and cognitive radio network can be written by multiplying the
process a packet header is called Processing delay. It is usu- above two components
ally constant as T proc . s !!L
Ps Ps
5) TOTAL DELAY T = Blog2 (1+ )× 1 − Q 2·
N 0 B+I p N 0B + I p
The total delay Ttotal occurred by a secondary user in an over-
lay and underlay cognitive radio network can be calculated (14)
by summing all the individual delay components:
IV. USING VLC AS CCC OF CR CAR-TO-CAR
T total = T trans + T queue + T prop + T proc (8) COMMUNICATION
In VLC communication applications such as car-to-car,
Substituting T trans with the effective transmission rate:
we use existing equipment to minimize the needed cost.
L L Therefore, this paper proposes taking advantage of modern
T trans = =   (9)
Reff Blog2 1 + Ps cars’ built-in LED lights as VLC transmitters and the parking
N 0 B+I p
sensor cameras as the receiver. In VLC communication, the
Thus, the total delay equation becomes: transmission is carried by laser or LED (i.e., visible light
L photon source). In each car, the headlight LEDs are divided
T total =
into a grid of independent transmitters in a 3 by 3 with a total
Blog2 1 + PsN 0B + I p
number of 9 LEDs. Each independent LED sends its data
ρ d
+ + + T proc (10) separately from the others’ transmission. The propagation
µ (1 − ρ) v channel carrying the information data is the air either indoors
or outdoors. The weather conditions (e.g., fog, sunlight, night
F. GENERAL THROUGHPUT EQUATION IN OVERLAY AND
stars, sandstorms, or rain) affect only outdoor communication
UNDERLAY COGNITIVE RADIO
such as in this car-to-car application. On the other hand, the
The throughput is essentially the rate at which data is success-
reception is carried by a camera or generally a photodetec-
fully transmitted from the sender to the receiver. It depends
tor (i.e., visible light photon absorber). The receiver camera
on some factors as power constraints, the available spectrum,
captures all of the LEDs’ illumination in each frame at once
and interference from primary users.
and then processes the received data. The LED ON/OFF
The general equation for throughput in an overlay and
sampling time must be equal to the camera FPS and syn-
underlay cognitive radio network is calculated as follows.
chronous with it. Commercial cameras capture from 30 FPS
up to 2000 FPS. Therefore, VLC is optimum compared to all
1) EFFECTIVE TRANSMISSION RATE (CAPACITY)
other aforementioned RF approaches to cost, effect on PUs,
The effective transmission rate for a secondary user in the outage probability, and BER. However, the VLC CCC must
existence of interference from primary users can be calculated include all VLC car-to-car links in the routing map generation
using the Shannon-Hartley theorem: that require alteration periodically due to the nature of the car-
Ps to-car application.
Reff = Blog2 (1+ ) (11)
N 0B + I p
A. LED APPLICATIONS AND CHARACTERISTICS
where: Two-lead semiconductors of a special type of PN junction
B is the bandwidth available for the secondary user. diode are used to form the light source which is commonly
Ps is the transmission power of the secondary user. named the light-emitting diode (LED) within an area of less
N0 is the noise power spectral density. than 1 mm2. Choosing the doping materials in the semicon-
Ip is the interference power from the primary users. ductors in both regions the P-type and the N-type changes

89346 VOLUME 12, 2024

N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

the characteristics of the LED’s illuminations as shown in TABLE 2. LED operation band versus the used doping material.
Table 2.
On the other hand, white light is generated by merging
LEDs of red, green, and blue in a small area while main-
taining the same intensity. However, to produce different
sensations of color, intensity quantities of the red, green,
and blue LEDs are changed. Each LED produces luminous
flux [lm] for a measured electrical power [W], and then The
LEDs’ luminous efficacy [lm/W] is the ratio between both of
them. The efficacy of the blue color is the lowest at 75 lm/W scattering and absorption in the channel medium (i.e., air) at
while the efficacy of the red color is the highest at 155 lm/W. various levels of visibility H_FOG. Since this paper focuses
The LED’s main advantages are listed as the following: on vehicle-to-vehicle communication in short-range, when
1) The LEDs have low cost and small area, and 2) the LED the fog visibility (V) is less than 0.5 km, Beer’s law can be as
intensity is controlled by microcontrollers and the electric- simple as
ity level. Moreover, LEDs are 3) Energy efficient, 4) Long  
Lifetimes, 5) Rugged, and 6) do not require a period to − 3.91
V
HFog = e (15)
warm up. While the LED’s main disadvantages are listed as
1) Temperature dependence and sensitivity, and 2) Electrical where V in kilometers is the visibility meteorologically.
polarity and Light quality. In addition, 3) Efficiency drops In addition, the coefficient of the free space channel HLOS
over time, and 4) Voltage sensitivity. Then the LED main at line-of-sight communication follows the general formula
applications are listed as the following: 1) in smartphones, of loss in free space:
2) using LED bulbs in industries and at homes, 3) as signals A cos (θ) cos (ϕ)
for the traffic light, and 4) in motorcycles and cars’ headlights HLOS = (16)
2π D2
and taillights.
where A in illuminance is the amplitude of light, ϕ and θ are
B. SPECIFICATION OF THE LED IN CARS’ HEADLIGHTS the receiver and transmitter angles from the centerline of the
The VLC system LED source in the car-to-car is a white light line-of-sight communication, and D is the receiver towards
with the following specifications. The bandwidth emission of the transmitter distance.
the visible light is from 400 to 700 THz then the white LED On the other hand, this paper adopts the Palmer and
covers all that band. The headlight of each car is composed of Marshal channel coefficient for rain attenuation HRain . Rain
independent LEDs in a grid of 3 by 3 with a total of 9 lights attenuation depends on rain size, transmission frequency, and
to form multiple transmission sources each of them indepen- temperature [5]. In the formula of Palmer and Marshal, the
dently uses ON OFF key (OOK) modulation to transmit its rain rates (RR) cause the attenuation in wireless optical com-
share of data. munication. Where rain rate is measured by meter/hour and γ
On the other hand, each shot of the receiver camera records in dB/km while the power-law parameters are B = 0.63 and
the reading of all LED illumination levels from the transmit- A = 0.365.
ted car headlight. Cameras’ shooting speed varies from as γ = A × RRB (17)
high as 2000 FPS for advanced cameras to as low as 30 FPS 4λ γL
for commercial ones. Consequently, we increase the data rate HRain = × 10− 10 (18)
π × hc
by choosing a camera with a higher FPS but suffers from
higher system costs. In addition, each captured frame detects where the wavelength is λ, the constant of Planck is h, and
the illumination of the 9 LEDs at once. To distinguish zeros the speed of light is c in free space. Note that we assume
from ones the received frame is compared to an illumination a spherical shape raindrop to make the Marshal and Palmer
threshold. assumption independent of polarization.
YRx = HLOS HRain HFog XTx + NAWGN (19)
C. VLC CHANNEL CHARACTERISTICS
VLC’s noise in the communication channel is usually mod- The camera captures the received light signal after passing
eled as additive white Gaussian (AWGN) with a certain by all the channel coefficients aforementioned (i.e., multiply
attenuation coefficient that depends on the distance between the transmitted modulated light by them) and then adding the
the transmitter and the receiver. The attenuation coefficient additive white Gaussian noise to the OOK signal as shown in
varies by different channel weather situations such as rain, FIGURE 2, then:
sand storms, or fog. Moreover, VLC communication is
affected by interferences from sunlight, commercial boards, D. SCALABILITY AND ADAPTABILITY
and street lights. Cars communicate with infrastructures such as road signs and
To determine the channel coefficient in foggy weather, this commercials as data sources. Then, distribute the received
paper uses Beer’s law [12]. Beer’s law describes the light information through car-to-car communications through the

VOLUME 12, 2024 89347

 N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

FIGURE 2. The Characteristics of the VLC Channel under various

conditions of weather such as fog, direct sunlight, rain, or snow.

road for cars non-neighboring to infrastructures. Higher scal-

ability for car-to-car communication leads to the need for
less infrastructure on the road which reduces the overall
system cost. Therefore, this paper adopts high scalability by FIGURE 3. The CCC data rate versus the number of cars in the CR network
prolonging the VLC communication distance through multi- for VLC, overlay, and underlay.
hop. The ability of the VLC system to withstand noise and
delay threshold is the main limit to prolong the multi-hop the road is discussed. For the VLC CCC, the line-of-sight
communication. Consequently, this paper proposal aims to communication coverage on the road is dependent on the
provide higher hop counts within the same noise and delay availability of cars to hop among them from the source to
level to increase our system scalability. the destination. Therefore, having a small number of nodes
The proposed system aims to adapt to the currently man- gives higher outage probability results (i.e., near 1 for less
ufactured car parts such as parking cameras and headlight than 10 cars). However, the VLC system operates with a
LEDs. Therefore, no need to install expensive parts to use the small outage probability (i.e., 0.1) and no visible problem
car-to-car communication feature. Adding higher complex at a higher number of nodes. In contrast, the RF CCC is
parts is backward compatible with the current parts. However, affected by the noise created by the adjacent nodes. Then, the
these complex parts achieve higher data rates when commu- outage probability of the RF at a lower number of nodes (i.e.,
nicating with each other. 0.1 for less than 5 cars) is lower than the outage probability
as the node number increases (i.e., 0.9 for more than 40 cars).
V. RESULTS AND DISCUSSION Moreover, the overlay RF shows a higher outage probabil-
A. THE CCC DATA RATE VERSUS THE NUMBER OF CARS
As shown in FIGURE 3,, the common control channel aver- ity than the underlay outage probability as the competition
age data rate versus the number of nodes (cars) within the road for the availability of secondary users’ slots increases with
is discussed. For the VLC CCC, the line-of-sight communica- no available slots (e.g., 4 slots available for 15 cars). The
tion coverage on the road is dependent on the availability of licensed type is not included in the figure as their outage
cars to hop among them from the source to the destination. probability is controlled by the service provider to be constant
Therefore, having a small number of nodes does not give at a low level.
the best data rate results (i.e., 1.2 kbps). This is because the
C. THE CCC BER VERSUS FADING EFFECT
average VLC communication distance is around 70 m at a low
number of nodes. However, the VLC system operates with As the fading effect increases the destructive interference
a high data rate (i.e., 1.8 kbps) and no visible problem at a leads to increased attenuation in dBs in the received signals.
higher number of nodes. This is because the average VLC This paper monitors the fading effect in the received signals.
communication distance is around 40 m at a high number In FIGURE 5, the common control channel average BER
of nodes. In contrast, the RF CCC is affected by the noise versus fading effect is discussed. For the VLC CCC, the BER
created by the adjacent nodes. Then, the data rate of the RF shows no visible effect as the fading effect increases because
at a lower number of nodes is higher (i.e., 2.5 kbps) than the there is little to no multipath in VLC. Therefore, VLC shows a
data rate as the node number increases (i.e., 1.5 kbps). This near-constant value as the fading increase (i.e., in the order of
is because, at a low number of nodes, the average overlay RF 10−6 ). In contrast, the RF CCCs are affected by fading effect.
communication noise level is around −110 dB, and at a high As the fading increases, the BER increases for the overlay and
number of nodes, the average overlay RF communication the underlay channels (i.e., from 5∗ 10−6 to 10−4 ).
noise level is around −70 dB due to the high ISI level.
Moreover, the overlay RF shows a higher data rate compared D. THE CCC DATA RATE VERSUS FADING EFFECT
to the underlay data rate by 25%. The licensed type is not As aforementioned in the previous result, the BER increases
included in the figure as their data rate is controlled by the with the fading effect increases. Therefore, the received data
service provider to be constant at a high level. suffers from data loss due to the dropping of the faulty packets
which decreases the effective data rate received. For the
B. THE CCC OUTAGE PROBABILITY VERSUS THE NUMBER VLC CCC, the data rate is not affected by the fading effect,
OF CARS therefore, it seems approximately constant (i.e., 2 kbps) as
As shown in FIGURE 4, the common control channel average shown in Figure 6. Because there is little to no multipath in
outage probability versus the number of nodes (cars) within VLC. In contrast, the RF CCCs are affected by fading effect.

89348 VOLUME 12, 2024

N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

FIGURE 4. The CCC outage probability versus the number of cars in the CR
network for VLC, overlay, and underlay.
FIGURE 6. The CCC data rate versus fading effect in the CR network for
VLC, overlay, and underlay.

FIGURE 5. The CCC BER versus fading effect in the CR network for VLC,
overlay, and underlay.
FIGURE 7. VLC CCC adaptability (Data rate versus number of LEDs for
different camera resolutions.
As the fading increases, the data rate decreases for the overlay
and the underlay channels (i.e., it decays from 2.5 kbps for
overlay and from 1.5 kbps for underlay CR). and the background noise greatly affect the recorded illumi-
nation in each pixel. Consequently, BER ranges from 6×10−6
E. THE VLC CCC ADAPTABILITY at a distance of 10 m and increases by 50% to 9×10−6 at
The relation between data rate and the number of LEDs for a distance of 70 m. For CR underlying CCC as the distance
VLC adaptability is discussed in Figure 7. For low resolution, increases multiple cars transmitting at the same underlay code
the data rate is small and seems constant with an increasing interfere with each other. such interference leads to increasing
number of LEDs. This is because the low-resolution camera BER from 3×10−6 to 5×10−6 for the same distance range.
can see a limited number of LEDs and any additional LED In contrast, the overlay CCC has BER 9×10−6 at a distance
will not contribute to the VLC communication. In medium- of 10 m and 12×10−6 at a distance of 70 m.
resolution cameras, as the number of LEDs increases the data
rate has increased to 2 kbps then seems to be constant with G. THE AVERAGE BER AT DIFFERENT WEATHER
an increasing number of LEDs beyond the camera resolution CONDITIONS
limits. While for high resolution, the data rate continues to As shown in FIGURE 9, the average BER versus the CCC
increase with increasing number of LEDs. of different types over different weather conditions are dis-
cussed. For clear weather, no environmental harsh conditions
F. THE AVERAGE BER VERSUS THE COMMUNICATION exist therefore, the minimum that can be achieved at clear
DISTANCE weather. consequently, BER is 4×10−6 for both underlying
As shown in FIGURE 8, the average BER versus the com- CCC and licensed one while it reaches 7×10−6 and 10×10−6
munication distance between cars is discussed. The BER for VLC and overlay CR respectively. In the case of heavy
increases as the distance increases in all cases except for fog, the humidity in the air leads to degradation in the
the licensed CCC type as it doesn’t depend on the distance received signal level for all RF communication techniques.
between cars and is controlled by the service provider. For the While the VLC suffers the most because the fog affects the
VLC CCC, as the distance increases between the transmitter light coming from the LED. Therefore, the BER is 4×10−6
car’s LED light source and the receiver car’s camera, the for licensed CCC and 5×10−6 for underlying one while it
received illumination decreases in both the power and the reaches 8.5×10−6 and 12.5×10−6 for VLC and overlay CR
number of covered pixels. Therefore, any interference source respectively. For heavy rain, all CCCs are affected by rain,

VOLUME 12, 2024 89349

 N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

FIGURE 8. The average BER versus the communication distance between FIGURE 10. The average delay for different CCC types.
cars.

FIGURE 9. The average BER at different weather conditions for CCC types. FIGURE 11. The cost of different CCC types including the initial
installation cost and the subscription fees.

while the overlay and VLC suffer the highest effect when TABLE 3. Rating the different types of CCC in comparison with each other.
they reach 7.5×10−6 and 10.5×10−6 respectively. BER is
6 × 10−6 for licensed CCC and 4.5 × 10−6 for underlying.

H. THE AVERAGE DELAY FOR DIFFERENT CCC TYPES

As shown in FIGURE 10, the average packet delay versus the
CCC of different types is discussed. The overlay CCC has
the highest delay of 10×10−4 sec compared to other types
because in overlay CR sensing cycle (i.e. sensing cooperation
decision adapt) requires a delay time before communication.
While VLC CCC has a delay of 7×10−4 sec, as it’s the aver-
age delay between single-hop and multi-hop VLC. In each
VLC repeater, the processing delay of receiving and then In Table 3, we present the highest average value for each
retransmitting data packets is added for each repeater in the metric (best or worst) as 100%. Then, we show the relativity
multi-hop chain. While the underlay and license have a delay of other CCC types with a comparison to it. For example,
of 4×10−4 and 5×10−4 respectively. Because they require no the licensed CCC type has the highest average data rate while
sensing phase and do not suffer from retransmission delay. the VLC CCC average data rate is 90% of the licensed CCC.
Therefore, the licensed CCC has the highest data rate and
I. COST AND OVERALL RATING cost while the overlay CR has the highest outage probability,
FIGURE 11 shows the cost of different CCC types including BER, and delay. The overall rating shows that the licensed
both the initial installation cost (i.e., fixed cost) and the CCC is the best choice followed by the VLC CCC while the
average subscription fees (i.e., variable cost). The VLC cost underlay is third place and the overlay is not suitable for this
includes the software only as the LEDs and cameras are application. However, given the future potential of the VLC,
considered included in cars. In both the overlay and underlay the 4% gap is expected to shrink over time.
CRs, the cost includes the purchase of telecom nodes and
attaching them to the cars. While in the licensed CCC, the cost VI. CONCLUSION
includes all the software, hardware, and subscription fees. We have demonstrated in this paper that the CCC is the
Therefore, VLC is considered the cheapest approach. bottleneck in the car-to-car CR network. Then, we explored

89350 VOLUME 12, 2024

N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

the different types of CCCs used to solve this bottleneck issue [13] U. Divakarala, K. Chandrasekaran, and K. H. K. Reddy, ‘‘A hierarchical
such as licensed overlay, overlay CR, underlay CR, and VLC. blockchain architecture for secure data sharing for vehicular networks,’’
Int. J. Inf. Technol., vol. 15, no. 3, pp. 1689–1697, Mar. 2023, doi:
A MATLAB simulation for a car-to-car framework has been 10.1007/s41870-023-01175-0.
built to demonstrate and compare these types of CCC through [14] F. Aghaei, H. B. Eldeeb, and M. Uysal, ‘‘A comparative evaluation
the chosen metrics. Our metrics included data rate, delay, of propagation characteristics of vehicular VLC and MMW channels,’’
IEEE Trans. Veh. Technol., vol. 73, no. 1, pp. 4–13, Jan. 2024, doi:
outage probability, cost, and bit error rate. The comparison 10.1109/TVT.2023.3302991.
analyses included comparing VLC versus the other CCC [15] S. Begum and N. B. Patil, ‘‘An intelligent vehicle control system for
types. Our results show that VLC can achieve up to 90% of enhancing road safety using optimal visible light communication net-
work,’’ J. Opt. Commun., vol. 44, no. 1, pp. 81–94, Jan. 2023, doi:
the licensed data rate with a small outage probability of 21.2% 10.1515/joc-2019-0049.
while maintaining moderate BER and delay. In addition, VLC [16] R.-R. Yin, X.-Y. Shen, P.-C. Zhao, and X.-Y. Ma, ‘‘Optimization of 2×2
presents the minimum cost, and as the introduced combined MIMO vehicular visible light communications model,’’ Acta Electonica
Sinica, Jan. 2023.
metric it placed second after the licensed type with a score [17] G. Singamsetty and S. Nallagonda, ‘‘Throughput performance analysis of
of 84.2%. With the VLC’s bright future of expansion and cooperative spectrum sensing network with improved energy detectors in
growing car-to-car application, the 4% gap is expected to Hoyt fading environment,’’ in Proc. 4th Int. Conf. Electron., Commun.
Aerosp. Technol. (ICECA), Coimbatore, India, Nov. 2020, pp. 720–725,
shrink over time. Therefore, we have proven that VLC is doi: 10.1109/ICECA49313.2020.9297530.
worthy of implementation practically in modern cars. [18] S. Nallagonda, A. Bhowmick, and B. Prasad, ‘‘On selection of param-
eters for cooperative spectrum sensing schemes over κ − −µ fading
channels,’’ IETE J. Res., vol. 69, no. 11, pp. 7684–7694, 2023, doi:
REFERENCES 10.1080/03772063.2022.203870019.
[1] M. Ahmed, M. A. Mirza, S. Raza, H. Ahmad, F. Xu, W. U. Khan, [19] S. Nallagonda, ‘‘Data fusion-aided cognitive radio network over gen-
Q. Lin, and Z. Han, ‘‘Vehicular communication network enabled CAV eralised fading channels,’’ Electron. Lett., vol. 55, no. 5, pp. 285–287,
data offloading: A review,’’ IEEE Trans. Intell. Transp. Syst., pp. 1–29, Mar. 2019, doi: 10.1049/el.2018.7727.
Aug. 2023, doi: 10.1109/TITS.2023.3263643. [20] G. Singh, A. Srivastava, V. A. Bohara, Z. Liu, N. A. Rahim, and
[2] H. Abuella, M. Elamassie, M. Uysal, Z. Xu, E. Serpedin, G. Ghatak, ‘‘Heterogeneous visible light and radio communication for
K. A. Qaraqe, and S. Ekin, ‘‘Hybrid RF/VLC systems: A comprehensive improving safety message dissemination at road intersection,’’ IEEE Trans.
survey on network topologies, performance analyses, applications, and Intell. Transp. Syst., vol. 23, no. 10, pp. 17607–17619, Oct. 2022, doi:
future directions,’’ IEEE Access, vol. 9, pp. 160402–160436, 2021, doi: 10.1109/TITS.2022.3156119.
10.1109/ACCESS.2021.3129154. [21] P. Kumar and A. Chakravarty, ‘‘Optimized resource management
[3] M. M. Céspedes, B. G. Guzmán, V. P. Gil Jiménez, and A. G. Armada, over connected vehicles in RF-CRN environment,’’ in Proc. 2nd
‘‘Aligning the light for vehicular visible light communications: High data Int. Conf. Adv. Technol. Intell. Control, Environ., Comput. Com-
rate and low-latency vehicular visible light communications implementing mun. Eng. (ICATIECE), Bangalore, India, Dec. 2022, pp. 1–5, doi:
blind interference alignment,’’ IEEE Veh. Technol. Mag., vol. 18, no. 1, 10.1109/ICATIECE56365.2022.10047516.
pp. 59–69, Mar. 2023, doi: 10.1109/MVT.2022.3228389. [22] M. A. Vieira, M. Vieira, P. Louro, and P. A. Vieira, ‘‘Cooperative vehicular
[4] M. A. Hossain, R. Noor, S. R. Azzuhri, M. R. Z’aba, I. Ahmedy, visible light communication in smarter split intersections,’’ Proc. SPIE,
K.-L. A. Yau, and C. Chembe, ‘‘Spectrum sensing challenges & their vol. 12139, pp. 28–29, May 2022, doi: 10.1117/12.2621069.
solutions in cognitive radio based vehicular networks,’’ Int. J. Commun. [23] E. M. H. Abouzohri and M. M. Abdallah, ‘‘Performance of hybrid
Syst., vol. 34, no. 7, p. e4748, Feb. 2021, doi: 10.1002/dac.4748. cognitive RF/VLC systems in vehicle-to-vehicle communications,’’
in Proc. IEEE ICIoT, Doha, Qatar, May 2020, pp. 429–434, doi:
[5] E. S. El-Mokadem, N. I. Tawfik, M. H. Aly, and W. S. El-Deeb,
10.1109/ICIoT48696.2020.9089610.
‘‘Design and performance evaluation of vehicular visible light communi-
[24] M. M. A. Momen, H. A. Fayed, M. H. Aly, N. E. Ismail, and A. Mokhtar,
cation system under different weather conditions and system parameters,’’
‘‘An efficient hybrid visible light communication/radio frequency sys-
Opto-Electron. Rev., vol. 31, no. 2, Nov. 2023, Art. no. 145580, doi:
tem for vehicular applications,’’ Opt. Quantum Electron., vol. 51, p. 364,
10.24425/opelre.2023.145580.
Oct. 2019, doi: 10.1007/s11082-019-2082-7.
[6] P. Gajewski, J. Łopatka, and P. Łubkowski, ‘‘Performance analysis of
[25] H. B. Eldeeb, S. Naser, L. Bariah, and S. Muhaidat, ‘‘Energy and spec-
public safety cognitive radio MANET for diversified traffic,’’ Sensors,
tral efficiency analysis for RIS-aided V2V-visible light communication,’’
vol. 22, no. 5, p. 1927, Mar. 2022, doi: 10.3390/s22051927.
IEEE Commun. Lett., vol. 27, no. 9, pp. 2373–2377, Jun. 2023, doi:
[7] S. Yahia, Y. Meraihi, S. Refas, A. B. Gabis, A. Ramdane-Cherif, and 10.1109/LCOMM.2023.3290025.
H. B. Eldeeb, ‘‘Performance study and analysis of MIMO visible light [26] J. Parikh and A. Basu, ‘‘Technologies assisting the paradigm shift from 4G
communication-based V2V systems,’’ Opt. Quantum Electron., vol. 54, to 5G,’’ Wireless Pers. Commun., vol. 112, pp. 481–502, Jan. 2020, doi:
no. 9, p. 575, Sep. 2022. 10.1109/gsmm.2017.7970330.
[8] Y. Zhang, W. Wang, X. Zhao, and J. Hou, ‘‘Performance analysis for mixed [27] H. Ouamna, Z. Madini, and Y. Zouine, ‘‘6G and V2X commu-
RF/VLC-based cognitive electric vehicle networks with imperfect CSI,’’ nications: Applications, features, and challenges,’’ in Proc. 8th Int.
Sensors, vol. 22, no. 5, pp. 481–488, Sep. 2022. Conf. Optim. Appl. (ICOA), Genoa, Italy, Oct. 2022, pp. 1–6, doi:
[9] E. Zadobrischi, ‘‘The concept regarding vehicular communications based 10.1109/ICOA55659.2022.9934407.
on visible light communication and the IoT,’’ Electronics, vol. 12, no. 6, [28] F. M. Alsalami, O. C. L. Haas, A. Al-Kinani, C.-X. Wang, Z. Ahmad, and
p. 1359, Mar. 2023, doi: 10.3390/electronics12061359. S. Rajbhandari, ‘‘Impact of dynamic traffic on vehicle-to-vehicle vis-
[10] Y. Yang, F. Zhang, Z. Zeng, J. Cheng, and C. Guo, ‘‘Visible ible light communication systems,’’ IEEE Syst. J., vol. 16, no. 3,
light communication for vehicular applications: A novel architecture pp. 3512–3521, Sep. 2022, doi: 10.1109/JSYST.2021.3100257.
with proof-of-concept prototype,’’ China Commun., Feb. 2023, doi: [29] N. M. Okasha, A. H. A. Zekry, and F. A. Newagy, ‘‘Hybrid VLC vehi-
10.23919/JCC.2023.00.021. cle to vehicle and LEO satellite communication system for highway
[11] Y. Xiao, P. D. Diamantoulakis, Z. Fang, L. Hao, Z. Ma, and road coverage,’’ in Proc. 6th Int. Tech. Conf. Adv. Comput., Con-
G. K. Karagiannidis, ‘‘Cooperative hybrid VLC/RF systems with SLIPT,’’ trol Ind. Eng. (CCIE 2021), Singapore, Jul. 2022, pp. 571–583, doi:
IEEE Trans. Commun., vol. 69, no. 4, pp. 2532–2545, Apr. 2021, doi: 10.1109/TAES.2021.3113880.
10.1109/TCOMM.2021.3051908. [30] A. Chaabna, A. Babouri, H. Chouabia, T. Hafsi, Z. E. Meguetta, and
[12] M. Karbalayghareh, F. Miramirkhani, H. B. Eldeeb, R. C. Kizilirmak, X. Zhang, ‘‘Experimental demonstration of V2V communication sys-
S. M. Sait, and M. Uysal, ‘‘Channel modelling and performance tem based on VLC technology for smart transportation,’’ in Artificial
limits of vehicular visible light communication systems,’’ IEEE Intelligence and Heuristics for Smart Energy Efficiency in Smart Cities:
Trans. Veh. Technol., vol. 69, no. 7, pp. 6891–6901, Jul. 2020, doi: Case Study Tipasa, Algeria. Cham, Switzerland: Springer, Nov. 2022,
10.1109/TVT.2020.2993294. pp. 653–661, doi: 10.1007/978-3-319-73192-6_5.

VOLUME 12, 2024 89351

 N. M. Okasha, F. A. Newagy: Car-to-Car Communication Using RF Cognitive Radio With VLC CCC

[31] S. Yahia, Y. Meraihi, A. Ramdane-Cherif, A. B. Gabis, and H. B. Eldeeb, NERMEEN M. OKASHA was born in Cairo,
‘‘Performance evaluation of vehicular visible light communication based Egypt, in 1986. She received the B.S. degree in
on angle-oriented receiver,’’ Comput. Commun., vol. 191, pp. 500–509, communication and electronic engineering from
Jul. 2022, doi: 10.1016/j.comcom.2022.05.025. the Modern Academy for Engineering and Tech-
[32] S. Mishra, R. Maheshwari, J. Grover, and V. Vaishnavi, ‘‘Investigating nology, Cairo, in 2008, and the M.S. degree in
the performance of a vehicular communication system based on visible electronics and communications engineering from
light communication (VLC),’’ Int. J. Inf. Technol., vol. 14, pp. 877–885, the Arab Academy for Science, Technology and
Jan. 2022, doi: 10.1007/s41870-020-00447-3.
Maritime Transport, Cairo, in 2016. She is cur-
[33] M. Meucci, M. Seminara, T. Nawaz, S. Caputo, L. Mucchi, and
rently pursuing the Ph.D. degree in communica-
J. Catani, ‘‘Bidirectional vehicle-to-vehicle communication system based
on VLC: Outdoor tests and performance analysis,’’ IEEE Trans. Intell.
tions engineering with the Faculty of Engineering,
Transp. Syst., vol. 23, no. 8, pp. 11465–11475, Aug. 2022, doi: Ain Shams University, Cairo.
10.1109/TITS.2021.3104498. Since 2009, she has been working as a full-time Teaching Assistant with
[34] S. Bahramnejad and N. Movahhedinia, ‘‘A reliability estimation frame- the Modern Academy for Engineering and Technology. Since 2016, she
work for cognitive radio V2V communications and an ANN-based model published three journal articles and one conference paper. Her research
for automating estimations,’’ Computing, vol. 104, no. 8, pp. 1923–1947, interests include fiber optics and visible light communications using different
Apr. 2022, doi: 10.1007/s00607-022-01072-7. techniques.
[35] A.-M. Cailean, C. Beguni, S.-A. Avatamanitei, and M. Dimian, ‘‘Experi-
mental demonstration of a 185 meters vehicular visible light communica-
tions link,’’ in Proc. IEEE Photon. Conf. (IPC), Vancouver, BC, Canada,
Oct. 2021, pp. 1–2, doi: 10.1109/IPC48725.2021.9592878.
[36] H. B. Eldeeb, M. Elamassie, S. M. Sait, and M. Uysal, ‘‘Infrastructure-to-
vehicle visible light communications: Channel modelling and performance
analysis,’’ IEEE Trans. Veh. Technol., vol. 71, no. 3, pp. 2240–2250,
Mar. 2022, doi: 10.1109/TVT.2022.3142991.
[37] K. Jadav, V. Sorathiya, W. El-Shafai, T. Altameem, M. H. Aly, V. Vekariya,
K. Ahmed, and F. M. Bui, ‘‘Shadow extraction and elimination of moving
vehicles for tracking vehicles,’’ Comput., Mater. Continua, vol. 77, no. 2,
pp. 2009–2030, 2023, doi: 10.32604/cmc.2023.043168.
[38] C. Beguni, A.-M. Căilean, S.-A. Avătămăniţei, A.-D. Potorac,
E. Zadobrischi, and M. Dimian, ‘‘Increasing vehicular visible light
communications range based on LED current overdriving and variable FATMA A. NEWAGY (Member, IEEE) received
pulse position modulation: Concept and experimental validation,’’ the B.Sc., M.Sc., and Ph.D. degrees in electronics
Sensors, vol. 23, no. 7, p. 3656, Mar. 2023, doi: 10.3390/s23073656. and communications engineering.
[39] E. Eso, A. Burton, N. B. Hassan, M. M. Abadi, Z. Ghassemlooy, and She is currently a Professor of communications
S. Zvanovec, ‘‘Experimental investigation of the effects of fog on opti- engineering with Ain Shams University. She has
cal camera-based VLC for a vehicular environment,’’ in Proc. 15th Int. more than 24 years of experience in communica-
Conf. Telecommun. (ConTEL), Graz, Austria, Jul. 2019, pp. 1–5, doi: tions and information technology. Throughout her
10.1109/ConTEL.2019.8848552. career, she held several positions, either academic
[40] S. Ghosh and M.-S. Alouini, ‘‘On the performance optimization of two- or managerial in several universities and organi-
way hybrid VLC/RF-based IoT system over cellular spectrum,’’ IEEE zations. To mention a few, Ain Shams University,
Internet Things J., vol. 9, no. 21, pp. 21204–21213, Nov. 2022, doi: Cairo University, The American University in Cairo, German University
10.1109/JIOT.2022.3177581.
in Cairo, and the Arab Academy for Science, Technology and Maritime
[41] G. Singh, A. Srivastava, and V. A. Bohara, ‘‘Visible light and reconfig-
Transport. She is the Supervisor of the Space Technology Laboratory,
urable intelligent surfaces for beyond 5G V2X communication networks
at road intersections,’’ IEEE Trans. Veh. Technol., vol. 71, no. 8, ASU–funded by Egyptian Space Agency (EgSA). She participated in differ-
pp. 8137–8151, Aug. 2022, doi: 10.1109/TVT.2022.3174131. ent ASRT activities, such as ‘‘Future Jobs in Egypt,’’ Updating the ‘‘National
[42] M. Mukherjee and K. Noronha, ‘‘Experimental analysis of received Strategy for Science, Technology, and Innovation STI-EGY 2030,’’ and Sug-
power for OOK-NRZ visible light communication system using off-the- gestions for ‘‘Sustainable Development Strategy–Egypt 2030.’’ She judged
shelf components,’’ Int. J. Inf. Technol., vol. 14, no. 6, pp. 2839–2853, Mega Projects and the Intel International Science and Engineering Fair (ISEF
Aug. 2022, doi: 10.1109/COMST.2019.2913348. 2019). She is the author/coauthor of more than 70 publications worldwide
[43] N. M. Okasha, A. H. A. Zekry, and F. A. Newagy, ‘‘Developing hybrid in the field of communications and information technology in refereed
car-to-car communication system based on MIMO visible light and radio international journals and conference proceedings. Her research and lecture
frequency,’’ EURASIP J. Wireless Commun. Netw., vol. 2022, no. 1, p. 114, interests include smart grids, energy efficiency, electrical vehicles, satellite
2022, doi: 10.1186/s13638-022-02192-6. communications, UAVs, AI, smart cities, security, and 6G.
[44] J. Tiwari, A. Prakash, and R. Tripathi, ‘‘An adaptive and cooperative MAC Dr. Newagy was a Steering Committee Member of MIT, ASU Center of
protocol for safety applications in cognitive radio enabled vehicular ad- Excellence–Energy founded by USAID, from 2018 to 2021. She is a member
hoc networks,’’ Ad Hoc Netw., vol. 138, Jan. 2023, Art. no. 103019, doi: of the Communications and Information Technology ICT National Commit-
10.1016/j.adhoc.2022.103019. tee, Egyptian Academy of Scientific Research and Technology (ASRT). She
[45] J. Gallego-Madrid, R. Sanchez-Iborra, J. Ortiz, and J. Santa, ‘‘The role of
is a member of the Space Research and Remote Sensing Scientific Council
vehicular applications in the design of future 6G infrastructures,’’ ICT Exp.,
and Specialized Scientific Councils, ASRT. She is also a member of the Arab
vol. 9, no. 4, pp. 556–570, Aug. 2023, doi: 10.1016/j.icte.2023.03.011.
[46] S. Kumari and B. Singh, ‘‘5G standard : The next generation wireless
Foundation of Young Scientists. She is an Editorial Board Member of many
communication system,’’ J. Interdiscipl. Math., vol. 23, no. 1, pp. 275–283, international journals, such as Ain Shams Engineering Journal (ASEJ) and
Jan. 2020, doi: 10.1080/09720502.2020.1721922. Elsevier. She has received many best paper awards at international confer-
[47] M. Vinodhini and S. Rajkumar, ‘‘Performance analysis of vehicle-to- ences. She is an Editor of the BOOK SERIES-Big Data for Industry 4.0:
everything communication using internet of LoRa computing for intel- Challenges and Application-Microgrids: Design, Challenges, and Prospects
ligent transportation system,’’ Intell. Decis. Technol., vol. 17, no. 12, (CRC Press, 2021).
pp. 577–594, May 2023, doi: 10.3233/IDT-220312.

89352 VOLUME 12, 2024