1

This work has been submitted to Engineering for possible publication.

A Survey on Performance, Current and Future Usage

of Vehicle-To-Everything Communication Standards
Falk Dettinger; Matthias Weiß; Daniel Dittler; Johannes Stümpfle; Maurice Artelt; Michael Weyrich

Abstract— Wireless communication between road users is communication paradigm is called Vehicle-to-Everything
essential for environmental perception, reasoning, and mission (V2X) communication [1, 2]. To realize V2X communication,
planning to enable fully autonomous vehicles, and thus improve two incompatible communication standards are currently being
road safety and transport efficiency. To enable collaborative used, namely (i) Wi-Fi-based and (ii) cellular-based V2X. Each
driving, the concept of vehicle-to-Everything (V2X) has long been
of them uses specific communication technologies with their
introduced to the industry. Within the last two decades, several
communication standards have been developed based on IEEE
own capabilities, protocol stacks, and supported use cases [3–
802.11p and cellular standards, namely Dedicated Short-Range 5]. The Wi-Fi-based approaches are divided into geographically
Communication (DSRC), Intelligent Transportation System G5 specific implementations, which in turn are incompatible with
(ITS-G5), and Cellular- and New Radio- Vehicle-to-Everything each other.
(C-V2X and NR-V2X). However, while there exists a high quantity Overall, new vehicular use cases such as remote driving,
of available publications concerning V2X and the analysis of the valet parking, platooning, and others envision the smart city. In
different standards, only few surveys exist that summarize these this paradigm, connected autonomous vehicles and smart
results. Furthermore, to our knowledge, no survey that provides infrastructure interact with each other, streamlining traffic flow
an analysis about possible future trends and challenges for the
and reducing the impact on citizens [6]. This vision requires
global implementation of V2Xexists. Thus, this contribution
provides a detailed survey on Vehicle-to-Everything
networking of many nodes and cyclic communication between
communication standards, their performance, current and future them. Due to the functional complexity, this leads to increased
applications, and associated challenges. Based on our research, we V2X communication requirements in terms of bandwidth,
have identified several research gaps and provide a picture about transmission speed and transmission reliability. Therefore, it is
the possible future of the Vehicle-to-Everything communication important to know the performance of approaches in different
domain. scenarios. In this way, the most efficient approach can be
selected for each specific use case to ensure the most resource-
Index Terms — C-V2X, Challenges on V2X, DSRC, IEEE 802.11, efficient data transmission.
ITS-G5, NR-V2X, Performance comparison, Vehicle-to-
Within the scope of this contribution our main objectives are:
Everything
- (O.1) the detailed analysis of the state-of-the-art V2X
I. INTRODUCTION technologies and their performance

A
- (O.2) the analysis of the market penetration of existing
utonomous vehicles have gained significant V2X approaches Dedicated Short-Range
prominence in recent years, becoming a central focus Communication (DSRC Intelligent Transportation
of research in the fields of transportation and System G5 (ITS-G5), Cellular- and New Radio-
technology. A high autonomy level can help to increase on-road Vehicle-to-Everything (C-V2X and NR-V2X)
safety and transportation efficiency and thus reduce accidents - (O.3) the summary of existing challenges for a
and traffic jams, while optimizing the fuel economy and the deployment of V2X in large scale real-world scenarios
travel time simultaneously [1]. To reach these goals, advanced The rest of the paper is organized as follow. In section II, the
vehicle functions with additional requirements regarding related work is summarized, whereas section III introduces the
environmental perception and decision-making show up. This methodology of this paper. The sections IV and V are dealing
leads to increasing information requirements of the functions, with the summary of the technical basics regarding the existing
necessitating a powerful exchange of data and information. V2X approaches and their performance. Furthermore, as there
Overall, the information exchange realizes the connected are several approaches used worldwide, their geographical
vehicle [5], which is able to communicate with other vehicles, usage is analyzed in section VI and the trend for their usage
roadside units (RSU), pedestrians, and networks. This

Corresponding author: Falk Dettinger Maurice Artlet, University of Stuttgart, Institute of Industrial Automation
Falk Dettinger, University of Stuttgart, Institute of Industrial Automation and Software Engineering (IAS), 70569 Stuttgart, Germany (e-mail: maurice-
and Software Engineering (IAS), 70569 Stuttgart, Germany (e-mail: [Link]@[Link]).
[Link]@[Link]). Michael Weyrich, University of Stuttgart, Institute of Industrial Automation
Matthias Weiß, University of Stuttgart, Institute of Industrial Automation and Software Engineering (IAS), 70569 Stuttgart, Germany (e-mail:
and Software Engineering (IAS), 70569 Stuttgart, Germany (e-mail: [Link]@[Link]).
[Link]@[Link]). Mentions of supplemental materials and animal/human rights statements can
Daniel Dittler, University of Stuttgart, Institute of Industrial Automation and be included here.
Software Engineering (IAS), 70569 Stuttgart, Germany (e-mail: Color versions of one or more of the figures in this article are available
[Link]@[Link]). online at [Link]
Johannes Stümpfle, University of Stuttgart, Institute of Industrial
Automation and Software Engineering (IAS), 70569 Stuttgart, Germany (e-
mail: [Link]@[Link]).
 2
This work has been submitted to Engineering for possible publication.

until 2030 is derived. In section VII, the currently existing that there is already a strong focus on investigating and
challenges regarding V2X applications in large scale real-world comparing security mechanisms and threats with respect to
scenarios are outlined. Finally, our results are discussed in cybersecurity. V2X standards and resource allocation as well as
section VIII and the conclusion is drawn in section VIII. applications and use cases have also been the subject of various
analyses. However, to the authors' knowledge, a detailed
II. RELATED WORK comparison of the performance of existing V2X approaches and
In recent years, a large number of publications on Vehicle- an analysis of the expected geographic deployment of the
to-Everything communication have been published, and approaches has not yet been performed. It is therefore the aim
surveys have been conducted to analyze and compare the of this paper to address this research gap.
content of the publications. In [7–11], the authors examine
methods of cybersecurity and privacy issues in the context of III. METHODOLOGY
V2X communication. In addition to common approaches [17], In the following, the basic information about the considered
vulnerabilities in standards [16] and possible attack vectors [15] review approaches will be summarized and the search strings as
are analyzed, and challenges and solutions are presented. well as the selection procedure of the relevant articles and
The existing standards and challenges related to V2X papers will be presented. Overall, the methodology used in this
communication are analyzed in [26, 27]. While Khezri et al. article was carried out in two stages:
[26] deal with the benefits and barriers of V2X communication, First, to define appropriate search terms for the literature
Gyawali et al. focus on cellular-based V2X communication C- review, an initial keyword search was conducted. For this
V2X and NR-V2X [27]. On the one hand, they focus on the purpose, the relevant terms and topics of 15 publications with
status of the V2X cellular standard, and on the other hand, they more than 30 citations in Google Scholar, dealing with vehicle-
describe specific challenges and their possible solutions in to-everything communication and trends, cooperative and
detail. collective perception, and connected vehicles were analyzed.
The surveys [25, 28, 29] examine the standards and use cases. The results of the initial keyword search, presented in Table 1,
In [29], Soto et al. describe cellular V2X standards and derive show heterogeneous cross-domain keywords, resulting in
communication types and applications in the context of unspecific search strings and thus a large number of
connected vehicles. In contrast, Alalewi et al. in [25] limit publications to be considered during the review.
themselves to NR-V2X use cases, where the authors describe During the literature review, a publication is considered if
the use case requirements. In [28], Wang et al. also investigate relevant terms are mentioned in the title, abstract, or keywords.
V2X applications, but also focuses on their possibilities for To focus on the performance analysis and technical trends on
testing and evaluation. In contrast, [30–32] discuss existing current events, only literature from 2017 onwards is considered,
approaches to resource allocation in V2X communication. since there is a higher number of active field tests involving
Looking at the overview of the analyzed surveys, it is evident V2X approaches starting in 2018. The Web of Science, Scopus,
Table 1: results of the keyword search considering relevant terms and ACM databases are used for the publication search because
and topics in [3, 12–25] they are more focused than Google Scholar. IEEExplore is not
explicitly considered, as its publications are directly
Relevant terms synchronized with and included in the Scopus and Web of
Top-Level Sub-Level
Science databases [33].
Vehicle-to-Everything (V2X)
Based on Table 1, we consider publications dealing with
IEEE 802.11p, bd
V2X and V2X-related approaches based on IEEE 802.11 and
ITS-G5
DSRC
cellular V2X and their performance analysis and global
C-V2X geographic trends during the literature review. This includes
LTE-V2X DSRC, ITS-G5, and C-V2X, LTE V2X, and NR V2X and their
NR-V2X performance analysis. To outline geographical trends, the use
V2X-Messages of V2X in the United States, Europe, and China will be
Performance analysis of V2X approaches highlighted. From the overall literature, the current challenges
Approach related overview about technical basics in V2X communications will be derived.
Cooperative perception / Collective perception The summarized search terms are shown in Table 2 along
Trend with the number of publications listed in the considered
Europe databases. Since all V2X and V2X-related approaches are
China relevant for this publication, the keywords V2X, IEEE 802.11,
United States DSRC, ITS-G5, C-V2X, LTE-V2X, and NR-V2X are considered
Standards alternatively. To focus the results of the literature review, three
ETSI
subtasks are defined to search for technical fundamentals,
3GPP
performance evaluation, and geographical trends. As this
Cooperative intelligent transportation systems (C-ITS)
contribution focuses on the vehicle domain, publications
Connected and autonomous vehicles (CAV)
focusing on the rail, maritime, and air domains are excluded. In
 3
This work has been submitted to Engineering for possible publication.

Table 2: Search strings and number of filtered publications for the initial keyword search with Google Scholar and the literature
review with the databases ACM, Web of Science (WoS) and Scopus.
Topic Search String ACM WoS Scopus Scholar
Initial keyword "Vehicle-To-Everything" OR V2X" OR "("V2X" AND "trends" OR - - - 8,330
search (2019 - (("cooperative" OR "collective") AND perception") OR "connected
2023) vehicles"
Technical basics ("V2X" OR "Vehicle to Everything" OR "NR V2X" OR "C V2X" 6 110 226 -
to V2X-related OR “DSRC” OR "ITS G5" OR "IEEE 802.11p" OR "LTE V2X")
approaches AND ("overview" OR "technical" OR "basics") NOT ("safety" OR
(2017 - 2023) "security" or "map" OR "Framework" OR "aerial" OR "rail")
Performance ("V2X" OR "Vehicle to Everything" OR "NR V2X" OR "C V2X" 30 1,897 2,144 -
evaluation of OR “DSRC” OR "ITS G5" OR "IEEE 802.11p" OR "LTE V2X")
V2X AND ("Performance" OR "Evaluation") NOT ("safety" OR
(2017 - 2023) "security" or "map" OR "Framework" OR "aerial" OR "rail")
Geographical ("V2X" OR "Vehicle to Everything" OR "NR V2X" OR "C V2X" 16 855 161 -
technological OR “DSRC” OR "ITS G5" OR "IEEE 802.11p" OR "LTE V2X")
trends in V2X AND ("trends" OR "Europe" OR "United States" OR "US" OR
(2017 - 2023) "China") NOT ("safety" OR "security" or "map" OR "Framework"
OR "aerial" OR "rail")

addition, to focus more on V2X communication, publications part of C-ITS. Due to a high dynamic environment, high travel
with specific frameworks and topics related to cooperative speeds and unexpected behavior of traffic participants, the
perception, such as map generation or point clouds, as well as interconnection between traffic participants requires three
safety and security aspects are excluded. specific needs. Besides a low latency for data transmission, a
Searching with, e.g., Web of Science, shows 2,054 high packet delivery ratio is needed, while maintaining the
publications dealing with V2X performance evaluation available maximum bandwidth of the communication channel.
excluding rail, maritime, safety and security applications and The related communication mechanism is called Vehicle-to-
frameworks. 172 are summarizing an overview about V2X and Everything communication.
its sub-approaches, while 1,125 address trends regarding V2X Currently, there are two different approaches for V2X
or geographical trends in China, Europe, or the United States. communication to enable the transmission of data and
Because of the large number of publications, we focus on a information between road users. Dedicated Short Range
semi-structured literature review [34, 35]. Communication (DSRC) is standardized by the United States
An initial semi-structured literature review is conducted to of America, while Intelligent Transportation System G5 (ITS-
summarize the state-of-the-art in section III and IV. To analyze G5) is standardized by the European Union. Both are based on
the performance of different V2X approaches, derive the the Wireless Local Area Network (WLAN) standard, IEEE
worldwide usage of these approaches beyond 2030 and to show 802.11p (i), but use a different communication stack. Cellular
existing challenges in section V to VII, a meta-analysis is V2X (C-V2X) and New Radio V2X (NR-V2X) (ii) use the 4th
practiced. Both, the semi-structure review and the meta- (Long Term Evolution (LTE)) and 5th generation (5G) cellular
analysis will consider the same literature basis, as most of the standards [4, 37]. In the following sections, the technical basics
publications introduce technical basics and a performance and the development timeline of both V2X communication
evaluation regarding the approaches. approaches are presented. For this purpose, the development
and associated communication stacks of DSRC and ITS-G5 as
IV. TECHNICAL BASICS part of the IEEE 802.11 standard are presented in Section A. C-
Technological progress in automotive applications is steadily V2X and NR-V2X are then introduced in Section B.
evolving. Advanced driver assisted systems (ADAS) as well as A. Wi-Fi based V2X
semi-autonomous systems increase both vehicles drivability Both DSRC and ITS-G5 are based on the IEEE 802.11p
and on-road safety. Due to this intelligent and autonomous standard [38]. This is a modified WLAN protocol introduced in
behavior, these systems are called intelligent transportation 2010, in which changes were made to the physical layer (PHY)
systems (ITS). Currently the usage of Level 3 autonomous and the data link layer, more precisely the MAC (Medium
systems like the Mercedes Benz drive pilot is allowed in Access Control) layer [3]. The initial requirements for IEEE
scenarios with lower or limited complexity, such as on 802.11p were a relative vehicle velocity of 200 km/h, a response
highways [36]. time of 100 ms and a communication range up to 100 m [39].
The interconnection of road participants enables the At the PHY layer, the operating frequency of the 802.11p
extension of ITS to cooperative ITS (C-ITS). They are expected standard has been increased from 5 GHz to 5.9 GHz compared
to be an indispensable step towards the development and the to the 802.11a standard [40]. Other changes were made to the
establishment of autonomous vehicles [36]. Thus, besides parameterization of the Orthogonal Frequency Division
vehicles, pedestrians and roadside units are also considered as Multiplexing (OFDM) modulation method [41]. The IEEE
 4
This work has been submitted to Engineering for possible publication.

802.11p standard uses Binary Convolutional Coding for error 802.11bd must also support 802.11p devices and be compatible
correction. However, its ability to recover erroneous messages with them at the communication layer [39, 47]. Among other
is lost as soon as higher Modulation Coding Schemes (MCS) things, to increase performance, the following modifications
are used for higher bit rates at distances greater than 50m [41]. were made [47, 48] to increase to a factor of 3, and to improve
More fundamental changes have been made to the MAC robustness and interoperability:
layer. In general, the IEEE 802.11p standard defines a Basic • Binary Phase-Shift Keying- Dual Subcarrier Modulation
Service Set (BSS) for network construction. This is a group of (BPSK-DCM) and 256 QAM as well as Multiple Input
network devices that allows the use of different network Multiple Output operation in unicast mode have been
topologies, such as mesh networks or access points. In order for added to the modulation schemes.
a device to send messages within a device group, it must be a • In addition, 52 data subcarriers are now supported
member of that BSS. However, membership involves instead of 48 data subcarriers in 802.11p.
procedures such as channel scanning, association, and • In addition to BCC, Low-Density Parity-Check (LDPC)
authentication. The time required for this is not available in the is also supported for error correction.
V2X context [42, 43], so an alternative mode called Outside the • The frequency band has been extended to the 60 GHz
Context of BSS (OCB) has been defined in IEEE 802.11p. In band and localization is supported to enable new high-
this mode, the BSS control procedures are omitted, resulting in performance applications.
a lower time overhead for message transmission. OCB is • One to three retransmissions of messages are also
particularly suitable for fast transmission of short messages. supported.
Network devices operating in OCB mode use Enhanced Currently there exist two specific implementations, namely
Distributed Channel Access to access the transmission medium. DSRC and ITS-G5. DSRC was developed in the USA, while
This is based on the Distributed Coordination Function (DCF), ITS-G5 was standardized by the European
which manages the access of multiple nodes to a wireless Telecommunications Standards Institute (ETSI) under ESI EN
network. In IEEE 802.11p, the access method used is 302663 [38, 42]. For the sake of completeness, it should be
CSMA/CA (Carrier Sense Multiple Access with Collision mentioned that ITS-Connect was standardized as the Japanese
Avoidance) [44]. version of DSRC/ITS-G5 by the Association of Radio
In CSMA/CA, the participants listen to the communication Industries and Businesses (ARIB) [2]. As mentioned at the
medium [40]. If the medium is free for a certain time, the station beginning, DSRC and ITS-G5 are based on the IEEE 802.11p
willing to transmit starts the transmission. If the channel is standard, but the approaches differ in some respects [2, 40, 49].
busy, the transmission is delayed by a random period In the USA [3], DSRC has been mandatory for new vehicles
(Contention Window, CW). Prioritization of messages is made since 2016, while in the EU [50] both DSRC and ITS-G5 are de
possible by assigning Access Categories (ACs) depending on facto standards. Here, the first V2X-capable vehicles were
the content of the message produced in 2019.
Because IEEE 802.11p does not meet the high latency and
reliability requirements of new Day2 and Day3+ use cases, the Dedicated Short Range Communication (DSRC)
IEEE extended the 802.11p standard to 802.11bd in 2018 [47]. In 1999, the Federal Communications Commission (FCC)
In doing so, the requirements for the standard were expanded reserved a 75 MHz frequency band in the 5.820 GHz and
and technical improvements introduced and tested in the 5.925 GHz range for automotive use [51]. In 2002 [49], the
802.11n/ac/ax standards were adopted in 802.11p with the American Society for Testing and Materials (ASTM) published
primary goal of achieving higher throughput and greater ASTM E2213, which proposed the use of a modified IEEE
resilience. The requirements of the IEEE 802.11bd standard 802.11a standard to implement DSRC. This was taken up by the
have also been adapted. For example, it requires a relative speed IEEE and the IEEE 802.11p standard was created. The
of 500 km/h, a mode with at least twice the range of IEEE transmission range is in the range of a few hundred meters
802.11p, and at least one mode for localization. In addition, compared to classic mobile radio technology [40].
The frequency band provided by the FCC is divided into
seven channels [2], each 10 MHz wide. One channel is designed
as a control channel (CCH), while the other six channels are
used for service delivery (SCH). The remaining 5 MHz, which
is not assigned to any channel, is used as a guard band to prevent
interference (Figure 1 (top)).
While the CCH is intended for broadcast transmission of
safety-related messages, messages for infotainment and traffic
purposes are transmitted via the SCH [2]. In addition to the
IEEE 802.11p standard, the DSRC communication stack also
Figure 1: DSRC (top) and ITS-G5 (bottom) band Channel includes the IEEE 1609.x standard, which together form the
frequencies according to [40, 45, 46]. For the ITS-G5 Band, no main part of the Wireless Access for Vehicular Environment
channel names are available. (WAVE) stack [40]. However, the term IEEE 802.11p is
 5
This work has been submitted to Engineering for possible publication.

usually used to refer to the entire WAVE protocol stack [51], as and to give priority messages a high delivery probability, the
shown in Figure 2 (IEEE WAVE). ITS-G5 standard also uses Decentralized Congestion Control
The IEEE 1609 standard [51] is used for direct (DCC) methods [2]. The GeoNetwork (GeoNet) protocol is
communication between vehicles and between vehicles and the used in the transport layer. This is a standard defined by ETSI
traffic infrastructure. In this context, IEEE 1609.2 implements EN 302 636 that allows ad hoc multi-hop communication
digital security for message transmission. It provides between nodes. It allows addressing and forwarding of
authentication mechanisms and message encryption using messages using geographical coordinates. Therefore, messages
digital signatures and certificates. Certificates do not contain can be sent and forwarded even if a communication partner is
driver-specific information and are changed cyclically, not directly in the communication range of the sender.
protecting the identity of the driver. IEEE 1609.3 defines Compared to the WAVE Short Message Protocol (WSMP),
WSMP (Wave Short Message Protocol) [40], which is a single- multi-hop communication via geo-addressing is more complex
hop network protocol with a small header and supports message and has a higher overhead. On the other hand, the Basic
multiplexing, where messages are only transmitted to Transport Protocol (BTP) is used to realize the end-to-end
communication partners one hop away from the current access transport of messages to the ITS devices.
point. The MAC sublayer extension allows DSRC systems to
B. Long Term Evolution (LTE)/5G based V2X
efficiently switch between communication channels when there
are multiple transceivers. With the advent of more powerful wireless communication
standards, C-V2X (Cellular V2X) has been developed based on
Intelligent Transportation System G5 (ITS-G5) 4G and 5G communication [4, 37]. This is divided into LTE-
The ITS-G5 standard also operates in a 75 MHz spectrum V2X, which uses LTE (Long Term Evolution) communication,
that is divided into seven 10 MHz channels [46]. However, in and NR-V2X, which uses 5G-NR (New Radio) communication.
contrast to DSRC, the frequency band is additionally divided C-V2X is intended to enable new use cases and applications
into functional groups. The band from 5.875 GHz to 5.905 GHz compared to IEEE technology. Unlike DSRC, for example,
corresponds to the main frequency band and is used for the which is only licensed in the 5.9 GHz band, C-V2X
transmission of safety-related messages and traffic efficiency communication is also supported in the licensed sub-6 GHz
messages. For non-safety related messages, the 5.855 GHz to band. In contrast to the 802.11 standard, the standardization is
5.875 GHz frequency band is used. The frequency range from done by the 3GPP (3rd Generation Partnership Project), which
5.905 GHz to 5.925 GHz is reserved for future expansion [46]. is responsible for the general standardization of mobile radio.
Figure 1 (bottom) shows the frequency bands structure. [54]. LTE was defined in 2008 with Release 8 (Rel-8) and has
Similar to DSRC, ITS-G5 uses the IEEE 802.11p standard in been enhanced with new features in recent years.
the PHY and MAC layers [53]. The structure of the ITS-G5 In C-V2X Release 12, also known as LTE-Advanced, the
stack therefore only differs in the layers above, which is why D2D (Device to Device), also known as Sidelink (SL)
ITS-G5 supports additional features compared to the WAVE communication, was introduced [1, 37, 55] using the PC5
components. The structure of the protocol stack is shown in interface to support Proximity Services (ProSe) without the
Figure 2 (ETSI ITS-G5). In order to limit channel occupancy need of the base station (eNodeB, eNB) for short-range

Figure 2: left: WAVE protocol stack considering the included sub-layers for security, MAC And PHY Layer according to [40, 52] and
ITS-G5 Stack containing physical and MAC layer as well as overlying layers according to [49, 52] (right)
 6
This work has been submitted to Engineering for possible publication.

communication between two or more network nodes using the During a TTI, 14 OFDM symbols are transmitted, nine of which
PC5 interface, through which V2V, V2I and V2P are used for data transmission and four for the transmission of
communication can be efficiently realized. On the other hand, Demodulation Reference Signals (DMRSs) [5]. The DMRSs
the Uu interface enables communication between the user are used to estimate the channel parameters and to suppress
equipment (UE) and the base station (eNodeB or eNB), which Doppler effects caused by the high vehicle speed. The
is directly used for V2N communication. [56]. To enable the remaining OFDM symbol is used as a guard symbol [1].
D2D communication, the eNB is bypassed by network nodes
and thus, the data traffic via the eNB is reduced. In addition, A group of RBs in the same subframe is called a subchannel
D2D communication can be used to forward data when and the number of RBs per subchannel can vary depending on
subscribers have no or poor connection to the eNB. Overall, two the configuration and the amount of data [59]. The subchannels
different modes [57] are defined for scheduling the message are used to transmit data and control information, where the
transmission of a network node (UE). Mode 1 is a centralized data is organized into Transport Blocks (TBs) that are sent over
connection approach. Here, the planning of the dedicated radio the Physical Sidelink Shared Channel (PSSCH). A TB contains
resources of the individual UEs is done by the eNBs in a complete message packet, such as a CAM or BSM, and
coverage, therefor called in-coverage mode. In contrast, in occupies a different number of subchannels depending on the
Mode 2 the radio resources are selected independently by the number of packets sent. In addition to the data, a corresponding
UEs from a defined resource pool and the transmission is Sidelink Control Information (SCI) message is sent per TB over
performed via the PC5 interface. In Mode 2, no UE needs to be the Physical Sidelink Control Channel (PSCCH) assigned to the
within range of the subscribers, which is why this mode is also TB, which requires two RBs. This message contains
referred to as out-of-coverage mode. Due to the high information such as the type of modulation, the priority of the
communication latency of D2D communication in Rel-12, it message, and whether it is an initial transmission or a blind
was not suitable for V2X communication [57]. Figure 3 shows retransmission of a message. If a TB requires retransmission,
the 3GPP release history for C-V2X technology. the same message structure is sent in a different subframe [57].
Building on the D2D communication in Rel-12, LTE-V2X In this case, the receiver uses the Hybrid Automatic Repeat
communication was specified in Rel-14 [2]. This is based on Request (HARQ) procedure to combine the two identical TBs
the time-division LTE (TD-LTE) based LTE-V presented by from transmissions one and two to form a resulting message.
the China Academy of Telecommunication Technology The repeated sending of a message by the sending UE is done
(CATT) in 2013. Similar to C-V2X, this includes two modes without prior knowledge of whether the message was
for centralized and decentralized data transmission and was successfully received the first time. This means that the sending
standardized by 3GPP as LTE-V2X in late 2016 [37]. LTE- UE performs the resending blindly [4]. Overall, this leads to an
V2X is specifically designed for use in the automotive increased reliability of message delivery.
environments [58]. Since the publication of LTE-V2X, the
development of C-V2X communication has been divided into
phases. Rel-14 marks the beginning of phase 1, which was
completed in March 2017 [37].
With Release 15 in 2017, Phase 2 of C-V2X was initiated,
introducing several improvements and enhancements to the
LTE V2X technology were introduced. These include support
for advanced V2X scenarios such as platooning or remote
vehicle control [3]. Performance improvements have also been
introduced. For example, 64-QAM was introduced as an
additional coding scheme [37], which increases the number of
transmitted bits per symbol and thus the data rate compared to
the originally used 16-QAM improving the performance of the
PC5 interface.
Like the LTE uplink [1], LTE-V2X uses the single-carrier
frequency-division multiple access (SC FDMA) method in the
PHY layer. The channel width is either 10 MHz or 20 MHz and
is divided into resource blocks (RBs). The overall structure of
LTE radio frames is summarized in Figure 4. These have a Figure 3: V2X-Timeline starting in 2014 with Rel-12, till the
width of 180 kHz and represent the smallest unit that can be end of 2023 with Rel-18 according to [54, 60, 61]. Timeline
allocated to a UE as a resource in the frequency domain. Each shown on the left, state of 4G and 5G radio communication
of these RBs contains 12 subcarriers, each with a bandwidth of (center) and V2X state (right).
15 kHz. In the time domain, the channel is divided into
subframes of 1 ms length, which are called Transmission Time
Interval (TTI), which is the smallest unit in the time domain.
 7
This work has been submitted to Engineering for possible publication.

Figure 4: left: Structure of Radio frames in LTE-V2X. In LTE each radio frame contains out of ten sub-frames, each 1 ms long while
each sub-frame consists out of two timeslots with maximum seven OFDM symbols per slot. Right: Structure of one time slot also named
sub-channel. These contain of two resource blocks (RB) with 12 carrier signals each 15 kHz wide and 0.5 seconds long. [1, 5, 41]
LTE-V2X introduces two new modes for C-V2X enhancing existing features such as sidelink communication,
communication, Mode 3 and Mode 4 [37, 62]. These are in power saving, and coverage and positioning improvements [60,
addition to the Mode 1 and Mode 2 modes introduced in Rel. 63]. In addition, some new features are provided, with a focus
12. In Mode 3, similar to Mode 1 of the initial C-V2X, resource on the integration of satellite access within the 5G network. The
management is controlled by an eNB. The UE requests changes are summarized in TR 21.917, which is not finally
resources from the eNB via the LTE-U (Uu) interface. The PC5 published in December 2023. Since Q4 2021, 3GPP has been
interface is then used for communication between the nodes. working on 5G "Advanced" in Release 18, which is expected to
In Mode 4, similar to NR V2X Mode 2, resource management be released in 2024 [63]. The focus is on system architecture
is not performed by the eNB but directly by the individual UEs. and related services as well as radio layers and interfaces [64].
Communication then takes place directly via the PC5 interface. Therefore NR-V2X is not supposed to replace LTE-V2X [5],
Al though the operation of modes 1/3 and 2/4 is similar, they rather than to extend it. As the 5G technology is more powerful
differ significantly in their structure and the way PSCCH and than 4G (LTE), classic use cases like the transmission of safety
PSSCH are allocated [44]. Mode 3 specifies two scheduling related messages like CAMs as well as enhanced V2X (eV2X)
approaches for resource selection. In dynamic scheduling, UEs – sometimes also called advanced V2X – use cases can be
request subchannels from the eNB for each TB. In contrast, in covered. 25 new use cases, assigned to four groups were defined
Semi Persistent Scheduling (SPS), the eNB reserves the in NR-V2X [37]. To support these new use cases two additional
subchannels for the UEs. In Mode 4, the Sensing Based Semi cast types, called unicast and groupcast were introduced, in NR-
Persistent Scheduling (SP-SPS) introduced in Rel-14/15 is V2X [57]. LTE-V2X only supports broadcasting, where all
used. The UEs then select their own subchannels. In Mode 3, nodes within the communication range of the sender can receive
higher performance can be achieved through centralized the message. In contrast, unicast allows communication
resource allocation, although there is a higher overhead due to between two dedicated UEs, while groupcast allows the
communication with the eNB [5]. message transmission from a UE to a specific group of
For a low density of communication participants, the defined subscribers. In NR-V2X, a UE can communicate via several
requirements for V2X communication via LTE-V2X can be cast types simultaneously using the PC5 interface [54].
achieved in terms of latency and bandwidth for simple data The NR-V2X SL in Rel-16 is designed for frequency range
transmission [57]. However, the Quality of Service (QoS) 1 [5]. This starts at 0.41 GHz and ends at 7.125 GHz. In
requirements are constantly increasing. The performance limits addition, the 5.9 GHz band, which is also used by LTE-V2X
of LTE-V2X have been exceeded, in particular, by enabling and DSRC, and the 2.5 GHz band for WLAN can be used.
new use cases such as platooning or extended sensors, as Frequency range 2 (24.25 GHz to 52.6 GHz band) is also
described in section. III.D, the performance limits of LTE-V2X supported, but was not the focus of Rel-16.
have been exceeded. the solution is a more powerful standard As introduced, the NR-V2X supports a variety of different
called New Radio (NR)-V2X. This is based on 5G New Radio communication types such as unicast, groupcast and broadcast.
(NR) technology. 5G NR was introduced by the 3GPP in V2X This provides greater flexibility and enables new use cases to
Rel-15, initially without sidelink communication [5, 57]. Based be realized. In NR-V2X, the modulation scheme of C-V2X in
on 5G NR, NR-V2X was then introduced in Rel-16, including Rel-15 is extended by the 16-QAM of Rel-14. This results in
the sidelink for D2D communication. The development of Rel- backward compatibility. In addition, the NR-SL supports 256-
16 started in June 2018 and was completed in July 2020 [4, 57]. QAM [65], which enables a higher throughput over the sidelink
With this progress, the V2X Phase 3 has been completed. channel. Since most V2X use cases require a communication
In 2020, the content extension for 5G Release 17 was started distance of less than 1000 m, the connection speed gains higher
and the release was frozen in 2022 [61]. The focus was on priority. Thus, the change from SC-FDMA to OFDM was
 8
This work has been submitted to Engineering for possible publication.

implemented, which increases the connection speed while D. Use-Cases

reducing the coverage at the same time. To increase the The 3GPP defines C-V2X and LTE-V2X use cases (in TR 22
reliability for the group and unicast communication [39], a 885) based on the ETSI (ETSI TR 102 638). These groups of
feedback channel is provided on the sidelink channel. This use cases require a maximum latency of 100 ms and packet
eliminates the need for blind retransmissions by the sender, as delivery ratio of 95 % [19]. Thereby the requirements for
the receiver confirms the correctness of the transmission. To different use cases are not identical. Some use cases require
reduce the latency the time division multiplexing is used in NR lower latency, other higher reliability, or bandwidth.
V2X. The receiver obtains all necessary information for With the more advanced communication technique of
decoding, followed by the actual data. Additionally, NR-V2X NR-V2X, 25 new use cases were defined by the 3GPP. They
supports a flexible frame structure with subcarrier spaces are divided into four groups namely Vehicle Platooning,
between 15 kHz and 480 kHz [1]. This leads to shorter advanced Driving, Extended Sensors and Remote Driving [5,
transmission intervals (TTI) and explicitly reduces latency. 66–68].
Furthermore two new modes for resource allocation were As shown in Table 3 there are three relevant parameters that
introduced with NR V2X. Namely SL Mode 1 and SL Mode 2 need to be considered in different use cases. Latency describes
[37]. SL Mode 1 is equal to the LTE-V2X Mode 3, where the the time between sending a message to a sender, processing it
so called gNB (instead of eNB in Mode 3) is responsible for the at the receiver, and receiving the results at the original sender.
resource scheduling. In SL Mode 2 (similar to LTE-V2X Mode The use case latency depends on the distance between the
4) UEs can communicate with each other directly using the vehicles and their speed, while the system latency depends on
sidelink channel. the network structure and the current network situation such as
C. Coexistence between LTE-V2X and NR-V2X traffic and number of participants. In particular, lower latency
can only be achieved with SL communication, as multi-hop
As introduced in the beginning, NR-V2X should not replace
communication increases latency due to routing. The packet
LTE-V2X, rather than extend it. So, with LTE-V2X and NR-
delivery ratio describes how many messages sent from a source
V2X different concepts are available [57]. NR-V2X is not
to a destination must be delivered correctly. It depends on the
compatible with LTE-V2X due to the usage of different
distance between vehicles, the maximum communication
numerologies like scalable Transmission Time Interval (TTI)
distance of the RATs, and the environmental conditions.
durations and sub-carrier spacings (SCS) in cases where the
Bandwidth requirements are described by the data rate and
higher performance of NR-V2X is required. Therefore, new
define the number of bits transmitted per second. The required
vehicles will likely be equipped with both radio access
data rate is directly related to the number of messages and their
technologies [39].
payload. Often, the data rate differs between the V2V data rate
While LTE-V2X covers classic use cases, NR-V2X is used
using the SL channel and the V2I data rate and the V2N data
for advanced ones [5]. It is intended that vehicles select the
rate.
appropriate Radio Access Technology (RAT) based on the
Vehicle Platooning: Vehicles in motion form a dynamic
current use case. To enable the coexistence of LTE-V2X and
convoy. Within a convoy, vehicles communicate by sending
NR-V2X, mechanisms have been defined in Rel-16, also
periodic messages to ensure the convoy's functionality. The
known as In-Device Coexistence. These are responsible for
goal is to optimize transportation and fuel efficiency. This
managing SL resources at the vehicle and device level.
requires latency between 10 ms and 500 ms, depending on the
Furthermore, the coexistence of both RATs at the network level
speed and number of vehicles in the convoy, and a message
is defined [39]. This mechanism is referred to as cross-RAT
delivery rate in the range of 90 to 99.99 %. As vehicles
control. In Rel-16, it is only applicable when LTE-V2X and
dynamically join or leave the convoy, maneuvering within the
NR-V2X operate on different channels. The challenges of in-
convoy and with road infrastructure or pedestrians needs to be
device coexistence lie in the management of resources such as
coordinated.
used channels and radio frequencies while both RATs are in use
Advanced Driving: Vehicles share sensor information and
[5]. The main challenges are the limited transmission power as
maneuver intentions to negotiate their trajectories. As vehicles
well as possible interferences between the RATs due to the
Table 3: ETSI Use Cases and their requirements regarding
simultaneous use of non-well separated frequencies. Rel-16
latency, reliability and bandwidth [66–68]
defined two possible solutions to mitigate interference between
V2X Use Latency Reliability V2N V2V V2I
LTE and NR-V2X, namely the use of Time Division case (ms) (%) Data Data rate Data
Multiplexing (TDM) and Frequency Division Multiplexing rate (Mbps) rate
(FDM). (Mbps) (Mbps)
With the TDM approach [5], only one RAT can access the Platooning 10 - 500 90 - 99.99 80 - 350 2 - 65 2 - 50
Advanced 3 - 100 90 - 99.99 10 - 50 0.5 - 50 0.5 - 50
transmitter at a time. Therefore, only alternate transmission of Driving
messages is possible and further synchronization is required. Extended 3 - 50 95 - 99.99 10 - 50 25 - 1000 -
With FDM [5], both RATs can be used simultaneously and Sensors (Map (collective
Sharing) Perception)
the available transmission capability is shared between the Remote 5 99.999 DL: 1 - DL: 1
RATs and can be allocated statically or dynamically based on Driving UL: 25 UL: 25
the message priority.
 9
This work has been submitted to Engineering for possible publication.

travel at different speeds, low latency data transmission is different traffic density scenarios, i.e., if there are many or only
required, ranging from 3 ms to 100 ms depending on the few partners are involved.
situation and speed. In addition, to ensure safe maneuver Across the observed papers, the following metrics were
coordination, a high packet delivery ration is required, ranging identified as common and are thus used for direct comparison:
from 90% to 99.99%. Therefore, 0.5 Mbps to 50 Mbps is • Packet Reception Ratio (PRR): The ratio between the
required for V2V and V2I communication, as status number of neighbors that correctly decode a
information and planned maneuvers need to be distributed transmitted message and the number of total neighbors
among different participants. within a given distance of the transmitting vehicle [1].
Extended Sensors: Enables the exchange of sensor data with • End-to-End (E2E) Latency: the time required to
nearby vehicles, RSUs, pedestrian devices, and V2X servers. transmit a message from the sender's application layer
This includes pre-processed data. The goal is to increase the to the receiver's application layer. Thus, it includes
awareness of individual road users and to extend the field of channel access [2].
view of the participants. Due to the variable speed and highly • Update Delay (UD): The time difference between two
dynamic environment, a low latency between 3 ms and 50 ms successive successfully decoded messages from a
is required. Due to the dynamic nature of the environment, a sender to a receiver located at a certain distance from
packet delivery ratio between 95% and 99.99% is required. the sender [56].
Depending on the level of data distribution, data rates between • Packet Delivery Ratio (PDR): The ratio between the
10 and 50 Mbps are required for collaborative map sharing number of correctly received messages at a receiver
and the number of retransmitted messages from a
between vehicles and networks, and 25 Mbps - 1000 Mbps for
reference transmitter.
collective perception between different vehicles.
Additionally, the following metrics are used by a partition of
Remote Driving: Enables the remote control of vehicles. This
the papers and mentioned if measured:
includes both driving without passengers and transporting
• Packet Loss Rate: The percentage of data packets that
passengers. Another important use case is the remote control of
are transmitted but fail to reach their destination,
automated vehicles that are in a situation where they can no
indicating the reliability of the communication system.
longer control themselves. This requires a low latency of about
• Average Bit Rate: The average amount of data
5 ms and a high reliability of 99.999% to enable safe remote transferred per unit of time, measured in bits per
control of a vehicle. The stream of sensor data from the vehicle, second (bps), reflecting the data transmission speed.
as well as infrastructure status information, must be transmitted • Average Latency: The average time taken for a data
to the remote driver. This results in an average uplink of packet to travel from the sender to the receiver,
approximately 25 Mbps. Since only the steering signals are indicating the responsiveness of the communication
transmitted to the vehicle, the downlink is about 1 Mbps. Table system.
3 summarizes different parameter values based on different • Average Range: The typical maximum distance over
literature references [66–68]. which communication can occur effectively without
significant loss of signal quality.
V. PERFORMANCE ANALYSIS • Channel Busy Ratio (CBR): The proportion of time
This section provides an overview of several different the communication channel is occupied with
research efforts of recent years that have focused on the transmissions, showing the channel's utilization and
performance of LTE-V2 and NR-V2X and the implementations congestion level.
of IEEE 802.11p, namely DSRC/ITS-G5. First, the • Propagation and Collision Error: Errors in data
performance metrics commonly used for comparison are briefly transmission caused by signal propagation issues and
explained in Section V.A. Subsequently, the results of the collisions when multiple transmissions interfere with
individual works are summarized in Section V.B. Finally, an each other.
assessment of the two technologies based on previous results is • Received Signal Strength Indicator (RSSI): A measure
of the power level received by the receiver, indicating
made in Section V.C.
the strength and quality of the received signal.
A. Performance metrics
B. Survey Results
Within the studies considered, various metrics are used for
Direct performance comparison between V2X
performance assessment. These metrics and evaluation
communication via IEEE 802.11p and LTE-V2X has long been
scenarios are briefly explained below.
a focus of research. An overview of studies comparing the
In general, two different environments are considered for
performance of communication technologies using IEEE
evaluation. For Line of Sight (LOS) measurements, two or more
802.11p standard and LTE-V2X with Mode 3 and 4 is given by
communication partners have a line of sight to each other that
A. Keyvan in [44]. The studies contained therein are from 2017
is not interrupted by objects. On the contrary, None-Line of
and 2018 and the technologies are compared in two different
Sight (NLOS) refers to two or more communication partners
scenarios with high and low traffic density respectively.
that cannot establish a line-of-sight connection because there
However, Keyvan does not explain how one technology is
are objects in between. Additionally, the papers consider
superior to the other, but he infers that neither technology is
 10
This work has been submitted to Engineering for possible publication.

superior to the other in all aspects. He also points out that most The study [71] by R. Molina-Masegosa and J. Gozalvez
of these studies are subject to limitations, such as being compares LTE-V2X and IEEE 802.11p technologies in terms
simulations under ideal conditions, and that while LTE-V2X of packet delivery ratio (PDR) at varying distances and
performs better than IEEE 802.11p in large areas, the potential transmission rates. LTE-V2X outperforms IEEE 802.11p at
improvements of IEEE 802.11p have been under-studied. distances beyond 200m, especially at higher transmission rates.
The authors in [37] investigate the performance of LTE-V2V However, IEEE 802.11p performs better at shorter distances
in modes 3 and 4 using simulation and consider the PDR as a and lower transmission rates. LTE-V2X generally experiences
function of distance in a highway scenario. They vary the traffic a lower CBR than IEEE 802.11p under the same traffic density,
density (in vehicles per meter), packet transmission rate (in especially when handling aperiodic messages. This is because
packets per second, pps), and transmit power. They observe a packet collisions, which are more frequent in LTE-V2X,
decrease in PDR in both modes when the packet transmission generate less channel load due to collided packets contributing
rate is increased, the traffic density increases, or the transmit less to the channel load.
power is decreased. However, the performance in mode 3 In [58], K. Jellid and T. Mazri examine PDR versus traffic
operation is always better than the performance in mode 4. density for DSRC and LTE-V2X technologies in a highway
Regardless of the operating mode used, the work shows that a scenario. LTE-V2X exhibits superior PDR compared to DSRC
PDR of over 90% can only be guaranteed at very short distances at shorter delivery times, lower latency and higher traffic
(< 100m). Metrics apart from PDR like average bit rate, average densities. Unfortunately, the authors do not provide any
latency, CBR, and propagation and collision error are information about the mode in which LTE-V2X is operated and
mentioned as future work considerations but are not specifically the distance between the vehicles. LTE V2X offers a much
compared in the provided simulations. higher bit rate (up to 1 Gbits/s) compared to DSRC (3-27
In [69], Guizar et al. present a system for cooperative Mbits/s), indicating that LTE V2X can support more data-
collision avoidance that involves the exchange of so-called intensive applications. Additionally, simulation results indicate
"occupancy maps" over LTE-V2X. For performance that LTE V2X consistently achieves lower latency than DSRC.
evaluation, these 700-byte messages are exchanged both using For example, even with an increase in the number of vehicles,
V2I communication with an RSU and V2V communication in LTE V2X maintains a latency below 100 ms, while DSRC's
a simulation. A large intersection in the city serves as the performance degrades significantly under congestion.
scenario, and the three transmission modes (V2I uplink, V2I T. Shimizu et al. compare DSRC and LTE-V2X PC5 Mode
downlink, V2V) are considered separately in terms of their 4 in [43] , analyzing different traffic density scenarios on a 12-
PDR. For the V2I uplink, a PDR of about 70 % and, for the V2I lane highway. The traffic density varies in the simulation from
downlink, a PDR of almost 96 % could be measured. With a 108 vehicles/km up to 423 vehicles/km at a speed of 100 km/h.
distance between the vehicles of up to 150 m, a PDR of almost LTE-V2X offers better PDR at longer distances, while DSRC
74 % was measured for V2V communication. For V2I demonstrates lower latency. This difference is explained by the
downlink communications, the latency to retrieve information two different media access methods (CSMA/CA for DSRC,
from the RSU was less than 120 ms. For V2I uplink, latency to Sensing-Based Semi Persistent Scheduling (SB-SPS) for LTE-
retrieve Cooperative Awareness Messages (CAM) was 106 ms V2X).
to 160 ms. For V2V communications, latency to share CAM V. Maglogiannis et al. [72] conduct a field test comparing
was 103 ms to 140ms for 150 m range. CBR was indirectly LTE-V2X and ITS-G5 technologies on a highway. LTE-V2X
addressed through the TTI (Transmission Time Interval) outperforms ITS-G5 in V2I communication at distances beyond
occupancy rate, which showed that MCS-15 (with 22 RBs per 200m, while ITS-G5 shows lower latency. For instance, 97% of
TTI) had fewer collisions and was more reliable under high packets transmitted by ITS-G5 have latency smaller than 5.5
vehicle densities. The authors conclude from these results that ms, while C-V2X PC5's latency for the same percentile is up to
their proposed system would be feasible using LTE-V2X 37.84 ms. In V2V communication, LTE-V2X offers longer
communication, provided that all UEs are interconnected. ranges than ITS-G5.
Ulbel investigates ITS-G5 in real road traffic in [70], where Results of another field test comparing IEEE 802.11p and
she focuses on the aspects of maximum range and connection LTE-V2X are presented by M. Shi et al. in [51]. In the field trial
quality. In her test drives, ranges of up to 3 km could be presented there, four different scenarios are considered, each
achieved for message transmission when the RSUs were with two vehicles. In the first scenario, both vehicles are at
mounted at the same height as traffic lights. Considering the distances of 100 m and 800 m from each other, while in the
link quality (signal strength min. -80 dBm), ranges of 90m second scenario, both vehicles follow each other at the same
(NLOS) – 300m (LOS) can be achieved. However, the speed at a distance of about 100 m. In scenario three, the two
connection quality was disregarded in the measurements. vehicles travel at the same speed but in opposite directions, and
Packet loss rate was significantly influenced by the line of sight. in scenario four, the two vehicles travel different routes and
Up to a distance of about 800 meters, no packet drop was then meet at an intersection at the same time. In this case, an
observed when the sender and receiver were in direct line of NLOS situation also occurs in places. All scenarios thus
sight. However, the smallest obstructions led to drastic packet represent situations with low traffic volumes. LTE-V2X
loss. demonstrates better PDR over distance, while 802.11p exhibits
 11
This work has been submitted to Engineering for possible publication.

lower latency. LOS situations favor V2V communication for collisions through centralized scheduling. Thus, the authors
safety, although collisions can still occur in NLOS situations. consider LTE-V2X to be the better technology for truck
In general, the paper indicates that both standards face platooning and high-density scenarios in general.
significant challenges in NLOS scenarios, leading to higher In [77], different MCSs for LTE-V2X and IEEE 802.11p
propagation and collision errors. LTE-V2X shows better were tried and compared. The authors were able to show that
performance in maintaining communication under these the use of different MCSs can improve the performance of IEEE
conditions compared to 802.11p. 802.11p especially for long ranges. For IEEE 802.11p, the bit
A. K. Gulia [2] evaluates the performance of the IEEE rate was improved to a range from 3 to 27 Mb/s. In contrast, the
802.11p standard and LTE-V2X in Mode 3 using a 4-lane bit rate in LTE-V2V ranges between 1.15 Mb/s and. 17.71
highway scenario with a length of 1 km and a maximum speed Mb/s. However, the authors still conclude that, while other
of 30 km/h. Furthermore, he varies the vehicle density in four metrics being equal, LTE-V2V offers a higher range and is
levels: 10, 30, 50 and 70 vehicles/lane/km. He calls these low, better suited when a high perceptual range is the goal.
moderate, high, and extremely high traffic density, respectively. In [49], the authors R. Sattiraju et al. compare LTE-V2X and
300-byte Basic Service Messages (BSM) or CAMs are sent IEEE 802.11p with respect to their link level performance of
with a frequency of 10 Hz and 20 Hz. For the IEEE standard, the physical layer (PHY) in different transmission channels. As
the default data rate of 6 Mbit/s per channel is used. Metrics a metric, they use Block Error Rate (BLER) vs. Signal-to-Noise
used include PRR and End-to-End latency. The results show Ratio (SNR). The BLER is the ratio between the number of
that IEEE 802.11p is better than LTE-V2X Mode 3 at a blocks transmitted with errors and the total number of blocks
transmission frequency of 10 Hz in almost all traffic densities. transmitted. Their results show that LTE-V2X offers
However, LTE-V2X Mode 3 has a better average PDR in significantly better performance in the physical layer than IEEE
extremely high traffic situations, even if the End-to-End latency 802.11p, and the authors also emphasize that LTE-V2X is better
suffers. suited for high vehicle speeds. The effective range is inferred
M. Karoui et al. [73] compare ITS-G5 with LTE-V2X in from the performance over different SNR levels and channel
mode 3, taking into account LTE traffic caused by applications conditions. LTE-V2X, with its better channel coding and
that are not related to V2X. Generally, ITS-G5 provides lower estimation, is likely to maintain reliable communication over
E2E delay, stable CBR, and consistent RSSI under varied longer distances compared to ITS-G5.
conditions, making it more suitable for critical ITS services that V. Mannoni et al. [42] compare LTE-V2X PC5 Mode 4 and
require lower latency. LTE-V2X Performs well under ideal ITS-G5 at both PHY and MAC levels. They find LTE-V2X
conditions but is more affected by concurrent non-ITS traffic, offers higher flexibility and better SNR performance at the
increasing density, and handover processes, leading to higher same data rate. However, LTE-V2X degrades faster with
E2E delays and variable RSSI. LTE-V2X offers broader increasing user density, making ITS-G5 preferable at densities
coverage but needs careful consideration of network load and >150 users/km². ITS-G5 exhibits lower latency over short
handover impact for reliable ITS service performance. In their distances, but LTE-V2X may have smaller latency at longer
paper, the authors also provide an overview of other works distances under certain settings. The average bit rate for ITS-
comparing C-V2X with ITS-G5 or DSRC. G5 was evaluated to be approximately 4.76 Mb/s using QPSK
In [74], the IEEE 802.11p standard and LTE-V2V are modulation a coding rate of ½. C-V2X on the other hand
compared in both modes. For this purpose, the performance supports variable bit rates. For instance, a configuration with 48
metrics PDR and update delay (UD) are used. In the paper, the Resource Blocks (RBs) and MCS 3 offers 2.4 Mb/s. Other
authors conclude that LTE-V2V in Mode 3 has better PDR than configurations can provide throughputs from 1.6 Mb/s to 6.4
IEEE 802.11p and that LTE-V2V in Mode 4 has the worst Mb/s depending on the message size and the number of RBs.
Update Delay. In another study by A. Bazzi et al. [75] the In [78], R. Roux et al. analyze the physical layer performance
authors also conclude that LTE-V2X Mode 3 is superior to both of ITS-G5 and LTE-V2X PC5 Mode 4. The paper compares
Mode 4 and IEEE 802.11p. However, IEEE 802.11p can be as ITS-G5 and LTE-V2X (PC5 mode 4) technologies focusing on
good as Mode 4 in terms of PDR and better in terms of UD up Packet Delivery Ratio (PDR), latency, and range. LTE-V2X
to a certain distance by a suitable choice of MCS. In contrast to shows superior PDR performance, achieving a range of 35
[73], the two papers [75] and [74] do not consider additional meters at a PDR of 0.9, compared to 23 meters for ITS-G5. The
occupancy of the cellular network by non-V2X applications. study highlights that LTE-V2X minimizes latency by design,
[76] investigates LTE-V2X Mode 3 and 4 and IEEE 802.11p whereas ITS-G5 can introduce significant latency under high
with respect to their suitability for truck platooning in a high- traffic conditions. Both technologies are evaluated for Channel
traffic scenario. The results show that LTE-V2X Mode 3 Busy Ratio (CBR) under high traffic, with configurations
performs better than 802.11p in terms of reliability for this use ensuring efficient congestion control. The results indicate that
case, and LTE-V2X Mode 4 performs only marginally better LTE-V2X outperforms ITS-G5 in terms of packet delivery and
than 802.11p. One possible explanation is provided by collision range, demonstrating better performance in handling
errors: IEEE 802.11p is more prone to collisions due to its propagation and collision errors.
contention-based mechanism. LTE-V2X Mode-4 also faces The authors in [38] analyze the performance of V2X
collisions but to a lesser extent. Mode-3 effectively avoids approaches regarding Ultra Reliable Low Latency
 12
This work has been submitted to Engineering for possible publication.

Communications and enhanced Mobile Broadband applications model for evaluating the performance which considers aspects
in simulation. NR-V2X outperforms all other standards in terms such as multiple access interference in vehicle platooning
of reliability, range, latency, and data rates, making it the most scenarios and places emphasis on resource allocation in LTE-
suitable for advanced V2V communications. IEEE 802.11bd V2X, which significantly influences the performance of
shows significant improvements over IEEE 802.11p with better Sidelink communication in real-world deployment scenarios.
packet error rates (PER), increased range, and higher The comparative study confirms the superior performance of
throughput. LTE-V2X, while better than IEEE 802.11p in data NR-V2X over LTE-V2X, especially in scenarios with a large
rates and reliability, is surpassed by NR-V2X and IEEE inter-vehicle distance and a high number of vehicles. The CBR
802.11bd in overall performance, particularly in latency and is indirectly evaluated through the effectiveness of the sensing-
packet reception ratio (PRR). Overall, NR-V2X and IEEE based semi-persistent scheduling (SB-SPS) scheme in
802.11bd emerge as the leading technologies, with NR-V2X allocating resources and minimizing interference. NR-V2X's
having a slight edge in data rates and latency. improved sensing mechanisms lead to better resource allocation
In [79], Zugno et al. analyze the end-to-end performance of and lower channel busy ratio compared to LTE-V2X.
NR-V2X communications at millimeter waves as a simulation The impact on co-channel coexistence between ITS-G5 and
with two vehicles. In that context, the authors investigate the LTE-V2X mode 4 is analyzed by Roux et al. in [82]. For the
network performance considering direct V2V communication evaluation of the influence of the coexistence, the authors
and the impact on antenna array, communication distance, consider only an urban scenario in a ns3-based simulation with
modulation coding scheme and others. Their results show a one 10 MHz channel shared between ITS-G5 and LTE-V2X
small latency by using mmWave with a modulation coding Mode 4. Their results show that with an equal number of 800
scheme MCS 28 in short communication distances because of ITS-G5 and LTE-V2X equipped vehicles, the ITS-G5
bad channel states. For small size packets as tested, MCS 0 communication benefits from applying the coexistence method.
results in a low latency for communication distances up to 500 LTE-V2X Mode 4 shows a degradation at the same time, as
m. Packet loss rate is analyzed through PDR, with higher losses resources for the resource selection is reserved for ITS-G5.
observed at increased distances, especially in NLOS conditions. CBR is discussed as well, noting that ITS-G5 has an average
Average latency is significantly affected by numerology and channel occupancy duration of 529 µs per CAM message,
MCS, with numerology 3 generally providing lower delays at whereas LTE-V2X (using MCS3) has an occupancy of 1500 µs
shorter distances, though latency increases with distance and per message. This higher occupancy time in LTE-V2X results
NLOS conditions. in more frequent collisions.
Todisco et al. [57] are focusing on the performance
C. Technology assessment
evaluation of the NR-V2X Mode 2 SL using an open-source
simulator. Thereby the authors show that blind retransmission Table 4 and Table 5 provide a condensed overview of the
(HARQ) is only useful at low traffic-density. In high traffic- performance comparisons between different communication
density scenarios, HARQ can further increase the network load. approaches. To summarize, this paper expands on [44] by
In addition, they showed the influence of the modulation coding adding a comparison between NR-V2X and other approaches
scheme on the possible package size and transmission distance as well as additional sources overall. While no conclusion can
at different traffic densities using a system-level simulator. be drawn for the comparison between LTE-V2X Mode 3 and
With an increasing packet size, the sub carrier spacing needs to NR-V2X due to the lack of related work, it appears as if NR-
be reduced, while the MCS needs to be increased for more V2X outperforms LTE-V2X and the IEEE standards in all low
resources per time. and most of the high traffic scenarios. However, due to different
Saad et al. compare the resource selection performance of approaches in conducting the measurements across papers, e.g.,
NR-V2X Mode 2 with LTE-V2X Mode 4 in [80] using ns-3 by measuring different KPIs, it remains challenging to create a
simulator framework. For their analysis, the authors use clear overall picture. Furthermore, while the general
Cooperative Awareness Massages (CAM) with variable size methodology of using traffic simulations for data acquisition is
according to the ETSI specification. The results show a slightly beneficial in terms of comparability between IEEE 802.11 and
decreasing packet delivery ration (PDR) for NR-V2X with an the other approaches, no conclusion can be drawn for the
increasing number of vehicles from 100 to 300 at subcarrier performance in specific V2X use cases or real world V2X
spacings of 30 kHz and 60 kHz. For LTE-V2X the PDR drops scenarios. As such, the authors follow that there still exists a
significantly from 70 % to 50 %. Furthermore, it can be seen in research gap in networking performance evaluation that is yet
the results that with an increasing number of Sidelink to be explored.
subchannels and a longer Resource Reservation Time, a higher
PDR can be achieved for NR-V2X and LTE-V2X.
Nevertheless, NR-V2X Mode 2 outperforms LTE-V2X Mode
4 in each of the executed tests.
[81] concentrates on the comparison of 5G NR-V2X
communication with LTE-V2X communication in the context
of the vehicle platooning use case. They present an analytical
 13
This work has been submitted to Engineering for possible publication.

Table 4: Literature overview about IEEE802.11 based approaches compared to Vehicle-to-Everything approaches.
Ref Method Compared Result
to
[83–88] IEEE - As vehicle speed and communication distance increase, packet delivery ratio decreases
802.11p while transmission latency remains constant. In addition, vehicle density only affects
packet delivery ratio for a given packet size in networks that are not fully connected.
[38, 41, IEEE IEEE 802.11bd consistently outperforms 802.11p in all scenarios for communication
89] 802.11 bd distance, latency, throughput, and packet delivery ratio for both small and large packet
sizes.
[38, 49, LTE-V2X LTE-V2X outperforms IEEE 802.11p in all scenarios for communication distance,
73, 90] throughput, and packet delivery ratio for both small and large packet sizes. For small
packet sizes, IEEE 802.11p has lower latency for short- and medium-range
communications. For larger packets, LTE-V2X outperforms IEEE 802.11p in terms of
latency.
[2, 75, LTE-V2X IEEE 802.11p is robust at limited distances while LTE-V2X is more reliable in mid- and
77, 91] Mode 3 long-range situation. The packet loss and the latency of IEEE 802.11p is lower than the
latency of LTE-V2X in PC5 and Uu communication.
[42–44, LTE-V2X IEEE 802.11p has a lower latency than LTE-V2X, while the PDR in LTE-V2X is higher
51, 58, Mode 4 compared to IEEE 802.11p in V2I scenarios. When vehicles are traveling in the opposite
72, 75, direction, the packet delivery ratio is higher and the latency is lower with IEEE802.11p.
76, 78, PDR for both approaches is affected by the number of vehicles and their speed in the
82, 91, communication area.
92]
[38, 93] NR-V2X NR-V2X outperforms IEEE 802.11p in all scenarios for communication distance, latency,
throughput, and packet delivery ratio for both small and large packet sizes.
NR-V2X Not available in literature
Mode 1
[94] NR-V2X NR-V2X Mode 2 outperforms IEEE 802.11p in terms of packet delivery ratio and
Mode 2 communication distance. While vehicle density reduces the PDR over the communication
distance, NR-V2X Mode 2 has little effect on the PDR. For both technologies, the PDR
decreases as the number of packets transmitted increases. NR-V2X Mode 2 is more
resilient to an increase in the number of packets transmitted.
IEEE - Not available in literature
[38, 93] 802.11bd LTE-V2X LTE-V2X outperforms IEEE 802.11bd in PDR. In small packet size scenarios, LTE-V2X
shows higher throughput. At the same time, the latency of IEEE 802.11bd is lower in short-
and medium-range scenarios. In larger packet size scenarios, 802.11bd shows higher
throughput and lower latency in near field communication.
LTE-V2X Not available in literature
Mode 3
LTE-V2X Not available in literature
Mode 4
[38, 93, NR-V2X NR-V2X outperforms IEEE 802.11bd in PDR. In small packet size scenarios, NR-V2X
95] shows higher throughput. At the same time, the latency of IEEE 802.11bd is lower in short-
and medium-range scenarios. In larger packet size scenarios, 802.11bd shows higher
throughput and lower latency in short-range communications.
NR-V2X Not available in literature
Mode 1
[39] NR-V2X No clear statement on performance. However, since NR-V2X does not require backward
Mode 2 compatibility, it can use the latest and most advanced MAC and PHY layer techniques.
 14
This work has been submitted to Engineering for possible publication.

Table 5: Literature overview about cellular-based approaches compared to Vehicle-to-Everything approaches.

Ref Method Compared Result

to
[83, 96] LTE- - Latency and Packet Delivery Ratio depend on relative vehicle velocity. Latency increases
V2X with higher speeds, while PDR decreases. Sidelink communication reduces latency and
increases PDR.
[38, 93] NR-V2X NR-V2X outperforms LTE-V2X in all scenarios in terms of communication distance,
throughput and packet delivery ratio for both small and large packet sizes. LTE-V2X has
lower latency than NR-V2X in long-distance communication scenarios with small packet
sizes.
LTE- - Not available in literature
V2X
[37, 75, Mode 3 LTE-V2X The Packet Delivery Ratio in LTE-V2X Mode 3 and Mode 4 depends on the number of
91] Mode 4 packets sent per second and the vehicle density. An increasing number of packets per
second and a higher vehicle density show that Mode 3 can maintain a high packet delivery
ratio over the communication distance. Mode 3 outperforms Mode 4 in terms of PDR in
all situations such as highway, rural or urban scenarios.
NR-V2X Not available in literature
Mode 1
NR-V2X Not available in literature
Mode 2
[56, 69, LTE- - Packet Delivery Ratio is affected by the Modulation Coding Scheme (MCS) and the
97–99] V2X number of vehicles in the communication range, resource blocks and message size. Lower
Mode 4 MCS results in higher PDR in long range situations. In high density situations, the Packet
Delivery Ratio decreases due to packet collision.
[1, 81] NR-V2X The Packet Delivery Ratio for LTE-V2X Mode 4 and NR-V2X is affected by the
modulation coding scheme, the vehicle density and the number of transmitted messages.
NR-V2X outperforms LTE-V2X Mode 4 in every situation.
NR-V2X Not available in literature
Mode 1
[80, 81, NR-V2X The Packet Delivery Ratio of both NR-V2X Mode 2 and LTE-V2X Mode 4 depends on
94] Mode 2 the vehicle velocity and the number of transmitted packets. However, NR-V2X Mode 2
outperforms LTE-V2X Mode 4 because the influence of velocity and transmission rate is
lower.
[79, 100, NR For a Modulation Coding Scheme (MCS) less than 27, vehicle velocity has no effect on
101] V2X transmission efficiency. For an MCS greater than or equal to 27, performance degrades as
velocity increases. In addition, a higher MCS improves throughput. For an MCS greater
than or equal to 27, throughput is dependent on vehicle velocity.
NR - Not available in literature
V2X
Mode 1 NR V2X Not available in literature
Mode 2
[57, 91, NR - The selection of packet generation rules from ETSI and 3GPP affects the Packet Delivery
102] V2X Ratio. Using the ETSI rules, the PDR is higher compared to using the 3GPP generation
Mode 2 rules. NR-V2X Mode 2 is affected by the numerology, the number of PSSCH
transmissions, the resource selection window, the modulation coding scheme, and the
maximum number of resources per reservation.
was the dominant technology until 2020, which slowed down
VI. GEOGRAPHICAL USAGE OF V2X both laboratory and field testing of other communication
This section introduces the worldwide distribution of IEEE standards. During this period, several DSRC field tests were
802.11 and C-V2X. The following sections are divided into conducted, such as the SPaT Challenge [103]. In SPaT, traffic
regions because some regularities are defined across countries. light cycle time was distributed using DSRC. According to the
In particular, the geographical regions of the United States of US Department of Transportation (USDoT), more than 70
America, Europe, and China are considered. DSRC applications were active in the US in 2018. [104, 105].
Due to the increasing importance of C-V2X and the low market
A. United States of Amerika penetration of DSRC, the Federal Communications
Commission (FCC), which regulates communications
In the United States, DSRC was introduced in 1999 [44]. It
frequencies, has decided to reuse parts of the 5.9 GHz spectrum
 15
This work has been submitted to Engineering for possible publication.

for other V2X-related applications in 2019 [106] and issued an focused on C-V2X. The first test of C-V2X was conducted in
order in November 2020 [107]. The frequency range between 2016 [117]. This was followed by interoperability of LTE-V2X
5850 MHz and 5895MHz is re-allocated for the expansion of devices from different manufacturers using the PC5 interface
midband spectrum operations, while the frequency range by 2018. Mass production of vehicles with V2X functionality
between 5895 MHz and 5925 MHz is now used for C-V2X began in 2020 [110]. In addition, C-V2X is considered a core
communication. In 2021, Audi started two C-V2X projects in technology to enable the development of a national traffic
the 5.9 GHz range [108] using LTE radio and the PC5 interface. control network connecting roads and vehicles using 5G
According to market analysts, even Toyota, a former advocate communication [114]. Therefore, China enacted a national
of DSRC in 2016 [19] changed his plans regarding the usage of strategy for the Internet of Vehicles (IoV) in 2021, which
C-V2X [109]. It is therefore reasonable to assume that C-V2X includes V2X communication [110]. The 20 MHz frequency
will replace DSRC in the long term [110]. Based on the current band of C-V2X is defined from 5905 MHz to 5925 MHz. In
changes, it can be expected, that the market penetration of C- general, China is already well advanced in terms of 5G network
V2X will increase until 2025. It can further be assumed that the deployment. According to Businesswire [118], by the end of
first commercial products equipped with NR-V2X will be 2021, all major cities and at least 97% of counties were covered
developed. As the publication of the next IEEE 802.11bd by 5G. The penetration of NR-V2X-enabled vehicles in 2021
standard is expected in 2030, it is likely that researchers and was less than 0.4% of the total number of vehicles with
companies also try to establish new DSRC use cases till then. communication capability but is expected to grow to 15% by
2025.
B. Europe
As the first NR-V2X equipped vehicles are already available
Instead of using DSRC, ITS-G5 technology based on IEEE in China, the fast rise of the technology is highly likely. Based
802.11p was introduced and tested in Europe. A first proposal on plans as of 2022, there shall be a market penetration of up to
from the European Union to use ITS-G5 as the main technology 50 % for C-V2X in 2030. Furthermore, it can be assumed that
was rejected in 2019 [111], where it was argued that defining a NR-V2X will be the mainly focused technology for vehicle
core technology would stifle technological progress. communication in 2030 and beyond, as the first commercial
Subsequently, the ETSI published a standard defining the C- applications shall be introduced in 2025 [118, 119].
V2X technology as the access layer for intelligent Some of the authors conclude that bandwidth and latency
transportation systems. In this standard, C-V2X communication requirements will become more relevant as new use cases for
over the PC5 interface uses the same frequencies as the ITS-G5 connected and autonomous driving, such as collective
technology [112]. For the sake of completeness, it should be perception or extended sensors, are deployed, leading to
mentioned that the coexistence of ITS-G5 and C-V2X is increased market penetration of NR-V2X by 2030 and beyond
possible. ETSI has published two technical reports in 2021 that
address the coexistence of ITS-G5 and C-V2X SL using the VII. CHALLENGES
PC5 interface [113].
V2X communication is a key technology to increase road
However, not all car manufacturers rely on ITS-G5. Market
safety and traffic efficiency [3, 72]. However, there are still
research companies consider C-V2X technology to be superior,
some open challenges pending regarding the overall vehicle
despite its possible coexistence with ITS-G5 [114]. As of 2020,
communication, including the required data rate and bandwidth
Volkswagen vehicles such as the VW Golf MK8 and VW ID
as well as the message generation rules. The following sections
models support ITS-G5. Frost & Sullivan [109] concludes that
describe these challenges in detail to clarify some open research
Volkswagen is likely to move to NR-V2X at a later stage when
areas and specific open topics from a use case specific
it is well developed and researched. This assumption can be
perspective.
supported by an analysis of recently started projects: As of 2021
To achieve the goals of autonomous driving, a detailed
[110], there are seven projects testing the viability and
perception of the environment is required [66]. This can be
performance of C-V2X in Europe. Thus, the trend towards C-
realized by data and information exchange [120]. But due to
V2X is also visible.
dynamic influences like lightning, weather and occlusion to
The mass deployment of ITS-G5 and C-V2X was planned for
participants within the environment, discrepancies in
2025, but according to recently published information to Euro
perception can occur [121], thus making it difficult to maintain
NCAP (2017) [115, 116], it is delayed to 2027. Nevertheless, it
the integrity of perceived environment (section VII.A).
is expected that Europe's technology openness will lead to an
Additionally, as several participants act in the same
almost equal distribution of communication systems using ITS-
environment at the same time, they will detect the same objects
G5 and C-V2X until 2027 and NR-V2X beyond 2030 due to the
and share them with others [122, 123]. This increases the
different interests of nations within the European Union. The
network load but does not necessarily provide any additional
technology openness allows flexible adaptation to use cases and
benefit. The rate of redundant information will increase with the
optimization of bandwidth requirements and transmission
number of participants. Thus. managing data transmission in
distance.
dynamic environments (section VII.B) is crucial. Especially in
C. China the context of V2X, various use cases (see section IV.D) that
China has never deployed DSRC. Instead, they have directly
 16
This work has been submitted to Engineering for possible publication.

Figure 5: Overview about open challenges in the field of V2X communication.

require the exchange of data and information between involved without any compression strategy, data pre-processing or
participants on different levels exist. Thus, the system intelligent data selection. The application of compression
efficiency and performance can be directly influenced by the strategies [134] or the intelligent selection of data can be helpful
communication architecture and function offloading (section to enable use cases requiring cooperative interaction (see
VII.C). Section IV.D), but still results in higher bandwidth usage
Figure 5 summarizes the above-mentioned points and (CA.2). Besides classical compression, a more efficient
clarifies the considered topics in different sections. The approach is to transmit the feature maps generated by machine
following subsections first derive the challenges (Cx.y) in the learning algorithms in connected vehicles [130], but still fail in
context of the specified topic and finally provide a summarized large environments (CA.3) due to alignment errors and
overview of open challenges. hardware and storage requirements. Since the transmitted maps
needs to be aligned and fused, the temporal and spatial domain
A. Challenge: Integrity of the perceived environment
needs to be considered [135]. Since the interaction time
To improve the efficiency and safety in the mobility sector, between several road users is short, the temporal aspect cannot
critical situations and ego-vehicles’ information needs to be be neglected, and high computational power is required to solve
transmitted between nearby participants. In addition, the task during the interaction time (CA.4). To enable
knowledge about all objects within a defined area in the cooperative use cases, information from different vehicles is to
vehicles’ environment is required. The currently established be fused. However, since sensors and detection algorithms can
V2X use cases try to fulfill these requirements through cause errors due to distance to an object, occlusion, weather
specified message formats that enable the transmission of condition and others, inconsistent information will occur, which
required information [124–127]. In this context, the vehicle needs to be resolved (CA.5).
status information or detected objects within the ego vehicle’s In conclusion, the following five main challenges are
environment are transmitted. A vehicle equipped with sensors identified regarding the integrity of the perceived environment:
and the capability to detect its environment is called ego • (CA.1) cooperative information exchange is limited by
vehicle. only considering identified objects,
To save bandwidth, aggregated information such as detected • (CA.2) the transmission of unprocessed sensor data
objects or identified situations are transmitted [126], in e.g. and feature maps results in a higher network load,
cooperative perception, 35 bytes are required to transmit an • (CA.3) cooperative SLAM and feature maps are not
object in a collective perception message. An object needs to be well appropriate for large scale maps,
identified before its transmission, so even with e.g. cooperative • (CA.4) short interaction time between participants,
perception, a driver or a vehicle can never assume that all • (CA.5) inconsistent information needs to be resolved.
present objects within a FoV are known [121]. Exchanging
identified objects within a vehicle's FoV can close blind spots B. Challenge: Managing data transmission in dynamic
caused by occlusion, but cannot increase detection confidence environments
[121, 128] (CA.1). To raise the efficiency and the confidence in Since the transmission of data and information is crucial for
the environmental model, new approaches to detect the automated and autonomous driving and it is assumed that an
surroundings in more detail are required. Possible ideas are the autonomous vehicle Is expected to generate up to 4 TB of data
transmission of sensor raw data [129] and the transmission of a day, the network resources have to be used as efficiently as
feature maps [130]. While the transmission of sensor data possible. As there are multiple vehicles perceiving a situation
results in a more detailed representation of the environment, as (and thus, the objects involved) in the same environment, data
all perception data is available, it requires high bandwidth for redundancy is to be expected according to [122, 136, 137].
data transmission at the same time [131]. As the estimated data Especially in medium- and high-density traffic scenarios, it is
generated by an autonomous vehicle is about 4 TB [132] per likely that information regarding specific objects will be
day and depending on the sensor set, between 50 MB and 70 transmitted and received several times (CB.1) [138], which
MB per second [133], this approach seems to be difficult dissipates rare network resources. This can lead to network
 17
This work has been submitted to Engineering for possible publication.

congestions, resulting in higher latency and thus delayed implement a performant communication architecture. While
message transmission. This behavior is not limited to specific some functions like cooperative mapping benefit from the
messages, but to all cyclically transmitted message packets. centralized provision of information, functions such as remote
Based on that, the Decentralized Congestion Control (DCC) driving require the instantaneous exchange of information. All
queues messages in overload situations using a First In First use cases have with specific requirements in terms of latency,
Out (FIFO) buffer [139]. According to Delooz et al. [123], this bandwidth, computational capacity and other factors [141,
mechanism can reduce network latency by reducing the number 142]. Thus, it is crucial to elaborate which use cases can be
of messages sent. However, relevant messages have to wait in applied to the different architectures and how this affects their
a queue for an undefined period of time, resulting in a delayed use case performance parameters (CC.1). The dynamic
transmission of information and, since the interaction time behavior of vehicles results in short interaction times between
between clients is short, invalid messages. Hence, the ETSI different participants, making it even more difficult to provide
specification defines the usage of four message generation rules accurate detection results and decision making through
for Cooperative Perception Messages to avoid network cooperative functions.
conjunction [124]. Due to the low market penetration of V2X- The corresponding morphological box is shown in Table 6.
equipped vehicles and roadside units, the verification of the The morphological combination of five parameters results in a
functions and the effects between them have only been tested in complex solution space that enables the determination of
simulations (CB.2). possible communication architectures through permutation.
Based on this review, it is shown that the different V2X The Main Component defines the type of nodes available in the
approaches (see Sections IV.A and IV.B) try to solve the same network, while V2X technology represents the approach itself.
problem by using different technologies and, most importantly, Considering only these two parameters, four architectures with
individually designed frequency channels, sometimes four different performances are possible. Additional parameters
overlapping with bands of other V2X approaches and protocol such as the number of participants and the communication
stacks. In a globalized world, it is likely that vehicles produced distance or the transmitted data itself will directly influence the
for the U.S., European, Chinese or any other market will be performance of the system, and thus resulting in a huge variance
driven on other continents, which would ultimately lead to a of use case specific architectures (CC.2).
loss in functionality. This is the case even though DSRC and Due to the progress in assisted and autonomous driving and
ITS-G5 are based on IEEE 802.11, since they differ in detail with respect to cooperative vehicle functions (CVF), the
like the channel’s frequency bands and different protocol complexity of vehicular functions is continuously increasing
stacks. In addition, the C-V2X Uu interface uses the same [143]. Thus, additional computational and storage resources are
frequency as the ITS-G5 (CB.3), which may result in a loss of needed to meet the requirements regarding processing time and
performance [82]. accuracy [144]. As modern vehicle functions are steadily
In conclusion, the following three main challenges are evolving, increasing computational and storage requirements
identified: are expected [143]. If these requirements cannot be fulfilled by
• (CB.1) reduce rate of information redundancy in mid- a vehicle, the latest updates or the whole functionality will not
and high-density traffic scenarios, be available for further usage [145]. CVF especially require
• (CB.2) Field tests in real environments are difficult, data exchange between all involved participants [138]. If the
• (CB.3) Continent specific and thus incompatible V2X CVF is deployed and executed locally, the same message needs
paradigms. to be transmitted between each participating vehicle. Thus,
increasing the network load and the information redundancy in
C. Challenge: Communication architecture and function
offloading the network. In this context, function offloading describes the
possibility to deploy vehicle functions to network elements
Based on the introduced use cases in section IV.D and the such as the edge, fog, or cloud, and can therefore mitigate these
available network components such as cloud, fog, edge and initial issues [143], but may result in higher functional latency
vehicle communication [140], various possibilities exist to [146]. With an increasing number of vehicle functions and their
Table 6: Possible parameters influencing use case performance complexity, it is hard to decide whether a vehicle function can
network architectures.
Parameter Characteristics be offloaded to the edge, fog, or cloud, or whether it needs to
be deployed in a vehicle (CC.3). Hence, each functionality and
Main
Cloud Edge Fog Vehicles use case requires a specific maximum latency and minimum
component
V2X technology ITS- bandwidth to maintain its functionality and a detailed
DSRC LTE-V2X NR-V2X
G5 knowledge of these thresholds is needed (CC.4). To gain the
Number of
1 10 100 1000 latter, extensive performance evaluations of networks for
participants specific use cases in real-world networks using real clients
Communication <100 while considering different Quality of Service parameters
<50 m <150 m <200 m
distance m
(QoS) are required (CC.5). Summarized, the following five
Transmitted Compressed
data objects data data or hybrid major challenges have been identified with respect to possible
features communication architectures and function offloading:
 18
This work has been submitted to Engineering for possible publication.

• (CC.1) impact of dynamic network parameters on research needs to explore how vehicle functions can be
dynamic use cases is unknown, deployed in distributed cloud/edge systems and how the
• (CC.2) huge variety on use case specific network associated data streams can be efficiently orchestrated and
topologies, managed. In this context, how to reduce the operational
• (CC.3) decision, what function can be offloaded to the complexity of such systems through continuous analysis and
network backend, optimization needs to be considered.
• (CC.4) detailed knowledge about performance
thresholds is required, REFERENCES
• (CC.5) use case specific evaluation of network [1] C. Campolo, A. Molinaro, F. Romeo, A. Bazzi, and A. O. Berthet, "5G
performance parameters in the real world necessary. NR V2X: On the Impact of a Flexible Numerology on the Autonomous
Sidelink Mode," in 2019 IEEE 2nd 5G World Forum (5GWF),
Dresden, Germany, 2019, pp. 102–107, doi:
VIII. CONCLUSION 10.1109/5GWF.2019.8911694.
In this paper, the most popular Vehicle-to-Everything [2] Aman Kumar Gulia, "A Simulation Study on the Performance
Comparison of the V2X Communication Systems: ITSG5 and C-V2X
communication standards based on IEEE 802.11 (DSRC and Communication Systems: ITS-G5 and C-V2X," Masterthesis, The
ITS-G5) and 4th and 5th generation cellular standards (C-V2X, School of Electrical Engineering and Computer Science, KTH Royal
LTE-V2X, NR-V2X) have been analyzed in terms of their Institute of Technology, Stockholm, Sweden, 2020.
[3] H. Zhou, W. Xu, J. Chen, and W. Wang, "Evolutionary V2X
performance, current and future geographical usage, and Technologies Toward the Internet of Vehicles: Challenges and
current challenges. The main findings can be summarized as Opportunities," Proc. IEEE, vol. 108, no. 2, pp. 308–323, 2020, doi:
follows: 10.1109/JPROC.2019.2961937.
1. The analysis of the existing literature shows that NR- [4] S. Chen, J. Hu, Y. Shi, L. Zhao, and W. Li, "A Vision of C-V2X:
Technologies, Field Testing, and Challenges With Chinese
V2X outperforms IEEE 802.11p and C-V2X Development," IEEE Internet Things J., vol. 7, no. 5, pp. 3872–3881,
respectively LTE-V2X in all scenarios (low- and high- 2020, doi: 10.1109/JIOT.2020.2974823.
density traffic). Compared to NR-V2X, IEEE [5] M. H. C. Garcia et al., "A Tutorial on 5G NR V2X Communications,"
IEEE Commun. Surv. Tutorials, vol. 23, no. 3, pp. 1972–2026, 2021,
802.11bd shows higher performance (latency and doi: 10.1109/COMST.2021.3057017.
throughput) in short-range communication as the [6] Falk Dettinger, Matthias Weiß, and Michael Weyrich, Eds., Future Use
packet size increases. Cases for Vehicular Communication based on Connected Functions.
2024 IEEE 100th Vehicular Technology Conference (VTC2024-Fall):
2. A gradual shift from DSRC and ITS-G5 to C-V2X by IEEE, 2024, doi: print.
2030 and NR-V2X after 2030 is expected globally. [7] R. Sedar, C. Kalalas, F. Vazquez-Gallego, L. Alonso, and J. Alonso-
3. We have identified three main challenges based on our Zarate, "A Comprehensive Survey of V2X Cybersecurity Mechanisms
and Future Research Paths," IEEE Open J. Commun. Soc., vol. 4, pp.
literature research: 325–391, 2023, doi: 10.1109/OJCOMS.2023.3239115.
a. maintaining the integrity of the perceived [8] A. Alnasser, H. Sun, and J. Jiang, "Cyber security challenges and
environment: Highly dynamic environment solutions for V2X communications: A survey," COMPUTER
NETWORKS, vol. 151, pp. 52–67, 2019, doi:
with short interaction time and heterogeneous 10.1016/[Link].2018.12.018.
clients perceiving the same environment [9] M. N.-E. Saulaiman, M. Kozlovszky, and A. Csilling, "A Survey on
from different perspectives. Vulnerabilities and Classification of Cyber-Attacks on 5G-V2X," in
b. managing cooperative data transmission in 2021 IEEE 21st International Symposium on Computational
Intelligence and Informatics (CINTI), Budapest, Hungary, 2021, pp.
dynamic environments: Optimal decision 235–240, doi: 10.1109/cinti53070.2021.9668440.
making requires large amounts of data, but [10] T. Yoshizawa et al., "A Survey of Security and Privacy Issues in V2X
because of a), data redundancy, Communication Systems," ACM Comput. Surv., vol. 55, no. 9, pp. 1–
36, 2023, doi: 10.1145/3558052.
contradictions and incompatible protocols [11] J. Huang, D. Fang, Y. Qian, and R. Q. Hu, "Recent Advances and
must be considered. Challenges in Security and Privacy for V2X Communications," IEEE
c. possible communication architectures and Open J. Veh. Technol., vol. 1, pp. 244–266, 2020, doi:
10.1109/OJVT.2020.2999885.
function offloading: The execution of [12] F. A. Schiegg, I. Llatser, D. Bischoff, and G. Volk, "Collective
cooperative tasks requires a stable network Perception: A Safety Perspective," Sensors, early access. doi:
connection with respect to specific quality of 10.3390/s21010159.
[13] I. Llatser, T. Michalke, M. Dolgov, F. Wildschutte, and H. Fuchs,
service requirements in real-world "Cooperative Automated Driving Use Cases for 5G V2X
environments. Communication," in 2019, doi: 10.1109/5GWF.2019.8911628.
Our analysis shows that NR-V2X currently has the highest [14] L. Hobert, A. Festag, I. Llatser, L. Altomare, F. Visintainer, and A.
Kovacs, "Enhancements of V2X communication in support of
performance. However, it is also clear that IEEE 802.11bd can cooperative autonomous driving," IEEE Commun. Mag., vol. 53, no.
provide an advantage in certain situations, but this is not yet of 12, pp. 64–70, 2015, doi: 10.1109/MCOM.2015.7355568.
great importance due to the late expected release period. Also [15] F. A. Schiegg, N. Brahmi, and I. Llatser, "Analytical Performance
due to the increasing degree of autonomy of vehicles, it is Evaluation of the Collective Perception Service in C-V2X Mode 4
Networks," in 2019 IEEE Intelligent Transportation Systems
expected that the focus in the near future will be on the use of Conference (ITSC), Auckland, New Zealand, uuuu-uuuu, pp. 181–188,
NR-V2X. doi: 10.1109/ITSC.2019.8917214.
To enable the use of V2X in different scenarios, function [16] C. R. Storck and F. Duarte-Figueiredo, "A Survey of 5G Technology
Evolution, Standards, and Infrastructure Associated With Vehicle-to-
offloading and data flow optimization are crucial. Future Everything Communications by Internet of Vehicles," IEEE Access,
 19
This work has been submitted to Engineering for possible publication.

vol. 8, pp. 117593–117614, 2020, doi: [35] H. Snyder, "Literature review as a research methodology: An overview
10.1109/ACCESS.2020.3004779. and guidelines," Journal of Business Research, vol. 104, pp. 333–339,
[17] H. Abou-zeid, F. Pervez, A. Adinoyi, M. Aljlayl, and H. 2019, doi: 10.1016/[Link].2019.07.039.
Yanikomeroglu, "Cellular V2X Transmission for Connected and [36] R. Perez, F. Schubert, R. Rasshofer, and E. Biebl, "Deep Learning
Autonomous Vehicles Standardization, Applications, and Enabling Radar Object Detection and Classification for Urban Automotive
Technologies," IEEE Consumer Electron. Mag., vol. 8, no. 6, pp. 91– Scenarios," in 2019 Kleinheubach Conference.
98, 2019, doi: 10.1109/MCE.2019.2941467. [37] E. E. González, D. Garcia-Roger, and J. F. Monserrat, "LTE/NR V2X
[18] K. Garlichs, H. Günther, and L. C. Wolf, "Generation Rules for the Communication Modes and Future Requirements of Intelligent
Collective Perception Service," in 2019 IEEE Vehicular Networking Transportation Systems Based on MR-DC Architectures,"
Conference (VNC), 2019, pp. 1–8, doi: Sustainability, vol. 14, no. 7, p. 3879, 2022. doi: 10.3390/su14073879.
10.1109/VNC48660.2019.9062827. [Online]. Available: [Link]
[19] M. Boban, A. Kousaridas, K. Manolakis, J. Eichinger, and W. Xu, [38] W. Anwar, N. Franchi, and G. Fettweis, "Physical Layer Evaluation of
"Connected Roads of the Future: Use Cases, Requirements, and Design V2X Communications Technologies: 5G NR-V2X, LTE-V2X, IEEE
Considerations for Vehicle-to-Everything Communications," IEEE 802.11bd, and IEEE 802.11p," in 2019 IEEE 90th Vehicular
Vehicular Technology Magazine, vol. 13, no. 3, pp. 110–123, 2018, Technology Conference (VTC2019-Fall): Proceedings : Honolulu,
doi: 10.1109/MVT.2017.2777259. Hawaii, USA 22-25 September 2019, Honolulu, HI, USA, 2019, pp. 1–
[20] S. Sun, J. Hu, Y. Peng, X. Pan, L. Zhao, and J. Fang, "Support for 7, doi: 10.1109/VTCFall.2019.8891313.
vehicle-to-everything services based on LTE," IEEE Wireless [39] G. Naik, B. Choudhury, and J.-M. Park, "IEEE 802.11bd & 5G NR
Commun., vol. 23, no. 3, pp. 4–8, 2016, doi: V2X: Evolution of Radio Access Technologies for V2X
10.1109/MWC.2016.7498068. Communications," IEEE Access, vol. 7, pp. 70169–70184, 2019, doi:
[21] V. V. Chetlur and H. S. Dhillon, "Coverage and Rate Analysis of 10.1109/ACCESS.2019.2919489.
Downlink Cellular Vehicle-to-Everything (C-V2X) Communication," [40] F. Arena, G. Pau, and A. Severino, "A Review on IEEE 802.11p for
IEEE Trans. Wireless Commun., vol. 19, no. 3, pp. 1738–1753, 2020, Intelligent Transportation Systems," JSAN, vol. 9, no. 2, p. 22, 2020,
doi: 10.1109/TWC.2019.2957222. doi: 10.3390/jsan9020022.
[22] E. Uhlemann, "Initial Steps Toward a Cellular Vehicle-to-Everything [41] B. Y. Yacheur, T. Ahmed, and M. Mosbah, "Analysis and Comparison
Standard [Connected Vehicles]," IEEE Vehicular Technology of IEEE 802.11p and IEEE 802.11bd," in Communication
Magazine, vol. 12, no. 1, pp. 14–19, 2017, doi: Technologies for Vehicles: 15th International Workshop,
10.1109/MVT.2016.2641139. Nets4Cars/Nets4Trains/Nets4Aircraft 2020, Bordeaux, France,
[23] N. Raza, S. Jabbar, J. Han, and K. Han, "Social vehicle-to-everything November 16–17, 2020, Proceedings (Springer eBook Collection
(V2X) communication model for intelligent transportation systems 12574), F. Krief, H. Aniss, L. Mendiboure, S. Chaumette, and M.
based on 5G scenario," in Proceedings of the 2nd International Berbineau, Eds., 1st ed. Cham: Springer International Publishing;
Conference on Future Networks and Distributed Systems, Amman Imprint Springer, 2020, pp. 55–65.
Jordan, A. Abuarqoub, B. Adebisi, M. Hammoudeh, S. Murad, and M. [42] V. Mannoni, V. Berg, S. Sesia, and E. Perraud, "A Comparison of the
Arioua, Eds., 2018, pp. 1–8, doi: 10.1145/3231053.3231120. V2X Communication Systems: ITS-G5 and C-V2X," in 2019 IEEE
[24] M. Hasan, S. Mohan, T. Shimizu, and H. Lu, "Securing Vehicle-to- 89th Vehicular Technology Conference (VTC Spring): Proceedings :
Everything (V2X) Communication Platforms," IEEE Trans. Intell. Kuala Lumpur, Malaysia, 28 April-1 May 2019, Kuala Lumpur,
Veh., vol. 5, no. 4, pp. 693–713, 2020, doi: Malaysia, 2019, pp. 1–5, doi: 10.1109/VTCSpring.2019.8746562.
10.1109/TIV.2020.2987430. [43] T. Shimizu, H. Lu, J. Kenney, and S. Nakamura, "Comparison of
[25] A. Alalewi, I. Dayoub, and S. Cherkaoui, "On 5G-V2X Use Cases and DSRC and LTE-V2X PC5 Mode 4 Performance in High Vehicle
Enabling Technologies: A Comprehensive Survey," IEEE Access, vol. Density Scenarios," in 26th ITS World Congress, Singapore, 2019.
9, pp. 107710–107737, 2021, doi: 10.1109/ACCESS.2021.3100472. [Online]. Available: [Link]
[26] R. Khezri, D. Steen, and L. A. Tuan, "A Review on Implementation of 336768425_Comparison_of_DSRC_and_LTE-V2X_PC5_Mode_4_
Vehicle to Everything (V2X): Benefits, Barriers and Measures," in Performance_in_High_Vehicle_Density_Scenarios
2022 IEEE PES Innovative Smart Grid Technologies Conference [44] K. Ansari, "Joint use of DSRC and C‐V2X for V2X communications in
Europe (ISGT-Europe), Novi Sad, Serbia, 2022, pp. 1–6, doi: the 5.9 GHz ITS band," IET intell. transp. syst, vol. 15, no. 2, pp. 213–
10.1109/ISGT-Europe54678.2022.9960673. 224, 2021. doi: 10.1049/itr2.12015. [Online]. Available: https://
[27] S. Gyawali, S. Xu, Y. Qian, and R. Q. Hu, "Challenges and Solutions [Link]/doi/pdf/10.1049/itr2.12015
for Cellular Based V2X Communications," IEEE Communications [45] ETSI EN 302 571: Intelligent Transport Systems (ITS);
Surveys & Tutorials, vol. 23, no. 1, pp. 222–255, 2021, doi: Radiocommunications equipment operating in the 5 855 MHz to 5 925
10.1109/COMST.2020.3029723. MHz frequency band; Harmonised Standard covering the essential
[28] J. Wang, Y. Shao, Y. Ge, and R. Yu, "A Survey of Vehicle to requirements of article 3.2 of Directive 2014/53/EU, ETSI EN 302
Everything (V2X) Testing," Sensors, early access. doi: 571, Feb. 2017. [Online]. Available: [Link]
10.3390/s19020334. en/302500_302599/302571/02.01.01_60/en_302571v020101p.pdf
[29] I. Soto, M. Calderon, O. Amador, and M. Urueña, "A survey on road [46] K. Kiela et al., "Review of V2X–IoT Standards and Frameworks for
safety and traffic efficiency vehicular applications based on C-V2X ITS Applications," Applied Sciences, vol. 10, no. 12, p. 4314, 2020,
technologies," Vehicular Communications, vol. 33, p. 100428, 2022, doi: 10.3390/app10124314.
doi: 10.1016/[Link].2021.100428. [47] "Next Generation V2X – IEEE 802.11bd as fully backward compatible
[30] A. Nair and S. Tanwar, "Resource allocation in V2X communication: evolution of IEEE 802.11p," CAR 2 CAR Communication
State-of-the-art and research challenges," PHYSICAL Consortium, Feb. 2023. Accessed: Jul. 15, 2024. [Online]. Available:
COMMUNICATION, vol. 64, p. 102351, 2024, doi: [Link]
10.1016/[Link].2024.102351. C2CCC_WP_2098_IEEE_802.11bd_TheV2XEvolution_V1.[Link]
[31] I. Brahmi, H. Koubaa, and F. Zarai, "Resource allocation for Vehicle‐ [48] V. Torgunakov, V. Loginov, and E. Khorov, "A Study of Channel
to‐Everything communications: A survey," IET Networks, vol. 12, no. Bonding in IEEE 802.11bd Networks," IEEE Access, vol. 10, pp.
3, pp. 98–121, 2023, doi: 10.1049/ntw2.12078. 25514–25533, 2022, doi: 10.1109/ACCESS.2022.3155814.
[32] A. Masmoudi, K. Mnif, and F. Zarai, "A Survey on Radio Resource [49] R. Sattiraju, D. Wang, A. Weinand, and H. D. Schotten, "Link Level
Allocation for V2X Communication," Wireless Communications and Performance Comparison of C-V2X and ITS-G5 for Vehicular
Mobile Computing, vol. 2019, pp. 1–12, 2019, doi: Channel Models," in 2020 IEEE 91st Vehicular Technology
10.1155/2019/2430656. Conference (VTC Spring): Proceedings : Antwerp, Belgium, 25-28
[33] IEEE Author Center Journals. "Abstracting & Indexing (A&I) May 2020, Antwerp, Belgium, 2020, pp. 1–7, doi: 10.1109/VTC2020-
Databases - IEEE Author Center Journals." Accessed: Feb. 8, 2024. Spring48590.2020.9129366.
[34] M. Templier and G. Paré, "A Framework for Guiding and Evaluating [50] "C-ITS FAQs: When will drivers benefit from cooperative V2X on
Literature Reviews," CAIS, vol. 37, 2015, doi: 10.17705/1CAIS.03706. European roads?" Accessed: May 24, 2022. [Online]. Available:
[Link]
 20
This work has been submitted to Engineering for possible publication.

[51] M. Shi, Y. Zhang, D. Yao, and C. Lu, "Application-oriented (VTC2021-Spring), 2021, pp. 1–5, doi: 10.1109/VTC2020-
performance comparison of 802.11p andLTE-V in a V2V Fall49728.2020.9348621.
communication system," Tinshhua Sci. Technol., vol. 24, no. 2, pp. [70] Andrea Ulbel, "Analysis of V2X Performance and Rollout Status with
123–133, 2019, doi: 10.26599/TST.2018.9010075. a Special Focus on Austria," Masterarbeit, Institute of Automation and
[52] B. Fernandes, J. Rufino, M. Alam, and J. Ferreira, "Implementation Control, Graz Univerity of Technology, Graz, 2021. [Online].
and Analysis of IEEE and ETSI Security Standards for Vehicular Available: [Link]
Communications," Mobile Netw Appl, vol. 23, no. 3, pp. 469–478, Analysis_of_V2X_Performance_and_Rollout_Status_with_a_Special_
2018, doi: 10.1007/s11036-018-1019-x. Focus_on_Austria
[53] M. N. Tahir and M. Katz, "Heterogeneous (ITS-G5 and 5G) Vehicular [71] R. Molina-Masegosa, J. Gozalvez, and M. Sepulcre, "Comparison of
Pilot Road Weather Service Platform in a Realistic Operational IEEE 802.11p and LTE-V2X: An Evaluation With Periodic and
Environment," Sensors, early access. doi: 10.3390/s21051676. Aperiodic Messages of Constant and Variable Size," IEEE Access, vol.
[54] M. Harounabadi, D. M. Soleymani, S. Bhadauria, M. Leyh, and E. 8, pp. 121526–121548, 2020, doi: 10.1109/ACCESS.2020.3007115.
Roth-Mandutz, "V2X in 3GPP Standardization: NR Sidelink in [72] V. Maglogiannis, D. Naudts, S. Hadiwardoyo, D. van den Akker, J.
Release-16 and Beyond," IEEE Comm. Stand. Mag., vol. 5, no. 1, pp. Marquez-Barja, and I. Moerman, "Experimental V2X Evaluation for
12–21, 2021, doi: 10.1109/MCOMSTD.001.2000070. C-V2X and ITS-G5 Technologies in a Real-Life Highway
[55] K. Ganesan, P. B. Mallick, J. Lohr, D. Karampatsis, and A. Kunz, "5G Environment," IEEE Trans. Netw. Serv. Manage., p. 1, 2021, doi:
V2X Architecture and Radio Aspects," in 2019 IEEE Conference on 10.1109/TNSM.2021.3129348.
Standards for Communications and Networking (CSCN): 2019 IEEE [73] M. Karoui, A. Freitas, and G. Chalhoub, "Performance comparison
Conference on Standards for Communications and Networking between LTE-V2X and ITS-G5 under realistic urban scenarios," in
(CSCN) took place 28-30 October 2019 in Granada, Spain, Granada, 2020 IEEE 91st Vehicular Technology Conference (VTC Spring):
Spain, 2019, pp. 1–6, doi: 10.1109/CSCN.2019.8931319. Proceedings : Antwerp, Belgium, 25-28 May 2020, Antwerp, Belgium,
[56] K. Sehla, T. M. T. Nguyen, G. Pujolle, and P. B. Velloso, "Resource 2020, pp. 1–7, doi: 10.1109/VTC2020-Spring48590.2020.9129423.
Allocation Modes in C-V2X: From LTE-V2X to 5G-V2X," IEEE [74] G. Cecchini, A. Bazzi, B. M. Masini, and A. Zanella, "Performance
Internet Things J., p. 1, 2022, doi: 10.1109/JIOT.2022.3159591. comparison between IEEE 802.11p and LTE-V2V in-coverage and
[57] Z. Ali, S. Lagen, L. Giupponi, and R. Rouil, "3GPP NR V2X Mode 2: out-of-coverage for cooperative awareness," in 2017 IEEE Vehicular
Overview, Models and System-Level Evaluation," IEEE Access, vol. 9, Networking Conference (VNC): 27-29 Nov. 2017, Torino, O. Altintas,
pp. 1–26, 2021, doi: 10.1109/ACCESS.2021.3090855. C. Casetti, N. Kirsch, R. Lo Cigno, and R. Meireles, Eds., 2017, pp.
[58] K. Jellid and T. Mazri, "DSRC vs LTE V2X for Autonomous Vehicle 109–114, doi: 10.1109/VNC.2017.8275637.
Connectivity," in Innovations in Smart Cities Applications Volume 4: [75] A. Bazzi, G. Cecchini, M. Menarini, B. M. Masini, and A. Zanella,
The Proceedings of the 5th International Conference on Smart City "Survey and Perspectives of Vehicular Wi-Fi versus Sidelink Cellular-
Applications (Springer eBook Collection 183), M. Ben Ahmed, İ. V2X in the 5G Era," Future Internet, vol. 11, no. 6, p. 122, 2019, doi:
Rakıp Karaș, D. Santos, O. Sergeyeva, and A. A. Boudhir, Eds., 1st ed. 10.3390/fi11060122.
Cham: Springer International Publishing; Imprint Springer, 2021, pp. [76] V. Vukadinovic et al., "3GPP C-V2X and IEEE 802.11p for Vehicle-
381–394. to-Vehicle communications in highway platooning scenarios," Ad Hoc
[59] R. Molina-Masegosa and J. Gozalvez, "LTE-V for Sidelink 5G V2X Networks, vol. 74, pp. 17–29, 2018. doi: 10.1016/[Link].2018.03.004.
Vehicular Communications: A New 5G Technology for Short-Range [Online]. Available: [Link]
Vehicle-to-Everything Communications," IEEE Vehicular Technology S157087051830057X
Magazine, vol. 12, no. 4, pp. 30–39, 2017, doi: [77] A. Bazzi, B. M. Masini, A. Zanella, and I. Thibault, "On the
10.1109/MVT.2017.2752798. Performance of IEEE 802.11p and LTE-V2V for the Cooperative
[60] M. M. Saad, M. A. Tariq, J. Seo, and D. Kim, "An Overview of 3GPP Awareness of Connected Vehicles," IEEE Transactions on Vehicular
Release 17 & 18 Advancements in the Context of V2X Technology," Technology, vol. 66, no. 11, pp. 10419–10432, 2017, doi:
in 2023 International Conference on Artificial Intelligence in 10.1109/TVT.2017.2750803.
Information and Communication (ICAIIC), Bali, Indonesia, 2023, pp. [78] P. Roux, S. Sesia, V. Mannoni, and E. Perraud, "System Level
57–62, doi: 10.1109/ICAIIC57133.2023.10067121. Analysis for ITS-G5 and LTE-V2X Performance Comparison," in
[61] X. Lin, "An Overview of 5G Advanced Evolution in 3GPP Release 2019 IEEE 16th International Conference on Mobile Ad Hoc and
18," IEEE Comm. Stand. Mag., vol. 6, no. 3, pp. 77–83, 2022, doi: Sensor Systems (MASS), Monterey, CA, USA, 2019, pp. 1–9, doi:
10.1109/MCOMSTD.0001.2200001. 10.1109/MASS.2019.00010.
[62] M. Wang et al., "Comparison of LTE and DSRC-Based Connectivity [79] T. Zugno, M. Drago, M. Giordani, M. Polese, and M. Zorzi, "NR V2X
for Intelligent Transportation Systems," in 2017 IEEE 85th Vehicular Communications at Millimeter Waves: An End-to-End Performance
Technology Conference (VTC Spring), 2017, pp. 1–5, doi: Evaluation," in GLOBECOM 2020 - 2020 IEEE Global
10.1109/VTCSpring.2017.8108284. Communications Conference, Taipei, Taiwan, uuuu-uuuu, pp. 1–6, doi:
[63] 3GPP. "Release 17." Accessed: Dec. 22, 2023. [Online]. Available: 10.1109/GLOBECOM42002.2020.9348259.
[Link] [80] M. M. Saad, M. T. R. Khan, S. H. A. Shah, and D. Kim,
[64] 3GPP. "Release 18." Accessed: Dec. 22, 2023. [Online]. Available: "Advancements in Vehicular Communication Technologies: C-V2X
[Link] and NR-V2X Comparison," IEEE Commun. Mag., vol. 59, no. 8, pp.
[65] W. Lei et al., "NR V2X Sidelink Design," in 5G System Design: An 107–113, 2021, doi: 10.1109/MCOM.101.2100119.
End to End Perspective (Wireless Networks), W. Lei et al., Eds., [81] A. Rehman, R. Valentini, E. Cinque, P. Di Marco, and F. Santucci,
Cham: Springer International Publishing, 2021, pp. 413–514. "On the Impact of Multiple Access Interference in LTE-V2X and NR-
[66] M. A. Khan et al., "Level-5 Autonomous Driving—Are We There V2X Sidelink Communications," Sensors, early access. doi:
Yet? A Review of Research Literature," ACM Comput. Surv., vol. 55, 10.3390/s23104901.
no. 2, pp. 1–38, 2022, doi: 10.1145/3485767. [82] P. Roux and V. Mannoni, "Performance evaluation for co-channel
[67] T. T. Thanh Le and S. Moh, "Comprehensive Survey of Radio coexistence between ITS-G5 and LTE-V2X," in 2021 IEEE 93rd
Resource Allocation Schemes for 5G V2X Communications," IEEE Vehicular Technology Conference (VTC2021-Spring), 2021, pp. 1–5,
Access, vol. 9, pp. 123117–123133, 2021, doi: doi: 10.1109/VTC2020-Fall49728.2020.9348517.
10.1109/ACCESS.2021.3109894. [83] F. Abbas and P. Fan, "A Hybrid Low-Latency D2D Resource
[68] E. Cinque, F. Valentini, A. Persia, S. Chiocchio, F. Santucci, and M. Allocation Scheme Based on Cellular V2X Networks," in 2018 IEEE
Pratesi, "V2X Communication Technologies and Service Requirements International Conference on Communications Workshops (ICC
for Connected and Autonomous Driving," in 2020 AEIT International Workshops), Kansas City, MO, USA, 2018, pp. 1–6, doi:
Conference of Electrical and Electronic Technologies for Automotive 10.1109/ICCW.2018.8403512.
(AEIT AUTOMOTIVE), Turin, Italy, uuuu-uuuu, pp. 1–6, doi: [84] C. A. Grazia, "On the Performance of IEEE 802.11p Outside the
10.23919/AEITAUTOMOTIVE50086.2020.9307388. Context of a BSS Networks," in 2018 IEEE 29th Annual International
[69] A. Guizar, V. Mannoni, F. Poli, B. Denis, and V. Berg, "LTE-V2X Symposium on Personal, Indoor and Mobile Radio Communications
performance evaluation for cooperative collision avoidance (CoCA) (PIMRC), Bologna, Italy, 2018, pp. 1388–1393, doi:
systems," in 2021 IEEE 93rd Vehicular Technology Conference 10.1109/PIMRC.2018.8580785.
 21
This work has been submitted to Engineering for possible publication.

[85] S. H. Lim, Y. K. Chia, and L. Wynter, "Accurate and cost-effective September 2019, Honolulu, HI, USA, 2019, pp. 1–5, doi:
traffic information acquisition using adaptive sampling: Centralized 10.1109/VTCFall.2019.8891534.
and V2V schemes," Transportation Research Part C: Emerging [100] J. Kim, G. Noh, T. Kim, H. Chung, and I. Kim, "Link-Level
Technologies, vol. 94, pp. 99–120, 2018, doi: Performance Evaluation of mmWave 5G NR Sidelink
10.1016/[Link].2017.10.017. Communications," in 2021 International Conference on Information
[86] K. Thomas, H. Fouchal, S. Cormier, and F. Rousseaux, "C-ITS and Communication Technology Convergence (ICTC), Jeju Island,
Communications based on BLE Messages," in GLOBECOM 2020 - Korea, Republic of, 2021, pp. 1482–1485, doi:
2020 IEEE Global Communications Conference, Taipei, Taiwan, 10.1109/ICTC52510.2021.9621128.
uuuu-uuuu, pp. 1–7, doi: 10.1109/GLOBECOM42002.2020.9322076. [101] I. Serunin, A. Pudeev, J.-Y. Hwang, S.-W. Lee, and A. Maltsev, "An
[87] M. Sepulcre, J. Gozalvez, G. Thandavarayan, B. Coll-Perales, J. Overview and Performance Analysis of CQI Reporting in 5G NR
Schindler, and M. Rondinone, "On the Potential of V2X Message Sidelink," in 2022 13th International Conference on Information and
Compression for Vehicular Networks," IEEE Access, vol. 8, pp. Communication Technology Convergence (ICTC), Jeju Island, Korea,
214254–214268, 2020, doi: 10.1109/ACCESS.2020.3041688. Republic of, 2022, pp. 34–39, doi: 10.1109/ICTC55196.2022.9952605.
[88] C. Bin Ali Wael, N. Armi, A. Mitayani, D. Kurniawan, A. Suryadi [102] C. Campolo, V. Todisco, A. Molinaro, A. Berthet, S. Bartoletti, and A.
Satyawan, and A. Subekti, "Analysis of IEEE 802.11p MAC Protocol Bazzi, "Improving Resource Allocation for beyond 5G V2X Sidelink
for Safety Message Broadcast in V2V Communication," in "Fostering Connectivity," in 2021 55th Asilomar Conference on Signals, Systems,
innovation through ICTs for sustainable smart society": Proceeding : and Computers, Pacific Grove, CA, USA, 2021, pp. 55–60, doi:
2020 International Conference on Radar, Antenna, Microwave, 10.1109/IEEECONF53345.2021.9723155.
Electronics and Telecommunications : virtual conference, 18-20 [103] A. V. Olteanu and M. Nicolae, "Using advanced V2X communication
November 2020, Tangerang, Indonesia, 2020, pp. 320–324, doi: technologies in self-organized VANETs," in 2021 23rd International
10.1109/ICRAMET51080.2020.9298654. Conference on Control Systems and Computer Science (CSCS), 2021,
[89] R. Jacob, W. Anwar, N. Schwarzenberg, N. Franchi, and G. Fettweis, pp. 254–259, doi: 10.1109/CSCS52396.2021.00049.
"System-level Performance Comparison of IEEE 802.11p and [104] United States Department of Transportation. "Intelligent
802.11bd Draft in Highway Scenarios," in 2020 27th International Transportation Systems - Communications." Accessed: Jun. 21, 2022.
Conference on Telecommunications (ICT), Bali, Indonesia, 2020, pp. [Online]. Available: [Link]
1–6, doi: 10.1109/ICT49546.2020.9239538. [105] Y. Yang and K. Hua, "Emerging Technologies for 5G-Enabled
[90] J. Hu et al., "Link level performance comparison between LTE V2X Vehicular Networks," IEEE Access, vol. 7, pp. 181117–181141, 2019,
and DSRC," J. Commun. Inf. Netw., vol. 2, no. 2, pp. 101–112, 2017. doi: 10.1109/ACCESS.2019.2954466.
doi: 10.1007/s41650-017-0022-x. [Online]. Available: https:// [106] L. Miao, S.-F. Chen, Y.-L. Hsu, and K.-L. Hua, "How Does C-V2X
[Link]/article/10.1007/s41650-017-0022-x Help Autonomous Driving to Avoid Accidents?," Sensors (Basel,
[91] H. Xing, B. Cimoli, I. Passchier, G. Kakes, V. Ho, and H. Nijmeijer, Switzerland), early access. doi: 10.3390/s22020686.
"Practical Challenges in CACC Communication: ITS G5, LTE Uu, and [107] Federal Communications Commision, FCC Modernizes 5.9 GHz Band
LTE Sidelink PC5," in 2021 European Control Conference (ECC), For Wi-Fi And Auto Safety, 2020. Accessed: Aug. 18, 2022. [Online].
Delft, Netherlands, 2021, pp. 1795–1801, doi: Available: [Link]
10.23919/ECC54610.2021.9655171. band-improve-wi-fi-and-automotive-safety
[92] J. Zhao, X. Gai, and X. Luo, "Performance Comparison of Vehicle [108] Audi. "Audi Newsroom." Accessed: Jun. 21, 2022. [Online].
Networking Based on DSRC and LTE Technology," in 2021 6th Available: [Link]
International Conference on Intelligent Transportation Engineering [109] "Strategic Analysis of the Global Vehicle-to-Everything (V2X)
(ICITE 2021) (Lecture Notes in Electrical Engineering), Z. Zhang, Ed., Market, Forecast to 2025," Frost & Sullivan, Rep. ME50-18, Aug.
Singapore: Springer Nature Singapore, 2022, pp. 730–746. 2020.
[93] W. Anwar, S. Dev, A. Kumar, N. Franchi, and G. Fettweis, "PHY [110] 5G Americas, Ed., "Vehicular Connectivity: C-V2X & 5G: A 5G
Abstraction Techniques for V2X Enabling Technologies: Modeling Americas White Paper," Sep. 2021. Accessed: Jun. 18, 2022. [Online].
and Analysis," IEEE Trans. Veh. Technol., vol. 70, no. 2, pp. 1501– Available: [Link]
1517, 2021, doi: 10.1109/TVT.2021.3053425. [Link]
[94] M. Montaño, R. Cajo, and F. Novillo, "Resource Allocation in C-V2X [111] Ofinno. "5G Cellular Technology for Connected Cars Receives Boost
and DSRC Technologies: Analysis and Simulation-based Evaluation from European Legislators - Ofinno." Accessed: Jun. 23, 2022.
for V2V Direct Vehicular Communication," in Proceedings of the Int'l [Online]. Available: [Link]
ACM Symposium on Design and Analysis of Intelligent Vehicular for-connected-cars-receives-boost-from-european-legislators/
Networks and Applications, Montreal Quebec Canada, R. Coutinho and [112] Intelligent Transport Systems (ITS); LTE-V2X Access layer
M. Bani Younes, Eds., 2023, pp. 115–122, doi: specification for Intelligent Transport Systems operating in the 5 GHz
10.1145/3616392.3623415. frequency band, European Telecommunications Standards Institute.
[95] A. V. Pestryakov, E. R. Khasianova, and S. I. Dinges, "Analysis of [Online]. Available: [Link]
Cellular Vehicle-to-Everything Physical Layer Parameters," in 2021 303699/303613/01.01.01_30/en_303613v010101v.pdf
Systems of Signals Generating and Processing in the Field of on Board [113] C2C-CC, Ed., "ITS-G5 and Sidelink LTE-V2X Co-Channel
Communications, Moscow, Russia, 2021, pp. 1–4, doi: Coexistence Mitigation Methods," Apr. 2021. Accessed: Jun. 18, 2022.
10.1109/IEEECONF51389.2021.9416036. [Online]. Available: [Link]
[96] N. Bonjorn, F. Foukalas, and P. Pop, "Enhanced 5G V2X services General_Documents/C2CCC_WP_2091_Co-ChannelCoexistence_
using sidelink device-to-device communications," in 2018 17th Annual MitigationMethods_V1.[Link]
Mediterranean Ad Hoc Networking Workshop (Med-Hoc-Net), Capri, [114] M. Demler. "C-V2X drives intelligent transportation." [Online].
2018, pp. 1–7, doi: 10.23919/MedHocNet.2018.8407085. Available: [Link]
[97] H. Su, M.-S. Pan, and S.-W. Kao, "A Subchannel Collision Reduction [Link]
Method for 3GPP C- V2X Mode 4 Network Based on Vehicular [115] "Euro NCAP 2025 Roadmap: In pursuit of vision zero," Euro NCAP,
Moving Status," in 2022 23rd Asia-Pacific Network Operations and [Link].2017. Accessed: Oct. 20, 2023. [Online]. Available: https://
Management Symposium (APNOMS), Takamatsu, Japan, 2022, pp. 1– [Link]/media/30700/[Link]
6, doi: 10.23919/APNOMS56106.2022.9919919. [116] "Euro NCAP Vision 2030: A safer future for mobility," Euro NCAP,
[98] J. Yin and S.-H. Hwang, "Variable MCS method for LTE V2V Nov. 2022. Accessed: Oct. 20, 2023. [Online]. Available: https://
Mode4," in 2021 International Conference on Information and [Link]/media/74468/[Link]
Communication Technology Convergence (ICTC), Jeju Island, Korea, [117] L. Miao, J. J. Virtusio, and K.-L. Hua, "PC5-Based Cellular-V2X
Republic of, 2021, pp. 1368–1370, doi: Evolution and Deployment," Sensors (Basel, Switzerland), early
10.1109/ICTC52510.2021.9620930. access. doi: 10.3390/s21030843.
[99] F. Eckermann, M. Kahlert, and C. Wietfeld, "Performance Analysis of [118] ResearchAndMarkets. "China Automotive and 5G Industry Integration
C-V2X Mode 4 Communication Introducing an Open-Source C-V2X Development Report 2022: Application of 5G-V2X Will Promote the
Simulator," in 2019 IEEE 90th Vehicular Technology Conference Realization of High-level Autonomous Driving -
(VTC2019-Fall): Proceedings : Honolulu, Hawaii, USA 22-25 [Link]." Accessed: Jun. 21, 2022. [Online].
 22
This work has been submitted to Engineering for possible publication.

Available: [Link] Berichte 2419), Düsseldorf: VDI Verlag; VDI eLibrary, 2023, pp. 513–
20220523005608/en/China-Automotive-and-5G-Industry-Integration- 524.
Development-Report-2022-Application-of-5G-V2X-Will-Promote-the- [134] F. Hawlader and R. Frank, "Towards a Framework to Evaluate
Realization-of-High-level-Autonomous-Driving--- Cooperative Perception for Connected Vehicles," in 2021 IEEE
[Link] Vehicular Networking Conference (VNC), Ulm, Germany, 2021, pp.
[119] T. Cui, L. Li, Z. Zhang, and C. Sun, "C-V2X Vision in the Chinese 36–39, doi: 10.1109/VNC52810.2021.9644667.
Roadmap: Standardization, Field Tests, and Industrialization," in [135] Q. Yang, S. Fu, H. Wang, and H. Fang, "Machine-Learning-Enabled
Vehicular Networks - Principles, Enabling Technologies and Cooperative Perception for Connected Autonomous Vehicles:
Perspectives [Working Title], IntechOpen, 2022. Challenges and Opportunities," IEEE Network, vol. 35, no. 3, pp. 96–
[120] G. Thandavarayan, M. Sepulcre, and J. Gozalvez, "Redundancy 101, 2021, doi: 10.1109/MNET.011.2000560.
Mitigation in Cooperative Perception for Connected and Automated [136] Q. Delooz et al., "Analysis and Evaluation of Information Redundancy
Vehicles," in 2021 IEEE 93rd Vehicular Technology Conference Mitigation for V2X Collective Perception," IEEE Access, vol. 10, pp.
(VTC2021-Spring), 2021, pp. 1–5, doi: 10.1109/VTC2020- 47076–47093, 2022, doi: 10.1109/ACCESS.2022.3170029.
Spring48590.2020.9129445. [137] G. Thandavarayan, M. Sepulcre, and J. Gozalvez, "Analysis of
[121] G. Cui, W. Zhang, Y. Xiao, L. Yao, and Z. Fang, "Cooperative Message Generation Rules for Collective Perception in Connected and
Perception Technology of Autonomous Driving in the Internet of Automated Driving," in IV19: 30th IEEE Intelligent Vehicles
Vehicles Environment: A Review," Sensors, early access. doi: Symposium : 9-12 June 2019, Paris, Paris, France, 2019, pp. 134–139,
10.3390/s22155535. doi: 10.1109/IVS.2019.8813806.
[122] H. Huang, H. Li, C. Shao, T. Sun, W. Fang, and S. Dang, "Data [138] G. Thandavarayan, M. Sepulcre, and J. Gozalvez, "Generation of
Redundancy Mitigation in V2X Based Collective Perceptions," IEEE Cooperative Perception Messages for Connected and Automated
Access, vol. 8, pp. 13405–13418, 2020, doi: Vehicles," IEEE Trans. Veh. Technol., vol. 69, no. 12, pp. 16336–
10.1109/ACCESS.2020.2965552. 16341, 2020, doi: 10.1109/TVT.2020.3036165.
[123] Q. Delooz, R. Riebl, A. Festag, and A. Vinel, "Design and [139] A. Balador et al., "Survey on decentralized congestion control methods
Performance of Congestion-Aware Collective Perception," in 2020 for vehicular communication," Vehicular Communications, vol. 33, p.
IEEE Vehicular Networking Conference (VNC), New York, NY, USA, 100394, 2022, doi: 10.1016/[Link].2021.100394.
2020, pp. 1–8, doi: 10.1109/VNC51378.2020.9318335. [140] B. Kar, W. Yahya, Y.-D. Lin, and A. Ali, "Offloading Using
[124] ETSI EN 302 637-2 - V1.4.1: Intelligent Transport Systems (ITS); Traditional Optimization and Machine Learning in Federated Cloud–
Vehicular Communications; Basic Set of Applications; Part 2: Edge–Fog Systems: A Survey," IEEE Communications Surveys &
Specification of Cooperative Awareness Basic Service, European Tutorials, vol. 25, no. 2, pp. 1199–1226, 2023, doi:
Telecommunications Standards Institute, Apr. 2019. [Online]. 10.1109/COMST.2023.3239579.
Available: [Link] [141] A, G. P. Wijesiri N. B., T. Samarasinghe, and J. Haapola,
30263702/01.04.01_60/en_30263702v010401p.pdf "Performance Enhancement of C-V2X Mode 4 Utilizing Multiple
[125] ETSI EN 302 637-3 - V1.3.1: Intelligent Transport Systems (ITS); Candidate Single-subframe Resources," Feb. 2022. [Online].
Vehicular Communications; Basic Set of Applications; Part 3: Available: [Link]
Specifications of Decentralized Environmental Notification Basic [142] F. Abbas, X. Yuan, M. S. Bute, and P. Fan, "Performance Analysis
Service, European Telecommunications Standards Institute, Apr. 2019. Using Full Duplex Discovery Mechanism in 5G-V2X Communication
[Online]. Available: [Link] Networks," IEEE Trans. Intell. Transport. Syst., vol. 23, no. 8, pp.
302699/30263703/01.03.01_60/en_30263703v010301p.pdf 11453–11464, 2022, doi: 10.1109/TITS.2021.3103974.
[126] ETSI TR 103 562 - V2.1.1.: Intelligent Transport Systems (ITS); [143] F. Milani, M. Foell, and C. Beidl, "A Data-based Approach to Predict
Vehicular Communications; Basic Set of Applications;Analysis of the the Response Time of Cloud-based Vehicle Functions," in 2019 IEEE
Collective Perception Service (CPS); Release 2, European International Conference on Connected Vehicles and Expo (ICCVE),
Telecommunications Standards Institute, Dec. 2019. [Online]. Graz, Austria, 2019, pp. 1–6, doi:
Available: [Link] 10.1109/ICCVE45908.2019.8965217.
103562/02.01.01_60/tr_103562v020101p.pdf [144] H. Wu, "Multi-Objective Decision-Making for Mobile Cloud
[127] ETSI TS 103 301 - V1.3.1: Intelligent Transport Systems (ITS); Offloading: A Survey," IEEE Access, vol. 6, pp. 3962–3976, 2018, doi:
Vehicular Communications; Basic Set of Applications; Facilities layer 10.1109/ACCESS.2018.2791504.
protocols and communication requirements for infrastructure services, [145] J. Liu, S. Wang, J. Wang, C. Liu, and Y. Yan, "A Task Oriented
European Telecommunications Standards Institute, Feb. 2020. Computation Offloading Algorithm for Intelligent Vehicle Network
[Online]. Available: [Link] With Mobile Edge Computing," IEEE Access, vol. 7, pp. 180491–
103399/103301/01.03.01_60/ts_103301v010301p.pdf 180502, 2019, doi: 10.1109/ACCESS.2019.2958883.
[128] X. Zheng, S. Li, Y. Li, D. Duan, L. Yang, and X. Cheng, "Confidence [146] L. Gillam, K. Katsaros, M. Dianati, and A. Mouzakitis, "Exploring
Evaluation for Machine Learning Schemes in Vehicular Sensor edges for connected and autonomous driving," in IEEE INFOCOM
Networks," IEEE Trans. Wireless Commun., vol. 22, no. 4, pp. 2833– 2018 - IEEE Conference on Computer Communications Workshops
2846, 2023, doi: 10.1109/TWC.2022.3214499. (INFOCOM WKSHPS), Honolulu, HI, 2018, pp. 148–153, doi:
[129] Q. Chen, S. Tang, Q. Yang, and S. Fu, "Cooper: Cooperative 10.1109/INFCOMW.2018.8406890.
Perception for Connected Autonomous Vehicles based on 3D Point
Clouds," May. 2019, doi: 39th. [Online]. Available: [Link]
pdf/1905.05265
[130] Y. Endo and S. Kamijo, "Deep Voxelized Feature Maps for Self-
Localization in Autonomous Driving," Sensors, early access. doi:
10.3390/s23125373.
[131] L. Ballotta, G. Peserico, F. Zanini, and P. Dini, "To Compute or not to
Compute? Adaptive Smart Sensing in Resource-Constrained Edge
Computing," 2022, doi: 10.48550/ARXIV.2209.02166.
[132] J. Wang, J. Liu, and N. Kato, "Networking and Communications in
Autonomous Driving: A Survey," IEEE Communications Surveys &
Tutorials, vol. 21, no. 2, pp. 1243–1274, 2019, doi:
10.1109/COMST.2018.2888904.
[133] F. Dettinger, H. Wei, M. Weiß, N. Jazdi, and M. Weyrich,
"Dateneffiziente Vervollständigung des Umgebungsmodells von
autonomen vernetzten Systemen mittels Sensorfusion," in Automation
2023: Transformation by Automation, 24. Leitkongress der Mess- und
Automatisierungstechnik, 27. und 28. Juni 2023, Baden-Baden (VDI-
 23
This work has been submitted to Engineering for possible publication.

Falk Dettinger, [Link]. is a PhD student and Maurice Artelt, [Link]. is a PhD student and
research assistant at the Institute of academic employee at the Institute of
Industrial Automation and Software Industrial Automation and Software
Engineering (IAS) at the University of Engineering (IAS) at the University of
Stuttgart, Germany. In addition, he is a Stuttgart, Germany. He received his
lecturer in the Faculty of Electrical bachelor’s degree in electrical engineering
Engineering at the Baden-Wuerttemberg in 2017 from the DHBW Mosbach,
Cooperative State University (DHBW) Germany and subsequently joined Bosch
Stuttgart, Germany. Rexroth AG, Erbach as industrie 4.0
He received his bachelor’s degree in electrical engineering in production planer. In 2018, he additionally became a student at
2017 from the DHBW Stuttgart, Germany and subsequently the University of Stuttgart, Germany, where he finished his
joined the Broetje Automation GmbH as commissioning master's degree in 2021.
engineer in the electrical engineering department. In 2019, he His areas of interest include remaining useful life prediction and
became a student at the University of Stuttgart, Germany, where anomaly detection with hybrid prediction models for electrical
he finished his master's degree with honors in 2021. His areas function groups.
of interest include function offloading between connected
vehicles and distributed cloud/edge systems and intelligent Prof. Dr-Ing. Michael Weyrich studied
function orchestration. electrical engineering, specialising at the
University of Applied Science Saarbrücken,
Matthias Weiß, [Link]. is a PhD student University of Westminster (London, U.K.)
and research assistant at the Institute of and at the Ruhr-University Bochum,
Industrial Automation and Software focussing on automation technology.
Engineering (IAS) at the University of Thereafter, he worked as a Research
Stuttgart, Germany. He received both his Assistant with Prof. Dr.-Ing. Paul Drews of
bachelor's and master’s degree in electrical the European Centre of Mechatronics and
engineering from the University of was awarded a doctoral degree in 1999 from the RWTH
Stuttgart in 2018 and 2021 respectively. Aachen.
His fields of research are the continuous Prof. Weyrich worked with Daimler AG for eight years, where
analysis and optimization of distributed cloud/edge systems he assumed the leadership of the “End-to-End Process Flexible
with the goal of reducing the operational complexity caused by Manufacturing” project and later had a technical management
frequent update deployments. function in the “CAx Process Chain – Production” of the
Information Technology Management division. From 2004
Daniel Dittler, [Link]. is a PhD student and onward, he was the head of “IT for Engineering” at
research assistant at the Institute of Industrial DaimlerChrysler Research and Technology Bangalore (India).
Automation and Software Engineering (IAS) After his return to Germany, he joined Siemens AG as
at the University of Stuttgart, Germany. He department leader with direct report to the Business Unit Head
received his bachelor's degree in mechanical of Motion Control in Erlangen for 2 years, focussing on
engineering from Pforzheim University in innovative technologies.
2019 and his master's degree in mechatronic In 2009, he was appointed as Professor for the Chair of
system development in 2021. His areas of Automation in Manufacturing at the University of Siegen and
interest are digital twins in automation as Director of the Automotive Centre Südwestfalen.
technology and the automatic model adaption of behavior Prof. Weyrich assumed the role of Director of the Institute of
models in the operational phase. Industrial Automation and Software Engineering at the
University of Stuttgart in 2013.
Johannes Stümpfle, [Link]. is a PhD student He is very interested in research in cyber physical systems for
and research assistant at the Institute of industrial application. He has published over 100 papers and is
Industrial Automation and Software very active in research and is a member of the Board of the
Engineering (IAS) at the University of German Engineering foundations VDI/VDE GMA.
Stuttgart, Germany. He received his Additionally, he is an appointed reviewer for the European
bachelor’s and master’s degree in electrical Commission, the German Research Foundation, DFG and a
engineering at the University of Stuttgart. His number of other institutions.
areas of interest include the evolution of software-intensive
systems and mastering the complexity in highly variant
automation systems.