Autonomous Vehicles in 5G and Beyond: A Survey

Saqib Hakak; Thippa Reddy Gadekallu; Swarna Priya Ramu; Parimala M; Praveen Kumar Reddy Maddikunta; Chamitha de Alwis; Madhusanka Liyanage

Source-linked AI summary

Autonomous Vehicles in 5G and Beyond: A Survey

Saqib Hakak, Thippa Reddy Gadekallu, Swarna Priya Ramu, Parimala M, Praveen Kumar Reddy Maddikunta, Chamitha de Alwis, Madhusanka Liyanage

arXiv:2207.10510v1Jul 21, 2022cs.NI

TL;DR

Integrating 5G with autonomous vehicles remains an early-stage effort facing security, infrastructure, sensing, and machine-learning challenges. This survey synthesizes AV technologies, 5G requirements, emerging integration techniques, security concerns, standardization, projects, and future directions, identifying requirements and research challenges for 5G/B5G-enabled AV applications.

Problem
5G–AV integration remains at an early stage, with unresolved requirements and challenges involving security, infrastructure costs, sensing, and machine-learning systems.
Method
The survey reviews AV features, automation levels, architectures, enabling technologies, communication requirements, emerging 5G/B5G techniques, security, standards, and projects.
Results
The survey identifies requirements across latency, security, privacy, bandwidth, mobility, scalability, availability, and reliability, while highlighting standardization efforts and research challenges.
Takeaways & Limitations
The synthesis frames MEC, SDN, network slicing, 5GNR, blockchain, federated learning, and related technologies as relevant to AV communication and coordination use cases.
Takeaways & Limitations
Autonomous-vehicle navigation and path planning remain difficult under uncertainty in dynamic environments.

Abstract

from arXiv · show

Fifth Generation (5G) technology is an emerging and fast adopting technology which is being utilized in most of the novel applications that require highly reliable low-latency communications. It has the capability to provide greater coverage, better access, and best suited for high density networks. Having all these benefits, it clearly implies that 5G could be used to satisfy the requirements of Autonomous vehicles. Automated driving Vehicles and systems are developed with a promise to provide comfort, safe and efficient drive reducing the risk of life. But, recently there are fatalities due to these autonomous vehicles and systems. This is due to the lack of robust state-of-art which has to be improved further. With the advent of 5G technology and rise of autonomous vehicles (AVs), road safety is going to get more secure with less human errors. However, integration of 5G and AV is still at its infant stage with several research challenges that needs to be addressed. This survey first starts with a discussion on the current advancements in AVs, automation levels, enabling technologies and 5G requirements. Then, we focus on the emerging techniques required for integrating 5G technology with AVs, impact of 5G and B5G technologies on AVs along with security concerns in AVs. The paper also provides a comprehensive survey of recent developments in terms of standardisation activities on 5G autonomous vehicle technology and current projects. The article is finally concluded with lessons learnt, future research directions and challenges.

I. INTRODUCTION … B. Features of Autonomous and Connected Vehicles

The paper surveys autonomous and connected vehicles, emphasizing how 5G can enable high-speed, low-latency communication for autonomous driving. It reviews integration techniques, security, standards, projects, and future research challenges while outlining ACV capabilities.

I. INTRODUCTION: Autonomous vehicles operate without direct driver input, enabled by high-speed networks, decentralized storage, edge computing, and related technologies.The driver is not expected to monitor the roadway constantly.
I. INTRODUCTION: 5G offers speeds up to 10 Gbps and latency of 1 ms, supporting delay-sensitive applications through eMBB, mMTC, and URLLC.URLLC is designed for applications such as autonomous driving that require very low error-bit rates.
I. INTRODUCTION: The survey provides a comprehensive review of AVs and enumerates technical aspects required for successful AV–5G integration.It also conducts a state-of-the-art review of integrating AVs with 5G.
I. INTRODUCTION: The paper identifies AV security concerns, highlights key 5G autonomous-vehicle projects and standardisation activities, and explores future research challenges and directions.These topics are presented among the survey’s stated contributions.
II. INTRODUCTION TO AUTONOMOUS AND CONNECTED VEHICLES: Autonomous and connected vehicles support Intelligent Transportation Systems by enabling vehicles to communicate and exchange critical information through communication infrastructure.Their growth responds to increasing urban mobility needs and transportation-technology changes.
A. Evolution of Autonomous and Connected Vehicles: ACVs are distinguished from conventional vehicles by two major properties: automation and connectivity.These properties define ACVs as a distinct vehicle category.
A. Evolution of Autonomous and Connected Vehicles: Connected vehicles use V2V and V2I communication, but heterogeneous communication levels create scalability and coverage-area issues that motivate ACVs.ACVs combine autonomous operation with connectivity to address these limitations.
B. Features of Autonomous and Connected Vehicles: ACVs manage maintenance and software updates, adapt to dynamic environments and failures, accommodate external policies, allocate resources by mobility, and protect against attacks.Their self-protection capability aims to mitigate failures and prevent failure of the entire network of systems.

C. Levels of Automation · D. Architecture

The paper describes six SAE-based automation levels, progressing from human-controlled driving to fully system-controlled cooperative driving. It also presents ACV architecture as a sensor-connected, three-layer system supporting perception, route planning, decision-making, actuator control, and cooperative driving.

C. Levels of Automation: Level 0 leaves driving decisions and road-use operations under human control, while Level 1 automates selective functions such as lateral or longitudinal motion control.Level 1 provides limited assistance to the human driver.
C. Levels of Automation: Level 2 combines multiple driving controls, whereas Level 3 provides conditional assistance during parts of a trip before the driver resumes control.Examples include adaptive cruise control, lane-maintaining assistance, and temporary driver assistance for other activities.
C. Levels of Automation: Level 4 manages the complete driving mechanism and environment monitoring but permits driver takeover during critical situations, unlike Level 5’s continuous human-free control.Level 5 also handles failures and dynamic decisions throughout the trip through cooperative driving.
D. Architecture: ACVs integrate technologies and onboard sensors to enable inter-vehicular connectivity for traffic safety, road assistance, efficiency, monitoring, congestion avoidance, maintenance, and failure management.Sensors communicate among themselves through the Controller Area Network bus and with infrastructure through V2X connectivity.
D. Architecture: The ACV architecture contains perception, planning/processing, and control-related functionality, with perception sensors gathering environmental data and sensor fusion generating locations and an environment map.The planning/processing layer determines routes using vehicle position, destination, road data, and traffic data.
D. Architecture: Planning supports cooperative driving and sends commands to actuators such as the steering wheel, gas pedal, and brake pedal, while connectivity enables critical decisions but increases deployment complexity.Connectivity is required among vehicles, infrastructure, and road users.

E. Key enabling technologies … 2) Multi-access Edge Computing:

The paper surveys enabling technologies for autonomous connected vehicles (ACVs), including heterogeneous sensing, cloud-based data access, high-bandwidth communication, and security. It then explains how 5G features such as ProSe and MEC support low-latency vehicular communication and emerging applications.

E. Key enabling technologies: Innovative sensor, cloud-computing, and artificial-intelligence technologies can support intelligent ACVs and their real-world deployment.Traditional methodologies proven in real time may not be suitable for implementing and deploying ACV features.
1) Sensing Environment:: ACVs use heterogeneous sensors, including detection sensors mounted on vehicles to identify features in and around the environment.Vehicular networks commonly focus on single-type or homogeneous sensors, whereas ACVs rely on multiple heterogeneous sensors.
2) Accessing Data:: ACVs generate large volumes of sensor signals that must be accessed by authorised authorities, other vehicles, and surrounding infrastructure for timely decisions.Temporary storage, archiving, and access require high-end resources and servers, including cloud-computing support.
3) Vehicular Communication:: ACV functionalities depend on heterogeneous sensor data, requiring gigabits-per-second bandwidth for high-performance communication.Millimeter-wave communication supports V2V, V2I, and intra-vehicular links, including real-time map and dynamic-environment stream downloads.
4) Security:: Vehicular communication security is a critical challenge in environments containing both legacy and autonomous vehicles, where inadequate protection can cause chaos.Physical layer security (PLS) has been designed as a replacement for cryptography to help maintain communication security and privacy.
1) Proximity Service:: 5G Proximity Service (ProSe) provides vehicles with awareness of nearby devices, infrastructure, objects, and locality information through spontaneous local communication.ProSe is particularly suited to discovering moving vehicles on the road.
2) Multi-access Edge Computing:: 5G vehicular communication targets latency up to 100 ms for safety measures and up to 1 ms for ACVs, with MEC moving core functionalities toward users.MEC brings services to network locations, while NFV can host mobile edge applications in multi-vendor edge environments; WiFi, 802.11p, and 5G provide access technologies.

3) Network Slicing: … 2) Existing Challenges/Limitations:

The paper presents network slicing as a way to manage heterogeneous access technologies in 5G-enabled autonomous vehicles, while identifying localization, uncertainty-aware planning, and realistic-road navigation as continuing challenges. It also outlines autonomous-driving technologies, their potential benefits, and planner architectures for handling static and dynamic obstacles.

3) Network Slicing:: Network slicing logically separates heterogeneous networks so 5G can manage diverse access technologies according to autonomous-vehicle application and requirement needs.Examples include safety applications and infotainment; slicing also supports network integrity and security in vehicular networks.
III. TECHNICAL ASPECTS OF AUTONOMOUS VEHICLES: Autonomous driving research aims to improve traffic safety, energy efficiency, public-resource use, vehicle capabilities, and control while reducing drivers’ physical and mental burden.Universities, research groups, automobile companies, and Internet auto companies participate in autonomous-driving competitions and technical challenges.
1) Introduction:: Autonomous vehicles are expected to replace ordinary vehicles with smart vehicles capable of decision-making, shortest-path selection, and optimal travel-route planning.5G can support local perception for short-range vehicle control involving safety, traffic control, and energy management.
A. Navigation and Path Planning: Autonomous navigation depends on localization, planning, and control, with reliable localization along the planned path as a basic requirement.Autonomous robots in UAVs and UGVs can provide more stable and robust systems, while UAV navigation may use an initial trajectory and waypoint-based guidance.
2) Existing Challenges/Limitations:: GPS localization performs well outdoors but becomes unreliable indoors and in dense urban environments, whereas laser range finders perform better across indoor and outdoor applications.In unstructured and complex environments, laser-range-finder readings support path planning to reduce localization errors and expose the vehicle to rich environmental information compared with existing techniques.
2) Existing Challenges/Limitations:: Navigation and path planning under uncertainty is difficult for UAVs, drones, and self-driving cars, motivating probabilistic methods and probabilistic decision engines for dynamic environments.The decision engine is presented as an alternative to traditional time-consuming Monte-Carlo methods for navigation under uncertainty.
2) Existing Challenges/Limitations:: Autonomous-vehicle technologies must support complex realistic road scenarios, provide effective collision warnings, and plan high-speed paths on both structured and unstructured roads.Many studies focus on structured roads, while free-form navigation remains important for unstructured roads.
2) Existing Challenges/Limitations:: The navigation planner seeks collision-free waypoints while reducing computational cost and travel distance, using global planning for static obstacles and local planning for dynamic obstacles.The described architecture includes a global planner using prior environmental knowledge and a local planner regulating paths around changing obstacles; TEB and Dijkstra are cited as planning methods.

3) How B5G help (with Related work): … 2) Existing Challenges/Limitations:

The section identifies sensing, mapping, navigation, data-sharing, latency, security, and legal limitations affecting autonomous vehicles. It also describes how 5G, B5G-related communication, cooperative localization, edge networking, and URLLC can mitigate selected challenges while introducing resource and security constraints.

3) How B5G help (with Related work):: AV sensors have complementary limitations: cameras struggle in poor climate, radar differentiates object types poorly, and lidar loses accuracy in fog or snow.These limitations arise from climatic conditions, radar’s longer wavelength, and degraded laser performance in low visibility.
3) How B5G help (with Related work):: Additional AV challenges include time-consuming signal-based map construction, inability to predict road-agent behavior, and legal issues related to accidents.These limitations are listed alongside sensing difficulties in the section’s challenge summary.
Summary:: Autonomous navigation requires timely obstacle localization, dynamic object detection, and path planning to avoid accidents and reach destinations safely.The navigation challenge combines detecting and localizing obstacles with adjusting vehicle paths at specific times.
B. Object detection/ Collision Avoidance: 5G edge networking addresses latency for AVs generating massive video, sensor, object-detection, and lane-condition data, while cooperative localization improves vehicle positioning.The described cooperative-localization framework uses vehicles as graph vertices and communication paths as edges.
B. Object detection/ Collision Avoidance: Long-distance and low-visibility sensing remains unreliable, while malicious vehicles can share fake data that causes AVs to change lanes or accelerate dangerously.Roadside equipment and intervehicle data exchange are proposed alternatives, but trusting manipulated source data creates collision-avoidance risks.
1) Introduction:: Vehicles and infrastructure must remain interconnected to transfer information, motivating communication technologies that satisfy autonomous-driving reliability and latency requirements.URLLC is designed for stringent reliability and latency requirements in critical packet transmission for connected autonomous vehicles.
C. URLLC: URLLC supports autonomous driving’s critical information exchange by targeting 1 millisecond latency and 99 percent reliability with end-to-end security.These requirements support communication between neighboring vehicles and road infrastructure for automated driving tasks.
2) Existing Challenges/Limitations:: Onboard processing is constrained by storage, power, cooling, and thermal limits, while cloud computing cannot guarantee low latency because server-client communication introduces delay.GPU cooling can reduce fuel efficiency, and SSD storage may fill within hours with sensor and device data.

3) How B5G help (with Related work): … 3) How B5G help (with Related work):

The section explains how 5G/B5G technologies support autonomous vehicles through URLLC, mMTC, V2V, V2I, and V2X, while addressing latency, reliability, interference, and connectivity challenges. It also highlights edge computing, redundancy, and vehicle–infrastructure awareness as mechanisms for safer and more efficient driving.

3) How B5G help (with Related work):: URLLC scheduling divides 1 ms time slots into minislots and promptly transmits received URLLC traffic by puncturing ongoing eMBB transmissions.Related work also uses guard zones around vehicle receivers to improve URLLC reliability and prohibit eMBB transmission within those zones.
3) How B5G help (with Related work):: Edge caching, computing, AI, and roadside BBU servers can deploy storage and computation near wireless networks to support URLLC in AVs.The proposed solutions also require reliability and redundancy across application, transmission, software, and networking layers.
3) How B5G help (with Related work):: URLLC enables AVs to make real-time decisions through faster communication, real-time connectivity, and 5G’s low-latency capability.The section identifies URLLC as a key technology for communication between autonomous vehicles and 5G networks.
1) Introduction:: mMTC connects billions of low-complexity, low-power devices, whereas uMTC provides reliable wireless links for services widely used in V2X.mMTC supports applications including automated industries, remote surgeries, and smart metering; its traffic is characterized by small packets and sporadic activity.
2) Existing Challenges/Limitations:: Traditional LTE-based technologies cannot satisfy the latency and reliability requirements of connected cars, autonomous vehicles, and industrial automation.They also cannot adequately handle IoT-specific sporadic transmission, power optimization, and uplink-centric transmission requirements.
2) Existing Challenges/Limitations:: V2V communication connects vehicles through mesh topology, while V2I and V2X communication support interactions between vehicles, infrastructure, and broader networked systems.Single-hop vehicular communication supports short-range applications, whereas multi-hop communication is used for traffic monitoring.
3) How B5G help (with Related work):: Asynchronous massive machine-type transmissions create inter-carrier interference because random base-station access lacks coordination.This creates a next-generation wireless-communication challenge involving activation ratio, subblock-size optimization, user clustering, and conflict reduction.
3) How B5G help (with Related work):: V2X and 5G connections help AVs detect objects and obstacles beyond corners, while vehicle–infrastructure connectivity provides advance traffic awareness.The section describes automatic speed reduction in slow-moving traffic and identifies Audi’s Traffic Light Information as a V2I case study in Europe.

E. eMBB … 1) Related work):

The surveyed work positions eMBB, together with URLLC and mMTC, as essential 5G capabilities for autonomous-vehicle applications, while MEC enables low-latency communication, real-time processing, and computational offloading. Related studies demonstrate MEC-enabled architectures and algorithms for vehicular communication and applications, but standardization, heterogeneity, privacy, and security remain unresolved.

1) Introduction:: eMBB enhances bandwidth-related Quality of Experience for in-vehicle applications.
2) Existing Challenges/Limitations:: Fixed Wireless Access provides wide-coverage eMBB using higher-spectrum bands and is expected to expand exponentially from 2018–2025.
3) How B5G help (with Related work):: Resource scheduling research assigns eMBB resource blocks using channel state and average data rate, then applies a chance constraint to maximize data rates.A two-dimensional Hopfield Neural Network and energy function solve the resource-allocation problem.
3) How B5G help (with Related work):: NOMA optimizes unicast and multicast distribution to support more AV data-transmission users, with complexity reduced through fewer injection levels and smart algorithm selection.The algorithm’s complexity is compared with Time Division Multiplexing (TDM).
3) How B5G help (with Related work):: eMBB and URLLC are treated as prerequisites for smart intelligent transportation systems, with eMBB scheduled at slot boundaries and random URLLC arrivals during transmission intervals.
3) How B5G help (with Related work):: URLLC, mMTC, and eMBB jointly address bandwidth, density, and latency limitations in applications including autonomous vehicles, smart cities, and augmented reality.
A. Multi-access Edge Computing (MEC): 5G and B5G technologies affect autonomous vehicles through prominent networking and computing capabilities, including MEC-enabled fast communication, real-time processing, and offloading of latency-sensitive, computing-intensive tasks.MEC reduces congestion and latency by bringing cloud and IT capabilities closer to the network edge.
1) Related work):: MEC-enabled 5G research supports V2I/V2V communication, mobility management, simulation offloading, tollgate selection, and motion planning, but standardization, heterogeneity, privacy, and security remain challenges.A proposed architecture provides guaranteed low packet delay and high scalability, while distributed mobility management makes IP handoff transparent and seamless.

B. Network Slicing … D. 5GNR/Physical layer stuff

The section surveys network slicing, SDN-enabled resource management, and 5GNR physical-layer technologies for autonomous connected vehicles (ACVs). It emphasizes support for heterogeneous communication demands, mobility, latency, throughput, and coordinated driving while identifying deployment challenges.

1) Introduction:: Network slicing creates multiple end-to-end logical networks over shared physical and virtual resources to support ACVs’ diverse communication requirements.It addresses needs that a generic one-fits-all pre-5G architecture cannot satisfy.
2) Related Work:: Dedicated V2X slices can separate safety-critical autonomous-driving messages from high-volume infotainment video-streaming services.Related work describes slices spanning devices, the radio access network, and the core network.
2) Related Work:: Network slicing remains immature for fully functional ACVs because vehicular mobility and changing resource demands complicate deployment and resource allocation.Requirements range from time-critical safety communication to high-volume infotainment data.
CONTROL PLANE: 5G-enabled SDN can manage ACV connectivity and edge computation dynamically to reduce communication overhead and latency while improving network-layer performance.The surveyed direction focuses on mobility support, reduced latency, increased connectivity, and greater intelligence for future ACV networks.
CONTROL PLANE: 15.9 percent improvement in end-to-end delay was achieved using a multi-objective evolutionary algorithm for edge-cloud communication.The approach also enhanced communication latency and reduced overhead through better bandwidth utilization.
D. 5GNR/Physical layer stuff: B5G vehicular communications require extremely high data rates, with transmission rates expected to reach up to 1 terabit/second.Massive MIMO with more than 100 antennas and millimetre-wave technologies are identified as spectrum-efficiency and bandwidth-enhancement approaches.
D. 5GNR/Physical layer stuff: Cellular V2X complements radar, cameras, GNSS, and other sensors by exchanging sensory data among vehicles, overcoming their line-of-sight limitations.The 5GNR physical layer must handle mobility from vehicles traveling up to 60 kilometers/hour to high-speed trains or cars exceeding 500 kilometers/hour.
D. 5GNR/Physical layer stuff: 5GNR supports ACV use cases requiring high data rate, throughput, and URLCC, including coordinated driving, local-condition updates, trajectory sharing, and raw-data sharing.Physical-layer design also involves channel coding, resource slicing, sidelink modes, and communication over challenging V2X channels.

E. Federated Learning · 1) Introduction:

Federated learning is presented as a way to address the privacy, scalability, volatility, and complexity challenges that limit traditional machine learning in connected autonomous-vehicle environments. It offers potential benefits in B5G networks, but poisoning attacks, false alarms, and energy constraints remain barriers to realizing its full potential.

E. Federated Learning · 1) Introduction:: Federated learning is motivated by the need to realize intelligent transportation systems while addressing high availability, data privacy, and scalability.The passage identifies traditional ML models as inefficient for these requirements.
E. Federated Learning · 1) Introduction:: Traditional machine-learning models struggle with the dynamic, volatile, and complex environments of intelligent transportation systems and the Internet of Vehicles.Vehicles constantly enter and leave otherwise stable roadside infrastructures, challenging static local intelligence.
E. Federated Learning · 1) Introduction:: Collaborative data sharing among vehicles can improve service quality and driving experience, but privacy, security, bandwidth, and unreliable communication impede participation and reliability.Vehicle-to-vehicle data sharing is described as a basis for collaborative analysis in IoV.
E. Federated Learning · 1) Introduction:: A proposed 5G solution combines local directed acyclic graphs, permissioned blockchain, and asynchronous federated learning for reliable and secured vehicle data sharing.The hybrid blockchain architecture is run by vehicles, and asynchronous FL is proposed to improve efficiency.
E. Federated Learning · 1) Introduction:: Federated learning supports privacy-preserving positioning applications because centrally collecting and training sensitive vehicle-trajectory data is difficult.Precise positioning for collision avoidance and autonomous driving can use V2I communications, sensing, and nearby landmarks.
E. Federated Learning · 1) Introduction:: In connected autonomous vehicles, federated learning may improve latency, address privacy issues, optimize resource utilization, and provide customized recommendations or predictions in heterogeneous B5G networks.These benefits are identified as part of FL’s potential in the B5G era.
E. Federated Learning · 1) Introduction:: Federated learning still faces poisoning attacks, high false alarms, and energy-efficiency challenges on low-powered autonomous-vehicle devices.Addressing these issues is necessary to realize the full potential of FL-enabled B5G networks for AVs.

F. Blockchain … V. SECURITY CONCERNS IN AVS

Blockchain offers immutable, transparent, distributed records and consensus-based trust for 5G-connected AVs, while related work applies it to event validation, VANETs, and UAV networks. B5G-era AVs require automated network management, but security threats and integration challenges such as standardization and scalability remain.

1) Introduction:: Blockchain distributes and duplicates transaction records across network nodes, linking each block to its own and the previous block’s hash.Its properties include immutability, transparency, and trustability.
1) Introduction:: Blockchain data is difficult to tamper with, while consensus algorithms require majority-node approval before adding a new node.
2) How it used in AV (with Related work):: 5G-connected AVs share environmental, blockage, and traffic information, but malicious AVs can inject false data, expose vehicle privacy, or modify shared information.Blockchain is presented as a response to these security and privacy issues.
2) How it used in AV (with Related work):: Private blockchain-enabled edge nodes can store AV event-driven messages and their multimedia evidence to support reliability and authenticity.Related work also proposes SDN- and blockchain-enabled 5G networks for VANETs.
2) How it used in AV (with Related work):: Blockchain integrated with MEC-enabled B5G networks is discussed for UAV scenarios requiring low latency, high data rates, and demanding QoS.UAVs provide on-demand mobile access points for applications including smart cities and smart manufacturing.
2) How it used in AV (with Related work):: Blockchain can protect UAV command-and-control exchanges, with UAVs creating blocks and validating transactions as miners; sharding is proposed to address scalability.The cited work targets better network manageability and security for defense applications.
2) How it used in AV (with Related work):: Integrating blockchain in B5G for AVs still faces standardization and scalability challenges as interconnected AV populations continue increasing.UAV and drone applications include crop or soil analysis, road surveillance, disaster monitoring, and product delivery.
G. ZSM(AI/ML): B5G-era AVs require ultra-high reliability, near-infinite capacity, global reach, massive machine-to-machine communication, and imperceptible latency, motivating fully automated network and service management.

A. Security Concerns via In-Vehicle Communication … 2) 5G for Connected and Automated Road Mobility in the European UnioN (5G-CARMEN):

The survey identifies security requirements and vulnerable in-vehicle and V2X components in autonomous vehicles, including sensors, ECUs, and OBD gateways. It also reviews 5G autonomous-vehicle projects, standards, and trials, including 5G-DRIVE and 5G-CARMEN.

A. Security Concerns via In-Vehicle Communication: AV security requires authentication, integrity, availability, and confidentiality/privacy, while active, passive, external, and insider attackers may target different security goals.Authentication identifies authorized users and grants feature access according to privilege; confidentiality/privacy prevents exchanged data from reaching unauthorized users.
1) Sensors and Actuators:: LiDAR, cameras, and RADAR are vulnerable to cyberattacks, while cameras can also suffer degraded performance in rain or fog and denial-of-service attacks.LiDAR supports obstacle detection, RADAR uses radio waves with comparatively long detection range, and cameras provide views for traffic-sign and lane detection.
2) Controller Area Network:: The controller area network connects electronic control units that operate essential autonomous-driving subsystems such as braking and engine control.ECUs are embedded systems controlling vehicle subsystems and electrical systems.
3) On-board Computer(OBD):: OBD-II interfaces can expose autonomous vehicles to attacks because OBD data lacks encryption and authentication, while the interface also permits firmware updates and software modification.OBD retrieves vehicle information such as emissions and speed and can modify software embedded in control units.
B. Security Concerns Via V2X technologies: V2X communication requires high bandwidth, low latency, and high reliability to exchange information among sensors, vehicles, infrastructure, and pedestrians.Its goals include improving traffic efficiency, reducing accidents through enhanced road safety, and saving energy.
B. Security Concerns Via V2X technologies: IEEE 802.11P and Cellular C2X are key V2X standards, offering characteristics such as low latency and performance under bad weather for vehicle-to-vehicle and vehicle-to-infrastructure communication.The passage describes progress in this area as being at an infant stage.
VI. PROJECTS AND STANDARDIZATION: Research projects and standardization activities play a vital role in realizing 5G autonomous vehicles and connected automated mobility.The survey presents key global research and development projects alongside related standardization activities.
A. Research Projects: 5G-DRIVE investigates 3.5 GHz eMBB and 3.5 GHz and 5.9 GHz V2X deployments using network slicing, NFV, MEC, and 5G New Radio, while 5G-CARMEN builds a 600 km trial network across three countries.5G-DRIVE is an EU-China Horizon H2020 collaboration; 5G-CARMEN focuses on 5G NR, C-V2X, and secure service orchestration along the Bologna-Munich corridor.

3) Fifth Generation Cross-Border Control (5GCroCo): … 1) European Commission (EC):

The surveyed initiatives span EU-funded 5G trials, connected-vehicle infrastructure, terahertz sensing, cybersecurity, and public acceptance, while EC-led standardization and policy activities support autonomous mobility across Europe.

3) Fifth Generation Cross-Border Control (5GCroCo):: 5GCroCo builds 5G trial networks and test sites across the France–Germany and Germany–Luxembourg borders to integrate telecommunications and automotive research.The project operates under the EU Horizon (H2020) framework.
4) 5G for cooperative and connected automated MOBIility on X-border corridors (5G-MOBIX):: 5G-MOBIX develops cross-border 5G corridors between Greece–Turkey and Spain–Portugal for connected automated vehicle use cases and international research cooperation.Use cases include cooperative overtaking, highway lane merging, truck platooning, remote control, see-through, and HD map updates.
5) ICT Infrastructure for Connected and Automated Road Transport (ICT4CART):: ICT4CART combines telecommunications, automotive, and IT developments for connected and automated transport, emphasizing hybrid connectivity, network slicing, privacy, security, and localisation.It includes a cross-border corridor between Italy and Austria and three small-scale trial sites.
6) Terahertz sensors and networks for next generation smart automotive electronic systems (car2TERA):: Car2TERA investigates 150-330 GHz sub-terahertz communications and radar for high-speed onboard data, high-resolution in-vehicle radar, and improved in-cabin and outdoor sensing.Its focus is autonomous automobiles and next-generation smart automotive electronic systems.
7) Fifth Generation Communication Automotive Research and innovation (5GCAR):: 5GCAR advances 5G C-V2X through radio access, spectrum, architecture, orchestration, security, privacy, edge computing, multi-connectivity, and business-model research.Its autonomous-vehicle use cases include lane-merge coordination and cooperative perception for connected-vehicle maneuvers.
B. Standards Developing Organizations (SDOs):: Standardization activities define autonomous-vehicle requirements and identify 5G technologies capable of realizing them through global standards-developing organizations.The section summarizes key global-level standardization activities related to 5G autonomous vehicles.

2) European Automotive - Telecom Alliance (EATA): · 3) CAR 2 CAR Communication Consortium(C2C-CC) : · 4) 5G Automotive Association (5GAA):

The paper highlights three organizations supporting connected and automated driving through cross-sector collaboration, safety solutions, 5G integration, standardization, and regulatory work. EATA focuses on European deployment barriers, C2C-CC on accident-free traffic and robust safety technologies, and 5GAA on 5G-based automotive platforms and applications.

2) European Automotive - Telecom Alliance (EATA):: EATA was formed after an EU Commission round-table discussion to promote EU-level autonomous-vehicle activities.Its vision is to support collaboration between automotive and telecommunications stakeholders across Europe.
2) European Automotive - Telecom Alliance (EATA):: EATA seeks to accelerate connected and automated driving deployment across Europe through automotive–telecommunications collaboration.The alliance explores deployment issues involving stakeholders from both sectors.
2) European Automotive - Telecom Alliance (EATA):: EATA mainly addresses regulatory and legislative obstacles affecting connected and automated driving deployment.The passage identifies regulatory and legislative barriers as a central focus of the alliance.
3) CAR 2 CAR Communication Consortium(C2C-CC) :: Founded in 2002, C2C-CC is a global organization comprising road operators, automotive manufacturers, IT service providers, telecommunications operators, and research organizations.Its membership spans the main transport, technology, and research sectors.
3) CAR 2 CAR Communication Consortium(C2C-CC) :: C2C-CC pursues accident-free traffic by supporting ultra-reliable, robust, and mature safety solutions.This goal is expressed as “vision zero,” meaning accident-free traffic as early as possible.
3) CAR 2 CAR Communication Consortium(C2C-CC) :: C2C-CC also supports innovation in 5G and wireless technologies, emphasizing spectrum efficiency and ad-hoc short-range communication.The supplied passage introduces these technical priorities but is truncated before completing the description.
4) 5G Automotive Association (5GAA):: 5GAA is a leading global standardization organization focused on integrating 5G with the automotive industry and enabling C-ITS and V2X.Its cross-industry membership includes automotive, IT, and telecommunications sectors, while its seven working groups cover requirements, architecture, testing, standards, business, regulation, and security.
4) 5G Automotive Association (5GAA):: 5GAA working groups develop standards, architectures, frameworks, and business cases for 5G-based autonomous vehicles and applications.Table XIV presents important 5GAA standardization efforts, with document dates based on the latest updates in SDO repositories.

5) European Telecommunications Standards Institute (ESTI): … 9) 5G Americas:

The surveyed standardization bodies address autonomous-vehicle communications, AI-enabled driving, cybersecurity, and 5G V2X. Their activities include cooperative ITS standards, C-V2X and NR-V2X development, global AI performance thresholds, connected-vehicle security, and critical vehicle information exchange.

5) European Telecommunications Standards Institute (ESTI):: ETSI’s Intelligent Transportation Systems Technical Committee develops standards for automated and connected vehicles across cooperative ITS, automotive radar, and anti-collision radar.Its work covers overall automotive communication architecture and related systems.
6) 3rd Generation Partnership Project (3GPP):: 3GPP develops C-V2X standards, replacing US DSRC and European C-ITS, with the initial standard included in Release 14, published in 2017.3GPP is a consortium of seven telecommunication standards development organizations.
6) 3rd Generation Partnership Project (3GPP):: NR-V2X, introduced in 2020 as part of Release 16, improves support for automated driving and complex interactions such as cooperative automated driving.The passage presents NR-V2X as an improvement supporting 5G and beyond.
7) International Telecommunication Union - Telecommunication (ITU-T):: ITU-T launched the FG-AI4AD focus group in 2019 to support standardization of AI-enabled autonomous and assisted driving systems and services.The group also seeks to harmonize global activities defining a minimal performance threshold for these systems.
8) Alliance for Telecommunications Industry Solutions (ATIS):: ATIS’s Connected Car-Cybersecurity Ad Hoc Group studies telecom–automotive collaboration and security threats in connected automated vehicles.It also discusses how future mobile networks could help prevent these attacks.
9) 5G Americas:: 5G Americas promotes 5G V2X as a critical technology for connected and autonomous vehicles across the American countries.It identifies critical information exchange among autonomous vehicles as a means to improve navigation and situation awareness and avoid road accidents.

10) Next Generation Mobile Networks (NGMN) Alliance: … 1) Lessons Learned:

The paper highlights ongoing 5G–automotive collaboration, the communication requirements and benefits of 5G-enabled autonomous vehicles, and future directions involving predictive, safety, and advanced in-vehicle services. It also identifies deployment costs and technical, infrastructure, privacy, and security challenges, with B5G technologies proposed to address AV functionality requirements.

10) Next Generation Mobile Networks (NGMN) Alliance:: NGMN formed a V2X task force in 2016 to accelerate C-V2X deployment and strengthen cooperation between telecommunications and automotive sectors.The task force also examines policies, business models, spectrum management, security, and privacy.
10) Next Generation Mobile Networks (NGMN) Alliance:: In 2018, NGMN published a white paper describing eight V2X use cases and their technological requirements.
11) Other SDOs:: Other standards organizations have 5G autonomous vehicles as a minor focus, including EUCAR’s automotive strategies and 5G-ACIA’s industrial 5G and automated-guided-vehicle activities.EUCAR supports common frameworks, research, and innovation, while 5G-ACIA studies cloud-controlled AGVs for industrial environments.
1) Lessons Learned:: The lessons learned emphasize that AVs rely on V2X and 5G for high transmission rates, latency less than 5ms, reliability, and responsive vehicle–infrastructure communication.Telecommunications and automotive industries are building an ecosystem around integrating 5G with AV technologies.
1) Lessons Learned:: 5G can support high-speed, high-bandwidth driving, minimal-time information delivery, advance hazard alerts, and AI-integrated autonomous decision-making.
B. B5G technologies — 1) Lessons Learned:: 5G-AV deployment faces expensive spectrum, high network-extension costs, and tedious geolocation mapping, while B5G technologies can support low latency, physical-layer improvements, privacy, and security.Relevant technologies include MEC, network slicing, SDN/NFV, 5GNR, blockchain, federated learning, and ZSM for functions such as detection, collision avoidance, navigation, and V2X communications.
2) Possible Future Directions:: Future 5G-AV directions include predictive maintenance, AI-enhanced AR/VR infotainment, and cloud-based traffic safety services that warn about hazards and congestion and recommend alternate routes.Predictive maintenance combines in-vehicle sensor data, cloud data, and AI to anticipate component failures.

2) Possible Future Directions: … VIII. CONCLUSION

The survey identifies security, standardization, explainability, and emerging computing paradigms as priorities for integrating 5G/B5G with autonomous vehicles. It concludes by consolidating AV requirements, enabling technologies, ongoing activities, and future research challenges.

C. Security Concerns in AVs: V2X expands AV attack surfaces, requiring confidentiality, integrity, availability, authentication, and effective countermeasures.The survey particularly emphasizes stronger security mechanisms for V2X-enabled AV communication.
2) Possible Future Directions:: Future authentication mechanisms should connect AVs with trusted entities while limiting latency and computing requirements that enable spoofing or man-in-the-middle attacks.The passage identifies authentication latency and computational cost as security conditions requiring attention.
D. Projects: AV research projects and SDO activities focus on MEC, network slicing, 5G NR, ZSM, and AI for deploying autonomous-vehicle applications.Network slicing and MEC receive particular focus because they support AV-related applications and services.
1) Lessons Learned:: Standardization efforts address path planning, mobility, service migration, security, and privacy, but full deployment may still require several years.The survey notes that these efforts contribute to relevant 5G/B5G technical aspects while deployment remains incomplete.
2) Possible Future Directions:: Core 5G SDOs lack dedicated AV standardization because AV remains treated as one 5G/B5G use case.The proposed response is greater AV-stakeholder participation and AV-focused subgroups within core 5G SDOs.
2) Possible Future Directions:: Stronger cooperation between AV SDOs and core 5G SDOs, supported by joint research and standards activities, is needed.Such cooperation is presented as a way to resolve standardization issues across the two communities.
1) Quantum Computing:: Quantum computing could support AV route planning, high-speed processing, smart transportation systems, and quantum security against security breaches.The passage links these benefits to AV demands for substantial computing power and protection from external threats.