Next Article in Journal
The EU and Sustainable Low-Emission Transport? Current State and Challenges of Environmentally Sustainable and Low-Emission Transport in the EU
Next Article in Special Issue
A Comparative Evaluation of Machine-Learning Models for Road Surface Roughness Forecasting in ITSs
Previous Article in Journal
From BRT to Multimodality: A Cost-Efficiency Comparison of Public Transport Systems in Curitiba and Lisbon
Previous Article in Special Issue
Dynamic Multi-Relation Learning with Multi-Scale Hypergraph Transformer for Multi-Modal Traffic Forecasting
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Future Transportation
Volume 6
Issue 3
10.3390/futuretransp6030103
futuretransp-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Cooperative Connected and Automated Mobility: A Survey

by
Ang Ji
1,
Xilu Ju
1,
Nieyangzi Liu
2,*,
Junxian Chen
3 and
Zhe Dai
3
1
School of Transportation and Logistics, Southwest Jiaotong University, Chengdu 611756, China
2
Key Laboratory of Transport Industry of Management, Control and Cycle Repair Technology for Traffic Network Facilities in Ecological Security Barrier Area, Chang’an University, Xi’an 710054, China
3
School of Transportation Engineering, Chang’an University, Xi’an 710054, China
*
Author to whom correspondence should be addressed.
Future Transp. 2026, 6(3), 103; https://doi.org/10.3390/futuretransp6030103
Submission received: 27 March 2026 / Revised: 30 April 2026 / Accepted: 5 May 2026 / Published: 7 May 2026
(This article belongs to the Special Issue Recent Advances in Artificial Intelligence and Big Data for Intelligent Transportation Systems)
Browse Figures
Versions Notes

Abstract

Cooperative Connected and Automated Mobility (CCAM) is a critical paradigm for overcoming the limitations of single-vehicle intelligence and enabling coordinated intelligent transportation. To address the lack of systematic reviews towards recent CCAM advances, this paper presents a comprehensive review of relevant publications from the past five years. First, we establish a unified framework spanning communication, perception, decision-making, and control, and clarify the associated core components and technologies. Then, we identify three major bottlenecks that constrain large-scale CCAM deployment: uncertainty propagation along the perception-decision-control (PDC) chain, misalignment between functional safety and SOTIF standards, and inadequate end-to-end cybersecurity protection. In the context of 5G-A/6G, edge computing, and large-language-model-driven intelligence, we further propose targeted research directions. This survey aims to provide a systematic reference for theoretical investigation and engineering implementation.

1. Introduction

The rapid development of intelligent transportation has established cooperative connected and automated mobility (CCAM) as a key solution for alleviating traffic congestion, improving road safety, and reducing environmental pollution, as demonstrated by extensive studies on cooperative control, vehicle platooning, and comprehensive CCAM overviews [1]. By integrating vehicle-to-everything communication, automated driving, and multi-agent cooperation, CCAM enables coordinated interaction among vehicles, infrastructure, and cloud platforms, thereby significantly improving traffic efficiency and sustainability, consistent with findings from recent CCAM research [2].
Against this background, this paper aims to provide a comprehensive and systematic survey of technologies in connected and automated vehicles (CAV)/CCAM systems. Compared to existing reviews, this paper integrates multi-domain advances to establish a technology-safety-application analysis framework for cooperative mobility, informed by recent work on functional safety and distributed control [3,4]. With emphasis on representative studies published from 2021 to 2026, this survey aims to provide actionable guidance for academic research and engineering deployment.
Vehicle platooning improves throughput and energy efficiency with mature control strategies [5,6]. Cooperative perception and decision-making overcome single-vehicle limitations and enhance robustness in complex traffic [7]. CCAM deployment also depends on high-definition mapping, edge computing, and vehicle-road-cloud integration for system-level optimization [8].
Despite substantial progress, challenges remain in safety assurance, uncertainty management, cybersecurity, multi-domain integration, system robustness, and large-scale deployment [9].
This survey summarizes progress in CCAM across key technologies, representative applications, safety, and its social impacts. It is intended to provide a comprehensive state-of-the-art reference for both researchers and practitioners. The remainder of this paper is organized as follows. Section 2 introduces the system architecture and technical pillars of CCAM. Section 3 reviews key enabling technologies, and Section 4 discusses core technical challenges. Some future research directions are outlined in Section 5. Finally, Section 6 concludes the paper by summarizing the key findings and practical implications.

2. System Architecture and Technical Pillars

This survey primarily covers research from 2021 to 2026, while tracing representative foundational studies from 2015 to 2020 to reflect technology evolution. Priority was given to publications in leading journals and conferences.
The CCAM system is organized around a vehicle-road-cloud-human architecture. Using next-generation V2X, it enables end-to-end collaboration across four pillars: communication, cooperative perception, cooperative decision-making and planning, and cooperative control, overcoming single-vehicle limitations. This section reviews these pillars, their technological evolution, methods, and applications, proceeding from architecture to enabling technologies and deployment challenges.

2.1. Communication Architecture

Vehicle-to-Everything (V2X) communication provides low-latency, high-reliability links across V2V, V2I, V2P, V2N, and V2D channels [10], serving as a core enabler for vehicle-road-cloud-human collaboration.
Three ranges are defined: short (<100 m) using Bluetooth/BLE, ZigBee, UWB for low-power local interaction [11]; medium (100–300 m) using DSRC, Wi-Fi 6/7 for intersection coordination and platooning [12]; and long (300 m–5 km) using C-V2X/5G-NR V2X, with 5G-A further enabling deep vehicle-road-cloud coordination [13,14].
CCAM communication has shifted from DSRC to C-V2X-centered architectures. 5G-NR V2X offers enhanced physical-layer design, broadcast/multicast/unicast, centimeter-level positioning, and security/privacy, enabling high-speed platooning and wide-area collaboration.
Since 2024, 5G-A V2X has been commercially deployed, reducing latency below 0.5ms with network slicing and edge computing [15]. Early 6G-V2X research explores terahertz, Space-Air-Ground Integrated Network (SAGIN), and IRS [16]. Dynamic resource scheduling based on multi-agent reinforcement learning (MARL) and related approaches has improved both communication reliability and resource utilization under high-density traffic conditions [17,18].
Recent studies have compared resource allocation algorithms for V2X, showing that MARL-based methods achieve the highest resource utilization (92–96%) at the cost of moderate latency, while heuristic algorithms offer lower latency but reduced utilization [19].
Beyond 5G-NR V2X, intelligent reflecting surfaces (IRS), also termed as reconfigurable intelligent surfaces (RIS), have emerged as a promising paradigm for enhancing V2X coverage and latency performance. As highlighted in the seminal survey [20], IRS can passively reconfigure wireless propagation channels by adjusting the phase shifts of incident signals, converting non-line-of-sight (NLOS) links into virtual line-of-sight (VLOS) paths. In dense urban vehicular scenarios, this technique can significantly extend the effective communication range and reduce end-to-end latency, primarily by mitigating signal blockage from buildings and lowering retransmission rates due to improved channel quality.
By reflecting signals around obstacles, IRS compensates for severe path loss in NLOS conditions, reliably extending the communication distance in typical urban intersections. Meanwhile, the shorter VLOS path established by IRS reduces propagation delay, while the improved signal-to-noise ratio (SNR) substantially reduces packet error rates, eliminating the latency overhead of retransmissions.

2.2. Cooperative Perception

Cooperative perception uses V2X to fuse on-board (LiDAR, camera, radar) and roadside sensor data, overcoming line-of-sight and blind spots [21]. As shown in Figure 1, the urban architecture centers on roadside units (RSUs) with a three-level flow: on-board sensors → RSU/adjacent vehicles → cloud [22].
Furthermore, the central technical mechanism of cooperative perception is multi-source data fusion [23]. Depending on the fusion stage, methods are classified into early, intermediate, and late fusion. These categories exhibit different trade-offs in perception accuracy, bandwidth demand, and communication overhead, and therefore map to different CCAM scenarios (as demonstrated in Figure 2).
  • Early fusion aggregates and fuses raw data from on-board/roadside sensors, achieving high perception accuracy but imposing strict requirements on communication bandwidth and data synchronization. Recent studies have reduced bandwidth requirements by compressing raw point clouds through point clustering and sparsification [24].
  • Intermediate fusion (feature-level) balances accuracy and bandwidth and is the mainstream approach [25], with variants including attention-based and topology-based methods [26].
  • Late fusion fuses detection results of each perception node, with a small communication overhead but certain information loss. Fusion reliability can be improved through simplified target representation and motion consistency verification, suitable for V2I collaborative scenarios between roadside units and vehicles [27].
From 2015 to 2026, cooperative-perception research has shown clear trends toward multi-modal fusion, scenario refinement, and full-chain adaptation. In sensor configuration, LiDAR-only pipelines and Camera+LiDAR multi-modal combinations have emerged as practical choices that balance perception accuracy and environmental adaptability [28]. Regarding scenario coverage, the field has expanded from early single V2V/V2I settings to full V2X adaptation by 2024, supporting cooperative-perception requirements across highways, urban roads, parks, and related scenarios [29,30].
Quantitative validation based on the KITTI and DAIR-V2X datasets further confirms the advantages of multi-modal fusion. LiDAR-only perception achieves 85% target detection accuracy under normal conditions but drops to 62% in rainy weather due to laser attenuation, while the Camera+LiDAR scheme maintains 93% (normal) and 81% (rainy) accuracy by leveraging complementary visual and 3D spatial information [31]. For foggy or nighttime low-visibility scenarios, the “infrared sensor + 4D millimeter-wave radar” combination exhibits superior penetration: infrared thermal imaging identifies thermal targets (e.g., pedestrians) over 300 m in dense fog (visibility < 50 m), while 4D radar suppresses multi-path “ghost targets” via AI anti-interference algorithms [32]. This fusion scheme achieves over 92% mAP@50 in fog and 95% at nighttime, outperforming traditional Camera+LiDAR pipelines [33].
However, in practical connected environments, factors such as sensor noise, communication latency, and adversarial interference introduce perception uncertainty that, if left unaddressed, directly affects the safety and stability of downstream tasks. Therefore, uncertainty quantification serves as a critical bridge connecting cooperative perception with subsequent decision-making and control. A nonparametric probability-based hybrid method quantifies both model and data uncertainties, dynamically adjusting multi-source sensor weights to reduce the impact of unreliable inputs. For cooperative perception in platooning, this method reduces uncertainty error by 40% and lowers collision-warning false-alarm rates by 35%, thereby enhancing system robustness against sensor noise and adversarial data [32].
Beyond the above uncertainty quantification, benchmark datasets have been instrumental in advancing cooperative perception. Widely used resources include KITTI (urban/highway, stereo camera + LiDAR) [34], DAIR-V2X (vehicle-infrastructure cooperative perception) [31], Waymo Open (large-scale urban/suburban LiDAR+camera) [35], and nuScenes (360° multi-modal sensor suite) [36]. Regarding fusion strategies, intermediate fusion provides the best trade-off between accuracy and bandwidth by transmitting compact feature representations, whereas early fusion demands high-bandwidth links and late fusion sacrifices feature richness for low communication overhead [25]. The choice of fusion method should therefore align with the specific communication and latency constraints of each CCAM scenario.

2.3. Cooperative Decision-Making and Planning

Cooperative decision-making fuses perception data to generate behavior strategies and trajectory plans that satisfy safety and efficiency constraints, overcoming single-vehicle limits and enabling multi-vehicle cooperation [37]. It resolves traffic conflicts and optimizes resource allocation via global information sharing [38].
CCAM cooperative decision-making and planning adopts a hierarchical recursive architecture, composed of three levels: route planning, maneuver planning, and motion planning. The computation timescale, decision granularity, and optimization objectives are progressively refined across these levels, jointly forming a strategy-tactics-operation decision system (as shown in Table 1).
The three-level hierarchical architecture (strategic-tactical-operational) of CCAM cooperative decision-making can be further mapped to representative game-theoretic models and corresponding application scenarios. The strategic level focuses on global network-level coordination and congestion evacuation, supported by cooperative and Bayesian games for system optimality. The tactical level addresses interactive driving behaviors (e.g., right-of-way negotiation), where Stackelberg games and Level-K models capture leader-follower or human-like decision logic. The operational level realizes fine-grained trajectory and spacing control, with differential games and auction models ensuring real-time, collision-free performance. This hierarchical mapping clarifies the application boundaries of different game-theoretic models in CCAM decision-making. It provides a structured framework for selecting appropriate decision algorithms based on specific scenario requirements (e.g., system coordination vs. local trajectory control).
Mainstream CCAM cooperative decision-making and planning methods can be generally grouped into four categories aligned with different application requirements. Numerous studies have validated their technical characteristics, scenario suitability, and practical advantages and are increasingly transitioning to engineering pilots:
  • Cooperative methods rely on V2V real-time communication to realize space-time resource reservation and distributed negotiation between vehicles [39]. Local behavior collaboration is achieved through interactive information synchronization, which is mainly suitable for high-frequency vehicle interaction scenarios such as lane changing, short-distance overtaking, and car-following [40], effectively reducing the probability of multi-vehicle conflicts and improving local traffic flow efficiency [41].
  • Game theory-based methods model interactive decision-making relationships between vehicles and between vehicles and roads through Nash equilibrium, Stackelberg game, and other models, accurately depicting the interest demands and behavioral logic of traffic participants. They are mainly suitable for scenarios with multi-agent non-cooperative games, such as right-of-way allocation [42,43], mixed traffic flow collaboration [44], and accident collision avoidance [45], yielding decision outputs that are more aligned with natural human driving interaction patterns.
  • Optimization-based methods take Model Predictive Control (MPC) and linear quadratic regulator (LQR) as core technical carriers. Trajectory schemes are dynamically adjusted through rolling optimization and feedback mechanisms [46], which can effectively handle vehicle dynamics constraints, road boundary constraints, and obstacle avoidance requirements [47]. They are mainly suitable for scenarios with strict requirements on real-time performance, such as high-speed platooning, precise obstacle avoidance, and complex merging scenarios [48].
  • Learning-based methods take MARL as the core, adopting the paradigm of centralized training and decentralized execution [49]. Through massive scene data training, adaptive decision-making for complex dynamic traffic environments is realized without relying on accurate dynamics and environment models. They are mainly suitable for dynamic and complex scenarios such as congestion evacuation, highway merging, and large-scale vehicle collaboration, and have become a research hotspot and breakthrough direction in recent years [50].
Table 2 summarizes the characteristics of the above four categories.
Multi-method fusion and full-chain optimization are key trends in CCAM decision-making. For example, MARL handles complex interactions, while MPC ensures trajectory feasibility and stability. With V2X and edge computing, end-to-end perception-decision collaboration becomes feasible through uncertainty quantification and robust design. Additionally, federated learning and meta-learning address data privacy and cross-scenario generalization for large-scale deployment [51].

2.4. Cooperative Control

Cooperative control translates planned trajectories into vehicle motion [52], integrating dynamics modeling, multi-agent control, and tracking optimization for safe, stable, and comfortable group execution [53,54,55,56]. Its stack spans architecture, mechanisms, and algorithms.CCAM control architectures are categorized as centralized, distributed, or hybrid. Centralized control offers high precision but is latency-sensitive and suits small-scale scenarios (e.g., parks, closed roads) [57]. Distributed control enables real-time V2V-based coordination for large-scale open roads and platooning [58,59,60]. Hybrid designs combine both, applying distributed cooperation routinely and centralized intervention in complex conditions, making them a mainstream choice [61,62]. Table 3 summarizes the mapping between architectures, methodologies, and application scenarios.
Reinforcement learning and model predictive control are highlighted as the dominant methodologies across all architectures, particularly in distributed control scenarios. Platoon control and state estimation are also critical for ensuring stability and reliability in cooperative control.
Efficient cooperative control relies on two enabling technologies: vehicle-state estimation and information-flow topology optimization [68]. State estimation fuses data-driven and physics-based models using AI [69]. Topology optimization uses undirected/directed/bidirectional structures to reduce communication redundancy and improve reliability [64,65,70].
Current research indicates that no single algorithm can fully satisfy multi-objective collaborative requirements. As a result, algorithm fusion has become a mainstream solution. For example, MPC-MARL integration combines the trajectory-tracking precision of MPC with the dynamic-environment adaptability of MARL, which enables further optimization of cooperative-control performance [71,72].

3. Key Technologies

Beyond the four core technical pillars reviewed in Section 2, the engineering deployment of CCAM systems relies on three enabling technologies: V2X communication, on-board computing units (OBUs), and high-precision perception fusion. This section briefly highlights their roles with an emphasis on implementation considerations. For a systematic overview, readers are referred to [73].

3.1. V2X Communication Technology

As detailed in Section 2.1, V2X provides low-latency, high-reliability links for CCAM [74]. Commercial 5G-A V2X achieves sub-0.5 ms latency and centimeter-level positioning [75], while security mechanisms are increasingly integrated into V2X stacks [76]. Table 4 summarizes the main technology categories, core techniques, and representative references.

3.2. On-Board Units (OBU)

The OBU is the primary computational carrier for CCAM, executing sensor data processing, perception fusion, decision planning, and control algorithms [89]. Mainstream platforms adopt heterogeneous CPU+GPU+FPGA architectures, often with DSP/MCU, to balance parallel acceleration and low-latency control [90]. Specifically, CPU handles scheduling and logic, GPU accelerates LiDAR point-cloud and deep learning inference, FPGA ensures microsecond-level response for control tasks, and DSP/MCU serve signal processing and vehicle dynamics.
Table 5 summarizes representative OBU architectures. As shown, heterogeneous CPU+GPU+FPGA platforms and edge-cloud collaborative schemes are the dominant trends, reflecting the growing demand for flexible high-performance computing and task offloading.
OBUs typically include hardware redundancy (dual power supplies, dual compute cores) and support edge offloading to reduce onboard load [94]. Multiple communication interfaces (5G, C-V2X, CAN bus, Ethernet) enable comprehensive vehicle-road-cloud integration.

3.3. High-Precision Perception Fusion

As detailed in Section 2.2, cooperative perception fuses on-board and roadside sensor data to overcome line-of-sight and blind-spot limitations. From an engineering deployment perspective, two practical choices dominate: LiDAR-only pipelines for structured environments (e.g., highways) and Camera+LiDAR multi-modal combinations for urban scenarios, with emerging hardware such as 4D radar and infrared improving performance in adverse weather [33,103]. Table 6 provides a structured classification of perception fusion studies by methodology.
The above technologies form a collaborative support system. Their coordinated integration and co-optimization jointly support engineering deployment of CCAM technologies [73].

4. Core Challenges

Although CCAM has made substantial progress in communication architecture, cooperative perception, and decision-making and control, the current technical system still faces critical challenges in three dimensions: system reliability, functional safety/safety of the intended functionality (SOTIF), and cybersecurity. These dimensions are strongly coupled and mutually constraining, and together constitute the major bottleneck for large-scale CCAM deployment.

4.1. System Reliability

System reliability is a central problem in CCAM deployment. At its core, reliability reflects the system’s ability to maintain accurate, stable, and continuous perception, decision-making, and control in complex dynamic traffic environments. The main challenge arises from uncertainty across the full PDC pipeline. Such uncertainty propagates through generation, evolution, transmission, and amplification processes, ultimately causing performance degradation and potentially safety-critical outcomes [121].
Uncertainty primarily originates from two sources: external sensing noise and limited internal model generalization. In the perception stage, measurement noise, environmental interference, and calibration errors lead to perceptual deviations [122]. In the decision stage, limited decision model generalization and environmental unpredictability reduce decision quality and amplify uncertainty. In the control stage, control errors, model inaccuracies, and execution delays further degrade trajectory-tracking accuracy, causing additional uncertainty propagation.
In addition, V2X communication latency, packet loss, and synchronization errors can further exacerbate uncertainty. Although existing studies have proposed several mitigation methods, system-level coordinated mitigation across the full PDC pipeline remains limited, making it difficult to fundamentally resolve system-reliability issues [112].

4.2. Functional Safety and SOTIF

Functional safety and SOTIF are key prerequisites for large-scale deployment of CCAM, with the shared objective of preventing accidents caused by electrical/electronic failures or fault-free hazardous behavior. Accordingly, current CCAM systems face two major safety challenges: insufficient adaptation of standards and inadequate integration of human-machine interaction considerations [123].
Functional safety (ISO 26262) [124] primarily addresses systematic and random hardware failures, but its adaptation to AI models in CCAM remains insufficient. CCAM systems widely use AI, such as deep learning and MARL. However, their black-box characteristics introduce limited explainability and difficult-to-predict failure modes. As a result, traditional test methods (e.g., FTA and FMEA) are difficult to apply directly, and ISO 26262 alone cannot fully cover AI-induced safety risks [125].
SOTIF (ISO 21448) [126] addresses non-intended functional risks under fault-free conditions, but adaptation to human-machine interaction in CCAM remains limited. Human-machine co-driving is currently a widely adopted operating mode, yet SOTIF does not fully consider driver behavior patterns, cognitive characteristics, and takeover responses, which can lead to misuse risks when system decisions deviate from driver expectations [127]. In addition, cross-standard collaborative modeling between ISO 26262 and ISO 21448 is still insufficient. Most existing studies adopt one standard in isolation, while integrated optimization has not been fully realized, making it difficult to establish a comprehensive safety-assurance framework [128].

4.3. Cybersecurity

Cybersecurity is a baseline requirement for large-scale deployment of CCAM. Its core objectives are to resist external malicious attacks, protect information authenticity and integrity, and ensure operational safety. As CCAM systems become increasingly interconnected, the attack surface continues to expand. A key current weakness is the lack of sufficiently strong authentication and encryption mechanisms in V2X communication and in-vehicle buses, which increases vulnerability to malicious attacks that may lead to information leakage, system compromise, and even large-scale traffic incidents [129].
The CCAM attack surface spans the full stack, including perception, communication, computing, and execution layers: perception-layer attacks (e.g., GPS spoofing and LiDAR point-cloud forgery) can distort perception outputs [130], communication-layer attacks (e.g., replay, forgery, and DoS) can destabilize cooperative control [131], computing-layer attacks (e.g., malicious code injection) can disrupt computational processes, and execution-layer attacks (e.g., CAN-bus attacks) can directly manipulate vehicle motion states [85].
Current protection schemes remain mostly single-point defenses (e.g., encryption, authentication, and intrusion detection) and lack an integrated full-chain protection framework from perception to execution [132]. Moreover, existing schemes still face limitations in real-time performance, robustness, and lightweight deployment, making adaptation to the low-latency requirements of CCAM difficult [86,88].
A critical trade-off in CCAM cybersecurity is the latency overhead of security mechanisms. For safety-critical V2X functions, the 5G system targets an end-to-end latency below 1 ms according to 3GPP requirements [133]. Even lightweight authentication schemes incur measurable delays: for instance, the SALT-V framework achieves an average end-to-end latency of 1 ms and 0.035 ms computation time [87], which already consumes a substantial portion of the latency budget, leaving insufficient margin for network fluctuations or retransmissions. This overhead can directly compromise the real-time performance of cooperative collision avoidance and platooning. Therefore, future solutions must co-design security with communication constraints, exploring directions such as physical-layer security or edge-based distributed authentication.

5. Future Directions

Recent field deployments have demonstrated measurable benefits of CCAM. In the EU, the L3Pilot project, a flagship Horizon 2020 pilot testing automated driving on public roads across seven European countries with 70 vehicles and 750 test drivers, provides comprehensive evaluation results covering user acceptance, traffic safety, and environmental impacts [134]. In China, the Wuxi city-level vehicle-road-cloud pilot achieved a 15% congestion reduction at 46% lower infrastructure cost according to the TM Forum case study [135], and the Shanghai 5G-A zone demonstrated 20 ms end-to-end latency with 99% reliability [136]. In the US, the Waymo Driver (rider-only service) reported 96% fewer injury-causing intersection crashes over 56.7 million miles [137]. Building on these experiences and the remaining challenges, future research should address several directions to support the safe deployment of CCAM systems.

5.1. Cross-Standard Collaborative Modeling of Functional Safety and SOTIF

A cross-standard fusion design framework is needed to address the fragmented functional safety and SOTIF standards. Future research should clarify their applicability, verification processes, and compliance paths in cooperative mobility systems [138]. For ML/data-driven systems, explainable safety modeling, extreme scene generation, and adversarial robustness verification are required. An integrated system with unified safety requirements, joint risk assessment, and collaborative testing can achieve full lifecycle compliance from module to system level.For on-board embedded platforms, lightweight safety monitoring, online risk assessment, and real-time fault diagnosis should be developed [139], along with an SOTIF-oriented full lifecycle management framework for closed-loop safety iteration.
Beyond modeling and management frameworks, future research must prioritize concrete verification methods that jointly address hardware failures and AI performance limitations. Specifically, three directions are critical for moving from engineering prototypes to public-road deployment:
  • Scenario-based testing with coverage guarantees: Develop systematic methods to generate and prioritize edge cases and corner scenarios that challenge both perception and decision modules, ensuring that the testing coverage is quantifiable and traceable to operational design domains (ODD).
  • Uncertainty-aware perception with confidence bounds: Integrate uncertainty quantification into perception pipelines (e.g., Bayesian deep learning, ensemble methods) to provide downstream modules with reliable confidence estimates, enabling safe decision-making under sensing ambiguity.
  • Runtime monitoring with safe fallback mechanisms: Design lightweight monitors that detect prediction errors or out-of-distribution inputs in real time and trigger pre-defined fallback behaviors without violating safety constraints.
Early efforts have proposed unified safety frameworks [128], but systematic integration of these verification methods into learning-enabled CCAM systems remains an open challenge.

5.2. Construction of Full-Stack Cybersecurity Defense System for Vehicle-Road-Cloud Collaboration

To mitigate the expanding attack surface and strict real-time constraints in V2X and on-board networks, a lightweight, vehicle-road-cloud collaborative cybersecurity system is needed. For on-board terminal security, hardware-level isolation, lightweight encryption, and runtime monitoring should be developed [140]. For in-vehicle networks, real-time intrusion detection and attack traceability are required [141]. For 6G V2X scenarios, access-side DDoS collaborative defense methods should be studied [142]. In addition, federated learning-based intrusion detection frameworks can be applied to edge and charging network scenarios [143]. Overall, lightweight, adaptive, and co-designed security mechanisms must be prioritized for future CCAM deployment.

5.3. Integrated Technical System Optimization of Vehicle-Road-Cloud Collaboration

Future research should prioritize integrated system-level optimization of vehicle-road-cloud collaboration to achieve end-to-end coordination across communication, perception, decision-making, control, and safety.
With 6G and space-air-ground integration, the communication architecture should be optimized for wide-area coverage, low latency, and high reliability, while network slicing and dynamic resource allocation methods are further developed. Distributed perception and decision-making frameworks should be established, offloading selected tasks to edge and cloud nodes to improve accuracy and efficiency.
A three-layer roadside-edge-cloud collaborative perception prototype should be developed for global environmental perception [92], and federated scheduling methods can improve resource utilization for roadside sensors. Joint service deployment and task offloading in vehicle-edge-cloud networks should be investigated to achieve low-latency, high-reliability scheduling [93]. Deep-learning-driven vehicle-road collaborative perception and intelligent scheduling methods are also needed for complex traffic conditions.
For engineering deployment, lightweight, interoperable, and cost-effective technologies should be enhanced, including lightweight algorithms, unified interfaces, and low-cost infrastructure. Mixed-traffic collaborative control strategies for CCAM and human-driven vehicles should be investigated to improve overall safety and efficiency [144].

6. Conclusions

This paper systematically reviews recent progress for CCAM and provides a comprehensive synthesis of the system architecture, enabling technologies, core challenges, and development trends within a vehicle-road-cloud integrated paradigm. The analysis indicates that CCAM has established a relatively complete technical chain, in which V2X communication provides the foundation, cooperative perception delivers key inputs, cooperative decision-making and planning constitute the core, and cooperative control executes the final motions. Supported by technologies such as on-board computing and multi-source perception fusion, CCAM has advanced from conceptual and theoretical exploration toward engineering applications.
However, substantial challenges remain for large-scale deployment and vehicle-level safety assurance. First, multi-dimentional uncertainty across the full PDC pipeline hinders unified modeling and proactive mitigation, directly constraining system reliability. Then, functional safety and SOTIF frameworks remain insufficiently integrated, while safety-verification mechanisms for machine-learning models and human-machine co-driving adaptation are still inadequate. Also, the attack surface in vehicle-road-cloud collaborative scenarios continues to expand, yet a lightweight, integrated, and full-lifecycle cybersecurity framework has not been fully established.
In the future, CCAM vehicle-control technologies are expected to evolve toward proactive full-chain uncertainty management, multi-standard safety integration, vehicle-road-cloud collaboration, and safety-oriented optimization for human-machine co-driving. Through coordinated advances in theory, architecture, and adaptation, the overall reliability, safety, and practical deployability of CCAM systems are expected to improve substantially.

Author Contributions

Conceptualization, A.J.; methodology, A.J. and X.J.; validation, N.L., Z.D. and A.J.; data curation, J.C.; resources, J.C.; writing—original draft preparation, A.J. and X.J.; writing—review and editing, N.L., J.C. and Z.D.; visualization, X.J.; supervision, Z.D.; funding acquisition, N.L. All authors have read and agreed to the published version of the manuscript.

Funding

The research was funded by the Fundamental Research Funds for the Central Universities, CHD [Grant number: 300102345506], the National Natural Science Foundation of China [Grant number: 52402398], the Natural Science Foundation of Shaanxi Province [Grant number: 2024JC-YBQN-0369], and the Youth Talent Support Program of Xi’an Science and Technology Society [Grant number: 0959202513001].

Data Availability Statement

No new data were created in this research.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Li, H.; Meng, W.; Han, Z.; Zhang, Z.; Yang, Y. Vehicle platoon in road traffic: A survey of modeling, communication, controlling and perspectives. Physica A 2025, 674, 130757. [Google Scholar] [CrossRef]
  2. Liu, B.; Han, W.; Wang, E.; Xiong, S.; Qiao, C.; Wang, J. An Efficient Message Dissemination Scheme for Cooperative Drivings via Cooperative Hierarchical Attention Reinforcement Learning. IEEE Trans. Mob. Comput. 2024, 23, 5527–5542. [Google Scholar] [CrossRef]
  3. Liu, Q.; Xiong, P.; Zhu, Q.; Xiao, W.; Li, G.; Hu, G. A cooperative lateral and vertical control strategy for autonomous vehicles based on multi-agent deep reinforcement learning. Proc. Inst. Mech. Eng. Part D J. Automob. Eng. 2026, 240, 665–680. [Google Scholar] [CrossRef]
  4. Dong, L.; Li, X.; He, X.; Hua, M.; Zhou, Q.; Sun, C.; Jiang, K. Robustness-enhanced cooperative adaptive cruise control for multi-task scenarios via generalised joint multi-agent reinforcement learning. Neurocomputing 2026, 664, 132036. [Google Scholar] [CrossRef]
  5. Hua, M.; Chen, D.; Jiang, K.; Zhang, F.; Wang, J.; Wang, B.; Zhou, Q.; Xu, H. Communication-Efficient MARL for Platoon Stability and Energy-Efficiency Co-Optimization in Cooperative Adaptive Cruise Control of CAVs. IEEE Trans. Veh. Technol. 2025, 74, 6076–6087. [Google Scholar] [CrossRef]
  6. Olivari, E.; Carboni, A.; Caballini, C.; Pasquale, C.; Dalla Chiara, B.; Sacone, S. Planning and Control Strategies for Truck Platooning: A Benefit-Driven Literature Review. Future Transp. 2025, 5, 187. [Google Scholar] [CrossRef]
  7. Han, Y.; Zhang, L.; Meng, D.; Zhang, Z.; Hu, X.; Weng, S. A Value-Based Parallel Update MCTS Method for Multi-Agent Cooperative Decision-Making of Connected and Automated Vehicles. IEEE Trans. Intell. Transp. Syst. 2026, 27, 1400–1415. [Google Scholar] [CrossRef]
  8. Li, S.; Wang, H.; Hu, J. Ecological driving at an actuated signalized intersection: A practical solution of Vehicle-Road-Cloud Integration system. Transp. Res. Part C Emerg. Technol. 2025, 177, 105198. [Google Scholar] [CrossRef]
  9. Chen, J.; Zhang, Y.; Zhou, Y.; Wu, Y.; Chung, E.; Sartoretti, G. Learning-Based Lane Selection and Driving Orders for Connected Automated Vehicles at Multi-Lane Freeway Merging Sections. IEEE Trans. Intell. Transp. Syst. 2026, 27, 3369–3382. [Google Scholar] [CrossRef]
  10. Cao, X.; Wang, S.; Ren, X. IRS-Enhanced V2X Communication and Computation Systems: Resource Allocation and Performance Optimization. IEEE Internet Things J. 2025, 12, 9180–9194. [Google Scholar] [CrossRef]
  11. Li, Q.; Shi, C.; Xiong, Z.; Xiong, J.; Chen, M. Cooperative Fault-Tolerant Positioning Based on Model-Mixed Robust Gaussian Belief Propagation. IEEE Internet Things J. 2025, 12, 21537–21551. [Google Scholar] [CrossRef]
  12. Wang, S.; Ding, Y.; Ding, X.; Tan, X. Solving the Vehicle Cooperation Problem at Signal-Free Intersection via an Asynchronous Deep Reinforcement Learning Approach. IEEE Internet Things J. 2025, 12, 23357–23372. [Google Scholar] [CrossRef]
  13. Xu, C.; Wang, G.; Wei, M.; Zhang, P.; Peng, B. Intelligent Transportation Vehicle Road Collaboration and Task Scheduling Based on Deep Learning in Augmented Internet of Things. IEEE Trans. Veh. Technol. 2025, 74, 2198–2209. [Google Scholar] [CrossRef]
  14. Gabilondo, A.; Fernandez, Z.; Martin, A.; Zorrilla, M.; Angueira, P.; Montalban, J. Dynamic Mobile Network Slicing Through Vehicular Traffic Analysis. IEEE Open J. Veh. Technol. 2025, 6, 1464–1480. [Google Scholar] [CrossRef]
  15. Mohammed, S.F.; Pan, Z.; Al-Gunid, H.M.; Qasem, Z.A.H. Optimizing end-to-end latency in C-V2X networks: A novel FD-RAN and MEC integration approach. Veh. Commun. 2025, 55, 100955. [Google Scholar] [CrossRef]
  16. Lu, Z.; Zhao, Y.; Xu, C.Z. RCSIL: RIS-Assisted Cooperative Channel State Information Localization for V2X System. IEEE Internet Things J. 2025, 12, 1016–1031. [Google Scholar] [CrossRef]
  17. Saad, M.M.; Tariq, M.A.; Ajmal, M.; Kim, D.; Srivastava, G. Federated Multiagent Reinforcement Learning for Resource Allocation in NR-V2X Mode 2. IEEE Internet Things J. 2025, 12, 23402–23417. [Google Scholar] [CrossRef]
  18. Xu, Y.; Zheng, L.; Wu, X.; Tang, Y.; Liu, W.; Sun, D. Joint Resource Allocation for UAV-Assisted V2X Communication with Mean Field Multi-Agent Reinforcement Learning. IEEE Trans. Veh. Technol. 2025, 74, 1209–1223. [Google Scholar] [CrossRef]
  19. Yang, C.; Kwong, C.F.; Chieng, D.; Kar, P.; Yau, K.-L.A.; Chen, Y. Navigating the Road Ahead: A Comprehensive Survey of Radio Resource Allocation for Vehicle Platooning in C-V2X Communications. IEEE Commun. Surv. Tutor. 2025, 27, 1326–1362. [Google Scholar] [CrossRef]
  20. Di Renzo, M.; Zappone, A.; Debbah, M.; Alouini, M.-S.; Yuen, C.; de Rosny, J.; Tretyakov, S. Smart Radio Environments Empowered by Reconfigurable Intelligent Surfaces: How It Works, State of Research, and The Road Ahead. IEEE J. Sel. Areas Commun. 2020, 38, 2450–2525. [Google Scholar] [CrossRef]
  21. Li, K.; Dai, Z.; Cui, H.; Wang, X.; Song, H. VRAR: Video-Radar Automatic Registration Method Based on Trajectory Spatiotemporal Features and Bidirectional Mapping. IEEE Trans. Circuits Syst. Video Technol. 2025, 35, 8707–8722. [Google Scholar] [CrossRef]
  22. Chen, S.; Wang, H.; Shi, Y.; Zhao, L.; Hu, J.; Xu, H.; Song, Y. S&C-V2X: Integrated Satellite-Cellular V2X Networks - Evolution of C-V2X from 5G to 6G. IEEE Commun. Mag. 2026, 64, 175–181. [Google Scholar] [CrossRef]
  23. Li, K.; Dai, Z.; Huang, X.; Dong, S.; Wang, S. HMT-SUV: Hierarchical Multi-sensor Tracking with Single-source Uncertainty Verification for Resolving Target Existence Ambiguity. Expert Syst. Appl. 2026, 310, 131297. [Google Scholar] [CrossRef]
  24. Zhao, C.; Ding, D.; Lei, C.; Wang, S.; Ji, Y.; Du, Y. Safety Field-Based Vehicle-Infrastructure Cooperative Perception for Autonomous Driving Using 3D Point Clouds. IEEE Trans. Intell. Transp. Syst. 2025, 26, 4676–4691. [Google Scholar] [CrossRef]
  25. Yazgan, M.; Graf, T.; Liu, M.; Fleck, T.; Zöllner, J.M. A Survey on Intermediate Fusion Methods for Collaborative Perception Categorized by Real World Challenges. arXiv 2024, arXiv:2404.16139. [Google Scholar] [CrossRef]
  26. Wang, K.; Ma, Q.; Jiang, Y.; Lu, J. Uncertainty Quantification for Safety of the Intended Functionality of Autonomous Driving: A Comprehensive Survey. IEEE Trans. Instrum. Meas. 2025, 74, 3602549. [Google Scholar] [CrossRef]
  27. Jiang, J.; Xu, X.; He, C.; Liang, C.; Chen, T.; Wang, K.; Zhou, M.; Bai, A. A review of Link-level uncertainty in the perception-decision-control pipeline of connected and autonomous vehicles: Generation, evolution, propagation, and amplification. Proc. Inst. Mech. Eng. Part D J. Automob. Eng. 2025. [Google Scholar] [CrossRef]
  28. Huang, C.; Feng, F.; Duan, C.; Chen, Z.; Mao, P.; Yuan, X.; Zhu, X. To lock the authentic signals: Mitigating GNSS spoofing with INS-aided tracking. Inf. Fusion 2026, 126, 103596. [Google Scholar] [CrossRef]
  29. Wang, Z.; Fu, Z.; Wu, Z.; Zhu, C.; Ji, B. Barrier Lyapunov Function-Based Prescribed-Time Cooperative Guidance Law with Field-of-View Constraint. IEEE Trans. Control Syst. Technol. 2025, 33, 2217–2229. [Google Scholar] [CrossRef]
  30. Abdi, B.; Rokhi, Z.; Vidal, C.; Emadi, A. Scene-Centric Vehicle Trajectory Prediction at Cooperative Intersection Using Decision-Aware Attention Graph Transformer. IEEE Trans. Intell. Transp. Syst. 2025, 26, 19322–19333. [Google Scholar] [CrossRef]
  31. Wang, H.; Niu, Y.; Chen, L.; Li, Y.; Sotelo, M.A.; Li, Z.; Cai, Y. DAIR-V2XReid: A New Real-World Vehicle-Infrastructure Cooperative Re-ID Dataset and Cross-Shot Feature Aggregation Network Perception Method. IEEE Trans. Intell. Transp. Syst. 2024, 25, 9058–9068. [Google Scholar] [CrossRef]
  32. Sedar, R.; Kalalas, C.; Dini, P.; Vazquez-Gallego, F.; Alonso-Zarate, J.; Alonso, L. Knowledge Transfer for Collaborative Misbehavior Detection in Untrusted Vehicular Environments. IEEE Trans. Veh. Technol. 2025, 74, 425–440. [Google Scholar] [CrossRef]
  33. Zhang, B.; Wang, Y.; Zhang, C.; Jiang, J.; Luo, X.; Wang, X.; Zhang, Y.; Liu, Z.; Shen, G.; Ye, Y.; et al. FogFusion: Robust 3D object detection based on camera-LiDAR fusion for autonomous driving in foggy weather conditions. Proc. Inst. Mech. Eng. Part D J. Automob. Eng. 2025, 240, 2312–2322. [Google Scholar] [CrossRef]
  34. Geiger, A.; Lenz, P.; Stiller, C.; Urtasun, R. Vision meets robotics: The KITTI Dataset. Int. J. Robot. Res. 2013, 32, 1231–1237. [Google Scholar] [CrossRef]
  35. Sun, P.; Kretzschmar, H.; Dotiwalla, X.; Chouard, A.; Patnaik, V.; Tsui, P.; Guo, J.; Zhou, Y.; Chai, Y.; Caine, B.; et al. Scalability in Perception for Autonomous Driving: Waymo Open Dataset. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020; pp. 2443–2451. [Google Scholar]
  36. Caesar, H.; Bankiti, V.; Lang, A.H.; Vora, S.; Liong, V.E.; Xu, Q.; Krishnan, A.; Pan, Y.; Baldan, G.; Beijbom, O. nuScenes: A Multimodal Dataset for Autonomous Driving. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020; pp. 11618–11628. [Google Scholar]
  37. Wang, J.; Ma, Z.; Zhu, X.; Bai, J.; Huang, L. Cooperative decision making for connected automated vehicles in multiple driving scenarios. IET Intell. Transp. Syst. 2023, 17, 2131–2142. [Google Scholar] [CrossRef]
  38. Peng, R.; Yang, M.; Tao, R.; Zhang, M.; Zhang, R. Hierarchical Control Strategy for Cooperative On-Ramp Merging of Connected and Automated Vehicles on Multilane Highways. IEEE Internet Things J. 2025, 12, 34513–34527. [Google Scholar] [CrossRef]
  39. Hidalgo, C.; Arizala, A.; Iturbe-Olleta, N.; Brazalez, A.; Zubizarreta, A.; Asua, E.; Rastelli, J.P. A Cooperative Driving Framework for Platooning Using V2X Messages in Urban Environments. IEEE Trans. Veh. Technol. 2025, 74, 16813–16826. [Google Scholar] [CrossRef]
  40. Feng, T.; E, W.; Kong, X.; Li, B.; Wang, X.; Hu, X.; Ding, Y.; Zheng, J.; Duan, X. Multivehicle cooperative control strategy in freeway merging areas under visible light communication environment. Transp. B 2025, 13, 2475216. [Google Scholar] [CrossRef]
  41. Prathiba, S.B.; Raja, G.; Kumar, N. Intelligent Cooperative Collision Avoidance at Overtaking and Lane Changing Maneuver in 6G-V2X Communications. IEEE Trans. Veh. Technol. 2022, 71, 112–122. [Google Scholar] [CrossRef]
  42. Jing, D.; Yao, E.; Chen, R. Decentralized human-like control strategy of mixed-flow multi-vehicle interactions at uncontrolled intersections: A game-theoretic approach. Transp. Res. Part C Emerg. Technol. 2024, 167, 104835. [Google Scholar] [CrossRef]
  43. Sun, Z.; Qin, Z.; Ma, R.; Huang, T.; Gao, Z.; Ji, A. Microscopic right-of-way trading mechanism for cooperative decision-making: Theories and preliminary results. IEEE Trans. Intell. Transp. Syst. 2025, 26, 2461–2477. [Google Scholar] [CrossRef]
  44. Cui, Y.; Fang, S.; Chen, Q.; Wang, Y.; Hang, P.; Sun, J. A Game-Theoretic Framework of Interaction and Cooperative Driving for CAVs at Mixed Unsignalized Intersections. IEEE Internet Things J. 2026, 13, 1524–1538. [Google Scholar] [CrossRef]
  45. Guo, W.; Song, X.; Zhang, W.; Li, J.; Wu, X. Game-Theoretic Shared Control Strategy for Cooperative Collision Avoidance Under Extreme Conditions. IEEE Trans. Veh. Technol. 2025, 74, 246–262. [Google Scholar] [CrossRef]
  46. Koehler, M.; Mueller, M.A.; Allgoewer, F. Distributed MPC for Self-Organized Cooperation of Multiagent Systems. IEEE Trans. Autom. Control 2024, 69, 7988–7995. [Google Scholar] [CrossRef]
  47. Li, Y.; Liu, C.; Zheng, F. Robust fuzzy model predictive control for connected and automated vehicles in mixed platoons using a bidirectional vehicle dynamics strategy. Expert Syst. Appl. 2025, 267, 126057. [Google Scholar] [CrossRef]
  48. Mohammadi, M.; Maleki, M.; Taghavipour, A. Multi-Parametric model predictive control for vehicle platoons. Proc. Inst. Mech. Eng. Part D J. Automob. Eng. 2025, 240, 962–975. [Google Scholar] [CrossRef]
  49. Zhao, R.; Sun, Z.; Ji, A. A deep reinforcement learning approach for automated on-ramp merging. In Proceedings of the 2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC), Macau, China, 8–12 October 2022; pp. 3800–3806. [Google Scholar] [CrossRef]
  50. Hua, M.; Qi, X.; Chen, D.; Jiang, K.; Liu, Z.E.; Sun, H.; Zhou, Q.; Xu, H. Multi-agent reinforcement learning for connected and automated vehicles control: Recent advancements and future prospects. IEEE Trans. Autom. Sci. Eng. 2025, 22, 16266–16286. [Google Scholar] [CrossRef]
  51. Parada, L.; Yu, K.; Angeloudis, P. IntNet: A Communication-Driven Multi-Agent Reinforcement Learning Framework for Cooperative Autonomous Driving. IEEE Robot. Autom. Lett. 2025, 10, 2478–2485. [Google Scholar] [CrossRef]
  52. Mu, Z.; Avedisov, S.S.; Moradipari, A.; Park, B.B. Formation and Investigation of Cooperative Platooning at the Early Stage of Connected and Automated Vehicles Deployment. IEEE Trans. Intell. Transp. Syst. 2026, 27, 5642–5656. [Google Scholar] [CrossRef]
  53. Li, M.; Wang, B.; Wang, S.; Ke, Z. Serial distributed reinforcement learning for enhanced multi-objective platoon control in curved road coordinates. Expert Syst. Appl. 2025, 269, 126493. [Google Scholar] [CrossRef]
  54. Yao, X.; Du, Z.; Sun, Z.; Calvert, S.C.; Ji, A. Cooperative lane-changing in mixed traffic: A deep reinforcement learning approach. Transp. A 2026, 22, 2343048. [Google Scholar] [CrossRef]
  55. Ji, A.; Huang, J.; Qin, Z.; Sun, Z.; Zhao, R.; Zheng, G. Cooperative merging for connected automated vehicles in mixed traffic: A multi-agent reinforcement learning approach. Artif. Intell. Transp. 2025, 2, 100007. [Google Scholar] [CrossRef]
  56. Berger, T.; Besselink, B. String Stability and Guaranteed Safety via Funnel Cruise Control for Vehicle Platoons. IEEE Trans. Autom. Control 2026, 71, 81–90. [Google Scholar] [CrossRef]
  57. Liu, W.; Hua, M.; Deng, Z.; Meng, Z.; Huang, Y.; Hu, C.; Song, S.; Gao, L.; Liu, C.; Shuai, B.; et al. A Systematic Survey of Control Techniques and Applications in Connected and Automated Vehicles. IEEE Internet Things J. 2023, 10, 21892–21916. [Google Scholar] [CrossRef]
  58. Fan, B.; Xie, H.; Li, T. Platoon Communication Power Control Under V2V Data Uncertainty: A Robust DRL Approach. IEEE Trans. Intell. Transp. Syst. 2025, 26, 21559–21573. [Google Scholar] [CrossRef]
  59. Gao, J.; Peng, Z.; Jang, S.; Piao, C. Platoon Control Leveraging Network Performance and State Estimation Under Dynamic V2V Network. IEEE Trans. Intell. Transp. Syst. 2025, 26, 12093–12105. [Google Scholar] [CrossRef]
  60. Liang, X.; Gao, W.; Hu, C.; Chai, T. Cooperative Adaptive Cruise Control of Connected and Autonomous Vehicles via Hybrid Iteration. IEEE Trans. Veh. Technol. 2025, 1–12. [Google Scholar] [CrossRef]
  61. Sun, Z.; Gao, Z.; He, X.; Li, Z.; Huang, T. Cooperative Merging in Mixed Traffic: A Mobile-Edge Hybrid Control Framework. IEEE Trans. Intell. Transp. Syst. 2025, 26, 4837–4850. [Google Scholar] [CrossRef]
  62. Rzig, I.; Jaafar, W.; Jebalia, M.; Tabbane, S. Energy-Efficient Vehicular Task Offloading Using Multi-Mode MEC and RIS-Equipped Aerial Platforms. IEEE Open J. Commun. Soc. 2025, 6, 7604–7619. [Google Scholar] [CrossRef]
  63. Qi, F.; Huang, Q.; Wang, Y.; Lin, P.; Hu, C.; Shi, X. IRS-Assisted Joint Broadcast/Multicast Resource Allocation and Cooperative Perception Optimization for 6G V2X Networks. IEEE Trans. Broadcast. 2025, 72, 106–116. [Google Scholar] [CrossRef]
  64. Li, S.; Zheng, H.; Wang, J.; Chen, C.; Xu, Q.; Wang, J.; Li, K. Influence of information flow topology and maximum platoon size on mixed traffic stability. Transp. Res. Part C Emerg. Technol. 2025, 171, 104950. [Google Scholar] [CrossRef]
  65. Wang, X.; Feng, F.; Zheng, J.; Hu, X.; Wenjuan, E.; Wang, X.; Xiao, Y.; Liu, Y.; Li, T. AutoML-BIMCTS: Optimizing Information Flow Topology for Heterogeneous Vehicle Platoons Under Communication Constraints. IEEE Internet Things J. 2025, 12, 40666–40685. [Google Scholar] [CrossRef]
  66. Yu, J.; Jiang, F.; Luo, Y.; Kong, W. Networked predictive control method of multi-vehicle cooperative control at communication-constrained unsignalized multi-intersection. IET Intell. Transp. Syst. 2023, 17, 929–942. [Google Scholar] [CrossRef]
  67. Li, Z.; Wang, Q.; Wang, J.; He, Z. A Flexible Cooperative MARL Method for Efficient Passage of an Emergency CAV in Mixed Traffic. IEEE Trans. Intell. Transp. Syst. 2024, 25, 8898–8912. [Google Scholar] [CrossRef]
  68. Gao, J.; Yu, B.; Chen, Y.; Wang, J.; Dai, Z.; Gao, K. Meta-MSCC: A foundation model for adaptive CAV control in highway weaving segments. Transp. Res. Part C Emerg. Technol. 2025, 181, 105397. [Google Scholar] [CrossRef]
  69. Chen, D.; Hu, J.; Sun, G.; Rong, F.; Zhang, P.; Huang, Y.; Cao, Z. Vehicle Sideslip Angle Redundant Estimation Based on Multi-Source Sensor Information Fusion. Mathematics 2026, 14, 183. [Google Scholar] [CrossRef]
  70. Hu, Z.; Chang, Y.; Chen, M.; Wang, B.; Yang, Z.; Qin, H. Cooperative Robust Fault-Tolerant Control for Vehicular Platoon Systems with Performance Limitations and Various Uncertainties. IEEE Trans. Veh. Technol. 2025, 74, 15065–15079. [Google Scholar] [CrossRef]
  71. Annu; Rajalakshmi, P. Towards 6G V2X Sidelink: Survey of Resource Allocation—Mathematical Formulations, Challenges, and Proposed Solutions. IEEE Open J. Veh. Technol. 2024, 5, 344–383. [Google Scholar] [CrossRef]
  72. Fan, X.; Bu, C.; Zhao, X.; Sui, J.; Mo, H. Incremental Double Q-Learning-Enhanced MPC for Trajectory Tracking of Mobile Robots. IEEE Trans. Instrum. Meas. 2025, 74, 3545523. [Google Scholar] [CrossRef]
  73. Liu, L.; Liu, S.; Shi, W. 4C: A Computation, Communication, and Control Co-Design Framework for CAVs. IEEE Wirel. Commun. 2021, 28, 42–48. [Google Scholar] [CrossRef]
  74. Maglogiannis, V.; Naudts, D.; Hadiwardoyo, S.; van den Akker, D.; Marquez-Barja, J.; Moerman, I. Experimental V2X Evaluation for C-V2X and ITS-G5 Technologies in a Real-Life Highway Environment. IEEE Trans. Netw. Serv. Manag. 2022, 19, 1521–1538. [Google Scholar] [CrossRef]
  75. Lang, P.; Tian, D.; Han, X.; Zhang, P.; Duan, X.; Zhou, J.; Leung, V.C.M. Towards 6G vehicular networks: Vision, technologies, and open challenges. Comput. Netw. 2025, 257, 110916. [Google Scholar] [CrossRef]
  76. Ferrag, M.A.; Lakas, A.; Tihanyi, N.; Debbah, M. LLM and AI Agents for Autonomous Systems: A Survey of Applications, Datasets, and Security Challenges. IEEE Open J. Intell. Transp. Syst. 2026, 7, 615–657. [Google Scholar] [CrossRef]
  77. Xue, S.; Gong, S.; Li, X. A Comparative Study of IEEE 802.11bd and IEEE 802.11p on the Data Dissemination Properties in Dynamic Traffic Scenarios. Appl. Sci. 2024, 14, 2099. [Google Scholar] [CrossRef]
  78. Srikanth, S.V.; Koshy, S.S.; Harin, R.; Prasad, M.; Krishna, C.M. India’s V2X Future: C-V2X vs. DSRC—A Comparative Study. In Proceedings of the 10th International Conference on Engineering and Emerging Technologies (ICEET), Dubai, United Arab Emirates, 27–28 December 2024; pp. 1–6. [Google Scholar]
  79. Khan, W.U.; Mahmood, A.; Bozorgchenani, A.; Jamshed, M.A.; Ranjha, A.; Lagunas, E.; Pervaiz, H.; Chatzinotas, S.; Ottersten, B.; Popovski, P. Opportunities for Intelligent Reflecting Surfaces in 6G Empowered V2X Communications. IEEE Internet Things Mag. 2024, 7, 72–79. [Google Scholar] [CrossRef]
  80. Liu, Y.; Zhao, J.; Xia, M.; Zhang, Q. Low-Latency V2X Communication with Non-Orthogonal Slicing: A Deterministic Network Calculus Approach. IEEE Trans. Veh. Technol. 2025, 74, 13759–13769. [Google Scholar] [CrossRef]
  81. Nair, A.; Tanwar, S. Resource allocation in V2X communication: State-of-the-art and research challenges. Phys. Commun. 2024, 64, 102351. [Google Scholar] [CrossRef]
  82. Huan, Z.; Sun, J.; Chen, Z.; Zhang, Z.; Sun, X.; Xiao, Z. Resource Allocation in V2X Networks: A Double Deep Q-Network Approach with Graph Neural Networks. Comput. Mater. Contin. 2025, 84, 5427–5443. [Google Scholar] [CrossRef]
  83. Xu, Y.; Zheng, L.; Wu, X.; Tang, Y.; Liu, W.; Sun, D. Energy-Efficient Resource Allocation for V2X Communications. IEEE Internet Things J. 2024, 11, 30014–30026. [Google Scholar] [CrossRef]
  84. Wang, J.; Topilin, I.; Feofilova, A.; Shao, M.; Wang, Y. Cooperative Intelligent Transport Systems: The Impact of C-V2X Communication Technologies on Road Safety and Traffic Efficiency. Sensors 2025, 25, 2132. [Google Scholar] [CrossRef]
  85. Yuan, J.; Sun, H.; Cao, R.; Huang, G. RL-IDS: A robust and lightweight intrusion detection system for in-vehicle network. J. Inf. Secur. Appl. 2026, 97, 104361. [Google Scholar] [CrossRef]
  86. Gong, B.; Li, Z.; Gong, M.; Zhu, H.; Meng, W.; Guo, C. Lightweight Continuous Authentication via IMU Fingerprinting for V2X. IEEE Trans. Intell. Transp. Syst. 2025, 26, 10483–10495. [Google Scholar] [CrossRef]
  87. Cao, L.; Wang, W.; Xie, Q.; Wei, D.; Zhang, L. SALT-V: Lightweight Authentication for 5G V2X Broadcasting. arXiv 2025, arXiv:2511.11028. [Google Scholar] [CrossRef]
  88. Rajasekaran, A.S.; Kumar Das, A.; Maria, A.; Sundar Kola, K.; Dharavat, N.; Jo, M. Lightweight Chebyshev Polynomial-Based Authentication and Signature Framework with Antenna Array for IoT-Enabled V2V and V2X Communications. IEEE Internet Things J. 2025, 12, 55227–55240. [Google Scholar] [CrossRef]
  89. Ren, R.; Zhao, J.; Zou, D.; Zhang, Q.; Wang, D.; Xu, W. Collaborative Computation in Integrated Sensing, Communication, and Computation System for Autonomous Driving. IEEE Trans. Intell. Transp. Syst. 2026, 27, 883–894. [Google Scholar] [CrossRef]
  90. Carballo-Hernandez, W.; Pelcat, M.; Bhattacharyya, S.S.; Carmona Galan, R.; Berry, F. Flydeling: Streamlined Performance Models for Hardware Acceleration of CNNs through System Identification. ACM Trans. Model. Perform. Eval. Comput. Syst. 2023, 8, 1–33. [Google Scholar] [CrossRef]
  91. Khan, M.J.; Khan, M.A.; Turaev, S.; Malik, S.; El-Sayed, H.; Ullah, F. A Vehicle-Edge-Cloud Framework for Computational Analysis of a Fine-Tuned Deep Learning Model. Sensors 2024, 24, 2080. [Google Scholar] [CrossRef]
  92. Xiao, W.; Li, P.; Tang, M.; Song, C.; Yu, D.; Liu, H.; Chen, D.; Chen, N. Real-Time Roadside 3-D Spatial Perception with LiDAR AIoT: An Edge-Cloud-Terminal Collaborative Sensing Prototype. IEEE Internet Things J. 2025, 12, 12624–12639. [Google Scholar] [CrossRef]
  93. Huang, M.; Zhang, W.; Zhang, G. Joint Service Placement and Task Offloading in Vehicle-Edge-Cloud Collaborative Networks. IEEE Trans. Intell. Transp. Syst. 2025, 26, 20577–20591. [Google Scholar] [CrossRef]
  94. Khattak, M.I.; Yuan, H.; Khan, A.; Ahmad, A.; Ullah, I.; Ahmed, M. Evolving Multi-Access Edge Computing (MEC) for Diverse Ubiquitous Resources Utilization: A Survey. Telecommun. Syst. 2025, 88, 71. [Google Scholar] [CrossRef]
  95. Zhang, J.; Wei, Z.; Liu, B.; Wang, X.; Yu, Y.; Zhang, R. Cloud-Edge-Terminal Collaborative AIGC for Autonomous Driving. IEEE Wirel. Commun. 2024, 31, 40–47. [Google Scholar] [CrossRef]
  96. Solouki, M.A.; Sini, J.; Violante, M. An Experimental Evaluation of Control Flow Checking for Automotive Embedded Applications Compliant with ISO 26262. IEEE Access 2023, 11, 51185–51198. [Google Scholar] [CrossRef]
  97. Liu, X.; Yan, R.; Kim, J.Y.; Xu, X. MPSO: An Optimization Algorithm for Task Offloading in Cloud-Edge Aggregated Computing Scenarios for Autonomous Driving. Mob. Netw. Appl. 2024, 30, 702–716. [Google Scholar] [CrossRef]
  98. Benmerar, T.Z.; Theodoropoulos, T.; Fevereiro, D.; Rosa, L.; Rodrigues, J.; Taleb, T.; Barone, P.; Giuliani, G.; Tserpes, K.; Cordeiro, L. Towards establishing intelligent multi-domain edge orchestration for highly distributed immersive services: A virtual touring use case. Clust. Comput. 2024, 27, 4223–4253. [Google Scholar] [CrossRef]
  99. Rahhal, M.; Gifre, L.; Armingol Robles, P.; Mateos Najari, J.; Zabala, A.; Jimenez, M.A.; Leira Osuna, R.; Munoz, R.; Gonzalez de Dios, O.; Vilalta, R. Edge Microservice Deployment and Management Using SDN-Enabled Whitebox Switches. Electronics 2026, 15, 246. [Google Scholar] [CrossRef]
  100. Dong, X.; Tian, W.; Ye, X.; Xu, Y.; Wu, T.; Wang, Z. A Federated Cloud-Based Auction Mechanism for Real-Time Scheduling of Vehicle Sensors in Vehicle-Road-Cloud Collaborative System. IEICE Trans. Commun. 2025, 108, 14–23. [Google Scholar] [CrossRef]
  101. Alcaide, S.; Kosmidis, L.; Hernandez, C.; Abella, J. Software-only based Diverse Redundancy for ASIL-D Automotive Applications on Embedded HPC Platforms. In Proceedings of the 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (DFT), Frascati, Italy, 19–21 October 2020; pp. 1–4. [Google Scholar]
  102. Hwang, K.Y.; Oh, S.Y.; Park, E.K.; Song, B.-K.; Kim, S.-I. Thermal Fault-Tolerant Asymmetric Dual-Winding Motors in Integrated Electric Braking System for Autonomous Vehicles. Machines 2024, 12, 708. [Google Scholar] [CrossRef]
  103. Liu, H.; Liu, J.; Jiang, G.; Jin, X. MSSF: A 4D Radar and Camera Fusion Framework with Multi-Stage Sampling for 3D Object Detection in Autonomous Driving. IEEE Trans. Intell. Transp. Syst. 2025, 26, 8641–8656. [Google Scholar] [CrossRef]
  104. Wang, M.; Li, D.; Casas, J.R.; Ruiz-Hidalgo, J. Adaptive Fusion of LiDAR Features for 3D Object Detection in Autonomous Driving. Sensors 2025, 25, 3865. [Google Scholar] [CrossRef]
  105. Wei, C.; Qin, Z.; Zimmer, W.; Wu, G.; Barth, M.J. HeCoFuse: Cross-Modal Complementary V2X Cooperative Perception with Heterogeneous Sensors. arXiv 2025, arXiv:2507.13677. [Google Scholar] [CrossRef]
  106. Wei, Z.; Zheng, L.; Liu, J.; Huang, T.; Han, Q.-L.; Zhang, W.; Zhang, F. MS-Occ: Multi-Stage LiDAR-Camera Fusion for 3D Semantic Occupancy Prediction. IEEE Robot. Autom. Lett. 2026, 11, 370–377. [Google Scholar] [CrossRef]
  107. Zhang, J.; Lu, Y.; Xu, X. PentaFusion: Differentiable attention weighting for real-time LiDAR-camera fusion in edge autonomous vehicles. Robot. Auton. Syst. 2026, 196, 105233. [Google Scholar] [CrossRef]
  108. Shi, K.; He, S.; Shi, Z.; Chen, A.; Xiong, Z.; Chen, J.; Luo, J. Radar and Camera Fusion for Object Detection and Tracking: A Comprehensive Survey. IEEE Commun. Surv. Tutor. 2026, 28, 3478–3520. [Google Scholar] [CrossRef]
  109. Lu, Y.; Jiang, B.; Liu, N.; Li, Y.; Chen, J.; Zhang, Y.; Wan, Z. CrossPrune: Cooperative pruning for camera–LiDAR fused perception models of autonomous driving. Knowl.-Based Syst. 2024, 289, 111522. [Google Scholar] [CrossRef]
  110. Cserni, M.; Rovid, A.; Szalay, Z. MM-VSM: Multi-Modal Vehicle Semantic Mesh and Trajectory Reconstruction for Image-Based Cooperative Perception. Appl. Sci. 2025, 15, 6930. [Google Scholar] [CrossRef]
  111. De Araujo, P.R.M.; Mounier, E.; Bader, Q.; Dawson, E.; Abdelaziz, S.I.K.; Zekry, A.; Elhabiby, M.; Noureldin, A. The NavINST Dataset for Multi-Sensor Autonomous Navigation. IEEE Access 2025, 13, 84659–84674. [Google Scholar] [CrossRef]
  112. Gong, H.; Li, Z.; Lu, C.; Du, G.; Gong, J. Leveraging Multi-Stream Information Fusion for Trajectory Prediction in Low-Illumination Scenarios: A Multi-Channel Graph Convolutional Approach. IEEE Trans. Intell. Transp. Syst. 2024, 25, 3854–3869. [Google Scholar] [CrossRef]
  113. Wang, J.; Wu, Z.; Liang, Y.; Tang, J.; Chen, H. Perception Methods for Adverse Weather Based on Vehicle Infrastructure Cooperation System: A Review. Sensors 2024, 24, 374. [Google Scholar] [CrossRef]
  114. Shi, P.; Yang, L.; Dong, X.; Qi, H.; Yang, A. Research Progress on Multi-Modal Fusion Object Detection Algorithms for Autonomous Driving: A Review. Comput. Mater. Contin. 2025, 83, 3877–3917. [Google Scholar] [CrossRef]
  115. Fang, Z.; Hu, S.; An, H.; Zhang, Y.; Wang, J.; Cao, H.; Chen, X.; Fang, Y. PACP: Priority-Aware Collaborative Perception for Connected and Autonomous Vehicles. IEEE Trans. Mob. Comput. 2024, 23, 15003–15018. [Google Scholar] [CrossRef]
  116. Wei, C.; Wu, G.; Barth, M.J. Cooperative Perception for Automated Driving: A Survey of Algorithms, Applications, and Future Directions. Proc. IEEE 2025, 1–27. [Google Scholar] [CrossRef]
  117. Li, Z.; Hu, C.; Zheng, X.; Wang, J.; Wang, H. V2V-APG: Adversarial Progressive Generalization for Vehicle-to-Vehicle Cooperative Perception. IEEE Internet Things J. 2025, 12, 54873–54884. [Google Scholar] [CrossRef]
  118. Ma, C.; Li, H.; Long, K.; Zhou, H.; Liang, Z.; Li, P.; Yu, H.; Li, X. Real-time identification of cooperative perception necessity in road traffic scenarios. Transp. Res. Part C Emerg. Technol. 2026, 185, 105547. [Google Scholar] [CrossRef]
  119. Ma, Y.; Zheng, Y.; Liu, Y.; Feng, Z.; Easa, S.M.; Wong, Y.D. Synthesizing perceived trajectory for placing cooperative roadside LiDAR: A hybrid simulation-deep learning framework. Transp. Res. Part C Emerg. Technol. 2026, 183, 105473. [Google Scholar] [CrossRef]
  120. Li, L.; Zhang, W.; Wang, X.; Cui, T.; Sun, C. NLOS Dies Twice: Challenges and Solutions of V2X for Cooperative Perception. IEEE Open J. Intell. Transp. Syst. 2024, 5, 774–782. [Google Scholar] [CrossRef]
  121. Le, V.A.; Chalaki, B.; Tzortzoglou, F.N.; Malikopoulos, A.A. Stochastic Time-Optimal Trajectory Planning for Connected and Automated Vehicles in Mixed-Traffic Merging Scenarios. IEEE Trans. Control Syst. Technol. 2025, 33, 1403–1417. [Google Scholar] [CrossRef]
  122. Han, T.; Chen, S.; Li, C.; Wang, Z.; Su, J.; Huang, M.; Cai, G. Epurate-Net: Efficient Progressive Uncertainty Refinement Analysis for Traffic Environment Urban Road Detection. IEEE Trans. Intell. Transp. Syst. 2024, 25, 6617–6632. [Google Scholar] [CrossRef]
  123. Kanak, A.; Ergun, S.; Atalay, A.S.; Persi, S.; Karci, A.E.H. A Review and Strategic Approach for the Transition towards Third-Wave Trustworthy and Explainable AI in Connected, Cooperative and Automated Mobility (CCAM). In Proceedings of the 27th Asia-Pacific Conference on Communications (APCC)—Creating Innovative Communication Technologies for Post-Pandemic Era, Jeju, Republic of Korea, 19–21 October 2022; pp. 108–113. [Google Scholar]
  124. ISO 26262-1:2018; Road Vehicles—Functional Safety. International Organization for Standardization: Geneva, Switzerland, 2018.
  125. Kochanthara, S.; Singh, T.; Forrai, A.; Cleophas, L. Safety of Perception Systems for Automated Driving: A Case Study on Apollo. ACM Trans. Softw. Eng. Methodol. 2024, 33, 1–28. [Google Scholar] [CrossRef]
  126. ISO 21448:2022; Road Vehicles—Safety of the Intended Functionality. International Organization for Standardization: Geneva, Switzerland, 2022.
  127. Swain, R.; Kaye, S.A.; Rakotonirainy, A. Shared intention and shared awareness for conditional automated driving: An online, randomized video experiment. Traffic Inj. Prev. 2025, 26, 398–406. [Google Scholar] [CrossRef]
  128. Soleimani, M.; Sari, A.A. A unified safety framework for automated vehicle development: Integrating ISO 26262, SOTIF, and UL 4600. Transp. Res. Interdiscip. Perspect. 2026, 36, 101831. [Google Scholar] [CrossRef]
  129. Cui, Y.; Zhou, H.; Zhang, Y.; Xue, F.; Zheng, J.; Lu, Z.; Chen, Y.; Li, C.; Jiang, Z.; Ma, H. An Iterative Intelligent Attack with Integrated Strategy for Resource-Constrained Internet of Vehicles. IEEE Internet Things J. 2025, 12, 12453–12467. [Google Scholar] [CrossRef]
  130. Sakhai, M.; Sithu, K.; Oke, M.K.S.; Wielgosz, M. Cyberattack Resilience of Autonomous Vehicle Sensor Systems: Evaluating RGB vs. Dynamic Vision Sensors in CARLA. Appl. Sci. 2025, 15, 7493. [Google Scholar] [CrossRef]
  131. Yousseef, A.; Lin, Y.Z.; Satam, S.; Latibari, B.S.; Pacheco, J.; Salehi, S.; Hariri, S.; Satam, P. Autonomous Vehicle Security: Hybrid Threat Modeling Approach. IEEE Open J. Veh. Technol. 2025, 6, 1774–1795. [Google Scholar] [CrossRef]
  132. Shang, C.; Cao, J.; Li, Z.; Niu, B.; Lam, K.-Y.; Chi, C.-H.; Li, H. A Hierarchical Encrypted Compression Scheme for Intra-Vehicle Network. IEEE Trans. Intell. Transp. Syst. 2025, 26, 11916–11930. [Google Scholar] [CrossRef]
  133. Lucas-Estañ, M.C.; Coll-Perales, B.; Shimizu, T.; Gozalvez, J.; Higuchi, T.; Avedisov, S.; Altintas, O.; Sepulcre, M. An Analytical Latency Model and Evaluation of the Capacity of 5G NR to Support V2X Services Using V2N2V Communications. IEEE Trans. Veh. Technol. 2023, 72, 2293–2306. [Google Scholar] [CrossRef]
  134. L3Pilot Consortium. L3Pilot Final Project Results. Available online: https://l3pilot.eu (accessed on 1 April 2026).
  135. TM Forum. China Mobile AI4Vehicle Case Study—Wuxi City Vehicle-Road-Cloud Collaboration v1.0.0. Available online: https://www.tmforum.org/resources/technical-specification/ig1476-china-mobile-ai4vehicle-case-study-wuxi-city-vehicle-road-cloud-collaboration-v1-0-0/ (accessed on 1 April 2026).
  136. China United Network Communications Co., Ltd. (Shanghai Unicom). Shanghai Unicom Builds China’s Largest 5G-A ’Double 20’ Connected Vehicle Demonstration Zone. Available online: http://vod.sasac.gov.cn/folder21/folder24/2025-06-19/X1ZFjGUuTZZCYo0P.html (accessed on 1 April 2026).
  137. Waymo LLC. Comparison of Waymo Rider-Only Crash Rates by Crash Type to Human Benchmarks at 56.7 Million Miles. Available online: https://pubmed.ncbi.nlm.nih.gov/40378124/ (accessed on 1 April 2026).
  138. Chia, W.M.D.; Keoh, S.L.; Goh, C.; Johnson, C. Risk Assessment Methodologies for Autonomous Driving: A Survey. IEEE Trans. Intell. Transp. Syst. 2022, 23, 16923–16939. [Google Scholar] [CrossRef]
  139. Abboush, M.; Ghannoum, E.; Rausch, A. An explainable hybrid deep learning-enabled intelligent fault detection and diagnosis approach for automotive software systems validation. Knowl.-Based Syst. 2026, 334, 114922. [Google Scholar] [CrossRef]
  140. Imghoure, A.; Omary, F.; El-Yahyaoui, A. Hybrid cryptography-based scheme with conditional privacy-preserving authentication and memory-based DOS resilience in V2X. Veh. Commun. 2024, 49, 100810. [Google Scholar] [CrossRef]
  141. Hellemans, W.; Le Jeune, L.; Rabbani, M.M.; Preneel, B.; Mentens, N. Toward a Real-Time Intrusion Detection System for Modern In-Vehicle Networks. IEEE Trans. Intell. Transp. Syst. 2025, 26, 18665–18679. [Google Scholar] [CrossRef]
  142. Aljabri, W.; Hamid, M.A.; Mosli, R. Lightweight and Adaptive Data-Driven Intrusion Detection System for Autonomous Vehicles. IEEE Trans. Intell. Transp. Syst. 2025, 26, 2282–2292. [Google Scholar] [CrossRef]
  143. Jiang, L.; Li, Q.; Che, X.; Chen, X. A Knowledge Distillation Enhanced Semi-Supervised Federated Learning Framework for Intrusion Detection in EV Charging Networks. IEEE Internet Things J. 2025, 12, 34360–34373. [Google Scholar] [CrossRef]
  144. Jiang, C.; Yin, S.; Yao, Z.; He, J.; Jiang, R.; Jiang, Y. Safety evaluation of mixed traffic flow with truck platoons equipped with (cooperative) adaptive cruise control, stochastic human-driven cars and trucks on port freeways. Physica A 2024, 643, 129802. [Google Scholar] [CrossRef]
Futuretransp 06 00103 g001
Figure 1. The cooperative perception framework for CCAM vehicle control based on V2X communication.
Figure 1. The cooperative perception framework for CCAM vehicle control based on V2X communication.
Futuretransp 06 00103 g001
Futuretransp 06 00103 g002
Figure 2. Different fusion stages in cooperative perception systems.
Figure 2. Different fusion stages in cooperative perception systems.
Futuretransp 06 00103 g002
Table 1. Core Framework for CCAM Cooperative Decision-Making and Planning.
Table 1. Core Framework for CCAM Cooperative Decision-Making and Planning.
Decision HierarchyComputing Time *Core TasksAdaptive Technologies
Route Planning (Strategic) 10 s < T < 1 min Global path planning, traffic flow coordinationRoad network optimization, deep learning prediction
Maneuver Planning (Tactical) 3 s < T < 5 s Vehicle behavior decision-making (lane change, overtaking, yielding)Game theory, multi-agent reinforcement learning
Motion Planning (Operational) 100 ms < T < 500 ms Local trajectory generation and optimizationModel predictive control, random trees
* All time ranges are measured under urban road conditions with V2X communication latency ≤ 50 ms.
Table 2. Comparison of CCAM cooperative decision-making methods.
Table 2. Comparison of CCAM cooperative decision-making methods.
Method TypeRepresentative TechnologiesApplicable ScenariosCore Advantages
CooperativeV2V communication + spatio-temporal reservationLane-changing, overtaking, car-followingReduced multi-vehicle conflicts, improved local efficiency
Game TheoryNash equilibrium, Stackelberg gameUnsignalized intersection right-of-wayHuman-like interactive decisions
Optimization-basedMPC (Model Predictive Control)High-speed platooning, obstacle avoidance, mergingReal-time dynamic trajectory adjustment
Learning-basedMulti-agent reinforcement learning (MARL)Congestion evacuation, highway merging, large-scale coordinationAdaptive to complex dynamic environments
Table 3. Classification of CCAM Cooperative Control Studies.
Table 3. Classification of CCAM Cooperative Control Studies.
Control ArchitectureCore Control MethodologiesTypical Application ScenariosReferences
Distributed ControlRL (DRL/MARL), MPC, CACC, State EstimationPlatoon formation/stability, lane-changing, power control, topology-aware estimation[24,52,53,54,58,63,64,65]
Centralized ControlFunnel Cruise Control, Networked Predictive ControlPlatoon string stability, multi-intersection coordination[56,66]
Hybrid ControlMARL, Mobile-Edge Hybrid Control, MEC+RISCooperative merging in mixed traffic, emergency CAV passage, vehicular task offloading[55,61,62,67]
DRL: Deep Reinforcement Learning; CACC: Cooperative Adaptive Cruise Control; MEC: Mobile Edge Computing; RIS: Reconfigurable Intelligent Surfaces.
Table 4. Classification of CCAM-oriented V2X communication studies.
Table 4. Classification of CCAM-oriented V2X communication studies.
V2X CategoryCore TechnologiesReferences
Technology GenerationDSRC/C-V2X, IEEE 802.11p/bd, LTE/5G/6G V2X[77,78]
Core Enabling TechnologiesIRS/RIS, network slicing, MEC, SAGIN[10,16,79]
Resource AllocationMARL/DRL, optimization, GNN, federated learning[80,81,82,83]
Performance OptimizationLatency modeling, reliability, localization, QoS[11,15,80,84]
Security & PrivacyLightweight auth, intrusion detection, DDoS defense, privacy[85,86,87,88]
6G V2X ProspectiveTerahertz, satellite-cellular, SAGIN[22,75]
Table 5. Classification of OBU Related Studies.
Table 5. Classification of OBU Related Studies.
ArchitectureKey FeaturesRepresentative References
Heterogeneous (CPU+GPU+FPGA)Parallel acceleration, low-latency control [91,92,93]
CPU+GPUDeep learning inference, sensor fusion [62,89,94,95]
FPGA/DSPDedicated accelerator, microsecond-level response [96,97]
Edge-cloud collaborativeTask offloading, resource pooling [94,95,97,98,99,100]
Redundancy & ReliabilityFault tolerance, self-diagnostics [101,102]
Table 6. Classification of Perception Fusion Studies.
Table 6. Classification of Perception Fusion Studies.
Fusion MethodKey FeaturesRepresentative References
Early fusionRaw data-level fusion (LiDAR+camera, LiDAR+radar, 4D radar) [24,103,104,105,106,107,108]
Intermediate fusionFeature-level fusion, attention mechanisms [28,31,109,110,111,112,113,114]
Late fusionDetection-level fusion, object tracking [115,116,117,118,119]
Cooperative perceptionV2X+on-board/roadside sensors [22,92,115,116,120]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ji, A.; Ju, X.; Liu, N.; Chen, J.; Dai, Z. Cooperative Connected and Automated Mobility: A Survey. Future Transp. 2026, 6, 103. https://doi.org/10.3390/futuretransp6030103

AMA Style

Ji A, Ju X, Liu N, Chen J, Dai Z. Cooperative Connected and Automated Mobility: A Survey. Future Transportation. 2026; 6(3):103. https://doi.org/10.3390/futuretransp6030103

Chicago/Turabian Style

Ji, Ang, Xilu Ju, Nieyangzi Liu, Junxian Chen, and Zhe Dai. 2026. "Cooperative Connected and Automated Mobility: A Survey" Future Transportation 6, no. 3: 103. https://doi.org/10.3390/futuretransp6030103

APA Style

Ji, A., Ju, X., Liu, N., Chen, J., & Dai, Z. (2026). Cooperative Connected and Automated Mobility: A Survey. Future Transportation, 6(3), 103. https://doi.org/10.3390/futuretransp6030103

Article Metrics

Back to TopTop