Source-linked AI summary
Vehicular Communications: A Network Layer Perspective
Haixia Peng, Le Liang, Xuemin Shen, Geoffrey Ye Li
TL;DR
Efficient vehicular communications face network-layer issues and diverse requirements. This paper surveys network-layer research across manually driving and automated driving vehicular networks, classifies automated networks by traffic-management strategy, and identifies under-explored issues and research opportunities.
Problem
Vehicular communications must address many network-layer issues and different requirements to provide efficient communications.
Method
The paper provides a comprehensive overview of network-layer research on manually driving and automated driving vehicular networks, including a three-category classification of automated networks.
Results
The survey covers vehicular communications from a network-layer perspective across manually driving and automated driving networks.
Takeaways & Limitations
The paper identifies under-explored issues in manually driving and automated driving vehicular networks and presents corresponding research opportunities.
Abstract
from arXiv · showhide
Vehicular communications, referring to information exchange among vehicles, pedestrians, and infrastructures, have become very popular and been widely studied recently due to its great potential to support intelligent transportation and various safety applications. Via vehicular communications, manually driving vehicles and autonomous vehicles can collect useful information to improve traffic safety and support infotainment services. In this paper, we provide a comprehensive overview of recent research on enabling efficient vehicular communications from the network layer perspective. First, we introduce general applications and unique characteristics of vehicular networks and the corresponding classifications. Based on different driving patterns of vehicles, we divide vehicular networks into two categories, i.e., manually driving vehicular networks and automated driving vehicular networks, and then discuss the available communication techniques, network structures, routing protocols, and handoff strategies applied in these vehicular networks. Finally, we identify the challenges confronted by the current vehicular communications and present the corresponding research opportunities.
I. INTRODUCTION
Vehicular communications support safety, intelligent transportation, and infotainment, but high mobility and changing environments create stringent network-layer requirements. The paper surveys these issues across manually driving and automated driving vehicular networks, including communication technologies, routing, handoff, structures, challenges, and opportunities.
- Vehicular communications can improve traffic safety and support infotainment services for manually driving and autonomous vehicles.
- Manually driving vehicles are distinguished from autonomous vehicles, which this paper treats as level 4 autonomy requiring no human intervention beyond setting the destination and starting the system.
- Autonomous vehicles require inter-vehicle information from sensors or communications for cooperative driving, collision avoidance, platooning, and awareness of hazards and surroundings.
- Efficient vehicular communications require network-layer decisions about communication entities, message-dissemination routing, technology selection, and handoff strategies.
- The survey covers MDVNET and ADVNET applications, characteristics, communication technologies, routing protocols, handoff strategies, communication structures, challenges, and potential solutions.
II. GENERAL VEHICULAR NETWORKS
General vehicular networks support safety, traffic management, and infotainment applications through information exchange among vehicles. Their applications must balance safety-critical delay and reliability requirements against non-safety service quality under difficult network conditions.
- Safety applications: Safety applications share warnings and vehicle-state information so drivers can detect dangers, act in advance, and reduce collision risks.
- Non-safety applications: Traffic-management applications use shared traffic and road information for rerouting, traffic-light scheduling, and reducing traffic jams.
- Non-safety applications: Electronic toll collection uses vehicle-identification information to eliminate toll-collection delays and save commuters’ travel time.
- Non-safety applications: Infotainment applications provide location-based services and Internet access for moving vehicles.
- Application requirements: Safety information receives higher priority because it requires low delay and high reliability, while non-safety traffic and infotainment services generally have less stringent real-time requirements.
- Application requirements: High mobility, variable network density, unstable topology, and complicated communication environments make these application requirements difficult to satisfy.
B. Vehicular Networks Characteristics
Vehicular networks inherit mobile ad hoc networking properties while adding distinctive effects from moving nodes. High mobility and stringent delay constraints hinder communication, whereas predictable routes and weak energy constraints can help it.
- Vehicular networks use spontaneous wireless data exchange and share MANET characteristics such as self-organization, short-to-medium range, broadcast, and low bandwidth.
- The paper classifies vehicular-network characteristics as detrimental or beneficial according to their effects on information exchange.
- Detrimental characteristics: High vehicle mobility frequently disconnects wireless links, reduces effective communication time, and dynamically changes network topology.
- Detrimental characteristics: Some applications require information exchange within a maximum source-to-destination delay to avoid accidents and preserve infotainment quality.
- Communication environments: Highway environments are relatively simple because vehicles follow straightforward directions and relatively fixed speeds, whereas urban roads and viaducts create more complex communication settings.
- Communication environments: Obstacles at intersections can prevent direct links between vehicles on different road segments, while multilayer viaduct traffic creates three-dimensional communication complexity.
- Beneficial characteristics: Vehicular networks benefit from weak energy constraints and vehicles’ substantial sensing, computing, storage, and processing capabilities.
2) Driving route prediction:
Vehicular networks can predict driving routes because vehicles are constrained to roads and route information can be inferred from maps and vehicle speeds. This prediction supports routing design under high mobility, whose effects differ across manually and automated driving networks.
- 2) Driving route prediction:: Vehicles are generally constrained to roads, making their driving routes potentially predictable from road maps and vehicle speed information.
- 2) Driving route prediction:: Driving route prediction plays an important role in vehicular-network routing protocol design, especially for high-mobility challenges.
- 2) Driving route prediction:: Manually driving vehicles move individually with high and heterogeneous mobility, creating significant impacts on manually driving vehicular networks.
- 2) Driving route prediction:: The paper reviews communication technologies, routing protocols, and handoff strategies for manually driving vehicular networks.
- 2) Driving route prediction:: MDVNETs use technologies including DSRC, cellular, Wi-Fi, UWB, and Bluetooth, with heterogeneous networks combining different wireless technologies.
- 2) Driving route prediction:: DSRC supports high-mobility links, safety-prioritized channels, and authentication, but faces broadcast storms, short-lived connectivity, and limited radio range.
- 2) Driving route prediction:: Cellular technologies provide relatively long-lived V2I connectivity and high capacity, while data costs and dense traffic can challenge their use.
B. Routing Protocols
Routing protocols in manually driving vehicular networks are classified into unicast, multicast, broadcast, and cluster-based categories. Their designs address mobility, disconnections, delivery scope, redundant transmissions, and security.
- B. Routing Protocols: The paper classifies MDVNET routing protocols into unicast, multicast, broadcast, and cluster-based categories.
- B. Routing Protocols: Unicast routing sends packets from one source vehicle to one destination using single- or multihop communication.
- B. Routing Protocols: Greedy unicast selects geographically farther neighbors, whereas opportunistic unicast supports frequent disconnections through store-and-forward and potentially parallel packet copies.
- B. Routing Protocols: Multicast routing supports one-to-many or many-to-many communication for geographically targeted applications such as roadblocks, congestion, and hazard warnings.
- B. Routing Protocols: Multicast protocols also face disconnection, mobility uncertainty, redundant packets, and authorization requirements for confidential group communication.
- B. Routing Protocols: Broadcast routing disseminates messages from one source to all receivers for neighbor discovery, traffic information, and safety or cooperative-driving applications.
- B. Routing Protocols: Broadcast protocols must address broadcast storms and disconnected networks, with suppression and network coding used to reduce contention and redundant broadcasts.
C. Handoff Strategies
Handoff strategies maintain vehicular communication as vehicles move among neighboring access points, cells, or heterogeneous networks. The paper distinguishes horizontal and vertical handoff and reviews triggers based on QoS, cost, throughput, fairness, load, and security.
- C. Handoff Strategies: Handoff is a major issue because vehicles move in and out of communication ranges, causing frequently disconnected links and changing network conditions.
- C. Handoff Strategies: Horizontal handoff transfers sessions between points using the same wireless technology, while vertical handoff uses different technologies.
- C. Handoff Strategies: Horizontal handoff supports V2V rerouting when neighbors change, D2D V2V continuity across cell boundaries, and V2I session transfer between infrastructures.
- C. Handoff Strategies: Vertical handoff maintains data transfer across network types and can be used in overlapping coverage areas of heterogeneous MDVNETs.
- C. Handoff Strategies: User-centric handoff triggers include QoS and financial cost, while network-centric triggers include throughput maximization, fairness, and load balancing.
- C. Handoff Strategies: Enhanced handoff schemes have been proposed to reduce tunneling burden, handover latency, packet loss, signaling overhead, and data-traffic cost.
- C. Handoff Strategies: Cloud computing and software-defined networking support centralized network selection and smoother handover decisions, while security mechanisms address changing connection keys.
IV. AUTOMATED DRIVING VEHICULAR NETWORKS
Automated driving vehicular networks (ADVNETs) are classified by traffic-management strategy into free, convoy-based, and platoon-based structures. These networks face stringent safety, information, and resource-management requirements, while platoon-based management can improve efficiency and safety.
- Network classifications: ADVNETs are classified into free, convoy-based, and platoon-based networks according to their traffic-management strategies.Free ADVNETs move independently; convoy-based networks coordinate vehicles across multiple lanes; platoon-based networks use centralized leader control.
- Network classifications: Free ADVNETs comprise independently moving autonomous vehicles controlled by computerized autopilots.
- Network classifications: Convoy-based ADVNETs group same-direction vehicles across multiple lanes and distribute lateral and longitudinal control to maintain spacing and aligned velocity.Convoys extend the platoon concept to multiple lanes.
- Network classifications: Platoon-based management can reduce energy consumption and exhaust emissions by minimizing aerodynamic drag while increasing driving safety through cooperation.
- Network classifications: Platoon-based structures can also occur in manually driving networks, but their spontaneous formation is sufficiently unlikely that related work is treated as cluster-based.
- Requirements and challenges: ADVNETs require stringent delay and reliable-delivery guarantees, more safety-related information, and more complex processing of sensory and communication data.These requirements are especially significant in intersection environments.
A. Communication Structures
ADVNET communication structures organize information exchange among vehicles, platoons, and infrastructure. They range from broadcast and infrastructure-assisted links to specialized intra-platoon and inter-platoon designs that support safety, control, and string stability.
- Free and infrastructure-based structures: Broadcast-based V2V lets autonomous vehicles share information with neighboring vehicles, including collision warnings for intersection and road-safety applications.Communication ranges may be adjusted for different scenarios.
- Free and infrastructure-based structures: V2I structures use roadside infrastructure to collect, broadcast, or relay information for vehicles and traffic management.Convoy systems may combine broadcast V2V for cooperative maneuvering with V2I for traffic-management efficiency.
- Platoon-based structures: Platoon-based ADVNETs require string stability, preventing inter-platoon and intra-platoon spacing errors from amplifying upstream between vehicles or platoons.
- Platoon-based structures: Leader and preceding-vehicle velocity and acceleration information must be shared to maintain platoon string stability.Inter-platoon communication can share information such as collision alerts between different platoons.
- Platoon-based structures: Intra-platoon communication includes leader-to-member or member-to-leader links, adjacent-vehicle exchange, and platoon multicast.These structures share control information and improve communication efficiency.
B. Technologies for ADVNETs
ADVNETs use established radio technologies alongside infrared and visible-light communications. Light-based methods offer directionality, high speed, and spectrum or deployment advantages, but their line-of-sight and environmental constraints often require complementary technologies.
- Technology combinations: DSRC and cellular technologies commonly support long-range ADVNET communication and communication in free ADVNETs, while infrared and VLC are also applicable to manually driving networks.
- Infrared communication: Infrared communication provides high directionality, confidentiality, and harmlessness, with potential speeds of 1 Gbps for large-volume data.It has also been used to measure intra-platoon spacing for gap control.
- Infrared communication: Infrared links are limited to roughly 10 meters, cannot penetrate obstacles, and are affected by weather conditions.They are therefore suited mainly to special scenarios such as two-adjacent-vehicle or intra-platoon communication and are combined with technologies such as DSRC.
- Visible light communication: VLC offers 360 Terahertz of license-free visible-spectrum bandwidth, avoiding RF interference and addressing radio-spectrum scarcity.VLC has been used as a backup for DSRC, cellular, Wi-Fi, and White-Fi to reduce congestion-related packet loss.
- Visible light communication: VLC can achieve multiple-Gbps speeds in research and around 100 Mbps under IEEE 802.15.7, while reusing lighting infrastructure for inexpensive deployment.
- Visible light communication: VLC signals cannot penetrate most obstacles and therefore support communication only between vehicles with line of sight.VLC feasibility has been studied for sharing information among platoon members.
V. CHALLENGES AND OPPORTUNITIES
Despite extensive research and available industrial products, efficient vehicular communications for practical intelligent transportation applications still face major challenges. The paper therefore discusses these challenges and corresponding research opportunities.
- Challenges and opportunities: Efficient vehicular communications for practical intelligent transportation applications remain challenging despite substantial research and existing industrial products.
- Challenges and opportunities: The paper presents major current challenges and identifies corresponding research opportunities.
A. Heterogeneous Driving Vehicular Networks
Heterogeneous driving vehicular networks combine autonomous and manually driven vehicles, creating communication and coordination challenges. Cooperative driving requires shared information, while mixed traffic introduces diverse message and QoS requirements.
- Heterogeneous driving vehicular networks: HDVNETs enable information sharing between autonomous and manually driving vehicles moving on roads simultaneously.They are needed to achieve information sharing across the two vehicle types.
- Cooperative driving: Cooperative driving in ADVNETs can improve AV safety by sharing real-time positions, velocity, and desired driving lanes within a platoon.These shared data support safety-driving patterns for intersection crossing.
- Communication requirements: Mixed traffic creates complex message structures because AVs and manually driven vehicles have different message types and QoS requirements.Examples include safety messages, periodic cooperative-driving beacons, and infotainment video.
- Communication requirements: Existing beacon scheduling considered two message types, whereas HDVNET scenarios may require simultaneous dissemination of safety, periodic, and video information.These requirements support platoon string stability while satisfying infotainment QoS.
B. Security Issue
Vehicular communications face security and network-management challenges across manually driven, autonomous, and heterogeneous networks. Low-latency cooperative driving, emerging technologies, and diverse applications create requirements for specialized solutions.
- Security issue: Cooperative driving depends on shared information, so message falsification, eavesdropping, jamming, and tampering can significantly affect driving safety.Security protection must cover both communication channels and sensors.
- Security issue: Current wireless communication security algorithms may not work well for ADVNETs and HDVNETs because cooperative driving requires lower latency.Reducing end-to-end delay is therefore an explicit security-design concern.
- Security issue: Group authentication can protect platoon or convoy communications, while authentication during formation and vehicle joining can reduce impacts on information sharing.The approach is discussed for clustered MDVNETs and platoon- or convoy-based ADVNETs.
- Other challenges: mmWave and VLC support remain challenging under high mobility and harsh weather, and heterogeneous networks need criteria for selecting among communication technologies.Communication information management, cooperative control, and resource allocation also remain complicated.
- Conclusion: The survey covers network-layer communications, communication technologies, structures, routing, handoff strategies, and under-explored issues in MDVNETs, ADVNETs, and HDVNETs.It also presents potential solutions for these issues.