Source-linked AI summary
5G Software Defined Vehicular Networks
Xiaohu Ge, Zipeng Li, Shikuan Li
TL;DR
Future intelligent transportation systems require vehicular networks that support stringent delay demands associated with applications such as pilotless vehicles. The paper proposes a 5G software-defined architecture combining cloud and fog computing, and simulations report minimum transmission delay under varying vehicle densities and higher fog-cell throughput than traditional transportation management systems.
Problem
Pilotless vehicles and intelligent transportation systems require vehicular networks with stringent performance, including transmission delay below 1 millisecond.
Method
The paper proposes a 5G software-defined vehicular-network architecture integrating SDN, cloud computing, and fog cells with separated control and data planes.
Results
The simulations identify a minimum transmission delay for different vehicle densities and show that fog-cell throughput is better than traditional transportation management systems.
Takeaways & Limitations
Fog cells provide the proposed network with a way to avoid frequent RSU–vehicle handovers while supporting the analyzed delay and throughput objectives.
Abstract
from arXiv · showhide
With the emerging of the fifth generation (5G) mobile communication systems and software defined networks, not only the performance of vehicular networks could be improved but also new applications of vehicular networks are required by future vehicles, e.g., pilotless vehicles. To meet requirements from intelligent transportation systems, a new vehicular network architecture integrated with 5G mobile communication technologies and software defined network is proposed in this paper. Moreover, fog cells have been proposed to flexibly cover vehicles and avoid frequently handover between vehicles and road side units (RSUs). Based on the proposed 5G software defined vehicular networks, the transmission delay and throughput are analyzed and compared. Simulation results indicate that there exist a minimum transmission delay of 5G software defined vehicular networks considering different vehicle densities. Moreover, the throughput of fog cells in 5G software defined vehicular networks is better than the throughput of traditional transportation management systems.
I. INTRODUCTION
Emerging 5G capabilities and pilotless vehicles create stringent requirements for vehicular networks, including transmission delay below 1 millisecond. The paper motivates integrating 5G, cloud computing, and SDN into a new vehicular-network architecture.
- 5G wireless technologies improve spectrum efficiency and energy efficiency through millimeter-wave and massive MIMO technologies.
- Pilotless vehicles impose stringent ITS requirements, including transmission delay below 1 millisecond.
- The paper argues that integrating 5G, cloud computing, and SDN is necessary to design a new architecture for future vehicular networks.
- IEEE 802.11p networks face poor scalability, low capacity, and intermittent connectivity.
- LTE technologies were proposed to support vehicular applications, while open issues in LTE vehicular networks remained under discussion.
associated requirements was presented and challenges were discussed. Besides, the past major
Prior work explored LTE, SDN, relay, and adaptive edge-computing approaches for vehicular networks, but handover and performance challenges remained. This paper proposes a 5G software-defined architecture with separated planes and fog cells, then analyzes delay and throughput.
- Related work: Existing vehicular-network approaches examined connectivity, relay, interference, and performance under varying base-station and vehicle densities.
- Related work: SDN-based studies addressed QoS adaptation and cooperative data dissemination through edge computing, admission control, queueing, and RSU coordination.
- Challenges: Frequent handovers can reduce RSU-based SDN performance when many vehicles connect to an RSU.
- Proposed architecture: The proposed architecture integrates SDN, cloud computing, and fog computing, separating control and data functions to improve flexibility and scalability.
- Fog cells: Fog cells avoid frequent RSU–vehicle handovers and use adaptive bandwidth allocation and multi-hop relay communication.
- Evaluation: The study analyzes transmission delay and throughput, reporting a minimum delay under different vehicle densities and higher fog-cell throughput than traditional transportation management systems.
A. Topology Structure of 5G Software Defined Vehicular Networks
The proposed network combines cloud and fog computing with SDN controllers, RSUs, base stations, vehicles, and users. Distributed fog clusters process edge data, while centralized cloud components coordinate network state and control information.
- Control structure: The architecture uses SDN to support flexible data forwarding and resource allocation, with its detailed logical structure illustrated in Fig. 1(b).
- Topology structure: The topology includes cloud-computing centers, SDN controllers, RSU centers, RSUs, base stations, fog clusters, vehicles, and users.
- Topology structure: The architecture supports infrastructure-to-infrastructure and vehicle-to-infrastructure links alongside vehicle communications.
- Fog computing clusters: Fog-computing clusters are distributed at the network edge to support prompt responses from vehicles and users.
- Fog computing clusters: Fog clusters include RSU centers, RSUs, base stations, vehicles, and users, and they save and process much of the edge data.
- Control structure: SDN controllers collect cluster-state information and forward it to cloud centers, which return control information to the clusters.
B. Logical Structure of 5G Software Defined Vehicular Networks
The logical structure separates 5G software defined vehicular networks into data, control, and application planes, with fog-cell coordination supporting vehicle connectivity and resource management.
- Logical planes: The data plane collects, quantizes, and forwards vehicle and infrastructure data into the control plane.It includes vehicles, base stations, and roadside units.
- Vehicle information: Vehicle information includes sensor-derived state and environment data, plus independent GPS and dependent inter-vehicle position information.Dependent position information can provide higher location precision than independent position information.
- Control plane: The control plane uses RSUCs and an SDNC to build global information maps, monitor network status, derive control results, and allocate resources among fog cells.RSUCs control individual fog cells, while the SDNC serves as the total control center.
- Fog-cell coordination: A fog cell groups vehicles with an RSU and uses millimeter-wave multi-hop relays so one vehicle can connect the group to the RSU.The arrangement is intended to avoid frequent handover between RSUs and vehicles.
- Application plane: The application plane converts user and vehicle requirements into rules and strategies for services including security, efficiency, and entertainment.These rules and strategies are forwarded to the control plane.
III. TRANSMISSION DELAY AND THROUGHPUT OF 5G SOFTWARE DEFINED VEHICULAR NETWORKS
The transmission and throughput analysis examines fog cells in the proposed network, where gateway vehicles relay traffic between other vehicles and RSUs while vehicles move along the road.
- Performance analysis: Transmission delay and throughput are investigated for vehicles operating in a fog cell of the proposed network.The fog cell is treated as the basic composition of the 5G software defined vehicular network.
- Fog-cell operation: A typical fog cell consists of an RSU and multiple vehicles, with a selected gateway vehicle connecting directly to the RSU.Other vehicles connect to the gateway through multi-hop relay routes.
- Fog-cell operation: Vehicles outside direct RSU coverage use multi-hop relay routes through the gateway vehicle to maintain connectivity within the fog cell.The gateway communicates directly with the RSU when it is within the RSU coverage region.
- Mobility handling: When the gateway vehicle departs, another vehicle is handed off to service as the gateway vehicle.This maintains wireless communication between vehicles and the RSU while the group moves along the road.
A. Transmission Delay of 5G Software Defined Vehicular Networks
The paper models transmission delay for multi-hop vehicle relaying within a fog cell and evaluates how vehicle density and transmission distance affect it. Delay first decreases and then increases with vehicle density, producing a minimum that depends on transmission distance.
- A vehicle data packet reaches the RSU through a multi-hop vehicle relay method within a fog cell.The distance between the RSU and the selected vehicle is denoted La.
- The fog-cell transmission delay is modeled as T = kThop +(k −1) Tretran, combining per-hop transmission and relay retransmission delays.Thop is the average one-hop transmission delay, while Tretran is the relay processing time at relay vehicles.
- Under the evaluated millimeter-wave model, transmission uses 60 GHz links, LOS interference is ignored, and the communication distance is limited to 50 meters.The default configuration uses WmmW ave = 2 GHz, Tretran = 5 microsecond, and tslot = 5 microsecond.
- At fixed vehicle density, transmission delay increases as transmission distance La increases.This relationship is reported for the vehicle-density analysis across different transmission distances.
- The minimum transmission delays are 0.32, 0.46, and 0.63 for transmission distances of 300, 400, and 500 meters, respectively.The numerical results identify a minimum delay for each evaluated transmission distance.
B. Throughput of 5G Software Defined Vehicular Networks
The section models fog-cell throughput under traditional and adaptive bandwidth allocation, showing that adaptive allocation improves utilization and throughput as vehicle density changes.
- Adaptive bandwidth allocation is proposed to optimize fog-cell throughput by reusing unoccupied bandwidth for other vehicles.The scheme reallocates unused bandwidth from vehicles with lower requirements to vehicles requiring more bandwidth.
- The model assumes N vehicles in a fog cell, available bandwidth B, maximum throughput C, and uniformly distributed vehicle bandwidth requirements.Each requirement follows Bi ∼ U(0, 2Bave), with 1 ≤ i ≤ N.
- The adaptive scheme always produces larger fog-cell throughput than average bandwidth allocation under the configured comparison.The comparison uses C = 1000 Mbps and Cave = 33 Mbps.
- Throughput initially increases as the number of vehicles in a fog cell increases.This trend is reported for either bandwidth allocation scheme once the allocation scheme is fixed.
- When the number of vehicles exceeds 30, fog-cell throughput becomes stationary because all available bandwidth has already been allocated.Additional vehicles cannot receive bandwidth after the available capacity is fully assigned.
IV. CHALLENGES OF 5G VEHICULAR NETWORKS
Future 5G vehicular networks face low-delay, handover, service-efficiency, and architecture challenges associated with pilotless vehicles and intelligent transportation systems.
- The low delay issues: Safety and road-information messages require very low transmission delay, while multi-hop relay networks can increase delay beyond a given threshold.The paper notes that extreme delay cases could contribute to fatal accidents and that route optimization remains important.
- The frequently handover issues: Handover among vehicles remains an issue for multi-hop relay links, especially when vehicles move between adjacent fog cells.Simultaneous handovers of fog cells and multi-hop links increase handover complexity and require minimizing warning-message propagation delay.
- The high service efficiency challenges: Pilotless vehicles are expected to generate massive wireless traffic because they must support different multimedia services.Improving service efficiency is identified as a major challenge for 5G vehicular networks.
- The architecture of 5G vehicular networks: The proposed architecture uses distributed networking for fog cells and centralized networking for the core network, connected flexibly through SDN.Scalability and compatibility remain challenges when these two network architectures coexist.
V. CONCLUSIONS
The paper motivates new 5G vehicular-network designs by linking future pilotless vehicles and intelligent transportation systems to demanding performance requirements.
- Pilotless vehicles create rigorous performance requirements for future intelligent transportation systems.
- 5G mobile communications, cloud computing, and SDN technologies are presented as potential solutions for future vehicular networks.
- The paper proposes a new architecture for 5G software-defined vehicular networks integrating 5G mobile communication technologies.
communications, cloud computing and SDN technologies. Moreover, fog cells are established
The proposed architecture uses fog cells and multi-hop relay networking to address handover, delay, and throughput requirements in 5G software-defined vehicular networks.
- Multi-hop relay networks are used at the edge to reduce frequent handovers between RSUs and vehicles.
- The simulations identify a minimum transmission delay under different vehicle densities.
- Fog-cell throughput is better than the throughput of traditional transportation management systems.
- After unresolved 5G vehicular-network challenges are addressed, the architecture could provide flexibility and compatibility for pilotless vehicles and ITSs.