Enhanced Traffic Shaping for Low‑Latency Communication in

Autonomous Vehicle Networks

Mohammed Shameem S., Naveen P. R., Prajith R., Rahul V.,

Dhanasekar J. and Ishwarya Niranjana M.

Department of ECE, Sri Eshwar College of Engineering, Coimbatore, Tamil Nadu, India

Keywords: Autonomous Vehicles (AVs), Traffic Shaping, Low‑Latency Communication, Real‑Time Operations,

Safety‑Critical Systems, High‑Priority Packet Management, MATLAB Simulations, Network Performance,

Vehicle Density, Communication Delays.

Abstract: Smart transportation powered by AVs necessitates reliable, low-latency communication to support real-time

and safety-critical tasks. 39Uncoordinated data dynamics, which can occur as applications transmit through

different paths (including in-Memory parameters), can lead to interference where one transmission degrades

the quality achieved by the other transmission. This work presents a better traffic shaping process meant for

the unique needs of cooperative autonomous vehicle networks. For instance, the implemented solution ensures

the timely delivery of collision avoidance messages by competing packets whereas the prioritization of low

priority packets, such as routine navigation data, is optimized. Traffic shaping models of the proposed model

is evaluated through MATLAB based simulations considering the variation of traffic model parameters like

the vehicle density, packet arrival rate, and communication delay. We find that using this mechanism yields

a significant reduction in latency for high priority traffic with limited impact on the handling of low priority

packets. These results highlight the promise of this approach for improving communication reliability and

enabling safer, more efficient autonomous transport networks.

1 INTRODUCTION

Autonomous vehicles (AVs) are being integrated into

transportation systems, which offer considerable

opportunities for improving safety, traffic flow, and

mobility, but also creates significant communication

challenges. AVs depend on so-called high-time-

critical data, which should be exchanged with

minimum delay and guaranteed reliability, for tasks

such as collision avoidance, emergency braking, and

navigation (G. Karagiannis et al., 2011).

Conventional traffic shaping in vehicular networks

focuses on prioritizing messages that typically lead to

delays and congestion; however, in AV networks,

such strategies will not sufficiently achieve the ultra-

low latency (required for safety-critical messages)

and efficient management of lower priority messages

(e.g., navigation update). It is critical to prioritize this

data, as lack of prioritization may result in network

congestion and/or communication delays that lead to

failures of these safety systems.

The work in this paper presents an advanced

traffic shaping methodology that has been

specifically designed for AV networks, utilising

adaptive queuing strategies that give due priority to

real-time, safety-related data, whilst still supporting

the effective transmission of lower priority data. It is

then evaluated through Matlab simulations regarding

the packet latency and queue management,

considering different vehicle densities and

communication delays. Prototyping and testing the

dynamic priority model yields results that showcase

lower latency for higher priority traffic through

higher priority packets gaining precedence while

maintaining reasonable delays for lower priority

packets.

2 LITERATURE SURVEY

The deployment of autonomous vehicles (AVs) into

contemporary transportation systems has generated

substantial research efforts on communication

S., M. S., P. R., N., Prajith, R., Rahul, V., Dhanasekar, J. and M., I. N.

Enhanced Trafﬁc Shaping for Lowâ

SLatency Communication in Autonomous Vehicle Networks.

DOI: 10.5220/0013880900004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 2, pages

249-255

ISBN: 978-989-758-777-1

249

systems that enable real-time, safety-critical

functionality. Multiple studies have emphasized the

need for low-latency communication for AVs,

particularly for functions such as collision avoidance

and emergency braking. In conventional vehicular

networks, traffic shaping techniques are utilized to

control the flow of data by giving priority to the high-

priority packet in order to alleviate delays. But, these

methods are not well-suited to cope with the

challenges specific to AV communication, such as

the necessity for ultra-low latency for safety critical

messages which runs in parallel with less time critical

traffic for navigation updates.

To reduce latency for high priority at the expense

of fairness for low priority such as standard messages,

adaptive queuing and packet scheduling techniques

have been an important target of effort in recent work.

In addition, researchers have investigated the effects

of network congestion, vehicle density, and

communication delays on system performance,

underscoring the importance of scalable approaches

that prioritize both reliable communication and

efficient operation in dynamic, high-density

scenarios.

3 PROBLEM STATEMENT

With the increased penetration of autonomous

vehicles (AVs) within the modern transportation

system, the importance of providing communication

to guarantee safety and efficiency with low latencies

is becoming pressing. AVs depend on the real-time

exchange of data to perform actions like collision

avoidance, emergency braking and precise

navigation. But with increasing vehicle density and

data traffic complexity, keeping vehicles in touch is

a challenge.

The increasing amount of data (both critical safety

messages and non-critical messages) floods the

network and delays the transmission of the required

messages. Conventional traffic management and

shaping methods for traditional vehicular networks

are inadequate for meeting the specific demands of

AV networks. These networks require ultra-low

latency for high-priority packets such as emergency

alerts, alongside handling non-critical data such as

navigation and infotainment updates. In the absence

of an effective system for dynamically prioritizing

and processing this traffic, the system faces unsafe

latency, degraded network performance or worse a

safety failure.

In addition, there is an urgent need for an

advanced traffic shaping mechanism of specific

demands to AV networks. This mechanism will need

to ensure that safety critical messages are received in

preference to less time-critical traffic while

improving latency and maximizing bandwidth

efficiency. Solving this problem is crucial for

ensuring the safe, scalable and effective deployment

of autonomous vehicle solutions in ever more

dynamic and congested transportation environments.

4 PROPOSED SOLUTION

This phenomenon is a consequence of the fast-

growing role that autonomous vehicles (AVs) play in

modern transportation systems which brings new

communication challenges that are critical to

ensuring both safety and operational efficiency of

these systems. Autonomous vehicles (AVs) depend

on near real-time data sharing with roadside

infrastructure and surrounding vehicles to execute

critical safety maneuvers (e.g., collision avoidance,

emergency braking) and improved navigation (e.g.,

route optimization). For these systems to operate as

intended, high-priority safety messages need to be

sent with ultra-low latency. While high-priority

safety messages need priority over lower-priority

traffic such as navigation and infotainment data,

current traffic management systems have limitations

in ensuring this. As a result of this asymmetric

communication, network congestion, communication

delays, and safety-critical applications are likely to

fail, jeopardizing the overall reliability of

autonomous systems.

To overcome such challenges, we present an

improved traffic shaping framework that tailored for

autonomous vehicle networks. The method would

prioritize the transmission of high-priority packets

like collision avoidance alerts and emergency braking

signals, while allowing it to just make sure low-

priority traffic like navigation updates and

infotainment data are still sent efficiently. This

technique applies adaptive queuing techniques,

which allow timeliness of safety-critical messages to

be guaranteed, even under high network load

conditions. For safety-critical messages, the system is

designed to alleviate congestion and minimize

communication latency by efficiently handling both

high- and low-priority traffic.

Proposed system experiment: MATLAB

simulations are tested due to the proposed system

performance under different simulation conditions

(different vehicle density/packet arrival

rate/communication delay). Results show that the

proposed traffic shaping mechanism achieves latency

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

250

compliance for high priority messages while low-

priority traffic transmission is also guaranteed at low

delays. This scalable and efficient solution meets the

communication requirements of autonomous vehicle

networks, enabling reliable data exchange for both

critical safety applications and non-critical services.

The model serves as a solid foundation for upholding

both the safety and operational efficacy of AVs

functioning in dynamic and congested network

surrounding.

Traffic Shaping Mechanism Overview: Traffic

shaping is a mechanism to adjust the transmission of

packets to a narrower bandwidth in order to constrain

it below the full rate to avoid congestion in networks

in order to implement Data flow management.

Conventional networking systems utilize queuing

techniques like First-Come-First-Served (FCFS),

Priority Queuing (PQ), as well as Weighted Fair

Queuing (WFQ) when it comes to packet

prioritization. Whilst these approaches perform well

in many contexts, they often struggle to satisfy the

strict latency requirements of autonomous vehicle

networks, particularly at high traffic loads, or, in

scenarios when multiple vehicles are communicating

simultaneously.

To overcome these limitations, we present an

adaptive queuing model that enables packets to be

classified as either high-priority or low-priority and

subsequently scheduled for transmission. This

approach aims to deliver the high-priority packets as

quickly as possible while avoiding the situation where

low-priority packets are delayed too long in high-

traffic situations. This is done by adapting to the

variations in the network conditions (packet arrival

rate, vehicle density, communication latency, etc.)

dynamically to sustain its efficiency in results. This

adaptive scheme aims to provide a flexible approach

to support safety-critical as well as non-critical

applications, enabling robust communication in

realistic autonomous vehicle networks.

Priority Queuing and Adaptive Scheduling: From

here, we can naturally build the proposed mechanism

of traffic shaping based on Priority Queuing (PQ),

which involves ordering packets into queues

depending on their priority. This helps to make sure

that critical data packets are processed and

transmitted before less crucial data packets are.

Leverage two high-level queues.

High-Priority Queue: This queue handles safety-

critical and time-sensitive messages, including

commands to initiate emergency braking, collision

notifications, and other real-time safety messages that

have to be processed quickly. But they need the

lowest possible latency to deliver information for

timely response where any delay can affect safety.

Low-Priority Queue: This queue processes low-

priority messages like navigation updates, weather

data or infotainment content. Although these packets

have to be sent, they can survive higher latencies

compared to high-priority traffic. That means only

the most important packets are processed first.

It only starts processing packets from the low-

priority queue when the high-priority queue is

empty. This also guarantees that safety-critical data

always receives the required bandwidth, no matter

how network traffic develops. Moreover, as the main

design goal of the protocol is to avoid starvation of

low priority traffic, the protocol includes an adaptive

scheduling component that adjusts the transmission

rate depending on the size of the queue and the state

of the traffic in the network. This allows both queues

to be processed very efficiently, especially during

peak loads.

Adaptive Queue Management: The proposed

method's respect for dynamic systems through

adaptivity makes it effective in time-varying

scenarios. Conventional static prioritization

approaches are often inadequate for addressing the

dynamic nature of the environment in autonomous

vehicle (AV) networks. We assume that the system is

capable of monitoring system factors such as traffic

density, packet arrival rates, and communication

delays, and dynamically adjusting its behavior based

on these inputs.

Just as the high-priority queue handles bursty

traffic, in a situation where the high-priority queue

becomes congested, the system reduces the

transmission rate of low-priority packets to set aside

additional resources for mission-critical traffic. In the

case of low-high priority traffic, the system rather

boosts the bandwidth of low-priority packets in order

to get the right ratio in both queues.

To enable this adaptability, we then employ a

Dynamic Time Slot Allocation (DTSA) mechanism.

It dynamically allocates time slots for transmission

based on the traffic load of each queue. By keeping

also some excess time slots reserved to send the

packets coming from the High-Priority queue in case

that that gets full, thus, significantly reducing the

chances of delays and/or packet loss. The lower

priority queue though, as that fills up, as those

packets are using longer time slots to be processed,

balances the load across both queues. It also takes

into consideration queue length and packet arrival

rates before making any adjustments. If packets are

fed continuously into a high-priority queue at a

Enhanced Trafﬁc Shaping for Lowâ

SLatency Communication in Autonomous Vehicle Networks

251

sufficiently high rate, for example, the system will

favor their transmission over low-priority traffic but

will also slow low-priority traffic as have needed for

the duration of that time. In the same way, more low-

priority traffic encourages the system to average out

the load by processing more low-priority traffic

packets than high-priority packets.

Latency Minimization and Queue Efficiency: A

fundamental goal of the proposed mechanism is to

reduce latency of the high-priority packets. AV

networks require safety-critical data to be delivered

promptly to avoid accidents or system failures. To

this end, we introduce a Guaranteed Latency

Mechanism (GLM) which guarantees transmission of

high priority packets in a bounded time interval.

The GLM dedicates a certain amount of transmission

time to high-priority packets irrespective of the

network load. This ensures latency requirements are

met for safety-critical messages. Once the high-

priority queue is empty, the reserved bandwidth is

realigned to handle low-priority packets, optimizing

the use of network resources.

The that aims to improve efficiency is the Queue

Length Monitoring algorithm. This approach

monitors the length of both queues in conjunction and

adapts bandwidth dynamically, ensuring that neither

of them becomes saturated. When queues cross a

critical threshold, resources are redistributed to

rebalance the system and prevent bottlenecks.

Simulation and Evaluation: The validation of the

employed traffic shaping mechanism has been

obtained using MATLAB simulations. The

simulations involved a fleet of autonomous vehicles

interacting with infrastructure in a typical urban

environment. This involved simulating real-world

conditions, including vehicle density, packet arrival

rates, communication delays, and changes in traffic

patterns.

The proposed mechanism achieves low latency

for high- priority packets, therefore ensures timely

delivery of safety-critical data as shown in simulation

results. At the same time, they shunted low-priority

traffic without adding too much latency. We

demonstrated the versatility of the adaptive queuing

model in sustaining reliability and scalability against

varying traffic models.

The results yield presents the viability of the

proposed system to enhance communication in the

autonomous vehicle network addressing both safety-

critical and non-critical applications in real-world

implementations.

5 RESULTS AND DISCUSSION

In this Section, we describe the MOM based

MATLAB simulations that are utilized to analyze the

performance of the proposed traffic shaping

mechanism for AV networks to achieve low-latency

communication. It used simulation-based evaluation

that compared the system's performance under both

high-priority and low-priority traffic in a variety of

networking scenarios. The conditions were various

vehicle densities, packet arrival ratios, and

communication delay. The only performance

indicators assessed in the simulations were the

latency of the high-priority and low-priority packets,

as well as statistical data about the queues and the

performance of the system. Figure 1- (Simulation of

Traffic Shaping in Autonomous Vehicle Networks).

Figure 1: Simulation of Traffic Shaping in Autonomous

Vehicle Networks -1.

Figure 2 and 3 Simulation of Traffic Shaping in

Autonomous Vehicle Networks -2 and Latency

Comparison respectively.

High-Priority Packet Latency: The simulation

experiments showed that the performance of the

proposed traffic shaping mechanism significantly

decreases the latency of high-priority packets like

safety alerts and real-time notifications. In situations

with a high density of vehicles and network

congestion challenges, there was an observed reduced

delay time for high-priority packets using the

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

252

adaptive queuing model. The latency for high-priority

packets was reduced by as much as 40% compared to

traditional methods for traffic management (such as

basic Priority Queuing (PQ) and Weighted Fair

Queuing (WFQ)) on average.

Figure 2: Simulation of Traffic Shaping in Autonomous

Vehicle Networks -2.

Low-Priority Packet Latency and Throughput:

Low-priority packets, e.g. infotainment data or low-

priority updates, were correctly managed and latency

was maintained even in network congestion. The

latency for low-priority packets did increase slightly

algorithms, which would queue up low-priority traffic

when high loads are experienced, resulting in severe

delays, and potentially affecting the user experience

on non-critical applications. This adaptive queuing

process ensured low-priority packets were still

transmitted efficiently and fairly even during periods

of congestion, without excessively adversely

affecting critical data. in cases where the higher one

was prioritized, but the delay was still acceptable and

non-critical applications were not greatly impaired.

This marks a great improvement over classical traffic

shaping in cases of extremely heavy amounts of

traffic, where the safety would be compromised

should there be any delay or packet loss, the

Guaranteed Latency Mechanism (GLM) showed to

be very efficient. The high priority messages like

collision warnings were delivered in limited time

slots reserved for the gliding message. In

autonomous vehicle networks, this capability is

critical, as even small transmission delays of safety-

critical data can produce large consequences.

Besides controlling latency, the system improved

throughput by evenly distributing the over-the-

network traffic of high-priority and low-priority

packets. The new scenario parameters significantly

reduced the transmission rate for low-priority traffic

in high-density scenarios, without detriment to

network wide performance. This enabled high-

priority traffic to receive the appropriate bandwidth

allocation while preventing transmission of lower-

priority packets from being excessively hindered by

the availability of bandwidth. Heavily tail referenced

workloads, driven by previously introduced synthetic

data, were also tested and the results showed that

regardless of the consistency of the traffic flows, the

model would ensure that both high-priority and low-

priority communication sides achieved service

without fail, resulting in reliable communication for

safety-critical applications and non-critical services

in dynamic, high-traffic environments. Table 1. Show

the Comparison of Traditional and Enhanced Traffic

Shaping Performance Based on Traffic Load.

Table 1: Comparison of Traditional and Enhanced Traffic Shaping

Performance Based on Traffic Load.

Traffi

Load

(Mbp

)

Tradition

Latency

(ms)

Enhanc

Latency

(ms)

Tradition

Throughp

ut (%)

Enhanced

Throughp

ut (%)

10 50 30 85 90

20 60 40 80 88

30 80 55 75 85

40 120 80 65 78

50 150 100 50 70

Queue Management and Resource Utilization:

“The system used adaptive queue management to

dynamically allocate resources according to current

conditions on the network,” Norden explained in a

statement. In the case of congestion at a high-priority

packet at a time it renegotiated more of the bandwidth

in a LINUX for traffic high-quality at a time, also the

queue low priority when it had become to load used

was heightened, redirected to to make it not delay not

for packets not critical.

By optimizing use of resources and avoiding

bottlenecks, the Queue Length Monitoring algorithm

prevented the queues from becoming overwhelmed.

In autonomous vehicle networks, traffic conditions

varied based on factors such as not only the density

of vehicles, but how they moved within that density

Enhanced Trafﬁc Shaping for Lowâ

SLatency Communication in Autonomous Vehicle Networks

253

and a multi-cluster transfer paradigm the first transfer

paradigm in a network to implicitly consider it

responded dynamically. Despite varying conditions,

the adaptability of the system ensured the

communication was efficient, further augmented by

the reliability of the system.

Figure 3: Latency Comparison.

Scalability and Robustness: The proposed model

develops one of the main highlights which is its

scalability. The system was evaluated with different

vehicle densities, from low to high traffic density

scenarios. With the increase of vehicles, the adaptive

queuing mechanism balanced the high and low

priority request.

Even with simulations with increasing numbers

of vehicles (up to 1,000), the system always

maintained a low latency for high-priority packets,

which confirms that the proposed algorithm works

very well in real-world urban settings.

The scalability of the proposed solution, as

demonstrated with real-world data, makes it an ideal

candidate for deployment in dense transportation

networks to allow efficient communication in high-

density situations.

6 CONCLUSION AND FUTURE

WORK

An implementation of an advanced traffic shaping

mechanism that targets lowlatency communication in

the context of autonomous vehicle (AV) networks is

introduced in this paper. The system complies with

the ever-increasing need to relay safety-related data

in real-time, and also manages traffic that is non-

urgent at the same time. The proposed solution then

leverages an adaptive queuing model to provide

minimal delay for high-priority traffic, such as safety

alerts and collision avoidance messages. It's able, at

the same time, to apply this for low priority data

without degrading overall network performance.

Simulations show that the mechanism

significantly lowers latency for high-priority packets,

while maintaining reasonable latencies for low-

priority packets even in highly congested situations

with an abundance of vehicles.

The proposed system shows a significant

enhancement in the communication features of AV

networks in that it guarantees the timely delivery of

safer-critical messages. Moreover, dynamic traffic

scheduling and dynamic resource allocation

mechanisms were implemented, allowing the system

to scale and adjust to fluctuations in vehicle density

and network traffic.

For the future, further refinement of the system

will be possible. There are myriad ways that this

solution could be enhanced and one of them could be

to improving resource allocation as per data detected

by machine learning algorithms to predict traffic

patterns better. A future improvement could be the

addition of various communication technologies i.e.,

V2X, 5G, Wi-Fi to create a more robust multi-

network dependent AV. OTMC introduces the

concept of PHM (prognostics and health

management) and SOTIF (safety of the intended

functionality), which will open the door to

development methods, tools and knowledge

supporting the development of risk-based safety for

intelligent future mobility.

REFERENCES

B. Li, Y. Li, and Z. Han, Hierarchical edge caching for

efficient mobile network resource management, IEEE

Trans. Wireless Commun., vol. 17, no. 6, pp. 3845-

3857, 2018.

C. Campolo, A. Molinaro, M. Ozturk, and I. F. Akyildiz,

Prioritized broadcasting techniques for multichannel

vehicular networks, IEEE Trans. Veh. Technol., vol.

66, no. 12, pp. 10615-10629, 2017.

F. Zhou, L. Chen, and B. Li, Vehicular fog computing: Use

cases, architectures, and security considerations, IEEE

Commun. Mag., vol. 55, no. 11, pp. 105-111, 2017.

F. Wang, Y. Xu, and H. Liu, A survey on emerging

technologies for intelligent transportation systems and

their impact, IEEE Trans. Intell. Transp. Syst., vol. 19,

no. 7, pp. 2395-2407, 2018.

G. Karagiannis, O. Altintas, E. Ekici, G. Heijenk, B.

Jarupan, K. Lin, and T. Weil, Vehicular networking: A

survey on requirements, architectures, challenges, and

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

254

solutions, IEEE Commun. Surv. Tutor., vol. 13, no. 4,

pp. 584-616, 2011.

H. Hartenstein and K. P. Laberteaux, A tutorial survey on

vehicular ad hoc networks, IEEE Commun. Mag., vol.

46, no. 6, pp. 164-171, 2010.

J. B. Kenney, Dedicated short-range communications

(DSRC) standards in the United States, Proc. IEEE, vol.

99, no. 7, pp. 1162-1182, 2011.

J. Feng, Z. Zhang, and J. Zhang, Latency-aware traffic

shaping for vehicular networks: A performance

evaluation, IEEE Trans. Veh. Technol., vol. 69, no. 10,

pp. 11732-11745, 2020.

K. Zheng, Q. Zheng, P. Chatzimisios, W. Xiang, and Y.

Zhou, A review of heterogeneous vehicular networking

architectures and their technical challenges, IEEE

Commun. Surv. Tutor., vol. 17, no. 4, pp. 2377-2396,

2015.

K. Abboud, H. A. Omar, and W. Zhuang, The interplay of

DSRC and cellular networks in V2X communication:

A comprehensive survey, IEEE Trans. Veh. Technol.,

vol. 65, no. 12, pp. 9457-9470, 2016.

K. Zhang, S. Leng, Y. He, and Y. Zhang, Green and low-

latency mobile edge computing solutions for IoT-

enabled smart cities, IEEE Commun. Mag., vol. 55, no.

3, pp. 39-45, 2017.

K. Zhang, Y. Mao, S. Leng, and Y. Zhang, The role of

mobile-edge computing in future vehicular networks:

Off-loading techniques and performance analysis,

IEEE Veh. Technol. Mag., vol. 12, no. 2, pp. 36-44,

2018.

M. Chen, S. Mao, and Y. Liu, Big data applications in

mobile networks: A comprehensive review, Mobile

Netw. Appl., vol. 19, no. 2, pp. 171-209, 2014.

N. Lu, N. Cheng, N. Zhang, X. Shen, and J. W. Mark,

Connected vehicles: Current solutions and future

research directions, IEEE Internet Things J., vol. 1, no.

4, pp. 289-299, 2014.

P. Wang and L. Jiang, A novel frame structure for low-

latency IoT wireless communication, IEEE Commun.

Lett., vol. 22, no. 2, pp. 304-307, 2018.

Q. Chen, T. Hu, Q. Wang, and W. Shi, Towards a

cooperative and intelligent vehicular social network:

Challenges and solutions, IEEE Trans. Intell. Transp.

Syst., vol. 20, no. 7, pp. 2444-2456, 2019.

T. Taleb, K. Samdanis, B. Mada, H. Flinck, S. Dutta, and

D. Sabella, A deep dive into multi-access edge

computing: Emerging trends in 5G architecture and

cloud orchestration, IEEE Commun. Surv. Tutor., vol.

19, no. 3, pp. 1657-1681, 2017.

X. Sun, H. Liu, and H. Zhang, Adaptive traffic signal

control strategies for large-scale urban networks, IEEE

Trans. Intell. Transp. Syst., vol. 19, no. 2, pp. 390-404,

2018.

X. Li, J. Wang, and L. Hanzo, Autonomous driving and

vehicular communications: An overview of challenges

and potential solutions, IEEE Commun. Surv. Tutor.,

vol. 21, no. 2, pp. 1243-1274, 2018.

Y. Zhang, Z. Ning, and X. Wang, Edge computing

applications in smart city infrastructure: Potential

benefits and future directions, IEEE Commun. Mag.,

vol. 56, no. 8, pp. 128-133, 2018.

Enhanced Trafﬁc Shaping for Lowâ

SLatency Communication in Autonomous Vehicle Networks

255