Algorithm development for Vehicle-To-Vehicle (V2V) communication

This paper presents the development of an algorithm for Vehicle-to-Vehicle (V2V) communication, a crucial technology in Intelligent Transportation Systems (ITS) that holds significant potential for enhancing road safety and traffic efficiency. One of the most common types of vehicle collisions occurs at intersections, particularly those without traffic lights. This study focuses on creating a V2V algorithm designed to prevent collisions in such scenarios. The findings were presented through visual simulations that depict various scenarios involving vehicles approaching an intersection. The algorithm follows a two-step process: Firstly, it utilizes Dedicated Short-Range Communication Systems (DSRCS) to accurately estimate the distance between vehicles. Leveraging this distance information, the algorithm dynamically adjusts the speed of each vehicle. The algorithm’s performance is assessed using Convolutional Neural Networks (CNN), which enables a comprehensive evaluation of its reliability and efficiency in V2V communication. The algorithm demonstrates notable enhancements in the reliability and efficiency of V2V communication. This paper serves as a validation of the feasibility of developing more advanced V2V communication algorithm and potentially making significant contributions to the advancement of ITS.


Introduction
Vehicles are used mainly for transportation from one place to another.It saves time for people by transferring them to their destinations.Traffic on the roads is increasing in metropolitan areas.Vehicles are producing pollution as well as increasing accidents.
Vehicle-to-vehicle (V2V) communication refers to the exchange of information between vehicles in order to improve safety, efficiency, and convenience on the road.This technology allows vehicles to "talk" to one another and share information such as location, speed, and trajectory.One of the primary benefits of V2V communication is the ability to detect and prevent accidents before they happen.By exchanging information about their surroundings, vehicles can alert each other to potential hazards and take evasive action to avoid collisions.This can significantly reduce the number of accidents on the road and improve overall traffic safety.
In addition to improving safety, V2V communication can also improve efficiency and convenience on the road.For example, it can be used to optimize traffic flow by allowing vehicles to coordinate their movements and avoid congestion.It can also be used to provide real-time traffic updates and route guidance to drivers, helping them avoid delays and arrive at their destinations more quickly.Vehicle to vehicle communication system uses the concept of internet of things (IoTs) and computing technologies [1].The focus of the study for this research work is at road-level flow as defined by Boukerche and Wang in [2].Some of the related studies especially discussing about the performance of the communication protocol in terms of range, reliability, and latency were discussed in [3,4].This study aims to develop an algorithm to avoid collision in a crossroad scenario.The algorithm was then evaluated and the result was presented by a visual simulation to represents several scenarios involve when vehicles approaching a crossroad.The objective of the research work is to reduce accidents and traffic congestion by interconnecting vehicles, especially ones approaching junction, and make decision based on set rules.The communication range of the vehicles was preset so that its start to develop a connection to other vehicles when the positions of each vehicle in the communication range.
Overall, V2V communication represents an exciting and innovative technology that has the potential to transform the way we drive and interact with the road.

Simulation Setup
A simplified algorithm is shown in the figure 1.The simulation setup will focus on a specific crossroad junction scenario, utilizing two cars as the primary entities of interest.Through the experimentation, a random speed of 30Km/h to 70Km/h and a range of 400M will be applied to the Vehicle-To-Vehicle (V2V) communication system.By systematically varying the speed parameter, we aim to observe and analyze the resulting outcomes, particularly the effects induced by alterations in speed settings.This approach will provide valuable insights into the dynamics and performance of V2V communication under various conditions within the context of a limited crossroad junction environment.
A comprehensive explanation of the setup settings and visuals is warranted to provide a clear understanding of their purpose and functionality.By delving into these aspects, we can establish a solid foundation for the subsequent discussion and analysis.It is essential to outline the key elements and configurations of the setup, as well as elucidate the visual components employed to enhance the overall simulation experience.This explanation will ensure that readers or users can grasp the intricacies of the setup settings and visuals, enabling them to make informed interpretations and engage meaningfully with the simulation.
Figure 2 provides a visual representation of the scenario wherein two cars are positioned near the entrance of the crossroad junction.Car 1 is depicted on the left side, while car 2 is situated at the top.The transparent blue region denotes the range of Vehicle-To-Vehicle (V2V) communication.Additionally, the top-right section of the figure displays the randomly generated speeds of the cars, which vary in each simulation run.This visual depiction offers a concise overview of the initial car positions, the V2V communication range, and the stochastic nature of car speeds, serving as a valuable reference for understanding the simulated dynamics.
In figure 3, a distinct visual cue is observed when the connection between cars is established.The circular boundary representing the Vehicle-To-Vehicle (V2V) communication around each car exhibits a green outline, indicating the presence of a successful connection between the vehicles.This visual indication provides a clear and easily interpretable signal of the established communication link between the cars, enabling observers to readily identify the connected state during the simulation.60 percent of road accidents can be avoided in the world if the vehicle driver gets a warning message [5].

Analysis
In this study, the analysis of data and the presentation of anticipated results will be facilitated through the utilization of a Convolutional Neural Network (CNN).The TensorFlow library, an open-source framework designed for building and training deep learning models, will be utilized in conjunction with Python.By integrating Python, TensorFlow, and Keras, this study aims to leverage the machine learning technologies to derive insightful results from the data analysis process.Both of these libraries are used to generate the confusion matrix (figure 4) and performance report.By taking the initial speed of both cars at the start of the simulation as input and the output as resulted crash state (crashed or not crashed).established, which can be utilized for subsequent analysis and reference.This component ensures the availability of comprehensive data for in-depth exploration and examination.Lastly, the analysis of the recorded data is conducted through the application of a convolutional neural network (CNN).This advanced machine learning technique enables the extraction of valuable insights from the recorded data.The analysis phase aims to uncover patterns, correlations, and significant findings within the collected data, ultimately enhancing the understanding and interpretation of the simulation outcomes.
Through the integration of these multiple outcomes, this project strives to provide a comprehensive approach that encompasses real-time visualization, data recording and extraction, and data analysis utilizing a convolutional neural network.

Crashes Range
As observed in figure 5, crashes begin to occur within a specific speed range, namely between 30 km/h and 70 km/h.Notably, crashes are more likely to transpire when both cars are traveling at similar speeds within this range.This occurrence can be attributed to the establishment of a connection between the cars when they are in close proximity to the center of the intersection.Consequently, this proximity introduces a delay in their response times, ultimately leading to a collision.Conversely, when one car is traveling at a higher speed than the other before the connection is established, crashes are avoided.This outcome can be attributed to the faster car reaching the intersection ahead of the other, thereby enabling it to safely avoid a collision.

Summary Statistic
Table 1 presents a statistical overview of our data, providing key summary measures for analysis.The count of our data is 8148, indicating a sizable dataset.The mean value is calculated to be 5, indicating the average value across the dataset.The standard deviation is 1, reflecting the dispersion of data points around mean.The minimum value observed is 3, while the 25th percentile is 4, the 50th percentile is 5, and the 75th percentile is 6.The maximum value recorded in the dataset is 7.These statistical measures offer a comprehensive understanding of the distribution and characteristics of our data, enabling meaningful interpretations and informed decision-making.junction environment.The recorded data, including car speeds and connection statuses, has served as a foundation for further analysis and evaluation.The application of CNN has been instrumental in uncovering hidden patterns and optimizing algorithms.Through extensive dataset analysis considering factors such as car speeds, distances, and connection statuses, informed predictions and crash prevention strategies have been developed.
Moving forward, future work in the field of V2V communication systems should consider expanding the simulation to encompass more complex junction layouts, such as roundabouts or multi-lane intersections.The integration of real-time traffic data from external sources and the development of intelligent decision-making algorithms using machine learning techniques hold potential for further optimization of traffic flow and safety.Exploring the integration of V2V communication with autonomous vehicles is another promising area to enhance coordination and efficiency in traffic management.
Practical implementation through real-world experiments and field trials is recommended to gain valuable insights into performance, usability, and acceptance of V2V communication systems.Gathering feedback from drivers, traffic authorities, and stakeholders is crucial for refining and tailoring the system to meet their specific needs.Additionally, addressing regulatory and policy aspects, including legal frameworks and privacy concerns, is essential to ensure compliance and create a favorable regulatory environment.
In conclusion, this study contributes to the existing body of knowledge on V2V communication systems, showcasing their potential for enhancing traffic management and safety.The findings and insights derived from this research provide a solid foundation for future advancements.As technology continues to evolve, the integration of V2V communication systems and advanced algorithms will play a pivotal role in promoting traffic efficiency, improving safety, and shaping the future of transportation.

Figure 1 .
Figure 1.Simplified flowchart for the algorithm.

Figure 4 .
Figure 4. Start of Simulation

Figure 5 .
Figure 5. Predicted Results of cars when travelling at different speed.Red dot shows high probability to crash.

Figure 6 .
provides a visual representation of the Model Loss.This model shows the ability of the machine to predict the expected output incorrectly, demonstrating the decreasing trend of the training loss across each epoch cycle.Indicating an increase in the ability of the machine to predict correctly.It is noteworthy, however, that beyond approximately 2000 epochs, the decrease in training loss becomes negligible, leading to diminishing returns in terms of further improvement.Continuing the training beyond this point consumes considerable computational resources without yielding substantial value or performance enhancement.This observation underscores the importance of carefully balancing the training duration to optimize computational efficiency while achieving satisfactory training outcomes.

Table 1 .
Data Exploration V2V) communication system within a simulated crossroad junction scenario.The effectiveness of the algorithms in regulating car speeds based on relative positions, optimizing traffic flow, and minimizing collisions has been demonstrated through the proposed methodology.Real-time simulation visualization has provided valuable insights into car behavior and interaction within the crossroad