Intelligent Driver Model-Based Vehicular Ad Hoc Network Communication in Real-Time Using 5G New Radio Wireless Networks

Vehicular ad hoc networks (VANETs) are self-organizing, open-structure inter-vehicle communication networks, along with wireless communication technology and transportation. Vehicle communication aims to improve the driver’s response-ability and ensure traffic safety when encountering traffic accidents. The 5G NR on V2V communication, compared to fourth-generation (4G) long-term evolution (LTE), is more sophisticated and has ultra-high reliability. This research proposes intelligent driver model-lane changes (IDM-LC) and intelligent driver model-avoidance (IDM-A) models, which enhance the performance accuracy of the V2V communication through the 5G NR system. In this case, the authenticity of the vehicle movement models is tested on different road scenarios, i.e., square, hexagon, heptagon and triangle. In these scenarios, the vehicles’ movements over the 5G networks receive widespread coverage whenever the vehicles receive signals from the roadside unit (RSU). Hence, the flexibility of VANET is used to improve the device-to-device (D2D) or vehicle-to-vehicle (V2V) communication efficiency in the fifth-generation (5G) new radio (NR) system. In addition, the VANET simulation platform, i.e., simulation of urban mobility (SUMO) and network simulator-3 (NS-3), simulate and evaluate the comparison of V2V through the 5G networks, which receives widespread coverage in 15 to 20 m/s. The simulation results and analysis show that the V2V communication through the 5G system performs better on the IDM models.

INDEX TERMS Vehicular ad hoc network, fifth generation, vehicle-to-vehicle, internet of vehicles, intelligent driver model, SUMO, NS-3.

I. INTRODUCTION
Vehicular Ad Hoc Network (VANET) is the wisdom of an important part of the traffic, which uses vehicles as mobile ad hoc networks. The nodes collect and transmit real-time road traffic information through collision rules avoidance and safety warning systems to improve the driver's safe driving environment. It can also reduce the number of casualties in traffic accidents and alleviate traffic congestion [1]. Despite being true, evaluating the performance of the in-vehicle network in the environment is critical, but due to the logical difficulty, economic issues, and technical limitations, the simulation of a proven vehicle routing algorithm is the best choice. When designing the simulation environment, one must define the appropriate vehicle shift so the dynamic model can properly reflect the real vehicle motion pattern. The increasing number of vehicles on the road has resulted in a drastic increase in traffic jams and road accidents. Therefore, for safety on the road, research in the vehicular ad-hoc network (VANET) has gained more importance in the recent past. The idea of VANET is to form an ad-hoc network with vehicles as the network nodes. VANET research has progressed mainly in two areas, i.e., vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communication. In V2I communication, connectivity between the vehicle and the roadside infrastructure is established, whereas V2V communication maintains the network among vehicles [2]. The institute of transportation systems in Braunschweig, Germany, focuses on research activities to increase the safety and efficiency of traffic. This includes developing assistance and automation systems within the automotive domain. Three different degrees of assistance are possible, i.e., information, warnings, and recommendations are provided to the driver; corrections of lateral and longitudinal movement of the vehicle; or the fully autonomous vehicle control. Therefore, the acceptance of these systems has to be evaluated. This is implemented at (Zentrum für Luftund Rundfahrt (DLR), which is a German aerospace center) using different simulators, test vehicles, and stationary test facilities [3]. For future assistance and automation systems, V2V and V2I, both together called vehicle-2-everything (V2X) communication, will be the underlying technology to develop and deploy a new class of assistance systems, which exchanges information like recommendations, warnings, and status information that vehicles could not gather on their own. Gupta et al. [4] established an indoor rapid test platform for autonomous vehicles based on vehicle-in-the-loop; Jaradat et al. [5] proposed and developed a human-vehicle road interaction mode virtual simulation system and driving data acquisition system, and realized the human-computer interaction and dynamic display of information in the simulation system. The vehicle industry, academia, and government agencies have jointly endeavored to realize the concept of vehicular communications [1], [6]; this field attracts researchers from different fields of development for applications, protocols, and simulation tools for VANET. Flack et al. [7] proposed a new hardware-in-the-loop (HIL) simulation platform. Marković et al. [8] proposed a multi-source sensing fusion system-in-the-loop virtual framework using radar and camera for V2X. Sun et al. [9] studied the virtualization technology of an intelligent networked vehicle computing platform. Duan et al. [10] proposed an active obstacle avoidance algorithm based on the lateral safe distance model. Li et al. [11] proposed a two-lane highway overtaking the safe distance model. Han et al. [12] improved the safe distance car-following model based on the optimal speed model. Aparow [13] systematically described an automatic generation method for autonomous driving simulation scenarios. Moayyed et al. [14] and Shirkolaei et al. [15] reported a resonator antenna with a different stack high bandwidth multilayer structure and microstrip patch antenna for high gain. Pourabdollah et al. [16] used three car-following models, IDM, Krauss, and Wiedemann, in traffic simulation for calibration and verification on the recorded trips of 200. VANET refers to an open mobile ad hoc network composed of communication between vehicles on the road and through the 5G fixed access points. It builds a self-organized, easy-to-deploy, low-cost, and open-structure vehicle-to-vehicle communication network on the road. VANET technology can realize applications such as accident warning, assisted driving, road traffic information query, inter-passenger communication, and internet information services through a 5G network system [2], [17]. With this, they face several challenges [2] since their objective is to provide services to users when they drive vehicles through certain elements that are part of these systems, such as sensors, computers, communications systems, location, and administration devices. To reduce traffic and accidents on the roads [18], one of the main challenges in these networks is finding and maintaining an effective route for information transport [17].
VANET builds an inexpensive, structurally open intervehicle communication network, self-organizing, easy to deploy, and low cost on the road. VANET is essentially a special mobile ad hoc network (MANET), which is similar to a general mobile Ad hoc network in that it has fast-moving nodes and changing network topologies. The road limits the distribution of vehicle nodes, showing a tubular shape state; network capacity is limited; vehicle node movement is subject to driver behavior and road traffic conditions. The impact of traffic control mechanisms; a relatively rich external information support, such as using global positioning system (GPS) to get the node's location information; the vehicle as energy conditions do not limit mobile nodes, and there is no need to consider energy saving when establishing routes. VANET in intelligent transportation system (ITS) the system plays a very important role, and its goal is to serve as the basis for future ITS. The basic part constitutes a unified no Line communication network used to transmit real-time information for assisted driving or avoiding accidents through the communication between the vehicleto-vehicle, the vehicle, and the roadside node. Data services such as entertainment information and Internet access are also available. With transportation infrastructure and IVC (intervehicle communication, Workshop communication) getting closer and closer, VANET is receiving research institutions worldwide. Close attention to researchers, a large number of research efforts to go through various ways to improve the performance of VANET. The first study was on communication between Japanese vehicles in the early 1980s and early 1980s [18]. Then, the European and American countries also started related research projects.
Some frameworks began to be developed from the standardization process carried out by the federal communications commission (FCC) by assigning 75 MHz for dedicated short-range communications (DSRC) [19]. The institute of electrical and electronics engineers (IEEE) adopted a draft as an amendment to the IEEE 802.11 family, which supports wireless communications. Said amendment is IEEE 802.11p created for vehicular environments to provide IVC and V2I in a speed range of 200 to 300 Km/h and with a coverage of up to 1000 m. IEEE 802.11p technology is strongly promoted by the vehicle manufacturing industry around the world, specifically in the United States through vehicle infrastructure integration (VII), in Japan with the vehicle safety communication consortium (VSCC) through the advanced project for safe vehicles (ASV), in Europe through the car-2-car communications consortium (C2C-CC) and in Germany through the secure vehicle communication (SeVeCOM) [1], [20]. The estimated cost of IEEE 802.11p implementation is expected to be relatively low compared to cellular technology. Therefore, this technology would have an advantage over cellular technologies and is much more suitable for VANETs [1]. However, it is currently carried out studies regarding heterogeneous networks applied to this field.
Lee [21] proposed an intelligent driving aid system that can assist drivers in recognizing the traffic objects around the host car, turning on the lights or wipers, and preventing collisions, which decreases the risk level of accidents and increases traffic safety. Road traffic safety applications [22] are the most important application that can best reflect the value of VANET in a 5G network system, including active accident warnings, accident-hidden danger reminders, etc. Whether it can effectively avoid and reduce traffic accidents depends on whether the accident warning information is reliable, fast, and error-free transmission. Currently, most VANET research is aimed at this application, and a reliable broadcast mechanism is often used to achieve accident alarms [23]. The IVC is an important component of ITS [24], [25]. It begins with the formation of the V2V and V2I communications network. In V2V communications, the vehicles communicate with other vehicles either directly or using intermediate vehicles through the hopping scheme (hop). In the case of V2I, the vehicles communicate directly with the roadside units (RSUs) to exchange messages. At present, some routing protocols are used in VANET. However, some MANET routing protocols have also been analyzed in these environments [26]. They have been affected in various aspects due to the dynamics of the vehicular network, such as routing, network delays, inefficient radio allocation, computational complexity in the network; and to face these problems, efficient and intelligent protocols for sending data are needed [26]. Based on the real vehicle operating characteristics, this paper expands the in-vehicle network mobility model IDM-LC based on considering the driver's car-following behavior and IDM-A, which considers the avoidance behavior. The models, through the network simulation tools SUMO and NS-3, complete the VANET scene construction and use the method of discrete parameter values to simulate and analyze the vehicle trajectory changes caused by vehicle avoidance behavior under different traffic densities. The simulation results show that the built scene can effectively simulate the avoidance behavior of vehicles through the 5G network system, and the vehicle trajectory is affected by the avoidance signal differently under different traffic densities. Our major contributions of this paper are as follows: • The VANETs have broad prospects in V2V communication to collect and transmit real-time road traffic information to improve vehicle node movement.
• Over 4G LTE, the popularity of the 5G new radio on V2V communication is more sophisticated and ultra-high reliability which performs better accuracy in different road scenarios by our approach.
• We consider IDM-LC and IDM-A models, which enhance the performance accuracy of the V2V communication through the 5G NR networks.
• The IDM-LC and IDM-A models also realize and extract the path using the SUMO and NS-3.
• The proposed technique uses different road scenarios, such as square, hexagon, heptagon and triangle. The vehicles' movement receives widespread coverage in 15 to 20 m/s to make vehicles' nodes more responsive.

ORGANIZATION OF PAPER
The remainder of the paper is organized as follows. Section II presents the model of vehicle movement and classification analysis. Section III describes the architectures and features of 5G vehicle networking. The simulation analysis and discussions are presented in Section IV. Section V concludes the paper with future scope.

II. MODEL OF VEHICLE MOVEMENT
A key factor in the simulation of VANET is real vehicle movement. The mobile model refers to the motion model of the mobile node at any time, and its extension pounces the position in the map. Early researchers working on VANET often used simple stochastic models in MANET research, graphically constrained moving models, etc. Models cannot describe vehicle mobility in a real way; they ignore vehicle traffic's special aspects of Tongzhong, such as vehicle acceleration and deceleration in the presence of a nearby car for the queuing at road intersections, traffic congestion caused by traffic lights, traffic congestion, etc. All of these realities greatly affect the network performance of VANET. Boukenadil [27] human proved that the real mobile model is very important for the correct simulation of VANET. The following summarizes the mobile models in the literature and presents real vehicles' conceptual framework for moving models.

A. MODEL OF VEHICLE MOVEMENT CLASSIFICATION ANALYSIS
The traditional classification method is based on the model expression for the classification of vehicle movement models. The degree of detail is divided into macro, mesoscale, micro, and sub-micro models. Macro model parameters such as vehicle flow, vehicle density, and vehicle-average speed usually contribute to fluid dynamics theory. To calculate, mesoscopic models use gas dynamics and queuing theoretical or macro-scale matrices, such as speed/density relationships.
To study the transport of vehicles, the microscopic model simulates the movement of every single vehicle on the street, mainly assuming the vehicle's behavior depends on two factors, namely the physical ability of the vehicle and the driver's control. Think of each vehicle as microscopic, but divide them further into smaller ones, such as the relationship between engine speed and vehicle speed. Gholamhosseinian and Seitz [25] moved the vehicle according to the source of the vehicle model. The model is divided into four categories, i.e., synthetic based on traffic simulation, trafficbased, survey-based, and trajectory-based models, as shown in Fig.1.
Model-based on traffic simulator: The history of traffic simulators is relatively long; some companies or research teams have developed a lot of famous software, such as PARAMICS, Corridor simulation (CORSIM), verkehr in städten (VISSIM), and TRANSMS, capable of simulating urban microscopic traffic [28], energy volume consumption, even pollution or noise level monitoring. These business model parameters are miscellaneous for VANET network simulation.
Survey-based model: Survey-based model based on surveys of a real social behavior model. The main large-scale survey is available from the US Department of Labor. They conducted a survey and collected a large amount of statistical data on the behavior of American workers, such as above class time, lunchtime, travel distance, etc., the typical representative of this type of model is UDel and survey-based models. This type of model can truly be the non-random behavior of urban traffic life; the shortcoming is that statistical data is not easy to obtain [29].
Trajectory-based model: There are two possibilities for the origin of trajectories in this type of model, i.e., collecting from various measurement activities; simulating real trajectories with a simulator. The measurement method consumes a lot of manpower and material resources and cannot obtain the trajectory for free, and the simulation method requires a huge amount of calculation.
Comprehensive model: The comprehensive model is actually a mathematical model that integrates all situations. In order to design a mathematical model that can reflect the real physical effects, many scholars have been conducting research. As per Fiore's classification, as shown in Fig.2, comprehensive models are divided into four categories, i.e., stochastic models, which include all models of purely random motion; traffic stream models, which treat vehicle mobility as a hydrodynamic phenomenon; vehicle-following models [30]. The models in which each driver's behavior is modeled according to the vehicle in front of him; are traffic flow and interaction models.
From the perspective of different classifications of vehicle models, the process of constructing a vehicle movement model is very complicated. Distinguishing between major classification parameters, such as topological maps, automotive production, the method of birth, the method of driver behavior, etc., cannot be randomly selected and must reflect the true match set. Then how can the model be configured to reflect the vehicle in the real VANET network? What about traffic? The next section is the author's summary of  the vehicle movement model proposed in the literature [31]. A conceptual framework built by real vehicle movement models built with key parameters.

B. INTELLIGENT DRIVER MODEL-LANE CHANGES
The intelligent driver model-lane changes (IDM-LC) considers the driver's following behavior, and the vehicle's speed is affected by the preceding vehicle's speed. IDM-LC uses the MOBIL model to control the lane-changing behavior of the vehicle. The lane changes are allowed when the following inequalities hold (1), where a l − a represents the amount of acceleration increase after the current vehicle chooses to change lanes to l, a new − a l new represents the acceleration loss of the following vehicle after the current vehicle enters the candidate lane, p is the politeness parameter, when the value of p is between 0 and 1, the more polite the driver, the smaller the value. a thr represents the minimum acceleration gain threshold allowing lane change; if its value is large, the vehicle will not change lanes. a bias is used to adjust the preferred movement behavior in the model. In IDM-LC, the vehicle lane change will introduce a bias to make the model conform to the behavior of the vehicle passing on the right side. It indicates that after the current vehicle changes to the lane, in the new lane, the deceleration of the vehicle behind the current vehicle must be greater than the safety value a safe .
Although the IDM-LC mobility model can well reflect the vehicle's overtaking and lane-changing behavior, it cannot effectively simulate the vehicle's evasive lane-changing behavior. The two main reasons are as follows.
(1) The IDM-LC mobility model, like most other mobility models, assumes that all vehicle nodes are in an equal state and abide by the same mobility rules without prioritizing vehicles.
(2) The IDM-LC mobility model only takes the current vehicle as the center and imitates natural driving behavior by collecting the driving status information (such as distance, position, speed, acceleration, etc.) of nearby vehicles and does not consider the information interaction between distant vehicles [32]. Therefore, it can only effectively reflect the feedback driving behavior of the driver to the traffic conditions in the field of vision and cannot properly reflect the information transmitted over long distances, such as sounds (avoidance signals and sirens).

C. INTELLIGENT DRIVER MODEL-AVOIDANCE STRUCTURE
The establishment of the driving safety distance monitoring and collision avoidance control model is affected by many factors, and the establishment methods are also diverse. There are traditional models based on mathematics and practice and models based on intelligent algorithms [33], such as the stereo vision algorithm [34]. The establishment and integration of a general virtual simulation system can only verify a model related to the simulation system. With the rapid development of intelligent technology, the establishment of smart vehicle safety distance monitoring and collision avoidance control models is also constantly updated and optimized [34]. So it is required that the safety distance monitoring and collision avoidance virtual simulation system can be embedded in multiple control models to verify the control effect under different models.
Li and Yu [35] used Zigbee wireless communication technology to collect the running state information of the front and rear vehicles in real-time and established an improved Berkeley model to dynamically select the safety warning distance according to the different collision avoidance times of the front and rear vehicles, to meet the requirements of the safety warning distance under different collision avoidance decelerations. Suppose it is assumed that the vehicle travels at a constant speed before the collision avoidance system intervenes when the active system takes effect. In that case, the vehicle's deceleration shows a linear trend for calculating the safety warning distance of the improved avoidance model as (2).
where D w is the safety warning distance, t x represents the driver's reaction time and the vehicle's delay lag time, a 1 is the maximum-collision deceleration of the vehicle; a 2 represents the maximum-collision avoidance deceleration of the vehicle in front, t s is the time required for the collision avoidance process, v c represents the speed of the vehicle, v rel is the relative speed of the two vehicles traveling front and rear (km/h) and d 0 represents the minimum safety distance. Aiming at the two major problems that IDM-LC cannot effectively reflect the evasive driving behavior, the structure of the vehicle movement model of IDM-A is constructed as shown in Fig.3. Vehicles are divided into two different types, which are special and ordinary vehicles, and it is assumed that each vehicle can receive the avoidance signal sent by the special vehicle through WiFi [33]. The method of the vehicle's driving flowchart is as follows.

III. 5G VEHICLE NETWORKING ARCHITECTURES AND FEATURES
Future 5G communication technology in the application of vehicle networking scenarios to make vehicle networking have more flexible architecture and new system elements, i.e., 5G vehicle on-board unit (OBU), 5G base stations, 5G mobile terminals and 5G cloud servers, etc. In addition to the car network, the car network, the car mobile internet implementation V2X, such as car, road, pedestrian, and internet. In addition to information exchange, 5G car networking will implement OBU, interconnection of base stations, mobile terminals, and cloud servers, giving them special features and communication methods. The characteristics of the structure are mainly reflected in the OBU multinetwork access and integration, OBU multi-channel internet access, and multiidentity 5G base stations [36], [37].

A. OBU MULTI-NETWORK ACCESS AND FUSION
In the IoVs, multiple networks coexist, including wireless local area network (WLAN) over the IEEE 802.11p standard protocol, cellular 2G and 3G, LTE, and broadcasting transmission networks. These standards and protocols apply to in-vehicle networking communications. Data processing and information interaction are not perfect. And the 5G car network will melt A variety of networks for seamless information interaction and communication switching. 5G mobile communication network is a macro layer designed as a two-layer network of spare layers [37]. Macro cellular layers and traditional cellular networks are similar and involve direct communication between the base station and the terminal device. D2D is an important group of 5G mobile communication technologies in device-level communication. In part, it is a kind of terminal and terminal without any network communication method in which the infrastructure directly exchanges information [18]. According to the base station allocation of resources and pair of origin, destination, and relay terminal nodes, the control situation, D2D terminal communication mode can be divided into four categories [37].
1) The base station controls the terminal forwarding of the link. Terminal equipment can operate in an environment with poor signal coverage, the letter through the neighbouring terminal equipment, information forwarding, and base station communication. The terminal device can achieve a higher quality of service (QoS) in this communication mode, where the base station and relay device control and establish the communication link.
2) The terminal of the terminal control link communicates directly. Between terminals Communication without the assistance of base stations and terminal equipment can control the chain of the establishment of the road, which helps to reduce the interference between devices.
3) Terminal forwarding of the terminal control link. The base station does not participate in the communication. The establishment of the link and the information exchange, the source terminal, and the destination terminal pass the relay devices coordinately control communication with each other.
D2D communication method of the 5G wireless communication network in the application of the IoVs, as shwon in Fig.4. Future 5G car networking D2D communication technology will provide a new communication model for the IoVs. Among them, in the car shift internet, OBU can directly pass 5G BS or relay (package including neighbouring OBUs, mobile user terminals) to access the internet quickly and realize the information interaction between the car and the cloud server. In the car network, the human-computer interaction between the user and the vehicle is realized, and the OBU is used as a medium. There is no BS or other terminal between the user and the 5G mobile terminal. With the device's assistance, the short distance is controlled by self-controlling link. Detached vehicle data transmission; in a D2D-based communication network, a 5G car unit can be on the edge of network communication or signal congestion to establish an ad hoc network based on single-hop or multi-hop D2D vehicle self-organizing network communication [37], [38] through the above analysis of 5G vehicle networking communication methods, as shown in Fig.4. The 5G car networking will change based on IEEE 802.11p standard car networking communication method, implemented between multiple entities (OBU and between OBU and owner mobile terminals, pedestrians, 5G BSs, information exchange between the internet to achieve multi-network connection of OBU [39]. When a traffic incident occurs, the traffic status message will be broadcast repeatedly in a short time and a short distance, which is easy to cause a heavy load on the limited link. Therefore, under the control of the base station, the vehicle establishes a D2D connection pair with its neighbor vehicles within its communication range and disseminates real-time traffic safety warning messages to warn of traffic accidents. Tehrani et al. [37] in the IoVs, it is assumed that vehicles can realize inter-vehicle D2D communication under the control of the cellular network base station, the communication range 4962 VOLUME 11, 2023 of each vehicle is R, and the vehicle periodically collects the traffic information of its neighbor vehicles, mainly including the current speed, direction, position and location accuracy, etc. The D2D connection between vehicles reuses the uplink spectrum resources of the cellular network, and the base station allocates orthogonal spectrum resources for the D2D users and the cellular users to reduce the co-channel interference with the cellular users.
The V2V communication mode of VANETs will cause link breakage due to sparse vehicles, reducing transmission performance. As shown in Fig.6, when the source node wants to send a packet to the destination node, no vehicle in the middle can assist in the packet transfer, so The link is interrupted so that the destination node cannot get information in real-time; instead, it is necessary for the vehicle to slowly catch up with the vehicle ahead at a faster relative speed in order to continue packet forwarding. As discussed in the previous subsections, 5G NR systems can serve as an important communication channel to supplement VANETs.
Suppose a vehicle collision accident occurs in lane 1, and the onboard sensors of the relevant vehicles generate a brief traffic accident warning message after detecting the accident, including the accident location, time, and accident type, as shown in the collision message in Fig.6. The message is uploaded to the base station, and the accident vehicle, that is, the source vehicle, requests the base station to establish a D2D connection pair for dissemination of the traffic warning message, and the base station establishes a D2D connection pair for the source vehicle according to the obtained link information. It is assumed that the transmission power required for D2D communication is adaptively controlled and is much smaller than the transmission power required for cellular communication [37].

B. MULTIPLE IDENTITY 5G BASE STATION
The traditional BS as a relay for terminal communication, in data forwarding and link control play an important role; a large number of 5G BS will be deployed in an ultra-dense network. The user can accurately locate and assist the terminal communication. Based on 5G in millimetre-wave communication networks, D2D technology involves terminals and BSs. Direct-to-base station (D2B) communication and the base station-to-base station (B2B) will communicate in a self-organizing manner [40]. D2B and B2B are significant breakthroughs that determine that 5G BSs will differ; it shows an important task. Application field in the IoVs, the 5G BSs will have the following features.

1) COLLABORATIVE RELAY
5G BSs have relay transfer of traditional base station that sends function, as a wireless access point, to assist the vehicle in communicating with the internet.

2) ACT AS RSU
In a high-speed environment, the vehicle 5G BS in networking communication will replace RSU in real-time with OBU communication, broadcasting to vehicles in the vehicle ad hoc network traffic information and assisting with V2V and multiple vehicles network communication. The construction cost of the car networking system also solves the multi-party faced by the V2I cooperative communication system [37], [39].

3) PRECISE POSITIONING
GPS, as the current OBU's positioning system, is fragile and vulnerable to deception, such as a blockage type attack, etc. And GPS signals are susceptible to weather, making it impossible to implement precise positioning [41]. Many futures 5G BSs use a higher frequency and signal bandwidth to implement a dense network and large-scale antenna arrays, making OBU complex loops in NLOS reduce positioning errors in the environment. Secondly, D2D communication fully uses the advantages of density terminal equipment connection, which are improved from the following two aspects of positioning performance [40]. On the one hand, many D2D links can be the pseudo-range between the vehicles that provide signal observation, such as (3) and (4). D2D communication allows the OBU to receive neighbors' vehicle and mobile terminal information. The synchronization and channel estimation unit entities, such as signal processing, can also be reused for signal transmission delays estimate. On the IoV, the D2D VOLUME 11, 2023  N (N − 1) for a network. On another side, the OBU's D2D transmission link is directly positioned to change the required data to speed up local decision-making further and improve the position estimate of the convergence time of the process. Fig.7 shows the D2D-based collaboration bit system; vehicle terminal OBU#1 receives signals from BS#2.
where η b2,v1 is the scalar factor used to describe the large-scale fading characteristics, s b2 is the transmitted baseband signal, c 0 describes signal transmission delay, and t v1 is the time base of the vehicle terminal OBU#1. With the time offset of the cooperative network, n b2,v1 (k) is additive white Gaussian noise. The pseudo-range between the vehicle terminal OBU#1 and the BS#2 as (3). When the base-band signal s v1 (k + T v1 ) is sent from the in-vehicle terminal OBU#1, the signal can be received at OBU#2, as (4).
5G mobile communication fusion cognitive radio (CR), millimetre wave, large-scale antenna array, ultra-dense networking, full-duplex (FD) wireless communication and other key technologies [42] significantly improved the communication system Performance. In the vehicle networking application scenario, IEEE 802.11p standard communication, 5G car networking features are mainly reflected in low latency efficient use with high reliability, spectrum, energy, and superior Communication quality.

C. LOW-LATENCY AND HIGH RELIABILITY
As the sender, receiver and relay section of the car network information point, the message passing process must ensure privacy, security and a high data transmission rate; communication has a strict delay limit [36]. Currently, real-time requirements are very high with intensive use of researched car networking communication data and frequent exchange changes. However, you can't support secure internet needs by wireless communication technology limitations (such as bandwidth, speed, domain name, etc.) and communication delays of less than milliseconds.
5G high/ultra-high-density networking and low equipment energy consumptions reduce the signalling overhead on the ground and solve the bandwidth and delay-related problems. As the 5G delay reaches the millisecond level, meeting the low latency and high-reliability requirements have become the most significant breakthrough in developing the IoVs. In 5G car networking communication, for better research and application of low latency and high-reliability link characteristics, literature [36] analyzed the adaptation to 5G adaptive antenna for 300 km/h speed mobile vehicle communication, high communication quality between OBU and BS, reducing channel estimation delay between the transmission and the data transmission [40]. Reynaud et al. [43] proposes to use the network-functions virtualization (NFV) and software defined networking (SDN) technologies to increase the flexibility of 5G network architecture and proposed a solution to achieve low latency services. Mainly including schedule and configuration to reduce the delay of IP address resolution, continuous service and time delay optimization. Optimizing 5G network services supports current application services, adapts to high-speed growth interest, and meets the needs of future diversity of services [36]-especially highly sensitive communications over time, such as the IoVs and V2X communication scenario.
The low latency and high reliability are 5G network architecture significant features of the application. Implementation based on D2D mode based on the key factors set as seen in Table 1, the V2V communication delay simulation yields the results shown in Fig. 5. As the amount of vehicles rises, the end-to-end (E2E) transmission interval is guaranteed stable, and 5G car networking will be based on D2D technology. The delay in the 5G mobile terminals is around 1 ms, and the E2E delay is controlled in milliseconds [44]. The delay performance is superior to the IEEE 802.11p standard. Effectively guarantee the reliability of communication [39].

D. SPECTRUM AND ENERGY EFFICIENT USE
Efficient use of spectrum and energy is one of the 5G user experiences important features. 5G communication technology in the application of the IoVs, and it is necessary to limit the current resources of the IoVs. 5G car networking 4964 VOLUME 11, 2023 spectrum and energy-efficient use are mainly reflected in the following aspects.

1) D2D COMMUNICATION
In 5G communication, D2D communication mode direct terminal communication is achieved by cellular multiplexing resources. 5G BS signal will be based on D2D technology and adjacent car unit, mobile terminal, car network, self-organizing network communication, and multi-channel mutual network access. In such a manner, the IoVs' communication spectrum is improved. Utilization [37], and vehicle networking based on the IEEE 802.11p standard compared with V2X communication, reduces costs and save energy.

2) FULL DUPLEX COMMUNICATION
5G mobile terminal equipment uses in full-duplex communication mode, allowing different terminals and 5G BSs channels in the same frequency band to simultaneously transmit and receive information to increase spectrum efficiency [42].

3) COGNITIVE RADIO AND 5G COMMUNICATION TECHNOLOGY
Cognitive radio is one of the essential technologies of the network [37], [42]. In the car networking application scenario, the vehicle terminal obtains efficient communication through the perception of the whole current wireless spectrum communication environment information, which accesses fast to idle spectrum, and other terminals. This dynamic spectrum access application meets in-vehicle users' spectrum requirements and increases the spectrum resources' utilization. The vehicle terminal can use cognitive radio technology to interact with other authorized users to share spectrum resources to solve the wireless spectrum resources problem of shortage.
In addition to the spectrum and energy-efficient applications mentioned above, most recent related research shows that without affecting communication performance, the deployment of large-scale antenna arrays for 5G BSs has a potential savings source role [36], [38]. Secondly, in the vehicle ad hoc network, 5G BSs can discover and communicate with neighbouring terminal devices in time and also reduce energy consumption between OBUs.

E. BETTER COMMUNICATION QUALITY
5G communication networks are estimated to have a higher-network capacity and provide a gigabit per second (Gbps) data rate per user to full of QoS requirements. Sekine et al. [45] proposes that the frequency band is 30∼ 300 GHz. Millimetre-wave (mmWave) communication system can be between 5G terminals and information exchange between BSs with better communication quality. Millimetre waves have great bandwidth, provide a very high data transfer transmission rate, and reduce various interferences in the environment, reducing the terminal between the probabilities of a connection being interrupted. As seen in Table 2, a 5G vehicle networking and IEEE 802.11p-based comparison of the key technical parameters of VANET in the IoVs [2], the results show that the 5G vehicle network has superior advantages over the current vehicle network and wireless link characteristics.
1) Communication distance, 5G vehicle networking V2V communication maximum distance approximately 800 m, which can solve IEEE 802.11p short-lived, discontinuous connection problems in vehicle ad hoc network communications. Especially it is a non-line-of-sight (NLOS) ring that encounters large objects occlusion during communication, under the border [31], [46].
2) Transmission rate, 5G vehicle networking provides high V2X communication fast downlink and uplink data rates maximum transmission rate with a range of 1 Gbps.
3) High-speed mobility communicates with the IEEE 802.11p standard. More than 5G car networking supports faster vehicle communication, where the maximum speed of the supported vehicle is approximately 350 km/h. 5G vehicle network communication technology considers advanced for the vehicles. The network field has improved the traditional car network's communication method and quality. It quantitatively optimizes the architecture of the IoVs for development with a major change, but the 5G vehicle network also faces a major challenge, mainly in interference management, communication security and driving safety, which has three aspects [37].

IV. SIMULATION ANALYSIS AND DISCUSSION
Due to the particularity of the vehicle network, the simulation of VANET considered not only the network-related issues but also the vehicle's mobility. For different test requirements, according to the interaction between the network simulator and the mobile model, the VANET simulation is usually divided into three forms, i.e., isolated, embedded, and integrated [46].
There is no doubt that an emergency vehicle will have an impact on vehicles on the same road segment after it joins the lane. In order to test the ability of evasive driving in the IDM-LC model to affect the VANET topology in different traffic density scenarios, in the 800 m highway scenarios of the IDM-LC and IDM-A models, when the traffic   distribution. Although the traffic density increases, the number of vehicles joining the middle lane does not increase strictly according to the traffic density. Therefore, the travel time of the special vehicle does not show an absolute increasing trend, but the two different mobile model scenarios can be compared and analyzed because of the same traffic density and the same vehicle distribution in the scene before the special vehicle enters. As shown in Fig. 6, when an emergency vehicle enters the scene, the number and position distribution of vehicles in the same lane in front of it will affect its speed. Especially when the traffic density is 4 to 6, the special vehicle in the IDM-A model scene travels 800 m away. The time required is obviously less than that of the IDM-LC model scene, and the vehicle avoidance effect is obvious during this time. When the traffic density is particularly small, the speed of the special vehicle can only reach the set maximum speed because there are few vehicles in the scene, and in order to ensure driving safety, even if there are no other vehicles in front of it, so the avoidance behavior will affect the driving of the special vehicle. If the strength is low, the effect of avoidance is not obvious. Before each vehicle is added to the road, it will be judged in advance whether there are other vehicles in the lane. If so, the newly generated vehicle will only enter the scene when the distance of the vehicle closest to the road entrance is greater than the minimum safe distance. Therefore, even if the traffic density increases, after the traffic density is greater than 10, the time required for emergency vehicles to travel in the IDM-LC and the IDM-A model scenarios tends to be a constant value.
The isolated approach allows the independent development of mobility and network modelling and is the main form of early VANET simulation, but it does not meet the extensive interaction needs in most VANET studies. Fusion is the current mainstream form of simulation, which enables state-ofthe-art mobility models or traffic simulators to work together with the latest and most efficient network simulators. For example, traffic network simulator (TraNS), which integrates the traffic simulator SUMO and the NS-2, we extract the path from SUMO through the interface and then transmits it to NS-3 and sends commands from NS-3 to SUMO's interaction. However, the amount of computation required is very large due to running two simulators simultaneously, and there is a communication delay between different simulators. Not ideal for VANETs with fast dynamic topology changes. The embedded method makes the mobile model and the network simulator merge into a single simulator, the interaction between the simulators can be simple and efficient, and the node positions can be synchronized more accurately. The simulation structure of this embedded method has higher simulation accuracy for VANET-related applications that require high real-time information transmission and reception, such as traffic safety and traffic efficiency. It is the future development direction of nautical ad-hoc network (NANET) simulation. The capacity limitation of network simulation has always been a difficult problem to solve. The reference information for decision-making is based on as follows.
It is assumed that the maximum distance the 5G NR base station can transmit is d max and the value with the smallest VOLUME 11, 2023 transmission bandwidth (occurring at the boundary of the transmission range) isb comm . The transmission bandwidth closest to the base station is assumed to be b comm (occurring at a distance from the base station). If the distance between the vehicle and the base station is x distance, according to the characteristic that the signal quality will be inversely proportional to the square of the transmission distance, we set the expected bandwidth (EB), i.e., EB 5G of the 5G NR system as (5).
After the above parameters are given, this paper proposes a method to calculate the expected bandwidth value EB VANET of packets transmitted through the VANETs system and the expected bandwidth value EB 5G of the 5G NR system, respectively. Finally, according to the calculated two, the relationship between the value and size is used as the routing selection decision. That is, if EB VANET >EB 5G , the vehicle will select the V2V mode of VANETs to continue transmission; on the contrary, if EB 5G >EB VANET , obviously it will be a relatively difficult task for the vehicle node to select the 5G base station as the transmission path. Since the EB 5G of 5G will change due to the distance change, the decision judgment can set the expected value judgment as each base station, as shown in Fig.7. The simulations in different road scenarios, such as square, hexagon, heptagon, and triangle, are shown in Fig.8, which shows the V2V communications with the designated RSU. The throughput variation over time in each RSU is presented to the proposed road scenarios, as seen in Fig. 9, and the models are applied to the VANET's highest performance. In the road scenarios, we have four active RSUs. The VANET models show how two or more RSUs (RSU-1 to RSU-4) communicate with vehicles simultaneously. In the RSU-1, there is higher throughput in the RSU-4. However, in this second, it can be seen that it has maximum peaks, but it is not a constant thought as it occurs in the RSU-1. The vehicles' movements received widespread coverage in 15 to 20 m/s whenever the vehicles received the signals from the RSUs. In Fig. 8, you can see points marked in red in the scenario, representing the uncovered area, and points marked  in green representing the covered area through the active RSUs resulting from the optimization process. And where the points marked in black represent the blank lane/road trajectory. In the simulation results, it can be verified that the square road scenario is the most used road style and have the most feasibility; over the other road scenarios, such as hexagon, heptagon, and triangle. The four active RSUs represent better coverage in these road scenarios. Therefore, in the case of IDM-LC and IDM-A through 5G networks, the real-time experimental results have been obtained at the time of vehicle node movement, as seen in Table 3. The vehicle node wants to send the packet to the other node, assuming that the vehicle is also located in the 5G coverage of the BS or RSU of the 5G system, but the bandwidth between the base station and the vehicle (source) is only 3 Mbps due to the long distance. At this time, the vehicle node decided and sent the data packet through the 5G system. V2V mode of VANETs is most suitable through the 5G NR system. 4968 VOLUME 11, 2023 Yang et al. [47] pointed out that the calibration of the warning time of the anti-collision control system should comprehensively consider the driving auxiliary control time and the vehicle duration. So, according to different driving styles, the warning time and alarm time calibration results are shown in Table 4. Due to the diversity and complexity of driving, many factors can characterize driving characteristics. It is challenging to quantify parameters such as driver emotions, vehicle interior and exterior vision, road conditions, etc., so the current research has yet to consider all disturbances fully. These factors will become an important research direction of the future collision avoidance control early warning system through the proposed method with a 5G NR network. The routing decision proposed in this paper is based on estimating the expected bandwidth of the 5G system and the VANET in the V2V system. Therefore, the experimental simulation compared the values of different parameters, as shown in Table 1. Fig. 10 shows the performance comparison of the proposed system when the road segment's vehicle density λ is set to 41. The horizontal axis in the figure is the change in the length of the monitoring distance, and the vertical axis is the change in the length of the monitoring distance bandwidth. As can be seen from the figure, 5G still has a high bandwidth when the distance value is set to less than 1585m, but beyond this distance, V2V of VANETs will become a better choice.

V. CONCLUSION AND FUTURE SCOPE
In this work, we proposed novel VANET models (IDM-LC and IDM-A) to enhance V2V communication using the 5G NR system. Our work is based on the fact that an autonomous vehicle perceived the driving environment better by the 5G communication system using the neighboring RSUs. The proposed technique tested different road scenarios, such as square, hexagon, heptagon, and triangle. The four active RSUs represented better coverage in these road scenarios. So, the vehicles' movement received widespread coverage in 15 to 20 m/s to make vehicles' nodes more responsive. VANET was used to improve D2D/V2V communication efficiency in the 5G NR network system and to improve the utilization rate of 5G network resources. The reliability of the VANET network simulation results evaluated the vehicle movement model's authenticity. In addition, the comparison of V2V through the 5G network system was simulated and evaluated through the VANET simulation platform, i.e., SUMO and NS-3. Our simulation results showed the feasibility and effectiveness of the proposed models. Due to the limitation of objective factors, there is still a big gap between the model and the scene environment in setting our approach and the real scene, which will be further considered in future research work.
ANAND NAYYAR received the Ph.D. degree in computer science from Desh Bhagat University, in 2017, in the area of wireless sensor networks, swarm intelligence, and network simulation. He is currently working with the School of Computer Science, Duy Tan University, Da Nang, Vietnam, as an Assistant Professor, a Scientist, the Vice-Chairperson (Research), and the Director of the IoT and Intelligent Systems Laboratory. He is a certified professional with more than 125 professional certificates from CISCO, Microsoft, Amazon, EC-Council, Oracle, Google, Beingcert, EXIN, GAQM, Cyberoam, and many more. He has published more than 150 research papers in various high-quality ISI-SCI/SCIE/SSCI impact factor journals cum Scopus/ESCI indexed journals; more than 60 papers in international conferences indexed with Springer, IEEE Xplore, and ACM Digital Library; and more than 40 book chapters in various Scopus and Web of Science indexed books with Springer, CRC Press, Wiley, IET, and Elsevier with Citations: 8000, H-index: 47, and i-index: 165. He has 18 Australian patents, seven Indian design cum utility patents, three Indian copyrights, two Canadian copyrights, three German patents, two Japanese patents, and one U.S. patent to his credit in the area of wireless communications, artificial intelligence, healthcare informatics, digital twins, cloud computing, the IoT, and image processing. He has authored/coauthored cum edited more than 40 books of computer science. He has reviewed more than 2000 articles for various Web of Science indexed journals. His current research interests include wireless sensor networks, the IoT, swarm intelligence, cloud computing, artificial intelligence, drones, blockchain, cyber security, network simulation, and wireless communications. He is a member of more than 50 associations as a Senior Member and a Life Member of ACM. He was associated with more than 500 international conferences as a program committee/chair/advisory board/review board member. He was awarded 38 awards for the Teaching and Research-Young Scientist, the Best Scientist, the Young Researcher Award, the Outstanding Researcher Award, the Excellence in Teaching, and many more. He is acting as an Associate  MOHD ASIF SHAH is currently working as an Associate Professor with Bakhtar University, Kabul, Afghanistan. He has with more than ten years of demonstrated Teaching Experience in the field of research, teaching. Further, he has the ability and capability to analyze both quantitative and qualitative data through valid software's like R, STATA, SPSS, MS Excel using different statistical methods and techniques. He has published more than 30 research papers which are indexed in Scopus and Web of Science with more than 50 citations. VOLUME 11, 2023