Abstract
Vehicular ad hoc networks (VANETs) are an eminent area of intelligent transportation systems (ITS) which includes vehicle tracking, positioning, and emergency warnings. For protective applications, vehicle localization in the urban area is a significant problem. Global Positioning Systems (GPS) are one of the many solutions that have been offered, although they do not offer accuracy. To locate a target vehicle accurately, a unique approach called received signal strength- (RSS-) based localization scheme has been proposed. By detecting signals within its range, it establishes communication through roadside units (RSUs) and determines the typical RSS. Following the discovery of the RSS, the RSS-based localization algorithm helped in determining the precise position of the vehicle. The high signal-to-noise ratio of the proposed algorithm, which is derived from neighbouring RSUs, is its key component. The vehicle’s Cramer Rao lower bound (CRLB) is examined after its position is discovered. Various experimentations considering dynamic vehicles ranging from 2 to 100, localization error, ranging error, and throughput have been conducted which demonstrates that the proposed algorithm (RSS) has better results (87%, 80%, and 120) than other well-established least squares (LS) and weighted least squares methods (WLS).
1. Introduction
VANET is a comprehensive and much more exact sort of advancement network. It incorporates several more technologies as well as communication among vehicles to vehicles (V2V) and vehicles with infrastructure (V2I). Although VANETs have numerous other uses, one of the most important is the efficient ubiquity of RSU and car connection, which is further connected to workplaces, homes, and other locations. As a result of effective V2V and V2I communication in VANETs, a wide range of corporate traffic monitoring system applications, such as ITS, are made possible [1]. There are several goals for users and vehicles provided by the ITS in VANETs. It provides drivers and passengers with safety and comfort when driving, which significantly reduces erratic movement in daily life and saves money and lives. Because all the drivers and passengers can easily communicate with one another, connect to the Internet and various networks using various types of gateways from their vehicles, and access a wide range of services and information, VANETs are particularly helpful for traffic management systems. Location information is key to substation applications that include location-based services, navigation, emergency response, and emergency support, thanks to the gateway found in VANET management [2]. Applications such as warning systems for car collisions, security cuisse management, dissemination of road info maps, automatic parking, and driverless vehicles are some of those that have been envisioned for this network and are currently possible in newly constructed vehicles. Another use of V2I and V2V communication is for the advanced driver assistance (ADAS) card structure. This is of particular concern due to the high usage, accuracy, and integrity burden with how the cars are positioned. The inexpensive or charge of the vehicle is typically partial when it comes to sensors; nevertheless, considering the specific ADAS constraints, the absent sensors are typically as effective as they can be [3]. The wireless device we used for home-based Wi-Fi installations for vehicle contexts is the 802.11p standard for V2V communication in VANETs. As seen in Figure 1, they can communicate directly and exchange knowledge without further discussion [4].
One of the most crucial methods for controlling a manually configured device’s location is localization, which is impossible for very large deployments or mobile systems, because of its significant material loss and unsuitability for interior and underground locations, GPS. Since ad hoc networks require the placement of high-density base stations, the local positioning system (LPS) hangs onto them.
The margin of the existing positioning system motivates new solutions for the network. Some specific nodes (anchors) identify the location of their global location, and the remaining nodes control or control their position by measuring the geographic information of their neighbours. This location is used in wireless multihop networks, known as corporate ad hoc localization networks or automatically located network nodes, and its location is determined by sharing its information [5, 6].
GPS is a very common technique for positioning systems for vehicle navigation systems. It produces very correct results in open and plane areas where communication is online, and at least four satellites are available for GPS. Localization gains information from the vehicle in this method navigating vehicle-automated vehicle Nemours and vehicle tracking; while in localization, GPS use the de facto standard of localization, and there are two error for localization: one is the global, and the other is local factor error localization algorithm that is executed in three stages like measurement updates and communication updates, and the other one is position estimation, and the planed localization scheme is to depend on RSU [7, 8].
Localization approaches are utilized for finding accurate, real-time location discovery of vehicles, such as altitude, longitude, and latitude, which is a key component in application success. In VANETs, vehicles could determine the position of the vehicle using a sensor such as GPS and an odometer. One of the most crucial elements in the calculation and verification is the localization as shown in Figure 2 of the positioning of the vehicle. The GPS is not the best localization approach for VANETs because it has an accuracy range of just 20 to 30 meters and cannot be utilized indoors, underground, or in populated areas because there is no direct link to the satellite. Due to this, we refer to this technique by combining GPS with additional positioning technologies like cruise control, cellular localization, image and video processing, etc. A system known as data fusion technology combines all of these positioning techniques. The data fusion model broadens the estimation of positions in numerous ways by using a range of techniques, such as the Kalman broadcast filter, particle filter, Bayesian filter, and belief theory [9]. Data fusion also makes systems more accurate and makes effective information. Data fusion technology, as opposed to weedy systems, calculates the position live and recovers from failure, allowing us to carry out independent real-time activities in VANETs [10]. The main key for VANET positioning is to prepare each node of the vehicle through a GPS receiver. This is the actual sensible solution as the GPS receiver can be simply fixed in the vehicle which has been provided by this technology. But with the development of VANET networks in key areas and the increasing reliance on positioning systems, GPS along with further satellite positioning systems (such as Galileo and GLONASS) are beginning to encounter difficulties such as lack of availability and insufficient availability of critical applications [11].
Now, in order to use and correctly calculate your location, the GPS receiver must be able to access at minimum three satellite indications for 2D locating and at minimum four satellite signals to evaluate the 3D position. At first instance, this does not appear to be the main problem because there are often between 2 and 11 significant satellites. Though, the problem is that these signals may be destroyed or congested by problems such as buildings, rocks, and the like. This results in inaccurate or hard-to-reach locations in dark urban scenes like tunnels, covered parking lots, jungles, and any other internal underground or underwater situation [12]. The main contributions of this study are as follows: (i)This paper’s contribution is the novel closed-form localization approach, and it suggests for VANETs, which enables us to estimate the positions of several vehicles(ii)The anchor RSS measurements at the RSUs are the key source of data for the proposed technique. It takes into account a completely connected VANET, which is a typical assumption for network localization [13](iii)The CRLB, which is the lower bound on the error variance, is also calculated for the suggested closed-form solution [14](iv)The theoretical mean square error for the suggested closed-form solution is also calculated. The suggested method clearly outperforms a number of other well-known localization methods, according to the numerical outcomes
Challenges in localization of VANETs are the following: (i)Line of sight (LOS) and non-LOS (NLOS) issues(ii)Dynamic changes in the network topology(iii)Inaccuracies in GPS measurements(iv)Interference and multipath effects(v)Heterogeneity of communication technologies(vi)Limited computational and energy resources in vehicles
Motivations for RSS-based localization of VANETs are the following: (i)Improved road safety through traffic and accident management(ii)Enhanced driver convenience through in-car services(iii)Intelligent traffic management and congestion reduction(iv)Improved fuel efficiency through optimized route planning
2. Related Work on VANET Localization
Due to its relatively precise results in open communication regions outside, GPS is a particularly common position and navigation system in VANETs. The VANET employs three processes for each positioning technique: measuring the distance, communicating, and estimating the position [15].
2.1. Evaluation of Centroid Localization CL and Weighted Localization WL
To assist the vehicle in determining its location, the authors of [16] offer two novel positioning strategies: a biased position (WL) using the signal-to-noise ratio (SNR) obtained from the exchange messages and a weighted position next to the vehicle (WLD) using the distance SNR. Their approach, which weights each surrounding vehicle coordinate built on SNR values and distances, develops the idea of LC centroid localization.
The importance of these cars having Internet access and providing drivers and passengers with a choice of ways to communicate over bridges is emphasized in [17]. The present location protocol for the VANET believed that locating security services offered our VANET a communication system and SEGAL fusion protection, which they named “security services,” rather than focusing on finding security gateway detection [18]. The strategy used combines information that is already available and works together to improve the accuracy of vehicle position data based on VANET location. In addition, it uses data fusion, several positioning techniques, and V2V communication. A full examination of the VANET network locations is given in [17], having an emphasis on the creation of INS and GNSS (global and satellite navigation). The algorithm in this book creates a camera-like image by directly transforming the given map function. The image prediction estimate is calculated using the logic value obtained from the structure sensor. Location accuracy is a serious challenge, according to the authors of [13] They assert that IoT (Internet of Things) position services are among the best. The time-of-arrival measured is raised by the author using a new subspace approach to linearize the nonlinear cost role used in typical location (TOA) situations. The magnitude detection characteristics and scalar creative matrix construction of this approach enable a closed-form answer that has been discovered to be resilient to severe measurement noise. The authors of [14, 19] detail a unique automobile network for VANETs. Nowadays that these vehicles are interconnected with one another and the road infrastructure, they can be used for a wide range of purposes, like driving Internet access and enhancing transportation security. Most protocols, algorithms, and applications in these networks require that nodes are located in real time. This assumption is based on the fact that GPS receivers can be easily placed in cars—some of which currently come equipped with such technology—and have been proven to work by studying and analyzing the location requirements of important VANET applications.
In [20], the authors proposed a new location solution that would benefit VANETs without GPS. The solutions they use include various positioning methods, data fusion, and data exchange between vehicles to combine existing data and work together to restore the accuracy of vehicle location information. The structure uses round-trip time (RTT) to measure the distance from the store and rely on improving the location of shippers and neighbours. They use the advanced Kalman filter (EKF) to measure the confidence that each vehicle is in its current position. In [21], the authors suggest using the round-trip time to calculate the distance among vehicles and to update the position of the vehicle in conjunction with the measured values of the inertial sensor which should also be relative to the new neighbour colocation plan. When calculating new positions, they use advanced Kalman EKF filters to limit the effects of errors on the sensor and adjacent locations. Compared with the existing colocation technology, their proposed colocation system does not rely on GPS updates in nearby areas making it more appropriate for urban canyons and tunnels. In addition, these characteristics include various speeds, sensor measurement errors, beginning positions, and automobile densities apply to a cooperative strategy that employs the ability to assess RTT. The authors [22, 23] pointed out that using the signal level indicator obtained by RSSI to estimate the distance to the location of the wireless sensor network is a daunting task. The anchor node at the inspection site accepts an assessment of the level of noise present in the environment. These anchor nodes establish a relationship between the actual distance between them and the calculated RSSI measurement distance. When considering information about the conditions of a link node, four different methods are provided for applying constant smoothing to the measured RSSI. The compression ratio is calculated for each procedure built on the area in which the unknown node and the communication node are situated and the number of communication nodes associated with the unknown node. In [24], the authors suggested a vehicle network location tracking method that allows widely used vehicles to detect sensors used by neighbours and smartphones at a lower cost and simulate location information to guess them. They use real data from urban scenes and four communication tools to test concepts; collect GPS, RSSI, and inertial data; and use existing road maps. The presented results assess the quality of the various combinations of these information sources. Therefore, we provided a mean location error of 7.7 meters during the trip including the main low GPS coverage area. In [25], the authors introduced a new type of security-sensitive routing technology called VANSec. The presented model is more resistant to various types of attacks and counteracts attempts to penetrate the network through malicious nodes. This is mainly based on the trust management method. The purpose of this architecture is to identify malicious data and fake nodes. From the point of view of TCE validity error, EED end-to-end delay, ALD averaging time, and NRO time-normalized routing, VANSec simulation results are compared with prior art known as trust and LT. From the point of view of TCE, compared with LT and trust, the efficiency of VANSec is 11.6% and 7.3%, respectively. A comparison of the EED shows that VANSec is 57.6% higher than the effectiveness of trust and 5% higher, 2% higher than LT efficiency. In [26], the authors predicted a new colocation scheme that uses RTT to calculate vehicle distance and updates the position of the vehicle in conjunction with the reduced RISE measurements of the inertial sensor system vehicle. The positioning of the car is also very good. We use an advanced Kalman EKF filter to calculate the effect of the error on the sensor and the end-to-end position of the new position. Considering that some vehicles can use GPS in a short time, their system has also been expanded. The ultimate goal of this effort is to efficiently achieve and organize this mixed sensor technology in a reliable accurate and powerful navigation system. Changed scenarios have been implemented at different speeds and environmental densities. GPS informs with different ratios and error tolerances were also presented to test the reliability of the proposed solution [27]. In [28], the authors solve this problem by studying and evaluating the position estimates as a normal way to progress VANET applications. They explore the proposed methods for locating tracking and calculating time sequence methods that can be used to estimate the future location of the vehicle. They also highlighted their strengths and weaknesses during logical discussions and presented possible scenarios for their use in VANET. In [29], the authors propose a method for optimally locating downstream vehicle units located on the roadside in an ad hoc network of the VANET networks. The layout of the roadside unit can significantly affect the presentation of the positioning algorithm. The planned approach is to search for the best RSU coverage location when using the minimum number of RSUs to ensure optimal positioning accuracy. [30] They used geometric dilution to evaluate the accurate measurement of the GDOP provided by the position derive from the expression of the GDOP to force the feed and estimate the mixing parameters from the received signal, respectively. The planned method has two steps. First, the optimal element model is obtained and applied to form a hedge investment. Second, based on the root cause for the K layer, the optimal position of the kth coverage plan is determined by reducing the average GDOP method using the APSO optimization algorithm, asynchronous group subdivision [31].
In [32], the authors propose a Bluetooth gateway method for detecting significant fluctuations in the RSSI of a Bluetooth node and uploading it to a cloud server which corrects RSSI in real time. The location of the terminal collects Bluetooth nodes around the RSSI and then adjusts the real-time RSSI fluctuation information according to the server. Adjusted RSSI can be used to calculate and reduce positioning errors. In addition, due to the difficult electromagnetic situation in the room, the long-term loss of the logarithmic model of the model is difficult to accurately compare the RSSI area. Therefore, the optimized network uses PSO-BPNN optimized network to build the RSSI model to reduce the error [33]. In [34], the authors demonstrate that the V2V and V2I Het Net environments have Wi-Fi DSRC and LTE technologies to ensure the best use of communication, minimizing the possibility of return connections for infrastructure needs, considering the requirements of the connected vehicles. Current CVT architectures such as CVRIA indicate that CVT applications can be used to maximize the potential of continuous V2V and V2I communications. [35] The performance of CVT applications using Het Net depends on the possibility of several wireless options’ acceptable communication latency data security and fast and reliable messaging. For a wide variety of CVT applications, namely, security, mobility, and environments, feasible communication options should contain V2V and V2I to optimally use the HET network to lose connectivity to the data network. This will require transferring the Het Net environment from the wireless network to another network. For protected applications, collecting traffic data does not require a very low delay of 200 milliseconds. [36] Table 1 is the summarization of all literature.
In [41], RMRPTS, a dependable clustering-based multilevel routing protocol, has been presented in VANETs. Clustering-based multilevel routing will provide the potential for self-organization and route maintenance even if this topology is continually changing. Additionally, it will use tabu search to address the issue of generating a trap in the local optimum. The suggested protocol, at the first level, is an extension of the AODV routing protocol that has been enhanced with fuzzy logic to produce dependable routing between cluster members. At a higher level, tabu search has been employed for routing between cluster heads and destinations. When solving hybrid optimization issues, the tabu search metaheuristic enhanced learning approach employs a cost function to choose one of several potential solutions. In [42], they introduce f-IVG, a fuzzy logic-based enhancement of the intervehicle geocast routing system. In the suggested approach, they employ fuzzy logic to select a link with the greatest link expiration time (LET) and probability adapts density (PAD) to prevent packet rebroadcasting. Effective parameters like distance, direction, and velocity are employed in the proposed routing protocol to determine the next-hop node. The dedicated short range communication (DSRC) channel is separated into three different sorts of priorities for beacon messages in the manner proposed in [43]. Each class has a dynamic receiver that will receive the message. To avoid broadcasting redundant information and make the best use of available bandwidth, a technique called considering the intermediate condition between the source node and the transmitter node is used. Hardware or a centralized controller is not necessary for the proposed solution. The proposed method is contrasted with ASPBT and adaptive data dissemination protocol (AddP) methodologies in order to assess the efficacy in [44]. They describe an authentication-capable fuzzy logic-based routing system for automotive ad hoc networks. The clustering phase, the routing phase between cluster head nodes, and the authentication phase are the three phases of the suggested routing approach. Vehicles are clustered in the first stage utilizing an effective method. We distinguish between immediate and regular data packets in the suggested technique. Phase 2 describes several route discovery procedures for various data packets. It should be noted that there are two categories for data packet types: simple and secure. There is no authentication mechanism for simple data packets. Secure data packets, on the other hand, employ a message authentication code (MAC) and symmetric key cryptography-based authentication system. In [45], the study proposed a routing approach for VANETs employing UAV nodes. Three phases make up the suggested method. The first stage focuses on the transportation of cars. Every time a vehicle needs to transmit a packet during this phase, it chooses the closest neighbouring vehicle to send the packet to base on distance, speed differential, and direction difference. The routing of operations between the vehicle and the UAV is the focus of the second phase. The car sends information to the UAV during this phase, and the roadside unit chooses the best one based on the distance and remaining energy. A greedy routing approach between the UAVs based on distance and energy is provided in the third phase, which is specifically dedicated to UAV routing activities. In [46], a three-dimensional (3D) evidence theory-based, opportunistic routing protocol, 3DEOR, is proposed in this research to overcome the aforementioned issues via a novel hybrid criterion known as PAL. The PAL criterion incorporates three metrics: level to increase network connection, packet advancement appropriate to 3D environments to reduce hop counts, and packet delivery probability to make up for unreliable dissemination environments. To choose the best relay node, order the members of the relay set according to the new criteria. The evidence theory can be a good option for merging the three metrics of the PAL criterion in situations when there is uncertainty.
3. Problem Formulation and System Model
3.1. System Modeling Localization Based on RSS
Ability to see a selection of V-goal automobile Ti denotes the actual location of each vehicle as a path, , where . If there are M anchor facts along the UAR, then , e2, etc., eM, and at some point. is the position’s kth anchor point, and M V, often M 3, is working to evaluate the 2D position [15]. The unidentified vehicle figure and the jth anchor have a route loss of where is the distance between the th anchor and its destination vehicle, () is the loss of path at , is the follower loss of path, and is the ranging error. The connection between path loss and RSS ZT is the transmission with the intended vehicle power, and is [16] . The received anchor signal strength is
From the calculation, it can be seen that () is the predictable signal power at a positional reserve tracked by path losses as assumed by
Combining (1), (2), and (3), we get (), where () is the predictable power at the th anchor from the RSS measurements [19]. or where ; the expected location of a marked vehicle is which is over on the RSS-ranging of the goal vehicle to the anchors with identified location , [21] where ; 2; M. Distance amid ith target vehicle and jth anchor is where “” is the relics of the ordinary operator.
As a result, the predicted noise range between an anchor vehicle ej and a target vehicle Ti is
Finding the precise position of the odd goal vehicle [21], which is aimed by employing the noisy RSS-based range, i.e., , is the most thrilling job in VANET localization.
3.2. RSS-Based Closed Method Explanation
RSS realistic range assessment from (1) to (6) is
Put on the square to both sides, and we get
For easiness, consider as and as , we obtain
In matrix form
or
Explaining (12), we get
Somewhere, is defined as persistent and is error amendment, [40] then,
The appreciated position of the goal vehicle remains
Somewhere, and
Here is unsigned so arrogant that , we get
To get
After repositioning the above equation, we now have a quadratic equation which is given by
The performance for locating the target vehicle is accomplished over the root mean square error (RMSE) [22] Consequently, the predictable position’s RMSE for a marked vehicle is defined as.
Given that the CRLB is computationally similar to the inverse matrix of the fisher information matrix (FIM), discovered using the probability density function, it serves as a bound for the estimation error covariance matrix for any unbiased estimation of an unknown limit (PDF). Through the use of a statistical signal processing model and an analysis of the real-world situation, the CRLB can be determined and calculated.
4. Research Methodology
This work proposes a new RSS-based positioning method for VANETs that can estimate the location of several cars. The proposed technique is based mainly on the RSS measurement of the anchor of the restricted unit. A completely connected VANET network is considered a mutual supposition of the location of the network. For the proposed solution, a CRLB is also derived, which corresponds to the lower limit of the error variance. Flowchart of the proposed method has been depicted in Figure 3. In addition, the theoretical square error of the proposed solution is calculated in closed form.
This experimental method is effective for conducting investigations. The positioning algorithm applies to the destination vehicles of different networks. Monte Carlo models were done in each state to prove the performance of the proposed technique. In this area, we are considering different conditions to calculate the performance of the best position algorithm for VANETs. The positioning algorithm applies to the target vehicles of different networks with the design focus on four different simulation configurations, as described in the following: (i)The localization error, ranging error, and throughput for two target vehicles(ii)The localization error, ranging error, and throughput for five target vehicles
4.1. Flow Chart Diagram
First of all, select the target vehicle. The target vehicle is the vehicle in which we found the position. After that, we select the anchors; the special kinds of vehicles call AP or work like an RSU; we use three anchor vehicles for a 2D positioning system. After that, we start the initialization of the anchor’s vehicles in initialization, and we chose the anchors that are suitable for the target vehicle or near the target vehicle.
After initialization of the anchors, we start to compute the RSSI within the range of the entire selected anchor vehicles that will be computed if RSSI is within range; then, the liner solution by the least square will be calculated. The least square technique is a form of mathematical regression inspection that invents the line of finest fit for a dataset, providing a graphic demo of the association between the data facts. Along with the weighted liner square solution WLS in this, all the weights are added to the least square, else, initialization will start again.
Now, if the signal is powered from -30 dBm to -70 dBm, the RSSI power is in dB or dBm because the signal strength is stated in decibels dB. Due to the reduction in low power levels and free space, the RSSI value is expressed as a negative number. The smaller the number, the weaker the signal, and the closer the number is to zero, the stronger the signal. Then, derivation of The CRLB lower limit of Cramer Rao can also be used to limit the change in the offset estimate for a given deviation. In some cases, the offset method can result in a mean-variance and a squared error, which are both lower than the unbiased CRLB. A two-dimensional positioning and evaluation of the target vehicle will then be performed; otherwise, the limiting amplifier will be used to increase the power. Limiting amplifiers are widely used in instant frequency and phase receivers. The most important requirement for the characteristics of a limiting amplifier is to minimize output power fluctuations and ensure constant output power over a wide dynamic input range. If the signal is amplified, it is sent to the next step; otherwise, it is sent to the RSSI for the three binding vehicles.
5. Simulation, Results, and Discussion
The proposed study examines several scenarios to measure the effectiveness of the improved positioning algorithm suggested for VANETs. The positioning algorithms are used to acquire the target vehicles in various conditions. The Gaussian noise vector was settled and aimed at two distinct simulation parameters under the assumption that the amount of error is zero. The industry-standard algorithm for error variance is Cramer Rao lower bound. If the outcome is close to the CRLB, it is regarded as an excellent result.
5.1. Ranging Error for Two Target Vehicles
This simulation scenario is carried out with 4 anchors and 2 target vehicles at the RSUs, where the exact positions of each target vehicle and anchor are illustrated in Figure 4. The ranging error for this scenario is designated by 10 log10 () as shown in Table 2. The RSS-based localization algorithm beats WLS and LS, respectively. And the result meets the standard CRLB and is greater than another algorithm.
5.2. Localization Error for Two Target Vehicles
This section presents a simulation scenario in which we have 4 anchors and 2 target vehicles at the RSUs, where the exact positions of each anchor and target vehicle are given in Figure 5. The localization mistake for this situation is shown by 10.log10 (RMSE). The robustness of the RSS-based localization algorithm shows better results as compared to WLS and LS, and its conformance to the standard CRLB. Table 3 shows that the algorithm in row two is 18% and 33% that is greater than the other algorithm.
5.3. Throughput for Two Target Vehicles
Calculated throughput for two target cars demonstrates that throughput rises when range and localization errors are reduced. Table 4 second row has a result that is 30% better than the other two rows, and Figure 6 demonstrates how better results are achieved; LS and WLS are not as effective as our RSS-based localization algorithm.
5.4. Ranging Error for Five Target Vehicles
In this simulation scenario, there are 5 target vehicles and 4 anchors at the RSUs. Table 5 shows that the algorithm in row two is 17% and 30% that is greater than the other algorithm. Where the exact positions of each target vehicle and anchors are given in Figure 7, the ranging error for this situation is denoted by10 log10. It is undeniable that the RSS-based localization method outperforms WLS and LS, respectively. The result is very similar to CRLB standards.
5.4.1. Localization Error for Five Target Vehicles
Five target vehicles and four anchors are considered in this simulation for localization at the RSUs in the designated zone, as shown in Figure 8. This example displays the actual locations of separate anchors at the RSU as well as the actual location of the targeted vehicle, demonstrating how accurately the RSS-based localization algorithm performs concerning the localization error. Table 6 compares the results and demonstrates that RSS-based localization is superior to other algorithms by 18% and 33%, respectively. The effectiveness of RSS-based localization for fixed positions of unidentified target vehicles is in comparison to LSs and WLSs. In the localization error statement for 10 log10 (RMSE), WLS and LS methods are inferior to RSS-based localization in rapports of performance.
5.4.2. Throughput for five target vehicles
Calculated throughput for five target cars demonstrates that throughput rises as range and localization errors are reduced. Table 7 demonstrates that the second row’s performance is 30% higher than that of the other two rows, and Figure 9 shows that our RSS-based localization algorithm is superior to WLS and LS in terms of performance.
5.4.3. Cost Analysis of Developed Algorithm
The proposed algorithm’s computational cost has been determined in this section. Let represent the volume of automobiles in the neighborhood RSUs. The suggested approach has an computational complexity, where is the length of the measurement vector. The matrix produced from Equation (8) was multiplied, causing the term, while other matrix multiplications in the Kalman update caused the term. The distance measured between vehicles and is averaged, making effectively equal to . If all measurements are used, will be equal to (). However, is modest in the vast majority of real-world circumstances. When there are up to 5 vehicles with additional RSU and a maximum of 4 anchor vehicles, and when the measurements are shared at a rate of once per second, the suggested technique implemented in MATLAB runs in less than 450 ms on a 6 GHz system. Only the point in the process where the cars share distance and velocity estimates with their neighbors involves communication overhead. The communication overhead is since measurements are being shared if we assume that nonoverlapping RSUs do not share the same channel due to range restrictions. In practice, each vehicle sends a packet with a size of that includes all the observed location estimation and fixed packet size for vehicular speed.
6. Conclusion
This study proposed to consider the fact that the vehicle captures signals within its estimates of the RSS (typical received signal power) for each power source and operates within its range. Through the use of an RSS-based positioning approach, the average RSS measurement is sent that pinpoints the vehicle’s position. The system uses only the RSS signals from the closest RSU because they have a larger signal-to-noise ratio and allow more accurate positioning. The RSS-based localization method was tested using Cramer Rao’s lowest bound. Many simulations have been performed to demonstrate that compared to the weighted least squares and least squares approaches, and the anticipated RSS-based solution is better. We compare the results based on range, localization, and throughput for each of the four situations in the simulation environment single for one vehicle, two for two vehicles, five for five vehicles, and ten for ten vehicles. We get to the conclusion that our RSS-based localization outperforms WLS and LS, and the output is comparable to traditional CRLB. Our best recommendation to fix these problems is a modern RSS-based localization approach for VANETs.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIT) (No. 2021R1F1A1063319).