Leveraging Big Data Analytics for Intelligent Transportation Systems: Optimize the Internet of Vehicles Data Structure and Modeling

doi:10.21203/rs.3.rs-1640856/v1

Download PDF

Research Article

Leveraging Big Data Analytics for Intelligent Transportation Systems: Optimize the Internet of Vehicles Data Structure and Modeling

https://doi.org/10.21203/rs.3.rs-1640856/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

An intelligent transportation system is an efficient and modern system to deal with big data, which is pertinent for a future smart city. It is the inherent challenge of data processing in classic vehicular systems. Therefore, we proposed Big data to optimize location-based operability and safety performances using federated sensor data. The data preprocessing and feature extraction process includes vehicle mobility, multi-source data acquisition, distributed computation, and multi-path data transmission in an analytical model to enhance performance and safety. The proposed approach is capable and scalable to manage large-scale sensor data from its source and line of information flow. The findings revealed the imperfect, complex, and challenging data transformation into actionable, safe, and usable information for the intelligent transformation system modernization and healthy digital ecosystems. Big data analytics for sensor data is complete and informative for the learning process of the physical systems in various situations and localizations. The experiment showed that leveraging big data technology is a systematic approach to enhancing safe and optimal vehicular system applications. It provides access to shared resources through connected devices and reduces costs using the response time and data structuring to the system stability and to accommodate future modifications in sensor technology.

Big data

intelligent system

intelligent cloud system

digital ecosystems

smart environment

Technology for human lives is a way to bring enormous benefits to a dynamic and sophisticated connected future. Big data analytics is leveraging sensor technologies and the Internet of Vehicles (IoV) to gain a cost-effective Intelligent Transportation System (ITS) with advanced and modern infrastructure. ITS provides sensor facilities to ensure intelligent safety and internal system operability. The Internet of Vehicles is intelligent and computer-based wheels, in which the sensor and wearable devices notify the objects’ movement facts. Applying big data analytics is to ‘earn the system ecosystems as the user preferences to take the proper actions and path as per the enabled sensor technology of ITS. The ITS sensor devices recording data are pertinent to big data analytics and recommendations for vehicular system safety and service performance improvements (Wei-Fan and Albert, 2013). Big data analytics is an advanced methodological computation to handle and analyze the ever-increasing sensor data generated from sensor devices that are independent and connected appliances of the IoV systems.

The Internet of vehicles encompasses various sensing and communication functioning features that rely on advanced tools to learn, predict, and make decisions to improve car safety, mobility, and efficiency through the digital ecosystems. Intelligent and interconnected systems are complex, and generating large-scale sensor data is sensible for new and powerful technologies to enable more complex data structure and analytics (Neha and Rakesh, 2012). The ever-increasing IoV sensor data are the characteristics of big data that brought precedent opportunities for unified management of the ITS and the smart devices. The sensor data is required to visualize the integrity and functionality of the intelligent system for the road vehicles’ safety (Nomusa et al., 2012; Yang et al., 2007).

The challenge for the Internet of Vehicle sensor data is becoming multiphase of the intelligent systems and analytics as to handling, transforming, and analyzing the sensor data deluge. The problem is how the transport industry can be advanced, safe, and efficient. How the tools and technologies are capable and scalable to the analytics of ever-increasing vehicular data? In this paper, I use real-time traffic data to focus on leveraging big data analytics for IoV intelligent facilities' sensor data structure and models. The proposed approach is to attain automated and reliable ITS systems operability for multi-use and management of the ever-increasing sensor technologies (Kashif et al., 2018; Gebeyehu et al., 2014; Bo-Wu, 2012). It is an augmenting big data for the fundamental of IoV intelligent cloud revolutions to optimize ITS performance (Huseni et al., 2013) and the knowledge-based decision-making process. Therefore, the main contributions to the paper are: Exploring the potential of big data analytics-based digital ecosystems for sensor data structuring and analytics.

Identifying the research issues and directions for IoVs and how big data analytics can support resolving such challenges:
Searching intelligence system applications for sensor car safety, security, efficiency, and convenience and the importance of big data analytics in such applications;
Providing an impetus for big data or artificial intelligence to investigate new exciting Big data-based ITS services;
Improving the situations of modern cities as the modernity of life of people and safety to deliver solutions to the most common traffic challenges in an empirical fashion.

The rest of the paper is organized as section 2 introduction for the paper contents and flow. In section 2, the related works of the Internet of Vehicles and the recent state-of-the-art discretion, including comparative analysis. Section 3 provides the concept and architecture of big data and the Internet of Vehicle integrity that focuses on big data for the Internet of Vehicle systems. The methodology and proposed approach have presented in section 4. Section 5 is the experiment, analysis, and discussion. Then, the conclusion and future work is summarized in section 6.

As the Internet of Vehicles, large-scale sensor data require more attention to structure and analytics processes that demand big data to formulate a robust digital ecosystem, which gains more attention in many fields, including industries and sensor technologies. Big data and intelligent systems can affect the current and future digital world. The data are characterized as large-scale in volume, versatility, and change at a quick pace. Data structuring is data categorization into a relational database to design proper data modeling for an efficient and effective decision outcome (Ahmed et al., 2018). Augmenting big data with the Internet of Vehicles is a methodological analysis of the integrity of smart devices that the sensor data modeling to predict and define the influence of future occurrences and the enhancement of the intelligent system standards (Indu, 2019). It is a modern approach to enhance the Internet of Vehicle’s technology in a safe and optimal functionality by gaining valuable information and knowledge of traffic management as needed. However, the success of big data in many research domains is still a gap and insufficient (Hany et al., 2018; Gang, 2015).

The intelligent car system is an integral sensor object working collaboratively with the digital environment (Mohamed et al., 2018). Internet of the vehicle is an emerging technology to create a more complex and sophisticated cloud system (Tejinder and Vijay, 2013). It empowers the intelligent systems sensor devices to make proper decisions that control the trajectory of cars in their direction (Amandeep and Sathans, 2019). The digital platform is persuasive to manage system operability for the cars’ safety within and secure network from the surrounding (Priscilla and Mathan, 2017) that the dramatic intelligent technology of the Internet of Things (IoT) connective ecosystem. The IoT (Shaik et al., 2019; Ovidiu and Peter, 2013; Bo-Wu, 2012; Shen-Bin et al., 2010) is efficacious with the ability to receive and send data to the endpoints via the traffic infrastructure and connected sensor devices (Park, 2018; Eugen et al., 2017). Big data analytics for large-scale data success, challenges, and future trends is summarized in Table 1.

Table 1

the summary of big data applications
Paper author(s) and date	Paper title	Objectives	Gap analysis
Bo-Wu (2012)	Internet-of-Vehicles as the technologies of Internet-of-Things,	To form vehicular networks to communicate with each other.	However, sensor vehicle connections, roadside and wearable devices, and their digital ecosystem didn’t address.
Gang Zeng, (2016)	Application of big data in Intelligent Traffic System	To solve the problems of traditional traffic data management systems.	This paper focused on traffic data management systems using an architecture of the big data platforms
Eugen et al. (2017)	Internet of Vehicles Functional Architectures -Comparative Critical Study	This paper aimed to offer novel advanced commercial and technical capabilities	This paper, as the title implies, didn’t discuss how the architecture enhances the capabilities of IoV
Priscilla and Mathan (2017)	Performance Improvement of a Secure Mechanism for Big Data Collection on the Large-Scale Internet of Vehicle Using MATLAB,	This paper was focused on performance improvement in wireless and mobile network service quality.	The authors’ focused on Big data collection in the Ad hoc network to address and evaluate performances in a very defined data structure.
Ahmed et al. (2018)	Big data technologies: A survey	It is a review of recent technologies for big data, and it aims to select and adopt different big data technologies.	Big data characteristics and challenges to present the application and opportunities for optimizations. It touches the component of the architecture but not as per other domains like intelligent systems.
Shaik et al., (2019)	A Review of Interoperability issues on the Internet of Vehicles (IoV)	The challenges of the interoperability of an intelligent vehicle system	However, this paper did not touch on the sensor device status, technology, compatibility, and even the solution for the digital ecosystem
This paper	Leveraging big data to optimize ITS digital Ecosystems for IoV Data Structure and Analytics	The research focused on big data for IoV incorporating intelligent systems to sensor data structure and analytics.	The study presents a comprehensive review of big data applications, data structure and analytics, and the intelligent transportation system status and future development.

The integrity of big data and the Internet of the vehicle is a matter of the wholeness of the Intelligence system (sensor device connectivity) and big data analytics (data structuring and analysis). The Internet of Vehicles is advanced sensor technology, and it enhances the ITSs performance. ITS is the content of sensor devices, systems, and data flow as the Internet of Vehicle unified environment and involves various tasks (Indu, 2019). Big Data support enhances the interoperability and performance of the intelligent system and the real-time processing of sensor data (Yang et al., 2018). The integration of sensor cars and Big Data visualizes the model shown in figure.1 and 2. The architecture revealed the sensing layer from smart devices connectivity and the data flow, which come from/to each smart device to provide valuable information (Hany et al., 2018). The data comes from the ITS’s sensor, and wearable devices provide information for users and vehicular systems to enhance services and safety by avoiding traffic problems. Big data supports the intelligent system by applying a machine-learning algorithm to develop a data model from a given vault and searching for a solution to real-time problems (Kayhan et al., 2019). The Internet of Vehicle’s digital ecosystem is an infrastructure of data communication between moving vehicles and fixed sensor equipment in a defined protocol and data transfer between the connected nodes, as shown in Fig. 1.

As shown in Fig. 1, the sensor data gained from connected sensor devices are semi-structured and unstructured, and I use a K-mean algorithm to label and feature selection for its subsequent analysis and searching for valuable information (Nomusa et al., 2012; Corcoba and Muñoz, 2011). Choosing K-mean clustering is to produce the best result in a repeated iterative experiment. K is an input for the number of clusters to determine the best performances of the data set using information criteria methods and the cluster centroids. Each vector or point similarity is measured from the cluster centroid with the minimum distance using Euclidean distance measurement. The elements of the class members are the collection of centroids (Ci) and the data point (x) assigned to a cluster as:

$$\mathop {\arg \hbox{min} dist}\limits_{{\mathop c\nolimits_{i} \in C}} \mathop {\left( {\mathop c\nolimits_{i} ,x} \right)}\nolimits^{2}$$

The cluster member similarities are recalculated using the mean of all points assigned to the cluster centroid as:

$$\mathop c\nolimits_{i} =\frac{1}{{\left| {\mathop s\nolimits_{i} } \right|}}\sum\limits_{{\mathop x\nolimits_{i} \in \mathop s\nolimits_{i} }} {\mathop x\nolimits_{i} }$$

These two steps (Eqs. 1 and 2) apply until no data points change clusters, and the outcomes are guaranteed to avoid the random starting centroids problems as the experiment is done. The cluster validation testing process uses sample sensor data patterns or purity measures to improve feature selections. The data set covered using Eq. 3 is as follows.

$$F\left( x \right)=F\left( {x,a} \right) \in F \to y=T\left( x \right)=\sum\limits_{{m=1}}^{M} {\mathop c\nolimits_{m} } I\left[ {x \in \mathop R\limits_{m} } \right]$$

Sensor data management is the principal concern behind big data machine learning analytics to structure and create the analytical models to permit the algorithms to learn continuously from generated data (Neha and Rakesh, 2012). It assures data modeling and testing processing for the ever-increasing massive and complex sensor data to handle the usability of the current ITS tools. The Big data broad applications include IoV systems enhancing to learn the intelligent circumstance and preferences of users’ safety and movement schedule. It optimizes through clustering methods that the connected vehicles and sensor devices pass and talk in the network regarding the real-time driving requirements (Eun-Kyu, Mario, et al., 2016).

To explore the training data set for the intelligent data to ensure a safe and structured movement within the roadway grid in a consistent and updated information flow. It includes mobile communication technologies (Gerard, 2012), real-time localization (Diansheng and Jeremy, 2009), ubiquitous computing, etc. (Shen-Bin et al., 2010). The model train to predict the moments of cars in a specified location. The training set learns the communication through the sensor devices standards and protocol (Wei and Albert, 2013), and the score criterion to judge the quality of fit of the model is defined as:

$$R\left( a \right)=\frac{1}{N}\sum\limits_{{i=1}}^{N} L \left( {\mathop y\nolimits_{i} ,F\left( {\mathop x\nolimits_{i} ,a} \right)} \right)$$

Where training set: $\mathop {\left\{ {\mathop y\nolimits_{i} ,\mathop x\nolimits_{{i1}} ,\mathop x\nolimits_{{i2}} ,...,\mathop x\nolimits_{{ik}} } \right\}}\nolimits_{1}^{N} =\left\{ {\mathop y\nolimits_{i} ,\mathop x\nolimits_{i} } \right\}$

K and N are the numbers of predictors and samples, respectively

Therefore, car’s poison and probability density in a fixed sample time t computed as:

$$\mathop p\nolimits_{i} (x)=\frac{{\mathop \lambda \nolimits^{x} }}{{x!}}\mathop e\nolimits^{{ - \lambda }} ,x>~0$$

Subject to x is the number of cars, and λ is a parameter that figures the number of vehicles arriving in sampling time t*. When λ = ⋅ α t*, α is the average arriving speed (veh/s) that depends on the different road sections and times. The data set includes numbers of car trajectories in a sampling time.

Big data analytics are applied to achieve ITS prediction performance in terms of traffic flow, travel time, roadside sensor device and vehicle circumstances, etc. It assures transportation benefits, including cost efficiency, ITS service, safety, and traffic database management (Jawdat, 2019; Shaik et al., 2019). The; paradigm of big data is an opportunity for intelligent systems enhancement in different aspects. (i) It enhances structure and analytics and integrates sensor data (Nomusa et al., 2012; Ying and Ling-Shu, 2011). (ii) It is competent to describe the Internet of Vehicle’s data in a spatial format associated with the location’s context and time series. (iii) To support the integrity of a federated Internet of Vehicle sensor objects’ fact, in semantics and inferred circumstances. (iv) The capability and scalability of big data analytics for the Internet of Vehicles are an advanced and dynamic technological fact (Raghava et al., 2016; Gang, 2015), which include: Address of federated IoV sensor objects’ data structuring to interact and maintain each other’s for car’s safe movements, and it is essential to

Data modeling and abstraction to filter the redundant data,
Data archive, index, scalability, and access control for IoV data,

Big data for IoV’s data structuring enhance the query language process in a multi-dimensional analysis, interoperability, and semantic intelligibility for heterogeneous data.

3.1 The paradigm architecture for big data and IoV’s system

Leveraging big data for an intelligent road system is the essence of support from one to another. Big data for sensor data acquisition, transmission, structuring, modeling, and analytics is a methodological approach to enhance the data value from the internal and external IoV’s environment. The architecture addresses the existing data processing and scales up to the global needs of the core of the data exploratory and management contexts (Jairo et al., 2021; Ying, 2013). The current IoV data analysis algorithm considers the sensor data and its dynamism in line with the ever-growing technologies, data deluge, which is the emerging paradigm of big data to apply for sensor data structuring and analytics. It performs in a multi-aspect, including (i) the need for the reliable design of popular algorithms makes integration with other techniques. (ii) Sequential design of the algorithms to parallel computing infrastructures to the optimal performance and sensor devices’ requirements. (iii) Enhancing technological development by integrating with intelligent cloud computing systems (Li et al., 2019; Gebeyehu et al., 2014). In parallel, the process gives secure information access and data trust to its sources–via vehicles and sensor devices (Mahmood et al., 2018; Wenchao et al., 2018) (Fig. 2).

Figure 2 is an architecture for a vehicular network that encompasses federated sensor data transformed into valuable information for the optimal intelligent system services and safety. It reveals the integrity of big data applications and IoV sensor data structure towards intelligent systems enhance performance. It is a verge of Big data as a core component of the future Internet of Vehicles to optimize large-scale data computation to extract valuable information (James, 2018; Diansheng and Jeremy, 2009). Big data for searching knowledge involves various steps, including feature extraction and selection to improve the intelligent system performances (Shen-Bin et al., 2010).

3.2 The success of the vehicular system and Big Data application

Big data for IoV data structuring and analytics is a notion that has a fundamental strength to the industrial efforts of devising intelligently connected vehicles for developing safe road environments. The paradigm application of big data is to enhance the ITS's digital ecosystem for IoV service privacy and safety (Longhua et al., 2017; Kate and Jason, 2014; Omer and Jules, 2013). Big data for IoV is a scientific and dynamic way to drive data patterns that gain from sensor devices. The data is processed using big data technologies from a data vault and event-driven approaches. The data received from multiple sensor devices, including vehicles, are a meaningful insight into the events, and bandwidth (Zhenyu et al., 2017; João, 2013), as shown in Fig. 3a. It applies the process information to optimize the response in the services triggered to the optimal vehicular system interface (Fig. 3b).

The paradigm of Big Data in an ITS applies to predict the Internet of Vehicles safety and optimal operability by the data analysis using machine learning techniques for safe and proper vehicle trajectories that include location and vehicle speeds (Elmustafa et al., 2021). This data flow in the network is pertinent to estimating the traffic circumstances along the road sectors using longitudinal and latitude coordination. The traffic flow, vehicle speeds, and direction are defined in spots of the specified time using (Shridevi and Manjunath, 2020) models (Eq. 7–9).

$q\left( A \right)=\frac{{d\left( A \right)}}{{\left| A \right|}}$ (7) $k\left( A \right)=\frac{{t\left( A \right)}}{{\left| A \right|}}$ (8) $v\left( A \right)=\frac{{d\left( A \right)}}{{t\left( A \right)}}$ (9)

Where q(A) is the flow of vehicles, k(A) is vehicle density, and v(A) is the vehicle speed. Annotations by d(A), t(A), and |A| represent the total distance traveled and vehicle time spent in that location and area covered by the Vehicles, respectively. Therefore, the integrity of big data for the Internet of Vehicles is to enhance road safety and vehicle trajectories time efficiency (Fei et al., 2017). As shown in Fig. 3, big data is associated with the fundamentals of the Internet of Vehicles for sensor data structuring and exploring vehicle movement safety (Noor et al., 2020). It ensures the Internet of Vehicle system security by learning and identifying the vehicles.

3.2 The Big data approach to exploit the hidden pattern of IoV’s

The big data for the vehicular system is the essence of effective intelligent technology for sequential vehicles’ information flow from one to another. A vehicle on the road will have signal information for its distance and road length, and the traffic controller can have the vehicle’s start and end distance in between (Sanghyun and Jonghwa, 2019; Young-Sik et al., 2019). The current signal information presents using Eq. 10, as follows:

$$\mathop T\nolimits_{M} =\mathop T\nolimits_{{current}} +\tau \times \frac{{\mathop R\nolimits_{{len}} - \mathop V\nolimits_{{dist}} }}{{\mathop R\nolimits_{{len}} }}$$

The Vehicle’s information message (M) transmitted at time t (T_M) is the current time t (C_urrent) and tau τ) of the unit time with the quotient of road length (R_len) minus Vehicle distance from its start (V_dist) to R_len. Big data for the traffic control system is methodological support for analyzing vehicle information as to their sequence or queues (Q) by Eq. 11. It allows the transformation of valuable information on the existing circumstance of road services. The roadside sensor devices or from the Vehicle work automatically to turn on the system and get notifications from or for the other sensor devices to remind the Vehicle of safety in its path and scenario at the car’s moment (Alesandra et al., 2018).

$$\mathop Q\nolimits_{{len}} =\left\{ {\begin{array}{*{20}{c}} {\mathop {\hbox{max} }\limits_{{\forall \mathop M\nolimits^{i} }} \left[ {\left( {\left\lceil {\frac{{\mathop V\nolimits_{{dist}}^{i} }}{{\mathop V\nolimits_{{len}} +\mathop V\nolimits_{{\operatorname{int} r - dist}} }}} \right\rceil +1} \right) \times k} \right],if\mathop V\nolimits_{{dist}}^{i} >0} \\ {1 \times k,otherwise} \end{array}} \right.$$

Where k is the number of lanes, Mⁱ is the i^− th message for Vehicle information, and V_{dist, i} is the distance of the Vehicle sending M_i. V_len is the length of the Vehicle, and V_dist is the distance between two vehicles.

Big data for the Internet of Vehicles are generic to concatenate the techniques of simulations, visualization, ontologies, and data-driven into one form of architecture (dashboard). It is an advanced approach to managing and exploring large-scale sensor data phenomena and technology that blends the issues and applications to allow organizations to get value from the ever-increasing data sets (Gang, 2015; Stefano et al., 2013). The analytic process in agents and the network or infrastructure learn and understand the real-world using wearable and sensor devices (Husen et al., 2013). It reveals the combination of the intelligent cloud system explosive, as shown in Fig. 4. Integration is a clear insight into the IoV technological revolution that passes the new ubiquitous connectivity, computing, and communication, which depends on technical innovations in several fields, from wireless sensors to nanotechnology (Razi, 2018; Ning et al., 2013).

3.3.1 Data structure and big data technology for The Internet of Vehicles data

Big data technology is a large-scale data structuring and handling using the techniques of data summarization and analytics. It can handle the massive data structures to extract value and new patterns through a dynamic analytic process. Big data for IoV’s data are sequential structuring approaches to know the speed and bytes associated with the data phenomena and technology in optimal performance. It allows a computer-based data model development and architecture (Ifeyinwa and Henry, 2019; Tasneem and Kamalrulnizam, 2018). Big data technology is a systematic approach to parametric data representation and correlation in the context of the intelligent systems shown in Fig. 4. It is vital to view the facts of the sensor device operation in a readable data model, as shown in Fig. 5.

The integrated architecture (Fig. 5) is a structuring of data in line with the system performance to a pre-hand solution to vehicular intelligent system functionality in the context of sensor nodes and smart objects (Abdul et al., 2020). Therefore, Hive-based structured data is pertinent to performing an integrated computational model in optimal disk spacing and processing time. The architecture is essential to demystify searching for a specific object within the group that processes the data items in any desired order (Fei, 2010).

3.3.2 Big data for the verge of IoV cloud system

Big data for IoV is technology-driven to a range of decision processes (Shridevi and Manjunath, 2020). The advantage includes enhancing the knowledge sharing for efficient system infrastructures, cost optimizations, and future predictions to a profound of intelligent cloud systems (Mohamed et al., 2018). It empowers system performance and service safety by performing an optimal precision of the Internet of Vehicle system functionality. Therefore, Big data analytics for the Internet of Vehicles is a technology for real-time sensor device connectivity to prove easy access, information flows security, and data analytics (Alesandra et al., 2018; Diebold, 2003).

The intelligent transportation industry is dynamic and incremental sensor technology, and the sensor data is pertinent to augment big data on the Internet of Vehicles. The data sets in terms of volume, velocity, variability, etc., revealed as big data are capable and scalable to handle challenges by considering the issues to propose various analytical tools, including Hadoop, Alteryx, distributed databases, complex event-processing frameworks, and others.

3.3.3 Big data for IoV data virtualization

Big data for IoV data visualization is the process of federating systems' relational databases, applications, file repositories, and website data sources integrity into single data access via the dashboard (Gebeyehu et al., 2019; Mingming et al., 2017). Big data applications need a cluster of nodes and communicate with defined protocols to solve large-scale data handling and transformation problems in a given network or connectivity (Kate and Jason, 2014). The distributed system works with the delegated agent or machine to use the federated large-scale sensor data (João, 2013). The data visualization framework allows for the data replication on a given Virtual Machine (VM) (Rateb et al., 2020; Hedi et al., 2019). The performance of various steps alters the system functionality depending on the frequency of data loading, replication, backups, etc. (Robert et al., 2014). As Zhou et al., 2020). IoV’s storage optimizations and access proposed correlation probability model demystify using Big data analytics and integrity for connected wearable devices. For example, a big data analytics tool such as Tableau uses data from connectivity than transforming or importing itself, as shown in Fig. 6.

Data visualization is an advanced methodological approach to sensor data virtualization through various steps, including virtualizing data sources into an abstract format. The physical and combined federated sources into virtual data and deploy to the web-based application via the networks, as shown in Fig. 6. The advantage of big data is to enhance a wide range of information access between devices.

3.3.4 Big data inside and outside the IoV’s data center

The Internet of Vehicle’s data structuring and analytics is a dynamic process, which evolves sensor data from different sources. The data captured in the central database and the data servers are sensor data, which comes from dedicated sensor devices in the system (Vijay, 2012) for information generating and flow across various sensor devices and the real-time process perform in connecting agents. Big data is a capable and advanced way of handling large-scale sensor data as data structuring and modeling (Robert et al., 2014). The data management system is a Big data application with its entire routing standards for a systematic enhancement of the Internet of Vehicle system performances. The sensor data handling is supported through big data, as showed it in Table 2.

Table 2

IOV intelligent data descriptions
Inside	Outside
The IoV’s data center is coming from applications and connected sensor devices -via the intelligent system, which includes: ♣ Clickstreams ♣ Application events ♣ Transaction records ♣ IoVlog/Monitoring data ♣ Virtual Machine (VM) Infrastructure monitoring	The data sources include phones, distributed sensors, and partners from outside of the center, which is massive, and endpoints and connections become the challenge, including: • Roadside and remote Telemetry • Sensors devices and nets • Smart Phones and wearable devices • RFID and barcode scanners • GPS Position Tracking • Point-of-Sale Terminals

3.3.5 IoV’s data modeling using big data analytics

Big data technology for IoV data modeling promotes novel approaches and technologies for structured and unstructured sensor data handling. An intelligent system sensor data is large and complex to manage and conduct a critical analysis using the classic approach (Wei and Albert, 2013). Dynamic processing support users’ queries repeatedly and improves the search results. Therefore, big data stand for putting analytics to search for valuable information or knowledge. The multi-layer platform agglomerated different tasks and protocols the integrating data, time, and other factors. The big data analytics approach provides advanced large-scale data processing (Eugen et al., 2017; Shen et al., 2010), as shown in Fig. 7, which is pertinent to analyzing events in IoV systems based on the given query and analysis in the layer.

AS the data processing in a multi-layer architecture of the Vehicle data, the big data technology allows being built data management and event processing of various event-based analyzing techniques for multiple applications, such as optimizations of managing data center and system performance. The data in IoV are always massive, heterogeneous, distributed, and time and position-related, which is challenging for classical or centralized mining architecture to extract valuable information (Mingming et al., 2017; Yang et al., 2017). Based on these facts, the advantage of big data over the traditional analysis tools presents the following fundamental issues. (i) Large-scale data of IoV are almost everywhere on its infrastructure and stations. (ii) The IoV data always needs preprocessing in real-time events. (iii) The complexity of the sensor data and sensor devices’ variables to analyze the data using traditional analytic tools. (iv) The structure of IoV is complex in architecture, which requires advanced sensor devices. (v) Moreover, the strategy of sending all data to the central nodes does not optimize the performance of the services.

Big data for the intelligent system is essential to carry out advanced and dynamic techniques to be scalable up large-scale data handling and analysis via federated data. The big data approach allows undertaking a preprocessing of the raw data in the distributed nodes (agents) (Paul et al., 2012), and users can send/receive information through distributed IoV’s system infrastructure. The proposed approach is a methodological augmentation of big data for ITS performance optimizations. Therefore, the big data approach is an agent-based Big Data Mining (BDM) mechanism of distributing the load among various sensor devices to improve the response time. And also it gives a clear insight into a situation where the devices are heavily loaded while others are idle or doing very little work (Shen et al., 2010). The data process and analysis are the core of the whole activity. An experiment is a methodological approach to combining tasks and navigation processes, as shown in the following pseudo-codes.

Algorithm 1

Step 1: defining the data sets for data analysis purposes,

Step 2: Data cleaning or preprocessing and filtering from its destination and real situations,

Step 3: Navigate the specific root system using its task,

Step 4: Assigning the devices, including the subelements to receive information or facts from various sensor objects,

Step 5: Saved data in the local data warehouse and the data model obtained from the data event filtering.

The Big data-based IoV’s data modeling is scalable to adopt large data sets from multiple sensor devices and data sources by accessing the infrastructure or the networks. It is the most advanced data analytics model to connect various sensor objects via the Internet. It becomes intelligent, context-awareness, and a long-range operable approach (Ovidiu and Peter, 2013). The model can also adopt the ubiquitous ways to access the digital system (Li et al., 2018; Kate and Jason, 2014) via the Internet, sensor devices such as RFID, WLAN/WiMAX, and others.

Nowadays, sensor data collected from various connected sensors and wearable devices get large and complex, which need advanced and proper structuring and handling in real-time for effective data classification techniques. Data modeling is pertinent to efficient data analysis to enable effective IoV operability and services. Big data support IoV’s system to provide safe and fast information flows between the connected devices for proper decision-making. It paves the way for improving the intelligent transportation system, which is pertinent to reducing costs, interferences, miscommunication, and challenges of having optimal energy and other factors. Leveraging big data is elucidating to combine both data phenomenon and technology to turn flat data into useful information. Visualization of the experiment results in terms of color, image, time, and location is adorable comprehensive and integrated into a single dashboard system to make a proper decision about the object moment (Omer and Jules, 2013). Integration is the process of searching for similarities to reveal the outcomes in a simple contextual analysis and prediction in mapping services of the vehicular digital ecosystems (Jimmy and Dmitriy, 2013; Shen et al., 2010).

Leveraging big data for IoV’s data set analysis is paramount in many forms. It includes heterogeneous, nonlinear, high-dimensional, and distributed data types (Diebold, 2003). The data sets are complex and evolve spatial, time series, and multimedia type data modes, which need a systematic representation and deep understanding to visualize the process as per the contexts and the data flow. The ITSs data structures and analyses are core tasks for vehicles and road safety in many aspects to benefit the system for optimal performance. Since a decision on a traffic jam and road safety requires visual evidence to investigate different events at a time, as shown in Fig. 8.

Figure 8 presents the raw data and the result visualization in a different sketch in a single dashboard. It rewards a far-reaching and meaningful insight into the best of the intelligent system operations and functionality. Big data remains in the system to be compatible with new technologies and data management processes. The IoV data handled and analyzed using big data is a systematic approach and ensures users need what they are looking for and enjoy a smooth discovery process. The dashboard-based result is an augmented application of big data technology for vehicular system decision-making to optimize intelligent systems and road service safety. The data mining/machine learning analytics tools support sensor objects' performance enhancement by retrieving and extracting valuable information. Big data being a solution is the content and context of the vehicular system (Elmustafa et al., 2021.

For Big sensor data handling and managing, the traditional data center is inefficient and challenging to understand the real-time traffic circumstance and situational analysis. A vehicular system is a solution for spacial sensor data, and easy to visualize their relationships and correlations (Paul et al., 2012), which demands dynamic applications and analytic processes. As it showed in Fig. 9, big data are capable and interacting with various connections. Users can have optimal information for their proper decision process by including the real-time car moments and trajectory. It means that users can gain a detailed description of the interaction of the components, which gives a clear insight into cars’ paths, locations, and alternative directions (Nomusa et al., 2012). The information shown in Fig. 9 is a single and complete visual of car moments and safety. It also gives a driving alternative and searching direction to the less traffic moment.

The car event path prediction at particular locations, speed, or other factors can collect from its neighbor car and synthesize at an integrated level. As shown in Fig. 9, the result gives an insight that will support the users to make the most proper decisions at various time intervals in a sequential process of the data coming from all sensor devices and applications in a given environment (Albert, 2013). The sensor nodes are essential to indicate the pivot for data similarity and formats in the context of the system by including attributes’ descriptions. The devices on the roadside perform various tasks, including sound sensing, image capturing, and spatial data recording. The image classifies based on contents, extracting sound sets, and recognizing and tracking objects through the data patterns (Rual et al., 2012) that look like Fig. 10.

Leveraging big data using machine learning techniques is a potential analytic tool and pertinent to optimizing object detection as:

Enable nodes and ages that allow visualizing the pivot and data similarities of the given object,
To guarantee the sequential delivery of events across sensor nodes and applications,
To handle the real-time de-duplication of records through the use of last value queues, sequence numbers, and content-based detections,
Capturing data to the proper storage location, which uses dynamic address space and Internode routing protocols to make sure every data routing properly,
Avoiding replications and data recovery and capturing data in a single location can expose the continuing problems.

This paper focused on Big data applications for the Internet of Vehicles' performance and safety. It focused on sensor data structuring and analysis for road safety. The integration between sensor cars and devices is through the cable of data exchange and flow within and around the intelligent systems and the surrounding environments. In the development of ITSs, smart cars often fast drive and demand to gain more communication abilities, which needs Big data technology for large-scale sensor data structuring and handling. It enhances various intelligent services and connectivity to deliver valuable information through the network onboard. The result showed big data for the Internet of Vehicles is a paradigm and leveraging big data for intelligent transportation system development, which addresses the fundamental of the cloud system. It explored big data mining models and frameworks to realize the future generation of sensor technology and inference sources for researchers and decision-makers in various aspects. It provides an additive value for intelligent vehicular systems' proper functionality and road service enhancement. Road safety and the sensor-based transportation system that dynamic and emerging research fields. The data storage and access facility is an extended work for this research.

Ethics: As the ethics of the conflicting interest, in this manuscript, there is no other body as an individual or company that will have a conflict of interest. In this statement, I declare that the work did not contain or involve human and animal participated study. The information related to the data, methodological procedures, and contents organized by the author is free from someone else intent.

Acknowledgments

I would like to thank the anonymous reviewers for their valuable comments, and constructive suggestions. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Abdul, Rehman et al, Context and Machine Learning-Based Trust Management Framework for Internet of Vehicles,Tech Science Press, Computers, Materials & Continua(2020) DOI:10.32604/CMC,pp 4125–4142
Ahmed O et al (2018) Big Data technologies: A survey. J King Saud Univ –Computer Inform Sci 30:431–448
Albert Bifet (2013) Mining Big Data in Real-Time, Informatica 37, pp. 15–20
Alesandra P et al (2018) Industry 4.0 Revolution in Autonomous and connected Vehicle a Non-Conventional Approach to Manage Big Data. J Theoretical Appl Inform Technol 96(1):10–18
Amandeep Singh, Suhag S (2019) Frequency Regulation in AC Microgrid with and without Electric Vehicle Using Multiverse-Optimized Fractional Order- PID controller, International Journal of Computing and Digital Systems ISSN (2210-142X) Int. J Com Dig Sys 8(4):375–385
Bo-Wu(2012) Internet-of-Vehicles based on Technologies of Internet-of-Things, ICLEM, Cheng Du, Oct.8–10, pp. 348–356
Corcoba V, Magaña, Organero MM (2011) Artemisia: Using an Android device as an Eco-Driving assistant, Cyber Journals: Multidisciplinary Journals in Science and Technology. Journal of Selected Areas in Mechatronics (JMTC)
Guo D, Mennis J(2009) Spatial data mining and geographic knowledge discovery: An introduction, Elsevier, Computers, Environment and Urban Systems 33, 403–408
Diebold FX (2003) Big Data Dynamic Factor Models for Macroeconomic Measurement and Forecasting Cambridge. Cambridge University Press, pp 115–122
Elmustafa Sayed A et al(2021) Machine Learning Technologies for Secure Vehicular Communication on Internet of Vehicles: Recent Advances and Applications, Wiley Hidawi, Security and Communication Networks, Volume 2021, https://doi.org/10.1155/2021/886835, PP. 1–24
Eugen Borcoci S, Obreja et al(2017) Internet of Vehicles Functional Architectures - Comparative Critical Study, AFIN 2017: The Ninth International Conference on Advances in Future Internet, pp. 12–19
Lee E-K, Gerla M et al(2016) Internet of Vehicles: From an intelligent grid of autonomous cars and vehicular fogs, International Journal of Distributed Sensor Networks, Vol. 12(9), DOI: 10.1177/ 1550147716665500, pp. 1–15
Fei-Yue, Wang (2010) Parallel Control and Management for Intelligent Transportation Systems: Concepts, Architectures, and Applications. IEEE Trans Intell Transp Syst 11(3):630–638
Fei-Yue, Wang et al (2017) Parallel Driving in CPSS: A Unified Approach for Transport Automation and Vehicle Intelligence, IEEE/CAA. J Automatica Sinica Vol 4(4):577–587
Gang, Zeng (2015) Application of Big Data in Intelligent Traffic System. IOSR J Comput Eng (IOSR-JCE) 17(1):01–04
Gebeyehu Belay Gebremeskel et al(2019) Architecture and optimization of data mining modeling for visualization of knowledge extraction: Patient safety care, Journal of King Saud University – Computer and Information Sciences 15(6), pp. 1–12
GB Gebremeskel, Y Chai, and Yang Z(2014) The Paradigm of Big Data for Augmenting Internet of Vehicle into the Intelligent Cloud Computing Systems, LNCS 8662, pp. 247–261, Springer International Publishing Switzerland
Gerard Goggin (2012) Driving the Internet: Mobile Internets, Cars, and the Social. Future Internet 4:306–321. doi:10.3390/fi4010306
Hany F, Atlam et al (2018) Internet of Things: State-of-the-art, Challenges, Applications, and Open Issues. Int J Intell Comput Res (IJICR) 9(3):928–938
Hedi HAMDI, Sabrine AMRI, Brahmi Z (2019) Managing Performance Interference Effects for Intelligent and Efficient Virtual Machines Placement based on GWO Approach in Cloud. Int J Comput Digit Syst Int J Com Dig Sys 8(4):317–332
Husen, Saboowala et al (2013) Designing Networks and Services for the Cloud, Cisco Systems, vol ISBN–10. Inc., 1-58714-294-5
Ifeyinwa Angela Ajah and Henry Friday Nweke (2019) Big Data and Business Analytics: Trends, Platforms, Success Factors and Applications, Big Data Cogn. Compute 3:1–30
Indu Sibaram Khara (2019) Internet of Vehicles (IOV): Evolution, Architectures, Security Issues, and Trust Aspects. Int J Recent Technol Eng 7:268–280Issue-6
Jairo R, Montoya-Torres et al (2021) Big Data Analytics and Intelligent Transportation Systems. Elsevier, IFAC PaperOnline, 54 – 2, pp 216–220
James J, Park (2018) Fusion algorithms and high-performance applications in vehicular cloud computing. J Supercomput 74:995–1000
Jawdat, Alshaer (2019) Towards Secure IoT: Securing Messages Dissemination in Intelligent Traffic Systems, International Journal of Computing and Digital Systems. Int J Com Dig Sys 8(4):333–341
Jimmy Lin, Ryaboy D (2013) Scaling Big Data Mining Infrastructure: The Twitter Experience. SIGKDD Explorations Vol 14(2):6–19
João, Gama (2013) Data Stream Mining: The Bounded Rationality. Informatica 37:21–2521
Kashif, Iqbal et al(2018) Intelligent Transportation System (ITS) for Smart-Cities Using Mamdani Fuzzy Inference System, (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 9, No. 2, pp. 94–105
Crawford K, Schultz J(2014) Big Data and Due Process: Toward a Framework to Redress Predictive Privacy Harms, 55 B. C.L. Rev. 93, http://lawdigitalcommons.bc.edu/bclr
Kayhan Zrar Ghafoor (2019) Millimeter-Wave Communication for the Internet of Vehicles: Status, Challenges and Perspectives, IEEE,Internet of Things Journal, (1–1), PP.1–23
Li MA et al (2018) Deployment of the Internet of Vehicle for Smart Cities: Applications, Architecture, and Challenges. IEEE Access 7:6473–6492
Li Z et al (2019) Big Data Analytics in Intelligent Transportation Systems: A Survey. IEEE Trans Intell Transp Syst 20(1):383–398
Guo L, Dong MX et al (2017) A Secure Mechanism for Big Data Collection in Large-Scale the Internet of Vehicle. IEEE J Internet Things 26–35. DOI 10.1109/JIoT.2017.2686451
Mahmood A, Zen H, Hilles SMS (2018) Big Data and Privacy Issues for Connected Vehicles in Intelligent Transportation Systems. In: Sakr S, Zomaya A (eds) Encyclopedia of Big Data Technologies. Springer, Cham. DOI: 10.1007/978-3-319-63962-8_234-1,
Zhang M, Wu T et al(2017) Carestream: An Industrial System of Big Data Processing, for the Internet of Vehicles, Proceedings of the VLDB Endowment, Vol. 10, No. 12 VLDB Endowment 21508097/17/08
Mohamed, Nahri et al(2018) IoV distributed architecture for real-time traffic data analytics, Elsevier, Procedia Computer Science 130 (2018) 480–487
Neha Verma, Kumar R (2012) A Method for Improving Data Delivery Efficiency in Vehicular Adhoc Networks. Int J Adv Sci Technol 44:11–24
Ning X et al(2013) Decentralized Information Flow Verification Framework for the Service Chain Composition in Mobile Computing Environments, IEEE 20th International Conference on Web Services, 978-0-7695-5025-1/13 $26.00, DOI 10.1109/ICWS.2013.81 , pp. 563–570
Nomusa D et al (2012) The State of Affairs in Internet of Things Research. Electron J Inform Syst Evaluation 15(3):244–258
Noor Afiza Mat Razali et al et al (2020) Conceptual Model for Connected Vehicles Safety and Security using Big Data Analytics. Int J Adv Comput Sci Appl 11(11):281–290
Tene O, Polonetsky J(2013) Big Data for All: Privacy and User Control in the Age of Analytics, 11 Nw.J. Tech. & Intell. Prop.239. http://scholarlycommons.law.northwestern.edu
Ovidiu Vermesan and Peter Friess (2013) Internet of Things: Converging Technologies for Smart Environments and Integrated Ecosystems, River Publishers, ISBN: 978-87-92982-96-4
Park DS (2018) Future computing with the Internet of Things and cloud computing. J Supercomputing 74:6401–6407
Paul C, Zikopoulos et al (2012) Understanding Big Data Analytics for Enterprise Class Hadoop and Streaming Data. The McGraw-Hill Companies, New York
Priscilla E, Juliet, Mathan K(2017) Performance Improvement of a Secure Mechanism for Big Data Collection on the Large-Scale Internet of Vehicle Using MATLAB, International Journal of Advanced Research in Computer and Communication Engineering, ISO 3297:2007, Vol. 6, Issue 10, pp. 244–250
Raghava, Mutharaju et al (2016) AI for Traffic Analytics, Feature Article. IEEE Intell Inf Bull 17(1):21–26
Rateb, Jabbar et al (2020) Blockchain for the Internet of Vehicles: A Decentralized IoT Solution for Vehicles Communication Using Ethereum. MDPI Sens 20:3928. doi:10.3390/s20143928., PP. 1–27
Razi, Iqbal (2018) Challenges in Designing Ethical Rules for Infrastructures on the Internet of Vehicles. Int J Adv Comput Sci Appl 9(3):11–15
Robert, Birke et al (eds) (2014) (Big) Data in a Virtualized World: Volume, Velocity, and Variety in Cloud Datacenters, USENIX Association 12th USENIX Conference on File and Storage Technologies pp. 177–190
Rual A. Gorcitz et al(2012) Vehicular Carriers for Big Data Transfers (Poster), IEEE Vehicular Networking Conference (VNC), Seoul: Korea, Republic
Sanghyun Ahn, Choi Jonghwa, ((2019)), Internet of Vehicles and Cost-Effective Traffic Signal Control, MDPI, Sensors, 19, 1275; doi:10.3390/s19061275, PP. 1–17,
Shaik et al (2019) A Review of Interoperability issues on the Internet of Vehicles (IoV). Int J Comput Digit Syst Int J Com Dig Sys 8(1):73–83
Shen B et al(2010) Research on Data Mining Models for the Internet of Things, IEEE, International Conference on Image Analysis and Signal Processing, Zhejiang, China, Apr 059 – 11, 78-1-4244-5555-3/10/$26.00
Shridevi Jeevan Kamble and Manjunath RK(2020) Machine Learning Approach on Traffic Congestion Monitoring System on Internet of Vehicles, Elsevier, Procedia Computer Science 171, PP. 2235–2241
Stefano, Ceri et al (2013) Towards Mega Modeling: A Walk-through Data Analysis Experiences. SIGMOD Record 42(3):19–27
Tasneem SJ, Darwish, Abu Bakar K(2018) Fog Based Intelligent Transportation Big Data Analytics on the Internet of Vehicles Environment: Motivations, Architecture, Challenges, and Critical Issues, IEEE Access, pp. 15679–15701
Tejinder Sharma and Vijay Kumar Banga (2013) Efficient and Enhanced Algorithm in Cloud Computing, International Journal of Soft Computing and Engineering (IJSCE), ISSN: 2231–2307, Vol. 3, Issue-1
Vijay Srinivas Agneeswaran (2012) Big-Data – Theoretical, Engineering, and Analytics Perspective, Springer-Verlag Berlin Heidelberg, BDA, LNCS 7678, pp. 8–15
Wei Fan, Bifet A (2013) Mining Big Data: Current Status, and Forecast for the Future. SIGKDD Explorations Vol 14(2):1–5
Xu W, Zhou H et al (2018) Internet of Vehicles in Big Data Era. IEEE/CAA J Automatica Sinica 5(1):19–35
Yang FH et al (2017) Architecture and Fundamental technologies for the Internet of Vehicles: a survey [J]. J Commun Inform networks 2(2):1–17
Yang Zhang et al., (2007), On Scheduling Vehicle-Road Side Data Access, ACM, 978-1-59593-739-1/07/0009…VANET’07, Canada
Leng Y, Zhao LS(2013) Novel Design of Intelligent Internet-of- vehicles Management System Based on Cloud Computing and internet-of-things, IEEE, International Conference on Electronic & Mechanical Engineering and Information Technology, 978-l-61284-088-8/11/$26.00, pp. 3190–3195
Yang X et al (2018) Advances in Vision-Based Lane Detection: Algorithms, Integration, Assessment, and Perspectives on ACP-Based Parallel Vision. IEEE/CAA J Automatica Sinica 5(3):645–661
YoungSik J et al (2019) Machine learning on big data for future computing. Springer 2925–2929. doi.org/10.1007/s11227-019-02872-z
Zhenyu, Zhou et al(2017) Social Big Data-based Content Dissemination on the Internet of Vehicles,IEEE Transactions on Industrial Informatics, Vol., No., pp.1–10
Zhou B et al (2019) Storage and access optimization scheme based on correlation probabilities on the internet of vehicles, Taylor and Francis. J Intell Transp Syst 1–17. https://doi.org/10.1080/15472450.1612247

Download PDF

Version 1

posted

You are reading this older preprint version

Read the latest preprint version →

Leveraging Big Data Analytics for Intelligent Transportation Systems: Optimize the Internet of Vehicles Data Structure and Modeling

Archived Versions:

Version 1

Abstract

Figures

I. Introduction

Ii. Related Works: The State Of The Arts

Iii. Methodology: The Architecture Of Big Data And The Internet Of Vehicle Integrity

Iv. Analysis And Discussion Of The Experiment

V. Conclusions

Declarations

Acknowledgments

References

Archived Versions:

Version 1