Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Department of Electronic and Information Engineering | en_US |
| dc.contributor.advisor | Ho, Wang-hei Ivan (EIE) | en_US |
| dc.creator | Magsino, Elmer Ramilo | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/11104 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | en_US |
| dc.rights | All rights reserved | en_US |
| dc.title | Efficient data dissemination for intelligent connected vehicles | en_US |
| dcterms.abstract | As the number of conventional, intelligent, and connected vehicles plying the roads and streets of major cities around the world increases, the need to provide comfortable, time-efficient, safe, environment- and energy-friendly travel becomes apparent. Daily trips of commuters, private, intelligent, and connected vehicles begin by anticipating traffic jams, accidents, road works, and other event-delaying circumstances to ensure that arriving at their destination is on time and convenient. To address these concerns and ensure the comfort of the daily voyage, intelligent and connected vehicles must obtain and share real-time, relevant, and accurate road information from/to other vehicles or roadside unit (RSU) infrastructure. However, as more vehicles use the road, the quantity of data sources increases, as well as the amount of available road information. In general, there are two categories of vehicular data: (1) control data and (2) environment data. Control data pertain to the safety applications for vehicles to abide by for achieving an organized traffic flow, such as the highÂway tollgate scenario. In contrast, environment data relate to non-safety and broad reports or applications for further processing, e.g., multi-junction city landscape. This thesis investigates the data dissemination topic for intelligent connected vehicles in vehicular networks to efficiently and effectively exchange control and environment data among vehicles and infrastructure in order to achieve travel objectives. Some of these travel objectives are: minimized queue time and length, optimal road map download based on vehicular demand under heterogeneous transmissions, maximized information sharing between vehicles and infrastructure, balanced roadside unit loading, and energy efficiency. We first examine the control data exchange in a tollgate scenario on highways where intelligent connected vehicles enter and exit the freeway for faster travel at the expense of paying toll fees for this convenience. Choosing and queueing at a certain tollgate becomes problematic for these intelligent connected vehicles since their current highway traffic and available tollgate service time knowledge are limited. Single-stage queuing system models the highway tollgate section with exponential service time, c servers, and K capacity, (M/M/c/K) to address the general problem of minimizing average queue time and length for all these vehicles entering/exiting the tollgate section. Also, independent and identical Poisson distributions characterize vehicular arrivals on highway lanes. A centralized fuzzy logic controller (FLC) is developed by considering the system's queue densities (number of vehicles per lane) and tollgate service times. The FLC facilitates the control data exchange where intelligent connected vehicles should line up. Extensive simulation results show approximately 50% improvement in reducing average queue waiting time and length when the central RSU employing an FLC is in-charge of allocating incoming vehicles to its appropriate servicing tollgate. The FLC controller has considered both the homogeneous and non-homogeneous arrivals of intelligent connected vehicles on the highway. As an added feature, the FLC allows early detection of forming high-density queues such that the service time of tollgate servers can be adjusted accordingly. | en_US |
| dcterms.abstract | We then investigate the general environment data dissemination problem considering the city-wide case by employing empirical mobility traces and real road environment LIght Detection And Ranging (LIDAR) data to represent intelligent vehicular mobility and sensed data, respectively. The framework follows the fog computing paradigm, where computing nodes are close to the vehicles, to immediately process received road information, then satisfy all vehicular demands. To initially reduce the file size of the sensed road segment data to be uploaded by the vehicles, octree compression, differential coding, and hashing techniques are implemented at the vehicular level. During the data dissemination stage, we propose an opportunistic index coding-based transmission scheme to optimally reduce the number of transmissions, transmitted data size, and overhead computations according to the vehicular data demand and availability. This setup considers heterogeneous modes of information download from an RSU fog node or base station to the vehicles. The objective of the proposed index coding-based transmission scheme applied at each of the RSU fog nodes is to reduce the reliance of intelligent and connected vehicles from long-range cellular transmissions and better exploit the short-range broadcast capacity at RSU fog nodes. Experiments involving mobile robots as intelligent connected vehicles have been tested to provide feasibility results for the implementation of this technique in the real world. To capture the city-wide data exchange using the proposed index coding transmission scheme, we employed empirical taxi mobility traces and evaluated the scheme's performance at each target junction. Our extensive simulation and mobile robot testbed results show a promising application of efficient data dissemination in an urban scenario utilizing the fog computing paradigm. Noting that the installation and maintenance of RSUs along public roads and highways can be quite expensive, RSUs must be strategically and economically deployed to support various vehicular fog applications. In the next two studies, we propose two strategies for maximizing information shared in a vehicular network. First, we develop an Information Sharing via Roadside Unit Allocation (ISRA) strategy to maximize information sharing between intelligent connected vehicles and RSU fog nodes. ISRA operates under the constraint that the number of RSU fog nodes is limited and only considers intersections as locations for possible RSU deployment. ISRA targets energy-efficient candidate locations during information exchange while balancing the load among the selected RSU fog nodes for better resource management. Given a set of candidate intersection locations, ISRA discriminates optimal locations from the set of candidate locations by implementing the proposed index coding-based transmission schemes and considering the vicinity's space mean speed and vehicular density. Simulations utilizing empirical mobility traces show that ISRA, on average, shares 20% more information at the energy efficiency of 83%, i.e., fewer packet transmissions, when compared to other deployment schemes. The second method, Enhanced Information SHAring via RSU Allocation (EISHA-RSU), also attempts to maximize shared information in a vehicular network but considers all locations as possible deployment spots. EISHA-RSU utilizes the Effective Region of Movement (ERM) concept to irregularly partition an urban setup based on the region's vehicular capacity and then discovers effective positions (EPs) within ERMs to deploy RSUs. We compare the performance of EISHA-RSU with three other benchmarks focusing on the amount of shared information, L-M measure, network starvation, vehicle count of 1-hop connectivity, and effectiveness. Extensive simulation employing three empirical urban mobility traces confirmed the efficiency of the proposed information-sharing scheme. In summary, this thesis investigated the data dissemination in vehicular networks for providing travel comfort and convenience to intelligent connected vehicles. This work has proposed adaptive and efficient data dissemination techniques to address the information exchange challenges in the data- and the source-rich vehicular environment. These techniques present methods of data exchange that are real-time, updated, and easily accessible to both intelligent connected vehicles and roadside infrastructure. It is desired that the thesis' results will provide practical solutions and approaches to the ever-growing and information-rich vehicular environment, particularly in promoting autonomous, connected driving. | en_US |
| dcterms.extent | xxviii, 159 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2021 | en_US |
| dcterms.educationalLevel | Ph.D. | en_US |
| dcterms.educationalLevel | All Doctorate | en_US |
| dcterms.LCSH | Data transmission systems | en_US |
| dcterms.LCSH | Intelligent transportation systems | en_US |
| dcterms.LCSH | Hong Kong Polytechnic University -- Dissertations | en_US |
| dcterms.accessRights | open access | en_US |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/11104