| Summary: | The development in telecommunications and particularly in wireless communications has been one of the most striking features of the contemporary world. The globalization only has been possible thanks to the evolution of communication technologies which increasingly have allowed to satisfy the constant people's needs of being "always connected" whatever the environment where they are. Concerning the evolution of technologies, vehicular networks have been one of the areas of great interest. This interest has been manifested both in research and in the development of the automotive industry that has produced innovative vehicles which are more and more equipped with new technologies. It is expected that communication in vehicular networks enable not only the communication between vehicles, but also a more comfortable and safe driving, making the user's experience of this type of networks richer and stimulating. The specific characteristics of vehicular networks, namely the high mobility, unpredictable routes, dynamic topology and the consequent and constant loss of connectivity, have been a challenge that has motivated studies to find solutions to these limitations. The work carried out for this dissertation is in the area of Vehicular Ad-hoc Networks (VANETs) and it is based on the Delay and Disruption Tolerant Networks (DTNs). With this project, identified as "Opportunistic Transmission of Information in Vehicular Networks", we aim to study the communication and transmission of information in these networks which do not allow communication without delays and disruptions. For this purpose it is studied the performance of DTN mechanisms in these networks. In this work it is used the implementation IBR-DTN to test DTN in VANETs. This implementation showed, in previous works, to be the one that presents the best performance comparing it with other existing implementations. The study involved, in an initial phase, reading and analyzing the implementation code so that it was possible to add instructions that allowed to observe the behavior of the implementation in the several tests carried out, as well as the correction of the bugs in the implementation. In the first phase, in laboratory, with fixed nodes and in a controlled environment, several scenarios were created to simulate the possible situations a node can meet: direct transfer with and without delay, indirect transfer (multi-hop) and indirect transfer with delay which corresponds to the store and transport of the bundles (set of information) until the next node. From the analysis of the collected information and observing the corresponding graphs, it was possible to observe that the implementation was working properly in the vehicles equipment for communication. Still in laboratory it was built an heterogeneous network with several devices (servers, NetRiders, Single Board Computers (SBCs), tablet, Raspberry Pi e Macbook) to show the integration of the IBR-DTN implementation and its extension in different equipments. During this test several files were sent among these devices, which were correctly received in the nodes previously defined as destination nodes. After testing and checking that everything was working properly in the laboratory, the same implementation was transferred to a testbed with 25 vehicles and 3 fixed infrastructures in Leixões harbor. In this testbed several DTN routing protocols were tested in order to check which of them showed better performance in the delivery rate of the bundles and of the collected information (the log files were also delivered via DTN) from the On-Board Units (OBUs) to the server, located in the Internet. The routing protocol with static routes to the Road Side Units (RSUs) proved a better efficiency compared to the other protocols. This was due to the the fact that this network is well covered with RSUs, and there is no relation between the historic of contacts and the probability that the vehicles will meet again in the future.
|
|---|