Please use this identifier to cite or link to this item: https://hdl.handle.net/11000/5757

Aplicación de técnicas para la mejora de la transmisión de contenido multimedia en entornos de redes vehiculares


Thumbnail

View/Open:
 TD Garrido Abenza, Pedro Pablo .pdf
2,06 MB
Adobe PDF
Share:
Title:
Aplicación de técnicas para la mejora de la transmisión de contenido multimedia en entornos de redes vehiculares
Authors:
Garrido Abenza, Pedro Pablo
Tutor:
Pérez Malumbres, Manuel José
Piñol Peral, Pablo José
Department:
Departamentos de la UMH::Ingeniería de Comunicaciones
Issue Date:
2019-12-16
URI:
http://hdl.handle.net/11000/5757
Abstract:
La transmisión de contenido multimedia en entornos de redes vehiculares, o Vehicular Ad-hoc NETworks (VANETs), puede tener un gran número de aplicaciones. Sin embargo, el uso de un canal inalámbrico compartido por todos los nodos de la red y el alto requerimiento de ancho de banda para el envío de ...  Ver más
The transmission of multimedia content in vehicular network environments, or Vehicular Ad-hoc NETworks (VANETs), is gaining high relevance in the automotive industry due to the attractive applications and services that may be developed. However, the use of a wireless channel shared by all the vehicles in a particular area, joined with the constrained requirements in terms of bandwidth and delay of video delivery applications, can make the quality of the received video not acceptable, especially in real time applications (e.g., video streaming), where low delay and bounded jitter are required. Furthermore, the topology of the network suffers frequent changes due to the mobility of the network nodes, and the high speed of the vehicles drastically limits the communication time between them or with the fixed infrastructure. By using video coding or compression techniques such as the High Efficiency Video Coding (HEVC) standard, the amount of data required for storage and transmission bandwidth are significantly reduced. However, the video quality perceived by the receiver can still be greatly affected due to transmission impairments, so other mechanisms are required to increase the robustness of the transmitted video sequences. This Doctoral Thesis is focused in the research of techniques that allow to significantly improve the video quality perceived by a user receiving an HEVC video stream in a vehicular network environment. In order to achieve this objective, we began with the development of some software tools to properly carry out the required experiments. As a first result, the GatcomSUMO application was developed. This application allows the generation of network scenarios and vehicles mobility for the OMNeT++ simulator, the Veins framework and the SUMO traffic simulator. Subsequently, the environment called Video Delivery Simulation Framework over Vehicular Networks (VDSF-VN) was also developed with the aim of creating a high detailed simulation environment including all the steps of a video delivery service: encoding of source video using the HEVC encoder, packetization of the encoded bitstream, simulation of network packets delivery, reassembling of the received packets, and finally, decoding of the reconstructed bitstream to play the video by the user. Also, several auxiliar utilities were developed like the ones focused in the automatization and statistics graphs generation, among others. With all of the above mentioned tools, various experiments were carried out by simulation to evaluate the quality of video sequences transmitted under different traffic conditions on the network. First, the parameters of the HEVC video encoder were tunned, in particular, using encoding modes with different Intra refresh patterns and varying the number of fragments (tiles) of each video image (frame). Encoding modes with higher Intra refresh rates (e.g., All Intra) require a higher bitrate for transmission, suffer a greater number of lost packets, a greater delay and delay variation (jitter) than other modes with less Intra refresh rates. However, in the presence of other traffic on the network (background traffic), it was evidenced that they are more robust than the other modes, as the loss of a frame can have a greater negative effect on the final quality of the received video due to the interdependencies between frames. Regarding the use of tiles, different uniform patterns were used together with various coding modes. As the number of tiles per frame increases, even though some encoding efficiency is lost, greater robustness is achieved in the presence of background traffic, that is, a better quality of the reconstructed video is achieved. However, the use of tiles introduces an overload in the resulting bitstream, and as from a certain value upwards the quality increase is not significant, in order to not saturate the network it was determined that the optimum is the use of intermediate values (between 4 and 8 tiles per frame). Secondly, and combined with the previous mechanisms, the Quality of Service (QoS) defined in IEEE Std. 802.11p-2010 was used, by assigning a higher priority to the packages corresponding to I frames only, and all the video frames. The best results were obtained in the latter case. In comparison to an scenario where QoS is not used, the losses corresponding to the priority frames are considerably reduced, without harming the rest of the lower priority traffic existing in the network (background traffic), obtaining a better use of the wireless channel.
Keywords/Subjects:
Ingeniería de Telecomunicaciones
Redes de comunicaciones
Tratamiento digital de imágenes
Knowledge area:
CDU: Ciencias aplicadas: Ingeniería. Tecnología: Ingeniería mecánica en general. Tecnología nuclear. Electrotecnia. Maquinaria: Ingeniería eléctrica. Electrotecnia. Telecomunicaciones
Type of document:
info:eu-repo/semantics/doctoralThesis
Access rights:
info:eu-repo/semantics/openAccess
Appears in Collections:
Tesis doctorales - Ciencias e Ingenierías



Creative Commons ???jsp.display-item.text9???