The selective use of redundancy for video streaming over Vehicular Ad Hoc Networks
Introduction
Video streaming over Vehicular Ad Hoc Networks (VANETs) is necessary or highly beneficial to a variety of envisioned services. A camera installed in a traffic light could stream the captured video chips of a local accident to either an incoming ambulance or to the nearest hospital. This could be done to provide valuable information to paramedics and physicians for a timely treatment response to improve the chances of survival in road accidents. Video chips of current traffic congestion, natural disaster or fire could be broadcasted to nearby vehicles to provide an accurate source of information to drivers, so that they are able to visualize road traffic and weather conditions ahead and take right decisions regarding safer or faster routes.
In this work, we refer to video streaming as the transmission of video content close to or in real time with no need for interaction between receivers and transmitters. This means that it is subject to more stringent requirements than those of stored videos (e.g., movies, shows) but that its requirements are slightly more lenient than those of interactive video (e.g., video conferencing).
Sending redundant information with the original packets is a common method that enables the receiver to reconstruct lost packets and to achieve a higher delivery ratio. The use of redundancy in VANETs has already been suggested [1], [2], [3], but, in this work, we go through an extensive analysis of its benefits for the specific case of video streaming. We discuss the advantages and disadvantages of using Erasure Coding and Network Coding for an efficient use of redundancy. This work covers the scenarios for both video unicast and video broadcast. Unicast covers the transmission from one point (source) to another one (destination), whereas broadcast deals with the transmission from a single source to multiple interested parties. This work also evaluates two distinct coding techniques: Random Linear Coding (RLC) and XOR-based Coding. RLC is a technique used to improve the network’s throughput, so it is an efficient method to add redundant information. On the other hand, XOR-based coding is another effective adding redundancy method based on pure XOR operation during coding computation.
The contributions in this work are divided in four parts: (i) a thorough study on the issues inherent in using redundancy for video streaming over VANETs for both unicast and broadcast; (ii) an evaluation on the benefits of two novel coding solutions for Network Coding and Erasure Coding: XOR-based Coding and Linear Coding; (iii) the demonstration of the improvements on a novel approach that uses additional redundancy selectively in I-frames data only; and, (iv) an extensive comparison provided on solution performance through the perspective of the receiving rate of unique video content.
The importance of the latter contribution is due to the strong correlation between the video quality that can be displayed to users and the data rates by which videos can be received at applications on receivers. The benefit on the use of redundancy to handle packet loss is always a trade-off between increases in delivery ratio and cost. Normally, cost is given by the increased used of bandwidth, however, as we discuss in this work, for video streaming over VANETs, the impact on the receiving rate of unique video content is even more crucial.
An analysis of the most relevant and current studies in the literature is shown in the next section. Section 3 provides a summarized discussion regarding the use of coding strategies for an efficient use of redundancy. The peculiarities involved in the use of redundancy for video unicast and video broadcast are discussed in Sections 4.1 Video unicast, 4.2 Video broadcast, respectively. Section 5 describes the selective use of redundancy in I-frames and presents a thorough performance comparison of the different strategies for both video broadcast and unicast. Finally, Section 6 presents our conclusions and future work.
Section snippets
Related work
The study of VANETs is a relatively recent topic that has been studied in the literature. In this section, we list and discuss some of the existing studies in the literature. We also evaluate those solutions suitability to video streaming and compare them to the solution we propose in this work.
In particular, studies on video streaming over VANETs are very limited, but data dissemination, a more general case of video streaming, has been widely explored. For instance, VIRTUS [4], LIAITHON [5]
Redundancy
Vehicular Ad Hoc Networks are strongly affected by packet loss mainly due to intermittent connections between nearby vehicles and communication congestion. Therefore, we need to design mechanisms able to recover from frequent losses. This is particularly important for video streaming that requires high rates of successful data delivery.
Interactive error correction mechanisms are not suitable for video streaming over VANETs because the exchange of messages between receivers and senders are not
Video streaming
Video streaming poses stringent requirements specially regarding successful delivery ratios and end-to-end delay [29]. Video streaming consists of streaming video content in a close to real-time manner in which delay cannot exceed a few seconds. It is also fundamental to provide solutions that delivery the vast majority of the content transmitted since error correction mechanisms require interaction between receivers and transmitters deteriorating prohibitively the end-to-end delay.
VANETs are
Selective redundancy
Through our preliminary analyses, we have noticed that Erasure Coding and our Hybrid approach have obtained a better result than the use of Network Coding (NC) by itself. In the case of video unicast, it is clear that NC is unfeasible since multiple transmissions by intermediary nodes are strongly avoided. In video broadcast, NC has presented some ability to take advantage of these multiple transmissions, but it is necessary to apply EC in the first hop of communication.
Another novel approach
Final remarks
In this work, we have provided a thorough study on the use of redundancy for video streaming over VANETs. We have investigated both EC and NC techniques, and designed XOR-based Coding with a strategy concerning the degrees distribution and segment selection suitable for the application of video streaming over VANETs. Furthermore, we have proposed the use of a novel Hybrid solution that uses EC for the first hop on video broadcast and NC for the remaining intermediary transmissions towards other
Acknowledgments
This work is partially supported by NSERC DIVA Strategic Research Network, Canada Research Chairs Program and ORF/MRI research Funds, and CNPq Brazil Research Fund.
Cristiano Rezende is a PhD candidate at the University of Ottawa, Canada. He has received his Bachelor’s Degree in 2005 and his Master’s Degree in 2007, both at the Universidade Federal de Minas Gerais in Belo Horizonte, Brazil. In 2005, he got the highest score on the Brazilian Students’ Performance National Exam in Computer Science (ENADE). He has years of experience in research in Vehicular Ad Hoc Networks, Mobile Ad Hoc Networks and Wireless Sensors Networks. His current research is in the
References (36)
- et al.
A receiver-based video dissemination solution for vehicular networks with content transmissions decoupled from relay node selection
Ad Hoc Netw.
(2014) - et al.
The impact of mobility on mobile ad hoc networks through the perspective of complex networks
J. Parall. Distrib. Comput.
(2011) - et al.
Blind XOR: low-overhead loss recovery for vehicular safety communications
IEEE Trans. Veh. Technol.
(2012) - C. Li, J. Jose, X. Wu, Distributed-fountain network code (DFNC) for content delivery in vehicular networks, in:...
- et al.
DDRC: data dissemination in vehicular networks using rateless codes
J. Inform. Sci. Eng.
(2010) - et al.
A reactive and scalable unicast solution for video streaming over VANETs
IEEE Trans. Comput. PP
(2014) - R. Wang, C. Rezende, H.S. Ramos, R.W. Pazzi, A. Boukerche, A.A.F. Loureiro, LIAITHON: a location-aware multipath video...
- et al.
Analysis of multipath routing—Part I: the effect on the packet delivery ratio
IEEE Trans. Wireless Commun.
(2004) - C. Rezende, R.W. Pazzi, A. Boukerche, A reactive solution with a redundancy-based error correction mechanism for video...
- et al.
Analysis of multipath routing, Part 2: mitigation of the effects of frequently changing network topologies
IEEE Trans. Wireless Commun.
(2004)
Understanding processing overheads of network coding based content distribution in VANETs
IEEE Trans. Parall. Distrib. Syst.
Vehicular communications: emergency video streams and network coding
J. Internet Serv. Appl.
Joint redundant and random network coding for robust video transmission over lossy networks
Int. J. Mob. Inform. Syst.
Cited by (0)
Cristiano Rezende is a PhD candidate at the University of Ottawa, Canada. He has received his Bachelor’s Degree in 2005 and his Master’s Degree in 2007, both at the Universidade Federal de Minas Gerais in Belo Horizonte, Brazil. In 2005, he got the highest score on the Brazilian Students’ Performance National Exam in Computer Science (ENADE). He has years of experience in research in Vehicular Ad Hoc Networks, Mobile Ad Hoc Networks and Wireless Sensors Networks. His current research is in the provision of video streaming capabilities of VANETs.
Azzedine Boukerche (FIEEE, FEiC, FCAE, FAAAS) is a Full Professor and holds a Canada Research Chair Tier-1 position in distributed simulation and wireless and mobile networking at the University of Ottawa. He is the Founding Director of PARADISE Research Laboratory at Ottawa U. Prior to this, he held a faculty position at the University of North Texas, USA. He worked as a Senior Scientist at the Simulation Sciences Division, Metron Corporation located in San Diego. He was also employed as a Faculty at the School of Computer Science McGill University, and taught at Polytechnic of Montreal. He spent a year at the JPL/NASA-California Institute of Technology where he contributed to a project centered about the specification and verification of the software used to control interplanetary spacecraft operated by JPL/NASA Laboratory. He was the recipient of several Awards including the IEEE Golden Core Award and IEEE Canada Gotlieb Computer Silver Medal Award.
M. Almulla is a Visiting Associate Professor at the University of Ottawa. He obtained his PhD, MSc and BSc from McGill University in 1995, 1990 and 1986 respectively. He as a Visiting Professor at the PARADISE Research Laboratory, University of Ottawa. His research of interests include intelligent dynamic systems, wireless networks, distributed and mobile computing and intelligent vehicular networks.
Antonio A.F. Loureiro received his B.Sc. and M.Sc. degrees in computer science from the Federal University of Minas Gerais (UFMG), Brazil, and the Ph.D. degree in computer science from the University of British Columbia, Canada. Currently, he is a full professor of computer science at UFMG, where he leads the research group in wireless sensor networks. He has been a Visiting Professor at the PARADISE Research Laboratory, University of Ottawa. His main research areas are wireless sensor networks, mobile computing, and distributed algorithms. In the last 10 years he has published regularly in international conferences and journals related to those areas, and also presented tutorials at international conferences.