Qoe enhanced social live interactive streaming

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 03 Issue: 05 | May-2014, Available @ http://www.ijret.org 764
QOE ENHANCED SOCIAL LIVE INTERACTIVE STREAMING
J.Ramya1
, D. Venkata Subramanian2
, R.Nedunchellian3
1
Dept of Computer Science & Engineering, Saveetha School of Engineering, Saveetha University, Chennai, India.
2
3
Abstract
Globalized live streaming services have got subscribers all around the world. The rapid emergence of the mobile devices has made it
a natural aspiration of such users to make social interaction with others who have got involved in the same application. The term
Social TV implies the integrated support of the television and the computer technology in order to provide a group viewing experience
to the users. The social interactions within the users who use such live streaming service needs to be spontaneous. The cloud
computing technology has triggered enormous opportunities to facilitate the mobile live streaming of the multimedia contents with an
extended support to interact with the users. The quality of service (QOS), storage and sharing are some of the issues that have to be
addressed in order to provide a mobile social television in a cloud environment. The cloud technology effectively handles some of
these issues by assigning proxies for the mobile users. These proxies (surrogates) for the users, operates on the base of transcoding
mechanism. Also the PAAS and IAAS cloud services are keys in providing such an effective interaction based live streaming.
Keywords— Mobile Social TV, Live Streaming, Quality Of Service, Social Interactions, Cloud Computing.
----------------------------------------------------------------------***--------------------------------------------------------------------
1. INTRODUCTION
Social TV started in the early 2000s with limited success as the
creation of the shared connections was cumbersome with a
remote control and the User Interface (UI) design made the
interaction disruptive to the TV experience. But social
networking has made Social TV suddenly feasible, since it
already encourages constant connection between members of
the network and the creation of likely minded groups. The
shared content and activities often relate to TV content. At the
same time, the smart phone market has been growing quickly.
Mobile TV is the technology providing multimedia contents to
the user by wireless communication. As our society is
becoming increasingly mobile, there is an emerging need for
content and services to become mobile as well. Within the
scope of our research, we define Mobile TV as real-time
transmission of traditional TV content to mobile handsets.
Mobile TV promises thrilling benefits to consumers and
increased revenues for mobile telecommunications’ operators,
equipment suppliers and television providers. The mobiale TV
device should produce an image quality which is not
significantly inferior to the standard established by traditional
TV, even if the screen size is much smaller. The network
coverage and the signal strength should be sufficiently good to
give the viewer an uninterrupted service of comparable quality
with traditional TV. As most mobile TV networks still are not
much less than prototypes there is still some way to go before
the transmission quality is sufficiently good.
This paper proceeds as follows: Section 2 introduces the
theoretical foundation of this study. Section 3 outlines the key
terms and technologies. Section 4 focuses on architectural
model; Finally Section 5 summarizes the results of study.
2. RELATED WORKS
Media Cloud is a general idea of integrating the multimedia
services with the cloud technologies, which enables the user to
access the multimedia content through the cloud computing
technology. Providing multimedia service through the cloud
technology faces the scalability, Qos and heterogeneity
challenges [4]. However the Content Delivery Network (CDN)
and the peer to peer multimedia computing have been worked
out to alleviate the problems in multimedia computing by
pushing the multimedia content to the edges and to the peers
respectively [4]. Rings et al has proposed the idea of
integrating the cloud computing with the multimedia services
without the Qos provisioning. Whereas Zhu et al proposed the
multimedia cloud computing which provides Qos provisioning.
The working of the cloud services for the multimedia content
can be differentiated and explained mainly by two ways, they
are: multimedia-aware cloud and cloud aware multimedia.
Here the former idea pertains to the quality of service(QOS)
services for multimedia content and the later concept deals
about how well the sharing, retrieval and storage of the
multimedia content can utilize the cloud-computing resources
for achieving a better Quality Of experience [2] More finer
framework like as Cloud Assisted Live Media Streaming have
been studied to utilize the cloud resources by leasing and
adjusting the cloud servers for the dynamic user
demands[15].Basically the video streaming using the cloud
technology relies on the transcoding mechanism, which has
two mapping options like Hallsh-based and Lateness-first

__________________________________________________________________________________________
Mapping for reducing the jitter in the transcoding [7]. Also the
Scalable Video Coding framework has also been proposed to
reduce the latencies in the transcoding and to adapt to the
dynamically changing network conditions [3]. The works of
Satyanarayanan et al [13] have suggested the implementation
of the dynamic virtual machine in carrying out the offloading
to the mobile devices computations. The work of the proxy
servers in the cloud technology for the multimedia or video
streaming service has effectively handled the issue of the
storage, processing challenges faced by some mobile devices
in accessing those services [1] have seen the idea of interactive
live streaming through the cloud computing technology, which
provides a co-viewing experience to the users with their friends
in different geographic locations. Thereby making the activity
more social able. The experiments by Oehllberg et al [6] on the
social activities of a human while viewing different video
contents have been inspiring but still those frameworks cannot
be applied in to the mobile environments directly. The design
proposed by the Coppens et al [4] have been intended to
elaborate the social interactions but it got constrained with the
broadcast program channels. Reflex [7] is a mechanism to
enhance spontaneity in the interaction between the users,
which makes use of the Google App Engine to achieve more
scalable and quality service in the social network, taking the
large amount users into consideration. These interactions can
be handled by the messenger service through the cloud
computing technology.
The video streaming through the help of cloud computing
technology has been facing another significant challenge as
studied by [2], which is the Quality of experience [2] have
already proposed the idea of enhancing the quality of
experience in viewing the live video content by introducing the
features like zooming, segmentation and panning but have not
extended to the interactive streaming environment. Here we
are proposing a framework for enriching the quality of
experience to the users in an interactive live streaming, with
the cloud computing as its backbone.
3. CLOUD-BASED INTERACTIVE STREAMING
The live video streaming has took a whole new dimension with
the evolution of cloud computing technology. In the case of
live video streaming, the cloud computing extends its support
for interaction within the users and for a better quality of
experience. The architecture workflow has been given as
below in figure. The proxy servers or surrogates are assigned
to each user who gets logged in to the cloud. These servers are
provided by the Infrastructure as a service cloud. The
surrogates will efficiently provide the offloading, it will be
enacting as a middleware between the mobile devices and the
video sources. It will be encapsulating the transcoding,
segmentation and content adaptation operations. In addition to
these operations the messenger, will also be handled by these
surrogates, which is a key component in delivering effective
interaction between the users.
The quality of experience for the users can be enhanced by the
features like segmentation, zooming and content adaptations.
The extensible messaging and presence protocol (XMPP), here
plays a significant role in operating the transportation of the
video segments and it is helpful in exchanging the metadata of
these video streams like title, description.
3.1 Transcoding
The surrogates assigned to the users, will be handling the
trancoder, which decide the encoding format for the video
stream dynamically. The bit rate and the dimension of the
video stream will also be decided in this module. Since MPEG-
4 has been the de-facto standard for the video delivery over
loosy medium, this stream will be generally followed for the
implementation. FFmpeg library is fundamentally utilized by
this service for generating the thumbnails.
The transcoding frameworks splits the video contents into
overlapping and non-overlapping group of pictures (GOP).
These pictures may encircle different encoding qualities and
resolutions, forming several layers.Such layers will be sliced as
a coding-independent contents. Here 16x16 macro-blocks are
framed for luma components and 8x8 MB is framed for
chroma components.The granularity is decided with the
parallelism in GOP level and MBs,which is divided into inter-
parallelism and intra-parallelism.The simple diagramatic
illustration in the fig [ ] of the transcoding framework is given
below will give a better understanding of the idea elloborated
above.
(i) Inter-node Parallelism: The Inter-node parallelism manages
the reduced picture nodes. The real-time transcoding is
achieved through this inter-node parallelism,the transcoding
jitter is an effect of the variatons in the time delay of the
encoding of the GOP(Group Of Pictures). The encoding time is
mainly estimated for the optimization problem .
(ii) Intra-node Parallelism: The Intra-node parallelism
manages the individual slice of the picture node. The Intra-
node parallelism is very significant in the GOP encoding and it
is vital in fixing the upper bound on the average computation
time spent on the GOP encoding. Moreover this parallelism is
not simply enough in reducing the access time
Fig. 1 Transcoding Framework

__________________________________________________________________________________________
3.2 Segmentation
The segmentation, is a feature to enhance the streamed videos,
here this feature will create the list of points of scenes by
identifying the scenes in the video. And by making use of the
transcoding service the thumbnails for such segments will be
created. Basically this feature is based on the OpenCV
algorithms.
3.3 Zooming
This service will enable the users to zoom into the middle of
the video stream. This will also be responsible in cropping of
the streamed videos. This service fundamentally follows the
object recognition service, which necessary in recognizing the
objects in the video. The object recognizing service can enable
to perform more complex zooming functionalities.
3.4 Messenger
This service is responsible in delivering the asynchronous
message to the user from the surrogates. User will make
queries periodically to the social cloud through this service,
which is in connection with the social cloud. This service is
also responsible in processing the plain text data, which are in
xml format.
3.5 Gateway
The authentication of the users logged in are checked by this
service, which also handles the logged in user list and
maintains it in a separate database. The gateway is also
responsible in reserving and destroying the surrogates based on
the current work load.
3.6 Social Cloud Service
The social cloud service will be maintaining the data stores like
Big-Table in Google App Engine, for storing all the details
regarding the user records, sessions and messages. A data store
interface is used to query and manipulate the Big-table which
indeed a multi-dimensional maps, stores data object as an
entity entitled with multiple properties. For the video streaming
it fundamentally follows the RTP protocol. The XMPP handles
the metadata of the video segments. The social cloud is
basically extended by the PaaS services.
Fig.2 Architectural Framework
3.7 Synchronizer
The synchronizer features the property of enabling the user to
view the concerned video content in the same window as the
other user in same session. This is achieved by retrieving
current playback position and guides the user to adjust to that
position. This is particularly concerned in providing the co-
viewing experience to the users.
4. PERFORMANCE METRICS
In this proposed framework there are certain factors which
affect its performance, where are here discussing some of the
possible factors and the predicted performance improvements
by this framework
4.1 Transcoding Latency
The Transcoding latency can make an impact on the entire
framework if had not handled carefully. Even with the modern
day’s multi core processors the process of encoding is highly
complex in nature. The transcoding delay can cause the delay
in the access of video content for the user and it could cause
even the freezing of the particular video content.
4.2 Power Consumption
The power consumption of the device, used for the access is
significantly important, since that might well have a prime role
in the working of this proposed framework in a considerable
margin. The playback segment size of 10, which has been
widely followed in streaming applications, was proposed by

__________________________________________________________________________________________
the Http Live Streaming protocol. However this segment has
found to have drained the battery life significantly. Normally
the power consumption factor is profiled by an Xcode tool
called “Instruments”.
4.3 Interaction Spontaneity
In this proposed framework the spontaneity of the interaction
between the connected users is measured by two levels of
factors. The first is the latency in the sending of the message to
the surrogates and the confirmation for the message being
registered in the cloud. The other is the latency in query sent
by the user to reach the assigned surrogates. Also the round-
trip time between the surrogates and the GAE which also need
to be taken into consideration for calculating the interaction
latency in this framework.
5. IMPROVEMENTS PREDICTED WITH THE
FRAMEWORK
5.1 Battery Efficiency
To eliminate battery usage of the mobile devices this proposed
architecture is aimed at providing an efficient burst
transmission technique, which enables the mobile devices to
operate on three states as High, Low and Intermediate. This
technique fundamentally follows the Http Live streaming
protocol, by which the video will be segmented by the
surrogates and sent to the mobile devices on request. The
mobile devices will be operating in High state when receiving
the video segments and will be in Low state when remaining
idle. The Intermediate state acts as the transition state between
these two states.
5.2 User Experience
Another important key aspect of this framework is the
enhancement of the user experience by the features like
zooming in and out the streamed video and the Scene by scene
segmentation of the video stream. This enhancement is
particularly achieved with the help of the XMPP protocol and
RTP protocol which handles the video streaming, exchange of
the metadata and video segments information. Here the
metadata handler will be responsible in fetching the video
segments and Playlist handler will be responsible in providing
the preview thumbnails of the videos. The thumbnails, having
a smaller resolution will get loaded very fast. The object
recognition service is a key part in providing the zooming
feature on the streamed videos.
5.3 Spontaneous Interactivity
Interaction is a key aspect of a social live streaming
framework and this aspect is effectively handled in this
proposed architecture by the means of the Messenger service,
which being operated in a asynchronous way will provide
spontaneous interactions between the connected users. A Big-
table like datastore will be made use to handle these data.
5.4 Scalability
As this proposed framework takes the implementation phase, a
challenge will be the ability of this entire system to handle the
large amount users who gets too logged into the service (i.e.)
the scalability measure of this framework. Being deployed on
to the cloud network, it should effectively handle this problem
as well but still this area has to be addressed in the future
works.
6. REMARKS AND FUTURE WORKS
Our work has concerned primarily in suggesting a framework
integrating the social interactive live streaming with the
Quality enhanced user experience. The enhancement in the
QoE of the user obviously shows great scope in the further
feature enhancement with the streamed video contents, like as
recording the live video contents to the device, noise removal
and so on. On the other side the interactive aspect of the
framework has several areas to be more deeply addressed like
applying memcache support and more efficient transcoding
mechanism.
7. CONCLUSIONS
We conclude by proposing a framework for enriching the
quality of experience to the users in an interactive live
streaming, with the cloud computing as its backbone. And
mobile users can import a live or on-demand video to watch
from any video streaming site, invite their friends to watch the
video concurrently, and chat with their friends while enjoying
the video.
ACKNOWLEDGEMENTS
The authors are immensely happy to extend their gratitude to
the reviewers for their insightful thoughts, which this paper
incorporates. We also thank the study participants for their
Invaluable time and appreciable enthusiasm.
REFERENCES
[1] Yu Wu, Zhizhong Zhang, Chuan Wu, Zongpeng Li,
Francis C.M. Lau:”Cloud MoV: Cloud-based Mobile
Social TV”, IEEE 2013.
[2] DejanKovachev, Yiwei Cao and Ralf Klamma,”Cloud
Services for Improved User Experience in Sharing
Mobile Videos”, IEEE 2012.
[3] Zixia Huang, Chao Mei, Li Erran Li, Thomas Woo,
“Cloud Stream delivering high-quality streaming videos
through a cloud-based SVC proxy”.
[4] Wen Hui1, Chuang Lin and Yang, “Media -Cloud: A
New Paradigm of Multimedia Computing”, KSII:
Transaction on Internet and Information Systems,
Vol.6, April 2012.

__________________________________________________________________________________________
[5] Ramesh.B, Savitha.N, Manjunath.A.E,”Mobile
Application in Multimedia Cloud Computing”,
Int.J.Computer Technology and Application, Vol 4(1),
97-103.Feb, 2013.
[6] Nicolas Ducheneaut, Robert.J.Moore1, Lora Oehlberg,
James.D.Thornton, EricNickell,“SocialTV: Designing
for Distributed, Sociable Television Viewing”.
[7] Zimu Liu, Yuan Feng, Baochun Li,”Socialize
Spontaneously with Mobile Applications”,IEEE’12,
INFOCOM proceedings.
[8] Hassnaa Moustafa and Nicolas Maréchal, Sherali
Zeadally, “Mobile Multimedia Applications: Delivery
Technologies”, IEEE, CS (September/October 2012).
[9] H. Schwarz, D. Marpe, and T. Wiegand, “Overview of
the scalable video coding extension of the H.264 AVC
standard,” IEEE Transaction on Circuits and Systems
for Video Technology, vol. 17, no. 9, pp. 1103–1120,
Sep. 2007.
[10] Z. Huang, C. Mei, L. E. Li, and T. Woo, “Cloud
Stream: Delivering High-Quality Streaming Videos
through a Cloud-Based SVC Proxy,” in IEEE
INFOCOM, 2011.
[11] Y. Wu, C. Wu, B. Li, X. Qiu, and F. C. Lau, “Cloud
Media: When Cloud on Demand Meets Video on
Demand,” in IEEE ICDCS, 2011.
[12] Z.Wu, C.Zhang, Y. Ji, and H. Wang. Towards Cloud
and Terminal Collaborative Mobile Social Network
Service. 2010 IEEE International Conference
[13] M. Satyanarayanan, P. Bahl, R. Caceres, and N. Davies,
“The case for vm-based cloudlets in mobile
computing,” IEEE Pervasive Computing, vol. 8, pp. 14–
23, 2009.
[14] A.Carroll and G.Heiser, “An analysis of power
consumption in as Smartphone,” in Proc. of
USENIXATC, 2010.
[15] Feng Wang, Jiangchuan Liu, and Minghua Chen,
CALMS: Cloud-Assisted Live Media Streaming
Globalized Demands with Time/Region Diversities”.
[16] R. Schatz and S. Egger, “Social Interaction Features for
Mobile TV Services, “in Proc. of 2008 IEEE
International Symposium on Broadband Multimedia
Systems and Broadcasting, 2008.
[17] H. Knoche, M. Papaleo, M. A. Sasse, and A. Vanelli-
Coralli, “The Kindest Cut: Enhancing the User
Experience of Mobile TV Through Adequate
Zooming,” in Proceedings of the 15th International
Conference on Multimedia, ser. MULTIMEDIA ’07.
ACM, 2007, pp. 87–96
[18] R. Pereira, M. Azambuja, K. Breitman, and M. Endler,
“An Architecture for Distributed High Performance
Video Processing in the Cloud,” in Proceedings of the
2010 IEEE 3rd International Conference on Cloud
Computing, 2010, pp. 482–489.

Qoe enhanced social live interactive streaming

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Qoe enhanced social live interactive streaming (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Qoe enhanced social live interactive streaming