distributed adaptation decision-taking framework and scalable video coding tunneling for edge and...
DESCRIPTION
Existing and future media ecosystems need to cope with the ever-increasing heterogeneity of networks, devices, and user characteristics collectively referred to as (usage) context. The key to address this problem is media adaptation to various and dynamically changing contexts in order to provide a service quality that is regarded as satisfactory by the end user. The adaptation can be performed in many ways and at different locations, e.g., at the edge and within the network resulting in a substantial number of issues to be integrated within a media ecosystem. This paper describes research challenges, key innovations, target research outcomes, and achievements so far for edge and in-network media adaptation by introducing the concept of Scalable Video Coding (SVC) tunneling.TRANSCRIPT
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
1
DISTRIBUTED ADAPTATION DECISION-TAKING FRAMEWORK AND
SCALABLE VIDEO CODING TUNNELING FOR
EDGE AND IN-NETWORK MEDIA ADAPTATION
Michael Grafl, Christian Timmerer, Markus Waltl, George Xilouris, Nikolaos Zotos, Daniele Renzi,
Stefano Battista, and Alex Chernilov
TEMU 2012, Heraklion, Greece, July 31, 2012
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
2
OUTLINE Introduction & Problem Statement Research Challenges ALICANTE Adaptation Framework
Adaptation & SVC Tunneling
Targeted Research Outcomes Proposed Integrated Test-Bed Scientific Results Achieved So Far
RC Modes for SVC Tunneling Results Result Evaluation
Conclusions
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
3
INTRODUCTION & PROBLEM STATEMENT
Universal Multimedia Access (UMA) Evolution of device and network infrastructure
Heterogeneity of devices, platforms, and networks Scalable Video Coding (SVC): bitstream consists of
cumulative layers that refine the video (resolution, framerate, bitrate)
SVC tunneling approach featuring edge and in-network media adaptation (for streaming)
Content-Aware Networking (CAN) as evolutionary approach towards the Future Internet
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
4
RESEARCH CHALLENGES Distributed adaptation decision-taking framework
Where to adapt? – at source, in-network, receiver, and combinations thereof
When to adapt? – at request and during delivery How often to adapt? – too often (risk: flickering), too seldom (risk:
stalling) How to adapt? – optimization towards resolution, framerate,
SNR (bitrate), accessibility, etc.; (too) many possibilities
Efficient, scalable SVC tunneling and signaling thereof Low (end-to-end) delay, minimum quality degradation, scalability
(# parallel sessions)
Impact on the Quality of Service/Experience (QoS/QoE) Trade-off (for certain use cases and applications); QoS QoE
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
5
ALICANTE ADAPTATION FRAMEWORK
FP7 ICT project "Media Ecosystem Deployment through Ubiquitous Content-Aware Network Environments"
Goal: New Home-Box layer and CAN layer with cross-layer adaptation enabling cooperation between providers, operators, and end-users
2 new virtual layers Home-Box (HB) Layer: enhanced home gateways CAN Layer: content-aware adaptation of SVC at
Media-Aware Network Elements (MANEs)
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
6
ALICANTE ADAPTATION FRAMEWORK
Michael Grafl et al.
CANCAN
Home-Box LayerHB
MANE
HB HB
HB
MANE MANE MANE
Autonomous System
End-to-End Multimedia Communication (MPEG-2, MPEG-4, AVC, SVC, ...)
...
...SVC (Layered-Multicast) Tunnel
HB
Autonomous System
Context-Aware
Adaptation
Dynamic, Network-Aware
Adaptation
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
7
ADAPTATION & SVC TUNNELING
Adaptation Decision-Taking Framework (ADTF) coordinating local adaptation decisions of modules at the content source; the border to the user (Home-Box); and within the network at MANEs
SVC (layered-multicast) tunnel Adaptation of scalable media resource at MANE At the border to the user (Home-Box), adaptation modules are deployed
enabling device-independent access
Key Innovations Better network resource utilization & maintaining
a satisfactory Quality of Experience Adaptation decision aggregation and propagation Distributed coordination with CAN layer for optimal adaptation & improved
bandwidth usage
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
8
TARGETED RESEARCH OUTCOMES
Guidelines for scalable media encoding/transcoding parameters (with SVC as example)
Guidelines for distributed adaptation decision-taking framework Enhancement of
decision-taking algorithm by exploiting active and passive monitoring SVC adaptation based on network load/conditions and QoS constraints using a
content-aware approach
Assessment of the performance and scalability (e.g., number of flows, flow traffic profile) computing resources utilized (e.g., CPU and memory) network related metrics (e.g., processing delay per flow, maximum achieved
bandwidth)
Mappings of network and device monitoring parameters Enable prediction of QoE; validation through subjective quality assessments
Holistic approach for in-network adaptation applying different adaptation policies per content-aware virtual network
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
9
PROPOSED INTEGRATED TEST-BED
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
10
SCIENTIFIC RESULTS ACHIEVED SO FAR
Achieved results Quality impact of SVC tunneling using MPEG-2 as
starting point: baseline for further research [3] Initial performance evaluations of SVC streaming and
real-time in-network adaptation [4] End-to-end QoS control including a model for
QoS-QoE mapping [5]
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
11
RC MODES FOR SVC TUNNELING Comparing rate control (RC) modes for SVC tunneling
Extended previous tests [3] to compare SVC tunneling for• Variable bitrate (VBR) constant quantization parameter (QP)• Constant bitrate (CBR)• Different codecs: bSoft, MainConcept
• SVC config: 4 medium-grained scalability (MGS) layers Procedure:
• Pixel-domain transcoding (PDT) from MPEG-2 to SVC• Transcode resulting bitstream back from SVC to MPEG-2• Measured Bjontegaard Delta (BD) Y-PSNR• Compared required bandwidths to MPEG-2 simulcast
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
12
RESULTS (1)
Sequence
bSoft (VBR) MainConcept (VBR)
MainConcept (CBR)
BD-PSNR [dB]
BD-bitrate
[%]
BD-PSNR [dB]
BD-bitrate
[%]
BD-PSNR [dB]
BD-bitrate
[%]
foreman -2.08 50.3 -2.03 53.7 -2.40 61.6container -1.57 38.2 -1.99 51.0 -2.91 66.9hall_monitor -0.75 22.6 -1.40 54.1 -1.82 73.6stefan -2.59 41.0 -2.09 32.1 -2.88 53.4Average -1.74 38.04 -1.88 47.7 -2.50 63.9
Michael Grafl et al.
Table 1: Bjontegaard Delta for SVC tunneling
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
13
RESULTS (2)
Michael Grafl et al.
Target Quality bSoft (VBR) MainConcept (VBR)
MainConcept (CBR)
SVC encoding
config
VBR [QP]
CBR [Mbps]
SVC tunnel [kbps]
MPEG-2 simulcast
[kbps]
SVC tunnel [kbps]
MPEG-2 simulcast
[kbps]
SVC tunnel [kbps]
MPEG-2 simulcast
[kbps]
Q1 16 3 5333 3041 3694 3454 3286 4721Q2 20 2 3446 2025 2418 2082 2242 3191Q3 24 1.5 2201 1452 1650 1277 1687 2093Q4 28 1 1438 1102 1132 900 1109 1287Average 3105 1905 2224 1928 2081 2823
Table 2: Comparison of required bandwidths for SVC tunneling vs. MPEG-2 simulcast
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
14
RESULT EVALUATION Less quality impact for VBR mode CBR mode: SVC tunneling more bandwidth
efficient than MPEG-2 simulcast (~26% reduction) Bandwidth efficiency of SVC tunneling depends
on number and configuration of SVC layers (mainly on quality of Base Layer)
Other scenarios: VBR mode SVC tunneling favorable to MPEG-2 simulcast if only server-side transcoding needed (i.e., client supports SVC)
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
15
CONCLUSIONS Research challenges and key innovations for edge
and in-network adaptation SVC tunneling Distributed Adaptation Performance evaluations of SVC streaming End-to-end QoS control & QoS-QoE mapping approach
CBR and VBR mode for SVC tunneling compared Integrated test-bed proposed Future work: HD content; subjective tests; integrate
QoS-QoE mapping; multi-video rate allocation
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
16
SELECTED LITERATURE[1] F. Pereira and I. Burnett, "Universal multimedia experiences for tomorrow,"
IEEE Signal Processing Magazine, vol.20, no.2, Mar. 2003.[2] European Commission, "ALICANTE, Annex I – Description of Work," FP7-
ICT-2009-4, Grant agreement no. 248652, 2009.[3] M. Grafl, C. Timmerer, and H. Hellwagner, "Quality Impact of Scalable
Video Coding Tunneling for Media-Aware Content Delivery," Proc. ICME’11, Barcelona, Spain, July 2011.
[4] N. Zotos et al., "Performance evaluation of H264/SVC streaming system featuring real-time in-network adaptation," Proc. IWQoS’11, San Jose, California, June 2011.
[5] B. Shao et al., "An Adaptive System for Real-Time Scalable Video Streaming with End-to-End QoS Control," Proc. WIAMIS’10, Desenzano Del Garda, Italy, Apr. 2010.
[6] G. Bjontegaard, "Improvements of the BD-PSNR model," ITU-T SG16/Q6, 2008.
Michael Grafl et al.
Distributed ADTF and SVC Tunneling for Edge and In-Network Media Adaptation
17
THANK YOU FOR YOUR ATTENTION!
Questions?
http://ict-alicante.eu/
http://itec.uni-klu.ac.at/~mgrafl
Michael Grafl et al.