OptiBand

Coordinator:Yossi BarsheshetDirector, Strategic AlliancesOrckit-Corrigent

Yossi Barsheshet, Orckit Communications Ltd Partners:

◦ Telecom Italia S.p.A (IT)◦ Thomson Grass Valley France SA (FR)◦ Irdeto Access B.V.(NL)◦ Lambdastream Servicios Interactivos SL (SP)◦ FTW Forschungszentrum Telekommunikation Wien Betriebs-GMBH (AT)◦ Fraunhofer Gesellschaft zur Foerderung der Angewandten Forschung E.V (DE)◦ Valtion Teknillinen Tutkimuskeskus (FI)◦ Optibase (IL)◦ Universidade da Coruña (SP)◦ Corrigent Systems Ltd (IL)◦ ARTTIC Israel International Management Services 2009 LTD (IL)

Duration 30 months: January 1st, 2010 to June, 2012 Funding scheme STREP

◦ Total Cost: € 4 539 492◦ EC Contribution: € 3 038 395

Contract Number: INFSO-ICT-248495

February, 2010OptiBand - Concertation Meeting 2

Playing in IPTV network◦ Technology may be expended toward Internet TV

Optimizing the bandwidth of IPTV for the delivery of multiple HD streams over ADSL◦ Providing multiple HD streams over DSL access


Worldwide IPTV service revenue

is expected to increase more

than x3

◦ 2009: 12B$

◦ 2013: 38B$

Limited DSL bandwidth◦ ADSL2+: few SD◦ VDSL2: one or two HD

A typical home requires:◦ Multiple STB per home◦ Growing demand for HD content


Very high cost of deploying FTTx◦ Challenging ROI◦ Need to get more of existing infrastructure under the

current economical climate

More HD STB per home ◦ ADSL2+ (10Mbps): 2 HD STB◦ VDSL2 (20Mbps): 4 HD STB

Scalable solution ◦ Centralized video processing at head-end

Per channel – not per subscriber – cost structure◦ Distributed video content aware packet drop at the network

edge Very low per-subscriber cost

Leverage the existing access infrastructure Integration with video distribution infrastructure

◦ STB, middleware, video servers, encryption, head-end equipment


OptiBand Head End Unit◦ Processing video content and tag the video

stream OptiBand Network Unit

◦ Performing content-aware packet drop over DSL◦ Personalization

Algorithms for video analyzing and data drop◦ Maintain QoE is mandatory

Maintaining echo-system◦ Encryptors, STB, Middleware, Video Server



SMART table: Providing more HD video over existing access technology

Specific3 HD streams for a single 15Mbps ADSL line (normally 3 HD streams require 24Mbps). Each video stream will suffer only minimal degradation of QoE (SSIM greater than 0.7 for 70% of the time, or a better – subjective – QoE measurement selected by WP2 and described in D2.3)

MeasurableThe PoC (Proof Of Concept) will demonstrate the operation of 3 HD streams over 3 TV screens fed by a single 15Mbps ADSL line. QoE degradation will be measured by SSIM metric, using measurement equipment (or a better – subjective – QoE measurement selected by WP2 and described in D2.3).

AchievableThis objective is achievable with the above measures of success as indicated by preliminary analysis and testing done in Corrigent labs.

RealisticThe system prototypes (D3.4.x & D4.5.x) will prove that the technology is realistic and can be built using current technology hardware components.

Time boundThis objective will be tested by simulation (see T3.3 description) of the content aware data dropping algorithm in M12 and later on tested on prototypes (D3.4.2 & D4.5.2) by the POC testing (WP8) in M27-M30.


SMART table: Providing Scalable Personalized Quality of ExperienceSpecificThis objective will be achieved through the data dropping algorithm that will allow adjusting, in real time, the

level of data dropping of each video stream according to the needs of the subscriber at any given time (e.g. adjustment to the number of TV screens that are switched on and to the type of video content (HD or SD) displayed on any of the screens).

MeasurableThe PoC will demonstrate that different levels of data dropping will be applied to different households, according to their individual preferences.

The test plan (D8.1) will include specific test cases to measure this objective. Examples of such test cases are:

1.Two TV sets are showing full bandwidth HD video. A third TV set is switched on. The OptiBand system responds by reducing bandwidth on each stream and enabling the 3 HD streams to the 3 TV sets

2.One of the 3 TV sets is switched off. The OptiBand system responds by restoring full HD bandwidth to the 2 remaining sets

3.The video quality is monitored by SSIM test equipment. In addition, the viewers of the TV sets are interviewed and asked if they have perceived any degradation of the video quality during the above 2 tests.

AchievableThis objective is achievable with the above measures of success as indicated by preliminary analysis and testing done in Corrigent labs.

RealisticPrototype (D3.4.x) will prove that the technology is realistic and can be built using current technology hardware components.

Time boundThis objective will be tested by simulation of the content aware data drop algorithm (D3.3.1) at M12 and later on tested as described above by the POC testing (WP8) in M27-M30.


SMART table: Providing content aware data drop algorithmSpecificThe content aware data drop algorithm will be able to reduce the data rate of a HD stream by

33% on the average. The measurement is “on the average” since the data dropping may vary during the time because of changing in preferences and video streams. Therefore only an average measure has any significance.

The bandwidth reduction will be achieved with minimal degradation of QoE (SSIM greater than 0.7 for 70% of the time, or a better – subjective – QoE measurement selected by WP2 and described in D2.3).

MeasurableThe PoC will measure the bandwidth of a HD video stream before and after the data dropping done by D3.4.x. This measurement will show a reduction of 33% in the bandwidth.

SSIM for the reduced bandwidth will be greater than 0.7 for 70% of the time (or a more accurate – subjective – QoE measurement selected by WP2 and described in D2.3 may be used).

AchievableThis objective is achievable with the as indicated by preliminary analysis and testing done in Corrigent labs.

RealisticThe implementation of the algorithm by the prototypes (D3.4.x and D4.5.x) will prove that the algorithm is realistic and can be built using current technology hardware components.

Time boundThis objective will be tested by simulation of the content aware data drop algorithm in M12 and later on demonstrated by the prototypes (D3.4.x and D4.5.x) at the POC (WP8) in M27-M30.


SMART table: Achieving the project objectives while maintaining the existing IPTV ecosystem intact

SpecificAll the other project objectives will be achieved without any need to upgrade the existing IPTV equipment such as home equipment (STB) or last mile equipment (DSLAM).

MeasurableThe PoC will demonstrate the achievement of the other project objectives without the need to upgrade any existing DSLAM or STB equipment

AchievableThis objective is achievable with the as indicated by preliminary analysis and testing done in Corrigent labs.

RealisticThe implementation of the algorithm by prototypes (D3.4.x and D4.5.x) will prove that the algorithm is realistic and can be built using current technology hardware components.

Time boundThis objective will be tested by simulation of the content aware data drop algorithm in M12 and later on demonstrated by prototypes (D3.4.x and D4.5.x) at the POC (WP8) in M27-M30.


WP1

-R

equirem

ents

(T

IS)

WP3 - Network Solution (CSL)

WP5

-S

yste

m A

rchitectu

re

(OP

T)

WP4 - Head end solution (HHI)

WP2 – Research of QoE Metrics (FTW)

WP7

-Sys

tem

Inte

gra

tion

(HH

I)

WP8

-Liv

e T

est (T

IS)

WP10 - Project Management (OCL)

WP9 - Dissemination (AIL)

WP6 - Internet TV Research (VTT)

Require-ments Research Development Integration POC

Project phases:

Documents

OptiBand