See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/224088488

IPTV and video networks in the 2015 timeframe: The evolution to medianets

Article in IEEE Communications Magazine · January 2010

DOI: 10.1109/MCOM.2009.5350371 · Source: IEEE Xplore

CITATIONS READS

30 548

2 authors, including:

Ali C. Begen
Özyeğin University
145 PUBLICATIONS 4,415 CITATIONS

SEE PROFILE

All content following this page was uploaded by Ali C. Begen on 24 February 2016.

The user has requested enhancement of the downloaded file.

BEGEN LAYOUT 11/18/09 11:58 AM Page 68

CONSUMER COMMUNICATIONS AND NETWORKING

IPTV and Video Networks in the

2015 Timeframe:
The Evolution to medianets
Khalid Ahmad and Ali C. Begen, Cisco

ABSTRACT fiber-to-the-home (FTTH)-based technologies

will prevail over currently emerging access trans-
The rapid progress being made in technolo- mission technologies such as very high speed
gies that enable video content delivery over IP digital subscriber line (VDSL2) for wireline net-
networks to consumers has prompted predictions works, or even DOCSIS 3.0 for cable networks.
that the evolution of IP-based next-generation Despite the pitfalls that inevitably accompany
networks will be largely driven by video service gazing into technological crystal balls, it is some-
delivery requirements. This article surveys trends times useful to step back and attempt to predict
in the underlying technologies, extrapolating out where the underlying video network technology
to the 2015 timeframe, and drawing on the devel- direction will likely take us, and what the future
opments in standardization for IPTV, cable net- landscape will look like based on the existing
works, and the IP NGN. These evolution trends technology trends. This article attempts this
lead to the notion of a medianet as a useful way daunting task, based on surveying current trends
to think of all of the enabling video and multi- in video networking technologies for both IPTV
media technologies. A medianet is essentially an and cable networks.
IP network that is optimized to deliver video ser-
vices to any or multiple display devices, and uses
any of optical, cable, wireline, and wireless net-
THE FUTURE OF IPTV TECHNOLOGY
works for this purpose. The enormous growth and interest in IPTV ser-
vices over the past few years is a trend that is
INTRODUCTION anticipated to continue over the 2015 timeframe.
Service providers globally view IPTV services as
Several forecasts on the future of IP-based net- a major area of revenue growth and preserva-
working have indicated that the evolution of tion, so investments in this area of technology
future networks, specifically the IP next-genera- deployments are likely to increase substantially,
tion networks (NGN), will largely be determined as new programming options and content choic-
by the demands of video content distribution es become available to an increasing number of
and services like IPTV. The recent explosive consumers all over the world. Such a projected
growth in the demand for video-based social net- growth will have a number of implications for
working applications such as YouTube, as well the evolution of IPTV related technologies, as
as the availability of ever more high-definition outlined below.
TV (HDTV) content, seem to be clear evidence It is useful to bear in mind that what we com-
pointing in that direction. If more and better monly refer to as IPTV is actually a complex
quality video content is what consumers want interplay of a number of technology areas, draw-
and are willing to pay for, the content distribu- ing from TV broadcast and storage technologies,
tion networks must be able to satisfy this demand video coding (e.g., MPEG2 and H.264) and
more effectively so that service providers can encryption (e.g., digital rights management
reap the revenue benefits that accrue. [DRM]), IP transport, quality of service (QoS),
It is not just a matter of predicting the use of quality of experience (QoE), Web services, and
more and more raw bandwidth capacity, for so on. Advances in IPTV technology will be
which there will obviously be the need, particu- based on progress in any one or more of these
larly in the access network bottleneck. It is well areas and its relationship with the other compo-
known that video content is bandwidth hungry if nents of the overall IPTV system.
any level of quality of service is in question, so An example of this interplay is the use of
the bandwidth requirements of delivering video MPEG encoding of TV signals. The digitization,
content to consumers will scale in proportion to encoding, and compression of video images
the number of users multiplied by the number of have been extensively studied and standardized
sessions in play. Extrapolating these numbers to over the past years jointly by the International
the 2015 time frame suggests that eventually Telecommunication Union — Telecommunica-

68 0163-6804/09/$25.00 © 2009 IEEE IEEE Communications Magazine • December 2009

BEGEN LAYOUT 11/18/09 11:58 AM Page 69

tion Standardization Sector (ITU-T) and Inter- layered codecs such as the scalable video codec
national Organization for Standards/Interna- (SVC) may also find more interest in the broad- In parallel with these
tional Electrotechnical Commission (ISO/IEC) cast world. Significant breakthroughs in this area
advances in com-
Joint Technical Committee (JTC1/SC29/WG11, are likely to have major impact in the evolution
also called the Motion Picture Engineering of IPTV and cable technologies. pression technology,
Group [MPEG]). This work has resulted in the we must also
development of a series of MPEG specifications
that are in widespread use in virtually all exist- OVERCOMING THE consider the whole
ing IPTV deployments in some form. Most cur- area of bandwidth
rently deployed IPTV services use the ITU-T
BANDWIDTH CRUNCH
capacity evolution,
Recommendation H.262/ISO/IEC 13818-2 In parallel with these advances in compression
MPEG2 standard to digitize, compress, and technology, we must also consider the whole area and how it is
encode TV signals [1]. The MPEG2 transport of bandwidth capacity evolution and how it is managed to ensure
stream is then used to multiplex the encoded managed to ensure IPTV QoS/QoE require-
audio, video, and auxiliary data in a synchro- ments. Scarcely a week passes without some study IPTV QoS/QoE
nized fashion [2]. The transport stream is subse- or trade article predicting the astronomical requirements.
quently packetized and encapsulated into the IP increases in bandwidth that will be required to
packets (increasingly with an additional Real- satisfy the demands of video usage over existing
Time Transport Protocol [RTP] encapsulation IP networks. Traffic volumes in the order of tens
[3]) for delivery to consumer set-top boxes or and hundreds of exabytes (one billion gigabytes)
other end devices. and zettabytes (1000 exabytes) are proposed, as
The introduction of the new MPEG4 part 10 experts envisage exponential increases in demand
codec (also called ITU-T Recommendation for video content, not only over the public Inter-
H.264 or advanced video codec [AVC] [4]) over net, but also for subscribed video-on-demand
the past few years has generated interest for (VoD) or IPTV services delivered to set-tops,
future IPTV deployments. AVC achieves a fac- mobile devices, and PCs. A detailed industry fore-
tor of two or more in compression efficiency cast is available on the Visual Networking Index
over the earlier MPEG2 codec, with little or no Web site at http://www.cisco.com/web/go/vni.
perceived degradation in video quality. Conse- The exact range of the astronomical traffic
quently, a typical standard-definition TV growth numbers may be debatable, but it is rea-
(SDTV) program encoded by MPEG2 will con- sonable to anticipate huge traffic growth in the
sume a bandwidth of between 3–4 Mb/s, whereas 2015 timeframe for video services, and its conse-
encoded in AVC it will use only between 1–2 quent impact on transport technology, including
Mb/s. A similar reduction in bandwidth con- traffic management aspects of the IP NGN. The
sumption is also achieved for HDTV programs IP NGN infrastructure technology evolution
encoded with the AVC codec. (multiprotocol label switching [MPLS], carrier
The implications of these advances in video Ethernet, and optical technologies) was
encoding technologies on IPTV services are described in detail in the March 2008 issue of
obvious: the bandwidth savings will enable this magazine.
more or higher-quality TV content to be made Without delving into the details of such tech-
available over the typically bandwidth-con- nologies, it is nevertheless useful to examine the
strained access links, whether it is wireline, likely impact of video transport on the potential
wireless, or cable. AVC technology is already evolution of the access part of the IP NGN based
being deployed in many consumer electronics on existing trends. Currently, the vast majority of
devices such as HD cameras and players, as the video content is delivered to consumers either
costs of the AVC chipsets respond to economies over cable or various digital subscriber line
of scale. (xDSL) access links, with a small but rapidly
It can therefore be reasonably predicted that growing proportion over optical fiber technolo-
in the 2015 timeframe, AVC codec technology gies (e.g., FTTx such as passive optical networks
will supplant the existing MPEG2-based encod- [PONs]).
ing in most applications of IPTV deployments The likelihood that optical transport infras-
including set-tops and other consumer electron- tructures will become the technology of choice
ics devices. This is not to say that MPEG2 will for the last-mile video content delivery does not
somehow disappear, since some level of back- preclude advances in either cable or DSL access
ward compatibility with existing content will technologies, or, for that matter, in wireless
likely still be a strong requirement. However, transport technologies such as Long Term Evo-
increasing amounts of new content will become lution (LTE) or WiMAX (refer to the February
available in AVC to enable IPTV service pro- 2009 issue of this magazine for the most recent
viders to leverage the resulting bandwidth sav- developments in the LTE and WiMAX stan-
ings, or reciprocally, the cost savings of dards). However, optical fiber transport tech-
providing a given amount of TV content in less nologies, such as Ethernet PON (EPON) and
bandwidth. Gigabit PON (GPON) or direct FTTH, clearly
It is also feasible that new more efficient have the requisite bandwidth capabilities to
video compression and encoding techniques may deliver multiple channels of HDTV and VoD
become available by the 2015 timeframe. content with high QoE that consumers demand.
Research continues in both academia and indus- It is therefore not surprising that many ser-
try on better algorithms and mechanisms to vice providers and cable companies are exploring
improve video compression and encoding tech- the potential of optical transport systems for
nology, which may result in further bandwidth their video delivery networks in some form. It
savings. In parallel to these efforts, the use of would be reasonably safe to predict that in the

IEEE Communications Magazine • December 2009 69

BEGEN LAYOUT 11/18/09 11:58 AM Page 70

2015 time frame, we would witness large-scale number of initiatives underway to develop a uni-
No discussion of the deployments of optical access transport net- fying underlying architecture for middleware
works. There are already clear indications of applications, notably in the ITU-T as part of the
future evolution of such a trend, notably in countries such as Japan IPTV Global Standards Initiative (GSI),1 as well
IPTV technologies and South Korea, where high-speed broadband as in CableLabs2 and the Digital Video Broad-
seems complete access infrastructure has relatively high penetra- casting (DVB) Project3 under the auspices of the
tion, and the demand for HD video content is European Telecommunications Standards Insti-
without considera- already substantial and increasing rapidly. tute (ETSI).
tion of the Such a trend would be equally applicable for Although it is unlikely that all aspects of
the evolution of either wireline or cable-based IPTV middleware, such as DRM or user control
middleware that technologies, since the underlying optical trans- interfaces, will be standardized to the point that
effectively glues IPTV port technology would be able to support either open application programming interfaces (APIs)
services together, type of video encoding technology transparently. will be available for all capabilities, the primary
For the case of wireless video content delivery, objective of specifying a standardized framework
despite the fact that the bandwidth constraints tend to be more severe architecture for middleware is to enable defini-
this area is arguably for the relevant air interfaces, so inevitably the tion of some of the open APIs to enhance inter-
need for transcoding of video signals for mobile operability and thereby reduce development
the most difficult handsets arises for conversion between HD and costs. It is anticipated that in the 2015 time
to categorize SD sources. It seems evident that as mobile net- frame these standardization initiatives will bear
or predict. works evolve to 4G/LTE and WiMAX-based fruit in the availability of lower-cost middleware
technologies supporting larger bandwidth capaci- applications and more user-friendly control
ty, the demand for mobile video content tailored interfaces.
for handset availability will grow, providing ser-
vice providers unprecedented opportunities for
revenue generation.
THE ROLE OF IMS IN IPTV
There has been significant interest and discus-
sion on the relationship of IP multimedia subsys-
THE WAY OUT OF THE tem (IMS) with IPTV and video content delivery
technology in the recent past, resulting in exten-
MIDDLEWARE MAZE sive work in several standardization bodies such
No discussion of the future evolution of IPTV as ITU-T IPTV GSI, ATIS IIF, 4 and ETSI
technologies seems complete without considera- Telecommunications and Internet Converged
tion of the middleware that effectively glues Services and Protocols for Advanced Networking
IPTV services together, despite the fact that this (TISPAN).5 As developed by the Third Genera-
area is arguably the most difficult to categorize, tion Partnership Project (3GPP), it is well known
or predict in terms of discernable technological that IMS was intended to provide an IP-based
trends. This is partially because IPTV middle- control plane architecture for the evolution of
ware is often the engine for service differentia- mobile networks, to enable session initiation,
tion of video offerings by service providers as authentication, authorization, and billing/charg-
well as equipment vendors, and hence resistant ing interfaces based on the protocols developed
to standardization efforts, and partially because by the Internet Engineering Task Force (IETF)
IPTV middleware is itself a somewhat nebulous such as the Session Initiation Protocol (SIP) and
area of capabilities. DIAMETER. Many service providers plan to
Despite these uncertainties, it is important to use an IMS control framework for evolution of
understand that developments in middleware their wireline and cable networks as well, so IMS
technologies will inevitably impact the nature of control of IPTV needs consideration.
future IPTV services significantly, so it is useful Most existing deployments of IPTV and VoD
to summarize some underlying architectural networks do not use IMS interfaces, so IMS is
aspects to elucidate potential trends. not an essential ingredient of IPTV or VoD.
The IPTV middleware typically comprises all However, many service providers plan to deploy
those aspects of the service that enable con- IMS for their wireless and wireline networks, so
sumers to access content from the provider and it is natural for them to require a common plat-
manipulate it for ease of viewing and/or storage. form for control and charging/billing for triple-
It may also include aspects of consumer entitle- and quad-play service offerings. With this in
ment/authorization for content use, the ability to mind, scenarios for evolution of IPTV/VoD ser-
store it or playback in real time as well as with vices toward IMS-based control have been
so-called trick modes, such as fast forward, pause, described by the ITU-T in Recommendation
1http://www.itu.int/ITU- or other manipulations of the video stream. The Y.1910, using the well established concept of an
T/gsi/iptv middleware is also responsible for capabilities interworking gateway between the existing
such as program discovery and selection, as well deployments and IMS-based IPTV/VoD
2http://www.cablelabs. as the type of display of the electronic program domains.
com guide (EPG) and user control interface. Depending on what specific IMS service
As may be expected, aspects such as the EPG needs to be supported, interworking of the sig-
3 http://www.dvb.org/ and user control interface are always likely to be naling messages via the gateway provides the
groups_modules/ very provider-specific and key to service differ- requisite protocol translation between the IMS
technical_module/tmipi/ entiation and packaging. However, technical domain and existing IPTV/VOD deployments.
developments that enable these aspects to This enables current deployments to support
4 http://www.atis.org/iif become more user-friendly for mass market IMS services as they are introduced while pro-
appeal will inevitably gain more widespread viding a graceful evolution path to introduction
5 http://www.etsi.org/tispan usage as IPTV services mature. There are a of the IMS control framework.

70 IEEE Communications Magazine • December 2009

BEGEN LAYOUT 11/18/09 11:58 AM Page 71

It is important to recognize that for the case companies to migrate from conventional analog
of video services, the IMS control plane is speci- programming to digital to accommodate more The evolution toward
fied to provide only the IPTV or VoD session long-tail content and HD programming. In the
initiation, authorization, and charging functions. 2015 time frame, based on current trends, it is all-IP-based service
All the other middleware and video stream con- safe to assume that substantially more video con- delivery will
trol capabilities like channel change, trick modes, tent will be distributed in digital format over transform virtually all
DRM, and encryption/encoding remain the HFC plants as consumers increasingly buy into
same. While it is feasible that this may change as the advantages of digital TV. aspects of cable
new control technologies are proposed, it seems Just as significantly, another major trend is networks, from the
likely that the well proven existing protocols likely to be the transition to all-IP-based service
such as Real Time Streaming Protocol (RTSP) delivery of video content. To some extent this development of
[5] and Internet Group Membership Protocol trend is already in evidence, with the increasing hybrid set-top boxes
(IGMP) [6] will continue to provide the underly- push to IPTV services, enabling the promise of to modular Cable
ing control plane architecture for video delivery enhanced interactivity and flexibility for content
in the 2015 timeframe. It can be expected that manipulation. However, it is anticipated that a Modem Termination
these technologies will be enhanced to support key driver for the transition toward all-IP service Systems (CMTS) and
additional video service requirements as they delivery will likely be the evolution of connected
arise, but the underlying control protocols are home technologies, which are primarily based on video headends.
unlikely to change radically. IP connectivity. An example of such an approach
to connected home IP networking is the increas-
ing use of architectures based on Multimedia
THE EVOLUTION OF over Cable Alliance (MoCA), 6 Digital Living
Network Alliance (DLNA),7 and Universal Plug
CABLE NETWORKS and Play (UPnP) Forum8 specifications for ubiq-
Many of the technology evolution trends uitous sharing of video content in residential
described above apply just as well to the devel- environments.
opment of cable networks as to conventional Essentially, the transition to IP video service
wireline or wireless networks. For example, the delivery is all about leveraging the flexibility
drive to enhance access bandwidth to enable made possible by IP connectivity. As this aspect
delivery of more HD content is likely to lead to is increasingly adopted in the consumer environ-
demand for optical transport for last-mile ment, the drive toward transitioning to IP-based
deployments even for cable operators, a trend service delivery will likely spread to the cable
already in evidence. distribution networks and in some cases all the
Cable networks have specific technology way back to the content provider. Such an evolu-
aspects that need to be factored in when evaluat- tion enables the advantages of IP connectivity
ing evolution trends for the 2015 timeframe. A for applications such as targeted advertising and
primary element of these is the evolution of cur- local content insertion, in addition to the poten-
rently deployed DOCSIS 1.1 and 2.0 technolo- tial for lower-cost transport and distribution,
gies toward the significantly higher-speed from both the capital and operational expendi-
DOCSIS 3.0 technology (and its potential future ture standpoints.
enhancements, which may lead to DOCSIS 4.0), The evolution toward all-IP-based service
which is capable of supporting upstream data delivery will transform virtually all aspects of
rates of 50–100 Mb/s or more for cable con- cable networks, from the development of hybrid
sumers. DOCSIS 3.0 also enables significantly set-top boxes to modular cable modem termina-
higher downstream data rates by a factor of tion systems (CMTSs) and video headends. To
three or more, although the actual data rates some extent such a transformation has already
available to consumers vary depending on the begun with the interest in cable IPTV, so by
number of customers served. The trend for high- 2015, it can be anticipated that most of the
er-speed Internet access in cable mirrors that major cable service providers will be deploying
already occurring for wireline/optical, and is like- some form of IP-based video delivery systems as
ly to accelerate as cable operators and IPTV ser- part of their cable networks.
vice providers compete to differentiate their
offerings with more HD content.
In addition to significantly increased deploy- ADVANCED (OR TARGETED)
ments of DOCSIS 3.0 technology in the 2015
time frame, we are also likely to see the further
ADVERTISING
transformation of the existing hybrid fiber coax New service provider offerings present signifi-
(HFC) secondary distribution networks toward cant opportunities and challenges for delivering
an increasing optical fiber component, in keep- advertising that matches delivered content with
ing with the trend to FTTx. The increasing use its audience. A solution that enables monetiza-
of FTTH as a migration path for conventional tion of new services via advertising is essential
cable networks is likely even though a cable since it not only justifies the service provider
plant has significantly more potential bandwidth investments in deploying a new service, but is
capacity than a conventional twisted pair access likely to be key in obtaining the necessary car-
plant. riage rights from the content providers. The 6http://www.
There are other major technology trends that addressability of the devices such as IP set-tops, mocalliance.org/
are likely to play increasing roles in the evolu- PCs, and mobile devices creates an ideal plat-
tion of cable networks. An obvious trend is the form for the targeted advertising media buyers 7 http://www.dlna.org
increasing penetration of digital TV content, value highly.
resulting from the push by most major cable In advertising, media buyers decide on the 8 http://www.upnp.org

IEEE Communications Magazine • December 2009 71

BEGEN LAYOUT 11/18/09 11:58 AM Page 72

developed a set of standards known as SCTE

130 [7]. As depicted in Fig. 1, this new standard
Inventory
defines a set of logical services that implement
system Content Back office extensible XML-based messaging interfaces to
store
communicate placement opportunities, content
and subscriber metadata, placement decisions,
and event status necessary for accountability
Placement op. Content Subscriber measurements. Industry adoption of SCTE 130
information information information is expected to ease integration between advertis-
SCTE 130-5 SCTE 130-4 SCTE 130-6
ing subsystems, enabling the rapid evolution of
interoperable multivendor advanced advertising
solutions.
Placement Campaign
SCTE 130 is designed to be highly flexible; in
Ad Ad
execution management decision management fact, no specific architecture, data model, or
SCTE
130-3 physical topology is imposed on the set of core
services. For the most part each defined service
has a public standard interface with which it
Figure 1. SCTE 130 services in an advanced advertising system. The figure is communicates with other services, and a private
reproduced from [7]. interface, which integrates with the vendor-/
provider-specific ecosystem components. Fur-
thermore, while initially targeted at ad place-
details of the ad campaign that define what audi- ment within the television domain, by virtue of
ence to reach with the desired message, and its IP-based control plane and transport domain
choose the period of the campaign and frequen- independence, SCTE 130 can easily be extended
cy. The service provider functions as the seller, to broadband and mobile content delivery plat-
who offers the ability to deliver the advertising forms, enabling multiplatform inventory buys,
message on its platform in association with the campaign management, ad placement, reporting,
content. In traditional broadcast 30 s advertise- and reconciliation.
ments appear interstitially during a particular
show, while in emerging addressable TV deploy-
ments the ad could be an ad on comedy genre
INTERNET TV
content delivered to viewers between 18 and 29 The term Internet TV is used to describe the
years of age in the New York metropolitan delivery of TV programming over the public
region. This is made possible by dynamic ad Internet, typically to personal computers as
decisioning. streamed or downloadable video content. A
In advanced advertising the campaign manag- growing number of broadcasting organizations
er, typically a software tool hosted by or on such as BBC and Swedish TV, as well as inde-
behalf of the service provider, needs three class- pendent video content providers are providing
es of information: direct access to TV programming over the Inter-
• Placement opportunities are temporal and/or net. This approach is also often referred to as
spatially constrained advertising instances over-the-top (OTT) video, since it essentially
defined by content owners and/or distribu- uses the Internet as a transport pipe to deliver
tors. the content. In some deployments solely server-
• Context refers to information that describes based delivery schemes are used, whereas in oth-
the content (e.g., title, genre, and rating). ers a hybrid delivery scheme is preferred to
• Audience is the available viewer and device simultaneously benefit from dedicated servers as
information, including demographic, geo- well as other peers in the system. The latter is
graphic, behavioral, and so on. achieved by running streaming-optimized forms
This information is used to provision the of the well-known peer-to-peer (P2P) file trans-
campaign manager and for an ad decision when fer protocols.
a placement request is made in response to a The OTT delivery of television is likely to
specific event such as the start of a VoD session, grow in popularity in the 2015 timeframe, as
or a viewer pausing the DVR or tuning to the more broadcasters and independents compete
on-screen EPG. The ad insertion system then for screen time or specialized content and niche
splices the ad into the stream and presents it in interests. Internet TV is a topic now being
some form as a companion to the content. Final- addressed by the European DVB Project for
ly, confirmation of ad delivery is reported back potential standardization, a move that may fur-
to the campaign manager for invoice and billing ther enhance its popularity as wider interoper-
purposes and potentially captured into a data ability and consequent lower costs are made
warehouse used to track measurement as well as feasible via economies of scale. It has been sug-
analyze viewer behavior, which can then be used gested that the rapidly increasing popularity of
to refine the audience data to enable more pre- Internet TV, as evidenced by the widespread use
cise targeted ad decisions. of social networks such as YouTube, and the
An advanced advertising solution has the amount of innovation the OTT providers are
potential to be both sophisticated and complex. pursuing even threaten to supplant conventional
To enable interoperability between components modes of TV programming such as cable or
and establish sufficient scale to attract both IPTV. Whether this can happen in the time-
advertisers and content providers, the Society of frame envisaged here remains an open question,
Cable Telecommunications Engineers (SCTE) but given the increasing popularity of large-
Digital Program Insertion (DPI) Working Group screen TVs and home theater systems capable of

72 IEEE Communications Magazine • December 2009

BEGEN LAYOUT 11/18/09 11:58 AM Page 73

providing subscribers a rich viewing experience,

it seems relatively unlikely that conventional Applications
viewing habits will change much in this time- Broadcast Telepresence Over-the-top
frame.
Today, a major struggle in deploying high- On demand Enterprise Mobile
quality OTT services is the disagreement in rev-

Application aware
Network aware
enue sharing between the OTT providers and Media delivery and control
the Internet service providers. An OTT service
Middleware Ad engines
offering HD content can easily chew up a lot of Data center
resources from the service provider’s network. virtualization
Headend Content delivery
Although poorly engineered P2P technologies
can result in significant additional costs to the IP network
service provider, users generally access the net-
Core Access
work indefinitely for a flat monthly fee, and, if Data center
applicable, only pay an additional premium to infrastructure
Aggregation Home/CPE
their OTT providers. This somewhat unbalanced
revenue sharing naturally discourages service
providers from allowing any OTT service to run Figure 2. medianet-enabled IP NGN framework.
in their networks. To enable the true potential
of OTT services, both sides must work together
and reach an agreement on how to share the end-to-end video and rich-media services evolu-
service costs as well as the revenues from premi- tion. It encompasses IPTV, cable, content deliv-
ums and ads. ery networks, and connected life at home as part
There are currently industry-wide efforts to of a media-aware IP NGN-based network that is
bring Internet service providers and OTT pro- optimized to support all envisaged rich media
viders together to develop network-friendly P2P services for the future. The medianet essentially
file transfer and distribution protocols. One such describes the evolution of an IP-based NGN
effort is the P4P Working Group 9 at the Dis- toward a rich-media-aware optimized network
tributed Computing Industry Association that enables virtualization of capabilities to
(DCIA). Another one is the Application-Layer enhance the scalability of the network, as well as
Traffic Optimization (ALTO) Working Group10 the potential to commercialize a range of multi-
at the IETF. The current work in these working media services, including video delivery (IPTV,
groups is focused on developing a set of opti- VoD, cable TV, and OTT services), enabling
mized protocols that will reduce the average service providers to transform from being simply
download times for users as well as the transport content providers to being overall experience pro-
costs for service providers. Early results from the viders.
preliminary field trials are promising [8]. Given The medianet vision is built on the recogni-
that many participants from the service provider, tion that video services and technology evolution
content provider, and P2P industries are playing have the potential to transform consumer experi-
an active role in these groups, we can anticipate ence by effectively revolutionizing the IP NGN
more widespread deployment of these new file into a media-aware network that manages scale
transfer and distribution protocols by 2015. and complexity by incorporating a high level of
This close interaction between service and virtualization, and enables generation of new
OTT providers as well as content providers is an revenue streams (monetization) by service veloc-
important development for the traditional con- ity enhancement. For video applications, what
tent delivery network (CDN) providers. Typical- this essentially implies is that the IP NGN needs
ly, service providers allowed one or more CDN to be endpoint-aware as well as media-aware, so
providers to physically operate within their net- services that rely on personalization, interactivi-
works to enhance the experience of their cus- ty, or social networking can be brought into play
tomers. Content providers also used to work more effectively. As shown in Fig. 2, the IP
with CDN providers to make their content avail- NGN architecture can be enhanced by the notion
able at as many places as possible. However, of a media-aware medianet to encapsulate a
with the introduction of newer service models vision for the future evolution of rich-media
and distribution channels, service providers can based services.
transform their networks in such a way that they Figure 2 shows that, as for conventional IP
can deliver the content directly from the content NGNs, a medianet relies on IP-based transport
providers. While this transformation may not be layers for media/content delivery, including
rapid, and existing CDN providers will evolve essential NGN attributes such as QoS, security,
solutions to stay competitive, by 2015 we can and survivability. Overall service control capabil-
expect some large service providers to collabo- ities are part of the media delivery and control
rate closely with content providers. layers, which include functionalities to support
middleware, headend processing, bandwidth
management, Tru2way services, and so on. As
EVOLVING THE IP NGN INTO A for the IP NGN, the application layer incorpo-
rates functionalities to support services such as 9http://www.dcia.info/
MEDIANET VISION telepresence, voice, multiscreen capabilities, and activities
The notion of a medianet was essentially envi- mobility. It should be noted that this layering is
sioned as a vehicle to encompass all of the evo- not intended to depict any particular implemen- 10http://www.ietf.org/
lutionary trends summarized in this article in a tation or deployment, but is simply a functional dyn/wg/charter/alto-
comprehensive architecture for the future of representation of typical medianet capabilities. charter.html

IEEE Communications Magazine • December 2009 73

BEGEN LAYOUT 11/18/09 11:58 AM Page 74

CONCLUDING REMARKS [3] IETF RFC 3550, “RTP: A Transport Protocol for Real-Time
Applications,” July 2003; http://www.ietf.org/rfc/
The growing rfc3550.txt
The evolution of IPTV and cable network tech- [4] ITU-T Rec. H.264, “Advanced Video Coding for Generic
consumer demand nologies capable of providing rich media and Audiovisual Services,” Mar. 2009.
for more and varied video services to multiple display devices, fixed [5] IETF Internet draft, “Real Time Streaming Protocol 2.0
(RTSP),” July 2009; http://tools.ietf.org/html/draft-
HD content will or mobile, can be viewed as leading to the emer- ietf-mmusic-rfc2326bis-22
gence of medianets, a concept that captures how [6] IETF RFC 3376, “Internet Group Management Protocol,
result in huge these capabilities can be architected within the Version 3,” Oct. 2002; http://www.ietf.org/rfc/rfc3376.txt
evolving IP NGN deployments worldwide. The [7] ANSI/SCTE 130, “Digital Program Insertion — Advertis-
demand for access ing Systems Interfaces,” 2008; http://www.scte.org/
growing consumer demand for more and varied content/index.cfm?pID=1485
network bandwidth, HD content will not only result in huge demand [8] H. Xie et al., “P4P: Explicit Communications for Cooper-
and improved quality for access network bandwidth, but also improved ative Control Between P2P and Network Providers,”
quality of experience, and ease of both temporal P4PWG Whitepaper, May 2007; http://www.dcia.info/
of experience and documents/P4P_Overview.pdf
and spatial control over the content available.
ease of both These challenges are likely to trigger signifi-
temporal and spatial cant changes in IP network deployments as we BIOGRAPHIES
look out to 2015, with increased penetration of KHALID AHMAD ([email protected]) is a technical leader
control over content. optical fiber to homes, higher-speed mobile data in the Industry and Technology Marketing Group at Cisco,
working on video, IPTV, and home networking technology
networks, as well as more sophisticated content standardization. He has over 25 years of R&D experience in
control and middleware architectures evolving telecommunications, data networking, and solid state elec-
from those currently being used for IPTV tronics in both project leadership and management roles in
deployments and cable networks. It is useful to Canada. He has chaired several technical groups for the
development of international standards on networking
think of all these capabilities as a part of video- technologies in the ITU-T, ATM Forum, FR Forum, and IETF.
optimized medianets that are designed to deliver He is the author of a book, Sourcebook of ATM and IP
the optimum video experience to consumers Internetworking (Wiley-IEEE Press, 2001). He holds a B.Sc.
using the attributes of IP NGNs and evolving (hons) in electronic engineering, an M.Sc. in solid state
electronics from the University of Manchester, and a Ph.D.
IPTV and cable technologies surveyed in this in solid state physics from the University of Cambridge,
article. United Kingdom.

ACKNOWLEDGMENTS A LI C. B EGEN [M] ([email protected]) is with the Video

and Content Platforms Research and Advanced Develop-
We thank our colleagues at Cisco for their con- ment Group at Cisco. His interests include networked
tributions to this article. entertainment, Internet multimedia, transport protocols,
and content distribution. He is currently working on archi-
REFERENCES tectures for next-generation video transport and distribu-
tion over IP networks, and he is an active contributor in
[1] ITU-T Rec. H.262, “Generic Coding of Moving Pictures the IETF in these areas. He holds a Ph.D. degree in electri-
and Associated Audio Information: Video,” ISO/IEC cal and computer engineering from Georgia Tech. He
13818-2:2000, Feb. 2000. received the Best Student-Paper Award at IEEE ICIP 2003,
[2] ITU-T Rec. H.222.0, “Generic Coding of Moving Pictures and the Most Cited Paper Award from Elsevier Signal Pro-
and Associated Audio Information: Systems,” May cessing: Image Communication in 2008. He is also a mem-
2006. ber of the ACM.

74 IEEE Communications Magazine • December 2009

View publication stats