Cisco Traffic Forecast

I have posted an article on Cisco's latest IP traffic forecast on my Telco 2020 blog. Cisco's forecast highlights the growing importance of video traffic on IP networks.

Cisco IPTV Video Delivery Test

Light Reading published an 18 screen article describing a test performed by EANTC of the ability of a Cisco infrastructure to deliver high quality IPTV service. The conclusion was that the Cisco network performed very well.

This is a very interesting effort. It does a good job of characterizing the performance of this network delivering broadcast/multicast service with a modest amount of video on demand. This are realistic assumptions for most of today's IPTV networks.

However, I am aware of IPTV networks that support network DVR services that are experiencing 80 percent penetration of video on demand services during peak periods. This puts the network into an entirely new dimension. It would have been much more interesting if such a heavy video on demand load was tested as well. This is where IPTV networks are going.

Time Warner Backing Off Usage Caps

Light Reading has published an article that states that Time Warner has shelved plans to extend its trial of usage caps currently underway in Beaumont, Texas to Rochester, N.Y.; Greensboro, N.C.; and San Antonio and Austin, Texas. Take a look a my previous posting to get information on the Beaumont trial.

Apparently, Time Warner is bowing to public and political pressure. It continues to believe in usage caps and believes that it needs to better educate its users.

Personally, I think usage caps and usage charging will not be accepted and is a tool that Time Warner is using to quash competition from over the top Internet TV services.

Cox Traffic Management Strategy

I have made a post on my Telco 2020 blog on Cox Cable's strategy for managing traffic on its broadband network. Its approach makes a lot more sense than the monthly usage caps that other operators are adopting.

Upgrading TCP for the Heavy User

I have made a post on my Telco 2020 blog that may be of interest. It discusses how changes to TCP will help the Internet handle traffic from bandwidth hogs and provide better service to casual users.

I think that this is a potential solution for the Net Neutrality problem, but I am no sure that it will help and IPTV network dominated by unicast IPTV traffic.

Telco IPTV Network Architecture

I just published a post on my Telecom 2020 blog that you may find interesting. It discusses the new IPTV networking announcements by Alcatel-Lucent and Cisco and their limitations.

Cisco Introduces IPTV Enabled Edge Router

Cisco introduced its Aggregation Services Router (ASR) 9000 that will provide significant increases in performance over its work horse 7600 routers with a total of 6.4 terabits per second of capacity.

The ASR 9000 incorporates the Cisco Advanced Video Services Module (AVSM) with 4 terabyts of flash RAM storage that offers content caching, ad insertion, fast channel change and error correction of video streams. Fast channel change and onboard error correction for both unicast and multicast video traffic helps ensure that errors can be detected by any set-top box and retransmitted within milliseconds.

The AVSM capabilities of the AVSM provide similar capabilities to Alcatel-Lucent's TPSDA 2.0 approach that provides similar capabilities. Both approach will require support from the IPTV middleware to get their full effect.

The 4 TB of storage will be enough to store about 1,000 hours of HD content. This will be enough for the most popular video on demand assets but not for the long tail of for an extensive NPVR service. Cisco will have to add a lot more storage to really offload video traffic from the network as it is currently claiming. I will write a column for my Telco 2020 blog that will discuss this issue in more detail in a couple of days. I just don't think that a distributed approach will hold up over the long term.

Zeugma Publishes Broadband Survey

Zeugma published the results of an IDC survey of U.S. 787 consumers that was conducted in August 2008. The survey produced the following conclusions:

On Premium Bandwidth Services:

• 94 percent of respondents saw value in broadband services that dynamically allocate premium bandwidth for certain types of traffic, such as video, VOIP, gaming, and telecommuter VPNs.
• 54 percent would actively seek to change service providers if another offered this service; 26 percent would be willing to pay additional fees for premium bandwidth services.

On Bandwidth Capping and Metering

• 81 percent do not like the idea of establishing a bandwidth cap and charging for use above the cap.
• 51 percent would try to change service providers if their provider imposed bandwidth caps. Interestingly, light and moderate users are even more opposed to capping and metering than are heavy users.
• 83 percent either do not know what a gigabyte is or have no idea how many they use.
• Only 5 percent said unequivocally that “those who use more should pay more

There is a growing gap between the broadband providers and their customers. The Telcos and the cable companies look at the issue from their perspective and are trying to limit traffic growth by punishing power users. This is bad business and policy. Why alienate your best customers?

There is a hidden agenda here. While these caps will not affect many data users, they significantly limit the amount of TV content that can be delivered over the top. The Telcos and cable companies want people to use their own TV services rather than viewing over the top content. This will become a major issue when it becomes apparent to people who want to watch Netflix or CinemaNow content regularly on their TV.

Alcatel-Lucent TPSDA 2.0 Steals Microsoft's Thunder

Alcatel-Lucent has introduced IPTV application awareness into its Triple Play Services Delivery Architecture (TPSDA) 2.0. It will add flash storage to its edge routers, its switches, and its access nodes that will support fast channel change, packet retransmission, and targeted ad insertion.

• For fast channel change the last few minutes of every multicast channel will be stored in the network and the new channel will be started immediately, eliminating the latency that typically occurs today. Alcatel-Lucent said that HD channel change time will be reduced from several seconds to well under a second.
• For packet retransmission the network will store enough packets to immediately service the packet retransmission request.
• For ad insertion targeted ads will be stored and inserted in the network.

If the service provider uses Alcatel-Lucent access nodes, these functions can be provided in the access network. These functions can also be provided in the Alcatel-Lucent switches or edge routers for service providers using Alcatel-Lucent's TPSDA 2.0 architecture.

Alcatel-Lucent has also discussed plans to increase the amount of storage in these network elements in order to support rewind TV and video on demand delivery.

TPSDA 2.0 can support these functions for service providers using access nodes from other manufacturers by implementing them in the switches and edge routers.

Microsoft has not announced any support for these network capabilities.

This is an interesting announcement. It enables middleware companies other than Microsoft to support fast channel change. This includes both the Alcatel-Lucent MiView and OMP middleware as well as other middleware from companies such as Thomson and Orca. This approach to fast channel change will eliminate one of Microsoft's most competitive features.

This approach will give service providers new flexibility in managing the bandwidth demands from an IPTV service, especially as more and more on demand and personalized services are provided. I think the flexibility that it gives in placing storage in the network will improve the ability of service providers to optimize their networks.

Cablevision Wins NPVR Ruling

Lightreading published an article about Cablevision winning an appeals court ruling that supports its NPVR service. The ruling allows Cablevisiion to offer its NPVR service that was blocked by a law suit by major U.S. TV networks and film studios. Cablevision has not announced any plans to restart the service.

This ruling is subject to an appeal to the U.S. Supreme Court. NPVR is a key feature for the cable and Telco IPTV services in the U.S. NPVR is needed to bring the personalization that viewers get from the Internet to the TV. It will also have a profound effect on the networks that support IPTV services.

Nokia Siemens Network Vision

Nokia Siemens held a webinar yesterday discussing its vision for Telco networks. It made a couple of interesting observations. One is that traffic will grow by a factor of 100. Since revenues are not likely to grow significantly, the cost of delivering this bandwidth has to decrease by a factor of 100. It said that somewhere around 60 percent or so of this cost decrease will come from Moore's Law and the increased density and performance of the chips. The rest of the savings will come from the flattening and simplification of these networks.

It also discussed the results of a service provider poll that it held with 900 respondents. Improving service quality is the number one factor over time. However, reducing cost and simplifying networks are expected to become the the number two and number three priorities, respectively, in three years.

Nokia Siemens also explained its decision to deemphasize GPON development. Its first reason is that the prospects for selling significant amounts of GPON over the next few years are not good. The market is sluggish and moving slower than expected. It also believes that VDSL is a very good alternative to GPON in the short run.

What Nokia Siemens does believe is that new technologies such as Hybrid PON and DWDM PON will start to appear in 2012 and will provide major improvements over PON. In addition to increased performance, these technologies will have the capacity to support unicast IPTV and to enable a significant network consolidation by eliminating the need for the current end offices.

There is a lot to what Nokia Siemens had to say. I think its PON strategy will be a good one if it can maintain its focus and be a leader in the introduction of these new technologies and if the GPON market does not heat up in the mean time.

Cisco IP Traffic Study Says Video Will Dominate

Cisco held a webinar yesterday to discuss its Visual Network Index that measures the degree to which video traffic of all types is contributing to IP networks. The details of this work can be found on its website. This work includes information gathered directly from major IP network providers. Cisco plans to provide quarterly updates to this analysis.

This work is worth following. It shows how much traffic Telco IPTV services will contribute to IP networks, especially in the metro areas. It also confirms the traffic assumptions that we used in our report Networking Strategies for TelcoTV Services,

Nokia Siemens Gives Up on GPON

Nokia Siemens announced that it will limit its investment in GPON because it does not expect mass market rollout of FTTH in the short term. Instead, Nokia Siemens will focus on ADSL-2+ and VDSL along with the development of next generation PON technology.

These next generation technologies include DWDM PON that can significantly expand the capacity of each fiber along with Hybrid PON that can extend the range of PON loops and facilitate the deployment of non-blocking IPTV aggregation/access networks.

DWDM PON will be enable the deployment of much higher speed PON services to each home and business. This will enable PON to meet future bandwidth requirements and extend PON into a broad set of business and enterprise applications.

Hybrid PON will permit the OLTs to be collocated in IPTV regional video headends, for example and create a network of dedicated connections to each subscriber. This will provide non-blocking access to each subscriber for unicast, on demand content. A 100 percent unicast service, such as this, will be required to support network PVR and interactive advertising applications.

This is a risky bet. If Nokia Siemens can beat it its competition to these new PON technology, it should be a winning bet. On the other hand, it will be hard for the company to maintain a strong focus on PON when it is not actively participating in the market.

Cisco Predicts Video Will Dominate IP Traffic

A Light Reading article has Cisco tables that gives their forecast for global IP traffic out to 2012. This forecast makes the following conclusions:

• Internet IP traffic dominates but non Internet IP will grow more quickly
• Consumer IP traffic dominates and will grow faster than business IP traffic
• Telco IPTV will be one of the fastest growing sources of metro IP traffic

This study confirms our analysis that Telco IPTV services will become a significant factor in carrier metro IP networks.

Alcatel-Lucent Improves IPTV Network Architecture

Alcatel-Lucent announced improvements to its Triple Play Service Delivery Architecture (TPSDA) that includes:

• Increased capacity and density
• Policy enforcement with per-subscriber resolution
• The equivalent of 19.2 million queues per telecom rack, which enables more fine-grain service control to ensure each service is delivered with the appropriate Quality of Experience (QoE) and uses network resources in the most efficient manner
• PPPoE termination, which facilitates the smooth migration of legacy networks to TPSDA
• A reporting and analysis system which provides tools to warehouse, aggregate and process per subscriber application-level statistics

Alcatel-Lucent's TPSDA has been successful and fifty service providers globally have adopted it along with Alcatel-Lucent products, including AT&T in the U.S. TPSDA is providing more granular control over flows in order to assure QoE. Zeugma is attacking the same problem with its new system.

These architectures will clearly be important for insuring QoE for VoIP and mobile services. It is still not clear to me that it will be effective for IPTV traffic as IPTV moves from a broadcast/multicast approach to an on demand/unicast approach. Unicast TV will saturate these networks and be quite difficult to manage using packet based techniques.

Zeugma Introduces Service Delivery Router

Zeugma, a Vancouver based start up, has introduced a service delivery router that combines deep packet inspection with the ability to support applications in a single box. Its objective is to support the deployment of value added services over broadband networks including managed Telco IPTV services and unmanaged over the top Internet TV and video.

Zeugma's market research indicates that more than 75 percent of the people surveyed are interested in buying or renting movies over the Internet. This research also indicates that people are not willing to pay much to begin watching the movie in a few minutes rather than waiting to download the entire movie first. While more than 55 percent would pay an extra $.25, only 40 percent would pay an extra $1.00 for a more immediate experience. On the other hand, more than 60 percent would be willing to watch a 30 second ad for the more immediate service. Zeugma stated that this shows that improved QoS can increase revenues.

The company states that its product's application awareness facilitates the classification of traffic. Its bandwidth manager can allocate or deny resources. It manages QoS and bandwidth on a per subscriber and per flow basis.

Zeugma's system is under evaluation at several major carriers including at BT.

Service management is clearly an important issue for broadband service providers. I don't think that best efforts will be adequate as the number and importance of these services increase.

However, the difficulties of managing video, especially Telco IPTV video should not be underestimated. We published a report Networking Strategies for TelcoTV Services that examines these issues in detail. You can request a free white paper on this subject.

Our conclusion was that as video on demand, replay TV, and other unicast forms of content from managed IPTV services will overwhelm broadband networks with video traffic. We half jokingly state that QoS will not work in this environment because there will not be enough low priority packets to discard.

While the effectiveness of the Zeugma and other similar system may be questionable for video traffic, especially for managed services, I think that they will be key for other services, especially for VoIP. I expect that they will be important on 4G mobile networks where all services, including voice, will be delivered over packet networks.

Siemens Developing Fiber IPTV Distribution Network

I had a hallway conversation at its analyst conference in Amsterdam this week, Nokia Siemens told me that it is developing a long reach PON system that will permit the OLTs to be located in the regional headends rather than the central offices serving the subscriber. There will then be an optical connection from the head end to each subscriber, which will eliminate any chance for the congestion that now can occur in Ethernet or IP metro/aggregation networks, especially when the demand for on demand unicast services is high.

This architecture is right in line with the report Networking Strategies for TelcoTV Services that we published a few months back. The conclusion in this report is that IPTV service providers will need to create a non blocking optical network rather than use an Ethernet or IP network to connect the regional headend to the OLTs in the serving central offices.

Few companies are stepping up to this requirement that is a direct result of the increasing amounts of personalization on on demand services that everybody expects to be part of the evolution of IPTV networks. It is good to see that Nokia Siemens is taking this seriously.

Light Reading/Cisco IPTV Scaling Test

Light Reading published an article that describes the results of an IPTV network scaling test that shows that a Cisco network can scale to support one million IPTV subscribers with QoS and resilience.

Light Reading asked the European Advanced Networking Test Center AG (EANTC) to run an end-to-end IPTV test on Cisco equipment. Cisco passed, showing its gear can handle a 60,000-user POP and, by extrapolation, a 1 million-subscriber IPTV service. It also demonstrated features such as Call Admissions Control (CAC) and Video Quality Experience (VQE).

EANTC simulated various failures and fiber breaks, and Cisco's routers managed to restore TV service in 625 milliseconds, worst case. And in QOS tests, Cisco correctly prevented the TV streams from getting disrupted when other types of traffic threatened to clog the network.

These results are interesting, but not surprising. There is little question that Cisco's routers can support IPTV networks.

The more interesting question is if this is the best architecture for IPTV. Our recent report Networking Strategies for TelcoTV Services comes to the conclusion that IPTV traffic does not fit a Layer 3 IP network well and that a Layer 1 optical network will provide a higher level of service at a lower cost. The issue is how to handle the on demand unicast traffic that everybody expects to dominate these networks. We are offering a free white paper on this subject.

Network

TelecomView has just published a report called Network Strategies for TelcoTV services. We found that TelcoTV service providers are moving from simple broadcast services to more and more services that use on demand techniques.

The problem that these service providers face as they make this change is that it will fundamentally change the nature of the traffic on their networks. Broadcast channels are transported using multicast technologies that send only one copy of each channel down each link independent of the number of viewers.

On demand technologies require that a unique video stream be sent across the network to each viewer, which significantly increases the load on the network. In fact, the unicast load from video on demand, NPVR, and advanced advertising services will cause video traffic to quickly dominate these networks.

Our conclusion is that TelcoTV service providers will have to move their video traffic off of their IP and Ethernet networks and on to a non blocking optical distribution network to be able to support Telco TV services, especially with the addition of NPVR and advanced advertising services.

You can get information about this report by clicking here.

You can order a free white paper that discusses our findings (as well as other free white papers) by clicking here.

I have been concerned about the load that video traffic will put on the networks that support TelcoTV networks for a long time. The recent announcement of new optically based systems from Sun Microsystems showed a new approach that became the inspiration for this report.

Most TelcoTV service providers are not thinking about this today, but they will be before long.