February 2005

A Triple Play Top 10 Service Provider Considerations For Platform Selection

By Ken Meaghe

The “Triple Play” has been a major topic of discussion and hype for a number of years now while all the real action has been around delivering best-effort high-speed Internet. Times have changed. The combination of regulatory changes and trends, along with the competitive landscape, has progressed to a point where all service providers are making firm plans on how to best move forward with the triple play. The decisions they make will impact their ability to compete over the next decade and as such, are fundamental to their future business. They need to move forward with plans on how to deliver multiple services over a single infrastructure. A multi-service network isn’t built using a best effort network model so a significant shift is required. This is a complex optimization equation balancing investment with revenue potential and there is no “one-size-fits-all” solution.

Here are the Top 10 things a service provider needs to consider when selecting a platform for delivering triple play services.

1. It all starts and stops with the “user experience”
The starting assumption has to be that the service offering must be as good as — or better than — what a customer is able to get through traditional service providers for video, voice, and Internet service today. Customers are not going to tolerate anything less than parity with existing benchmarks. The channel surfer is not going to accept long channel change times or issues with picture quality. The voice caller is not going to put up with long delays in conversation.

There will be a willingness on the part of the consumer to move to service providers that offer new features that come from the integration of voice, video and Internet into a single bundle that interwork and enhance the overall experience. This is where the key to new service revenue and subscriber growth lies.

2. Building a differentiation capability
Service providers cannot focus too narrowly on the service offerings and bundles, as they exist today. The objective needs to be extended to include building a capability to deliver new and differentiated services. The current service offerings have been commoditized, and so will a “bundle” of these services quickly become a commodity. Convergence will enable a service provider to roll out new services that provide differentiation. Differentiation is the answer to new sources of revenue and customer retention. This, however is much more than a platform decision, it has to be built into the organizational model as well if the service provider is looking to build a sustainable competitive advantage.

A service provider that appreciates this will build an infrastructure and organizational capability to deliver new services, harvest and renew them over and over again.

3. Standards-based interfaces and protocols will reduce risks and costs
It is essential to build a network solution consisting of a set of components that have standards- based interfaces and protocols. There are a large number of moving parts in a network that will deliver the triple play and to reduce the technology and vendor risk, a standards- based approach must be taken. The additional benefit is that costs are easier to control if there are multiple system choices available.

Even with a standards-based approach to building a service delivery platform for the triple play, interoperability between the various components will still be an issue for service providers and vendors. Standards compliance does not necessarily equal interoperability. It takes time to work out the subtleties of the interfaces and protocols. Service providers and vendors will have to work together in this area to achieve full interoperability.

4. End-to-end quality of service (QoS)
QoS is probably the most talked about piece of the triple play equation and perhaps also the most misunderstood. In simple terms, it represents the ability to ensure that a subscriber is able to get what they have paid for.

The network must be able to reserve the required end-to-end network resources necessary to deliver the service requested by a subscriber at the time the service is requested. This is very different from just being able to prioritize traffic flows in a best effort network and becomes increasingly important with interactive services where delay, jitter, and packet loss may render a service inoperable.

For Internet telephony, people are not going to tolerate long delays in their voice conversations such that it interferes with the rhythm of conversation. Nor will they put up with too much echo. For video delivery, the use of higher compression techniques will cause video to suffer significantly from packet loss. The delivery platform must have deterministic characteristics for end-to-end delay, and must not drop high-priority packets under congestion conditions.

The service provider needs to think about the trends in services offerings and what that means with regards to network performance parameters such as delay and packet loss.

5. Subscriber/Policy Management
Subscriber management or policy management is also going to change significantly in this next-generation network. In the traditional “best effort” Internet model, subscriber management represents the classic PPP termination point where the “AAA” was performed and subscriber traffic was passed into the Internet. In the multi-service network, subscriber management must be extended to include policy management. There is great debate within the industry as to whether the “BRAS” function should be centralized or distributed.
For the service provider, what is important is getting revenue for the resources used to deliver the subscribed services and gaining additional revenue by offering new services that subscribers can select even temporarily when network resources are not fully utilized.

The subscribers want easy access to the services that they have subscribed to and to have the ability to select enhanced services if the network resources are available. Since they will select different service bundles and performance attributes, the network must have the ability to shape/police traffic based on a subscriber’s service agreement and have the ability to change those policies on the fly in order to maximize the revenue potential of the network.

6. Flexibility for deployment and evolution
Having the capability to work into legacy networks and a variety of softswitches gives the service provider an important degree of flexibility in terms of how to migrate and evolve the network. Working into the existing voice infrastructure is especially important to the incumbent providers today while service providers without existing infrastructure are likely to deploy softswitching technologies. In either case, a platform that offers flexibility in this regard will mitigate the risk of selecting technologies that may become obsolete quickly as technologies and the services that they are based on continue to evolve.
At the physical layer, the service provider is faced with numerous physical plant and deployment scenarios. A platform that can easily address a variety of deployment needs will reduce the number of variables required to deliver the total solution. On the access side this may mean supporting a variety access technologies, xDSL and Fiber (PON and FTTC), as well as a number of network facing technologies such as SONET and Optical Ethernet. A single platform or product family that can address a variety of deployment needs will be easier from a deployment, maintenance, and interoperability perspective.

7. The Profitability Equation — Revenue Versus Costs
In November 2004 The Wall Street Journal published an article entitled “Meet the New TV Guy” in which Ed Whitacre discussed SBC’s evolution to being a provider of TV services. In this article, he describes a $100 service bundle that includes TV, Internet, and voice and wireless services. In order to operate at a price point of $100 it will be crucial that service providers consider the Total Cost of Ownership and move beyond the CAPEX piece of the equation. While everyone may not agree with a $100 price point for the bundle, it is certain that with increased competition, the lowest possible operational cost model will be critical. Consumers have a certain amount of disposable income to spend on these services. This discretionary spending will go to the service provider best able to balance the value/price equation. Given the competition for this disposable income, a low operational expense model will be increasingly important. Service providers should take an “Activity-Based-Costing” approach to refining operational expenses.

8. Scalability — more than just size that matters
Every vendor claims to offer products that scale easily to support multiple users and services. Service providers need to spend time understanding how the network should scale to support thousands of subscribers using multiple services and how performance and operational costs may be impacted by subscriber growth and load levels. Vendors make a lot of claims about system performance and scale and service providers need to be careful about what is being quoted and under what conditions. They need to determine how the system will be used, what the performance needs to be under various conditions to ensure that the selected solution is capable of delivering a positive user experience. Scalability claims may be impressive but the service provider needs to get behind the numbers to ensure they closely match their needs. The next step is testing the solution and stressing it to reflect the real world application.

9. Distributed intelligence
In order to deliver multiple services in this network, more intelligent endpoints are going to emerge. The networks of the past had central control elements that were very large, costly, and usually single purpose. They also suffered from being expensive and difficult to upgrade to provide new features.

Video is the significant driver for change in the broadband network today and the most significant impact is on network bandwidth. In order to economize on network resources, distributing a multicast capability into network elements makes sense. By having the capability to join and leave multicast groups made by elements within the network, the service provider is able to minimize network bandwidth and channel selection times.

In order to support the variety of multimedia applications and their evolution, it is likely that the media processing and control will be distributed on a service basis. There are multiple control planes that evolve here. The first is the network control plane that manages and reserves network resources required to support a specific service and the second would be the service control plane (IGMP for Video, SIP for communications). They will have to work together to deliver the service but they are likely to be distributed into various network elements.

10. Demarcation point between service provider and the customer becomes blurred
As the service infrastructure becomes IP-based and subscribers receive multiple service offerings over their home networks, the demarcation point between the service provider and the customer is going to blur. A set-top box will be used for video delivery and its network interface will be an Ethernet port that is connected to a hub or router that is part of the customer’s home network. Other home networking gear including wireless local-area networks (LANs) are already an important element of the home network and if not engineered properly, may impair the service offering.

Like it or not, service providers, are going to have to deal with customer network issues. Service providers need to consider this carefully as they move to a single IP network to deliver multiple services especially as home networks become increasingly complex and varied. Types of networking equipment and the quality of it will all factor into delivering a quality experience.

Conclusion
This article has tried to highlight the top ten issues facing a service provider deploying a network for the triple play. There are others, but one thing is clear, the triple play changes the game significantly from the best-effort Internet model that has been deployed over the last several years. The technology exists to build a true multi-service network than can deliver a quality end user experience. Service providers need to focus attention on building a capability to deliver new and differentiated services over a common infrastructure. Only by building such a capability will they create a sustainable competitive advantage for the future.

Ken Meagher is currently product marketing director within the Broadband Access Group at CIENA. For more information, please visit the company online at www.ciena.com.

[ Return To The February 2005 Table Of Contents ]