Cable operators are preparing for a great and necessary migration. To provide voice service to their customers in recent years, they have installed millions of lines that conform to PacketCable 1.0 and 1.1 specifications. With the emergence of IP Multimedia Subsystem (IMS) and Session Initiation Protocol (SIP) standards and the PacketCable Release 2 specifications, MSOs are now planning for deployments of new multimedia services based on a converged network architecture.

Step 1

MSOs are making the move to a standards-based IMS architecture that supports revenue-generating enhanced voice and multimedia services. Over the past few years, some of the major MSOs in the U.S. have deployed large amounts of PacketCable 1.0/1.1-based endpoints. Unfortunately, none of the equipment they have already deployed in the field can support next-generation voice and SIP-based multimedia services. However, MSOs will have to continue to support their existing Voice-over-IP/Voice-over-cable (VoIP/VoC) customers who use existing gear as operators roll out IMS infrastructures.

MSOs, therefore, will need a migration path to an IMS infrastructure that supports continued service to existing customers and a smooth transition to enable new multimedia services. As mentioned above, PacketCable 1.0/1.1 gear uses NCS, a variation of the Media Gateway Control Protocol (MGCP) signaling, to manage endpoints. On the other hand, PacketCable Release 2 supports SIP, as does all of the IMS equipment that service providers will be putting into their next-generation IMS networks.

Because there is so much PacketCable 1.0/1.1 gear in the network, flash cuts to next-generation IMS gear are highly unlikely. MSOs deployed VoC based on PacketCable 1.0/1.1 to offer their customers bundled voice, data and video services (triple-play) in an effort to win business from the ILECs and reduce customer churn. As a result, most VoC customers today are interested in the service primarily because it provides them with good quality, low-cost voice service.

These triple-play cable customers could be motivated to upgrade to SIP-based eMTAs if they could gain the ability to access enhanced voice services, including Dual-Mode Handset (DMH) service. DMH would enable them to combine their cellular and VoC services to create a converged voice platform that supports seamless mobility (i.e. seamless roaming and handover functionality) between VoC and cellular access. Customers who upgrade also could use their new eMTAs to access a host of new multimedia services. Despite these advantages, most MSOs believe the majority of their existing VoC customers will migrate to these new IMS/SIP-based services over an extended period of time. Furthermore, many MSOs have only just begun to map out their IMS strategies. It's still early in the game, but the game is definitely on.

Smooth sailing

When they begin their IMS and PacketCable Release 2 migrations, MSOs can perform remote upgrades of some of their inventory of PacketCable 1.0/1.1-based eMTAs. However, MSOs likely will have to roll a truck to replace any older endpoint devices in the field.

MSOs realize that any next-generation services they offer must work properly for their customers the first time, and every time, they use them. In fact, most MSOs are already in the habit of making house calls to install eMTAs when customers are switching their voice service from the Public Switched Telephone Network (PSTN) to VoC. While at the premises, technicians usually check home wiring systems and ensure that all the phone jacks in the home are able to deliver voice. The same diligence will be needed to support SIP-based services and to ensure a satisfactory customer experience for all who decide to upgrade.

Fortunately, how MSOs choose to accomplish this and the timeframe in which they will migrate from PacketCable 1.0/1.1 is up to them. This is because MSOs can reuse PacketCable 1.0/1.1 network components as they migrate legacy gear onto an IMS infrastructure.

In the PacketCable 1.0 and 1.1 architecture, the Call Management Server (CMS) function provides NCS-based call control of eMTA endpoints and the Media Gateway Control Function (MGCF) controls the PSTN facing media gateways. Both functions are typically performed by a softswitch in today's networks.

In order to leave the existing PacketCable infrastructure in place and operating seamlessly alongside their growing IMS infrastructures, MSOs can separate the CMS call control function from the existing MGCF function. When separated from the MGCF, the call management function can be moved to the application layer of the IMS architecture (see Step 1). Taking this action smoothes migration because MSOs do not have to perform any database conversions for the NCS-based legacy end-points.

Alternatively, MSOs could leave existing subscribers on legacy softswitches with integrated CMS and MGCF functions and add new voice application servers at the application layer that are capable of providing call management server function for both NCS and SIP endpoints. Depending on the existing network configurations, the most simple and efficient solution could be to leave the existing PacketCable layer as is, and add a new voice application server that connects via SIP to the MGCF of the existing softswitches. Here, SIP acts as a switch-to-switch interface that connects the existing CMS/MGC, while a new application server can be installed to handle voice and multimedia features for new subscribers.

During this stage of migration, the interface between the PacketCable infrastructure and the IMS infrastructure is the softswitch. Based on the requirements of the existing PacketCable infrastructure, the softswitch supports both SIP and NCS endpoints along with interswitch SIP interfaces for IP peering and intra-network connectivity. The use of SIP enables the voice application server function to be moved into the IMS domain. Existing PacketCable endpoints connect to the IMS infrastructure at the network layer via the interswitch SIP interface.

Introducing IMS

The next part of the migration path is actually very straightforward. The magic comes later as the MSO begins to introduce IMS core elements. As these elements are added, the CMS function can be moved into the IMS application layer. There, the CMS application platform can continue providing support for any remaining NCS endpoints in the network as well as serve new IMS-based SIP endpoints. The beauty is that the voice application server is able to operate in both domains and simultaneously provide call control for IMS/SIP-based endpoints as well as NCS-based endpoints (see Step 2).

Step 2

This, in turn, enables the seamless migration of services as well as equipment. The MSO has the ability to support NCS-based devices and offer additional features for SIP-based components that support advanced multimedia applications. Furthermore, it is much easier to upgrade and maintain databases of legacy and next-generation devices if they are supported on the same platform.

Bringing IMS core components into the network is somewhat more complicated than the initial steps of IMS migration. These components include the CSCF and the Home Subscriber Server (HSS) database. Adding these two components requires database migration. The SIP subscriber database must be moved to the HSS in order to introduce the CSCF. And just as the SIP interfaces have been moved between the CMS and the existing PacketCable devices, control must be moved from the CMS application server in the IMS domain to the CSCF in order to manage SIP-based endpoints. In the IMS environment, the SIP endpoints are directly controlled by the CSCF and HSS (HSS and S-CSCF jointly provide SIP-based location and registrar functions). The CMS in the application layer becomes a pure voice application server, but it continues to support legacy PacketCable 1.0/1.1 endpoints.

When preparing their existing infrastructures to support IMS components, some service providers might decide not to put any SIP functionality on the CMS before they introduce the core IMS components in their networks. The choice belongs to the service providers. They can migrate control of those elements to the CSCF when they introduce the IMS core components. Alternatively, MSOs can choose to continue to deploy PacketCable 1.0 and 1.1 end devices, begin to migrate them, and then start to deploy SIP-based eMTAs when they begin to deploy IMS core elements. Many service providers believe the most cost-effective approach is to begin deploying next-generation eMTAs as soon as possible.

Step 3

When MSOs deploy their core IMS infrastructures, they will have the ability to provide much more than basic voice services to SIP-based endpoints (see Step 3). When they have migrated their current PacketCable infrastructures, they will have the ability to deliver personalized media-rich applications, such as push-to-talk, buddy lists, integrated Instant Messaging and chat, presence-enabled multimedia content sharing, and seamless handover between cellular and Wi-Fi networks, to name a few. These services cannot be delivered to existing PacketCable endpoints today. But, because they have a converged IMS network architecture that reuses existing PacketCable core components, MSOs that are in migration mode will be able to cost-effectively support existing customers while preparing for the future with a new generation of SIP-based multimedia services and applications.