The beauty of the IP Multimedia Subsystem (IMS) architecture is that it enables “hard” switches to take on a “softer” nature through software-based technologies. An IP network is a dumb network comprised of generic switches that handle only transport. With IMS, the services that run on top are implemented via software rather than specialized hardware.
That tack differs from the traditional networks, where the same companies that built the switches also packaged applications and feature sets. With IMS, carriers can make application decisions independent of switch decisions. That is somewhat of a reincarnation of the intelligent network promise of innovation outside proprietary systems, but with better technologies to back it up.
“Once you separate the switch from the applications, you can divert traffic to wherever you want,” says Emerino Marchetti, director of Sprint’s network development group. Sprint is using components of IMS to establish consistency in service delivery, whether over dual-mode handsets, Wi-Fi, CDMA, PBX, or DSL. With its roll-out of portfolio sets for enterprise business customer desktops requiring PBX and mobile services, the goal is to enable customers to extend any feature to any device.
“We don’t want customers to have to learn different ways to do things. We want them to press the same key to access voice mail or to read messages or to pay their bills,” says Marchetti. “If a parent wants to restrict text messaging or phone calls, we want them to pick from one menu and for that feature to work whether on a wireless connection or landline.”
Despite its potential to facilitate such features, IMS is not a silver bullet. It does not provide application definitions or standards for specific features.
“IMS doesn’t alter the situation in a fundamental way; however, it allows real-time-oriented media services to move on IP networks. That is what opens the door to monitoring usage in real time associated with real-time charging,” says Kai Michaelis, senior director for strategy and communication in Siemens’ Mobile Networks Charging and Care organization.
“While the engineers may disagree, the view from a charging perspective means you handle IMS like any other application server or network element,” says Michaelis. “It talks to online charging systems via standardized interfaces like Diameter, which serves as an extensible baseline that enables carriers to add specialized functionality.”
“You have to look at IMS as something on top of which the service delivery will sit,” says Marchetti. “If you want to make it simple for customers to access multiple applications, IMS will provide the means for having a single data store and concentrated view into preferences.”
That’s particularly important for an operator in a multi-vendor environment, as the rapid introduction of services will require a network and services architecture to have open “hooks” for billing and OSS applications relying on real-time data access.
“Because billing is just too complex to be standardized, 3GPP2’s IMS documentation for charging and the capturing of data elements for usage, number of bits, duration of session and so on will be the key,” says Cathal Kennedy, Telenor’s billing architect. “With IMS, you derive the data that billing needs, as billing will remain beyond the scope of any standards body.”
For billing and OSS to procure the data in real-time, mediation and charging will take on more important roles.
Active Mediation, Charging
Active mediation is the real-time mediation that must occur among pre- and post-paid billing solutions.
The key to real-time mediation is the service delivery platform (SDP) used to implement services that the IMS framework supports.
SDPs support standards such as OSA Parlay and JAIN SLEE (service logic execution environment) to open up service implementations on IP networks. “Carriers need solutions that incorporate complementary standards work, such as JAIN and Java OSS, to make sure components of security, service level management and charging are all taken care of in a IMS,” says Chris Couch, chief marketing officer for Ace-Comm.
Couch observes that the mediation function is moving from a “collect-and-transform” function to a two-way real-time enablement model: “Two-way mediation is necessary because you are working with the online charging system and other service elements in real time,” says Couch, noting that batch mode mediation will be phased out over time.
Because mediation will connect directly to the network to acquire information in real time, OSS and billing will have to accommodate numerous IMS interfaces in parallel to contemporary interfaces. “Carriers will do more then collect and correlate data; they will have to apply business analytics to make real-time decisions,” says Couch, noting that Ace-Comm’s newly released MarketMaker is designed to assist in that purpose.
Active mediation will sit between operators’ networks and charging systems to integrate service delivery infrastructure with charging systems. That enables the real-time exchange of charging information for control over calls, sessions and events.
Online charging systems (OCS) are at the core of active mediation, as they establish a standard set of active mediation products for real-time control of chargeable events. The purpose of OCS is the creation of unified rating for pre- and post-paid systems through modular or “componentized” systems that help carriers break free of single-vendor solutions.
“To ensure their OCS works with different network types, carriers have to look for rating and charging systems that are designed to handle transaction rating in real-time from a plurality of event triggers such as SS7, CAMEL, deep packet inspection, Diameter CCA and so on,” says Jeff Popoff, Redknee’s vice president of marketing. Redknee has implemented and deployed Diameter’s credit control application (CCA) between the Gateway GPRS Support Node (GGSN) and its OCS. Thus far, Redknee has completed Diameter interoperability testing with Starent, Nortel, and Cisco.
For a comprehensive rating, charging and policy management solution, Diameter’s CCA comprises a material link between the Traffic Plane Function (TPF) and the OCS.
Other interfaces also exist to foster better granularity. For example, SIP’s Session Description Protocol in conjunction with charging mechanisms in Diameter come together to provide additional information on media flows for more granular rating and charging.
SIP is important as a “session” occurs between a subscriber handset and the network, as various elements communicate and transport data via SIP.
“To foster reuse of what already exists in carriers’ infrastructure, SIP and existing Diameter infrastructure are quickly evolving so the two elements work together for service differentiation or prioritization of voice or video to happen,” says Siemens’ Hannes Tschofenig, division scientist and a member active in the IETF’s new working group called Diameter Maintenance & Extensions, which will focus on QoS mechanisms through Diameter extensions. “The IMS system consists of SIP signaling and network elements that handle the prioritization. Diameter extensions are used as a means of enabling those systems to interact,” he explains.
“In a scenario where a subscriber wants to see the streaming video of the latest episode of a TV show, the subscriber initiates a request to the network, which contacts the SIP application server for permission, rating and control,” explains Luther Rudisill, senior product manager for bcgi’s real-time billing and network product lines. “The SIP application server that is to provide the requested content is contacted by the SGSN/PDSN/GGSN, and the stream happens.”
Enter Diameter
Diameter is important, as it has become the base protocol in IMS for charging and credit control (see Standards Watch, pg. 22). It is Diameter that enables IMS to foster “call control” in an IP world, which is why the 3GPP, 3GPP2, ETSI, OMA, TISPAN and ITU-T have been working on their respective pieces of IMS documentation to enable service delivery through coordinated offline and now online charging, as well as QoS capabilities.
Where CAMEL has been used for voice, it is Diameter that is used for IP traffic by procuring information from rating and accounting systems.
For example, call control can be realized by building CSCF (call server control function) switching layers, HSS (home subscriber server) user data planes and media functions independent of access points. “That enables IMS to ‘aim’ the customer into your various vendors’ switches, or gateways and PBX farms,” says Marchetti.
HSS is used for all charging information sent for AAA to offline or line systems using Diameter. And CSCF is the packet server that is processing or managing the session and invoking services and application. That means application servers will be increasingly important, as they generate the chargeable events and send data for AAA.
The 3GPP2 has published a specification to support the charging data sent to offline charging systems, and it is now focusing on real-time charging and flow-based charging to enable prepay and pay-as-you-go accounts to function.
Because mediation works with records coming off the content server, it would be nice to have something analogous to TAP3 or TAP4 to define formats for all the various application servers in the world. But there will be different interpretations for different components of IMS, which means carriers must beware of how their partners proceed with IMS.
Beware Political Jockeying
Carriers in contemporary 2.5G-based environments looking for full convergence will look to Diameter, Parlay, SIP and myriad of standards comprising IMS to help address inputs from multiple sources. In the drive toward a comprehensive rating and charging solution, carriers must be cognizant of the risk of “gelling” standards in interoperability testing environments that leave too much open for interpretation.
The term “IMS” is ambiguous in and of itself and each vendor defines it slightly differently.
When you look at the approximately 80 RFCs for SIP, it’s easy to see how problems can occur with the content of CDR-like records if the formats from multiple application servers are different as SIP communicates with various elements of architecture.
Just as GPRS did not evolve as originally intended, it will be the carriers who pay the price for a legacy of proprietary systems to maintain if vendors “gel” standards in a closed environment. If they are not careful, carriers may be headed toward multiple administrative consoles, product catalogs and subscriber databases because vendors didn’t open up interfaces in the race to be first to market.
“It’s up to the operators to make sure their network equipment providers open up the interfaces defined by the standards in IMS,” says Amdocs’ Tal Gavoly, chief scientist and a chair in the accounting management task force under ATIS. He advocates that operators remain very active in the development of the IMS charging architecture, AAA, RADIUS extensions, IPv6 and other charging and accounting aspects of IMS development.
He believes the “nuancing” will lead to proprietary interpretations, which then will lead to stovepipes.
Nevertheless, some operators are rushing to the forefront with “lower-cost /lower-risk” solutions that get them to market first.
“There has to be an opening up of network elements from vendors with IMS so we can plug content on the Web to a voice network,” says Kurt Silverman, the CTO of converged billing solutions for Comverse. But for that to happen, the standards networks and e-commerce on Web sites for content have to be handled simultaneously by one infrastructure that handles pre- and post-pay. “That means being in the authorization path as a session control point, checking balances and knowing subscriber transaction history, as well as new acquisitions and orders,” says Silverman.
Comverse is looking at IMS as an enabler for soup-to-nuts solutions, so the company is evaluating the control, service and applications layers in the IMS framework to handle pre- and post-pay convergence at a large scale.
As vendors look to offer integrated suites attached to the signaling network and pertinent data for real-time services, it is important for carriers to determine if their vendors are “future-ready.” Key to that will be breaking architecture down into components so that provisioning, Web acquisition systems, pre-pay, post-pay and other systems all interoperate. That means all vendors have to maintain open interfaces and pre-established adapters, so that carriers can leverage investments and phase out legacy systems as they depreciate.
Email this article
Add a comment
Printer version
Order reprints
RSS Feed
Bookmark article