The preceding article in this series covered the current battleground between the walled garden carriers and what the market is referring to as the virtual telecom operators (Skype, Google, Yahoo! and AOL). This month we take a look at the IMS value proposition in the market and some of the early adopter feedback from actual operators who are testing and deploying it. The reviews are mixed, at best.
IMS Hype Defined
Anywhere you turn these days in the telecom industry youll hear someone mentioning IMS (define - news - alerts). It has become one of the hottest topics in telecommunications today. It has been collectively embraced by operators, network equipment providers, and analysts alike. Each group focuses on its own conclusions about why IMS is great. An example from In-Stats Web site, IMS will deliver the Holy Grail of convergence of access to multimedia services/applications across any end user device that all service providers are seeking to offer their customers in the future, says Henry Goldberg, In-Stat analyst. But providers have a long list of challenges facing them that must be overcome to fully migrate to a converged network architecture for their entire wireline and wireless businesses.
The important message here is that IMS, at its core, provides two immensely valued concepts ubiquity of access to applications via IP and reduced applications delivery cost through common billing and customer information control (via the HSS), no more silos of information! This is great news for carriers that deploy IMS and network equipment vendors that sell it, but is it great news for consumers? That depends on how these benefits are exposed to end customers in applications that provide unique value and whether or not there exist insurmountable business and/or technical problems.
Intra-Operator, Interoperability is What Counts
Typical IMS solutions today are provided from a single vendor, top to bottom. This means that the various IMS layers, transport, session and control, and application services will most likely work seamlessly and provide the deploying carrier a solution to their needs. This is in effect creating just a bigger stove pipe, one that can serve a carrier reasonably well. However, this is not the originally hyped intent of IMS, which was to enable best of breed solutions to be integrated by an operator to meet their needs. Unfortunately for the near term and, perhaps, forever, single vendor solutions will likely be the rule not the exception. But this is not the biggest issue with IMS. Early test reports from operators like Cingular indicate that IMS as a delivery vehicle for applications that need to follow users as they span multiple operators has many real world challenges built in.
Remember the operator is purchasing IMS from one vendor to eliminate interoperability issues, so what happens when one operator, say Cingular, fields an IMS application to its subscribers, which they purchase and then proceed to try and access in a T-Mobile region. Well, it just doesnt work. There are both technical and business issues to solve here. For one, its not likely that two different operators open service access gateways (OSA-G/W) will interoperate and that their home subscriber server (HSS) content will be shared. Remember that each operators HSS holds subscriber profile data, which they will guard with every fiber in their corporate body. So, for example if youre a Cingular customer and happen to subscribe to an IMS-based application service, its likely not to work outside of your home network. Will consumers care? The answer is yes.
Examples include combined presence, location-based services, or being unable to authenticate the purchase of Digital Rights Management encoded music outside of your network. So intra-operator interoperability is a big technical issue and a big business issue. What two operators would like to sit down and give a competitor their most important private asset: customer data? Its not going to happen. Whats needed is some sort of third-party clearinghouse, not unlike the services that interconnect GSM operators for roaming and billing today. So the least common denominator (a voice call) will always go through, but an application will not traverse the service layer of a foreign operator. Can this be fixed? Its possible, but many, many issues need to be thought through and agreed upon in the industry.
IMS Use of Session Initiation Protocol (SIP) is Great, Right?
SIP is an ASCII text-implemented signaling protocol that is easy to modify and extend to absorb new service models, which is part of the reason it was chosen to be used in IMS. It also has a lot of unique protocol methods that were originally developed for services in the VoIP world and are now being adapted to the IMS application space. There is a notion in SIP of the ability to register a notify request that reports back to the originator/requestor when some state has changed on the clients end for example, when a mobile user is off the phone and available to take a call. Another lab report from Cingular clearly shows that the protocol overhead alone in conventional SIP methods like notify if made available to a moderate percentage of users would swamp the data carrying capability of most mobile IMS networks. Thats just one example of how much more work needs to be done to define the reality in IMS.
Making IMS Real
For operators and end users, IMS can bring positive values. Its really about how well developed the technology is and how quickly the business realities of an all-IMS world can be resolved. To be clear, there is an ominous message from outside the walled garden: solve these types of problems or alternatives will be brought in their place. An example of this is the recent MVNO announcement from Disney on top of the Sprint mobile network. It proves that it is possible to successfully deploy targeted mobile advanced applications without the need to wait for IMS. IT
Mike Katz is director of product marketing for NMS Communications. For more information, please visit the company online at www.nmscommunications.com (news - alerts).
If you are interested in purchasing reprints of this article (in either print or PDF format), please visit Reprint Management Services online at www.reprintbuyer.com or contact a representative via e-mail at [email protected] or by phone at 800-290-5460.