IMS has been discussed for a decade and specifications have been available since 2002, however adoption is minimal. There are partial IMS deployments for specific new services, but mobile voice telephony continues to rely on traditional GSM protocols. It was expected that deployment of LTE (News - Alert) radio networks (which are completely packet-based) would require the widespread adoption of full IMS, but now several developments suggest IMS will be further delayed.

 

First, the widespread adoption of 3G USB modems means most data passing over 3G networks is going to and from the public Internet. For mobile Internet access, there are no QoS expectations and thus no need for IMS. Second, the success of Apple’s (News - Alert) iPhone “App Store” has prompted similar application stores from Google, Microsoft, Nokia, Adobe and others. An expanding community of application developers is producing thousands of new applications — far more than any mobile operator could dream of two years ago. But these new applications access the mobile Internet with no need for IMS. As a result, it’s increasingly likely that initial LTE networks will be used for bulk data access while voice telephony remains on existing 3G networks. With LTE used only for mobile Internet access, there’s no need for QoS and no need for IMS.

 

Finally it appears the GSM community is looking for ways of providing voice services on LTE using traditional GSM signaling, i.e., without IMS. Several efforts are on the table including 3GPP TS 23.272 Circuit Switched (CS) fallback in Evolved Packet System (EPS) and a study document TS 23.879 for CS services over EPS.

 

So is IMS finally dead? No. All of the large equipment providers have developed IMS equipment and partial IMS deployments are in service for specific new services like Push-to-Talk over Cellular (PoC) and Fixed Mobile Convergence (News - Alert) (FMC). While no GSM or CDMA operator is using IMS for regular voice service, the equipment is available and has gone through trials. All that’s required is an opportunity and that may exist in mobile WiMAX (News - Alert). WiMAX was designed for data. There is no prior voice telephony support, no circuit switching. If voice is to be offered on a WiMAX network, it will need a system like IMS. Furthermore, like any cellular network, WiMAX cells are shared and subject to congestion during peak traffic, so some QoS will be beneficial for voice telephony — again IMS is a potential solution.

 

WiMAX networks are being rolled out in many parts of the world. Today these are mostly data networks for Internet access, but most WiMAX operators are eyeing mobile voice services as voice is still the killer application. So indeed, WiMAX may save IMS.