It's been a decade since the notion of wholesale changes to public infrastructure based on IP convergence was first introduced, and the truth is that there has been little done to converge voice in most markets. The reason for this is that while VoIP gear might be less costly than Class 4 and 5 switches when considering a new purchase, those voice switches are already purchased and installed. In fact, worldwide, carriers estimate they have nearly a decade of useful life remaining. Many public carriers who offer both TDM and IP voice actually charge more for the latter; Verizon told me my business voice lines would be more expensive in IP form than with TDM! This makes TDM voice an institution in the near term...maybe.
The open question in carrier voice today is the timing and scope of fixed-mobile convergence (FMC) and mobile voice evolution. Network operators are more likely to use IP voice for mobile service than for wireline, and the cost of FMC to the operator is lower if wireline voice is also based on VoIP. Mobile voice, however, is already experiencing price capping worldwide, and that's a sure sign that voice ARPU will be under pressure. This means that any FMC or VoIP transition by network operators will be under pressure to reduce cost to the minimum.
One development out of this price/cost squeeze for carrier voice is the decision by a dozen or more worldwide Tier 1 operators to examine the P2P voice model. The hope of this group, who are keeping a very low market profile so far, is that a P2P or Skype-like model would distribute voice signaling and call-directing functions, and thus reduce the amount of centralized equipment to buy and manage. P2P is really a model of service signaling and not a kind of traffic-relay process, and so this makes sense on the surface.
Another development is the increased interest in collaborative and UC services as direct offerings. This is coming so far more from outside the carrier space than within it; Google, Microsoft, Yahoo, Cisco, IBM, and even a couple social network sites are looking at becoming hosts to UC and collaborative tools that would include VoIP.
Both these developments could be helpful, obviously, but they could also pose some major dilemmas for the enterprises who want to link public services with their internal VoIP and UC systems. The problem is that there are no real standards associated with these offerings so far, and that can spell major headaches in integration with enterprise solutions in the same area. Most of the advanced offerings are based on APIs rather than on signaling, meaning that some software solution would need to bridge external services to internal ones, and that might mean a software tool for every such service that a customer or partner might use. Standards in this area are likely to develop only very slowly, too, so there probably won't be a solution to this problem any time soon.
The lack of standardization includes the handling of even seemingly simple things like SIP calls. Enterprises already know that where SIP trunking or VoIP is offered by carriers, the interfaces and signaling isn't always identical, and so there may be some tweaking of the connecting devices needed to get things to work. A broader problem is the lack of consistent use of things like caller ID with VoIP; even if your own operator promises to pass accurate information on the caller (essential for many incoming call handling and call center applications), the problem is that VoIP peering may link your operator with one who doesn't even assure that the calling number is authentic.
In fact, peering and standards issues are the major problems with carrier VoIP services already. They're disguised by the fact that most VoIP services today interoperate with each other via the PSTN, for the simple reason that most handsets are still on either wireline PSTN or mobile PSTN services. VoIP peering, like the Internet, can be the wild west as far as adherence to common standards and practices are concerned, and one weak enforcement of standards by one VoIP peering partner contaminates everyone in the peering relationship. As the transition to carrier VoIP occurs, there are likely to be increased issues, and these could limit the credibility of IP-based carrier services, which in turn would impact the effectiveness of IP services used by enterprises for their own UC deployments.
You might wonder why UC standards could not be set by the UC providers themselves, and the answer is that with profit pressure driving the need for feature differentiation, no vendor in the UC space is truly anxious to see their industry lock features and interfaces into a common model quite yet. There is also a lack of enforcement of standards for customer-premises tools, including UC, where carrier services are highly regulated and thus carriers could be forced to adopt industry standards where they exist.
The hope for harmony in the carrier space to create UC adoption in the enterprise may come, ironically, from fixed-mobile convergence. FMC isn't a hot topic for most enterprises, but where it's necessary to create services that are accessible from multiple types of voice system (fixed and mobile), the meaning of "presence" has to be expanded, and presence may be the concept central to both UC and to future voice service profitability.
The competition for UC hearts and minds is expanding, and so is the carrier drive for new higher-layer feature revenues to offset declining revenue per bit. The future of UC may depend on the carrier trends creating the standards that will build a worldwide UC-ready community from the current voice users.
