We are a species that likes things settled. There are no hoops we will not jump through or haystacks we will not scour to place a big fat period where a question mark used to be.

The riddle of the Sphinx -- solved. The big heads of Easter Island – done deal. Stonehenge – case closed.

Most recently, our lack of tolerance for loose ends has brought scientific theory to bear on perhaps the last great unsolved mystery of the 20th century: Is double dipping simply rude or a legitimate health risk? In case you missed it, researchers from Clemson University resolved that George Costanza did indeed deserve admonition from his girlfriend’s grieving brother.

Another mystery solved. The human race can once again rest easy.

But hold on. This respite from resolving open-ended questions will be short-lived in the telecommunications industry. Operators and equipment vendors are already growing anxious over a brewing conundrum in the access portion of carrier networks.

To answer subscribers’ cravings for communications and entertainment services that will flow across any type of access network and display on any type of device, carriers need to reconstruct the edges of their networks to perform the session management acrobatics required to fulfill these end user expectations. It no longer makes sense – in terms of operational/capital costs or service delivery efficiency -- to erect separate termination and session management subsystems for each type of access network coming into multiservice networks. And we’re not just talking about two flavors here -- fixed and mobile -- but a Baskin-Robbins-like roster that includes Wi-Fi, picocells/femtocells, cellular (three generations), DSL, cable, fiber and more, all of which come with their unique sets of protocols, interfaces and security requirements.

While operators generally agree it makes sense to cluster all the technology and functionality required to manage sessions -- things like authentication, security, deep packet inspection and seamless handoff -- in the access portion of their networks, there is absolutely no consensus on how that functionality should be packaged. The argument, at a high level, has two sides. One alternative is to tie together discrete equipment through standardized interfaces. The other is to concentrate all of this functionality into a single device, controlled by a single management system and with a single interface out to the rest of the network.

If you’re paying attention, that last option should sounds a lot like the much-reviled Godbox, which seems to unsuccessful rear its ugly head about once a decade, only to be sent back to purgatory by non-believers. Though proponents of this approach, which has been tagged by some with the banner “convergence gateway,” downplay the supernatural aspect of the equipment, they make the case that the unique nature of fixed mobile convergence (FMC), which requires unprecedented levels of coordination in the signaling and bearer paths, screams out for a single anchor point in the network for session management. Networks of the future will not only tolerate a Godbox, they will demand one, advocates claim.

Though it’s convenient to boil down this conundrum to two options, as I have done above, in reality there are dozens of options for clustering session management and security functionality. Just do the math. When you consider that there are at least seven or eight functions that fall into a session management solution, the laws of probability point to a dizzying number of ways to group session management functionality. These shades of grade, as we’ve already established, are driving both operators and equipment suppliers batty.

In terms of actual product development, the field is pretty split. Some equipment makers, presumably influenced by RFPs, are offering up integrated equipment that includes the kitchen sink, two kits and a caboodle, for good measure. Others, such as Acme Packet, have drawn clear lines of demarcation, limiting its bundling efforts to adding security gateways and a tunnel termination gateway (TTG) to its SBC.

Given our propensity for seeking solutions to riddles, the questions on operators’ and equipment makers’ lips are: Who is right? How long will it take to fill in the missing pieces of the puzzle?

The answer to both questions is unsatisfyingly the same: Too early to tell. Unfortunately, this is a mystery that will remain unsolved until at least the end of 2008 and, most likely, well into 2009. In the meantime, why not apply your powers of deduction to a more accessible conundrum. How about: What’s the sound of one hand clapping?

Joe McGarvey is Principal Analyst, Carrier IP Telephony, at Current Analysis.