By Manuel Vexler, IMS Forum

Recently, IMS joined a list of “misused” standards which, under the effect of market forces, changed direction by 180 degrees. These standards were originally intended for other applications, however they found their mass market appeal in other directions than the one originally intended by their creators.

IMS is not the first and for sure not the last case where standards are designed for one application, just to find that they are used for another.

As examples, let's start with the Ethernet, which at 10MB/s was running at a hard-to-imagine speed when modems were still running at a few KB/s. Of course, it was destined to be a LAN working over very short distances. Ethernet is now known as a 10/100 GB Ethernet point-to-point connection and can run thousands of miles on a fiber optic cable. Session Initiation Protocol (SIP) started as a peer-to-peer (P2P) music download standard, just to morph into a signaling standard used in IMS. And finally, IMS started life as a mobile broadband standard and is now used as a replacement for ... you guessed, softswitches. Used for VoIP delivery, the softswitches are used in both fixed and mobile networks to deliver a single, narrowband service.

IMS, however, will see at least one more turn as market forces will push convergence and multimedia in telecom networks worldwide. The writing is on the wall. Many service providers and in some cases regulators, are busy combining fixed and mobile voice business units in converged organizations. That will make the introduction of IMS easier and faster as it simplifies both the business case and maximizes the OPEX savings. It also forces the regulators to reconsider basic things such as the definition of a voice call — since for now TDM calls, cellular calls and VoIP calls are regulated separately.

The change to IMS also is helped (sadly) by the economic crisis, as companies have less financial means to keep competing point solutions in operation; for example, VoIP and PSTN fixed line are being delivered on multiple non-IMS platforms. With major and mid-sized service providers entering into new markets such as video delivery and content delivery, there is still a question why the IMS business case is still made on softswitch replacement. Maybe we should wait for a couple of more years so LTE will finally provide a competitive amount of wireless bandwidth (when compared with 3G) and demand the level of quality-of-service required by new interactive media.

Also, for a final comparison, let's go back to the Ethernet model, and talk about the mutation of standards – that is when a standard jumps from one transport medium to another. After all, Ethernet is closely related to the WiFi and WiMAX architectures. So how will IMS mutate?

Manuel Vexler is chair of the technical working group in the IMS Forum. He is well known for his expertise in voice and multimedia over Internet, bringing more than 20 years of experience to roles such as the CTO of CopperCom and vice president of IMS Interoperability at the IMS Forum. He drove M&A at Cisco, and launched new technologies at CopperCom, AMD, Alcatel (Newbridge) and Nortel.

A week ago today, on the morning of May 14, there was a “hiccup” in the Google network that lasted about two hours, during which 14 percent of Google users found themselves inexplicably and helplessly transported from cloud computing to a cloud of confusion.

The good news for the IMS community is that Google is not even looking at IMS as a technology to strengthen its search engines, which include VoIP, e-mail, IM and virtualized tools, just to mention a few of the applications and services that are devouring a good chunk of the Internet bandwidth.

Arbor Networks posted some analysis and graphs regarding the event that show the chilling effect of Google traffic disappearing.

The first question is how much revenue Google lost in the process and how the outage affected the thousands of small businesses and individuals using Google mail. A more important question, in my opinion, is if Google is finally starting to see a trend in the series of outages that affect its best-effort cloud computing.

OK, so what’s wrong with a best effort service? It’s free — we all love a free service — but there’s no live customer service number to call and ask for help. So our love of the free service lasts until that service breaks. Then we start to think that reliability and availability may be worth some money — or, depending on your business, it may be worth a lot of money. Since Google services are free there are no metrics for service quality, availability or customer satisfaction.

The problem is not the outage; the problem is the fact that Google, so far, has not been able to demonstrate a level of control and management over its network that is consistent with the offerings of a Tier One applications service provider. Which causes one to ask whether Google, Yahoo, MSN and other large ASPs should consider adding a level of control and management support to their respective “clouds.” The phone companies endeavored to provide reliable phone service. To date, this quality of service element is not an intrinsic part of what we call Web 2.0 services. How important do users consider the reliability of e-mail, VoIP and IM to be, both socially and financially? As they would say in Twitter, “Essential!”

If we agree that Internet should no longer be “best service,” then control and management are “Essential!”

Manuel Vexler is chair of the technical working group in the IMS Forum. He is well known for his expertise in voice and multimedia over Internet, bringing more than 20 years of experience to roles such as the CTO of CopperCom and vice president of IMS Interoperability at the IMS Forum. He drove M&A at Cisco, and launched new technologies at CopperCom, AMD, Alcatel (Newbridge) and Nortel.

The current recession has exposed the difference between the “utility' business model” — revenue based on recurring monthly contributions by millions of subscribers — and the “Web 2.0 business model” — indirect revenue based on advertising. With marketing and advertising dollars rapidly shrinking, Google 's stock went down by 30 percent as compared to AT&T and by 15 percent as compared with Verizon. The difference is even higher when we look at a whole year — about a 70 percent loss of Google's value when compared to AT&T.

First let's acknowledge that this will likely change in the next six to 12 months as the economy (we hope) will recover and Web 2.0 and Web 3.0 will demonstrate again their market power. Huge amounts of bandwidth will be delivered by fourth generation GSM (aka LTE) and services from Google and others will quickly fill the broadband pipes.

Second, AT&T and Verizon valuations show that investors and customers understand the value of the network in the Internet world — in other words, the long-term investments service providers are making in their infrastructures. Also, this is supported by the economic stimulus programs with broadband taking the front seat in the government's plans.

So where is IMS/NGN coming into this picture? With broadband becoming ubiquitous, service providers can no longer afford to have separate networks for fixed and mobile voice, DSL, fiber to the home (or to the kerb), mobile data (with overlays for GPRS, HSDPA, etc), IP video (e.g. FIOS) and so on. A good example is the fixed voice network. While still a good revenue generator, it competes with free services from Skype, low cost services from Vonage and alike and cable telephony, not to mention cellular users. While much debate focused on the IMS business case, or more precisely the inability to charge an entire network cost to a single service or “killer app,” the debate should really focus on whether is really a good investment to continue with building parallel networks for what can be simply defined as Web 2.0 services (voice, video, internet access, mobility, presence).

Let's draw a parallel with the road system. Will it make sense to have five or six separate highways to keep truck traffic separated from buses, vans, cars and motorcycles? Of course not, and no one proposes this model for highways. So why would this model make sense for communications service providers? Not to mention that investors that continue their support for service provider are going to benefit along with customers, from lowering the cost and increasing the revenues with converged services delivered by “all-IP” IMS/NGN technologies.

Manuel Vexler is chair of the technical working group in the IMS Forum. He is well known for his expertise in voice and multimedia over Internet, bringing more than 20 years of experience to roles such as the CTO of CopperCom and vice president of IMS Interoperability at the IMS Forum. He drove M&A at Cisco, and launched new technologies at CopperCom, AMD, Alcatel (Newbridge) and Nortel.