NEW YORK-The interplay between intelligent networks and Internet Protocol is a battle of the Titans, according to David Berndt, director of wireless mobile technologies for The Yankee Group.
Carriers must avoid getting caught in the crossfire while emerging from this transition as winners that take the best parts of both worlds.
“The rapid rise to prominence of IP has surprised many players in the telecom industry. It has caused a radical re-think of how networks are designed and implemented … The technology of this next-generation network is extensible to cellular networks as well,” said a recent Ovum report on Mobile IP from Iain Stevenson.
Over the longer term, IP offers three transformational promises, said Pulin Patel, co-founder and chief technology officer of IPmobile.
It will lower barriers of entry, enabling a host of new players, including carriers and software applications developers, to add value to the telecommunications marketplace.
It will permit seamless and ubiquitous access by end users to various networks that are now distinct, including cellular, digital subscriber line, cable television, wireline and fixed wireless.
It will offer billing tied to an individual person so that, for example, a person seeking to recharge a prepaid account could do so “as if he were buying gas and found a gas station, instead of having it done on a provisioning basis through an IN platform,” Patel said.
IN, developed in the early 1980s for wireline voice communications over circuit-switched networks, is predicated upon a hierarchical system of intelligence posited in centralized databases, said Eric Zimits, managing director of Chase H&Q’s communications technology group.
“To the extent service providers can maintain intelligence in their networks, this is to their advantage,” he said.
“But the next generation will be IN, with intelligence embedded in nodes or decentralized control centers. Intelligence is getting cheap enough that carriers can decide where to allocate it.”
Centralized agglomerations of always-available network intelligence are a huge waste of money, in the view of Anand Parikh, vice president of marketing and business development for Appian Communications Inc.
“IN creates tremendous unnecessary overhead because you don’t need all that brain power for every call. You should determine how much you need and when. You will need more for voice than data,” he said.
Andrew Silver, senior manager of wireless Internet solutions for Nortel Networks, defined IN as a service architecture and IP as an enabler of service architecture.
“The problem is we see data traffic rising at 1,000 percent a year and voice at 4 percent a year, with data overtaking voice, and this is starting to occur in wireless,” he said.
At the same time, while IP is optimized for data and IN for voice, the concurrent problem carriers face is that they may gain 80 percent of their revenues from voice communications while data accounts for the same percentage of their network traffic, Silver added.
However, telecommunications services providers literally have spent billions of dollars on intelligent networks, and they are not likely to rip them out until they have achieved a full return on that investment.
In the move from the vertically integrated networks dictated by IN to the layered and dispersed networks envisioned by IP, “the preservation of investment in IN is important,” said Bob Stozek, director-advanced network products for Motorola Inc.
“The challenge is to skate in both ponds, supporting both IN capabilities and IP’s distributed intelligence architecture.”
The extent of that investment depends on whether the carrier is an incumbent or a newcomer. However, intelligent network/Signaling System 7 protocol “has come a long way and is an open architecture on its own,” said Sandip Mukerjee, director of mobile Internet product management and marketing for Lucent Technologies Inc. Were this not so, wireless roaming among networks would not be possible, nor would calls between wireless and wireline phones.
That kind of interconnectedness, this time between IN and IP, will be key to a seamless and cost-effective evolution from one to another, he said.
Another big challenge is to raise IP’s quality of service well above “best effort, which works OK for data applications, but isn’t so good for voice,” he added.
Cynthia Heyn, North American director of product unit service capability servers and applications for Ericsson Inc., sees IP as fulfillment of the original IN vision.
“IN was a very big addition to networks back in the early ’80s because it allowed new services in much quicker ways. But it never met all the objectives we were dreaming of, that you could sit down at lunch hour and design a service,” she said.
“Our expectations of IN were a bit too high. Nevertheless wireline and mobile operators have made enormous amounts of money off of IN’s telephony number translation services.”
IP will reach the objective set by IN because call control will be separated from underlying transport, said Bob Wienski, director and general manager of IN services for Illuminet.
“IN removed intelligence from the switch because software vendors controlled the introduction of new features. But by and large, call control remained part of the central office switch, the underlying transport,” he said.
“IP lets the switch switch and, when it needs call control intelligence, it goes to offloaded service logic … IP introduces open standards associated with open systems, the advanced development associated with the Web and WAP’s roaming Internet capability … Innovation will happen at an accelerated rate inside the network and on the periphery … IN will evolve to take advantage of IP.”
Already, wireless carriers are using asynchronous transfer mode and other packet-based protocols, of which IP is one, to “route information in backhaul, but this has not happened yet from the switch into the mobile SS7,” said Berndt of The Yankee Group.
In fact, the proliferation of new competitive local and interexchange carriers building IP networks at a rapid pace is good news for wireless carriers in the wireline part of their transport.
“All else being equal, if I’m a wireless services provider and need to buy capacity from a landline carrier, I will shop on price. The new competitors will be in a better position to drive down the underlying price of bandwidth,” said Appian’s Parikh.
No one can say for certain today what the IN/IP endgame will be, but several offered some possible scenarios.
“Today, fixed, mobile, cable TV and data are four separate networks, but they may not be. In my personal opinion, there will (also) be many more services for vertical markets that would not have been cost-effective for carriers before,” Ericsson’s Heyn said.
Lucent’s Mukerjee called the role of IP in unifying diverse communications networks one of “getting different access points to act as a network of networks so the network knows each device and code is swallowed into the network infrastructure instead of going through gateways.”
The Yankee Group’s Berndt called this stage of evolution the “fourth generation,” one that could arrive in five-to-10 years.
Referring to the competition between intelligent networks and Internet Protocol, Heyn said, “I don’t believe one will take over the other. There will be location-based services, Internet-based services and number translation services.”
However, Berndt said he envisions a day in the fourth generation when “everything has an Internet address.
“Instead of the IN registering the phone, the servers on the IP network will find the nearest base station and ask `is this guy cool?’ “
