NEW YORK-In order to handle the converged communications revolution, the vision of L.M. Ericsson is the evolution of network reliability to a degree that today is only a dream on a drawing board.
The so-called 10 nines, or 99.99999999 percent reliability, is a Holy Grail far removed from the “best effort” requirements for the first phase of third-generation wireless, said Keith Shank, Ericsson vice president of business management. Even five nines, or 99.999 percent reliability, likely will not be available for several years.
The natural question is why this should matter. For one thing, “90 percent of the American population considers a 33 (kilobits per second) modem high-speed,” Shank said during a recent interview. Furthermore, the human eye cannot even read as fast as data speeds contemplated by General Packet Radio Service, which in Ericsson’s view is the only true 2.5-generation wireless technology.
However, the ultimate goal is universal access for customers in a converged environment in which carriers will deliver a variety of wireline and wireless services. The melding of these diverse communications will create inherent instabilities. To compensate for and override these, a much higher degree of network reliability than what is available today will be required.
While data generally is considered more forgiving of those invisible blips that interrupt transmissions briefly, the kinds of streaming video and audio services envisioned for 3G wireless are far less fault tolerant, Shank said.
Internet Protocol networks pose another challenge in quality control. On one hand, they are desirable because, by design, they decentralize intelligence. Consequently, they are more stable and far less subject to massive and disruptive outages. However, the inexorable transition to Voice over IP means that equipment manufacturers need to develop ways to reduce the latency in Voice over IP that delays and interrupts conversations and slows downloads of information.
CDMA already is an IP-based technology, Shank noted. For GSM technology, Ericsson has developed nodes that permit GSM communications over the Internet, he said.
The overarching goal in all of these plans and developments is provision of ubiquitous communications availability and a seamless, painless end-user experience.
With core networks supporting packet data, “each air interface would have a different structure for the radio controller, but that slides into the core network,” said Michael Coyne, Ericsson’s technical director of satellite systems, at a public forum earlier this summer.
“This will mean that satellite and terrestrial data operators will use the same network and the same backhaul for customer care. Satellite is just another transport backbone. One of the more interesting aspects is that operators working with GSM/UMTS can do provisioning for both transparently,” he said.
“This can also be applied to other access technologies, like plugging in a GPRS handset that gets you to the home network transparently. At least I hope so, because personally I am tired of trying to figure out access when I travel.”
There are other kinks that also need to be worked out. A substantial proportion of wireless data usage, as much as 48 percent by one estimate, occurs inside buildings. And yet, in-building penetration often is spotty at best.
Ericsson, as one example, already has a product that is a small switch installed inside a building in tandem with radio jacks dispersed at strategic interior locations. Today, “Wireless Advantage” only works for voice communications, but the company is working on a version for data, Shank said.
To help carriers manage network capacity, Ericsson also is in the process of developing “Always Best Connection,” a system that will let wireless operators select the preferred way of routing transmissions, whether by Bluetooth, 3G wireless, cable television or digital subscriber line, for example.
“Bluetooth will be one of the key elements to wireless access because it opens up data roaming, provides a standard interface for third-party developers and will create something I call `body area networks.’ It will make the tie-in to information seamless and invisible,” Shank said.
“You won’t need one device for everything, and there never will be one device that does everything for everyone. This will take the onus off of individual devices, which for one thing, will extend battery life. We make phones and don’t want to stop making phones, but they will become information hubs.”
Location-based services also will take some of the impediments away from speedy receipt of useful information, Shank said. At its 3G Wave Lab in Texas, Ericsson has engaged groups of consumers in tests of services that use who and where they are for targeted content delivery.
“If operators supply things like portals do, then there will be mass customization,” he said.
Wireless electronic commerce is another promising development on the horizon in the United States. An entrenched and usable system of automated teller machines in this country has stymied widespread deployment of mobile commerce so far. However, Ericsson expects that to change soon.
“It took us 10-12 years to get used to ATMs. By the time third-generation systems go online, you’ll see SIM cards not just for GSM but also for TDMA and EDGE,” Shank said.
Besides expediting the transfer of features and programs from one handset to another, subscriber identity modules can serve as smart cards used for e-commerce and banking transactions. With a view toward the potential of mobile commerce, Ericsson will include single and double slots for smart cards in the EDGE phones it plans to introduce late next year.
“3G is about combining components to make things simpler and easier. The pieces are starting to come together,” he said.
