Ultimate mobility

Wi-Fi is the rage today, but the real mobile story lies in the convergence of multiple network types, application persistence and multi-modal devices.

You wake up and, after sufficient amounts of caffeine, turn on your mobile network device of choice. Five e-mails grace your in-box, three of them urgent. A quick glance at your workgroup buddy list reveals your boss is already in the office, but the icon next to her name indicates she doesn’t want calls. Your calendaring application shows a customer meeting at 9 a.m., so you tap the “Today” icon and an alert goes out to your staff and colleagues that you’ll be out of the office till late morning.

Later, arriving for your appointment, you finish a voice mail to a friend about basketball tickets. Your mobile device, detecting your rights on the customer’s Wi-Fi network, seamlessly moves the call from the your cellular network to the wireless LAN. You don’t get dinged for the roaming, thanks to the chip-level authentication your mobile device uses and the corporate roaming agreement. When your meeting starts, your mobile device knows not to ring audibly, although color-coded instant messages might pop up.

Cool, yes. Feasible, no. You’ve got all the pieces at your disposal and can use some in combination today, but the applications, the devices and the network infrastructure are not yet agile enough to support such a connected user. Such convergence is at least three years away, analysts and users estimate.

As George Surdu, director of IT infrastructure for Ford Motor in Detroit, says: “We’ve got it all – Wi-Fi, bar-code scanners, cell phones, PDAs – but it’s tough plugging it all together.”

From MAN and LAN to WLAN

Disparate wireless networks likely will come together before convergence at the application and device levels.

In today’s wireless WAN world, AT&T Wireless and Cingular could support roaming between their networks, which rely on Time Division Multiple Access (TDMA) technology. But they couldn’t offer roaming between their networks and the Code Division Multiple Access (CDMA)-based network of either Sprint PCS or Verizon Wireless. The TDMA and CDMA systems send and receive information differently. But as silicon changes and multi-mode phones become available – already happening in Europe and expected stateside next year – users will be able to roam between TDMA and CDMA networks worldwide. Carrier adoption of the International Telecommunication Union 3G digital cellular technology also will facilitate such roaming. Besides being able to handle TDMA and CDMA interfaces, 3G networks can carry data at speeds ranging from 128K to 2M bit/sec. U.S. carriers are expected to begin 3G rollouts next year.

Cellular providers, ever keen to exploit a revenue opportunity in an otherwise flat commodity market, also now are focusing outside their traditional WAN purview. Most are trying to figure out how to get into the Wi-Fi hot-spot business so their users will be able to tap the cellular network for voice calls and high-speed hot-spot access to corporate resources or the Internet. IEEE 802.11-based hot spots offer access at rates of 11M or 54M bit/sec, depending on what version of the wireless Ethernet standard is supported.

“We view 3G and Wi-Fi as complementary,” says Bill Stone, executive director of technology development for Verizon Wireless. To that end, the carrier in August said it would work with hot-spot builder Wayport to provide Verizon users with access to hundreds of hot spots nationwide. Sprint PCS cut a similar deal with Wayport and Airpath Wireless for 2,100 hot spots by year-end.

Hot spots might work well for itinerant users, but they will remain as access points for fixed data applications. Those users cannot expect to walk, or drive, out of hot-spot range and have the Wi-Fi provider hand off their e-mail sessions to their cellular operators, for example. Neither the roaming nor the application persistence works.

For users, the application persistence seems to be the bigger of the two problems. Late last year, Sprint surveyed several of its users in healthcare, telemetry and transportation who said resoundingly that they need application persistence, not seamless roaming. “Those who needed to be able to move between 3G and Wi-Fi were few and very far between,” says Wes Dittmer, senior director of business development for Sprint PCS.

The IETF’s proposed Mobile IP (MIP) specification (RFC 2002) goes toward solving these problems. MIP will let users roam by giving them a fixed “home” IP address and a “care-of” address that changes at each point along their wireless trajectory. And while it might not provide for wireless persistence single-handedly, MIP will go a long way toward enabling it. While router vendors are working on their own versions, these will not likely be ready for deployment or widescale use for a few years. Again, many security issues arise over the discovery, registry and tunneling of the care-of address.

Mobilize those apps

The persistence/roaming isn’t holding back deployment of applications such as e-mail, instant messaging – with or without presence – and Short Message Service (SMS) on wireless networks. Gartner says wireless messaging is today’s fastest-growing enterprise wireless data application. The research firm predicts that mobile SMS and instant messaging will become a regular part of mobile communications in the next two years.

Other applications will move increasingly into the wireless realm, too. “For every vertical, there are two or three key applications that are essential to running the business and would benefit from mobilization,” says Dan MacDonald, head vice president of Nokia Internet Communications. “Make it accessible and useable with small screens, and you’ve got a killer app.”

Toward that end, Nokia in May acquired Eizel Technologies, whose software platform “mobilizes” applications for the wireless environment, MacDonald says. Other efforts to make business applications suitable for use over mobile networks include Motorola’s Cellular Digital Packet Data standard that ported messaging and dispatch applications into the wireless realm.

Any presence-based application will require making signal availability more automated, says Dave Werezak, vice president of marketing for Research in Motion. If users have to keep resetting availability and presence data, then as a practical matter people won’t use wireless applications. “But we can combine someone’s location with presence and use data to say, ‘If I’m not at my desk or am in this meeting room, consider me unavailable,'” he says, explaining that General Packet Radio Service (GPRS), a 3G cellular technique for sending short bursts of data such as e-mail and instant messages, might be one approach.

Because GPRS can be used on TDMA or CDMA networks, and presence is being built into handsets, “We could see it sooner rather than later,” perhaps within the next few years, he adds.

Having the operating system on the device, as opposed to using an appliance, should make presence more widely available and convergent applications more real, says Mike Wehrs, director of technology and standards for Microsoft. The operating system deals with the network, as it does for the location-based E911 services mandated for cellular networks, rather than the user, he says. AT&T Wireless is demonstrating presence technology and is expected to offer a commercial service, although it hasn’t set a date.

Recipe for clunky devices?

Chips might get smaller, less expensive and more sophisticated, but users still need the right mobile tools. As one analyst notes, convergence might help mobile users eliminate a PDA, cell phone, pager, laptop or PC card, but will never get them down to one device. Voice telephony can be integrated into anything, but data needs a larger display and keyboard than many mobile devices offer. Ask early users of the so-called wireless Web of a few years ago-they’ll still likely gripe about slow downloads and crummy graphics.

Multi-mode chips that will handle functions like least-cost routing are most likely to go into PDAs first and high-end mobile phones, says Greg Collins, a director for Dell’Oro Group. “But frankly, we’re not seeing a lot of integration on the chip side between Wi-Fi, GSM and TDMA,” he says.

GSM is TDMA-based and a widely used digital cellular standard everywhere but North America. At least right now vendors want chips that combine DSL and Wi-Fi so they can provide landline and wireless connectivity from a single device, he adds.

Nokia’s MacDonald is a bit more bullish, noting that the PC card is a transitional device. He says mobile phones with 3G, Wi-Fi and Bluetooth capabilities will be available within the next two years. Bluetooth, originated several years ago by Ericsson, IBM, Intel, Nokia and Toshiba, is a specification for so-called personal-area networks. Its signal has a short reach and enables connectivity among different wireless device types and networks on a more dynamic and personal level, for instance as a user progresses through a shopping mall. “It’s only the early adopters who can stomach it right now to accomplish a lot of this,” he adds. “It has to get simpler for the adoption to take off.”

Nokia and others are watching to see how users receive the Mio8380, a mobile phone that plays music, sends e-mail and takes pictures. It’s hardly revolutionary, but it is the first to use Microsoft’s Smartphone 2002 operating system and Intel’s PXA255 200-MHz processor. The phone will be available this fall in Europe for $600.

While the unit is more a consumer play, its ability to handle multiple applications of varying bandwidth requirements portends good things for enterprise users looking for a richer all-in-one or more-in-one device. And as the debut of Microsoft and Intel into the handset realm, it likely will force Ericsson, Motorola, Nokia and others to hasten their own convergence developments for business users.

RIM’s Werezak says the industry can combine as many technology types into silicon as it likes, but for each different radio technology added to the chip, another antenna, code stack, software and electronics have to be added to the hardware.

“We anticipate a converged wireless handheld that would support WAN and Wi-Fi subscriptions, but it would be bigger, more expensive and have a shorter battery life than the BlackBerry or any other existing mobile device,” he cautions.

So while convergence of wireless networks, applications and devices is still a few years out, customers, vendors and service providers all are focused on the potential.

Sweeney is a writer and editor in Los Angeles who has covered IT and networking for 20 years. He can reached at terry@tsweeney.com.