Manhasset, N.Y. With the introduction of a single-chip wireless-LAN and cellular RF transceiver, a small startup may have laid the fundamental hardware building block for ubiquitous wireless connectivity. Quorum Systems Inc. believes its QC2530 could upend today's coverage model of high-cost basestations for every square mile by allowing mobile systems to switch from one network type to another based on location, application and available bandwidth.
In the Quorum scheme of things, a single mobile unit could make low-bandwidth voice calls over a wide-area GSM network while using a Bluetooth headset, then initiate a multimegabyte data download from a nearby Wi-Fi network. It could transfer a GSM voice call to the Wi-Fi network using voice-over-Internet Protocol (VoIP) technology, freeing up the cellular operator's wide-area resources and making it possible to handle more traffic with fewer busy signals or dropped calls.
While most large chip makers are taking a wait-and-see approach to single-chip WLAN/cellular transceivers, a rash of startups are working on solutions. For example, Mobilian Corp., which was recently purchased by Intel Corp., has a dual-radio device on its road map.
"Our goal is to enable revenue opportunities in the Wi-Fi/cellular market," said Bernard Xavier, founder, president and chief executive officer of Quorum Systems (www.quorumsystems.com). "Carriers are looking at mobile applications and voice services like GSM, as well as at the enterprise with video and VoIP and also hotspots with VoIP." The carriers, he said, "want a piece of the $25 billion [application and gaming] purchasing revenue."
The concept of seamless roaming has been around for some time. Before they merged in March 2002, Telia and Sonera had embraced cellular and Wi-Fi integration at a time when other major operators viewed Wi-Fi as a potential sapper of yet-to-be-realized third-generation (3G) cellular revenues. That's changed over the last 18 months, however, and many operators now believe that Wi-Fi and cellular networks may in fact be complementary.
"It's all about simplicity," said Anders Igel, CEO of TeliaSonera AB, a wireless carrier based in Stockholm, Sweden. "This plus a good service level is the biggest driver for growth." Igel believes Quorum's single-chip transceiver is the kind of technology that's needed to "help ease of use and lower overall cost."
"It makes sense for carriers to continue pushing this [convergence]," said Allen Nogee, principal analyst for the wireless-networking group of In-Stat/MDR, citing T-Mobile and AT&T as examples. Merged Wi-Fi/cellular handsets, combined with the move to VoIP, will help carriers reduce infrastructure costs as well as customer churn by capturing enterprise users with unified cellular/land line billing, he said.
"It's especially hard to cover enterprise buildings, and this avoids having to put micro- and picobasestations inside." Then too, wireless-LAN voice calls cost less than cellular calls. Finally, as customers start to use 3G services, high data rates won't be available in dense regions, "and that's where a WLAN really shines," Nogee said.
But even a convergence scheme like Quorum's doesn't go far enough, said Gerald Maguire, a professor of computer communication systems at Sweden's Royal Institute of Technology. Maguire advocates using whatever wireless technology suits the task and taking a two-chip route to high-capacity versions of the tiniest basestations, known as femtocells.
"So much of the complexity and cost of basestations comes from trying to overcome nonline-of-sight issues over wide areas," he said. "It would be much easier to make use of advanced VLSI technology and build a low-cost, two-chip femtocell and put them anywhere coverage is needed," a setup complemented by the ability to smoothly roam in and out of whatever is the most efficient technology at hand.
Already, the move toward merged Wi-Fi/cellular handsets is gathering momentum. Nogee of In-Stat/MDR predicts a worldwide total of 50,000 Wi-Fi-equipped cellular handsets will ship this year, most of them for trials, growing to 525,000 next year and 4.88 million in 2006. "I think there's a pretty big motivation for carriers to deploy this in the home market so they really do start to displace the home line operator," Nogee said. "While the initial target will be enterprise, I think the home market is much bigger. People at home are tired of having two phones."
Forward Concepts (Tempe, Ariz.) is even more bullish, predicting shipments of 4 million or more Wi-Fi-enabled handsets by the end of 2004. "Three leading companies are doing it Nokia, Motorola and Samsung," said Will Strauss, president of Forward Concepts, in an e-mail to EE Times. "Even Qualcomm will sample 802.11b capability in 2004." Thanks to a 78 percent compound annual growth rate, Strauss predicts between 25 million and 30 million Wi-Fi-enabled handsets will ship by 2007, rising to 40 million in 2008.
Though the infrastructure is not yet in place to support roaming and handoffs, "it'll happen over the next number of years," Nogee said, citing T-Mobile's recently announced single-billing scheme for cellular and Wi-Fi hotspot customers.
Startups confronting those obstacles greeted news of Quorum's integrated transceiver with excitement. "I think one chip that does both [Wi-Fi and cellular] is exactly the direction we're going to see," said Mike Klein, president and CEO of wireless-network security specialist Interlink Networks Inc. Jasbir Singh, president and CEO of Pronto Networks Inc., a Wi-Fi infrastructure-software provider, called this type of chip "the next level of convergence."
But chip makers think the device might be too early for the market. "I think it's a good idea," said Frank Ferro, manager for wireless-LAN product marketing at Agere Systems Inc. "However, the attach rate [adoption of Wi-Fi by handset manufacturers] just isn't there yet to justify a single chip. Nor are the usage models and services." Bill Krenik, chief wireless architect at Texas Instruments Inc., said, "I think it's cool that someone has done this, but it depends on the availability of the baseband to go with it and also on the uptake of VoIP."
While Quorum's QC2530 has been initially implemented in a silicon germanium process, CEO Xavier said the plan is to go to CMOS in future generations. Krenik, however, is skeptical. "We tried that before and we found that it doesn't translate well," he said. "If you want to do CMOS you really have to architect from the start for a digital process."