Whats the point of a wireless data network if you cant access it everywhere? On the other hand, what good is such a network if its slow speed limits you to abbreviated, text-only Internet access? Such are the arguments posed by the supporters of two rival wireless data technologies: 3G cell-phone services that operate at speeds similar to that of low-end DSL connection and WiFi, the ultrafast but limited-range technology that millions of laptop owners use to log onto the Net wirelessly. The one clear edge 3G has had is the convenient combination of voice and data services in a single device. But new technologies that will allow WiFi users to make cheap, high-quality phone calls over the Internet are likely to prove so irresistible that 3G could end up being relegated to use in the outer suburbs beyond the reach of WiFi.

3G phones have some obvious advantages for wireless data networking, particularly ubiquitous access. Last week Czech carrier Eurotel Praha launched the first commercial European 3G service, and more will follow. Boasting long-range coverage from each cell tower, 3G is expected to cover most of Europe and Japan over the next few years while making considerable headway in the United States and elsewhere.

But 3G data transfer speeds currently top out at about a half megabit per second, 100 times slower than WiFis top speed. Though individual WiFi access points offer a range of only 30 to 90 meters, overlapping “hot spots” are already approaching near ubiquitous coverage in those urban areas likely to have 3G services. WiFi also represents an open standard that bridges numerous technologies, whereas 3G represents a variety of proprietary and incompatible technologies designed for smart phones (think Sprint versus T-Mobile).

The introduction of phones that offer voice calls via wireless networks rather than a cellular network, using a technology known as voice over Internet protocol (VoIP), has stirred the debate. Useful now only within a corporate network, such handsets are already being sold by companies such as Alcatel, Cisco, Nortel, and SpectraLink. Even more exciting are coming handsets that will combine WiFi data access with cellular voice. Models from Motorola and Hewlett-Packard are due this fall, with Nokia, NEC, and others planning to follow suit. Most of the early models arent optimized for voice, limiting the WiFi connection to data retrieval. But WiFi/cellular handsets enabled with WiFi VoIP, also known as VoWiFi, should become commonplace within the year.

Admittedly, making VoIP calls over WiFi will be an adventure for several years to come. As 3G boosters are happy to point out, the 802.11 WiFi standards are not designed to handle voice. This fall, however, should see the finalization of the standards wireless multimedia extensions, which tag different types of traffic with various priority levels. The extensions automatically give phone calls priority over data access on the same WiFi network. Without it, voice calls are likely to be garbledif possible at all. Another problem with VoWiFi is security. Existing WiFi security protocols will keep your VoWiFi calls private, but they add latency, creating annoying delays in conversations. A set of protocols due later this year will modify the security algorithms to work effectively for both voice and data. With these problems solved, VoWiFi users will likely enjoy far better voice quality than cellular callers while talking indoors and around tall buildings.

Even so, VoWiFi will never seriously compete with cellular voice until it can seamlessly hand off data access sessions and voice calls from one hot spot to another. For example, a user making a VoWiFi call as she drives down the street needs to be able to keep talking without interruption even as she moves between overlapping WiFi networks. An open standard is expected in a year or two, but proprietary solutions will likely emerge sooner. There are also a number of emerging schemes for doing handoffs between WiFi and cellular/3G for both voice and data. Motorola claims that by early next year, callers using its WiFi/cellular phones will be able to move continuously and smoothly between the two types of networks, starting a VoWiFi conversation in the office, switching to cellular in the car, and finishing up with another VoWiFi network at home. Sorting out billing issues during roaming handoffs is a thornier problem, but T-Mobile and others are working on solutions that should arrive within the next two years.

Manufacturers of WiFi chipsets must also fix a more basic problem: battery power. When WiFi is enabled for voice, it drains huge amounts of power in order to detect, or “listen for,” incoming calls. Because cell phones were designed from the start to listen for calls, they are far more power-efficient. WiFi chipset and device vendors are working on better sleep modes to conserve energy.

With all these obstacles, cellular providers may well be confident that the same VoIP technology that is threatening the profits of long-distance carriers wont soon threaten them. But some have already seen the folly of bucking a technology as flexible, affordable, and multifaceted as WiFi. T-Mobile, which has already acquired the worlds largest network of WiFi hot spots and made it even larger, also recently announced a partnership with HP to resell its upcoming cellular/WiFi-enabled Ipaq. If VoIP service takes off in homes and small businesses the way it has in corporations, other cellular providers may follow. Rather than being an enemy, VoWiFi could be the cellular industrys best friend, giving it an effective tool for supplanting local phone providers. More and more consumers are already using their cell phones in lieu of landline phones at work or at home. If you can take the same cell phone you use at work on the road and even link it up to your home WiFi network to make cheap phone calls, why not cut the cord?