Three major U.S. wireless carriers recently announced their support for an NFC payment system called Isis, which would let Americans pay for items at retail stores with their mobile phones rather than using physical credit cards. But the history of NFC in the U.S. isn't encouraging for Isis's success.

NFC, or near-field communications, is a way for two devices to communicate small amounts of data when they're placed about four inches apart. Similar technologies are used for "mobile wallet" services such as Japan's popular Mobile Felica system, where a mobile phone stores encrypted credit-card data, transit pass information, or retail coupons, and can transmit them to readers at stores or train stations with a tap.

One way to look at NFC is that it's like RFID, but smart on both sides. RFID chips like those in Mastercard's PayPass credit cards, Visa's PayWave cards, and San Francisco's Clipper transit cards are dumb chips that store data which can be read or altered by card readers. When you tap the chip on a reader, it performs a transaction. NFC takes this a step further by putting the RFID chip in something that can do its own computing, like a mobile phone. So the phone itself can download coupons, for instance, and put the coupon data onto the RFID chip before it's tapped.

NFC can work with the RFID readers currently installed in places like New York City's Duane Reade pharmacies and San Francisco's Muni stations, with some software reprogramming. Other retailers and transit systems will need new RFID/NFC hardware. Existing U.S. mobile phones also won't work with NFC. You need to build the RFID hardware and reader into the phone, which requires an additional radio and other hardware. So to work with NFC, everyone in the U.S. would have to get a new phone.

This week both Google's Eric Schmidt and RIM CEO Jim Balsillie showed off prototype mobile devices with NFC embedded, but neither device is ready for the market yet.

The M-Wallet That Cried Wolf

NFC causes various technology companies to get all gleeful about hideous neologisms like "m-wallets" and "m-commerce." But NFC has had so many false starts in the U.S. that it's hard to get too excited about the latest initiative.

The technology has been around since 2003. In 2006, Cingular Wireless and Citibank did an NFC trial in the New York City subway. In 2007, Nokia released an NFC phone for the U.S., the 6131, which went nowhere. In 2008, Sprint did a similar trial with San Francisco's BART system that also went nowhere.

NFC trials in the U.S. tend to stall because various powerful interests can't align behind one standard. For NFC to take off, it needs to involve enough wireless carriers, enough banks, and enough retailers to be successful. Previous U.S. trials haven't been able to bring competing carriers and banks together, so retailers and mobile-phone makers couldn't find enough of a market to start rolling out NFC hardware.

There are also security concerns. Back in 2006 there was a healthy debate around whether RFID tags could be hacked to reveal user data, or even skimmed from a distance by wireless thieves. Putting the data in a smart mobile phone can make it harder to skim, as the phone could, in theory, detect attempts to read the data. But hackers could install malicious software on smartphones to get around phone-based security, as antivirus company McAfee explains on its corporate blog.

Isis may have more luck than previous U.S. NFC attempts, as it has backing from AT&T, T-Mobile, and Verizon. The question remains whether Isis will be a major standard or an optional, niche feature. If each carrier only carries one or two Isis phones and it only works with one or two banks, Isis could go the way of previous U.S. trials. NFC pay-by-phone systems, after all, are just like money—and money only works if everyone accepts it.