Wireless LANs as WANs - The New Threat to Carriers

August 17, 2001
by Derek Kerton

The current times are tough for wireless carriers and hardware manufacturers alike. The market downturn, combined with expensive and limited licensed spectrum, and technological headaches have made 3G successes seem further and further away. The balkanization of 3G technologies makes it unlikely that any effort will get global economies of scale, and it further threatens the customer's likelihood of being able to roam across multiple networks. It seems that it will be years before the carriers can offer reliable broadband-speed wireless connections to the mobile worker.

In the meantime, the wireless LAN world has seen extremely favorable conditions emerge. A clear standard, 802.11b (or Wi-Fi), has all but eclipsed its would-be rivals (HomeRF, OpenAir). Although Wi-Fi already enjoyed the support of most networking companies like Cisco, 3Com, and MANY more, recent announcements of support have came from operating system behemoth Microsoft and chip giant Intel (who abandoned the HomeRF camp). With a clear standard, interoperability becomes possible, consumers can be assured that their brand of hardware will work on any Wi-Fi network, and this encourages purchases. More purchases create scale economies, and prices drop. An ad from Outpost today promoted Wi-Fi PC cards for $90 and a home gateway-access point for $180. That price is not only reasonable for business customers, but also for consumers. In fact, I installed 802.11b network in my home last year, and have been enjoying it ever since. So prices are dropping, cross-compatibility is ensured, a speed of 11MB (vs. 3G's 384k - 2MB) is here today, the spectrum is unlicensed, and usage is essentially free.

So a question forms: are these two technologies, 3G and Wi-Fi, competitors or not? The answer is contradictory. They were not designed to be competitors, one was designed for LOCAL Area Networks, which means an office building or such, and the other was designed for Wide Area Networks, like a multi-state cellular network. However, with the current negative pall cast on 3G, and the glow cast on 802.11b, it seems that the latter is about to make an assault on the turf of the former.

A new breed of company, the Wireless ISP (WISP), is beginning to offer Wi-Fi based Internet service in high-usage locations around the globe. Mobilestar, the best known, is deploying in Starbucks around the country (as I discussed in #23, 1/01). Other firms have deployed in other locations such as Wayport (airports and hotels), SurfAndSip.com (Mom'n'Pop coffee shops), Airwave (Bay area locations) and more. For now, each firm has a wildly different business model, with prices ranging from free to $2.50 for 15 minutes (information highway robbery.) While it will take some time for business models and pricing to shake out, the key is that they all share the same underlying technology. Expect rapid growth followed by consolidation.

Each Access Point (such as one Starbucks) requires a relatively cheap installation of a DSL modem, a gateway, and a wireless access point. This can run less than $500, with additional monthly DSL fees. A user with a PCMCIA Wi-Fi card need only adjust a setting to connect to the access point. From that point, a browser request for ANY web page would return the WISP's login page. After login, the Internet would be fully accessible.

The threat to existing phone carriers is that this low-cost, easily used technology could mount a joint attack for the high-margin customers. Essentially a WAN is nothing more than many interconnected LANs. If the WISPs were to cooperate or consolidate, or if one becomes an 800-pound gorilla, Wi-Fi could offer access points in most of the locations where connectivity is desired.

A case: Let's say that a high-margin user commutes from Jersey to NYC every day, and carries a laptop so as to plug in at work, use a cell-modem to do email on the PATH train, and a DSL connection to work at home. This user is valuable to the carrier, as he burns high-margin prime minutes of cellular data use while on the train. Now imagine that a WISP strikes a deal with PATH to connect the train to the Internet, and install a wireless LAN on the train. Our user's needs are fully met by a Wi-Fi solution. A WAN solution that also works in Tulsa or Houston is not valuable to this user, but faster, cheaper, more reliable is. This user will abandon the cellular carrier in favor of the WISP.

Here's my wrap-up: The Carriers: hate Wi-Fi. They have invested heavily in 3G and 2.5G, and don't need any new problems. While they should try to adopt Wi-Fi into a hybrid solution, they will instead choose to try to resist it.

The Networking Giants: love Wi-Fi. They finally have networking products that they can sell to the consumer, the revenues are needed, and it's a brand new market for them. Their only problem is that they don't want to offend the carriers, to whom they sell their cash cow products.

The Consumer: loves Wi-Fi. Finally a technology that's fairly easy to use, works as well as promised, and is cheap. Now one can walk within range of an access point, and be up and e-mailing in two minutes.

The WISP: obviously loves Wi-Fi, as it is the core of their business. BUT, they have a difficult conundrum. Should they try to win the market and offer the most access points, or should they make deals to inter-operate with their brethren to offer more access points? I expect that in the short term, they will go it alone. Soon, it will become clear that to offer true value to consumers, they will need to have seamless interoperability (shared access rights, or "roaming") and they will consolidate or cooperate. Of course, like IM or auctions, there will be one 800-pound gorilla that chooses to go solo. Look for Mobilestar in this role.

802.11b vs. Bluetooth: These two are often incorrectly positioned as competitors; not so. They are complementary. The only similarities are they use the same unlicensed frequency and they are both radio protocols. Bluetooth was designed with specific parameters, namely cheap chip cost, low power requirements, short-range, and small size. These parameters make it well suited for manufacture into cell phones and PDAs for interoperability connections. Wi-Fi is better suited for installation into laptops for LAN connections. For more on this: Table Comparing 802.11 to Bluetooth (PDF)