Wireless

April 2, 2003
Wi-Fi or Wide-Area Wi-Fi?
Andrew M. Seybold

UNITED STATES -- There are some within the Wi-Fi community that have decided that the technology is cheap enough, fast enough and popular enough that they can take this extension of a wired LAN and turn it into a high-speed WAN.

It started with "hotspots" and the rush is on to build out thousands of them around the world. Cometa is going to build 20,000 and Toshiba is going to build a bunch more. Wayport and CNN are going to unwire airports that have not already been unwired and T-Mobile is going to build out more hotspots. AT&T Wireless and the rest of the wide-area wireless folks are going to build out even more, and Intel is funding companies that are going to build out yet more.

The leap from thousands of hotspots to wide-area coverage is a leap that neither Evil Kenevil nor his son would take. It has already cost Paul Allen and WorldCom $300 million each (remember Metricom and Ricochet?). Now we are getting ready to hear about it all over again.

Let's look at the facts:

• Wi-Fi or 802.11b makes use of unlicensed spectrum that is shared with microwave ovens, cordless phones, video cameras and all sorts of home devices, including TV-to-TV links, weather stations and home WLANs.
• There are three channels available for 802.11b -- more if you offset them, but three prime channels.
• The FCC in the US mandates power limitations, while other countries such as Australia can run higher power. Output is limited in most places.
• Systems hand off from one access point to another and do not keep track of a mobile unit over an entire network. (Access points capable of handing off a mobile unit are much more expensive.)
• Initiated solely by the mobile user, there is no push capability as there is no way to "find" a user across a Wi-Fi network until he or she signs on.
• Each access point requires AC power. In a power outage the system stops working.
• Unless a mesh network is used, each access point will require a broadband connection back to the Internet or some central point.

Wi-Fi to WAN makes no sense
Trying to turn Wi-Fi, a short-area, unlicensed wireless communications media technology that works great as an extension of a wired LAN inside a building as a hotspot, into a WAN makes no sense to me. Different types of wireless networks are better suited for different types of wireless access from different types of devices. Wi-Fi is a short-range, high-speed extension of a wired LAN that is valuable and has a place in the overall mobile landscape. Wi-Fi hotspots are great for those who enter an area and sit down, open their portable computers and connect to the Internet or back to their corporate information centers via a VPN.

There are some Wi-Fi capable PDAs and some others have a Wi-Fi slot for inserting cards. However, there are a lot fewer PDAs than notebooks equipped with Wi-Fi. After Intel announces its Centrino products, the number of Wi-Fi-capable notebook computers will increase. For the foreseeable future, Wi-Fi is a world of portable notebooks, not mobile devices. While there is talk of VoIP via Wi-Fi and wireless phones capable of Wi-Fi operation, Wi-Fi is about providing portable notebook users with a desktop experience while they are stationary!

WANs combine voice and data, and work with truly mobile devices - small, light phones, smartphones, PDAs and data-centric devices such as BlackBerry email units, as well as notebooks with PC Card wireless modems. WANs only provide 20Kbps to 80Kbps data speeds (compared to Wi-Fi at 5Mbps), but you don't have to spend five minutes booting up a computer and requesting your email. You simply turn on a wireless device and receive your email automatically.

Can Wi-Fi challenge WANs?
Consider this: Soon we will have at least one, and possibly more, wide-area wireless network in the US capable of DSL and higher speeds to notebook computers, PDAs and wireless phones. Data speeds of up to 2.4Mbps will be possible, and truly mobile speeds of 384Kbps will be commonplace to a vehicle moving at fifty miles per hour. If you believe that Wi-Fi will challenge these wide-area networks you are not living in the real world. Wi-Fi systems have their place. They can provide higher-speed access to more people in a confined area than WANs. They can provide near-desktop capabilities to fixed portable devices in airline clubs, at airline gates, in hotels, in conference centers and in coffee shops.

However, Wi-Fi cannot provide ubiquitous coverage across the US and those who believe it can haven't done the math. There is a need to tie all Wi-Fi access points to a back-end network, unless they "talk" to each other. While WANs are capable of providing wireless data service to the "last mile," Wi-Fi is capable of providing higher-speed data service to the last 300 feet.

Even with claims of longer-distance Wi-Fi, a single wide-area cell site provides coverage greater than a large number of Wi-Fi "cells" and a Wi-Fi wide-area network will require the same type of backend infrastructure as WANs (which isn't cheap).

I am sure that some companies will try to make wide-area Wi-Fi work though it makes no sense to me. I am also sure that we will have some high-speed WANs available in the next few years. The marriage of Wi-Fi and these WANs will provide us with the best of both worlds by using the best wireless technology in each instance and tying it all together on the backend. Wi-Fi is about local, portable computer access. WANs are about anywhere, anytime mobile communications that is both push and pull. These are complimentary -- not competitive -- forms of wireless communications!