| Some philosophers claim that true happiness is wanting what you get rather than getting what you want. Obviously, these geniuses never had to make a living satisfying customers. And what the customer wants right now is a practical way to achieve wireless Internet access. With billions of dollars chasing both spectrum auctions and overheated Internet stocks, you would think there would be a way to unite these forces. Although many of the ingredients are within r
each, the recipe itself proves elusive. Whether the laws of physics or the laws of economics present bigger obstacles remains a toss-up, yet search we must in the hope that a sprinkling of Internet fairy dust might make the moribund wireless data business take wing.
The problem is not as simple as it sounds. Dribbling TCP/IP over a radio link does not a wireless Internet make. Potential solutions that could support substantial systems deployments are predicated on some fairly ugly compromises, any of which could substantially undermine the system's attractiveness.
Until Daddy Takes The T-1 Away
The image of surfing the Interne
t while cruising down the highway at 65 mph is the ideal, but it's the first thing that's got to go. High mobility, high capacity and high bandwidth just don't mix. You can achieve two of the three, depending on how much you're prepared to invest, but you can chant "Someday You Will" until you're blue in the face, and you'll still never get them all.
Why? High mobility
requires both a large coverage footprint, lest you surf out of range, and
substantial immunity to multipath fading, giving your signal a prayer of getting through. A large coverage footprint requires cell sizes on the order of miles. Blanketing the landscape with minicells sounds cute, until you actually go shopping for sites. While the FCC may be pondering preemptive regulatory relief, local zoning boards are calling out the militia as irate citizens fight to keep their backyards from turning into unsightly antenna farms. Even if you could subdue the natives, minicells at 65 mph mean macro handoff-headaches, one of the primary sources of dropped connections and performance degradation.
So, big cells it is. Unfortunately, multipath fading, that insidious form of destructive interference where multiple signal copies bounce a
round the landscape and pile up on each other, gets worse as path lengths get longer. The only way to fix it? Lower the data rate. You just can't have it all.
No doubt my blanket dismissal of network nirvana will generate the usual quotient of hate mail from true believers, especially the CDMA crowd, which continues to promise that its technology will do everything for everybody. (Go count the number of fingers in your Rake receivers before you come back and tell me you can solve all the multipath problems.) For now, we're either going to have to stop moving or figure out how to live on the Net with narrowband connections.
|
REPORTS
ANALYTICS
Follow 
