|
Imaginary Infrastructure
By Larry Mittag
Everywhere you turn, someone's promising some amazing new application can be accomplished wirelessly. Problem is, while the products may have the technology to back up these kinds of promises, today's infrastructure just won't support the claims.
Wireless communication is all the rage. In fact, it's about the only classification of tech stocks showing any life at the moment, and everyone is scrambling to declare that they "get it." Conspicuously absent is a clear definition of what "it" is, though.
This situation is exciting to the investment community, since it is a harbinger of change and a useful distraction from the embarrassing failure of the "new economy." This group thrives on anything that will build new empires and move money into their hands and out of those of the people that are stuck in the old way of doing business.
This is fine for the investors, but where does it leave the engineers that are stuck trying to deliver on the promises of a technology that has huge expectations and uncertain infrastructure deployment? How do you explain to your CEO that there is nothing you can do to make your company's new WAP application work better because it relies on the infrastructure of a carrier that has promised far more than it can deliver?
Same old story
This is not a new situation, of course. I have had a long-running battle trying to get a simple residential ISDN line to work dependably out here on the fringes of civilization in Ramona, California.
The problem here is not a subtle one. It is simply a case of too many homes sharing too small a cable plant. A friend of mine inside the industry tells me that when trouble is reported on a line, they have no new cable pairs to use, so they switch it out for someone else's. This closes the trouble ticket very quickly, making them look good to the Public Utilities Commission, but it's bad news to whoever got swapped out. I have ended up on a line that is just good enough that they won't admit it has a problem, but one that requires a resynchronization if I try to send too much data over it. The good news is that Cox Cable just bought our local cable provider, so I get to try a brand new path for data communications.
The frustrating thing is that wireless communication is just the thing to solve this problem. One cell tower could service our entire valley, providing voice capability and data communications. This is a case where there is a crying need, technology that can answer that need, and a client base that can afford to pay for that technology. So where's the problem?
What's the diagnosis?
I ran into at least part of that problem when I picked up a recent edition of our local newspaper. There was a story about deployment of cell towers that was an ill-disguised editorial against the intrusion of technology into the rural atmosphere of our area. The concerns raised were the usual ones regarding health effects and aesthetics, neither one of which was impartially addressed. These are the kind of local politics that have little if anything to do with engineering, but they can play a significant part in the acceptance of the fruits of our labor out in the real world.
Most of the readers of this magazine have little concern for events in Ramona, but this is indicative of what is happening all over the US, if not the world. Products that we are developing are often dependent on an infrastructure to support them, and if that infrastructure is not deployed, then the products simply don't work. In a sense, this is the current state of wireless communications overall.
While it may be frustrating for engineers to wait and wait before the world can see their latest efforts, it can be fatal to companies that depend on revenue from those products. Right now there are hundreds of companies that are betting their investors' money on the hopes that the wireless infrastructure will be put into place in time for them to deliver products and services that can utilize it.
On the other side of the fence, there are carriers that are trying to decide if compelling products will be there if they build out that infrastructure. Each of these two groups is trying to convince the other that they are serious and will deliver on schedule, but quite frankly, neither of these two groups has a great track record on timely delivery./P>
Wireless data status
This is setting up an interesting situation for wireless data. There is really only one technology out there that is fast and robust enough for serious business applications and that is actually shipping today. That is wireless LAN technology using 802.11b. Aside from LAN installations, it is also being used as a development tool and for demonstrations of the possibilities of wireless data. Granted, it would be nice to be able to do wide-area networking so that such demos would be portable, but on the other hand laptops can be equipped with access-point software that makes for very portable networking. These laptops can serve up content to a variety of devices that can also be carried to customer sites.
But isn't this cheating? Why show off 11-Mbps communication speeds that are only active as long as you are within about 100 feet of a laptop? This is hardly a worldwide data access solution. Doesn't this just add to the hypefest around wireless communications?
I don't think so. By showing a potential solution that can be deployed today, technology integrators are very up-front about the limitations of this solution. The fact is that it could be quite interesting for a number of applications around a facility that the company controls.
For example, one company I know of will be using just such a solution with tablets in their warehouse. This is a perfect example of wireless capability within a relatively small area. They are not interested in having the data travel outside of their facility, but within it they will have the capability to access and update the information on their inventory. The best part about it is that they don't have to wait for a carrier to upgrade and they don't have to write a check to that carrier every month. Once they deploy it there are no more charges.
This is a very different type of wireless communications than that pictured in commercials by big-time carriers, but it is just as legitimate an application. Keep in mind that our customer can also provide Internet access through that link, making for a much more useful connection than any wide-area link available today - as long as you stay within that 100 feet.
Carrier alternatives
There are other alternatives to carriers as well. The entire San Diego area recently sprouted a series of odd devices hanging from lamp posts that implement Internet access at a nominal rate of 128 kbps. This is Metricom's Ricochet modem service, which offers wireless Internet service at a fixed rate of $75 to $80 per month. This is offered through a handful of carriers and promises a clean way of going wireless over a number of metropolitan areas.
Wireless data communications are coming, whether the cell-phone carriers bring it to us or not. It is also rapidly going down the pricing curve. The concept of paying for connection time or amount of data transferred is disappearing very quickly. I am quickly being convinced that the wireless carriers may end up missing the boat completely, installing infrastructure too late to capitalize on the early end of the market. This will relegate them to the status of handling bulk business services at razor-thin margins, just the situation the long-distance carriers are facing today.
But what does that mean to engineers? The bottom line is that faster, more-ubiquitous communications are coming. These capabilities will enable a new generation of portable products that will utilize the Internet backbone to operate with data well beyond what the products can carry by themselves. There is a very real tendency to create products that will work within the existing infrastructure and to put off ones that require things that are not in place yet. The future time frame for the latter category always seems so far off.
Solutions are coming
This is the situation I was referring to earlier, where both infrastructure and applications that utilize that infrastructure are required. My point is that the traditional infrastructure names are floundering at this point, but that doesn't mean that the infrastructure is not being built. Other players are appearing on the scene that are not as concerned with protecting existing franchises, and these players will end up on top. But this can only happen if we create compelling applications to use that bandwidth.
Take a moment to sit back and imagine a world where there is a sea of connectivity surrounding us wherever we go. What is possible in such a world? What is desirable? How do we want to control that connectivity so that we don't drown in it? That connectivity is coming, whether it is built from above by thunderstorms created by the major carriers or by massively connecting puddles of connectivity that are created by smaller startups. The question is, what are we going to do with it?
| 
Puzzled by a network processing design issue?
