Friday, July 1, 2011

Cellular capacity expanded... or not?

Of course, as soon as I put a stake in the sand, one of my associates comes up with a different perspective. Vince D'Onofrio made some excellent points, so I am re-posting here what he wrote to me directly -- I think his commentary warrants a wider audience...
-------------
I have seen all kinds of distributed radio technologies and do not believe result of DIDO technology is really possible the way it has been reported in this article. Except for the bit about overcoming Shannon's Law. I believe it is possible for DIDO to add additional capacity in the right environment, but it still must follow mother natures laws. So it has to have some limitations. Each person getting full cell capacity regardless of the number of users is never going to happen. There are always choke points in the network unless there is unlimited money to spend on infrastructure and unlimited radio spectrum available. Undesired signals from other in-band users will be perceived as noise by the radio receiver and noise chokes throughput, thus limiting data rates. There is no easy way to test the impact of unrestrained capacity without a fully loaded network. Nether Steve Perlman or his cohorts have the capital to do that. Anything can be proven on paper with enough math. But there are reasons why most theoretical patents never get implemented in the real world.

Even if we were able fix radio link capacity problems by distributing the transfer nodes, the switching centers are going to hit a wall. The only way I see this technology working is if all of the content comes from other local mobiles are being used as relay stations. This also means distributing backhaul data traffic (PSTN and Internet traffic) to local fixed connection points or miniature base stations. Otherwise no real-time transmissions will be possible due to interconnection delays in the network.

Though, it is possible that new technology could address the needs of certain high bandwidth users better that what is out on the street today. Since 99% of users would be happy with only 1% of what a base station has to offer, there might be enough surplus capacity for the bandwidth hungry gamer after all, despite the lack of high throughput for all users. We may approach this capacity anyway with LTE Advanced technology which is well on its way to being implemented globally.

Today, most users are not actually annoyed with the speed of their wireless networks. What they perceive as a slow network is the delay associated with provisioning and setup time, which can be horribly long (one minute or more when the network is congested). This delay is not a radio link problem. You can experience these delays even when there is enough channel capacity. LTE takes care of this problem with new protocols that create an Instant On mobile Internet experience. Users will perceive faster network speeds because the overhead delays will be greatly minimized.

My guess is that DIDO is best used for small local networks like college campuses where everyone is talking or gaming with their neighbors and there are relatively few high bandwidth consumers. In a dense urban environment, with 1,000+ simultaneous users possible in a few city blocks, the technology will likely fail to deliver enough throughput for all because of the combination of excessive load and noise.

Though, hats off to Steve Pearlman if he can prove me wrong. I would love to see something truly disruptive change the industry landscape.

-Vince D'Onofrio
Wireless Telecom Consultant
www.radiofrontier.com

No comments:

Post a Comment