Several news outlets today are covering entrepreneur Steve Perlman’s DIDO technology, which claims to support “impossible” levels of capacity over wireless networks. It’s heady stuff. The descriptions promise 10 to 1000-fold gains by delivering individually formed beams to each device, in contrast to current sharing mechanisms that suffer from interference.
For me, what’s important is how this illustrates the point that I learned from innovators like David Reed, Tim Shepherd, and Paul Baran: our conception of what’s possible through wireless communication has been radically restricted by radio engineering practice. The limits we take for granted aren’t dictated by the laws of physics. They are artifacts of the way we design systems. WiFi and other unlicensed technologies are just the first example of what can happen when we start to relax our assumptions. And we’re going to need more such breakthroughs to keep up with skyrocketing demand for wireless data.
Steve Perlman is a genius at developing and commercializing novel technologies, so DIDO shouldn’t be dismissed as a fantasy. There is apparently a white paper coming out today, but based on the press reports, the technology sounds intriguing. I can see some limitations: it depends on massive levels of cloud processing to form and reform beams as network topology changes. And as far as I can tell, it’s a forklift upgrade: it would seem to require dedicated new wireless transmission systems as well as dedicated frequencies, as the interference avoidance relies on a completely centrally managed spectral environment. These aren’t fatal problems, but they would limit the deployment path and applications of the system. And as Perlman admits, getting wireless carriers to sign on will be a big challenge, given how much investment they have in the status quo.
One should maintain a healthy skepticism of predictions like, “We believe DIDO wireless will completely transform the world.” Steve has had commercial success with a stunning range of innovations, but it’s worth keeping in mind that none of them — WebTV, Moxi, advanced motion capture for films, and Onlive — quite reached the lofty heights that were predicted. And remember all the excitement about Dean Kamen’s Segway? Entire cities would be rebuilt around it, we were told by the likes of Bill Gates and Jeff Bezos. Like most of Perlman’s technologies, it wound up being a remarkable invention that had a big impact… just not nearly at that scale.
But these are cautionary notes of degree, not of kind. I have no independent basis to know whether DIDO will work. But it could. And that’s exciting. We simply can’t get to where we need to be in terms of wireless capacity with incremental improvements. We need to free up more spectrum through incentive auctions, and we need to promote innovation in white spaces technology, and we need to do a whole bunch of other things. We also need to look toward fundamental advances in the means of wireless communication. As I said yesterday, experimentation is crucial for the future of the broadband ecosystem.
I’m hopeful that DIDO will be one such example. Let’s have more.
Update: EE Times quotes a Berkeley engineering professor expressing skepticism about the novelty of DIDO. Commenters on the article do as well. Doesn’t change my view about the need for creative experimentation outside the dominant wireless paradigms.