Amazing new wireless tech: Breaking Shannon's Law | Hacker News Reader

Amazing new wireless tech: Breaking Shannon's Law

onlivespot.blogspot.com

56 pointsjedi_stannis26 comments

"They don't use wireless energy."

Please tell me more about your magical energy-free signal propagation!

In all seriousness, Shannon's entropy limit only applies to a single variable, modulated at a certain frequency. The only reason we apply this limit to radio is because we treat each radio frequency in a given zone as a single variable. In fact, you're perfectly free to use directed antennas or interferometry to establish multiple communication variables on the same frequency.

Ah, and indeed, at 1:19, Perlman mentions each cell phone on a separate spatial domain.

Since they say it's a single cell phone antenna, I'm guessing we're looking at multiple towers and interferometry between them, possibly taking advantage of the intensity dropoff with distance and timing data for spatial discrimination.

That explains the "bubble" comment. Things could be addressed somewhat by location. However, a moving target would be tough to track.

Aren't we using some interferometry techniques to communicate with the Voyager spacecraft?

In context, I think he meant that they don't use wireless energy in the same way as other radios.

I very much hope I'm wrong but it sounds pretty fishy to me. When the guy in the audience asked the question about what modulation this new radio uses he got a marketing babble answer unworthy of a university audience.

Working in sensor networks we could really use one of these breakthrough technologies and in the past year or two I've heard at least three vendors give the exact same pitch but as of today there's no technical detail that gives me much hope.

Also, transmitting 30 miles. Is this an unlicensed frequency?

They have developed on amateur frequencies since the FCC denied them an experimental license.

So, 5.8GHz or 2.4GHz or 900MHz, or what? Are the 40s (MHz) still available?

I wonder how this beats the curvature of the earth, or is it being bounced of the ionosphere, or penetrating through the earth?

It is either a lost work of Tesla or a bad version of this Douglas Rushkoff novel...

http://www.amazon.com/gp/product/1887128905/ref=as_li_ss_tl?...

Amateur, not unlicensed:

http://en.wikipedia.org/wiki/Amateur_radio_frequency_allocat...

While the 40M band beats the curvature of the earth quite reliably, I would expect that they are doing something in the UHF band if they plan on deploying this in licensed cell-phone frequencies.

Also the amateur band lets them transmit at much higher power than the ISM bands:

http://edocket.access.gpo.gov/cfr_2010/octqtr/47cfr15.245.ht...

Sounds like interferometric based MIMO from the description. Not breaking Shannon's law in that case.

It is, I've been doing a bit of poking around in the patent applications filed by this guy recently and that seems to be the gist of it. Unlike normal MIMO though, only the base station has multiple antennas, the mobile stations have one antenna each.

This leads to the downstream bandwidth being much higher than the upstream, but given that video is going to drive mobile bandwidth growth that is ok.

Basically what's happening is that a training signal from each of the base station antennas is constantly being monitored by the mobile station, channel information gets calculated and sent back to the base station.

The base station uses the channel vectors for each of the mobile stations (updated frequently because this is spatial multiplexing and mobile stations... move) to compute the signals for the entire antenna array.

The odd seeming stuff about energy not being transferred has to do with using a sort of inverse spatial coding where you can aim signal nulls at mobile stations that don't want to receive particular signals.

Relevant patents: http://patents.com/us-20110044193.html http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sec...

Can get the pdf here. Nice, Mvandenbergh! http://www.google.com/patents?hl=en&lr=&vid=USPAT741... and here http://www.google.com/patents?hl=en&lr=&vid=USPATAPP...

The "interesting" bit starts at 55 minutes.

So are we supposed to believe that incumbent tech companies have no investments in new tech?

Canon is not working (or already passed) on the same stuff as Lytro? Samsung is not working (or already passed) on the same stuff as Nanosys? Cisco (or whoever) is not working (or already passed) on this type of stuff?

Anybody have examples of modern (within 10 years) stealthy startups/incubators that invented new hardware that revolutionized something?

Lots of networking and database companies have invented new hardware+software combinations, upon which they are promptly acquired by Cisco and Oracle, respectively.

Roku? revolutionize is a strong word, but they've had a lot of impact.

I'd throw makerbot out there as revolutionary, but that's still a very young market. - not stealthy though.

Tesla motors?

When I look at Lytro I think of E-Ink and FingerWorks, both companies that (like Lytro) grew out of academia and both of whose technology has been significant (arguably revolutionary) in the last decade.

Those are pretty good examples but it basically means that there is no reason to get excited about this now since it is many years away from making any kind of impact.

Anybody have examples of modern (within 10 years) stealthy startups/incubators that invented new hardware that revolutionized something?

Tivo (just outside your 10 year limit though)

Flip Video

Roomba

SpaceX

Telsa

Can you really call something like Roomba a revolution? I've met a single person in my life with a Roomba. People everywhere still sweep manually.

You have to look at it context - not every market works as quickly as the consumer tech market.

In the vacuum cleaner market there have been two major revolutions in the last 30 years.

The first was the Dyson. It was launched in 1986 and has slowly taken over the high end vacuum cleaner market.

The Roomba seems to be following a similar path. I know a few people who use it, and it has been a huge success for the company.

I see most of those as "revolutionary" applications of existing technologies/hardware. When I said hardware maybe I meant "technology"?

Seemed more like a plug for opposition of America Invents Act than some new wireless tech to me..

Not to say they haven't made some astounding break-through, but I didn't get the impression that they revealed anything of substance. They claim to have some experimental radios working that are able to beat Shannon's law, but reveal no details around it.

I'm really excited for the potential of wireless technologies like this to break the last mile death grip companies like AT&T and Comcast have. We are at the point where spectrum regulation/management isn't even necessary, except for legacy devices.

I think the meaning of 'law' has been lost here. When I was doing my masters in Info./Comm theory I remember coming across a much ridiculed paper that claimed to break Shannon's law. It was pretty funny but of course I can't find it now since a google search only turns up hits for this.

What are the odds the FCC allows it?

Reply on news.ycombinator.com