A couple of clarifications and comments (new whitepaper is coming with a lot more info):

The iPhones (and dongles) we demoed have no software modifications at all, and there is no additional overhead beyond the standard LTE protocol. pCell works with unmodified Android LTE phones, too. (We had wanted to post a lab demo we did of the Galaxy S4 and an Xperia, but we just ran out of time.) Out-of-the-box compatibility is essential for rapid deployment. And, yes, we are meeting many of the core goals of 5G today (e.g. unlimited 4K UltraHD streaming) using Rel. 8 LTE devices in LTE spectrum.

pCell is indeed protocol agnostic, and can concurrently support different protocols in different pCells in the same spectrum. For example, we can support unmodified LTE phones in their own pCells, while concurrently supporting lower cost/lower power devices with far lighter protocols (that are lower latency), since we don't need all of the complexity of LTE. For example, there are no cells, no cell edges (and no need for CoMP), and no cell handoff.

I'm waiting for something to connect the dots and realize this is a vastly more efficient way to use white spaces than anything currently on the table. White spaces will be full overnight with current techniques. With pCell, they will never be full.

More data is coming later. Apologies. We have been utterly overwhelmed in incoming inquiries since launch.

But, I will confirm this: pCell is indeed a much bigger deal than anyone has yet touched on. The "tubes to transistors" analogy is not just marketing speak: Compared to cellular, pCell is far more reliable, enables much smaller and lower power device and can be continually extended in density. Tubes had physical constraints that limited their reliability and scalability. Transistors did not. Cellular (and other interference avoidance protocols like Wi-Fi and cognitive radio) have a physical constraints that limit their reliability and scalability. pCell does not (as far as we know). Cellular has stalled in scalability. There is an entire era of innovation in front of us with pCell. - Steve Perlman

I think the idea of focusing multiple low power radios onto a single point so that they coalesce is simply being rediscovered. While we could do this before the real ingenuity is being able to rapidly update all the parameters when the point you want to focus onto is rapidly changing.

Another scary idea: While we could beam wireless power imagine if this was in any way weaponized. With all these stations beaming power to everyone what would it take for an overwhelming number of them to direct it at a target with the purpose of eradicating it? Certainly not a laser in the traditional sense but surely something just as destructive.

Nice explanation, does it bother anyone else that you could set up a bunch of antennas such that they could create a lethal 1 cm ball of energy anywhere in the space they can see? Sure they think about powering devices but what about a couple of Watt-Seconds of energy appearing in your frontal lobes? If Perlman can do that its kind of the ultimate 'border' fence is it not?

(I'm tempted to write, "All you would need is an unscrupulous engineer to develop a targeting system ...")

If the latency of pCell is low enough then you could have a thin-client phone tunneling into a cloud OS. Now get rid of most of the CPU/memory hardware complexity on the phone (which should lower power requirements for the phone too). If you can wirelessly send energy then get rid of most of the battery.

Suddenly you've got a very, very thin phone with infinite battery life and computational power only limited by the server you're hooked up to. It's the perfect mobile device. Do the same thing for a laptop and I'm in paradise.

Assuming pCell works as advertised it's a huge deal.

It allows you to cover a city in cells for insanely cheaper than current cell technology and provide far, far better reception and bandwidth to far larger numbers of customers AND it works with existing LTE gear while also affording simpler, cheaper, and lighter-weight new gear and reducing power consumption across the board. What's not to like?

I don't think anyone who has heard about pCell thinks it isn't a huge deal, if it works.

One example of a major potential problem: how well does it track erratically moving objects (e.g. most cell phones?), in built-up areas with lots of signal reflection?

Steve Perlman always comes across as a snake oil salesman, but I didn't think he could do OnLive either, and that technology works fantastically. As a result, I'm pretty excited about pCell, even though it's being preceded by a wave of marketing hype.

As many positive things as I have heard about it I have one question: will we ever actually see this as an actual product? I feel like I hear about so many cool things which never come to fruition and I have a nagging feeling that this might be one of them. Why? Because, as I've read, the companies would need data centers to manage all of these connections. What interest do these companies have to risk all of this money (a recurring cost, might I add) when people probably wouldnt spend more for it? I dont profess to know, so can someone who has actual information enlighten me?

"With this technology, it's conceivable that 5G wireless could displace both cable and DSL connections within a few years, as is claimed in the presentation."

A few years?! This statement should give landline data providers, like Cox and Comcast, some serious pause.

The original Columbia presentation is actually pretty cool ( https://news.ycombinator.com/item?id=7274288 ) I thought I'd mention it as it only got a couple of votes and sarky comments at the time but having just watched it all I think it really look like a major breakthrough, maybe the biggest one of the year and could be giving us 4k video all over the place before long. One of the interesting things is it will work with existing unlicensed spectrum, existing handsets and the backhaul can be done with line of site links on rooftops so it can be rolled out pretty much straight off - well they are talking a year or so.

I wonder if wireless power like this could be used to provide power for launching spacecraft and this avoiding the tyranny of the rocket equation?

> Will it work for WiFi as well? It's protocol agnostic, so it could work in unlicensed spectrum as well. The issue is that you don't have complete control over all the other transmitters, so you can't coordinate them.

From the moment I heard of it, the first thing I thought about was: this could make long range Wi-Fi possible (and therefore meshnets and carrier disruption possible).

Seeing this tech being used by wireless carriers would be cool, but I really want to see it (or something like it) being used in new mile long Wi-Fi standards. Now the Wi-Fi Alliance only needs to make one for us.

Long range Wi-Fi could also make communication between self-driving cars much more feasible (got your attention now, Google?) - although that presents a pretty huge security risk, too, but I imagine the industry already wanted them to be connected to the Internet, so it wouldn't be much worse.

    > Humans use <todo>.

Indeed.

Wireless power transmission will change everything in hardware if someone can pull it off. I'd bet Apple is waiting until they can do low power wirelessly to launch their smart watch.

Don't you still have large power losses from the inverse square though?

This is a very nice explanation of not only pCell but of what modern communication systems are like (802.11n has MIMO/diversity but also channel estimation, etc)

People may find this interesting: http://en.wikipedia.org/wiki/Space–time_block_code

Is there any reason this same technology wouldn't work with light? Wouldn't that be a phased light array?

With seemingly unlimited mobile bandwidth and speed, the cloud gets a new dimension too. Perlman has already proven that even the most demanding applications, realtime games, work well in a streaming scenario even today. With something like this, i could see my desktop in the cloud being accessible from anywhere through a mobile thin client in my pocket. Hook up multiple monitors/keyboard/mouse to it or just use the device screen itself, with computing power only limited by the cloud provider. Technologies like OnLive or PCoIP show thats its possible even today, all thats missing is the infrastructure to support it and pCell could be the answer to that.

Linked paper:

http://www.rearden.com/DIDO/DIDO_White_Paper_110727.pdf

While I think charging cell phones and Telsas is a problem, getting cheaper than petrochemical energy is the key. So I have always thought that some sort of space-based energy source was the solution but how do you get it back to earth? 1000's of small solar cell beaming back energy concentrated at one point would be a great solution. How much would this cost would be interesting to find out?

Sort of reminiscent of ESWL, which breaks kidney stones, except in that case they "aim" the sound waves with an elliptical waveguide.

At first I thought the technical challenges inherent in maintaining state for a web of arbitrarily placed transmitters cooperatively targeting thousands of moving devices would make this technology impractical for deployment in the wild, but now I believe. pCell is going to be revolutionary. I can't wait to get rid of my consoles and stream new, triple-A games over it.

USA Federal Communications and other gov agencies EU's for example wanted to move towards telecomms spectrum being a commodity where the device hands off to other carriers seamlessly to handle spectrum scarcity

pCell takes a different tact ..seems that the way to get telcomms to move would be through government mandates tied ot say spectrum sales.

Sorry, its just beamforming. Its not new, its just marketing.

its not all that different from having an 802.11n MIMO hotspot in your house.

I like to think of it as that conventional cellular is using a broadcast medium for "narrowcast" communications.

If you put it that way it's obvious that we're currently Doing It Wrong, provided there's a practical alternative. It also helps people get over their reflexive "But Shannon says…" reaction.

A friend at CERN commented with this: " This is EXACTLY the same procedure we're using to correct experimental data from distortions introduced by detector effects (unfolding) - http://cds.cern.ch/record/1600778 "

The idea of deliberately burning out devices is interesting, and probably feasible, given a dense enough array of powerful enough transmitters (bear in mind that one of the key benefits of pCell is that it uses far lower-powered transmitters). But it seems to me the inverse square law is going to make it a very inefficient weapon.

It doesn't have to be high power to be effective. Google search on "russian microwave US embassy" (not Bing search, unfortunately), which yields some interesting background on the Russian's beaming microwaves at the US Embassy. It was within the US legal limit at the time. Health problem ensued. Making an adversary sick without noticing may be more effective for some situations.

First of all, your body would likely screw up the interference patterns enough that it wouldn't be a "ball of energy" in the center of the person - it would likely mostly diffusely heat them.

Second, the amount of power to burn a person is many orders of magnitude higher than needed to receive a signal. The transmitters would have to be extraordinarily powerful. I can only see this happening if they plan on making a wireless power transmission service.

Also, a 1cm ball is extremely optimistic, given that LTE frequences have several cm wavelengths. It would also require an enormous number of pCell transmitters to concentrate their transmission into such an area. For communications though, even concentrating into a 20m ball would be a huge improvement over existing technology, so I don't know if there is incentive to place more antennas than required for that.

I just replied to another thread with this exact same thought, then looked down to see yours. And yes, it is scary... If I understand it correctly, the targeting system just consists of normal receivers hacked to offset their targeting feedback by a distance that would park the energy ball inside the victim.

In any of the applications where this is deployed broadly, that is, where we cover our cities in transmitters, then we'll have effectively given over this remote kill capability to anyone who can do trilateration.

You would basically already do that with mobiles. When people use for talking they keep them <1cm from their head. And while 70mph is indeed great speed people sometimes take high speed trains (200mph) or don't turn off their mobile on an airplane (~300mph in vicinity of ground)

Tin foil hat? (Ok, maybe a Faraday cage hat.)

If this kind of technology were to fall into the wrong hands would we even know? Would it appear as though the victims had a heart attack or stroke?

The first question that comes to mind is this: how can we defend ourselves from being brain burnt?

You can already cause injuries with phased array radars, right?

What seems to me a more realistic concern is that you could create a beam of energy that would be lethal to an individual device. Brains are typically a lot more durable to EM radiation than electronics.

And more generally, it seems like relying on coordinating devices you don't physically own opens you up to all sorts of new kinds of DoS attack.

Actually if they could make it that small wouldn't it also be useful for surgery? Imagine non invasive surgery.

Nope (well, yes, as distance -> infinity)

Imagine three boats making waves such that the ripples constructively interfere in some spots and destructively at others.

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

I think with the system they are demoing for communication that is true. They seem to use non directional transmitters of about 5W which cover an area of 'four or five blocks' ( http://www.youtube.com/watch?v=5bO0tjAdOIw&list=PLbOnG9yNO5u... )

A system for transmitting power, which I don't think they have built on any scale, you'd probably want to focus a beam towards the receiver with the power loss being proportional to the amount the beam spreads out.

This is more like radiating narrow transmission beams, not symmetrical spherical transmissions in all directions. There will be losses with distance, sure, but nowhere near as bad as omnidirectional broadcast.

Yes, but constant factors matter.

Consider the simple case of a flashlight bulb by itself, versus in the flashlight (with a reflector). They are both inverse square brightness, but the constant factor in the flashlight case is much smaller.

Similarly this will make coherence areas that will be very small. Wasted power will be approximately inverse-square, but the falloff will be small enough that for short distances you can ignore it.