Jump to content

Steve Perlman claims to have a new approach to revolutionize wireless networks w pCell/Artemis


TaiKing

Recommended Posts

Great questions. We're glad you're interested.

 

at Columbia University for the Engineering faculty and students might answer a few of them...

 

Definitely more information provided there, but it would have been nice to see the Q&A. Though I must say the more I learn about this and come to understand what this is doing and where it could go I find myself excited to start to see it. I'm even hopeful that Sprint could jump on this relatively quickly with their past history of embracing new technology quickly and the fact they are really just getting started with the deployment for band 41 could give them an opportunity to pursue this route.

Link to comment
Share on other sites

I am pretty sure that at some point, the tech will be tested in a Faraday cage and analyzed via a scope. Then tested in a real world scenario as well.

I want 20 scopes/analyzers watching, spaced out around the locations.  I want to make sure there isn't too much similarity in the signal.

Link to comment
Share on other sites

I don't know if I would call bs on this just yet.  If it is for real then watch out.  This guy has been behind some nice projects in the past and it they have some pretty slick demos.  That said there is a lot of vaporware in the world, with any luck this will be concrete.

And there has been some nice vaporware with his name on it.  Like I said, I would like to see this work, and if it does, for someone to put it into service.

 

It just feels too much like a Quixtar seminar to me, especially in how much is talking tells so little about what's really (or allegedly) going on.  I wonder how many radios one will have to send to hit gold diamond whatever status.

Link to comment
Share on other sites

Great questions. We're glad you're interested.

 

at Columbia University for the Engineering faculty and students might answer a few of them...

Thanks for the link, it provides slightly more information than the initial demo video posted yesterday, even though we'd love to know more about the underlying technologies.

 

Because of the scalability, backwards compatibility with 3GPP/3GPP2 and potentially lower cost, this could be a dream come true for smaller nationwide operators like Sprint and T-Mobile since their footprint and business model is heavily depended on densely populated metro areas. If pCell technology is as good as it seems, operators jumping on it early will have a clear competitive edge in terms of performance which they could use in their marketing campaign.

 

I'd also like to know more about the logistics of the actual deployment (does it require Artemis contractors or only wireless provider), and does the business model call for recurrent licensing fees or one time upfront cost of deployment? 

 

Thanks again!

  • Like 3
Link to comment
Share on other sites

Thanks for the link, it provides slightly more information than the initial demo video posted yesterday, even though we'd love to know more about the underlying technologies.

 

...

Very slightly more, unfortunately.

 

 

 

Here's a simple question...  How will this work in conjunction with more broad-reaching LTE solutions?  The system as presented would work well in a city, but not in the middle of nowhere.

 

 I live in a state with large rural areas which are covered by sites spaced at the limits of their radio technologies.  I see lots of areas with 5-10 mile site spacing with 300-400' antenna heights, whose radios are running at their maximum possible design power.  Some of these cells should be split due to changes in capacity requirement, but it rarely happens.  This leads to LTE speeds of pretty much ether 25 Mbps or 5Mbps, with little in between.  Either the cell is overloaded but not quite enough to be split, or it's just sitting there providing great coverage to the farmer's cattle and seeing 10-20 actual device connections per day.

 

If this truly is compatible with conventional LTE technologies at the phone's end, how will it handle the handoff with conventional networks?  I would hate to see a separate band required, but would hate even more to see a zone of alienation between the two LTE technologies to prevent interference.  Even though your system is intended to leverage interference to its own benefit, I cannot see it being able to do so with that which is caused by neighboring conventional cell sites, as the signaling would not be precisely known.  This could present some major spectrum issues, thereby substantially increasing operator expense.  Frequency reuse is a headache now.  I would hate for it to become a migraine.

Link to comment
Share on other sites

I want to know if there is anything technically unique about this product that separates it from 'plain vanilla' Small Cells/HetNet, CoMP, etc. solutions being built into upcoming 3GPP releases. And if there isn't, what is to stop the ALUs, Ericssons, Airvanas. etc. of the world from offering identical products that also integrate better with their existing deployments?

Edited by CaptainSlow
Link to comment
Share on other sites

Very slightly more, unfortunately.

 

 

 

Here's a simple question...  How will this work in conjunction with more broad-reaching LTE solutions?  The system as presented would work well in a city, but not in the middle of nowhere.

 

 I live in a state with large rural areas which are covered by sites spaced at the limits of their radio technologies.  I see lots of areas with 5-10 mile site spacing with 300-400' antenna heights, whose radios are running at their maximum possible design power.  Some of these cells should be split due to changes in capacity requirement, but it rarely happens.  This leads to LTE speeds of pretty much ether 25 Mbps or 5Mbps, with little in between.  Either the cell is overloaded but not quite enough to be split, or it's just sitting there providing great coverage to the farmer's cattle and seeing 10-20 actual device connections per day.

 

If this truly is compatible with conventional LTE technologies at the phone's end, how will it handle the handoff with conventional networks?  I would hate to see a separate band required, but would hate even more to see a zone of alienation between the two LTE technologies to prevent interference.  Even though your system is intended to leverage interference to its own benefit, I cannot see it being able to do so with that which is caused by neighboring conventional cell sites, as the signaling would not be precisely known.  This could present some major spectrum issues, thereby substantially increasing operator expense.  Frequency reuse is a headache now.  I would hate for it to become a migraine.

His video claims that a phone can switch between p-cell and cellular.  See around 45:00.  How it does it, I don't know.

Link to comment
Share on other sites

His video claims that a phone can switch between p-cell and cellular. See around 45:00. How it does it, I don't know.

Yeah, I saw that, but didn't see details. LTE and CDMA2000 hand off too, but it isn't a handoff of equals. Same with LTE and UMTS, or UMTS and GSM. 800 MHz LTE hands off with 1900 gracefully, but now we are talking about two different bands. If this is the case, and I am over at Sprint trying to deploy this, will it use up all of my 2500 MHz spectrum and preclude me from deploying conventional LTE on anything but 800 and 1900?

Link to comment
Share on other sites

Yeah, I saw that, but didn't see details. LTE and CDMA2000 hand off too, but it isn't a handoff of equals. Same with LTE and UMTS, or UMTS and GSM. 800 MHz LTE hands off with 1900 gracefully, but now we are talking about two different bands. If this is the case, and I am over at Sprint trying to deploy this, will it use up all of my 2500 MHz spectrum and preclude me from deploying conventional LTE on anything but 800 and 1900?

My understanding is that it only uses as much spectrum as you want the end user to have access to.  So if you want each end-user to have 10 mhz then you only deploy 10 mhz.  Now backhaul is a big beast with this, but if I am understanding his claims correctly, sprint could basically get away with using almost none of their spectrum and still have more end user bandwidth than they have now.  I will definitely be following this.

Link to comment
Share on other sites

My understanding is that it only uses as much spectrum as you want the end user to have access to. So if you want each end-user to have 10 mhz then you only deploy 10 mhz. Now backhaul is a big beast with this, but if I am understanding his claims correctly, sprint could basically get away with using almost none of their spectrum and still have more end user bandwidth than they have now. I will definitely be following this.

actually, backhaul won't be as big a beast as you might think it would be. With the processing they're using their data center for, it looks like this would do double duty with its backhaul, just like it would do with his radio signals.

 

I, on the other hand, am talking about spectrum reuse between technologies. Namely, this network, and conventional LTE networks.

Link to comment
Share on other sites

Hi, I let Steve Perlman know about your questions and he just got to me and said he read through all of you discussion posted so far. First, he wants you to know an updated and far more detailed white paper is coming. He and Antonio had planned to have it done before the announcement, but they just ran out of time (and hit limits of human exhaustion), largely due to an unexpected ramp up demos and tech due diligence meetings covering some of the very topics you are discussing. He apologies personally for not having the whitepaper completed. This is a very different approach to wireless and he knew there would be significant questions asked that we are happy to answer. The technology not only works as advertised, but we've spent years not only making it computationally tractable and scalable, but also making it extremely practical and fast and inexpensive to deploy.

 

These answers from Steve do not answer all of your questions (the ones require long answers are best answered in the white paper), but he wanted to toss over a few. What's below is directly from Steve:

 

1.       pCell has far greater range than cellular. We can crank up power to reach a user. Cellular had to limit power to avoid interference with adjacent cells. There's some good Matlab figures in the Columbia University talk that show this (e.g. the Serendipitous antenna layout figures)

 

2.       Yes, pCell tech will give any operator (small or large) an immense competitive edge (and its subscribers far better service, in terms of throughput, consistency, coverage, lower latency and device power consumption)

 

3.       There is no modification, software or hardware, to standard LTE devices. If you take LTE test equipment made to receive and analyze a standard LTE signal and connect an antenna to it in a pCell coverage area, the device thinks it is receiving a standard LTE signal, just one that is extremely high SINR with no other users sharing the eNodeB (the LTE protocol transmitted from an LTE cell tower). An LTE phone's antenna sees the same thing: from its point of view, it is in a standard LTE cell with very high SINR, and it is only user in the cell.

 

4.       We expect to work with other infrastructure companies who will manufacture Artemis-compatible pWave radio heads and deploy them, setup front haul, and set up data centers for the C-RAN. We'll supply the software that runs on generic Linux-based servers. For example, the entire Columbia University demo, including 8 virtual eNodeBs and generating their physical layer waveforms that fed into pWave processing, and generating the physical layer waveforms for 8 radio heads was implemented in our own software-defined radio (SDR) code running on two dual 8-core Intel motherboards. It was completely real-time, with room to spare. So, we actually see ourselves as a cloud software radio company, and we're happy to work with anyone who is actually going to deploy the tech (and not interested in working with anyone planning to just sit on it).

 

5.       pCell works in any band, not just mobile and not just licensed. We can't comment on our carrier relationships, but we are working in more than the one licensed band referenced in this discussion group. And, we also are working in unlicensed spectrum. As it turns out, we ended up using 3 bands, 1 UHF, 2 microwave (can't say precisely which, because it would identify the partner that owns the spectrum) across the demos in the various videos and Columbia, just because what was set up when we had a window to shoot and/or where we had a license to do the demo (licenses are location-specific). We work in unlicensed spectrum (compatibly with LTE in 900MHz, which is unlicensed in the Americas), and also have done extensive testing in HF for ionospheric pCell up to about 45 miles away, but traveling 300 miles up and back from the ionosphere. pCell tech works great at even such long distances, and it also verifies that pCell size remains the same regardless of distance (it's proportional to wavelength). So, the range is really only limited by the MRC  of the power of the pWave radios reaching the user.

 

6.       The Columbia University demos were all done with only about 1mW of power from each pWave, including the 10MHz multiple 4K/1080p video demos. As you can see from the Matlab sims, we achieve enormous efficiencies over cellular. Cellular seeks to maintain SINR across the entire coverage area as best it can, resulting in most of the coverage area having low SINR. We only maintain high SINR exactly where the LTE devices' antennas are. The white paper will go into this in more detail.

 

7.       Lastly, the white paper won't answer every question down to the last detail (if it fully described all the layers of the pCell system as well as all the techniques we developed to make it computationally tractable and commercially practical), it would be volumes. For example, simply the technology we developed to achieve perfect reciprocity with uncalibrated RF chains (e.g. iPhones, dongle, Android phones) is an entire body of work itself (and certainly has many applications outside of pCell) that would take more than a whitepaper to explain. It was just one of the many things we had to develop to get pCell to be a drop-in wireless tech for carriers (which we knew was the only way we'd ever get something deployed rapidly).

 

So, again, apologize for announcing something "too good to be true" without the white paper backup. But we had already scheduled the Columbia demo and had to announce then. Sincerely, we had hit human levels of exhaustion. it is no exaggeration that we've been all getting 2-3 hours of sleep at night for the last few weeks. But it is coming. It won't answer everything, but keep asking, and we'll try to get you answers. (If you can't wait, feel free to dig through the patents that cover a lot of it.)

 

Bottom line: pCell works superbly as advertised. We appreciate and invite your hard questions. And, also, it may well be there is something we didn't think of that your questions may expose that we have to address, and indeed we will.

 

More coming later.

 

*SGP*

  • Like 18
Link to comment
Share on other sites

I must say I like that he is taking the time and addressing the questions that are being brought up here. With everything that must be going on having just announced this and the excitement surrounding it, the fact that he would take the time to look through this little thread we have going and personally begin to address the questions being brought up is something I would have never expected.

 

I also respect that you talk so openly so early about not working with groups who would just sit on something that could shake up the market like this. As Sprint fans here I'm sure many of us have wondered at some point what could be done to unseat VW and AT&T at the top of their mountain or at least level the playing field for the smaller operators and I could actually see a way here to at least significantly level the playing field. I imagine with as significant of a change as this technology would be, if any of the major operators adopt it all of them would have to if they wanted to stay competitive, and I would hope it would be easier for the smaller operators to reach a point of network parity with their competitors so that the selling point would become less about the network quality and more customer focused but i'm afraid the duopoly will do everything they can to complicate the process.

  • Like 6
Link to comment
Share on other sites

I must say I like that he is taking the time and addressing the questions that are being brought up here. With everything that must be going on having just announced this and the excitement surrounding it, the fact that he would take the time to look through this little thread we have going and personally begin to address the questions being brought up is something I would have never expected.

 

Ditto!

  • Like 2
Link to comment
Share on other sites

I wonder if any of the carriers are looking to purchase the company outright?

I'm sure duopoly would absolutely LOVE to monopolize this, I just hope that never happens.

Staying independent benefits all of us, including Artemis Networks as they can make this solution available to every single operator globally and push for innovation.

 

But with everything that Artemis seems to be bringing to the table, all other mobile tech companies should really think twice before they ever use the term "innovation" again...

Link to comment
Share on other sites

I'm sure duopoly would absolutely LOVE to monopolize this, I just hope that never happens.

Staying independent benefits all of us, including Artemis Networks as they can make this solution available to every single operator globally and push for innovation.

 

But with everything that Artemis seems to be bringing to the table, all other mobile tech companies should really think twice before they ever use the term "innovation" again...

Woah. I'll leave the Pcell concept aside and say that MSS-TDMA and subsequent options for closed loop MIMO have just begun to demonstrate wireless throughput 1gbps + 

Through an xr 365, xr9 Mikrotik box and UBNT radios I have aggregate streamed 260mbps+  via 2x2 (aggregate 4x4) 2 spatial, so essentially standard 3x3 MiMO with iPV6 L2L. 

Perhaps I should end ranting. 

If this technology is going to make it, it needs to be deployed effectively;quickly.  That means working with current and NEW operators to deploy on available spectrum.  Who has money, and spectrum is critical. Engineering and crews are a whole other issue in which I have a direct connection to several keystone people in the ISP and Wireless industry and think that a huge partnership with backhaul provision and spectrum sharing would be a huge success for the adaptation of pCell.  I hope that Sprint is able to realize the value.

-William

  • Like 1
Link to comment
Share on other sites

I must say I like that he is taking the time and addressing the questions that are being brought up here. With everything that must be going on having just announced this and the excitement surrounding it, the fact that he would take the time to look through this little thread we have going and personally begin to address the questions being brought up is something I would have never expected.

 

I also respect that you talk so openly so early about not working with groups who would just sit on something that could shake up the market like this. As Sprint fans here I'm sure many of us have wondered at some point what could be done to unseat VW and AT&T at the top of their mountain or at least level the playing field for the smaller operators and I could actually see a way here to at least significantly level the playing field. I imagine with as significant of a change as this technology would be, if any of the major operators adopt it all of them would have to if they wanted to stay competitive, and I would hope it would be easier for the smaller operators to reach a point of network parity with their competitors so that the selling point would become less about the network quality and more customer focused but i'm afraid the duopoly will do everything they can to complicate the process.

 

 

Sprint has plenty of spectrum, so even without pCell they will not be starved. With all of the spectrum they own and all the capacity inherent in this technology, is it time for Sprint to become an OTT video provider. I mean, what are you going to do with all this bandwitdh except stream video? Wireless cable, anyone? Would it make sense to team up with Dish since Dish already has the content deals?

Edited by bigsnake49
Link to comment
Share on other sites

I'd like to know how differently would pCell address capacity issues in stadiums vs regular DAS that wireless operators already have installed. Obviously we can't expect pCell delivering maximum throughput of any given LTE channel to 20,000 people unless you have 20,000 pCells, but how will it improve median downlink rates? I understand that the coverage will be substantially improved, but I'm interested in throughput within cell center and high amount of UE.

 

Thanks!

Link to comment
Share on other sites

This is a very interesting video interview from Bloomberg TV:

 

http://www.bloomberg.com/video/sculley-pcell-changes-rules-for-mobile-wireless-t1i3VbxKR8OhFkGOOThRsw.html

 

Btw, I can totally see Charlie getting all over this technology.

 

 

 

Exactly, charlie has been very quiet as of late, and this could be his way of making a big splash.

 

Possibly something similar to what sprint planned with clear? And possibly a little investment from Sprint?

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...