I’m still working-away on my IBM book and it is still a week from being finished (the well-known second 90 percent syndrome). The book, if I am allowed to sell it on Amazon, will cost a whopping $3.99 and will be worth the money. But I’m still a columnist of sorts so here are my thoughts on pCell, an impressive new technology for increasing performance of LTE mobile data networks. It was invented by WebTV founder Steve Perlman, introduced two weeks ago in New York (very impressive video here, but fast-forward to 5:30) and was the talk of the Mobile World Congress in Barcelona the following week. pCell is amazing. It is also probably a security nightmare waiting to happen.
This is not me being a bad-ass or somehow wanting pCell to fail. I think it is great and I want it to wildly succeed, but there are a couple things about pCell that have been going over the heads of most reporters, security being one of them. I’ve read all the stories about pCell and the word security doesn’t appear in any of them, none.
Typically with new technologies like this the inventors haven’t thought much about security or they rely on a small installed base to keep the product or service under the radar of the bad guys. But pCell, for all it’s high tech loveliness, is a Software Defined Network proudly running in a data center on plain old Linux servers. So if there is a security vulnerability found in Linux (Red Hat found one just this week that I’ll be writing about later tonight) then pCell is vulnerable. Not might be vulnerable, IS vulnerable.
This vulnerability will probably show itself first as a disruption of service. Since the personal (that’s the p in pCell) cells are defined and controlled from the servers through transceivers scattered around town, if you mess with the server you mess with the network and can potentially take it down. So pCell, as Linux software, is vulnerable to Linux problems.
It just worries me that nobody is mentioning this because I expect we’ll see a ton of pCell announcements shortly with major mobile carriers covering densely-populated areas where this service will shine… until it is hacked.
The other point that is glossed over in the pCell discussion is total bandwidth: if you are going to have every pCell user watching his or her own HD (not to mention 4K) video signal, the network is going to have to be provisioned with enough bandwidth to cover those total users. You see pCell is about the efficient utilization of radio spectrum, not network bandwidth. If it takes two megabits-per-second (Netflix recommends five mbps by the way so my numbers here are probably too low) to practically serve what the video network vendor will claim is a 1080p HD video stream using H.264 encoding, then you’ll need two megabits for every mobile user watching video on the pCell network. So when they talk in the video about a single pCell radio helping to create a clear channel for up to 1000 users, then each of those users had better be provisioned with at least two megabits of backbone or cached server bandwidth.
Yes, an LTE connection is faster than two megabits, but unless there are two million actual bits on the other end of that connection (called provisioning) having an empty pipe is meaningless.
If there are two million people in Manhattan, 10 percent of those are on the phone and 10 percent of those are watching video on their phones, then the aggregate bandwidth for all those users had better be 20,000*2,000,000=40 gigabits-per-second (100 gigs according to Netflix). This isn’t much if video is cached and delivered from the data center, but if it is coming across the Internet backbone it’s a lot.
There are probably limitations, too, on how much bandwidth each of those pWave radios (the ones on buildings, not the ones in your phone) can handle. If as suggested by Perlman’s claim of single-virtual-channel operation, then each of those radios probably maxes out around 75 mbps based on the 10 MHz LTE standard.
So how many pWaves would it really take to serve Manhattan? At least 40gigs/75megs=533 radios. Looking at it from a different angle (Perlman’s claim of 1000 users per radio) then the number under the scenario I described is lower (200,000/1,000=200). That isn’t a lot but it is one radio for every 27 acres (one radio for every 700 acres in the latter example — did you know Manhattan has 14,526 acres?). This explains why they had several radios in the demo room on the pCell video in order to show dual 4K video streams running in 10 MHz.
Where I am going with this is that nobody really knows how many cell towers there are in Manhattan but only nine are registered with the FCC and the largest number I can count on a cell tower monitoring site is 38. Whichever of my pCell numbers is right is going to require some significant changes in the way mobile carriers do business in order to best utilize this new tech in the dense areas where it is needed most.
Based on the fact that Comcast, for example, is building opt-out public WiFi access points into its cable modems, I suspect the mobile carriers are going to try piggybacking on customer bandwidth, too. Because the more the merrier when it comes to pCell. If Verizon, say, offered some of its FiOS customers free Internet in exchange for hosting a pCell, their mobile network could be locally transformed overnight. Who’s going to turn down a deal like that?
It’s worse than that, to get the full bandwidth, you need a pWave for every simultaneous user.
In their patents, they do describe piggybacking on whatever is “serendipitous” for a given location, including DSL/cable. Which they’ll have to, given how many of these things they will need to roll out in densely populated areas. There is some cool things like self adapting network where distant towers can increase power to help ease congestion is other areas but that only gets you so far with transmit limitations.
Between the number of pWaves and data centers for each area, seems like there’s a very real risk of out of control CAPEX. Someone like Dish who already pays $800/install for new customers and just recently got spectrum and IP streaming rights might be a good partner in that regard.
That said, I really do hope they can make it work.
Any system could be vulnerable to a zero day exploit of course.
.
The key will be how quickly can they patch when things go wrong.
Looking forward to the book. As an IBM employee of more than 25 years, I can tell you morale has never been lower. I currently work in IGS IGA (IBM Global Account). It’s at the point where I am hoping for an RA even though I am not of retirement age and will have to find another job. IBM has become a toxic place in which to work, and there is no light at the end of the tunnel. IBM is deliberately making the workplace untenable because they are trying to encourage attrition. Their latest stick of encouragement has been to hand out more negative appraisals. Two negative appraisals in a row will mean we no longer qualify for a decent severance package.
I’m a reformed, er, former IBMer, too. After I got out, I found the sun shining brighter, the stress stressing lower, and the air smelling sweeter.
In all seriousness, though, there are much, much better places to work. Why people willingly continue to work there is simply beyond me. It can’t be for the money, prestige or job satisfaction any longer. I can only imagine it is fear of the unknown.
Some people live misery.
I have never heard so much upper middle class employment related whining as I have from long-time IBMers. Yes, your company is treating you rather poorly, but your tone is petulant, entitled, and frankly disempowered.
Don’t like it? You own your career… get a better new job. Think the US 401K thing was a raw deal? Well, it was but that’s done so move on. Think the company has terrible long-term strategy? Go compete against them.
I largely agree with the criticism of IBM by the way, but the IBMers who b*tch and moan come off as pathetic and it weakens their position.
I have to agree with you. I have former colleagues who will call me and start the conversation with, “You’ll never guess what they took away today…” Really? And you still work there? Frankly, I think it’s a version of the Stockholm Syndrome.
I agree. There are IT jobs all over the country. If you don’t like IBM and it’s the only player in town, move to another town with better employers and job prospects. IT skills are highly portable and if yours are not, it’s not IBMs problem, it’s your problem.
Long-term employees don’t “own their own career”. Their dedication guarantees that they have become cogs in their employer’s machine. Take away the machine and the cog is useless.
I find those here who make snide remarks about the IBM situation and are of the “so move to another town” mindset to be pathetic, short-sighted, and narrow minded. Some folks MIGHT have families, spouses with careers, etc, and can’t just “up and move to antother town” for any number of reasons. What pompous self-righteous smug fools. Just because they might have that freedom or lack of other commitments they think everyone else in the country is the same. Wake up.
Things are pretty nice on my team in SWG.
I’m well paid. Management values my contributions. I have a flexible work schedule. I get to work from home or the office–my choice.
The solution for unhappy IBMers is simple–either find a better fit within IBM or leave altogether. I don’t want to come off as jerk, but I hate seeing people who are unhappy or feel they’re being mistreated.
How many people report to you, directly or indirectly?
Your IBM writedowns have been a drag but not as much a drag as some of the smart young PMs we have to work with who can sit through 2 patient explanations of anything before it becomes our problem.
There is however a smart new solution AND it delivers on share price too: get rid of most of the “no/bandwidth partners” & their bitches ( APs) & hey guess what you get your 78% cutbacks & you get a really happy workforce ready to deliver mainframe 3,near real-time messaging,overnight batch or whatever from home. That’s right from home because the best of us are happiest spending tea breaks nurturing our plants & interacting from time to time with ex colleagues from their combo sys admin/plantsmen or landscape gardening businesses, yachts or studios,managing entire mainframes or multi-tenanted clouds with ease.We still rule ok!
Have you set up “early bird” ordering on Amazon? Will there be a print copy?
.
If you want IBM’s exec’s to read the book you’ll need a print edition. I don’t think they can handle an electronic book. Make sure you use simple words. Pictures will be helpful too.
.
If we set up a collection could you send a print copy of your book to every US congressperson and senator?
.
I read the stories on pCell. My compliments to Steve Perlman! I think it is a tremendous idea and I hope it is wildly successful. I too am worried about security. About once a week we’re hearing about another retailer that has been hacked. We’re on our 7th set of bank cards in a year. Security is now a really, really big problem. The problem is everywhere. My cell phone was hacked twice at Wrigley Field in Chicago. (No I am not a Cubs fan — I was cheering for the other team.)
As an individual author I can’t do pre-sales on Amazon. Only publishing companies can do that. So if my print publisher decides to grab this book I’m sure they’ll start taking orders the next day, though ironically the book itself will take longer to reach the market because, well, publishers are slow. I’ll get paid a lot slower, too. Damn!
As an individual author whats to stop pre-sales on cringley.com?
Keep up the good work dude.
BTW, it would have been nice to measure the ERP of those PCell units. It is after all directing the cellular signal into focused beams onto the target handsets and those handsets are where? Next to people’s heads.
Microwaves were discovered when a guy working on a magnetron noticed his chocolate candy bar melting inside his shirt. Wonder that prolonged exposure to PCells be like?
Just wondering. Adam kb2jpd
In the video they say the radios are running at one milliwatt, but I think that referred to the radios in the phones which would be mainly receiving, not the pWave radios. So you could be right about the ERP. Remember guys in the Air Force up on the DEW Line in Greenland used to stand in the radar beams to get warm!
Microwaves are non-ionizing radiation. There will always be people scared of them, like with vaccines, powerlines, etc. but there is really no evidence that they are harmful. Sure, there are anecdotes, like everything else, but I’m not worried. Your Banquet frozen dinner doesn’t become radioactive, and any cell disruption occurring is because of the heat generated by vibrating the water molecules, not atoms being split apart.
My Great Uncle was part of the design and deployment of the DEW Line (as well as the development of “trunk lines” in Germany’s reconstruction (to extend the use of fewer primary lines for long distance runs) and the translation headset concept created for the Nuremberg Trials that is still used in places like the U.N.
All of that to get back to the microwave stories (and a host of others). Great fun to sit at the knee of one of The Great Generation’s techno nerds. To hear those stories growing up and then see them crop up in the history books gives them a special, personal touch.
Probably a little like Sean Lennon when someone starts telling Beatles’ stories.
Oh yeah… the point. It’s great to be part of a wave of tech revolution. To “inherit” the family nerdism as it were. His was an analog wave, ours was the digital wave through mainframes, minis, PCs, LANs, WANs, and now mobile.
I look forward to whomever is next in my family to take on the nerd mantle.
In the pcell demo, he moves the machines slowly, they are very close to the base station and we do not see how many base stations there are or how they are spaced around the room.
Does it scale with things like pcell density, the number of pcells per base station, the distance of p-cells from base stations, etc.? Will it work if the clients are moving rapidly — perhaps walking? Driving?
What are the tradeoffs and “hidden gotchas?” I’d like to show pcell to my students, but neither they nor I are qualified to view it skeptically.
It scales as long as number of transmitters > number of receivers if we’re talking full bandwidth. He believes it can scale 1000x+. Doppler is accounted for and will use pWaves experiencing lowest doppler. Whether that works as advertised at scale can only be bourne out in real world trials which we have to wait to see.
Larry-
I didn’t move the phones slowly because I was concerned about Doppler. They are just small and hard to see from the audience.
And, MD is correct. If you are on a train platform as an express train zooms through while sounding its horn, you’ll notice the pitch rising and falling because the train is going directly toward, and then away, from you. You experience high Doppler. But, if you are 1/4 mile away, the angular displacement of that same motion is only a few degrees and the horn’s pitch will not change noticeably at all. You experience low Doppler.
The same is true for wireless Doppler. If the user motion is directly towards or away from an access point, there will be very high Doppler. If user motion is orthogonal to an access point, there is low Doppler. Well, with a cellular system, you only have one access point (the cell tower), so if you have bad Doppler, you are stuck with it. For pCell, you are surrounded by pWaves, and there are always some pWaves experiencing low Doppler.
“I didn’t move the phones slowly because I was concerned about Doppler. ”
That sentence can be parsed two ways yielding opposite meanings:
1) I was concerned about Doppler, therefore, I moved the phones slowly.
2) The reason I moved the phones slowly was not because I was concerned about Doppler.
PCell works fine at high speed, but that is better shown by looking at raw received data bits:
This video: https://www.youtube.com/watch?v=Lv-vkBNzZwE shows me moving multiple tablets while showing the received constellations (the data bits). The constellation is stable as I move the tablets and pCell adaptation (the technology that moves and reshapes the pCells to track the LTE device antennas) is on. When we turn off adaptation the pCell bubbles freeze in space and you can see that even a few millimeters of motion will move the antenna towards the edge of the pCell and the constellation begins to fall apart, and then completely falls apart in less than a cm when the antenna exits the pCell. When we turn adaptation back on, the pCell is locked to the antenna and the constellation doesn’t even begin to fall apart, which means the adaptation is keeping the antenna near the center of the pCell.
Hope this helps.
My friends generally think that I’m a fast walker, but I’m not at all worried about Doppler effect when carrying electronics. Train whistles are one thing, because there you’re thinking in terms of the speed of sound. Wifi is speed of light, and that’s another thing entirely.
If he’s moving slowly, it’s because of a normal, generally rational in the real world, fear of screwing things up.
We have a whitepaper coming (sorry for the delay…utterly swamped since the announce) that will provide a lot more details about pCell, including data sizing and so on. But, you are correct that if you start streaming data at that scale, you either need the data cached locally or you need a large enough pipe coming in.
First, this is a “high-class problem” that the mobile carriers would LOVE to have. They’re in the business of selling mobile data, people badly want that product, but cellular has hit a “physics limit”, a phrase from Verizon’s CEO that is accurate.
Second, if you become such a major ISP, then the major video services, like Netflix will happily host their content in your data centers. It’s cheaper for them than paying a CDN.
Lastly, in thinking about the data centers having access to content, bear in mind that the current incumbent carriers are not the only ones interested in (or capable of) deploying pCell.
Regarding the number of pWaves, you are correct that the right way to think about them is each one adds aggregate capacity to the coverage area. If one pWave can deliver 75 Mbps, then 10 will provide an aggregate of 750 Mbps that can be shared among any number of users. So, in this example, with 750 Mbps of aggregate capacity, you could have 750/5 = 150 concurrent 5 Mbps users from 10 75 Mbps pWaves. If the average data rate is 1 Mbps, then you’d serve 750 concurrent users with10 75 Mbps pWaves, and so on.
In terms of pWave deployment, cellular and Wi-Fi people are shocked to find out how easily pWaves are deployed. pWaves don’t need a high vantage point like cellular and don’t have to be visible or even outdoors. Typically what we’ve been doing is putting about 8 of them on a rooftop, maybe 1 or 2 on a pole, etc. With cellular and Wi-Fi, you are so used to planning and aiming antennas, it’s hard to get used to the fact you can put pWaves anywhere in the coverage area. Remember, the user devices are intermingled with the pWaves. No one’s aimed at anyone. It’s just random. So, what drives pWave placement is convenience and low cost of deployment.
For example, in San Francisco, we have access to 350 rooftops all chosen because they happen to have line-of-sight radio links of another building already in the mesh and the landlord/building assn is reasonable. There are only 6 fiber drops in the entire mesh back to the data center. The rest is free fronthaul in line-of-sight unlicensed spectrum. The reason cellular can’t use those rooftops is their locations are completely random. Some buildings are close together, some are far apart, some have obstacles that would be bad for cellular, etc. Cellular needs to strictly follow a plan. pCell doesn’t care where the pWaves are placed.
If we put 8 pWaves per each of the 350 roofs, that’s 2,800 pWaves. In your example of 75 Mbps/pWave. That’s an aggregate of 210 Gbps. And, as discussed, with the high-class problem of being such a hugely successful ISP, the streaming movie folks would love to eliminate CDN costs and host their movies in the same data center creating the pWave waveforms. So, in summary Manhattan would not be problem for coverage.
Better yet, it’s easy to start small and expand coverage. In the cellular world, it is very hard to add new ones to fill in coverage because they interfere (in fact, even indoor femtocells spill out and interfere with the outdoor coverage). With pCells, just toss a few pWaves up, and as you get more subscribers, toss more pWaves into the area wherever it’s convenient and cheap. Serendipitous deployment.
One comment on Linux security. Bear in mind that the way we are using Linux is just as an OS for pCell SDR code that crunches insanely complex math, not for serving web pages or running any 3rd-party apps. So, there is no direct user (or Internet) access to the Linux machines. If security became a concern, we could easily move the code to another OS (e.g. a commercial real-time OS). We’re not using any Linux-specific features.In fact, we strip it down to its bare bones and custom-tune the drivers to get tight real-time performance on all cores and all threads. We’re not just generating and demodulating full-bandwidth LTE waveform in real-time, we’re doing multiple LTE waveforms per machine, then we pCell encode them, and then synthesize the multiple encoded waveforms and output them to pWaves. The actual interfaces to the Internet are conventional, data center-grade hardened gateways.Nothing from the outside touches the pCell servers.
Anyway, the whitepaper has a lot of good info (and if there are more questions we’ll write another one).
– Steve Perlman, Artemis CEO
Steve, thanks for those details, that cleared up a few things for me. Absolutely crazy potential here, though like most I’ll be skeptical it can be done at scale and cost effectively until it’s got at least one large public deployment under it’s belt.
Good luck, very much looking forward to the whitepaper and the results of the initial deployments.
Steve, wow, thanks. That was a great, open, honest, and candid discussion. I wish you great success.
.
Regarding security — those of us in this industry are really scared right now. We know firms in Eastern Europe are now selling sophisticated tools and services to break into networks and computers to steal data. Someone managed to install a memory scraper on every one of Target’s cash registers. Those registers ran a stripped down OS, no user or Internet access, no third-party apps, etc. Yet someone managed to steal 10″s of millions of credit card accounts from them.
.
We are sure your team is writing the best and most secure applications possible. We’re scared. The potential of these security problems is alarming. Every few days we learn of another retailer who has been hit. The list of firms not at risk could be very short. Please continue to be sensitive to this problem.
How do you know that the cash registers were hacked? This is not the story that I have heard. Rather, the network was hacked thru use of a stolen password.
Good point. The key to security is no network access. But I can imagine a pCell-knowledgeable hacker social engineering his way into the computer room for the sole purpose of adding a network card. Having done that, he can remotely hack Linux at his leisure, adding whatever software is necessary to watch free movies. Coming to think of it, there is nothing to worry about since at that point he’ll be arrested for dmca violations. 🙂
You can be sure I will buy your book. I have Accidental Empires and loved it, so I have no doubt that your new one will be good.
One way to aid in the security aspect would be to build it on OpenBSD. Security is the first thought of that community. Linux is just not that secure.
Wow! To heck with mobile, I’m dreaming that Comcast would have a competitor in my neighborhood. Dish or Direct TV feeds a local data center and distributes through the pCell network with much lower infrastructure cost than stringing coax or fiber. On the downside, this will probably ensure that we never get upgraded to FIOS. My worry is that Comcast decides to spend $ gazillion to buy Artemis before the FTC can understand how disruptive pCell can become.
Second, I think that pCell system can provide very precise location service, maybe not so good for vertical measurement but very good near the plane of the transmitters. Those little carts in the Amazon distribution center would always know where they were without any floor markings, and the drone would put things on your doorstep with better accuracy than the newspaper boy ever had.
If Perlman is involved, it will fail.
This article has a very telling quote from Perlman himself: “Perlman has been unable to tempt venture capitalists with the technology. “They invariably bring in experts who say it doesn’t really work,” he says. “I am showing them a demo, but they remain convinced that it’s something else.” It sure sounds like he said he can’t convince the experts: https://www.businessweek.com/articles/2014-02-18/steve-perlmans-amazing-wireless-machine-is-finally-here
Of course the theory shouldn’t be dismissed out of hand. I discovered a long discussion on Shashdot from a few weeks ago about how DIDO is related to MIMO and phased arrays in general. Found this great comment: http://mobile.slashdot.org/story/14/02/19/1747246/new-pcell-technology-could-bring-next-generation-speeds-to-4g-networks which concludes with “Yeah, right. It will be out by Xmas, I’m sure.” (Look for “I’m Feeling Deja Vu”)
Steve, Bit late now, just wondering why OpenBSD was not used ? Available developer skills ?
I don’t understand the panic over linux bugs. Bob is of course talking about the TLS and SSL bugs. Within 4 days of their discovery, my Debian-based release had new libtls and libssl updates, and one command later, my system was updated. No other OS can match that. Bugs happen, get used to it, but when it comes to fixes, you can’t beat linux.
You bring up a good point. Bob said “This vulnerability will probably show itself first as a disruption of service.” So far I don’t recall much talk in the press about cell phones experiencing a “disruption of service” due to a vulnerability in any of the popular OSs used by consumers and websites, whether Windows, OSX, or Linux. This makes me wonder how current cell phone systems operate that they don’t currently have these types of problems. Wouldn’t it be more important that they be protected from running executable programs coming over a network card than which OS they used?
test me
[…]although sites we backlink to beneath are considerably not related to ours, we feel they may be actually worth a go via, so have a look[…]
motocourier
[…]Here are a few of the websites we suggest for our visitors[…]
product review
[…]The information and facts talked about inside the report are a number of the ideal obtainable […]
visit site
[…]we came across a cool web site which you could possibly take pleasure in. Take a look in the event you want[…]
follow me
[…]here are some hyperlinks to web pages that we link to for the reason that we consider they are worth visiting[…]
www
[…]usually posts some extremely intriguing stuff like this. If you?re new to this site[…]
Green Smoke ecigs
[…]Wonderful story, reckoned we could combine some unrelated data, nevertheless really really worth taking a look, whoa did a single study about Mid East has got more problerms too […]
Diet plan for Weight Loss
[…]that could be the finish of this article. Here you will find some websites that we assume you will value, just click the links over[…]
enlevement epave gratuit 78
[…]we came across a cool web page that you could get pleasure from. Take a search in the event you want[…]
Rock
[…]we like to honor a lot of other online web sites on the internet, even if they aren?t linked to us, by linking to them. Below are some webpages worth checking out[…]
poncelet or poncelet brian
[…]we like to honor many other internet web pages on the web, even when they aren?t linked to us, by linking to them. Underneath are some webpages worth checking out[…]
weddings band perth
[…]please pay a visit to the web sites we adhere to, including this one particular, as it represents our picks in the web[…]