Talk:Very minimum-shift keying

From Wikipedia, the free encyclopedia

Remember that article talk pages are provided to coordinate the article's improvement only, and are not for engaging in discussion of off-topic matters not related to the main article. User talk pages are more appropriate for non-article-related discussion topics. Please do not use this page as a discussion forum for off-topic matters. See talk page guidelines.

I started getting interested in the VMSK topic after seeing this article, and by following the links I got to see both H.R. Walker's views as well as the views of the many debunking publications.

It's actually strange how Mr. Walker seems to have published a lot of papers (most with cooperation from academics), got regular publications and apparently continues working with VMSK, plus there's this mysterious "Prof. Howe" for whom or his works I couldn't find any references other than Mr. Walker's own publications. Can anyone help expand/clarify and/or provide sources for Mr. Walker's published papers (not those on the VMSK site) as well as some data on this "Prof. Howe" ? EpiVictor 00:36, 28 May 2006 (UTC)

  • Without trying to insult you, what part of "Such claims are to communications what Perpetual motion is to energy production, as they run afoul of firmly established mathematical principles such as the Shannon-Hartley theorem and the Nyquist-Shannon sampling theorem." didn't you understand? This is a Fool's Errand, plain and simple. --DV8 2XL 00:49, 28 May 2006 (UTC)
I did understand them, don't worry (I mentioned "debunking", not "skeptical" publications), plus due to my academic studies I would not allow myself to be so easily fooled by such a scam.
The only thing that, let's say, drew my attention to VMSK, was the sheer amount of papers on VMSK, although most of them come from Mr. Walker himself (assuming he really exists and is an engineer) and from the fact that there are even other academics (some of them I'm familiar with, may I add) who have, at least at a certain period, supported his positions or at least worked with him. Plus I'd like to know if that "Prof. Howe" really existed or if he is entirely fictional (the people mentioned on Mr. Walkers oldest paper, especially in the 1997-1999 period are, in good or bad, real). EpiVictor 13:46, 28 May 2006 (UTC)

Sorry for misunderstanding.

I cannot answer your questions in detail, however you should note that there are legitimate ultra narrowband modulation schemes that stay within the theoretical limits in applications where a very slow data transfer rate is tolerable. I would say off hand that Walker is citing some work done in that area. But this is just a guess. --DV8 2XL 14:12, 28 May 2006 (UTC)

Contents

[edit] Disputed science infobox?

Someone removed the "disputed science" infobox summarizing the contradiction with the Shannon coding theorem. Why? Is there some kind of edit war going on here between proponents of crank claims to have broken this theorem and information theorists? ---CH 16:58, 29 May 2006 (UTC)

CH are you sure it was there? I can't seem to find it in the edit history. There were several redirects done on this topic, perhaps it was lost by accident during one of them. Anyway I agree it should have one. Do you want to do the honors? --DV8 2XL 17:11, 29 May 2006 (UTC)

Hmm... sorry, it must have been another page. Please, go ahead and add the infobox to this article. ---CH 19:47, 29 May 2006 (UTC)

[edit] Edits by H.R. Walker himself?

Apparently the latest edits to the article have been made by someone using the Pegasusdat@aol.com username, which suspiciously coincides by Mr. H.R. Walker's own email address as it appears on his VMSK page, or at least it's the one of "Pegasus Data systems" which Mr Walker owns/represents (?).

Trying not to be critical of Mr. Walker's VMSK analysis or the validity of his claims, there's still the wikipedia policy of not allowing a person/corporation/group from editing their own articles, for neutrality policy reasons.

Perhaps "his" analysis and references can be included for completeness, but they can't be the article's only contents, especially since his modulation scheme is still surrounded by controversy . EpiVictor 22:25, 31 August 2006 (UTC)

Here's what appeared in the body of the article:


“Very Minimum Shift Keying”, or more correctly, “Very Minimum Frequency Shift Keying”, is one of several “Ultra Narrow Band Modulation” methods which are based on an elementary mathematical analysis of Armstrong’s “Frequency Modulation” method published by Professor Howe in "Wireless Engineer", Nov. 1939. pp 547. ( England ).
This analysis showed that if a rectangular waveform was used instead of a sine wave, as used by Armstrong, the phase modulation could be preserved in a single frequency without any frequency shift, or useful sidebands. Since only a single frequency bearing phase modulation is transmitted, this single frequency can be passed though a very narrow bandpass filter and the phase modulation can be recovered. In no way does this concept violate Shannon’s limit or Nyquist’s sampling theorem. There is one essential caveat.
The bandpass filter used must have near zero group delay, which is not a feature of most commonly used filters. There are perhaps a dozen or so filters which will exhibit this characteristic at a single frequency, while rejecting all other frequencies, by using the vector addition principle. Signal to noise ratios equal to or better than those obtained using “Bipolar Phase Shift Keying” are being obtained.

The problem is that the validity of this analysis is disputed, and no references to this "Professor Howe" can be found. On the other hand, it's hard to find "official" rebuttals of Mr. Walker's analysis and claims, and most such rebuttals are done at personal level at best, from what can be found on the internet, while Mr. Walker can actually indicate some acadmics in support or at least endorsing his scheme. So...could his analysis be included after a certain cleaning up and several factual accuracy warnings, and reference digging? EpiVictor 17:52, 1 September 2006 (UTC)

[edit] Reply in support of VMSK

These two paragraphs were added by User:Hyperspace1, evidently supportive of VMSK:

First I must tell you that the sceptics are just that. You sound like the gas men that told Edison the light bulb would never work and the stable men who told Ford his machine would only work if the devil blessed it.
Yes Dr. Howe did exist. An easy bit of research finds George William Osborne Howe, DSc, LLD. Dr Howe became a professor in 1921 and in 1938 dean of Engineering at the University of Glascow. This information can be verified at www.glascow.ac.uk

At least the anon contributor provided some extra info on that Howe guy...hope anyone can put it to good use. EpiVictor 22:26, 1 September 2006 (UTC)

[edit] Remember, this is science/math, not politics or philosophy

That means there is only one correct side to this story, and there is no need to be "fair and balanced" to Walker's mathematical claims when they are provably wrong. Any competent independent analysis of his claims will stand on its own, whether or not a "Prof Howe" ever existed.

By way of disclaimer (not that it really matters, given what I just said), I am probably Hal Walker's main critic.

Karn 13:57, 9 December 2006 (UTC)

Mr. Karn has made it his "duty" to bash and dash VMSK every chance he gets. He may turn out to be right, and I have no doubt he is convinced of his righteousness. But now you have xG Technologies, which he said, "at first looked like VMSK" - that he clearly does not understand - raising $100 million from merchant bankers and building mobile VoIP networks. Their market cap on the London Exchange is now probably about $2 BILLION, and believe me, this is a drop in the bucket for them if they really have the "goods". See xMax on Wikipedia. Now, is it POSSIBLE that Mr. Karn's knowledge in the field does not permit him to see breakthroughs like these technologies? Could there be a gap - could the leap forward be so grand as to blind him?

Eagle Comtronics, which "verified" VMSK is the #1 company of its type in the world. Here's the blurb from their site:

"In 1975, Eagle Comtronics introduced the CATV industry's first cylindrical Pay TV trap. From our humble beginnings in a basement workshop we have expanded to approximately 150,000 square feet of manufacturing and administrative space. Today, Eagle provides traps, scrambling systems and custom filters for telecommunications systems in more than 60 countries. We also design custom filters in a broad range of frequency responses, sizes and specifications."

So finally, things get interesting.....Phil saying its Hokum, and Eagle Comtonics "verifying" a "promising technology". Is Phil our Quixote, or is Hal? Wish I understood all of this enough to approach it from the scientific and engineering point of view.....But I have certainly followed it all for years and am invested on the side of VMSK. I realize I could lose, but the risk/reward is astronomical in the favor of reward......I look forward to Phil's "it's impossible" follow-up to this. But I also look forward to the first cable network licensing VMSK instead of spending billions laying fiber....

One last tib bit - Phil's employer, Qualcomm, is rumored to be in discussions with.........xG Technologies. Wouldn't that be priceless:)?

Being an electrical engineer myself, I'm sure Mr.Karn is right "bashing" VMSK. It's not impossible to send data using VMSK (something Mr Karn has always precised, and is stated in the article), it's impossible to do it as fast and efficiently as "promised", and probably it's already hard reliably sending data within the theoretical limits, seeing how much transmitter energy VMSK wastes. So your investment may result in some kind of practical (even marketable) working equipment, but you should indeed THANK those lying fiber, for the waste of bandwidth using VMSK will be just appalling. EpiVictor 22:58, 13 April 2007 (UTC)
xG's claims are as equally nonsensical as those of VMSK. Any technology based on claims that defy Shannon's limit is necessarily nonsensical; this is no different to any technology that claims to be a perpetual motion machine. Raising VC is of no relevance; this does not affect the underlying physics or mathematics. Any competent information theoreticist should be able to tell you that the claims of xG or VMSK are based on fantasy; any paper or patent presented from the xG/VMSK camp can easily be picked apart as nonsense. If these companies manage to produce products and services that work, then good for them, but I have no doubt that they will not be based on the principles promised in their patent documents, etc.
As for Eagle Comtronics, a brief glance at their website should be enough to assure you that they are an irrelevance; clearly not qualified to pass any sort of judgement on the issues at hand. Oli Filth 02:12, 14 April 2007 (UTC)

Here is some later news regarding xG Technologies: Major investment funds are buying xG stock - Stormur Holdings AB has 18.2% and ACH Securities SA of Geneva 17.7%. Others wanted in but there were not enough shares to make purchases at reasonable prices. The market cap is over $2 billion. The company initially plans to open in 90 markets, and eventually go global. "The first 10,000 handsets were designed in Sweden at a cost of $285 each, but later headsets will be made in China for $150." Guys, isn't it time to start rethinking your science? It's past time to re-evaluate certain pre-conceived notions about these technologies still held by a few....

Why would any competent comms engineer do that? Here are a few reasons why not (referring to VMSK here, not xMax):
  • There have been no public (or otherwise) demonstrations of VMSK working successfully in a realistic setting.
  • Walker's papers, etc. are child-like, badly-written, and teeming with fundamental errors that any competent first-year undergrad could spot.
  • Walker's claims directly violate Shannon's limit.
  • VMSK can be analysed perfectly as BPSK + a square wave clock component. The maths easily proves it doesn't work as claimed.
  • Walker dismisses all such (dis)proofs with claims such as "VMSK cannot be analyzed at baseband unless you use the full Nyquist BW"; not only does he give no rationale for these claims, but in fact they implicitly prove that VMSK does not do what he thinks it does.
Phil Karn has described this far better and in far more detail than I can do justice here. Oli Filth 17:44, 20 April 2007 (UTC)

CTLG is my bet on VMSK - it's still under $1.00 and only has a market cap of about $60 million. They are focusing on cable, encryption, satellite and radio. Very interesting to say the least.

By the way, as I understand it, Hal Walker has never claimed to violate Shannon or Nyquist - I believe he has stated that some engineers do not understand the technology well enough to know why those laws are not violated.

Of course, that's what "inventors" of perpetual motion machines also DON'T claim: none of them boasts of having "broken" or "hacked" the laws of thermodynamics, it's always the others that misinterpet them. Sounds lame, right? Probably because it is. EpiVictor 17:20, 20 April 2007 (UTC)
By making claims as to the bandwidth/information-rate/SNR performance of VMSK, Walker has implicitly claimed to violate Shannon's law. Whilst in his analysis he has shirked this issue simply by re-defining "bandwidth", this does not affect the underlying maths or physics implied by Shannon. As Karn puts it:
"This is a bit like proving that 2 + 2 = 5 if we simply redefine the numbers 2 and 5." [1]
Please be under no illusions, Walker does not understand the implications of Shannon's limit (or is being intentionally deceitful). Oli Filth 17:44, 20 April 2007 (UTC)
I've exchanged countless emails with Harold Walker since 2000. I am personally convinced that he honestly believes in this stuff. He's completely impervious to logic and rational argument, but he does have just enough knowledge to be dangerous. So I do not think he is being intentionally deceitful. Karn 05:35, 29 April 2007 (UTC)
He would probably bring up the old story about prof. Howe (I would really like to SEE that publication he's citing) and how if it doesn't work it's all just because of the non-existant zero-group delay filters...another way of saying that he's got his stuff right, it's us bumpkins that don't get it...lame. EpiVictor 23:09, 20 April 2007 (UTC)
Which raises another question: most fundamental information theory theorems were formulated as early as the 1920s but many weren't proved until the 1940s, 50s or 60s, so it's entirely possible that Mr. Walker is just following an "old school" current of thought (if there ever was a group of scientists openly disputing the various bandwidth/information theorems in the decades following the original formulations). This would place VMSK in the same category as polywater or N-rays, e.g. pathological science, as well. EpiVictor 23:21, 20 April 2007 (UTC)
I think I've figured out the basis of Walker's obsession with "zero group delay" filters. UNB methods vary, but he generally puts his small, narrow (and thus wideband) data pulses at the zero crossings of his very strong carrier signals. His "magic" filters are simple crystal resonators with broad, flat skirts typically 30 or 40 dB down from the resonant peak that he tunes to the carrier frequency. This is actually the correct matched filter for his modulation, but I doubt he knows anything about the matched filter theorem. He seems to have found it entirely by trial and error.
His filter merely attenuates the "grass" (as he calls his wideband data spectrum) without distorting it, so subsequent amplification and limiting can easily restore it for demodulation. However, the receiver filter must maintain the carrier/data phase relationship or the data pulses will move away from the strong carrier's zero crossings and be clipped off by the limiter. Away from resonance, where the "grass" lies, his filter has a very small group delay, so the group delay at the carrier frequency must also be low to match. Hence his insistence on "zero group delay filters". He doesn't understand that "zero group delay" per se isn't important, but rather a matching group delay between the carrier and the data sidebands.
It's been obvious from the beginning that VMSK cannot possibly work, but the real fun has been to figure out how Walker and his friends have fooled themselves. It all seems so obvious now, though of course Walker will never figure it out. Karn 19:43, 10 May 2007 (UTC)

[edit] Who is the VMSK supporter here?

Somebody has been adding comments supportive of VMSK, but I can't tell who or when because they are unsigned. You can use a pseudonym if you want to remain anonymous, but it would be nice to delimit what's written by each person, and to know when it was written.

I say again, this is math and science, not religion or philosophy. The only opinions that count are those backed up by facts, logic and empirical evidence. Those not so backed up are useless. You don't have to take my personal word for it that VMSK is flawed. In fact I encourage you not to take my personal word for it. Take the time to learn the math and science and verify my analysis for yourself.

As for the large amount of money supposedly flowing into companies developing VMSK and UNB, this proves nothing except that humans can be ignorant, greedy and foolish, and they often move in mindless herds. Just witness the 1999-2000 dot-com bubble. You can fool people, maybe even many people, but you can't fool nature. This is exactly why I've spent a considerable amount of my own personal time for the past 7 years debunking and speaking out against VMSK/UNB. This has nothing to do with my job at Qualcomm; it's just that Qualcomm employs a lot of communication engineers like me who understand comm theory. Because I understand this stuff a lot better than the man in the street, I feel a moral obligation to speak out when I see something like VMSK/UNB taking real money from real people.

It'll all be worthwhile if I can help some investor somewhere avoid losing his money. My more cynical friends have asked me why I don't just keep quiet and heavily short-sell companies like CTLG, but my conscience won't let me to do that. It would make me complicit in fleecing people who just don't know any better. Debunking can be a thankless task. You get a lot of ad-hominem attacks and the occasional legal threat. And not just from the companies promoting it -- some of the strongest defenders of Nigerian scam artists have been their own victims who cannot admit to having been fooled. But it's something I just feel I have to do.

Karn 21:30, 28 April 2007 (UTC)

The history shows that the edits originated from IP address 71.52.105.197. According to [2], this address belongs to EMBARQ Corporation, who are some sort of ISP. I don't know if there's any underlying connection here... Oli Filth 22:18, 28 April 2007 (UTC)
I hope not. It would hurt the company's reputation if one of its employees or even worse, a key executive had invested in this technology. What next, Megabit internet over POTS thanks to VMSK? EpiVictor 21:40, 2 May 2007 (UTC)
As an interesting addendum, this IP address appears to be located in Gainesville, FL, which is just down the road from where CTLG are based (Tampa, FL). Coincidence? Possibly. Oli Filth 18:31, 13 May 2007 (UTC)

[edit] If you read too much

into the anonymous nature of my posts, you're making a mistake. It's just that I've spoken to some of you before on these things (on the phone and e-mail exchanges) and I would just like to keep things from getting personal. Here's the latest from 2 days ago (4/30). xG's press release announced the first commercial base stations have been sent, and that they are preparing to offer mobile broadband services. http://www.xgtechnology.com/news_pr_4-30-07.asp. Soon enough, they will either be offering the service to customers in Daytona Beach (close to my house), or they will be confessing this all is a huge ruse. Or you might argue that they will actually offer service, but not be using the technology they say they are using. If you are correct, of course you will be remembered well as the first whistleblower on some massive deception involving a company with a market cap of around $2 billion.

However, if you are wrong, it has been a serious error, of many years standing. I think you deserve a footnote in history either way:) I have no certainty as to who's right or wrong. I say once again, as an investor, that I hope I am right for financial reasons and because I know some of these people and believe them to be good people. If you're right, I lose my money. But as time goes on, it appears you are the one being backed into the corner, not VMSK. They are the ones that have announced successful third party testing after weeks of trials. That third party, in spite of what you have said about them, should certainly have the ability to perform tests that you yourself have said are simple. They've been in their business for 32 years.

If you are wrong, it will be a story that has been repeated throughout time when new technologies knock on the world's door. "It can't be done" is a well-worn phrase that has been heard from the telescope to nanotechnolgy. The naysayers may have had mathematical reasoning, the "latest thinking" in physics or whatever, but it turned out that other scientists were reasoning out issues from a different perspective. I ask again, is there any way you're missing something? As far as I know, the patent office doesn't have the reputation for being a bunch of wild and crazy naked people with a rubber stamp. xG and Walker both have plenty of patents purportedly representing advances in their field. I wouldn't want a single dollar risked because of my views. I am up here because the comments had just gotten too one-sided, and there seem to be important developments weekly that no one here ever mentions.....I am here in the interest of balance. This should be a very fascinating place to stay tuned for anyone who understands what these technologies could mean.....

At risk of turning this page into a discussion forum (which it isn't), here are some replies to your points. (Oh, and please at least sign your posts (using ~~~~), so that people can tell where your comments finished and the next person's began).
There is nothing to suggest that "Eagle Comtronics" has any pedigree in this sort of due diligence process, or in the design of physical-layer communiation standards. Getting them to verify the validity of a modulation scheme would be like getting my local greengrocer to verify the long-term effects of genetically-modified carrots on the human immune system. If they don't understand the issues, then they will be unable to critique the subject matter, and so will have nothing useful to say.
If I (or Phil Karn, or anyone else dismissive of VMSK) were "missing something", then it would require the fundamentals of Fourier analysis and comms theory to be wrong. VMSK simply cannot work the way its advertised if Fourier, Nyquist and Shannon are correct. The maths explicitly prevents it, without resorting to any ambiguity, assumptions or approximations. If VMSK does indeed work, then the implication would be that every communications system in existence today would not be working the way we thought it was.
But it's not even as if Walker has tried to point out some fundamental flaw in our current understanding of comms theory (e.g. something along the lines of "although Fourier analysis would seem to indicate X, this analysis is in fact invalid because of Y"); indeed, from reading the contents of his website, he doesn't seem to be able to present a coherent case at all, and seems unable to deal with rational and logical arguments (see the conversations, etc. listed at Karn's website). Simply re-defining the meaning of variables and blatantly misapplying rock-solid mathematical principles does not make the result correct! So there's literally nothing to go on; nothing to make someone go "ah, right, maybe I'm missing something!".
Quoting examples of experts being wrong in other fields at other points in history does not make Walker's "analysis" any more correct, nor Karn's any less correct. (The examples you quoted were not in fact dismissed for mathematical reasons, but simply because people did not belive that humans were capable of the precision, etc. required to manufacture such things. A modulation scheme, however, can be analysed completely mathemetically. It is, after all, a bunch of mathematical operations on a bunch of mathematical functions. It's even possible to analyse modulation schemes that aren't possible; this is implicitly what Shannon's 1948 paper does!)
As for money that you may have invested in VMSK, ask yourself the following:
  • Why does CTLG have its stock price displayed live on its front page, and create press-releases along the lines of "Market Advisors Issues 12-Month Price Target of XXX"? Do you think they're trying to make something out of nothing?
  • Why do their CEO and CTO, both of who apparently have 25 years of high-profile experience in the industry, have zero internet presence (other than that produced by CTLG's syndicated press releases)?
  • Why does their "technology" page only have images of breadboard-based circuitry?
  • Why are there non-sequiturs, such as referring to their patent (#6748022) as a "cryptology transmission method" that "will prove to be the most secure methodology for the safe transmission of all types of sensitive data", when in fact this patent has nothing to do with cryptography at all?
  • Why are there links to Walker's scrappy incoherent web material, rather than well-presented professional-quality material? If they were indeed at the point where they were implementing this technology, it should not be a challenge for them to produce some decent-quality papers that discuss VMSK.
  • Why do the partners mentioned in press-releases (e.g. "Ludwig Enterprises" and "Balmoral Financing") have zero internet presence, or equally amateurish-looking websites?
To me, this all screams "amateur", certainly not indicative that CTLG consists of a bunch of highly intelligent, innovative comms engineers.

Oli Filth 19:54, 2 May 2007 (UTC)


[edit] I'll do my best.....

As far as displaying the stock price and chart on the first page, so does Novartis (http://www.novartis.com/) on the New York Stock Exchange and many other companies. That IS an area I'm familiar with, and I can assure you nothing is out of place there. Companies need investors, and companies need to keep investors.

As far as Internet presence for the company's leaders, I can tell you the CEO has had high-responsibilty jobs around the world for major corporations, and the CTO has a very impressive resume, including building and selling an early comm company to MCI. He was recently in the Ivory Coast on a government-sponsored trip regarding VoIP tie-ins there. He has "served on the President's Oversight Committee of the U.S. Small Business Administration for Telecommunications." And so much more.

I can't tell you why a common board was chosen for the technology photo. I'm not sure that is significant.

The cryptology potential is inherent within VMSK technology, and not separated by additional patents, to the best of my knowledge.

Walker has a 130-some page book available online. Surely he must speak to some of your issues. I wish I could understand it - its not my field.

The firm handling certain of their European rights negotiations is "Balmoral Capital Holdings Inc." (mentioned in the initial CTLG release concerning the relationship), which must be the parent of "Balmoral Financing". I found this release involving a Spanish telecom acquisition - http://uk.biz.yahoo.com/09022007/290/sentex-sensing-technologies-inc-signs-binding-mou-acquire-vitelcom-mobile.html, in which they acted as investment bankers.71.52.105.197 20:05, 3 May 2007 (UTC)71.52.105.197 18:48, 3 May 2007 (UTC)

As for the stock prices, fair enough. I'm still of the opinion that this is a ruse though, of the form:
  • Force numerous vague, vapid press-releases through as many syndicates as possible.
  • Unknowledgable investors do no due diligence and buy stock, forcing prices up.
  • Other investors go to website, literally the first thing they see is that prices have gone up, and so invest, forcing prices up further.
  • Repeast ad nauseum.
Though this, of course, is pure speculation on my part.
As for the company's officers, do you have any independent proof of their "achievements", or just assertions made in CTLG's publications, e.g. [3] (p.25)? (I'm not asking because I need this proof, I'm asking for your own peace of mind!) Google any of the companies listed on that page; you won't find mcuh at all, which is odd considering they're all telecoms companies! (that's not to say they don't/didn't exist, merely that they were failures). Hardly a high-profile resume to be proud of.
I think the photo is entirely relevant. This is supposed to be a company promising groundbreaking technology, and the best example they can come up with is a few components cobbled together on a hobbyist's board?
None of the VMSK patents nor Walker's web material mention anything about cryptography. It's certainly not "inherent" within the technology.
Luckily, physical-layer comms is precisely my field. I've read through the whole of Walker's online book, and I must say that it's been a struggle. I say "struggle" not because it is complicated or advanced (quite the opposite, in fact), but because 99% of it is incoherent rambling, completely unstructured, with no discernable train of thought to follow. Any of it which is possible to follow inevitably turns out to be wrong, usually for very trivial reasons. I therefore know for a fact that VMSK is nonsense. If I turn out to be wrong, I might as well give up my job, as I use this maths and theory on a daily basis, so everything I've done up until now would by implication have been designed on incorrect principles. Unlikely though, because the stuff I work on actually functions correctly and as intended!
I appreciate that for a layperson the subject must be hard to follow (Walker doesn't help by writing stuff that's structured like a 10-year old's science-experiment write-up), and therefore it's difficult to know who to believe. However, a casual skim through Walker's material ought to be enough to convince anyone that this is not the work of a genius (not least because he's being working on this for over 7 years - more than enough time to put together a decent, readable publication!).
Oli Filth 22:04, 3 May 2007 (UTC)

[edit] Debunking in layman's terms

Hoping Mr. Karn and any fellow electrical/telecom engineers will forgive me for the extreme simplifications and some degree of term abuse for the sake of clarity, I'll try to write a short debunking in layman's terms, for someone entirely unfamiliar with modulation or signal theory. The only thing required is understanding the concept of frequency, and bandwidth.

As a practical explanation, a standard phone line cannot plainly carry a frequency of 5000 Hz because it has a bandwidth such that it allows only frequencies from 300 to 3400 Hz to "pass", everything else is either strongly weakened or cut off. It is possible to convert a certain "zone" of frequencies, e.g. from 5000-7000 Hz to 1000-3000 Hz to make them "fit" through the channel but I can't "squeeze" a zone larger than the bandwidth itself, e.g. I can't generally "squeeze" a signal with frequency range from 1 to 7000 Hz inside a medium with a band of 300-3400 Hz unless special conditions apply (e.g. I encode the signal differently, essentially trying to compensate the lack of available frequencies by using a more precise amplitude encoding).

That's what Mr. Walker indirectly claims to be able to do, but instead of compensating for frequency with amplitude and more sophisticated modulation, he uses a simplistic modulation scheme based on the delusion that somehow embedding high frequencies "as they are" will go "unnoticed". That just doesn't work.

"Modulating" a carrier with a certain frequency, e.g. changing an FM Stereo carrier with a signal of a certain frequency, will make that same frequency (or even worse, multiples of it) appear in some form inside the original carrier. Thus, if the channel transmitting the modulated carrier does not have the bandwidth to carry both frequencies (carrier's + signal's) the modulation scheme just can't be used, and will result in distortion and loss of information.

Mr. Walker instead, claims to be able to do just that: the concept of "slightly marking" a low frequency carrier with a (much) higher frequency one while keeping the OVERALL frequency requirements equal to the carrier ONLY is just plain wrong: the higher frequency will have to appear in some form, and thus it will need a channel width at least as wide as the frequency itself. If (oversimplifying VMSK) I try to "slightly change" the phase of a 1 kHz carrier a million times per second (1 MHz), that 1 MHz will have to appear somewhere, plain and simple. It can't be "packed" with such a naive modulation scheme.

There are ways to "pack" information contained in a 1 MHz carrier into a 1 KHz one, but VMSK isn't one of them, plain and simple, as any and all valid methods require transferring the modulation complexity into the carrier's amplitude (essentially, trading bandwidth for signal/noise ratio, which implies being able to discern between increasingly tiny signal level variations. That's another reason why e.g. 1 Mbit/sec cannot be carried by a POTS phone line). EpiVictor 08:56, 7 May 2007 (UTC)


[edit] Well-known Ph.D. in support of VMSK?

On Walker's site, he quotes a from a "recent book" by Dr. W.Y.C. Lee, formerly Vice President of Vodafone:

"Dr. Lee witnessed an over the air demonstration held at Vodaphone's facility in Walnut Creek, CA.. He has this to say in his recent book (1). "VMaxSK technology can also be used. It can send 48 kb/s data through a 2 kHz filter and receives with good quality "----. He also refers to the 1 Hz transmitted bandwidth. 48 kb/s was used as a comparison with IS136, which presently uses 48 kb/s data in a 30 kHz channel. The demonstration he witnessed at Vodaphone's lab. was actually transmitting 273 kb/s through a 2 kHz receiver filter noise bandwidth. This data rate was later raised to 812.5 kb/s to use EDGE software. Data rates as high as 6.0 Megabits/sec are presently being transmitted in adjacent 30 kHz Cellular channels. This has now been raised to 12 Mb/s in the lab., with 5 channels operating at once with 30 kHz separation. Dr. Lee is quite capable of judging a technology by its merits alone."

What do you make of that, and is anyone familiar with this book?71.52.105.197 —The preceding unsigned comment was added by 71.52.105.197 (talk) 14:39, 7 May 2007 (UTC).

I'm not familiar with the book, although one may peruse the table of contents and index at Amazon. There is no mention of VMSK in either. To me, it seems unlikely that the author would discuss a technology (especially in acronym form) and then fail to mention it in the index. It also seems unlikely that a prominent researcher in the field (Lee) would blithely talk about a "1 Hz transmitted bandwidth", something that is essentially impossible according to standard engineering wisdom.
It also seems unlikely that Walker demo'd at Vodafone (which would be a phenomenonal achievement), but yet is incapable of spelling it correctly. Oli Filth 17:59, 7 May 2007 (UTC)
Oh, and the claim of 12 Mbit/s in 30 kHz equates to a spectral efficiency of 400 bit/s/Hz; an absurdly huge number. Plugging this into Shannon's equation, we get that the absolute minimum required SNR is 2400, a number so big that it doesn't fit on my calculator. This is equal to approximately 1200 dB. To put this in perspective, the ratio of an yottawatt to a yottowatt (see Watt#SI multiples) is only 480 dB. As another comparison, the estimated number of particles in the entire universe is around 1079, equivalent to only 790 dB. Oli Filth 18:22, 7 May 2007 (UTC)
Indeed. The fact that these schemes reportedly do pass data with less power than would vaporize the galaxy is itself conclusive proof under Shannon that they use much more bandwidth than is claimed. But you don't even need Shannon to dismiss Walker's claims. Because he says his modulation is binary, the simpler Nyquist Sampling Theorem is all you need to reject them. Nyquist says that you cannot achieve more than 2 symbols/sec per Hz of bandwidth at baseband, or more than 1 symbol/sec per Hz of bandwidth at IF or RF. For binary modulation, a symbol is a bit, so this means no more than 2 bps/Hz at baseband or 1 bps/Hz at IF or RF. A quick look at his waveforms with Fourier analysis shows exactly what's going on. There is no mystery here except how Walker can manage to attract fans who ought to know better. Karn 21:42, 8 May 2007 (UTC)

[edit] I noticed that four of Lee's books

are not searchable at Amazon, so the reference could be in one of those. VMSK's value, or lack thereof, should be settled pretty quickly right here. Which book, and what exactly does it say, if anything? It may also be included in an upcoming book of Dr. Lee's, as CTLG recently announced that Wiley & Son's, the publisher of at least one of Lee's books, will be including a chapter on VMSK in a scheduled book (this was confirmed by Wiley). It may take a little time to find out which book to order, plus time to order and receive the book. I think Dr. Lee's positive assessment would carry very, very significant weight, should it exist. It would be most unusual for Mr. Walker to make up or alter a quote from an eminent source such as Dr. Lee, and have it rest unchallenged on his site for 10 months: http://www.vmsk.org/Comments.pdf. I think we all can all agree that, if real, the value of VMSK would be hard to calculate.71.52.105.197

The book that Walker references is the one I linked to; "Lee's Essentials of Wireless Communications". I haven't got the book, so obviously I'm speculating as to whether the quote is real or not. But if it is made up, who exactly is going to challenge it?
IMO, it would be much more "unusual" for an eminent source such as Dr. Lee to publicly endorse a technology that makes claims as absurd ("revolutionary", if you will) as a spectral efficiency of 400 bit/s/Hz, and no-one to challenge that!
It would also be "unusual" if the Vice President of Vodafone endorsed such a revolutionary technology after a demo, but yet failed to invest in it in any way; i.e. Vodafone purchasing the rights to the technology and developing it.
Incidentally, do you have a link or something that describes the chapter in the upcoming book? I can't find anything via Google.
One last thing, it's not necessary to create a new heading for every message that you leave here! Indentation is the preferred way of indicating a reply (see Wikipedia:Talk_page_guidelines#Technical_and_format_standards). Oli Filth 19:45, 7 May 2007 (UTC)


[edit] Dr. Lee's VMSK reference located

For some reason, if you search the book using his quote "good quality", please refer to #19 in Amazon's search results referencing page 289: http://www.amazon.com/gp/reader/0071345426/ref=sib_dp_srch_bod/102-9621408-8692967?v=search-inside&keywords=good+quality. It is only a partial quote, and thus leaves the issue unresolved for a short time, but clearly Dr. Lee is familiar with VMSK. Until the book arrives, that's all I can locate at this time. Possibly someone reading this discussion owns the book and can shed a more immediate light on this and other possible references to VMSK within Dr. Lee's book.71.52.105.197

Hmm, yes, so the book certainly does mention it. Through some "magic", I've acquired the full quote:
"AlphaCom technology also can be used. It can send 48-kbps MD-3 data stream through a 2-kHz filter and receives with good quality. The idea is to find ways to slightly mark the states of the modulation and VMSK (variable phase shift keying) on the carrier wave such that less distortion of the carrier waveform can be achieved. Of course, we know that undistorted CW carrier only needs 1-Hz filter in principle."
There are no other mentions of VMSK in the book.
AlphaCom was a previous company involved with VMSK; it no longer exists. Quite what "MD-3" data is, I have no idea. Exactly what the third sentence means, I'm not entirely sure; it's written very bizarrely (in short, one does not normally consider applying modulation as "distortion" of the carrier, the whole point of modulation is to alter it!). The final sentence is equally bizarre; an "undistorted" (presumably "unmodulated") carrier theoretically requires no bandwidth at all, there's nothing magic about 1 Hz. Either way, it in no way backs Walker's claim that VMSK requires only 1 Hz.
Nevertheless, I do accept that this stuff has somehow made it into a publication by a fairly prominent figure within the industry (although never to be mentioned again by Lee, as far as I can see). But I really don't understand how it got this far! The maths, etc. that Phil Karn uses in his debunking is rock solid, so quite what Lee was thinking when he wrote that section, I don't know.
Phil, if you're out there, do you have any thoughts? Oli Filth 23:43, 7 May 2007 (UTC)
That quote from Lee has been there, unchanged, for quite a few years. I don't think it's as old as my awareness of VMSK (spring 2000) but it's close. Lee is a real person; I met him very briefly during a CDMA demo at Qualcomm in the early 1990s, but I certainly don't know the man and cannot judge his technical capabilities. If he is as sharp as his resume suggests, and if the quote is accurate, then I can only say that he must have had an off day or been really busy or distracted when he wrote this, and he probably hasn't thought about it since (there apparently being no further quotes from him on the subject). Even smart people sometimes make dumb mistakes.
Remember what it was like in 2000; it was a heady time with investors tossing truly mind-boggling sums of cash at anybody and everybody with a wireless or Internet-themed business idea, no matter how outlandish, impossible or just plain silly. When the inevitable correction came, the market fell by far more than the market cap of CTLG.PK, XGT.L and any other company supposedly developing "ultra narrowband". Karn 06:58, 8 May 2007 (UTC)

[edit] Dr. Lee's Credentials

Dr. William C.Y. Lee, Director Dr. William C.Y. Lee is the Chairman of LinkAir Communications, Inc. Dr. Lee is renowned for his leading contributions in making analog and CDMA technologies commercially viable. He has published more than 200 articles and seven technical books on CDMA theory and technology. His mobile cellular telecom systems book published in 1989 was the first cellular system book ever published. An expert in developing marketable communications technologies, Dr. Lee is the inventor of Microcell, a leading technology that increases frequency reuse factors and boosts capacity by 250 percent. He holds more than 27 US patents, with 11 more pending.

For 15 years, Dr. Lee was one of a team of pioneers in developing advanced wireless technology—AMPS—for Bell Labs. He then joined the ITT Defense Communications Division, where he headed the advanced mobile communications system. He was Vice President and Chief Scientist during his subsequent tenure with Vodafone Airtouch where he assisted in CDMA research and the initial trial of the technology.

The pioneer in personal communications network (PCN) technology, Dr. Lee led PacTel’s PCS experimental trial, and under his leadership, the first CDMA phone call was completed in Los Angeles in 1995. Dr. Lee has been selected as an IEEE Fellow and has served as a member of numerous Councils, including the California State Council on Science and Technology, the US Council on Competitiveness, and was a former member of the FCC Technical Advisory Council from 1999-2001. He has earned many prestigious industry awards, including the IEEE VTS Avant Garde Award, the CTIA Award, the CDMA Industry Achievement Award, the SATEC Award, a Bell Lab Award and, most recently, the IEEE Third Millennium Medal Award. Dr. Lee has received the Telecommunication Achievement award from Chinese Historical Society of America and the Stuart Meyer Memorial award from IEEE Vehicular Technology Society.

Dr. Lee earned his Ph.D. in electrical engineering from Ohio State University.71.52.105.197

Here is a link to Amazon's page of Dr. Lee's books: http://www.amazon.com/exec/obidos/search-handle-url/104-7460485-3659131?%5Fencoding=UTF8&search-type=ss&index=books&field-author=William%20C.%20Y.%20Lee71.52.105.197

[edit] I think you may all be in for a surprise

When you find out who's been working - very successfully - with VMSK.

So... do you have any information you'd like to share? Or are you just going to add unhelpful unsubstantiated fluff to the discussion? Oli Filth 18:45, 11 May 2007 (UTC)
 Gentlemen, I would certainly like to do so, and there are no big secrets here, but it isn't
 my place to share something that others prefer to announce themselves. Let's just
 wait about 60 more days (maximum), and I think you will be very satisfied with the
 news.71.52.105.197 14:35, 14 May 2007 (UTC)

Inasmuch as the laws of physics don't change with time, and proved mathematical theorems can't be unproved, it's hard to imagine somebody actually showing VMSK outperforming conventional modulation methods. (Though I can certainly imagine clueless suits thinking that's what they're doing.) So maybe CTLG will throw in the towel and reimburse their investors? The SEC will announce an investigation? Stranger things have happened.

Then again, I've been following pseudoscience long enough to know that coy promises of big surprises -- always just a few months off, maximum -- are the crackpots' stock in trade. Karn 06:08, 15 May 2007 (UTC)

  Then again Phil, you're hanging your hat on Dr. Lee having "a bad day" that actually was so
 bad it lasted a couple of months right through the editing process and made it into
 his book.....I will just predict that one day the famous, and possibly soon to be
 imfamous, VMSK Delusion page will be brought down by your own hand, and that you will
 personally want to rewrite the main article here on Wikipedia with mea culpas. If
 that happens, then the discussion should turn to the possible conflicts involved with
 your work at QUALCOMM when you have stated right here that you were aware of Dr.
 Lee's quote for years. Just an opinion for now. No, I'm not a scientist, but I do
 know with some 99+% accuracy who is having excellent results working with VMSK.
209.16.117.50 22:35, 15 May 2007 (UTC)

...I may have missed something here, but sending a "3.67Mbps signal over 18 miles with only 35 mw of power" doesn't seem a very great achievement, considering DIY enthusiasts have achieved 11 Mbps@15mW and 200m with several dB room to spare, and even a 2Mbps connection over 2km with unmodified wi-fi equipment (that means 15mW max) and with DIY directional tin can waveguide antennas with not perfectly matched cables. [4] Just my 2 cents. And let's not even ask how that nW figure was obtained... EpiVictor 15:54, 15 May 2007 (UTC)

No, I certainly don't "hang my hat on Dr. Lee having a bad day". I hang my hat on the very large body of work by Joseph Fourier, Harry Nyquist, Claude Shannon, and the many other luminaries I learned about in great detail during my undergraduate EE education, and who laid the foundation for just about all of modern electrical engineering and communications on which I have managed to pursue a successful professional career for over 30 years. I also hang my hat on mathematical proofs, once made and verified as correct by countless others, and at least understood by me as to their implications, not suddenly being undone by an anonymous person who admits to no relevant training or background, acting coy about some vague surprise announcement a few months hence. But hey, if I could be shown to be wrong on this and that the carrying capacity of the radio spectrum is far greater than I ever thought possible, I'd actually be quite happy. It's just that, unlike you, I consider it important not to let my wishful thinking control my perceptions of reality.

Oh, and I suppose I could ask how someone who admits that he is not a scientist or engineer would even know what would constitute "excellent results working with VMSK", as opposed to smoke and mirrors designed to wow the technically clueless, but I guess that would be far too obvious a cheap shot, no? Karn 11:25, 16 May 2007 (UTC)

"Joseph Fourier, Harry Nyquist, Claude Shannon, and the many other luminaries I learned about
 in great detail during my undergraduate EE education" is a very relevant reference.
 But something tells me that they will not be remembered as the last people who
 learned anything new about RF technology. I have a vague feeling others in the field
 will continue to enhance and advance their work. And while your undergraduate work no
 doubt serves you well and puts bread on the table, Dr. Lee earned his Ph.D. in

 electrical engineering from Ohio State University, and his team literally made
 the first CDMA call. I think he may be slightly - just a little - more advanced in
 his understanding of how VMSK works. If every EE understood it,
 it wouldn't a potential breakthrough. Sometimes we get locked into a mindset that just brings
 the thought process to a halt. You see VMSK as an impossibility, and something tells me you
 reached a point, made a decision, and are not going to look anymore. Maybe that's a pretty
 standard approach for all of us. Maybe it shouldn't be once people like Dr. Lee publish
 their views. At that point, comments like "perpetual motion machines" just become...mysterious. (anonymous comment)

I never said there could be no further advances in RF technology. If that were true, I'd be out of a job. But an essential prerequisite to advancing any technology is to know the crucial difference between what violates fundamental mathematical and physical principles and what is possible but just hasn't been done yet. That's the most important function of an engineer's education, which you concede you don't have. Funny you should mention CDMA; I came to Qualcomm in 1991 precisely because I could see that it was entirely feasible and a significant advance over what had already been done. Lee was one of our early customers, but we did the actual work. We're grateful for his early support of CDMA, but he's still completely wrong about VMSK. Even smart people sometimes make mistakes, and Lee's are his responsibility to explain, not mine. Karn 23:16, 16 May 2007 (UTC)

Well, now that 60 days has passed, I wait with baited breath to see what news our anonymous editor will surprise us with... Oli Filth 19:01, 12 July 2007 (UTC)


[edit] Let's Recap

We have the exalted Dr. Lee in one corner, along with xG supporter & Electrical Engineering Professor Stephen Schwartz of Princeton. Plus my knowledge of who continues (successfully) to work with VMSK.

In the other corner, we have you guys who I know very little about. Even if I know nothing about anything - who would I believe? They have the more weighted credentials, I think you may agree.

On the other hand, I am absolutely certain you gentlemen are seriously grounded in the fundamentals of RF technology. No question about it. But, if it's so easy to calculate the flaws in these technologies - and you repeatedly refer to them as almost childish efforts - tell me how that could possibly escape these capable men? How could that be? You seem convinced, they seem convinced. I am now not sure where to take this other than wait out xG & CTLG developments. They will sink or swim on their own merits or lack thereof, and in the end, that is the out of the control of any us at our keyboards.

It would have been wrong to leave the other side of the case without a voice. —Preceding unsigned comment added by 71.52.105.197 (talk • contribs)

Just a friendly reminder - claiming credentials on Wikipedia is troublesome, please have a look at Essjay controversy. May I also remind all parties here that suggestions are to be taken on their own merit, to be discussed in details and implemented if worthy. Regards,xC | ☎ 15:20, 17 May 2007 (UTC)
I strongly agree. Don't take my word for it that VMSK is bunk, learn the math yourself. Karn 22:42, 17 May 2007 (UTC)