Late and Probably Never

Hilarious press release this morning. Remind me never to use the same courier service that xG use: Finally we have delivery of the Army trial equipment... 140 days into a 180 day trial. And that, according to xG, is a "timely delivery"!

Now, I'm a big fan of the armed forces, but this press release took it a little far in its praise dontcha think: "dedicated", " resourceful", " leadership", "exceptional" soldiers. Maybe Marc's fellatio fantasies are more common than I thought.

The last of the quote is pretty telling. Basically it says, even though we've only given you 40 days to test it - it doesn't currently work : "[xG will] listen to them, to learn from them, and then revise our trial product to reflect their recommendations and better meet their needs."  

Edit:
and the last line of the 'about xG' is also notable. Nothing new, but in this context boasting about the number of your patents isn't a great idea. Some inconvenient person may compare yours to, say, Harris communications who is the current vendor of most Army radio systems. Have a look HERE at the US Patent Office filings for Florida. Harris are near the top. Scroll down to see where xG are.....

"...iDigi Works as Advertised.."

Now that we have a new brokers' note, let's have a look back at the last one and compare.  In 2006 xG obtained an impressive 50 Mbps in their trials. ("These are actual figures, not theoretical" their broker Hitchens noted.) This was well in advance of any modulation available at the time. Now in 2010, following a technical review from PA consulting on behalf of Allenby, they come up with a maximum theoretical figure of 6 Mbps - that's only 12% of their previous "actual" speed...

But wait. Hindsight is a wonderful thing. This has all happened before. In 2000 iDigi were claiming 3 Mbps for their system. Again, this was well above any other figures the industry were quoting. Again, these were real figures, tested (by xG themselves - in a fraudulent demo.) But in 2001 when the system was tested by Katapult it only gave performance as low as 0.04 Mbps - that's only 1% of their claimed speeds, lower than a basic modem!

I see a pattern emerging...

What gives? Could it be that they haven't got a scoobie? Or could it be that they lie to investors to steal their money from them? I'm going with the latter, but if you are an investor the result is the same - you will lose all your cash...

Rich Tehrani: If you think xMax works "as advertised" I've got a lovely perpetual motion machine I'd like to sell you. Sure, it's only 1% efficient but...

Docket 03-201

The FCC are considering 'Spectrum Etiquette' regulations that would severely degrade xG's proposals to use the 902-928 Mhz band. Docket 03-201 is supported by Wireless Internet Service Providers Association (WISPA) and others. In their webpage they state:

"[spectrum etiquette] may promote more efficient spectrum utilization by reducing interference between different license-free and licensed operators and should be considered for newly-certified devices"

Big Problem for xG. Big BIG problem. They via their lawyer plead (see addendum):

xG is concerned that adoption of the spectrum etiquette proposal, in its present form, will have an adverse effect on the deployment of its xMax technology.

xMax employs what can best be described as frequency agility. In the 902-928 MHz band in
which xMax operates, the band is divided into 18 discrete channels. When potential interference is detected on a channel on which the xMax device is operating, due to the presence of another low power device in close proximity operating in the 902-928 MHz band, the xMax device is capable, in milliseconds, of jumping to one of the other 18 channels, hence the term "frequency agility." The likelihood of encountering other devices operating on all 18 channels in a mobile environment at the same time is probably close to zero statistically. Thus, xMax has been designed to greatly minimize, if not eliminate entirely, the prospect of causing interference to other users of the 902-928 MHz band.

There is accordingly no need for a duty cycle, insofar as xMax is concerned, to allow other users access to the unlicensed 900 MHz band, despite the fact that xMax is designed to operate
continuously.

While adoption of the spectrum etiquette proposal would not be fatal to the deployment of the
xMax technology, it would require extensive redesign and development with the attendant additional costs, further delays in deploying the technology and reduction in system capacity.

If docket 03-201 is adopted, can you see Bohman putting his hand in his pocket to pay for 'extensive redesign and development'? Especially now that he's made Rick cough up some of his own cash, and now that Rick is passing the hat round for $10m? Me neither.

If docket 03-201 is adopted then newMax is dead.

Its father, xMax, died a few years ago. For the technically minded it's instructive to compare the system outlines in the FCC link above with the one xG made to the FCC back in 2003. They are two very different systems. Funny. I missed the stock-market announcement about the change...

NewMax Vs xMax

There's an interesting post over at Brough Turner's Blog HERE. He's an acknowledged radio expert who seems to know his stuff. Back in 2006 he had much the same conclusions as Phil Karn that to work as they then claimed xMax would have to violate physical laws, and/or FCC regulations. Then the claims were of 10,000 times less power than other systems for the same data-rate, cities covered with 4 basestations, and networks run on watch batteries. It was madness. To make matters worse the company was promising an imminent launch of this impossible system: The R&D is over said Mooers in March 2006 "We’re at the commercialization and product stage now." The launch was promised before the end of 2006, with 4G networks coming shortly after. Well, that went well.

This obviously made them a laughing stock and exposed the ridiculous lack of knowledge (like the Shannon Bound) of basic communications theory of the inventor Joe Bobier, and the psychotic optimism of Arsehole Rick Mooers. The fact that xG were touting for investors to buy shares in an invention that couldn't possibly work was fraud pure and simple. And they could never claim they were ignorant of this: Karn told them, I told them, it was all over the Internet.

Then, somehow, Joe seemed to have been introduced to the real world. This culminated in the FCC certification of BSN250. This was a basestation that seemed to use perfectly ordinary modulation methods. Nothing spectacular. Nothing unusual. Boring, but it would at least work.

The problem with that is twofold
1) Investors were promised a Ferrari, xG unveiled a bog standard hatchback. It was the most outrageous misrepresentation.
2) Nobody thought of telling the stockmarket that xG's only product had changed 180 degrees in violation of many stockmarket regulations.

To xG's credit they have used this conventional radio modulation system to make a network. Gone was the Wavelet Pass Filter which Bobier and Schwartz had claimed was the "key to xMax", which meant that interference was 'eliminated' (unless you were conducting a xMax demo obviously), and in was a conventional system that switches frequency many times a second.

In July 2010 Brough actually went to visit Joe to see the system. As we imagined the modulation was standard, and running at 2G speeds, not the 4G promised four years ago. There's nothing special. Most fascinating for me though is an entry in the comments:

Brough

Thanks for the fair overview of where xG is today. It did take a reinvention of the company to move into our new direction of offering carrier class cognitive radios. Hopefully we can get this technology into the standards at some point.

Rick Rotondo
VP Marketing
xG Technology

That, to me is incredible. He's admitted the bait and switch - The 'reinvention'. He's as good as said what they have now is a standard modulation, not the amazing low power/high data rate modulation they were flogging initially. The system around it may be cute, but there are many other cute systems out there.

I'll email Mr Rotondo to comment here. I wonder if he will.

Edit: Surprise! No news from Mr Rotondo. There's another blog over at 4G Wireless Evolution that mentions xG. Carl Ford says:

The chip makers for WiMAX are all finding their way into LTE - mostly through Time Division over Frequency Division, which according to our friends at xG Technology, is a secret sauce that will allow beam forming and cognitive radio to live in noisy spectrum.

TMDA and FDMA (Time and Frequency Division Multiple Access) are fairly common among modern radio systems. The 'Key to xMax' - the wavelet pass filter - is gone. The incredible modulation is gone. And so is the $1.5Bn valuation.

Don't look at the man behind the curtain. Move along, nothing to see here.

Real Life Performance Data

Quite a few people over at iii.co.uk, Raustovitch steadfast among them, have held on heroically to their xG shares in the distant hope that mathematics was wrong and Bobier was right. 'Lets wait for real-life performance data', seems to be their call.

Well chaps, there's good news and there's bad news.

The good news is; I have stumbled upon some actual performance data for you, done in a real-life test with real xMax equipment. Yey! The bad news is; umm, sorry, Bobier was ~wrong~. xMax is just ~nothing special~. It works, sure, but (according to people who know) its performance is 15 years out of date. Ouch! There's two websites below with the data. They are tests done in Florida with the xDrive system. The first one shows two tests done 13/3/08 and 11/3/08, you can play with the settings - a figure of 34 is required for acceptable voice, and you have to plug in the antenna location to get proper readings. On the second site, the setting required are shown as 34(signal strength) 29.30(latitude) -81.1127(longitude) 1(antenna sector). I don't have the second antenna location for site 1. The second site shows pins all over the 'Birds Eye' view of the tests area. Passing your mouse over the pins shows the signal strength.

http://www.wirelessexpressions.com/xdrive.aspx

http://www.wirelessexpressions.com/virtualearth.aspx

Aside from showing xMax's radio performance is 15 years out of date, the tests don't actually show how well an xMax system would work in real life. It's probable that performance is even worse that 15 years out of date because these tests don't include any interference. xMax is meant to work in unlicensed spectrum so *will* be subject to interference. "But", says Mr Raustovitch, "What about the Wavelet Pass Filter - the 'Key to xMax' according to Bobier - which means xMax can bypass interference?" Well have a look HERE just to remind yourself. It can't.

It won't work better in unlicensed spectrum. It won't carry more users. Their BER curve does show their true performance - despite their hurried retraction. It won't send signals further on less power. It won't need less basestations to cover an area. It won't make a network cheaper than other systems. All the revolutionary performance claims made to fool money out of investors in this stock scam are FALSE. xMax is nothing special in theory. xMax is even less special in practice. xMax is dead. xMax only ever existed so that Mooers could con people like Raustovitch out of their hard-earned money. xMax only ever existed as a scam.

What we have now with all the changes highlighted by the FCC tests is something quite different to xMax. Something new. Something regular. Something pedestrian. Something that could work. Let's call it "newMax". Raustovitch: Sorry mate. They've offered a brand new rocket-powered Lamborghini to entice you in, and switched it for a 15 year-old Ford Mondeo and hoped you wouldn't notice.

Dealer Program PPT

So now we know that the performance of the BTS250 bears no relation to what they originally claimed for xMax. Wireless Guy over at iii seems to know his stuff: "This looks like it will do a fine job of supporting a 100 or so VOIP calls and perhaps text messaging or emails..." BUT as this PowerPoint show proves they were telling the dealers that it would support 780 simultaneous calls, and that OPEX and CAPEX would be lower than other systems. Not might. Not may. Not could. But will support 780 simultaneous calls

Another interesting thing is (slide 4) that they claim the carrier (Telcentris) rates are specially negotiated - but then send me a Cease and Desist when I post them as they are too embarrassing...! And the final hilarious kicker is the launch date: April/May/June 2007!

Unless they announce they have changed their technology to the
stockmarket, they have ONCE AGAIN broken the AIM rules.

I've highlighted interesting things in red. Not all the slides are shown as quite a few are just marketing blurb. First up is the properties of the doc so you can see who wrote it.