X-Message-Number: 21058
Date: Wed, 5 Feb 2003 05:18:27 EST
Subject: Data flow

Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit

"Every complex, difficult problem has a simple, easy solution, which
is wrong."

Extracting informations from  a partially autolysed brain is a difficult 
complex problem. I have suggested two solutions: An intensity interferometer 
using quasi-particles such phonons and a quantum non-demolition x-ray 
interferometer. My estimates are that the first would need some months to 
scan a brain at the small molecular level. The QND interferometer would do 
the job in some seconds. So at first sight, the second seems far better. To 
be sure, the first would be simpler to build and a part of its technology 
would be requested for the second...

There I would see them on another angle: The data flow and processing. Assume 
a brain volume is 1,500 cm^3 or in round value: 1000 cm^3. If there is one 
molecule for each nanometer, then the scan must looks at: 10^24 places. The 
data for such a mesh could be 100 bytes, so the total data would be: 10^26 b. 
The slow intensity interferometer taking 4 months for a scan (10 millions 
seconds) would sustain a data flow of 10^20 bit/second. Even with a 
compression factor near 100, this would let one billion gigabits/s.

Assume we can handle on an electronics circuit a 10 terahertz waveband or 
nearly 10^13 bits/s, we are short by a factor of 100,000. A solution would be 
to use an optical fiber technology. A cable can contain up to 1,000 fibers, 
this is what can be found in astronomical fiber feed spectrographs. Each 
fiber can have up to one hundred propagation mode, each is, in fact an 
independent light beam. For each mode in each fiber there can be 100 carrier 
frequencies. So in the end, an optical system could boost the electronics 
modulation capacity by up to ten millions. All data could be routed without 
compression. Another option would be to limit the modulation frequency to 100 
GHz. This is 10 - 20 times what can be found today and may be the technology 
on the market ten years from now.

The intensity Interferometer data flow seems so a manageable problem. The QND 
interferometer on the other hand would have a data flow one million time 
larger. Even with 1,000 propagation modes in 10,000 optical fibers, the 
electronics modulator can't work under 10 THz with the maximum data 
compression. the carrier wave would have a frequency in the 100 THz range. 
Clearly, a mere extension of current electronics devices can't cope with 
that. What are the possibilities?

Collective electron gas waves?
Free electron lasers?

Whatever the selected solution, it will take time to be implemented. These 
requirements are so specific to a brain reader than we can't simply wait for 
the general progress to produce it. We have to create that technology or it 
will never materialize.

This is not a philosophical question: The answer will define where the money 
will go 2 or 3 years from now.

Yvan Bozzonetti.


 Content-Type: text/html; charset="US-ASCII"


Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=21058