From: jim bell <jimbell@pacifier.com>
Date: Wed, 15 May 1996 19:52:42 +0800
To: cypherpunks@toad.com
Subject: Past one terabit/second on fiber
Message-ID: <199605150618.XAA28824@pacifier.com>
MIME-Version: 1.0
Content-Type: text/plain


In the April 1996 issue of Fiber Optic Product news, there is an article on 
a Lucent Technologies (formerly Bell Labs...no relation...sigh...) product 
which wavelength-multiplexes quantity 8, 2.5 Gb/second signals on a single 
fiber, for a total of 20 Gb/sec.  This is a real, purchaseable system.  On 
the same page is a somewhat more experimental system, done by Corning and 
Siemens, in which eight channels at 10 Gb/sec each were transmitted on a 
single Corning fiber.

"Wow", I said.  Far faster than the 2.5 Gb/sec transmission that is 
currently fairly standard for long-haul fiber trunks. 

I wasn't prepared, however, for page 38, in an article titled "Research 
Teams Achieve 1 Trillion bits a Second."  In fact, three separate groups did 
this.  I copy the article below.

CP relevance?  Well, the justification the government uses to regulate the 
airwaves, via the FCC, is that the available bandwidth is limited, which it 
is.  But that argument has never been true with fiber, at least in theory, 
and is becoming even less true in practice.  For example, that recent flap 
over Internet-based long-distance telephone interconnects (LD companies 
don't want competition) is based on the fact that the normal providers of 
these services want to get their dime a minute rates come hell or high 
water. Sure, that's a might cheaper than it was a decade ago.   But with 
fiber transmission probably less than 1/100th the cost of older coaxial 
transmission systems, per connection, it is unclear why they're even 
continuing to meter LD phone calls.  

Even if we only consider that 20 Gb/second fiber from Lucent, that is 
equivalent to about 300,000 simultaneous voice calls.  With a standard, 
36-fiber cable, that represents 18x300,000 two-way calls, or about 4.8 
million calls.  This is probably far greater than the maximum number of 
people on LD in the US at any given time, and that's just a single cable trunk.

If we assume that the fiber cable costs $1/meter per fiber, and the cost of 
trenching, burial, and interconnects raise this to $10/meter/fiber, and if 
we generously assume that the average LD call goes 3000 miles (5,000,000m), 
that call occupies 1/150,000th of a $50 million fiber for a few minutes.  If 
we suppose that the fiber has to gross $100,000,000 per year to pay for 
itself, and even if it's only operating at an average 10% load level(both 
assumptions are pessimistic, that only works out to a cost of 1.3 cents per 
minute per call.  That's why these LD phone companies are so scared:  If we 
can transmit Internet on fiber, that fiber can accept this extra traffic at 
very low marginal cost.


Part of article follows:

"Research teams Achieve 1 Trillion bits a second"

Debra Norman, Editor in Chief.

Three research teams achieved their ultimate goal by sending the most 
information possible over optical fiber. 

The scientists, including a 12-member group from AT&T Research, Bell 
Laboratories, Lucent Technologies, reported in post-deadline papers at the 
Optical Fiber Conference held recently in San Jose, Calif., that they had 
sent one terabit of information over non-zero-dispersion fiber in a second's 
time.  In short, it is similar to transmitting the contents of 1,000 copies 
of a 30-volume encyclopedia in one second.  The researchers had not expected 
to send that much data until at least the year 2000.

In the paper, the group described a 1 Tb/s transmission experiment that 
utilized WDM [wavelength division multiplexing] and polarization multiplexing.

The outputs of 25 lasers were multiplexed using star couplers and waveguide 
grating routers.  The wavelengths ranged from 1542 nm (channel 1) to 1561.2 
nm (channel 25) with 100 GHz channel spacing.  All lasers were 
external-cavity lasers except for channel 16, which used a DFB laser.  Four 
of the laser outputs (channels 10,11,17, and 25) were amplified and filtered 
before multiplexing.  The multiplexed wavelengths were then amplified and 
propagatedthrough an polarization beamsplitter to align al the 
polarizations.  Polarization controllers at the output of each laser allowed 
independent polarization control for each source.

The 25 co-polarized wavelengths were split by a 3-dB coupler, separatedly 
modulated by LiNbO3 Mach-Zehnder modulators, and then recombined with 
orthogonal polarizations in a PBS.  The modulators have a small-signal 
bandwidth of 18 GHz and built-in polarizers.  The 20 Gb/s NRZ drive signals 
were produced by electronically multiplexing two 10-Gb/s 215-1 pseudorandom 
bit streams using a commercial GaAs multiplexer.

Two other groups from Japan, Fujitsu and Nippon Telephone and Telegraph Co., 
also submitted papers reporting that they reached the terabit mark.  All 
three groups achieved the record with different experiments.  

Scientists from NTT demonstrated 100 Gb/s x 10 channel (1 Tb/s), error-free 
transmission of all the 10 channels over a 40 km dispersion-shifted fiber 
using a low-noise single supercontinuum WDM source fitted with a newly 
developed arrayed-waveguide grating demultiplexer/multiplexer.  By fully 
utilizing the super-broad bandwidth of the SC spectra over 200 nm, up to 5 
Tb/s would be possible.

Fujitsu researchers achieved 1.1 Tb/s (55 wavelengths x 20 Gb/s) WDM 
transmission over 150 km of 1.3 mm [?] zero-dispersion singlemode fiber 
using preemphasis and dispersion compensating fiber with a negative 
dispersion slope.  BER [bit error rate] degradation was not observed in any 
channel, even without channel-by-channel dispersion adjustment.

[end of quoted portion]

Jim Bell
jimbell@pacifier.com