Quantum dice debut
January 14/21, 2004
By Eric Smalley, Technology Research News

Page One

Quantum dice debut

Researchers have overcome a major obstacle to generating random
numbers
on quantum computers by limiting the possibilities in the otherwise
unlimited randomness of a set of quantum particles.

Random numbers play a key role in classical computing by providing an
element of chance in games and simulations, a reliable method for encrypting
messages, and a means of accurately sampling huge amounts of data.

Researchers from the Massachusetts Institute of Technology and the
National Atomic Energy Commission in Argentina have shown that short
sequences of random operations -- randomly shifting laser pulses or magnetic
fields -- acting on a string of quantum bits can, in effect, generate random
configurations of qubits.

Being able to generate random numbers in quantum computing could make
quantum computers easier to build by countering the noise that eventually
destroys qubits, which represent the 1s and 0s of computer information.
Quantum computers promise to be fantastically fast at solving certain types
of large problems, including the mathematics that underpins today's security
codes.

Quantum random numbers could also be useful for increasing the
efficiency of quantum secret-sharing schemes, quantum encryption and various
forms of quantum communications.

Qubits can represent not only 1 and 0 but any number in between; a
string of 100 qubits can represent every possible 100-digit binary number,
and a single set of operations can search every possible answer to a problem
at once. This gives quantum computers their power, but also poses a problem
for generating random numbers. The nearly infinite number of possible qubit
configurations theoretically requires an impossibly large number of
calculations.

In the quantum world, no outcome is certain, and in most aspects of
quantum computing, the goal is to reduce the uncertainty in order to get a
definite answer to a problem. The researchers' scheme, however, aims for
uncertainty. It limits the possible outcomes without making them
predictable.

The scheme generates quantum states in such a way that the
probabilities of the limited set of outcomes are as evenly distributed over
the nearly infinite range of possible outcomes as quantum theory allows,
said Joseph Emerson, one of the MIT researchers who is now a fellow at the
Perimeter Institute for Theoretical Physics in Canada. "These pseudo-random
transformations are a practical substitute for truly... random
transformations," he said.

The number of operations required to represent a truly random
configuration increases exponentially with the number of qubits in the
configuration. For example, if the quantum equivalent of generating random
numbers takes 22, or four, operations for two qubits, 15 qubits would
require 215, or 32,768, operations.

The researchers' pseudo-random number method could be used to help
build quantum computers by providing a practical way to estimate
imperfections or errors in quantum processors, said Emerson. "This is
addressing a very big problem -- imperfections such as decoherence and
inadequate control of the coherence between the qubits are the main limiting
factors in the creation of large-scale quantum computers," he said.

A quantum particle decoheres, or is knocked out of its quantum state,
when it interacts with energy from the environment in the form of light,
heat, electricity or magnetism. Researchers are looking for ways to fend off
decoherence for as long as possible in order to make qubits last long enough
to be useful.

A way to estimate decoherence would allow researchers to assess the
strength and type of environmental noise limiting the precision of a given
quantum device, said Emerson. Random quantum operations can be used as
control operations that, when subjected to the noise affecting a prototype
quantum computer, will generate a response that depends only on the noise,
he said. This way the noise can be characterized with many fewer
measurements than existing methods, which are dependent on the interactions
of the qubits and so require a number of measurements that increases
exponentially with the number of qubits, he said.

In addition to helping build quantum computers, random operators would
be useful for quantum communications tasks like encryption, said Emerson.
"The idea is to randomize a specific configuration of qubits containing the
message, and then transmit this randomized state," he said.

In this case, if each bit that makes up the message is encrypted, or
changed randomly, it is not possible for an eavesdropper to find any type of
pattern that may lead to cracking the message.

The researchers tested the method on a three-qubit prototype liquid
nuclear magnetic resonance (NMR) quantum computer. The computer consists of
a liquid sample containing the amino acid alanine, which is a molecule made
of three carbon-13 atoms. The qubits are the atoms' spins, which are
analogous to a top spinning clockwise or counterclockwise. The two
directions, spin up and spin down, can be used to represent 1 and 0. The
qubits are controlled by magnetic fields generated by the nuclear magnetic
resonance device.

Being able to diagnose faulty quantum computer components in a way
that is independent of the number of qubits is very important, said Daniel
Lidar, an assistant professor of theoretical chemical physics at the
University of Toronto. "For this reason alone I suspect random [operators]
will find widespread applications as quantum computer benchmarking becomes
an experimental reality," he said.

It is also likely that future quantum algorithms will make increasing
use of pseudo-random operators, said Lidar.

The researchers are working on making the random-number-generation
system more precise, said Emerson. "Right now one can only estimate very
coarse properties of the noise, such as [its] overall strength," he said. "I
would like to devise methods to get a much more detailed analysis of the
noise operators."

Complete noise-estimation experiments could be implemented in
rudimentary quantum computers within the next few years, said Emerson.
Researchers generally agree that practical quantum computers are a decade or
two away.

Emerson's research colleagues were Yaakov S. Weinstein, Marcos
Saraceno, Seth Lloyd, and David G. Corey. The work appeared in the December
19, 2003 issue of Science. The research was funded by the National Science
Foundation (NSF), the Defense Advanced Research Projects Agency (DARPA) and
the Cambridge-MIT Institute.

Timeline: 2 years, 10-20 years
Funding: Government; University
TRN Categories: Quantum Computing and Communications; Physics
Story Type: News
Related Elements: Technical paper, "Pseudo-Random Unitary Operators
for Quantum Information Processing," Science, December 19, 2003
--
Do Wah Ditty

"Our country puts $1 billion a year up to help feed the hungry. And we're by
far the most generous nation in the world when it comes to that, and I'm
proud to report that. This isn't a contest of who's the most generous. I'm
just telling you as an aside. We're generous. We shouldn't be bragging about
it. But we are. We're very generous."

- Washinton D.C., July 16, 2003
President Bushisms