A Code-Breaking Method Found a Pulsar Hiding in 15 Years of NASA Data

A team of astronomers has found a pulsar in NASA's public gamma-ray archive that the previous best search method had failed to find, using an approach borrowed from code-breaking mathematics. PSR J1622-7204, an isolated neutron star spinning 261 times a second, turned up in 15 years of data from NASA's Fermi Gamma-ray Space Telescope after about ten CPU hours.

The authors, writing in Physical Review Letters, say the method may turn up many more gamma-ray pulsars, including ones orbiting a companion star, a case they describe as out of reach until now unless the orbit was already known.
Nothing new was observed — the data have been public for years, and what changed is the search. Finding a gamma-ray pulsar means guessing its spin rate, how fast that spin is slowing, and where it sits in the sky, all at once. "We have a four-dimensional grid, and we must search each direction with millions of trials in it," said lead author Daniel Pook-Kolb of the Max Planck Institute for Gravitational Physics.
Dotan Gazith and Barak Zackay, both at the Weizmann Institute of Science, recast that search as a geometry problem: finding the shortest vector in a high-dimensional lattice, a regularly spaced array of points. Those algorithms come from cryptographers trying to break encryption designed to resist quantum computers. "We have nothing to do with cryptography in our analysis here, but those algorithms can help us solve the problem of searching for pulsars in our data," Gazith said in NASA's release.
The paper reports that the earlier attempt ran for roughly 600 CPU years without finding it. NASA's release puts the gain at about a million times faster than before, and notes that conventional searches, run on computing time donated by volunteers, have typically needed the equivalent of hundreds of computer-years to find one pulsar.

Colin Clark, a co-author at the Max Planck Institute, said tens or perhaps hundreds of pulsars are probably still hidden in the same data, and that the team will next run the algorithm on the volunteer computers of the Einstein@Home project.
Sources
- Physical Review LettersPeer-reviewed
- science.nasa.gov
