Astropulse: A Search for Microsecond Transient Radio Signals Using Distributed Computing. I. Methodology,” submitted to Astrophys

We use the Arecibo telescope to search for brief, wideband radio pulses on timescales of microseconds to milliseconds. Such pulses could come from explod-ing primordial black holes, RRATs (rotating radio transients), pulsars, or new phenomena yet to be discovered. If black holes explode via the Hawking pro-cess, the resulting fireball might lead to a single EMP once the black hole has dwindled to a mass of 1011 g. We observe a bandwidth of 2.5 MHz centered on 1420 MHz, and we can detect single pulses of strength 36 Jy µs at a 0.4 µs time resolution. When the survey is finished, we will be able to establish a limit of 3 · 10−13 pc−3 yr−1 on 0.4µs events with energy 1032 erg. We will also have access to coherent records of our results on a 0.4µs time scale, whereas other surveys take data incoherently, typically on a 64µs scale or longer. Our data comes from a seven-year and continuing commensal sky survey, of which the most recent data uses Arecibo’s new 7 beam receiver. The survey covers 30 % of the whole sky and comprises 100 terabytes of data. We search for both single pulses and regularly repeating pulses, with dispersion measures ranging from 50pc cm−3 to 800pc cm−3. Because our time resolution is better than 50 µs, we must use coher-ent dedispersion, which is computationally intensive, requiring 1020 floating point operations in total for the project. We utilize the distributed computing power o