Detecting collisions in sets of moving particles: a survey and some experiments

Detecting and responding to collisions between particles is an important requirement for building simulations in computational science. Due to the large number of potential collisions it is impractical to check all possibilities, so the development of algorithms which narrow down the number of possible searches to a small number is important. In this paper we review various algorithms for this task, and give results from a number of experiments which demonstrate the relative efficiency of these algorithms on a fundamental problem of detecting collisions between particles undergoing Brownian motion. The general slant of the paper is towards the development of algorithms for simulating microbiological systems.