Project Details
Projekt Print View

Long cycles of spatial random permutations: Size, geometry and dynamics

Subject Area Mathematics
Term from 2014 to 2019
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 255533188
 
Final Report Year 2019

Final Report Abstract

Spatial random permutations (SPR) are a model that is simple to describe in plain words, but difficult to analyse. A special version of SRP is as follows: imagine a cube-shaped subset of Zd with side length L and consider the set of all permutations on it with the property that each point in ΛL is either mapped to itself or to one of its neighbours. Let n(π) be the number of fixed points (points that are mapped to themselves) under a permutation π. The model consists in choosing a random permutation from the set of all possible permutations, and the law of this randomness should only depend on n(π). More precisely, the probability of choosing a certain permutation π should be proportional to eαn(π) . Thus, the parameter α regulates how likely permutations with many fixed points are. The interest in such models comes mainly from their connection with the theory of Bose-Einstein condensation. In this context, the important question is the existence of long cycles: when L is very large, is it true or not that with reasonably high probability, a randomly sampled permutation will have cycles that contain a fraction (e.g. more than 1/2, or more than 1/100) of the available points)? Does this answer depend on the parameter α? This is a very difficult question to answer rigorously, although there is overwhelming numerical evidence for the existence of some α0 such that for α > α0 , no long cycles exist, while for α < α0 , long cycles exist. In this project, we made the first steps towards solving this question in a special case. One of the most important outcomes is a paper where, using a method (’reflection positivity’) that had to be adapted to the current situation, it was shown that SRP indeed have long cycles in the (degenerate) case α = −∞, i.e. in situations where we do not allow fixed points. This result may seem quite special, but it is an important first step towards understanding more general models of SRP.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung