We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.DS

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Data Structures and Algorithms

Title: The Search Efficiency of Intermittent Lévy walks Optimally Scales with Target Size

Abstract: L\'evy walks are random walks processes whose step length follows a long-tailed power law distribution. Due to their abundance as movement patterns of biological organisms, significant theoretical efforts have been devoted to identify the foraging circumstances that would make such patterns advantageous [Viswanathan et al. Nature, 1999]. Despite numerous attempts, there is currently no conclusive analytical evidence that L\'evy flights are preferable strategies in higher dimensions than one. Here we show that the optimality of inverse-square L\'evy walks in two-dimensions becomes striking when targets are sparse and unpredictable in size, and when detection is weak. Specifically, we prove that under the intermittent model, in which detection is not possible while moving ballistically [B\'enichou et al. Reviews of Modern Physics, 2011], this strategy optimally finds sparse targets of any size and shape. That is, in a square torus of area~$n$, and assuming that the detection radius is normalized to $1$, the strategy finds any connected set of diameter $D$ in $\tilde{O}(n/D)$ expected time, whereas $\Omega(n/D)$ is an unconditional lower bound on the expected time, that holds even when assuming that the shape and size of the target are known. Furthermore, this particular L\'evy process stands in stark contrast to many other basic intermittent processes, including all other L\'evy walks, whose competitive ratio is shown to be polynomial in $n$, for wide ranges of target scales. Our results thus provide strong theoretical support for the optimality and robustness of intermittent L\'evy walks under general conditions.
Subjects: Data Structures and Algorithms (cs.DS); Quantitative Methods (q-bio.QM)
Cite as: arXiv:2003.13041 [cs.DS]
  (or arXiv:2003.13041v2 [cs.DS] for this version)

Submission history

From: Amos Korman [view email]
[v1] Sun, 29 Mar 2020 14:45:02 GMT (116kb,D)
[v2] Thu, 9 Apr 2020 10:14:36 GMT (117kb,D)

Link back to: arXiv, form interface, contact.