English

On the randomised query complexity of composition

Computational Complexity 2022-04-05 v1

Abstract

Let f{0,1}n×Ξf\subseteq\{0,1\}^n\times\Xi be a relation and g:{0,1}m{0,1,}g:\{0,1\}^m\to\{0,1,*\} be a promise function. This work investigates the randomised query complexity of the relation fgn{0,1}mn×Ξf\circ g^n\subseteq\{0,1\}^{m\cdot n}\times\Xi, which can be viewed as one of the most general cases of composition in the query model (letting gg be a relation seems to result in a rather unnatural definition of fgnf\circ g^n). We show that for every such ff and gg, R(fgn)Ω(R(f)R(g)),\mathcal R(f\circ g^n) \in \Omega(\mathcal R(f)\cdot\sqrt{\mathcal R(g)}), where R\mathcal R denotes the randomised query complexity. On the other hand, we demonstrate a relation f0f_0 and a promise function g0g_0, such that R(f0)Θ(n)\mathcal R(f_0)\in\Theta(\sqrt n), R(g0)Θ(n)\mathcal R(g_0)\in\Theta(n) and R(f0g0n)Θ(n)\mathcal R(f_0\circ g_0^n)\in\Theta(n) - that is, our composition statement is tight. To the best of our knowledge, there was no known composition theorem for the randomised query complexity of relations or promise functions (and for the special case of total functions our lower bound gives multiplicative improvement of logn\sqrt{\log n}).

Cite

@article{arxiv.1801.02226,
  title  = {On the randomised query complexity of composition},
  author = {Dmytro Gavinsky and Troy Lee and Miklos Santha},
  journal= {arXiv preprint arXiv:1801.02226},
  year   = {2022}
}
R2 v1 2026-06-22T23:38:40.093Z