Efficient Algorithms for Sorting in Trees

TL;DR

This paper introduces a randomized algorithm for sorting tree-structured partial orders that improves query complexity bounds, and also presents a deterministic algorithm with better overall complexity than previous methods.

Contribution

It provides the first optimal randomized algorithm for bounded-degree trees and a new lower bound, advancing the understanding of sorting in tree posets.

Findings

Randomized algorithm achieves $O(dn\,\log n)$ complexity.

Improves previous bound of $O(dn\,\log^2 n)$.

First deterministic algorithm with lower total complexity for tree posets.

Abstract

Sorting is a foundational problem in computer science that is typically employed on sequences or total orders. More recently, a more general form of sorting on partially ordered sets (or posets), where some pairs of elements are incomparable, has been studied. General poset sorting algorithms have a lower-bound query complexity of $\Omega(wn + n \log n)$, where $w$ is the width of the poset. We consider the problem of sorting in trees, a particular case of partial orders, and parametrize the complexity with respect to $d$, the maximum degree of an element in the tree, as $d$ is usually much smaller than $w$ in trees. For example, in complete binary trees, $d = \Theta(1), w = \Theta(n)$. We present a randomized algorithm for sorting a tree poset in worst-case expected $O(dn\log n)$ query and time complexity. This improves the previous upper bound of $O(dn \log^2 n)$. Our algorithm is the first to be optimal for bounded-degree trees. We also provide a new lower bound of $\Omega(dn + n \log n)$ for the worst-case query complexity of sorting a tree poset. Finally, we present the first deterministic algorithm for sorting tree posets that has lower total complexity than existing algorithms for sorting general partial orders.