arXiv subjects

Kaito Harada

Publications and source records attributed to Kaito Harada.

3 recordsLinked to original sources

arXiv2026

A Simple Algorithm for the Directed Multiple Source Replacement Paths Problem

In the replacement paths (RP) problem, we are given a graph $G = (V, E)$ with $n = |V|$ and $m = |E|$, together with two vertices $s, t \in V$, and are asked to compute the shortest-path distance from $s$ to $t$ in $G \setminus e$ for every failed edge $e \in E$. The multiple source replacement paths (MSRP) problem is its natural generalization: given a set $S \subseteq V$ of $σ$ sources, compute the replacement path distances for all pairs in $S \times V$. In this paper, we present a randomized combinatorial algorithm that solves MSRP on unweighted directed graphs in $\tilde{O}(m\sqrt{σn} + σn^2)$ time, with all the output distances correct with high probability. This improves the best known bound $\tilde{O}(m\min\{σ\sqrt{n}, n\} + σn^2)$ for directed graphs, which is obtained either by running the single source RP algorithm of Chechik and Magen [ICALP'20] from each source separately or by constructing and querying the all-pairs distance sensitivity oracle of Bernstein and Karger [STOC'09]. Our running time is essentially tight among combinatorial algorithms because Gupta, Jain, and Modi [PODC'20] proved a lower bound of $m{(σn)}^{1/2-o(1)}$ for such algorithms, which holds even on undirected graphs, and the additive term $σn^2$ is proportional to the time needed to write down the $Θ(σn^2)$ output distances. The algorithm is also remarkably simple.

cs.DS↗

arXiv2026

Constant Factor Optimal 2-Resilient Local Failover Routing Scheme on Directed Graphs

Local failover routing delivers a packet from a source $s$ to a destination $t$ using only pre-computed, per-node forwarding rules together with a small rewritable packet header, remaining correct even when up to $k$ arcs of the network fail. We study this problem on directed graphs with $n$ nodes, measuring efficiency by the number of rewritable bits carried in the header, and focus on the case of $k \ge 2$ failures. Our first contribution is an improved upper bound for general $k$. We design a scheme using at most $\lceil \log\!\binom{2n+k-4}{k-1} \rceil+1$ bits, improving the best previously known $\lceil \log \binom{2n+k-3}{k} \rceil$ bit scheme for every $k < 2n-3$. For the special case $k=2$, we further improve this to a scheme using at most $\left\lceil \log\!\left(1+2\lfloor 2(n-1)/3 \rfloor\right)\right\rceil$ bits. Our second contribution is a family of improved lower bounds, obtained via a new path-gadget construction. For $k=2$, we prove a lower bound of $\left\lceil \log \lfloor n/3 \rfloor \right\rceil$ bits, and for general $k$, writing $k_2 = \lfloor k/2 \rfloor$, a lower bound of $\left\lceil k_2 \log \left\lfloor (n-1+k_2)/(3k_2) \right\rfloor \right\rceil$ bits, improving on the best previously known bound. In particular, for $k=2$, our upper bound of $\left\lceil \log\!\left(1+2\lfloor 2(n-1)/3 \rfloor\right)\right\rceil$ bits and our lower bound of $\left\lceil \log \lfloor n/3 \rfloor \right\rceil$ bits differ by at most $3$ bits.

cs.DS↗

arXiv2024

A Nearly Linear Time Construction of Approximate Single-Source Distance Sensitivity Oracles

An \emph{$α$-approximate vertex fault-tolerant distance sensitivity oracle} (\emph{$α$-VSDO}) for a weighted input graph $G=(V, E, w)$ and a source vertex $s \in V$ is the data structure answering an $α$-approximate distance from $s$ to $t$ in $G-x$ for any given query $(x, t) \in V \times V$. It is a data structure version of the so-called single-source replacement path problem (SSRP). In this paper, we present a new \emph{nearly linear-time} algorithm of constructing a $(1 + ε)$-VSDO for any directed input graph with polynomially bounded integer edge weights. More precisely, the presented oracle attains $\tilde{O}(m \log (nW)/ ε+ n \log^2 (nW)/ε^2)$ construction time, $\tilde{O}(n \log (nW) / ε)$ size, and $\tilde{O}(1/ε)$ query time, where $n$ is the number of vertices, $m$ is the number of edges, and $W$ is the maximum edge weight. These bounds are all optimal up to polylogarithmic factors. To the best of our knowledge, this is the first non-trivial algorithm for SSRP/VSDO beating $\tilde{O}(mn)$ computation time for directed graphs with general edge weight functions, and also the first nearly linear-time construction breaking approximation factor 3. Such a construction has been unknown even for undirected and unweighted graphs. In addition, our result implies that the known conditional lower bounds for the exact SSRP computation does not apply to the case of approximation.

cs.DS↗