Title: Structured $(\min,+)$-Convolution And Its Applications For The Shortest Vector, Closest Vector, and Separable Nonlinear Knapsack Problems

Abstract: In this work we consider the problem of computing the $(\min, +)$-convolution of two sequences $a$ and $b$ of lengths $n$ and $m$, respectively, where $n \geq m$. We assume that $a$ is arbitrary, but $b_i = f(i)$, where $f(x) \colon [0,m) \to \mathbb{R}$ is a function with one of the following properties:
1. the linear case, when $f(x) =\beta + \alpha \cdot x$;
2. the monotone case, when $f(i+1) \geq f(i)$, for any $i$;
3. the convex case, when $f(i+1) - f(i) \geq f(i) - f(i-1)$, for any $i$;
4. the concave case, when $f(i+1) - f(i) \leq f(i) - f(i-1)$, for any $i$;
5. the piece-wise linear case, when $f(x)$ consist of $p$ linear pieces;
6. the polynomial case, when $f \in \mathbb{Z}^d[x]$, for some fixed $d$.
To the best of our knowledge, the cases 4-6 were not considered in literature before. We develop true sub-quadratic algorithms for them.
We apply our results to the knapsack problem with a separable nonlinear objective function, shortest lattice vector, and closest lattice vector problems.