p_0 \approx 2.1397$, there no $2^{n/C_p}$-time algorithm for $\SVP_p$ for some explicit constant $C_p > 0$ unless the (randomized) Strong Exponential Time Hypothesis (SETH) is false. $\bullet$ For any $p > 2$, there is no $2^{o(n)}$-time algorithm for $\SVP_p$ unless the (randomized) Gap-Exponential Time Hypothesis (Gap-ETH) is false. Furthermore, for each $p > 2$, there exists a constant $γ_p > 1$ such that the same result holds even for $γ_p$-approximate $\SVP_p$. $\bullet$ There is no $2^{o(n)}$-time algorithm for $\SVP_p$ for any $1 \leq p \leq 2$ unless either (1) (non-uniform) Gap-ETH is false; or (2) there is no family of lattices with exponential kissing number in the $\ell_2$ norm. Furthermore, for each $1 \leq p \leq 2$, there exists a constant $γ_p > 1$ such that the same result holds even for $γ_p$-approximate $\SVP_p$.">

(Gap/S)ETH Hardness of SVP (original) (raw)

View PDF

Abstract:$ \newcommand{\problem}[1]{\ensuremath{\mathrm{#1}} } \newcommand{\SVP}{\problem{SVP}} \newcommand{\ensuremath}[1]{#1} WeprovethefollowingquantitativehardnessresultsfortheShortestVectorProblemintheWe prove the following quantitative hardness results for the Shortest Vector Problem in the WeprovethefollowingquantitativehardnessresultsfortheShortestVectorProbleminthe\ell_p$ norm ($\SVP_p$), where nnn is the rank of the input lattice.
bullet\bulletbullet For "almost all" p>p0approx2.1397p > p_0 \approx 2.1397p>p0approx2.1397, there no 2n/Cp2^{n/C_p}2n/Cp-time algorithm for SVPp\SVP_pSVPp for some explicit constant Cp>0C_p > 0Cp>0 unless the (randomized) Strong Exponential Time Hypothesis (SETH) is false.
bullet\bulletbullet For any p>2p > 2p>2, there is no 2o(n)2^{o(n)}2o(n)-time algorithm for SVPp\SVP_pSVPp unless the (randomized) Gap-Exponential Time Hypothesis (Gap-ETH) is false. Furthermore, for each p>2p > 2p>2, there exists a constant gammap>1\gamma_p > 1gammap>1 such that the same result holds even for gammap\gamma_pgammap-approximate SVPp\SVP_pSVPp.
bullet\bulletbullet There is no 2o(n)2^{o(n)}2o(n)-time algorithm for SVPp\SVP_pSVPp for any 1leqpleq21 \leq p \leq 21leqpleq2 unless either (1) (non-uniform) Gap-ETH is false; or (2) there is no family of lattices with exponential kissing number in the ell2\ell_2ell2 norm. Furthermore, for each 1leqpleq21 \leq p \leq 21leqpleq2, there exists a constant gammap>1\gamma_p > 1gammap>1 such that the same result holds even for gammap\gamma_pgammap-approximate SVPp\SVP_pSVPp.

Submission history

From: Noah Stephens-Davidowitz [view email]
[v1] Mon, 4 Dec 2017 07:50:40 UTC (60 KB)