NLF:一种用于凸网络流均衡的电阻网络框架与线性时间求解器
NLF: A Resistor-Network Framework and Linear-Time Solver for Convex Network-Flow Equilibria
AI总结:
提出NLF框架,通过阻尼弦牛顿迭代与近线性拉普拉斯求解器,实现凸网络流均衡的线性时间求解,在单商品和多商品问题上显著优于内点法和L-BFGS。
摘要:
我们提出了NLF(非线性拉普拉斯流),一个用于凸网络流均衡的统一框架和线性时间求解器。拥塞路由、最小延迟路由和最大流共享一种形式:非线性图拉普拉斯算子 $B\rho(B^T\phi)=\alpha d$,其中单调边律 $\rho_e$ 编码了物理规律(无向图;有向变体是未来工作)。NLF通过阻尼弦牛顿迭代求解,其冻结线性化——一个加权图拉普拉斯算子——通过近线性拉普拉斯求解器(默认:近似Cholesky,LAMG+可互换)求逆。非线性求解需要 $2$--$4$ 次线性拉普拉斯求解,使得实际运行时间在边数 $m$ 上经验地为 $O(m)$(非证明界)。在单商品拥塞(BPR成本)问题上,NLF在所有2003个SuiteSparse语料库图(最多 $1.8\times10^7$ 条边)上收敛。与最先进的内点法相比,NLF在两者都收敛的情况下中位数快 $2.6\times$,在IPM的直接核心为超线性的难分离图上快 $>45\times$;与L-BFGS相比,中位数快 $4.2\times$,并且是唯一在90个最难实例上完成的求解器。多商品扩展通过一个共享层次结构路由 $K$ 个商品,每步复杂度 $O(Km)$。同样的机制通过一系列拉普拉斯求解恢复精确最大流,并附带割势作为副产品。代码:此 https URL
英文摘要:
We present NLF (Nonlinear Laplacian Flow), a unified framework and linear-time solver for convex network-flow equilibria. Congestion routing, minimum-delay routing, and maximum flow share one form: the nonlinear graph Laplacian $Bρ(B^Tϕ)=αd$, where a monotone edge law $ρ_e$ encodes the physics (undirected graphs; directed variants are future work). NLF solves it by a damped chord-Newton iteration whose frozen linearization -- a weighted graph Laplacian -- is inverted by a near-linear Laplacian solver (default: approximate Cholesky, LAMG+ interchangeable). The nonlinear solve costs $2$--$4$ linear Laplacian solves, making the wall-clock empirically $O(m)$ in the edge count $m$ (not a proved bound). On single-commodity congestion (BPR cost), NLF converges on all 2,003 SuiteSparse corpus graphs up to $1.8\times10^7$ edges. Against a state-of-the-art interior-point method, NLF is a median $2.6\times$ faster where both converge and $>45\times$ on poorly-separable graphs where the IPM's direct core is superlinear; against L-BFGS, a median $4.2\times$ faster and the only solver to finish on the 90 hardest instances. A multicommodity extension routes $K$ commodities through one shared hierarchy at $O(Km)$ per step. The same machinery recovers the exact max-flow as a short sequence of Laplacian solves, with the cut potential as a by-product. Code: https://github.com/orenlivne/nlf
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。