Improved Bi-criteria Approximation for the All-or-Nothing Multicommodity Flow Problem in Arbitrary Networks

This paper addresses the following fundamental maximum throughput routing problem: Given an arbitrary edge-capacitated $n$-node directed network and a set of $k$ commodities, with source-destination pairs $(s_i,t_i)$ and demands $d_i> 0$, admit and route the largest possible number of commodities -- i.e., the maximum {\em throughput} -- to satisfy their demands. The main contributions of this paper are two-fold: First, we present a bi-criteria approximation algorithm for this all-or-nothing multicommodity flow (ANF) problem. Our algorithm is the first to achieve a {\em constant approximation of the maximum throughput} with an {\em edge capacity violation ratio that is at most logarithmic in $n$}, with high probability. Our approach is based on a version of randomized rounding that keeps splittable flows, rather than approximating those via a non-splittable path for each commodity: This allows our approach to work for {\em arbitrary directed edge-capacitated graphs}, unlike most of the prior work on the ANF problem. Our algorithm also works if we consider the weighted throughput, where the benefit gained by fully satisfying the demand for commodity $i$ is determined by a given weight $w_i>0$. Second, we present a derandomization of our algorithm that maintains the same approximation bounds, using novel pessimistic estimators for Bernstein's inequality. In addition, we show how our framework can be adapted to achieve a polylogarithmic fraction of the maximum throughput while maintaining a constant edge capacity violation, if the network capacity is large enough. One important aspect of our randomized and derandomized algorithms is their {\em simplicity}, which lends to efficient implementations in practice.