218

Tight local approximation results for max-min linear programs

Algorithmic Aspects of Wireless Sensor Networks (AAWSN), 2008
Abstract

In a bipartite max-min LP, we are given a bipartite graph \myG=(VIK,E)\myG = (V \cup I \cup K, E), where each agent vVv \in V is adjacent to exactly one constraint iIi \in I and exactly one objective kKk \in K. Each agent vv controls a variable xvx_v. For each iIi \in I we have a nonnegative linear constraint on the variables of adjacent agents. For each kKk \in K we have a nonnegative linear objective function of the variables of adjacent agents. The task is to maximise the minimum of the objective functions. We study local algorithms where each agent vv must choose xvx_v based on input within its constant-radius neighbourhood in \myG\myG. We show that for every ϵ>0\epsilon>0 there exists a local algorithm achieving the approximation ratio ΔI(11/ΔK)+ϵ{\Delta_I (1 - 1/\Delta_K)} + \epsilon. We also show that this result is the best possible -- no local algorithm can achieve the approximation ratio ΔI(11/ΔK){\Delta_I (1 - 1/\Delta_K)}. Here ΔI\Delta_I is the maximum degree of a vertex iIi \in I, and ΔK\Delta_K is the maximum degree of a vertex kKk \in K. As a methodological contribution, we introduce the technique of graph unfolding for the design of local approximation algorithms.

View on arXiv
Comments on this paper