Why locally-fair maximal flows in client-server networks perform well

  • Authors:
  • Kenneth A. Berman;Chad Yoshikawa

  • Affiliations:
  • Computer Science Department, University of Cincinnati, Cincinnati, USA 45221;Computer Science Department, University of Cincinnati, Cincinnati, USA 45221

  • Venue:
  • Journal of Combinatorial Optimization
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Maximal flows reach at least a 1/2 approximation of the maximum flow in client-server networks. By adding 1 additional time round to any distributed maximal flow algorithm we show how this 1/2-approximation can be improved on bounded-degree networks. We call these modified maximal flows `locally fair' since there is a measure of fairness prescribed to each client and server in the network. Let N=(U,V,E,b) represent a client-server network with clients U, servers V, network links E, and node capacities b, where we assume that each capacity is at least one unit. Let d(u) denote the b-weighted degree of any node u驴U驴V, Δ=max驴{d(u)|u驴U} and 驴=min驴{d(v)|v驴V}. We show that a locally-fair maximal flow f achieves an approximation to the maximum flow of $\min\{1,\frac{\varDelta^{2}-\delta}{2\varDelta^{2}-\delta\varDelta-\varDelta}$ }, and this result is sharp for any given integers 驴 and Δ. This results are of practical importance since local-fairness loosely models the steady-state behavior of TCP/IP and these types of degree-bounds often occur naturally (or are easy to enforce) in real client-server systems.