Methods for Flow Graph Selection in Integral Network Coding

Abstract

The search for optimal multicast subgraphs for network coding is considered. We assume unit link capacities and binary flow rates. In the first version of the problem, there is no constraint on the acyclicity of the subgraphs, whereas such constraints are imposed in the second version. These problems are known to be NP-hard. We provide heuristics to deal with both versions of the problem. The heuristics are based on well known optimization algorithms and they are therefore easy to implement. Through numerical experiments, we demonstrate that the proposed algorithm performs close to the lower bounds obtained from relaxed solutions.