Next generation tactical military network will be based on mobile ad hoc networks (MANET). These networks require efficient spatial channel reuse in order to provide high spectral efficiency and this is only achieved by efficient channel assignment optimization. For a clustered network topology the basic goal is to assign different channels to adjacent clusters, i.e. a graph coloring problem. Unfortunately, is the optimal solution for graph coloring problems intractable, the problem is NP-hard. As a consequence heuristic methods must be applied, which provide solutions with as close to optimal result as possible. In this article the grouping genetic algorithm is applied for solving the channel assignment problem in a cluster based mobile ad hoc network. The used multi objective function minimizes interference and maximizes the spectral efficiency.