A decomposition algorithm to determine the maximum flow in a generalized network

Generalized networks are characterized by the presence of positive arc multipliers not all equal to one. Generalized network flow algorithms are known to have high computational complexity. This paper presents a procedure which decomposes the network into smaller networks and solves the maximum flow problem on the subnetworks. This procedure enables us to handle large networks which can not be solved using the available computers in the work place. Maximum flows for pure networks can also be determined by setting all arc multipliers equal to one.