Distributed Solution of Network Programming Problems

A method is presented for the solution of network programming problems through the iterative execution of the same algorithm for each station of the network. At a given station, the algorithm requires data only from adjacent stations of the network. The theory of the method is developed through the use of electrical models of network programming problems. The techniques should be useful for the communications message-routing problem and similar cases in which the disadvantages of relying on a computation performed at a master station must be avoided.