In this paper we study optimum routing strategies that minimize the degrading effects of system noise in noisy communication networks. Previous papers have treated problems related to finding an optimum single communication path through a network; we consider the problem of establishing several communication paths simultaneously. Networks with fixed link characteristics and networks with links that may adapt their number of channels to traffic requirements are considered. Primary emphasis is on digital communication networks, but it is also pointed out that the same techniques can be used for a certain class of linear, continuous signal networks. Using the concepts of information measure from information theory, we discuss the maximization of information transfer in a communication network. We also treat the problem of finding the maximum number of channel paths with signal transmission capabilities better than some lower bound, and the problem of finding

channel paths that minimize the use of the poorest paths.