Algorithms for multiplex bundling in a telecommunications network

A two-phase approach based on heuristic programming is used to derive a solution to the problem of multiplex bundling. The first phase constructs a model that performs static edge rerouting to derive results initially for relaxed capacity constraints, followed by a zero to one multiconstraint knapsack feasibility problem. The results for this phase closely approximate the optimal solution. The second phase uses a single-stage algorithm based on dynamic programming to solve a multiperiod problem for phase one rerouting. An exact solution that minimizes state space is derived for the second stage. Heuristic programming is useful to data transmission systems to obtain fast and easily modifiable solutions to problems of optimization in multiperiod networks.