Local routing algorithms based on Potts neural networks

A feedback neural approach to static communication routing in asymmetric ne tworks is presented, where a mean held formulation of the Bellman-Ford meth od for the single unicast problem is used as a common platform for developi ng algorithms for multiple unicast, multicast and multiple multicast proble ms. The appealing locality and update philosophy of the Bellman-Ford algori thm is inherited, For all problem types the objective is to minimize a tota l connection cost, defined as the sum of the individual costs of the involv ed arcs, subject to capacity constraints, The methods are evaluated for syn thetic problem instances by comparing to exact solutions for cases where th ese are accessible, and else with approximate results from simple heuristic s. In general, the quality of the results are better than those of the heur istics. Furthermore, the computational demands are modest, even when the di stributed nature of the the approach is unexploited numerically.