TEAM THEORY AND NEURAL NETWORKS FOR DYNAMIC ROUTING IN TRAFFIC AND COMMUNICATION-NETWORKS

The dynamic-routing problem in traffic and communication networks is a ddressed. Routing nodes must accomplish the following tasks: (i) gener ate routing decisions on the basis of local information (i.e. the cont ents of their queues) and possibly of some messages received from neig hbouring nodes, and (ii) compute (or adapt) their routing strategies b y measuring local variables and exchanging a small number of messages with neighbouring nodes. The first task leads to regard routing nodes as the cooperating decision-makers of a team organization. The second task calls for a computationally distributed algorithm. Such tasks and the impossibility of solving team functional optimization problems un der general conditions suggest that each routing node be assigned a se t of multilayer feedforward neural networks able to generate routing d ecisions. The weights of such neural networks are then adjusted by mea ns of a gradient algorithm based on backpropagation.