Learning-based strategies for message-initiated constraint-based routing

摘要

A method is presented for a learning-based strategy utilized within message-initiated constraint-based routing for digital message communication among nodes in an ad-hoc network, in which each node includes attributes. The method includes determining local attributes for each of the nodes and defining constraints on the attributes. Each node is provided access to the attributes of each neighboring node. Each message transmitted over the network has a message type, which includes a destination specification, route specification, and objective specification. Constraint checking and cost estimation checking are performed for each message type. Cost estimation is utilized to converge on an optimal message path. The message that is routed within the network includes the address of a sending node, the address of the source node, route constraints, destination constraints, the number of route constraints, the number of destination constraints, message identification number, sequence identification number, and routing objectives.