A Genetic-Neural Approach for Mobility Assisted Routing in a Mobile Encounter Network

Mobility Assisted Routing (MAR) is a concept, where the mobility of a network's nodes is used to physically carry data to its destination. Traditionally, MAR algorithms have been based on few simple rules, often limiting the performance of these algorithms. This paper proposes an architecture in which a trained neural network is fed information about the message and the encountered peer, and which then decides whether to forward the message to the encountered peer. This algorithm, called NeuroRouter, is capable of utilizing the most efficient routing strategies in different environments by adapting its behavior based on environmental variables.