Network connectivity buildup by adaptive learning

A method for sub-optimal traffic routing in a wireless network (or any other unreliable network infrastructure) by means of adaptive learning is presented. It is aimed to allow data transfer, using the terminals themselves as relays, without network manager interference. Routing decisions are based on the acquisition of parameterized knowledge that encodes a limited view of the network connectivity as seen from each of the terminals. The research was restricted to the master/slave (MS) control approach. The method allows one to spread a wireless network without a-priori knowledge of the connectivity and topology of the network, but all the same allow the network to operate.