backward learning
Among the many different routing methods, a distinction is made between adaptive and non-adaptive methods. In non-adaptive or static routing, a fixed route is specified without regard to traffic density, congestion, or disruptions.
Backward learning is a simple adaptive routing algorithm. The information about the traffic situation is not obtained directly, but by evaluating the experience of the backward flowing traffic. For example, if network node A receives a data packet from network node "B" via intermediate node "C", then the routing algorithm in Baackward Learning assumes that node "B" is optimally reached from "A" via "C".