Where does routing collide?

In operations research and logistics, routing is the problem of designing least-cost routes for vehicles that must serve a set of demands and return to a depot, balancing distance, time, capacity, and the number of vehicles. Dantzig and Ramser's 1959 The truck dispatching problem gave the first mathematical formulation of what became the vehicle routing problem. The referent is the optimisation of physical delivery tours over a road network, measured in total distance or cost; it is a close cousin of how data packets are routed (we flag that kinship as our reading, not the sources') but a different thing from a factory's fixed route sheet.[1]

Unit

Total travel cost (distance or time) of the vehicle tours that serve all demands

Recovery means

Not applicable

Measured by

Evaluated by total route cost, number of vehicles, and distance or time against the optimal or best-known solution

In communication networks, routing is the process of choosing the path a message or packet takes from source to destination across intermediate nodes, and of surviving the failure of any node or link along the way. Baran's 1964 On distributed communications networks argued for distributed, redundant routing in which messages find their own way, a foundation of packet switching. The referent is path selection for information through a network, measured in hops, delay, and robustness; it shares the directing-flow-through-a-network idea with vehicle routing (that kinship is our reading) but is applied to data rather than trucks.[2]

Unit

The path (and its cost in hops or delay) assigned to each packet across the network

Recovery means

Routing reconverges after a failure as the protocol discovers new paths around the broken node or link

Measured by

Tracked by path length (hops), delay, delivery ratio, and convergence time after a topology change