Development of efficient network structures for freight transport is fundamental in the current global market. Demands need to be fluid quickly, meet customer needs in shortest possible time, congestions and delays need to be minimized, CO2 emissions have to be controlled and lower transport costs must to be offered to customers. Hub-and-spoke structure is actual network model used both in regional transport as intercontinental, offering an economy of scale for aggregated demands in hub nodes. But, delays, congestions and long delivery time are drawbacks of this network. In this thesis, a new concept, “sub-hub”, is added to classic hub-and-spoke network. In the proposed network models, economy of scale and shorter alternative paths are implemented, minimizing the transport cost and delivery time. Sub-hub is a connection point between two routes from distinct and close regions. Transhipments without passing through hub nodes are possible in sub-hubs. Congestions are avoided and consequently delays are thus minimized. Four binary integer linear programming models of hub location and routing problem are developed in this thesis. Networks with sub-hub and networks without sub-hub taking into account circular hub routes or direct connections between hubs are compared. These models are composed of four sub-problems (location, allocation, service design and routing) that makes complex the solving. A cutting plane approach is tested to solve small instances of problem whereas a Variable Neighborhood Decomposition Search (VNDS) composed of exact methods (matheuristic) has been developed to solve large instances. Implemented VNDS explores each sub-problem by different operators. Important gains in objective function are provided by models with sub-hub thus confirming the development of more competitive networks.