Joint optimization of scheduling and multicast trees by column-generation

In a wireless network, transmissions from the various nodes have to be scheduled so as to avoid mutual interference. The pattern of interference induced by active transmissions depends on the routes along which the link-transmissions have to be scheduled; i.e., the interference that scheduling has to accommodate depends on the routing. Both routing and scheduling are mechanisms to promote the efficient use of network capacity, and in view of their interdependence, it is important to consider their joint optimization. Considering one without the other can create limitations for each function and is non-optimal. We develop an algorithm based on the column-generation technique of Linear Programming for the joint optimization of routing and scheduling for multicast flows for maximizing network capacity, and demonstrate the benefit of the joint optimization in increased capacity over the case where the routing and scheduling are separately considered.