Computing capacity bounds of ad hoc networks, as well as optimizing radio resource allocation in order to approach those bounds is a hard task that have recently attracted the interest of the research community. In this paper we address the problem of jointly optimizing the routing of flows, the transmission scheduling over wireless links along the path, and the emitted power for bandwidth guaranteed traffic demands. We propose a mixed integer linear programming model which considers the signal to interference and noise ratio at receiver and accounts for the effect of adaptive transmission rate through a set of discrete link capacity and power values. To find the optimal solution of the proposed problem we provide an algorithm based on column generation. Reported numerical results on networks with up to 50 nodes and 170 links show the effectiveness of the proposed approach even with quite bit problem instances.