Focusing on the joint relay selection and power control problem with a view to maximizing the sum-rate, we propose a novel sub-optimal algorithm that iterates between relay selection and power control. The relay selection is performed by maximizing the minimum signal-to-interference-plus-noise-ratio (as opposed to maximizing the sum-rate) and the power control is performed using a successive convex approximation. By comparing the proposed algorithm with existing solutions via extensive simulations, we show that the proposed algorithm results in significant sum-rate gains. Finally, we analyze the two-user multi-hop network and show that optimum transmit power of at least for two transmitting nodes can be found using binary power allocation.