The downlink capacity of a unicast network with a large number of users than the base station transmit antennas depends on user selection and interference among the selected users. Various suboptimal selection schemes in combination with suboptimal or optimal precoding have been proposed in the literature, and some of these techniques asymptotically achieve the sum capacity of DPC, as the number of users goes to infinity. In the previous works, the joint design problem is addressed as a decoupled problem of selection and precoding either at the design level or the solution level. In this work, we focus on the design of a joint solution to the joint design problem, which has the potential to achieve better sum capacity than the aforementioned designs. First, we develop a framework to formulate the joint design problem, with the objective of maximizing Weighted sum rate (WSR). Further, we reformulate the resulting NP-hard WSR maximization problem as a difference-of-concave (DC) functions with novel rate reformulations. Finally, the DC maximization problem is solved using the convex-concave procedure, which is guaranteed to converge to a stationary point. Through simulations, we show the efficacy of the proposed solution over the state-of-the-art designs.