We consider the problem of sum rate maximization with joint resource allocation and interference mitigation by multiantenna processing in wireless networks. The denominators in the users' signal-to-interference-plus-noise expressions are assumed to be representable in the form of matrix-based, concave interference functions. It is shown that the problem of interest for this system model can be readily rewritten as a minimization of a difference of convex functions. Based on this representation, an iterative algorithm with guaranteed convergence is employed to calculate possibly suboptimal solutions of the main problem, which is known to be NP-hard. The proposed technique enables achieving a large portion of the globally optimal sum rate. It is also very efficient and rather general in terms of allowing interesting extensions, compared with the related results from the literature.