Abstract
For two-user multiple-input-multiple-output interference channels with simultaneous wireless information and power transfer, we investigate robust transceiver design problem where channel uncertainties are modeled by worst case model. Our objective is to minimize worst case sum mean square error under individual transmit power constraints and worst case energy harvesting (EH) constraints. We propose to transform the quadratic EH constraints into linear matrix inequalities combined with two equality constraints. Furthermore, we propose to employ an algorithm based on constrained concave-convex procedure to alternatively optimize precoding matrices and postprocessing matrices. Simulation results demonstrate that our proposed robust transceiver design has significant performance gain over the nonrobust one.