We address joint design of robust precoder and equalizer in a MIMO communication system using the minimization of weighted sum of mean square errors. In addition to imperfect knowledge of channel state information, we also account for inaccurate awareness of interference plus noise covariance matrix and power shaping matrix. We follow the worst-case model for imperfect knowledge of these matrices. First, we derive the worst-case values of these matrices. Then, we transform the joint precoder and equalizer optimization problem into a convex scalar optimization problem. Further, the solution to this problem will be simplified to a depressed quartic equation, the closed-form expressions for roots of which are known. Finally, we propose an iterative algorithm to obtain the worst-case robust transceivers.