This paper proposes a low-complexity and robust H-infinity channel estimator for Multiple-input multiple-output (MIMO) Orthogonal frequency division multiplexing (OFDM) systems. To lessen the complexity, the Expectation maximization (EM) and Space-alternating generalized EM (SAGE) algorithms are adopted respectively. To treat with the effect of Non-Gaussian noise (NGN) channels, an Equivalent signal model (ESM) is used to enhance the robust of the EM-type H-infinity estimator. Furthermore, joint estimation using Phase shifted orthogonal (PSO) pilot sequences is applied for improving the estimation and the requirement for pilot. Simulation results show that the H-infinity estimator has almost the same Mean square error (MSE) as Maximum a posteriori (MAP) estimator. By using the ESM, the robustness of EM-type based H-infinity estimator is enhanced against NGN channels. In addition, the MSE and pilot overhead are reduced by applying the joint estimation with PSO.