On the efficient mitigation of phase noise in MIMO-OFDM receivers

In this paper, an algorithm is proposed for the joint phase noise (PN) estimation and data detection in or- thogonal frequency division multiplexing signals pertaining to spatially multiplexed multiple-input multiple-output channels. Severe oscillator PN gives rise to inter-carrier interference that becomes a limiting factor in sustaining reliable communication link. Considering a received signal perturbed by PN, we derive an iterative Bayesian algorithm based on the statistical modeling of the PN process. This approach however needs computationally intensive inversion of large data-dependent matrices. Taking advantage of the approximate sparsity pattern inherent to the PN spectral composition and exploiting decoupled compensators, the proposed scheme enhances the receiver performance while retaining a modest complexity. Furthermore, we consider how these estimators perform in the case of multiple-antenna RX front-ends with a generalized oscillator signal distribution to the radio paths and numerically evaluate the performance achieved by the proposed algorithm.