A receiver design for MIMO systems over rayleigh fading channels with correlated impulse noise

A Middleton Class-A (MCA) density is well known to model impulsive interference. The statistical-physical extension of this model for multiple receive antennas is currently limited to two antennas. An algebraic extension of the univariate MCA model leads to a multivariate MCA distribution, which can be used for an arbitrary number of receive antennas. Since recent studies show a significant level of noise correlation in several wireless systems, we develop MIMO receivers for Rayleigh fading channels in the presence of spatially correlated MCA interference. We derive an upper bound pairwise error probability (PEP) for orthogonal space time block codes (OSTBCs). We show that the performance improvement of OSTBCs is highly dependent on the impulse noise environment and it becomes minor as the number of transmit and receive antennas increases. In the design of MIMO receivers, the maximum likelihood (ML) detection has a high computational complexity. Since the MCA model can be seen as a multivariate Gaussian distribution conditioned on the knowledge of noise state, we introduce a simple approach to estimate the state of noise at the receiver, which subsequently reduces the complexity of the ML decision rule.