A physically-based impulsive noise model for array observations

Impulsive interference has been measured previously in some wireless communication environments. Physically-based impulsive interference models have been lacking for cases where the antenna elements in an array are closely spaced. Such cases are important when spatial processing is employed. In these closely-spaced element cases, the interference observations may be statistically dependent from antenna to antenna. A model for these closely-spaced element cases is developed using some ideas originally proposed by David Middleton. The model developed here is very general and so it appears to be applicable to a wide variety of physical situations. It assumes interference sources which are Poisson distributed in space and time, and it includes an additive Gaussian background component.