Efficient identification of impulsive channels

Summary form only given. A physically meaningful parametric model for impulsive interference is the so-called Class A Middleton model, whose parameters A and Γ can be adjusted to fit a great variety of non-Gaussian noise phenomena occurring in practice. A, the overlap index, is a measure of the average overlap of successive emission events. Γ, the Gaussian factor is the ratio of the intensity of the independent Gaussian component of the input interference to the intensity of the non-Gaussian component. A batch estimator of the Class A parameters with good small-sample-size performance has been obtained. The estimator is based on the EM algorithm, a two-step iterative technique which is ideally suited for the Class A estimation problem since the observations can be readily treated as incomplete data. For the single-parameter estimation problem (A unknown, Γ known), a closed-form expression for the estimator is obtained. It has been shown that the sequence of estimates obtained by the EM algorithm converges, and if the limit point to which the sequence converges is an interior point of the parameter set of interest, then it must necessarily be a stationary point of the traditional likelihood function. The small-sample-size performance of the estimator has been examined by an extensive simulation study. For both the single-parameter and two-parameter estimation problems, the results indicate that this likelihood-based scheme yields excellent estimates of the Class A parameters (in terms of attaining the Cramer-Rao lower bound) for sample sizes on the order of 10³.