Fundamentals of statistical signal processing: estimation theory
Fundamentals of statistical signal processing: estimation theory
An introduction to signal detection and estimation (2nd ed.)
An introduction to signal detection and estimation (2nd ed.)
Asymptotic spectral distribution of Toeplitz-related matrices
Fast reliable algorithms for matrices with structure
Toeplitz And Circulant Matrices: A Review (Foundations and Trends(R) in Communications and Information Theory)
IEEE Transactions on Information Theory
Neyman-pearson detection of gauss-Markov signals in noise: closed-form error exponentand properties
IEEE Transactions on Information Theory
Hi-index | 35.68 |
An approximate implementation is formulated and analyzed for the detection of wide-sense stationary Gaussian stochastic signals in white Gaussian noise. For scalar processes, the approximate detector can be realized as the correlation between the periodogram of the observations and an appropriately selected spectral mask, and thus is termed the frequency-domain correlation detector. Through the asymptotic properties of Toeplitz matrices, it is shown that, as the length of the observation interval grows without bound, the frequency-domain correlation detector and the optimum quadratic detector achieve identical asymptotic performance, characterized by the decay rate of the miss probability under the Neyman-Pearson criterion. The frequency-domain correlation detector is further extended to the detection of vector-valued wide-sense stationary Gaussian stochastic signals, and the asymptotic optimality of its performance is established through the asymptotic properties of block Hermitian Toeplitz matrices.