Parameter estimation of chirp signals using the metropolis-adjusted-langevin's algorithm
Yan Lin, Xintan Wang, Peng Yingning · 2005
This paper addresses the problem of parameter estimation of chirp signals in additive Gaussian white noise. A new Markov chain Monte Carlo (MCMC) method called the metropolis-adjusted-Langevin's (MAL) algorithm is employed to solve this problem, which is faster to converge than the random walk metropolis-hastings (MH) algorithm. The initial values for the method are obtained by the discrete polynomial-phase transform (DFT). Simulations show that the Cramer-Rao low bound (CRLB) can be attained by the proposed method even at low signal-to-noise ratio (SNR) and the MAL algorithm is more efficient than the random walk MH algorithm.