Asymptotic efficiency of distributed estimation from constant modulus sensor transmissions

A distributed estimation problem using sensor networks is considered. A parameter is observed in noise by sensors, phase-modulated using complex exponentials and transmitted over a Gaussian multiple-access channel to a fusion center. The received signal is used to estimate the parameter under observation. The system is evaluated using the asymptotic variance of the estimate. The relationship between the Fisher information and the characteristic function is presented and used to show that the estimator is efficient if and only if the sensing noise is Gaussian.