A spectral moment approach to velocity estimation in mobile communications

Estimation of the maximum Doppler spread, or equivalently the vehicle velocity is useful in improving handoff algorithms, and necessary for the optimal tuning of parameters for systems that adapt to changing channel conditions. We provide a novel velocity estimator based on the spectral moments of the in-phase and the quadrature-phase components or the squared envelope of the received signal. We characterize the joint effects of the Ricean K-factor, the directivity and the angle of non-isotropic scattering, and the effects of additive white noise on our estimator and other covariance-based velocity estimators analytically. Simulations illustrate our approach and compare with existing techniques.