Doppler Shift Estimation for TD-SCDMA System over Railway Environments

In this paper, a new Doppler shift estimator is proposed for Time Division-Synchronous Code Division Multiple Access (TD-SCDMA) over railway environments. Different from other existing methods, which commonly assume that the fading channel coefficients are available for the Doppler estimators, the proposed estimator is developed based on the signals which contain unknown frequency selective fading channel coefficients, unknown Rice factor and random receiver noise; it contains three steps. The statistical properties of the training symbols which are called near independent identically distributed (i.i.d.) training sequences, i.e., midamble, are first analyzed. Afterwards, by the help of a training sequence, the second-order and fourth-order statistics of the received signals are calculated. Then, by exploiting these statistical results, the Doppler shift estimator is obtained. The proposed estimator employs a smoothing function with corrective factor. By optimizing the corrective factor, an accuracy solution is acquired within the Doppler shift range of possibilities. Simulation results indicate that the proposed estimator is greatly reliable and effective to estimate Doppler shift up to 400 Hz.