An Algorithm for Estimating the Doppler Speed

Adaptive algorithms are obligatory for tracking the channel or its inverse in mobile communication systems. Typical are either the Least-Mean-Squares (LMS) algorithm or the Recursive-Least-Squares (RLS) algorithm. Their performance considerably improves when the step-size for LMS (or forgetting factor for RLS) is chosen according to the Doppler speed, that is the speed of a vehicle in which the receiver is located relative to the base station. A higher Doppler speed usually requires a higher step-size. However, it is not straightforward to derive Doppler speed estimates from the received data. Theoretical analyses have been done in the context of zero-crossings [1, 2], a method that typically requires several thousand samples in order to deliver satisfying estimates. In IS-136 based TDMA phones an estimate is required every slot (162 symbols). This paper describes a very simple scheme for Doppler speed estimation that is well suited for real-time implementations in xed point and derives its properties.