Doppler measurements play an important role in highsensitivity receivers for a few of reasons. They canprovide some level of immunity over communicationchannels, convey user velocity, and attitude information,and can also be used to smooth pseudorangemeasurements. Regarding velocity, conventionalestimation methods in high sensitivity receivers first getthe Doppler by block processing the incoming samplesand then compute the velocity using least squares (LSQ)or Kalman filtering (KF). However, in harsh indoorenvironments, Doppler measurements are often biasedand very noisy due to multipath and fading, thusconventional velocity estimation methods are not veryaccurate and the quality is often very poor.This paper first reviews the implementation of a directvelocity estimation algorithm, which is basically themaximum likelihood estimate (MLE) of velocity, andthen presents the Cramer-Rao lower bound (CRLB) of theDoppler estimated with high-sensitivity receivers inindoor multipath environments. The CRLB of theconventional method and the direct method are thenpresented and discussed. Data collected inside a concretebuilding is processed in order to evaluate performance. Inaddition, the effect of adding signals from another GNSS(GLONASS in this case) is also assessed. Results showthat jointly and directly processing GPS and GLONASSsignals outperforms the conventional method, and thebenefits of GLONASS are also apparent in some cases.
展开▼