Developing a precise estimation system of pedestri-an tracks inside buildings in order to qualify others navigation methods is targeted in this project. The aim of this contribution is to improve the existing solution thanks to novel GNSS observations. GNSS based velocity estimate was studied using two dif-ferent observations: Doppler frequency and Time-Differenced Carrier-Phase (TDCP). Following a per-formance comparison of the velocity estimated with both observables, the TDCP was chosen for the hy-bridization filter. The contribution of GNSS TDCP is found to be significant on the existing solution. Achieving a better biases estimate improves the overall quality of pedestrian foot tracks estimation