A database of simulated pedestrian trajectories and exact sensors readings is presented. The trajectories are based on the RuDaCoP database of experimental data. The experimental data was processed with ZUPT -aided navigation algorithms in the smoothing mode and then artifacts were removed from the solution with the help of certain additional information. Adding noise and changing sampling rate allows one to model different real situations. For example, analysis of attainable accuracy can be performed statistically. Different implementations of the extended Kalman filter were compared with the help of this database, some results are discussed in the paper.