The progress of MEMS sensors technology in size

cost

and power consumption contributes to the development of personal navigation system. Attitude estimation has been a fundamental challenge in the GPSdenied environments for MEMS—based personal navigation system

and accurate heading estimation can effectively restrain drift of navigation error. To obtain accurate heading information and reduce the localization error

the data fusion algorithm based on error quaternion is proposed

and error quaternion is estimated using Kalman filtering algorithm

effectiveness of algorithm is verified by field experiments. Results indicate that the localization error can be reduced to 3 % of the total travelled distance

while walking distance is 80 m.