To provide a solution robust and accurate navigation , GPS receivers must operate in optimal conditions , that is to say have a direct line of sight with at least four satellites, which is hard to find in an urban environment where GPS signals may be contaminated with significant multipath errors . Coupling GPS data with those from an inertial navigation system (INS) can significantly reduce these errors, however, this technique is generally valid for high-end systems. Indeed, for systems with low cost inertial sensors used generally have significant measurement errors causing a rapid divergence of the navigation solution . There is a real need for the development of new technologies that enable self calibration of inertial sensors at low cost.

The objective of this project is to improve the performance of integrated navigation systems INS / GPS low cost based on the use of inertial MEMS sensors to enable an accurate and robust positioning in urban environment .
For this , this paper proposes the following methodology:

1) Development of a robust calibration procedure for real-time error compensation deterministic
2) Error Compensation stochastic nature of low-cost inertial sensors using a Gauss- Markov model (GM ) based on the parameter of the autoregressive (AR) model

Read the detailed site SUBSTANCE article.
SUBSTANCE blog is a graduate of the École de technologie supérieure interactive allowing members of the community of graduate students , professors and researchers from ETS to submit articles on their academic activities .