Importance sampling Kalman filter for urban canyon navigation
Yeongkwon Choe, Chan Gook Park, Jin Woo Song · 2018
This work presents a complementary navigation algorithm to Global Navigation Satellite Systems (GNSS) in an urban canyon. In urban canyon environment, it is difficult to use GNSS because its signal is frequently denied. Thus, we perform navigation by matching highly detailed 3D maps to measured distance to nearby buildings in GNSS-denied environment. Because the proposed measurement model of distance measurements for 3D map referenced navigation is nonlinear, a filter considering nonlinearity is required. To consider the nonlinearity of the measurement model while reducing computational load, importance sampling technique is fused to conventional Kalman filter. Simulation results are presented to show the feasibility of the proposed system.