This paper explores a way of combining conventional inertial sensors with cold atom interferometers (CAI) inorder to reduce the drift of the navigation solutions in velocityand orientation. Instead of complementing and improving theCAI with conventional sensors, in this approach the conventionalIMU will be used as main sensor for a prediction of the kinematicstate. The CAI is then used for the correction of systematic errorsand offsets in the framework of an extended Kalman Filter.Monte Carlo simulation studies demonstrate an improvementof the navigation solution precision. In addition, most drifts ofvelocity and orientation can be eliminated and the uncertainty ofthe velocity solution can further be reduced by a factor of 30 ormore compared to the conventional strapdown. The observability of the error states is discussed.