| We present four vehicle localization (position estimation) algorithms. These algorithms can be broken into two classes: those which minimize the mean square error in the position estimates (i.e. the Kalman filter and the nonlinear least squares constrained algorithms), and those which minimize the maximum level of error in the position estimates (i.e. the Hinfinity filter and the constrained minimax algorithm). We show that gains in accuracy and reliability can be achieved over existing GPS-based approaches by making use of radio ranging based inter-vehicle distance measurements. Also, we show that the accuracy of previously proposed radio ranging based localization can be improved upon by taking into account additional information that is available to vehicles (e.g. digital road maps: vehicle kinematics). Then, after showing the benefits of using our algorithms, we present a detailed mathematical analysis and series of experiments that highlight the advantages and disadvantages of each algorithm. |
