Positioning With Mixed Signals Of Opportunity Subject To Multipath And Clock Errors In Urban Mobile Fading Environments

Positioning with signals of opportunity (SOOP) encounters two major technical challenges in urban mobile environments, namely, multipath and clock errors. The presence of signal fading, multipath dominance over direct signal, and non-line of sight (NLOS) signal can occur more than 50% of the time, as evident in the experimental data presented in this paper. As such, conventional delay lock loop (DLL) based on correlation peak tracking can be easily destabilized. As a remedy, we set forth a first arrival estimation method based on the constant false alarm rate (CFAR) detection and matching pursuit (MP) algorithm, which is shown to provide better pseudorange measurements. Furthermore, not all commercial signals of opportunity strictly align their clock to a stable frequency reference such as the GPS atomic time but rather may apply local frequency adjustment on their own. Such an adjustment may result in higher order oscillatory clock errors. We propose to use a six-state Kalman filter to calibrate both quadratic and sinusoidal error terms so as to prevent long-term pseudorange divergence and error growth in the position solutions. Finally, we present experimental data illustrating standalone and inertial-integrated position solutions to demonstrate the performance of urban mobile positioning with mixed signals of opportunity (digital television and cellular phone signals).