Scalable low-complexity GPS and DGPS positioning using approximate QR decomposition

• GPS/DGPS with dynamically scalable on-demand positioning accuracy.
• QRD using CORDIC-based approximate rotations.
• Adjusting positioning accuracy by choosing the number of optimal CORDIC angles.
• The presented method is highly suited for hardware implementation.
• The presented method is particularly suited for recursive positioning.

In this paper we discuss the efficient implementation of GPS and DGPS positioning using COordinate Rotation DIgital Computer (CORDIC)-based approximate rotations for solving the required Least Squares (LS) problems. By changing the number of optimal CORDIC angles (approximation accuracy), different accuracies of the positioning results can be obtained. In particular, we show that coarse approximations are sufficient for obtaining the required positioning accuracies. Experimental results are obtained using Rinex files from high-end GPS receivers and raw GPS data from low-end GPS receivers. The accuracy of the positioning results is compared for varying numbers of iterations and varying approximation accuracies. The presented positioning methods with scalable accuracies allow cost- and power-efficient software/hardware implementations for both GPS and DGPS positioning, while the required accuracies of different Location Based Services (LBS) are obtained.