A Dynamically Configurable Decimator for a GNSS Baseband Receiver

Abstract

Modern global navigation satellite system (GNSS) front ends sample the received radio-frequency signal at an intermediate frequency (IF) and, then, downconvert it to baseband in the digital domain. A locally generated spreading code is subsequently correlated with the baseband signal at the sampling frequency. This can mean that the receiver correlates at an unnecessarily high frequency (sometimes more than four times the chipping rate). The power consumption is, thus, high. If the baseband signal is downsampled after appropriate filtering, the correlator can process at a fraction of the sampling frequency. The power dissipation is considerably reduced as a result. Furthermore, by resampling at an integer submultiple of the sampling frequency, not only is downsampling jitter eliminated, but also architectural configurability in the correlator is possible, allowing further reductions in power and resource consumption. In this paper, a dynamically configurable decimator for a matched filter correlator is presented. The mathematical model of the signal processing is analyzed for various downsample rates. The effective postcorrelator signal-to-noise-plus interference ratio is estimated in theory and measured by experiment, illustrating that the proposed decimator correlator outperforms the resample IF correlator.