CMOS RF Notch Filter for GNSS Receivers

Bandpass sampling architectures, also called subsampling architectures, exhibit several advantages over super-het architectures. First, the complexity of subsampling architectures is significantly lower since no Phased-Locked-Loop (PLL) is required. A direct consequence is that downconversion from RF to IF can be achieved with significant power savings as compared to the super-het architecture. Another significant benefit of such architectures is the capability to process multiple signals at the same time. The property to simultaneously handle multiple carriers make subsampling architectures particularly suited for GNSS applications, as downconversion of multiple frequency bands is required in GNSS environments. With the advent of the new civilian GPS signal L2C and L5 in 2007 and 2009, respectively, and the onset of the new Galileo signal, a receiver that can process multiple signals simultaneously without adding complexity, is highly desired. In fact, a bandpass sampling GNSS receiver front-end made of discrete components was proposed and investigated by Akos et al. [3], [4],; a strategy for determining the optimum sampling frequency for such a GNSS receiver was found by Psiaki et al. [5]. As Psiaki showed, it is possible to downconvert the GPS frequencies L1, L2 and L5 and the Galileo signals simultaneously with a sampling frequency of 55.51, 77.33 or 99.23 MHz. It is evident that Interference in GPS receivers can come from either spurious harmonic mixing products from adjacent transmitters or from intentional jamming. Therefore, in most practical GPS down-conversion receivers, some sort of filtering is applied at the RF input. In practice, off-chip Surface Acoustic Wave (SAW) filters provide excellent blocking capabilities of unwanted out-of band interference. SAW filters exhibit a narrow passband response of less that 20 MHz and typically attenuate out-of band interferes and other out-of band signals by more that 25 dB. However, such off-chip filters are bulky, expensive and can be lossy, degrading the overall Noise Figure (NF) of the system.