Optimal Filtering for Jamming Suppression
John W. Betz, The MITRE Corporation
Date/Time: Wednesday, Aug. 25, 4:45 p.m.
This presentation provides theory that predicts the effect of jamming on correlation processing in satnav receivers, and describes the design and performance of optimal linear time-invariant filtering to mitigate effects of jamming on satnav receiver processing. It describes the most efficient (degrading receiver performance the most for a given jamming power) power spectral density shapes for degrading effective C/N0 (the metric for initial synchronization, carrier tracking, and data message demodulation) and for degrading the accuracy of tracking the spreading waveform’s time of arrival (code tracking) for a bandlimiting correlator—the conventional receiver processing that bandlimits the input waveform and then crosscorrelates against a replica of the signal. It also provides numerical results for the degradation of a bandlimiting correlator’s effective C/N0 and spreading waveform tracking accuracy using bandlimited white noise jamming and matched spectrum jamming, including when the jamming has different bandwidths from the receiver bandwidth. Throughout the presentation, numerical results are provided for BPSK-R(10) and BOC(10,5) spreading modulations, representing the original and modernized GPS signals, respectively.
The presentation then describes the processing, called the whitening correlator, that adapts to the power spectral density of the noise plus jamming to maximize effective C/N0 for a given jamming power and power spectral density shape. The whitening correlator filters the bandlimited input waveform to whiten the sum of noise and jamming in the received waveform before correlation, also filtering the signal replica using the same filter design. The whitening correlator also turns out to be the processing that minimizes the error variance in tracking the spreading waveform’s time of arrival.
The benefits of the whitening correlator are explored for different types of jamming, different signal spreading modulations, and different receiver functions. It is shown that the whitening correlator completely removes the effects of the most efficient jamming of a bandlimiting correlator, rendering this jamming completely ineffective. The whitening correlator also can significantly improve receiver performance when the jamming power spectral density is bandlimited white noise, but only when the jamming bandwidth is less than the receiver bandwidth. When the bandlimiting white noise jamming bandwidth is the same as the receiver bandwidth, however, the whitening correlator provides no benefit. For matched spectrum jamming, the whitening correlator always provides at least some benefit. For matched spectrum jamming, the whitening correlator’s greatest benefit is when the jamming bandwidth is less than the receiver bandwidth. Even when the matched spectrum jamming bandwidth is the same as the receiver bandwidth, however, the whitening correlator provides several decibels of receiver performance improvement, with different values of improvement for different signal power spectral densities.
Many implementation considerations for the whitening correlator are identified, posing a variety of challenges and complexities that receiver designers would need to address. Interestingly, however, an alternative implementation of the whitening correlator is presented that avoids the need to filter the signal replica used in the correlator, thus retaining the implementation simplicity of using a one-bit replica. The results inform the design of jamming waveforms, including for testing receivers, and apply to the selection and assessment of jamming mitigation in receiver development.
Approved for Public Release; Distribution Unlimited. Public Release Case Number 21-0199
NOTICE: This technical data was produced for the U. S. Government under Contract No. FA8702-19-C-0001, and is subject to the Rights in Technical Data-Noncommercial Items Clause DFARS 252.227-7013 (FEB 2014)
©2021 The MITRE Corporation