Interferer Synchronized Sub-sampling for Continuous Wave Tone Suppression in Direct Sequence Spread Spectrum Systems such as GPS
Sakib Abdullah and Izzet Kale, University of Westminster, Applied DSP and VLSI Research Group, UK
Alternate Number 2
Global Navigation Satellite Systems (GNSS) underpin many of today’s technologies which are reliant on accurate positioning awareness. In many cases the reliability of such systems is crucial to ensuring safety of life as well as financial security. Unfortunately, the rapid adoption and spread of GNSS technologies is confronted with the ever-increasing threat from readily accessible civil jammers. Despite restrictive legislation, these devices continue to prove popular. To combat this, it becomes necessary to relentlessly pursue any avenue of opportunity which may lead to a more robust receiver.
This paper explores the feasibility of exploiting the temporal separation between the zero crossings of an interferer signal and the received spread spectrum signal.
The nature of Direct Sequence Spread Spectrum (DSSS) ensures a generous spreading of the information bandwidth such that isolating the zero crossings can provide sufficient data samples to fulfil the subsampling criteria. Based on this knowledge, if the subsampling instances could be caused to coincide with the estimated time instances where the interferer is at or near zero, then this component of the received signal could be significantly ‘attenuated’. What this amounts to, is a need to accurately estimate the frequency and phase characteristics of the interferer which can then be used to control the sub-sampling action. In conveying this technique, this paper will focus on Continuous Wave (CW) tones and their effects.
A combination of Simulink and MATLAB modelling is used to gauge the accuracy to which the interferer must be estimated in terms of its frequency/phase to ensure significant suppression of the interferer such that subsequent correlation stages perform adequately. This is first evaluated in an ideal context followed by the introduction of Additive White Gaussian Noise (AWGN) for a more realistic assessment. In addition, the technique is assessed as a whole, for its effectiveness in interferer suppression and is weighed against the technical effort involved in achieving satisfactory performance. Evaluation is carried out in the context of GPS technology whereby the signals are also modelled in MATLAB.
Initial results indicate the effectiveness of this technique against a stationary in-band single tone interferer provided the tone and signal carrier frequencies do not coincide. A difference in the phase between the interferer and the sampling operation manifests as a ‘bias’ in the output which acts to reduce the SNR going into the de-spreading operation. Nonetheless, signal acquisition remains possible within a margin of phase dictated by the signal’s amplitude. A more stringent requirement is placed on the accuracy of the estimated frequency of the interferer. This level of sensitivity has shown that slight deviations are sufficient to cause the output to grow too large within the length of a CA code period for the simulated case of GPS. This work will aim to more accurately identify these thresholds and the relationships which govern them.
Although interference, particularly from deliberate sources can be quite varied in terms of their time/frequency characteristics, only the case of stationary CW tones is explored in this work in order to convey the technique and provide a foundation for further research into its feasibility for more complex scenarios.