Signal Processing - September 2017 - 104

should be detected with a probability of missed detection
^ Pmd h less than 10 -9 to ensure user safety. The subset of the
threat space that is left unmitigated with DSIGMA or ground
CCD monitor is denoted by the yellow and green regions in
Figure 5. While the threats in the green region are not mitigated with the addition of the ground IGM only, there is no
unmitigated region remaining if the Pmd values of DSIGMA
and IGM are combined. The combined Pmd is accepted for
threat mitigation by considering that ground IGM is treated
as statistically independent from DSIGMA, because IGM
primarily uses carrier-phase measurements, unlike code measurements that dominate a DSIGMA test statistic. The results
shown in Figure 5 assume specific parameters for each monitor, including the test statistic bounding sigma and detection
threshold, which will not be discussed in detail, and thus their
performance is subject to change depending on monitor design
and simulation conditions.

Signal processing for ionospheric scintillation
monitoring and mitigation
Scintillation monitoring
To develop appropriate signal processing algorithms to mitigate the impact of scintillation on SBASs and other GNSSbased safety-critical systems, it is necessary to understand the
characteristics of scintillation based on high-quality amplitude
and phase scintillation measurements. Since it is challenging for a conventional GNSS navigation receiver to maintain
code and carrier tracking under strong scintillation, a specially
designed GNSS scintillation-monitoring receiver is usually
used for a data collection campaign. A scintillation-monitoring
receiver can utilize signal processing techniques that may not
be applicable for a GNSS navigation receiver because the scintillation monitoring receiver is stationary at a known location,
and postprocessing algorithms are allowed.
An early effort in developing a scintillation-monitoring receiver was a 1992-1996 Small Business Innovative Research
(SBIR) project funded by the U.S. Air Force Research Laboratory [39]. The objective of the project was to modify
low-cost commercial GPS receivers to monitor ionospheric
scintillation using GPS L1 signals. This project offered much
insight into the characteristics of equatorial scintillation
through field experiments. Filtering techniques developed
through this project to extract amplitude and carrier-phase
scintillation parameters from GPS tracking loop outputs are
still widely used.
Today, there are commercial scintillation-monitoring
receivers available from major GNSS receiver manufacturers
[40]. Unlike a typical navigation receiver, a very low-noise
oven-controlled crystal oscillator (OCXO) is used for precise
carrier-phase measurements, and raw amplitude and phase
measurements are provided in a high data rate (e.g., 50 Hz).
Recent products support multifrequency, multiconstellation
GNSS signal tracking, which basically track all visible GNSS
signals with independent tracking channels. Multipath mitigation techniques are also important because multipath fading
104

can contaminate the measurements of amplitude fading due
to scintillation especially for low-elevation satellites. Commercial scintillation-monitoring receivers have proprietary multipath mitigation algorithms.
Because real-time processing is not required for scintillation monitoring, raw intermediate frequency (IF) sampled
data, right after the analog-to-digital converter of the receiver
front end, are preferred particularly on the academic side [41],
[42]. Once IF-sampled data are collected, various novel tracking algorithms can be applied by replaying the data using a
software-defined GNSS receiver. For example, scintillationmonitoring receivers with an open-loop [41] and semiopen
[43] tracking architecture have been proposed. These architectures utilize the known location of the stationary antenna of
the scintillation-monitoring receiver and the precise positions
of satellites obtained by postprocessed precise ephemerides.
Various receiver delays and biases also need to be estimated
and compensated to improve the performance of the openloop architecture.
Numerous commercial or custom scintillation-monitoring
receivers and IF sample recorders have been deployed in the
low- and high-latitude regions where strong scintillations are
expected. Recent activities include the DemoGRAPE project in
Antarctica [44], the CIGALA/CALIBRA network in Brazil [45],
and the SAGAIE network in the Western African region [46].

Scintillation mitigation techniques
Research on scintillation mitigation has been focused on minimizing errors in carrier-phase and carrier-frequency estimation, on reducing the occurrence of carrier cycle slips, and
on maintaining lock of the scintillation signals. For receivers
on dynamic platforms, phase scintillation itself may introduce additional signal phase variations on top of the platform
dynamics. An alternative approach for this type of applications
is to integrate GPS carrier tracking with measurements from an
inertial sensor colocated on the receiver platform. In this section, signal processing techniques for ionospheric scintillation
mitigation are presented, with a focus on standalone receiver
algorithms suitable for SBASs and other GNSS-based safetycritical applications.

Mitigation techniques based on carrier tracking
loop parameter optimization
There are only a limited number of optimization options in a
conventional carrier PLL design: correlator integration time, discriminator type, and loop filter type. This section will provide a
summary of some of these parameter optimization approaches.

Integration time
Signal intensity fading during amplitude scintillation is associated with low signal carrier-to-noise ratio C/N 0 . Longer correlator integration time, in general, will reduce thermal noise
effects and enhance the SNR as shown by (5), which relates the
thermal noise phase error v Th generated by the popular Costas
PLL with the integration time T, noise bandwidth of the PLL
B PLL, and signal carrier-to-noise ratio C/N 0 [20]:

IEEE SIGNAL PROCESSING MAGAZINE

September 2017

Table of Contents for the Digital Edition of Signal Processing - September 2017