Application of the Phase Correction Method in Tracking Modules of Navigation Receivers
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Electrical Engineering. The day in electrical engineering: signals, systems and the hardware that carries them.
Rosa: We're looking at the paper: "Application of the Phase Correction Method in Tracking Modules of Navigation Receivers".
Dev: Here's the gist of the paper. Tracking modules are very important parts of GNSS receivers, providing precise monitoring of satellite signals and evaluating time delays and frequency shifts necessary to calculate ultimate position The pivotal idea of the method supposes correction of the phase deviation between the received signal and its local copy at the end of every integration period.
Rosa: This approach reduces inertial effects crucially.
Dev: Tracking Module Architecture and Inertia Issues Common architecture contains two tracking loops: PLL and DLL, generating local copies of the PRN code and carrier frequency in order to ensure close correlation with the received signal.
Rosa: One of the problems of these loops is their significant inertia in case of narrow bandwidth, which is desirable to mitigate signal noise.
Title and authors: Rosa: So, we're looking at this paper today titled "Application of the Phase Correction Method in Tracking Modules of Navigation Receivers." It’s about those little components in a GPS receiver that keep track of the satellite signals to figure out where you are.
Dev: Right, it focuses on those tracking modules which use things like the phase locked loop and delay locked loop to get copies of the signal and the carrier frequency. The main problem they point out is that these loops have this significant inertia when you use a narrow bandwidth, which is actually what we want because it helps reduce noise.
Taro: Inertia in a tracking loop means that if you integrate for a long time, like for better accuracy, the phase and frequency tracking gets sluggish and becomes really sensitive to any small jitters in the receiver's clock oscillator. That’s a real failure mode we have to watch out for when we move toward longer integration times.
Rosa: Exactly. So this paper proposes this method where they correct that phase deviation between what the receiver got and its local copy at the end of every integration period, which is supposed to reduce those inertial effects crucially by doing it directly over the phase accumulator of a digital direct sequence system, or DDS system.
Dev: That direct correction over the phase accumulator is smart because it’s supposed to improve both the accuracy and those dynamic characteristics of how fast the tracking process moves. It basically acts like an additional frequency offset for that local oscillator.
Taro: I'm interested in how this change affects things when we look at the dynamics. Does this correction help smooth out those transient processes, like that overshoot they showed in their simulation, where it started with equal frequencies but different initial phases?
Rosa: Yeah, the simulations show that by doing this phase correction directly over the phase accumulator with a value of EPLL multiplied by two to n, they let the tracking modules use narrower filter bandwidths while still keeping things stable. That’s what they showed in Figure fifteen and their conclusion.
Title and authors: Dev: Narrower filter bandwidth is key because it means we can reduce the noise bandwidth without losing stability, which is a big win for reducing noise interference in the system. They demonstrated that this can allow for a PLL loop bandwidth reduction from about twenty five hertz down to about zero point five to one hertz when integrating over twenty milliseconds.
Taro: That's a big reduction in the loop bandwidth, going from something quite fast down to near one hertz over twenty milliseconds. How does that translate into real-world performance when the signal starts fluctuating?
Rosa: The paper shows that using deviation values obtained from unfiltered input signals can be a drawback if you use relatively short integration intervals, like just a few milliseconds. But they suggest that using narrower filter bandwidth and longer timing, which comes from less inertia, outweighs that initial drawback. They recommend starting tracking with integration intervals of a few milliseconds and then moving to longer ones once the PLL has stabilized and the signal frequencies aren't fluctuating much anymore.
Dev: So it’s a staged approach, right? You start fast when you need stability, but you push for longer integrations later because the inertia is reduced by this correction method. It makes sense from a control system standpoint to manage that loop rate carefully.
Taro: I wonder about the robustness against initial errors. They mentioned that this phase correction at the very beginning of tracking eliminates that overshoot we see in transient processes, like what they showed in Figure two which means the tracking module becomes significantly less sensitive to those initial frequency deviations when you use longer integration intervals <ref:2611.14007#pg2>.
Rosa: That’s a key point for me because if the receiver starts with a big initial error, this method helps it recover without overshooting as much. It makes the system more robust when you integrate for longer periods.
Dev: From an engineering standpoint, that robustness is important because it affects how we design the safety case for real-time loops. If the system handles those initial transient errors better with longer integration times, that means less chance of instability during those extended tracking periods.
Title and authors: Taro: So, when we think about what this means for autonomy and real-world applications, it suggests that we can rely on these tracking modules to handle longer observation times more effectively without worrying about them spiraling out due to initial conditions.
Rosa: Right. So to wrap up the paper "Application of the Phase Correction Method in Tracking Modules of Navigation Receivers," it shows that correcting the phase deviation directly over the phase accumulator is a way to reduce inertial effects and allow for much narrower noise bandwidths without losing stability, especially when you use longer integration intervals.
Dev: And they show this method can help handle initial frequency deviations more robustly at the start of tracking by eliminating that overshoot seen in transient processes. This leads to a tracking module that is less sensitive to those initial conditions when you integrate over longer times like twenty milliseconds.
Taro: I just think it’s about making the long-term operation smoother and more predictable, which is essential for reliable navigation systems.
Rosa: Yeah, so that’s what it boils down to for the tracking module itself. It's a way to make those core elements of GNSS receivers more accurate over longer observation times than before.
Dev: And they open the door to extending this idea with third-order filters, provided you adjust the filter parameters appropriately, which suggests there’s still room for further refinement in the design of these loops.
Taro: I’m curious to see how those third-order filters behave when we push them beyond what they were originally tested on.
Rosa: Well, that’s what this paper is about—applying this correction method to improve tracking modules and show how it can handle the practical issues of inertia and initial transients in GNSS receivers.
Dev: It seems like a solid piece of work for improving the fundamental tracking loop design in these systems.
Taro: I think making those loops more resilient to time instability is what really matters for dependable positioning.
The paper's summary: Rosa: So, this paper is basically saying that by fixing that phase difference between what you get and what your receiver thinks it got at the end of every time period, you can drastically cut down on those inertial effects we talked about before.
Dev: Right. It’s about doing this correction directly over the phase accumulator in a DDS system, which is supposed to make the whole tracking process more accurate and dynamic.
Taro: The core idea here is that this isn't just a small tweak; it’s meant to improve how well the module handles things when it starts up or when things get noisy.
Rosa: Yeah, and they show that this method lets you use much narrower filter bandwidths while still keeping the tracking stable, which is pretty significant because usually narrowing the filter means more noise gets in.
Dev: They quantify that improvement by showing a PLL loop bandwidth drop from about twenty five hertz down to something around zero point five to one hertz when you integrate over twenty milliseconds.
Taro: So, if we're talking about real-world scenarios, that means the system can handle a lot more time integration before things get really sluggish and sensitive to small clock jitters.
Rosa: Exactly. And they showed it helps with those initial frequency errors too; it eliminates the overshoot you see during those early transient processes when you start tracking.
Dev: So, if a receiver starts up slightly off, this correction at the beginning smooths out that initial error, making the module much less sensitive to those starting conditions when you use longer integration times.
Taro: That robustness is important because it changes how gracefully the system recovers when things aren't perfect right from the start.
Rosa: It does. And they suggested a practical approach: start tracking with short time intervals, like a few milliseconds, and then move to longer ones once you see that the PLL has settled and the frequencies aren't fluctuating much anymore.
Dev: That sounds like a sensible way to manage the loop rate—start conservative when you need stability, then push for longer integrations later.
Taro: But what about the limits? The paper mentions using deviation values from unfiltered input signals can be a problem if you only use short integration intervals, like just a few milliseconds.
Rosa: They do flag that; if you're integrating for very short periods, using those raw, unfiltered deviations can actually give worse results compared to this corrected approach.
Dev: So the trade-off is clear: you need longer timing and narrower bandwidths for stability, but you have to be careful about how long you integrate before the initial correction helps versus when it might introduce other issues.
Taro: It seems like a solid way to make those tracking modules more resilient over time, especially for longer observation periods where we need that precision.
Rosa: That's what it boils down to—making the core of the GNSS receiver smoother and more accurate across longer observation times by doing this phase correction directly in the hardware.
The paper's improvements: Rosa: So, we're looking at how they suggest improving this method further; it’s not just about the correction itself but how they build on it to handle different situations better.
Dev: They talk about extending this phase correction over the phase accumulator by multiplying that value by two to n, which is supposed to let tracking modules use much narrower filter bandwidths while maintaining stability.
Taro: Narrower filter bandwidth is good because it means we can reduce the noise bandwidth without losing stability, and they show this allows for a PLL loop bandwidth reduction from about twenty five hertz down to something around zero point five to one hertz when integrating over twenty milliseconds.
Rosa: That's a big change in terms of how fast the loop is reacting, which is important for real-time performance.
Dev: And they mentioned that the tracking module can handle initial frequency deviations more robustly at the very start of tracking because this phase correction eliminates that overshoot we saw during those early transient processes.
Taro: So when you're running a system with longer integration times, it means if the receiver starts off with a big error, it recovers much better without overshooting as much.
Rosa: It makes the whole module significantly less sensitive to those initial frequency deviations when you use those longer integration intervals we discussed earlier.
Dev: That robustness is key for our safety case because it means less chance of instability during those extended tracking periods in real-time applications.
Taro: So, by improving the correction at startup, the system is more resilient to real-world initial conditions that aren't perfect right away.
Rosa: And they also mentioned that there are no obstacles to extending this method to use third-order filters for tracking modules, as long as you adjust the filter parameters correctly.
Dev: So the authors have shown a path forward for making these loops even more complex and accurate by using higher order filters.
Taro: I'm curious how those third-order filters would behave in practice when we push them beyond what they were originally tested on, especially in noisy environments.
Rosa: That’s exactly what this paper is about—showing that this correction method can handle the practical issues of inertia and initial transients, and then it opens the door for further refinement with these higher order filters.
Conclusion: Rosa: So, to wrap up this paper on the "Application of the Phase Correction Method in Tracking Modules of Navigation Receivers," they’re showing that this phase correction over the phase accumulator really helps reduce inertia and lets you use much narrower filter bandwidths without losing stability.
Dev: It seems like a solid method for improving tracking accuracy when you're dealing with longer integration times, which is a key concern for our real-time loops.
Taro: I just think making those loops more resilient to time instability is what really matters for dependable positioning in the field.
Rosa: That’s right. And they also showed that this method can help handle initial frequency deviations more robustly at the start of tracking by eliminating that overshoot we saw during those early transient processes.
Dev: So, if a receiver starts off slightly off, this correction at the beginning smooths out that initial error, making the module much less sensitive to those starting conditions when you use longer integration times like twenty milliseconds.
Taro: That robustness is important because it changes how gracefully the system recovers when things aren't perfect right away.
Rosa: It does. And they suggested a practical approach: start tracking with short time intervals, like a few milliseconds, and then move to longer ones once you see that the PLL has settled and the frequencies aren't fluctuating much anymore.
Dev: That sounds like a sensible way to manage the loop rate—start conservative when you need stability, then push for longer integrations later.
Taro: But what about the limits? The paper mentions using deviation values from unfiltered input signals can be a problem if you only use short integration intervals, like just a few milliseconds.
Rosa: They do flag that; if you're integrating for very short periods, using those raw, unfiltered deviations can actually give worse results compared to this corrected approach.
Dev: So the trade-off is clear: you need longer timing and narrower bandwidths for stability, but you have to be careful about how long you integrate before the initial correction helps versus when it might introduce other issues.
Taro: It seems like a solid way to make those tracking modules more resilient over time, especially for longer observation periods where we need that precision.
Rosa: That's what it boils down to—making the core of the GNSS receiver smoother and more accurate across longer observation times by doing this phase correction directly in the hardware.
Dev: And they open the door to extending this idea with third-order filters, provided you adjust those filter parameters appropriately, which suggests there’s still room for further refinement in the design of these loops.
Taro: I'm curious to see how those third-order filters would behave in practice when we push them beyond what they were originally tested on, especially in noisy environments.
Rosa: Well, that’s what this paper is about—applying this correction method to improve tracking modules and showing how it can handle the practical issues of inertia and initial transients in GNSS receivers.
Kudryavtsev I.A., Shafran S.V., Balmasova M.A.
Samara National Research University
eess.SP
Submitted: 2026-11-10
Updated: 2026-11-10
Comments: 15 pages, 17 figures. The paper is accepted for publication in the journal "Radiotechnika" (Radio engineering)
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 74/100
The gist: Tracking modules are very important parts of GNSS receivers, providing precise monitoring of satellite signals and evaluating time delays and frequency shifts necessary to calculate ultimate position…
Key concepts
- Tracking Modules
- These are crucial parts of GNSS receivers that monitor satellite signals. They evaluate time delays and frequency shifts to calculate the final position. The paper focuses on improving how these modules track signals over time.
- Inertia in Tracking Loops
- Common tracking loops, like PLL and DLL, have inertia when using narrow bandwidths. Increasing integration time increases this inertia, which can make the receiver sensitive to clock instability and affect dynamic performance.
- Phase Correction Mechanism
- The method corrects phase deviation by calculating a shift value ($ riangle ext{φ}$) from in-phase and quadrature channel integration results. This correction is applied directly to the DDS phase accumulator, effectively acting as an additional frequency offset for the local oscillator.
Terminology
Summary
Tracking modules are very important parts of GNSS receivers, providing precise monitoring of satellite signals and evaluating time delays and frequency shifts necessary to calculate ultimate position The pivotal idea of the method supposes correction of the phase deviation between the received signal and its local copy at the end of every integration period. p. 0 Samara National Research University
Abstract
Tracking modules are very important parts of GNSS receivers, providing precise monitoring of
satellite signals and evaluating time delays and frequency shifts necessary to calculate ultimate position. Common architecture contains two tracking loops: PLL and DLL, generating local copies of the PRN code
and carrier frequency in order to ensure close correlation with the received signal. This correction
is performed directly over the phase accumulator of a DDS system. This approach reduces inertial effects
crucially.Where the paper says it
Where the paper says it
Tracking Module Architecture and Inertia Issues
Common architecture contains two tracking loops: PLL and DLL, generating local copies of the PRN code and carrier frequency in order to ensure close correlation with the received signal. One of the problems of these loops is their significant inertia in case of narrow bandwidth, which is desirable to mitigate signal noise. p. 0 Common architecture contains two tracking loops: PLL and DLL, generating local copies of the PRN code
and carrier frequency in order to ensure close correlation with the received signal. One of the problems of
these loops is their significant inertia in case of narrow bandwidth, which is desirable to mitigate signal noise.Where the paper says it
Phase Correction Mechanism
The method describes a method of phase correction, improving accuracy and dynamic characteristics of the tracking process in GNSS receivers. The correction is performed directly over the phase accumulator of a DDS system. p. 0 The pivotal idea of the method supposes correction of the phase
deviation between the received signal and its local copy at the end of every integration period. This correction
is performed directly over the phase accumulator of a DDS system. The correction value is added directly to the phase accumulator. The phase shift value ∆𝜑 is calculated using the integration results in the in-phase and quadrature
channels (IP и QP). 7
In the presence of a frequency offset (let 𝜔𝑖𝑛 = 𝜔𝑙𝑜𝑐 + ∆𝜔) we can write:
𝐼𝑃 + 𝑗𝑄𝑃 = ∫ 𝑒
𝑗(𝜔𝑙𝑜𝑐𝑡+∆𝜔𝑡+𝜑𝑖𝑛)
∙
𝑇
0
𝑒
−𝑗𝜔𝑙𝑜𝑐𝑡𝑑𝑡 = ∫ 𝑒
𝑗∆𝜔𝑡 ∙ 𝑒
𝑗𝜑𝑖𝑛𝑑𝑡
𝑇
0
Hence, we obtain:
𝑃 + 𝑗𝑄𝑃 = 𝑒
𝑗𝜑𝑖𝑛 ∫ 𝑒
𝑗∆𝜔𝑡𝑑𝑡 = 𝑒
𝑗𝜑𝑖𝑛 ∫(𝑐𝑜𝑠(∆𝜔𝑡) + 𝑗𝑠𝑖𝑛(∆𝜔𝑡))𝑑𝑡 =
𝑒
𝑗𝜑𝑖𝑛
∆𝜔
𝑇
0
∙ [𝑠𝑖𝑛(∆𝜔𝑇) − 𝑗(𝑐𝑜𝑠(∆𝜔𝑇) − 1)]
𝑇
0
𝐼𝑃 + 𝑗𝑄𝑃 =
𝑒
𝑗𝜑𝑖𝑛
∆𝜔
∙ [2 sin (
∆𝜔𝑇
2
) cos (
∆𝜔𝑇
2
) + 𝑗2𝑠𝑖𝑛2 (
∆𝜔𝑇
2
)] =
=
2𝑒
𝑗𝜑𝑖𝑛
∆𝜔
∙ sin (
∆𝜔𝑇
2
) ∙ [𝑐𝑜𝑠 (
∆𝜔𝑇
2
) + 𝑗 sin (
∆𝜔𝑇
2
)]
It follows that the output signal of the phase discriminator in the Costas loop is proportional to the
sum of the initial phase deviation and ∆𝑓𝑇
2
. In other words, the application of the phase correction described
above can be interpreted as an additional frequency offset of the local oscillator.Where the paper says it
Where the paper says it
Where the paper says it
Filter Bandwidth and Integration Interval Trade-offs
The results of numerical simulations and physical experiments display the efficiency of the method in the case of tracking using long-time integration. Using deviation values obtained from unfiltered input signals is a drawback of the approach, so in the case of relatively short integration intervals, such correction can provide worse results; however, using narrower filter bandwidth and longer timing, which are available due to less inertia, outweighs this drawback. p. 0 The results of numerical simulations and physical experiments display the efficiency of the method
in the case of tracking using long-time integration. Using deviation values obtained from unfiltered input
signals is a drawback of the approach, so in the case of relatively short integration intervals, such correction
can provide worse results; however, using narrower filter bandwidth and longer timing, which are available
due to less inertia, outweighs this drawback. Using narrow filter bandwidth requires attention to transient processes, described in the paper. It is
recommended to start tracking with the integration intervals of a few milliseconds and proceed to longer ones
after some time, when the PLL has stabilized and the signal frequencies are not fluctuating significantly.Where the paper says it
Where the paper says it
Experimental Validation and Conclusion
The experimental research shows that the proposed method of phase correction in the DDS-based local carrier oscillator positively affects the operation of the tracking module, especially when using integration intervals longer than 1 ms. The main effect is achieved by narrowing the filter bandwidth in the PLL loop and reducing the probability of phase difference accumulation. p. 3 The obvious limitation for
the Costas loop can be described by the expression: |∆𝜔| ∙ 𝑇 <
𝜋
2
, where ∆𝜔 isthe frequency difference (rad/s),
and t is the integration interval (s). In addition, the desire to set the filter noise bandwidth as narrow as possible in order to reduce tracking
errors leads to the inertia of the circuit as a whole, which in turn can lead to a loss of lock, even if the above
limitation is met, due to the slow compensation of the phase mismatch, which leads to the accumulation of
error. Application of phase correction method with narrow filter bandwidth is much more efficient when
FLL-assisted PLL tracking is used. It is shown that in this case a separate FLL discriminator is not needed,
and the resulting filter structure can be significantly simplified. The obvious limitation for
the Costas loop can be described by the expression: |∆𝜔| ∙ 𝑇 <
𝜋
2
, where ∆𝜔 isthe frequency difference (rad/s),
and t is the integration interval (s). In addition, the desire to set the filter noise bandwidth as narrow as possible in order to reduce tracking
errors leads to the inertia of the circuit as a whole, which in turn can lead to a loss of lock, even if the above
limitation is met, due to the slow compensation of the phase mismatch, which leads to the accumulation of
error. All the experiments were carried out with a second-order
filters, however the method can be used with a third-order filter architecture as well. Keywords
GNSS, PLL, DLL, NCO, Phase correction, Bandwidth, PseudorangeWhere the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Phase correction in a DDS based local carrier oscillator reduces inertial effects by performing correction directly over the phase accumulator with a value of EPLL multiplied by 2 to n. This allows for tracking modules to use narrower filter bandwidths while maintaining stability, as shown in Fig 15 and Conclusion.
-
The proposed method allows for narrowing the noise bandwidth without losing stability, enabling a PLL loop bandwidth reduction from about 25 Hz to about 0.5-1 Hz when integrating over 20 ms. This improvement is demonstrated experimentally with a filter noise bandwidth of 5 Hz for an integration time of 1 ms and approximately 0.5-1 Hz for an integration time of 20 ms.
-
The tracking module can be improved to handle initial frequency deviations more robustly, as the phase correction at the beginning of tracking eliminates the overshoot observed in transient processes like Fig 2. This results in a
significantly less sensitive
tracking module to initial frequency deviation when using longer integration intervals.
Abstract
One of the problems of the tracking modules is their significant inertia in case of narrow bandwidth, which is desirable to mitigate signal noise. The paper describes a method of phase correction, improving accuracy and dynamic characteristics of the tracking process in GNSS receivers. The pivotal idea of the method supposes correction of the phase deviation between the received signal and its local copy at the end of every integration period. This correction is performed directly over the phase accumulator of a DDS system. This approach reduces inertial effects crucially. Application of phase correction method with narrow filter bandwidth is much more efficient when FLL-assisted PLL tracking is used. It is shown that in this case a separate FLL discriminator is not needed, and the resulting filter structure can be significantly simplified.
Related papers
- Open-H-RF: A Large-Scale Ultrasound Raw Channel Dataset Initiative
- Scalable Attention for 5G NR Channel Estimation
- Low-Altitude Wireless Networks: A Comprehensive Survey
- Sensing-Assisted OFDM with Index Modulation: A Waveform for Optical Wireless Integrated Communication and Localisation
- Iteration-Free Candidate Pruning and an Index-Domain VLSI Architecture for Sort-Free MIMO Box Decoding up to 1024-QAM
- A Foundation Model for Large-Scale Wireless Network Planning, Operation and Optimization