1. Introduction
Global Navigation Satellite Systems (GNSSs) provide essential Positioning, Navigation, and Timing (PNT) information across various sectors, including high-precision maritime applications. These systems, encompassing global (GNSS) and regional (RNSS) configurations, rely on satellite-transmitted radio signals for precise location and time determination. However, GNSS signal quality is susceptible to various threats and influencing factors, necessitating robust monitoring systems, as discussed in the following sections.
One of the foremost challenges for Assured PNT (APNT) solutions users is the threat to GNSS RF signals, particularly in environments completely denied, disrupted, intermittent, or limited (Denied/Disrupted/Intermittent/Limited). These conditions are commonly encountered in conflict zones but also in coastal maritime areas, such as the Black Sea and the Mediterranean Sea, where commercial and military activities create a congested RF environment [
1].
Any disruption to GNSS signal transmission can lead to a partial or complete loss of the PNT solution, meaning that navigation (whether by car, aircraft, or vessel) occurs without geo-referencing. Typically, the fundamental tasks of GNSS receivers are to receive and separate satellite signals, calculate pseudo-ranges for each satellite based on the signal reception time, demodulate the navigation message to obtain ephemeris data, and finally estimate the PVT solution. In this context, GNSS signal threats, whether unintentional or intentional, can cause signal denial or falsification. Intentional transmission of interference signals within the GNSS frequency spectrum is known as jamming.
Spoofing involves deliberately transmitting false GNSS signals to a receiver, causing it to compute an incorrect position and leading the user to believe they are in a different location or time than they are [
2]. This type of threat is a significant concern in safety-critical applications across all industries, whether commercial or military, that rely on precise PNT solutions [
3]. From the user’s perspective, the distinguishing factor between the two types of GNSS interference—jamming and spoofing—is their effect on the receiver’s ability to provide a PNT solution.
GNSS jamming disrupts positioning, navigation, and timing (PNT) by overpowering the genuine GNSS signal, causing loss. Spoofing, conversely, feeds the receiver false PNT data, leading to incorrect location information. Spoofing is a more serious threat to GNSS users due to its deceptive nature. While both jamming and spoofing disrupt GNSS, the methods and tools used to carry them out differ.
Although illegal in most cases, very-low-power jamming can be carried out using devices known as PPDs (personal privacy devices), which can be purchased online and used for this purpose. Even simple, low-power jamming can effectively disrupt real/authentic GNSS signals over a relatively large area [
4]. However, the effectiveness of these GNSS applications, particularly in critical maritime operations within the Black Sea, is threatened by several factors. Therefore, the effectiveness of jamming primarily depends on its transmission power and range (i.e., the distance to the target receiver). A nearby low-power jammer could have the same effect as a high-power jammer transmitting from a much greater distance. Other characteristics of jamming include the frequency bands in which it is emitted (narrowband or wideband) and whether the signal is transmitted as a continuous wave (either on a selected frequency or across multiple frequencies in a spectrum) or as pulsed signals at a lower power level (chirping) [
5].
The threats to GNSS signals, namely jamming and spoofing, represent only one category of factors that can degrade the accuracy and reliability of GNSS-based positioning. Atmospheric phenomena also play a significant role in GNSS signal propagation and must be carefully considered in any comprehensive GNSS monitoring system. The following sections will discuss the impact of these atmospheric effects, highlighting their relevance to the Black Sea region.
Maintaining the same focus on the quality and integrity of GNSS signals, the impact of atmospheric effects is a significant factor that can lead to errors in positioning and timing solutions. Two primary atmospheric phenomena that impact GNSS signal propagation are ionospheric scintillation and tropospheric delays. While users of Assured Positioning, Navigation, and Timing (APNT) solutions, particularly in military and safety-critical applications, have the most stringent requirements for GNSS signal integrity and are, therefore, a key focus when discussing GNSS vulnerabilities, it is important to emphasize that all GNSS users can be affected by interference. Ionospheric scintillation, for example, occurs due to irregularities in the ionosphere, which can cause rapid fluctuations in the amplitude and phase of GNSS signals. This phenomenon is particularly pronounced during periods of high solar activity, when the total electron content (TEC) in the ionosphere increases, resulting in more significant signal degradation [
6]. Jamming and spoofing, although potentially more targeted in specific scenarios, can also disrupt GNSS signals for civilian users, resulting in navigation errors or a complete loss of positioning. Research indicates that scintillation can lead to positioning errors exceeding 10 m, especially for low-elevation satellites, which are more susceptible to these disturbances. Furthermore, monitoring ionospheric conditions is essential to mitigate these effects effectively.
Tropospheric delays arise from the interaction of GNSS signals with the neutral atmosphere, which consists of water vapour, temperature, and pressure variations. The troposphere can introduce delays of up to 2.6 m in the zenith direction, significantly affecting GNSS positioning accuracy [
7]. The complexities of modelling tropospheric delays are well-documented, noting that variations in atmospheric conditions can lead to substantial errors in GNSS measurements.
Specifically, for the Black Sea region, monitoring Global Navigation Satellite System (GNSS) signal quality poses specific challenges due to geographical and environmental features.
To understand the factors affecting GNSS signal quality in the Black Sea region, it is essential to consider both atmospheric and interference effects and the specific applications of GNSS technology in this environment. One such application that is particularly relevant is the use of GNSS Interferometric Reflectometry (GNSS-IR) for monitoring sea levels. GNSS-IR utilizes reflected GNSS signals to derive information about the Earth’s surface, including sea level. Therefore, understanding the factors that affect the quality of both direct and reflected GNSS signals is crucial. While this paper focuses on a GNSS signal quality monitoring system designed to detect interference, the discussion of GNSS-IR provides essential context for the broader challenges of ensuring reliable GNSS positioning in the Black Sea.
One of the primary applications of GNSS technology in the Black Sea is monitoring sea level changes. For instance, Avsar and Kutoğlu (2020) utilized data from six GNSS stations near tide gauges to assess vertical land movements and seasonal sea level variations [
8]. Their findings indicated significant vertical land movements at specific tide gauge locations, which is crucial for understanding the region’s relative sea-level changes. This highlights the importance of accurate GNSS data for coastal monitoring applications.
The GNSS Interferometric Reflectometry (GNSS-IR) technique has emerged as a promising continuous sea-level monitoring method. Peng et al. (2019) highlighted the advantages of GNSS-IR in providing weather-independent sea level information while simultaneously detecting vertical land motions [
9]. A critical consideration in GNSS signal quality monitoring, particularly for applications such as GNSS-IR, is the potential difference in received signal power between direct and reflected signals. As GNSS-IR relies on reflected signals, which are inherently weaker, the system’s sensitivity to interference becomes a key concern. The monitoring system employed in this study utilizes a receiver with a high dynamic range, enabling it to effectively capture and analyze both direct and reflected signals within the same power spectrum.
The analysis of the power difference between direct and reflected signals in the Black Sea region indicates that distinct interference detection thresholds may not be essential. The existing literature provides substantial evidence regarding the power characteristics of GNSS signals, highlighting that the power levels of reflected signals tend to be lower than those of direct signals. For example, Larson et al. observed that the reflected GNSS signals maintained an approximately 3 dB lower signal-to-noise ratio (SNR) than direct signals in their study on coastal sea-level measurements using a single geodetic GPS receiver, which suggests that these power levels are comparable enough not to warrant separate thresholds for interference detection [
10]. This is pivotal since the core principle in GNSS Reflectometry (GNSS-R) is the ability to measure the time delay between the direct and reflected signals, which operate effectively as long as the SNR remains sufficiently high [
11]. However, further investigation into optimizing interference detection thresholds for GNSS-IR applications in diverse maritime environments remains a valuable area for future research.
This technique utilizes reflected GNSS signals to estimate sea level changes, making it particularly valuable in coastal areas where environmental conditions may hinder traditional measurement methods. The application of GNSS-IR in the Black Sea context is further supported by studies that demonstrate its effectiveness in detecting storm surges and other hydrodynamic phenomena [
9].
Moreover, the integration of GNSS data with numerical models has proven to be beneficial for understanding the dynamics of the Black Sea. For instance, Dorofeyev and Sukhikh (2017) employed atmospheric forcing functions in their modelling of Black Sea circulation dynamics, revealing the influence of various environmental factors on hydrophysical fields [
12]. This modelling approach, combined with GNSS observations, enhances the understanding of long-term changes in the Black Sea’s hydrodynamics, essential for effective maritime navigation and environmental management.
The quality of GNSS signals can be significantly affected by multipath effects and environmental conditions, particularly in maritime settings. Research by Wang et al. (2018) indicates that the use of dual-frequency GNSS signals can mitigate some of these issues, leading to improved sea level estimation accuracy [
13]. These studies highlight the importance of signal quality monitoring in ensuring the reliability of GNSS applications, particularly in the context of rising sea levels and coastal management in the Black Sea region.
While the preceding discussion has highlighted the diverse applications of GNSS technology in the Black Sea, it is essential to recognize that a complex interplay of factors affects the accuracy and reliability of GNSS signals in this region. These include atmospheric effects, unique oceanographic conditions, and unintentional and intentional interference sources. The subsequent sections will delve into these challenges, providing a comprehensive overview of the factors affecting GNSS signal quality in the Black Sea.
In addition to sea level monitoring, GNSS technology is increasingly being applied to assess wave energy potential in the Black Sea. Rusu (2015) conducted a comprehensive assessment of wave energy based on a 15-year hindcast, utilizing GNSS data to identify relevant energetic features and patterns [
14]. This research highlights the potential for GNSS technology to contribute to renewable energy assessments in the region, which is particularly relevant given the growing interest in sustainable energy sources. The impact of environmental factors on GNSS signal quality is also a critical area of investigation. Studies have shown that atmospheric conditions, such as humidity and temperature variations, can introduce errors in GNSS measurements [
15,
16].
Advanced algorithms and machine learning techniques are gaining traction in GNSS signal quality monitoring. Recent research has demonstrated the effectiveness of these approaches in enhancing the accuracy of sea surface wind speed retrieval and other geophysical parameters [
17]. It is also important to note that while jamming and spoofing are distinct forms of interference, they can sometimes be executed using similar tools, such as software-defined radios (SDRs). SDRs offer the flexibility to generate and transmit a wide range of radio signals, making them adaptable for jamming and spoofing purposes.
As the challenges posed by climate change and human activities continue to evolve, the role of GNSS technology in understanding and managing the Black Sea’s dynamic environment will become increasingly important.
Extensive research has examined the various factors that deteriorate GNSS signal quality. Several investigations have focused on characterizing atmospheric impairments, including analyses of ionospheric scintillations and associated delay models, as well as tropospheric delay effects, which are critical for understanding signal degradation [
18,
19]. In parallel, the development and deployment of dedicated monitoring networks, exemplified by the GNSS Finland Service, have demonstrated the value of continuous reference station data and real-time corrections in evaluating GNSS performance under coastal conditions [
20]. These initiatives have provided extensive datasets that validate innovative techniques, such as GNSS interferometric reflectometry (GNSS-IR), for coastal sea-level monitoring and beyond [
21].
Despite these significant advances, the literature reveals a persistent gap in developing integrated, multi-station systems capable of simultaneously analyzing atmospheric and oceanographic parameters in real-time. While prior studies have individually addressed the atmospheric challenges—emphasizing phenomena such as scintillation and delay modelling—and the oceanographic influences—through sea-level multipath and coastal refraction effects [
18,
21] —a holistic approach combining these factors has yet to be fully implemented. Recent comparative analyses between GNSS-derived sea-level measurements and traditional tide gauge data have reinforced that these techniques provide complementary insights into tidal dynamics and sea-level changes, enhancing our understanding of climate variability [
19,
22].
Additionally, operational practises established through the GNSS Finland initiative provide a valuable blueprint for extending integrated approaches to coastal regions, such as the Black Sea [
20]. Comparative investigations have shown that GNSS-derived sea-level measurements are consistent with conventional tide gauge data and contribute additional perspectives on tidal asymmetry and coastal refraction effects [
21,
22]. Beyond sea-level assessment, the GNSS Finland Service has successfully expanded the scope of GNSS applications to include remote sensing tasks such as soil moisture estimation and vegetation cover monitoring. Studies employing GNSS interferometric reflectometry in these domains have demonstrated that such techniques can seamlessly incorporate into routine monitoring frameworks, broadening environmental remote sensing applications without introducing significant processing complexity [
23].
Furthermore, while pulsed signals (commonly referred to as “chirping”) are one method for jamming, it is important to note that broadband jamming via alternative modulation techniques can also severely disrupt GNSS signals across a wide frequency range. This highlights the need for systems that enable real-time interference detection and signal quality analysis across multiple stations, particularly in the challenging maritime environment of the Black Sea [
20]. The present study introduces a novel multi-station monitoring system that integrates atmospheric parameters, including ionospheric and tropospheric conditions, with oceanographic influences such as sea surface multipath and coastal refraction. This integrated approach enhances the reliability of GNSS-based applications in coastal regions and improves our overall capability to develop real-time mitigation strategies for interference sources [
21,
22].
The effectiveness of GNSS applications, particularly in critical maritime operations within the Black Sea, is threatened by several factors. Current Global Navigation Satellite Systems (GNSSs) consist of various satellite positioning configurations that provide global or regional coverage. The radio frequency (RF) signals transmitted by GNSS satellites are essential for PNT (Positioning, Navigation, and Timing) across multiple industries and in the military domain, including land, air, and maritime applications, many of which require high precision and reliability.
2. Materials and Methods
2.1. GNSS Jamming Analysis
The increasing reliance on Global Navigation Satellite Systems (GNSSs) for critical infrastructure and applications has heightened concerns about their vulnerability to intentional or unintentional interference, known as jamming. GNSS jamming can disrupt the reception of GNSS signals, leading to positioning errors, loss of accuracy, and even complete denial of service. Understanding the characteristics and patterns of GNSS jamming is crucial for developing effective mitigation techniques and ensuring the resilience of GNSS-based systems.
To effectively address the challenges posed by GNSS jamming, it is essential to develop robust analysis techniques, as detailed in the following subsections. This section will detail the methodologies used to analyze GNSS jamming incidents, focusing on data collection, processing, and visualization techniques. It will describe the sources of jamming data, the methods used to process and filter the data, and the visualization techniques employed to represent the spatial and temporal distribution of jamming activities. The analysis will leverage data from various sources, including online jamming incident reports, GNSS monitoring networks, and satellite-based observations.
Figure 1, derived from data from [
24], presents a monthly overview of GNSS jamming incidents detected by commercial aircraft GNSS receivers over the Romanian coastline in the L1/E1 frequency band throughout the year 2024. These monthly maps reveal both spatial and temporal patterns in jamming activity. Spatially, the data suggest a concentration of jamming incidents along specific stretches of the coastline, particularly in the southern regions. This indicates the possibility of localized jamming sources or activities that are more prevalent in these areas. In contrast, other coastal areas experience less frequent or intense jamming, as indicated by the varying colour codes on the maps.
Temporally, the monthly progression reveals fluctuations in jamming frequency and intensity. Some months show widespread jamming affecting a significant portion of the coastline (e.g., March, June, August), suggesting periods of heightened jamming activity. In other months, the jamming appears to be more intermittent or confined to smaller areas (e.g., April, May, and October), indicating variations in the nature and extent of the interference.
It is important to note that aircraft-based detection, as shown in
Figure 1, provides a broader perspective on jamming events due to the high altitude and extended line of sight of aircraft GNSS receivers. This contrasts with ground-based monitoring systems, which typically have a more limited detection range. Consequently, this allows aircraft to detect jamming signals from greater distances, potentially including sources beyond the immediate coastal area, providing valuable regional context. In contrast, fixed ground stations have a more limited detection range, primarily capturing interference in their vicinity. Nevertheless, the broader jamming patterns observed from aircraft provide valuable regional context for interpreting the localized interference events detected by the ground-based monitoring system, aiding in assessing the overall threat level to maritime navigation.
Existing GNSS monitoring solutions include terrestrial networks (e.g., IGS) with anomaly detection and satellite-based augmentation systems (SBASs) like EGNOS and WAAS, which enhance GNSS accuracy and integrity. Emerging trends involve AI-based spoofing detection and hybrid infrastructure development. Next-generation user devices are also gaining interference detection capabilities.
2.2. Experimental Setup
This section details the experimental setup of the GNSS signal quality monitoring system, focusing on the scientific rationale behind its operational parameters and functions. The system is designed to continuously monitor GNSS signals and detect potential interference events, with specific functions activated based on predefined thresholds and conditions. The system’s interference detection capabilities are primarily driven by real-time analysis of signal power, spectral characteristics, and signal-to-noise ratio (SNR). For instance, the wideband interference (WBI) mitigation system is automatically engaged when the received signal power exceeds a predetermined threshold, indicating the presence of potential wideband interference. This threshold is dynamically adjusted based on background noise levels to minimize false positives while ensuring effective interference mitigation. Similarly, narrowband interference detection triggers the activation of notch filters when spectral analysis reveals distinct peaks exceeding a defined amplitude within specific frequency bands, suggesting the presence of jamming signals.
These functions’ specific thresholds and activation criteria are determined through manufacturer specifications, empirical testing, and established signal processing techniques to optimize detection accuracy and responsiveness. These parameters are crucial for accurately identifying and classifying interference events, which the system reports and logs. The system’s design primarily focuses on direct GNSS signals when applying these levels and thresholds to GNSS-IR. While the receiver’s high dynamic range allows for capturing both direct and reflected signals, the current interference detection mechanisms are optimized for direct signal characteristics. Further research is needed to investigate the specific requirements for GNSS-IR signal monitoring and whether separate thresholds or algorithms are necessary to account for the unique properties of reflected signals.
The Maritime Hydrographic Directorate (DHM) [
25] currently operates four permanent stations that provide DGPS and RTK corrections, located near the lighthouses of Sfântu Gheorghe (DHMS), Gura Portiței (DHMG), Midia (DHMM), and Tuzla (DHMT).
Figure 2 presents the positioning of GNSS receivers within the monitoring system. The primary role of these permanent stations is to transmit RTK (Real-Time Kinematics) corrections via UHF to offshore vessels, enabling them to reduce positioning errors to less than 10 cm.
In addition to this capability, the permanent stations are equipped with state-of-the-art Septentrio (Leuven, Belgium) PolaRx5e GNSS receivers, which integrate a spectrum analyzer and can store raw data for post-processing. This allows for recalculating positioning solutions using different combinations of GNSS constellations and frequencies. Additionally, these stations enable the determination of key performance parameters of global positioning systems, such as availability, continuity, and accuracy.
Figure 3 shows the receiver’s web interface, which provides information on receiver status, satellite tracking, and logging. The system is configured to log raw data for post-processing and real-time interference analysis.
Regarding interference detection and mitigation, Septentrio PolaRx5e receivers are equipped with an Advanced RF Interference Monitoring and Mitigation system (AIM+). Narrowband interference effects are mitigated using three-notch filters (band-rejection filter), which can be configured in either automatic or manual mode. These notch filters effectively remove a narrow portion of the RF spectrum around the interference signal. The L2 band, being open for amateur radio use, is particularly vulnerable to this type of interference. Both intentional and unintentional wideband interference effects can be mitigated by enabling the WBI (wideband interference) mitigation system. The WBI system also reduces the effects of pulsed interference more effectively than traditional PB (Pulse-Blanking) methods.
2.3. Data Analysis Methods
The signal processing techniques employed in the system extract relevant information from the raw GNSS signals to ensure accurate interference detection and signal quality assessment.
GNSS signal data collected from the Sfântu Gheorghe (DHMS), Midia (DHMM), and Tuzla (DHMT) stations, located along the North-Western Black Sea coast (
Figure 2) and observed between 5 April and 12 April 2025, were used to evaluate the system’s interference detection capabilities. This dataset was used to estimate the performance of the GNSS interference detection system in the Black Sea region.
2.3.1. Raw Data Acquisition
Operating across L1, L2, L5, E5a, E5b, and E6 frequency bands, Septentrio PolaRx5e GNSS receivers acquire high-rate raw signal data from GPS, GLONASS, Galileo, and BeiDou. These data are sampled at a high rate to ensure precise analysis and are recorded in a standardized SBF format.
2.3.2. Interference Detection
The system employs various algorithms to detect and classify different types of GNSS interference.
Power Monitoring: The received signal power is continuously monitored across different frequency bands. Any significant deviation from the expected power levels, either an increase or decrease, can indicate the presence of interference.
Spectral Analysis: The system performs spectral analysis to identify unusual patterns in the frequency domain. Narrowband interference, such as jamming signals, appears as distinct peaks in the spectrum. Wideband interference may manifest as a general rise in the noise floor across a wider frequency range.
Signal-to-Noise Ratio (SNR) Analysis: The SNR of the received signals is continuously monitored. A rapid drop in SNR can indicate the presence of interference, as the interfering signal masks the authentic GNSS signals.
By combining these techniques, the system detects and classifies various types of interference.
To assess signal quality, the system employs specific algorithms and metrics to detect and characterize ionospheric scintillation. This phenomenon, indicative of irregularities in ionospheric electron density, manifests as rapid fluctuations in both the amplitude and phase of GNSS signals. The detection process leverages the high-rate raw signal data the Septentrio PolaRx5e GNSS receivers acquired across L1, L2, L5, E5a, E5b, and E6 frequency bands. Amplitude fluctuations are quantified by calculating the rate of change in signal power over short time intervals, often expressed as the S4 index, which represents the normalized standard deviation of the received signal power. Phase fluctuations are assessed by computing the standard deviation of the detrended transporter phase. Scintillation events are flagged when these metrics exceed predefined thresholds, empirically determined based on established signal processing techniques and receiver specifications. It is important to emphasize that this scintillation analysis is performed independently of the positional data processing that produces latitude, longitude, and height information.
2.3.3. Signal Quality Assessment
The system utilizes several metrics to assess the quality of the received GNSS signals:
Carrier-to-Noise Ratio (C/N0): C/N0 is a fundamental measure of signal quality, representing the ratio of the carrier signal power to the noise power. Higher C/N0 values indicate stronger and more reliable signals.
Pseudorange Accuracy: The accuracy of the pseudorange measurements, which are the estimated distances between the receiver and the satellites, is a key indicator of signal quality. Errors in pseudorange measurements can be caused by various factors, including interference, atmospheric effects, and multipath.
Doppler Shift: The Doppler shift in the received signals is analyzed to assess the relative motion between the receiver and the satellites. Unusual Doppler shifts can indicate interference or other signal anomalies.
3. Results
This section presents the results of the GNSS signal monitoring system, focusing on its capabilities and the preliminary data analysis conducted during its implementation. The initial subsections describe the system’s features and data visualization tools, while subsequent subsections will present a more detailed analysis of observed GNSS signal characteristics and interference events.
The integration of data from receivers into the GNSS signal monitoring system considered their capability to detect and mitigate “jamming” and “spoofing” attacks, with their history archived in a log file. Thus, data regarding the detection time, central frequency, and bandwidth of the disturbing signal can be stored.
For preliminary analysis during the implementation process of the monitoring application, data sources from the GNSS receivers were integrated. The data included information regarding the monitored frequency spectra, which will be analyzed within the application’s modules to issue alerts in specific jamming cases.
Preliminary data can be visualized graphically for each receiver (for example: signal power at the receiver, spectrum processing histogram, etc.) based on the signal bands emitted by the GNSSs.
In histogram analysis, GNSS signals are typically represented in the complex domain (I/Q) to allow for more efficient processing, including demodulation and Doppler effect correction. The I (in-phase) component represents the signal’s real part, while the Q (quadrature-phase) component represents the imaginary part of the signal.
For example, in
Figure 4, at the Sf. Gheorghe station, the distribution shapes are approximately Gaussian, which is normal for an AWGN (Additive White Gaussian Noise) signal. The I component has a mean of −0.18, which means the values are slightly shifted toward the negative, and the Q component has a mean of 0.55, suggesting a slight asymmetry and a possible bias error (DC component). The RMS (Root Mean Square) values for I and Q are close (4.36 for I and 4.32 for Q), indicating a balanced energy distribution between the two components. The histogram shape suggests that the received signal is well-distributed and has normal statistical characteristics for a valid GNSS signal.
The subsequent analysis, facilitated by dedicated RxTools v24.0.0 GNSS interference monitoring software provides a more detailed examination of the system’s capabilities and the nature of observed interference events.
Regarding the values in the tables below the graphs, the specific parameters are as follows:
Gain (dB) (automatic gain) ensures automatic gain control, adjusting the signal amplification to maintain levels within the optimal range for processing. Therefore, the amplification values in the table range from 15 dB to 24 dB, which is normal for a GNSS receiver. The specific GLO L1 signal shows higher amplification (24 dB), indicating that its received signal was weaker, while for B3I and E6 signals, lower amplifications (around 15 dB) suggest that these signals were stronger at reception.
Sample Variance shows how much the signals vary for each monitored frequency spectrum. Relative to the amplification values, lower variance values can be observed for weaker signals at reception.
Following this, an additional analysis of these data is performed using a GNSS interference monitoring software, which provides the following capabilities:
A map, in
Figure 5, showing the layout of the four receivers that provide data on the quality of GNSS signals at their respective locations (
Figure 2), along with their status, as follows:
Green indicates the absence of interference;
Red indicates the presence of interference;
Orange indicates a lack of connection with the GNSS receiver;
Grey means that the station is not generating data or is not communicating with the GNSS orchestrator.
Figure 6 presents the time series analysis of GNSS positional divergences at the Sfântu Gheorghe (DHMS), Midia (DHMM), and Tuzla (DHMT) stations from 5 April to 12 April 2025. The y-axis represents the differences in latitude (Lat Diff), longitude (Long Diff), and height (Height Diff) in metres, while the x-axis shows the time in UTC. It is important to note that this figure displays the positional variations over a relatively short observation period of seven days, focusing on a period proposed to show the system’s stability under typical operating conditions. The analysis highlights that within this specific timeframe, the offsets in each coordinate component show minimal temporal variation, suggesting the presence of systematic rather than dynamic errors. These systematic errors are attributed to inaccuracies in the reference coordinates for calculating positional differences, internal receiver calibration errors, or unmodeled local effects that constantly distort signal reception. It is crucial to emphasize that while
Figure 6 illustrates the stability of positional measurements over this short period, it does not depict the system’s ability to detect transient interference phenomena, such as jamming. Jamming events, which induce increased variability or noise in the measurements, are detected through separate analyses using spectral analysis and SNR monitoring techniques, as detailed in
Section 2.3.2.
The time series analysis presented in
Figure 6 highlights a key characteristic: the near-constant nature of deviations in latitude, longitude, and height during the seven-day observation period. As shown, the offsets in each coordinate component display minimal temporal variation within this time frame, further supporting the interpretation that predominantly systematic errors occurred during this interval. This observation indicates that the discrepancies detailed in
Figure 6 do not indicate transient interference phenomena such as jamming, which would typically manifest as increased signal variability or noise. Instead, these persistent biases likely originate from factors affecting the inherent accuracy of the GNSS solutions.
Several potential sources may contribute to the systematic errors identified within the GNSS data. These sources include, but are not limited to, the above-mentioned inaccuracies in reference coordinates, internal calibration errors within the GNSS receivers, and unmodeled local effects that constantly distort signal reception. For example, if not adequately corrected, multipath propagation or atmospheric refraction can introduce dependable offsets in the computed positions. Such systematic errors highlight the importance of implementing calibration procedures and applying appropriate error correction methodologies to ensure the reliability of GNSS-derived positioning within the Black Sea monitoring network.
The consequences of these findings extend to the overall accuracy of GNSS-based applications within the region. While the system demonstrates the capacity to acquire GNSS signals, as illustrated in
Figure 6, uncorrected systematic biases can compromise the integrity of positional data utilized for critical applications such as maritime navigation, hydrographic surveying, and sea-level monitoring. Consequently, these errors necessitate mitigation through differential GNSS (DGNSS) methodologies or other suitable correction models. Further investigation into the precise origins of these biases is necessary to refine error strategies and enhance the accuracy of GNSS positioning along the Black Sea coast.
Figure 7a–c illustrates the temporal variation in 2D and 3D differences from the mean value for the GNSS measurements at the specified stations. A consistent observation across all three stations is the absence of any deviation from the mean for both 2D and 3D representations. Throughout the observed time span, both the 2D and 3D difference values remain precisely at zero, indicating that the measurements exhibit no displacement or offset from the calculated average. This suggests high stability and agreement with the mean value in the data’s horizontal and three-dimensional components.
The time series plots further reveal the time-invariant nature of the observed differences. The lack of fluctuation or trend in the 2D and 3D differences signifies that this precise agreement with the mean is maintained consistently over the measurement period. Moreover, the perfect overlap of the 2D and 3D difference lines in each subplot emphasizes that both representations’ deviations from the mean are identical. This concordance between 2D and 3D results implies that the vertical component of the measurements does not introduce any additional variability or displacement relative to the mean value, at least within the resolution of this analysis.
The finding of zero deviation from the mean in both 2D and 3D measurements carries significant implications for the accuracy and stability of the GNSS data at these stations. It suggests that the factors influencing the measurements, such as atmospheric effects or multipath interference, are either negligible or effectively mitigated by the processing techniques. However, it is crucial to acknowledge that this analysis reflects deviations from the mean and does not preclude the presence of systematic errors that might shift the mean itself. Further investigation, potentially involving comparison with independent reference data, would be necessary to assess the absolute accuracy of the GNSS positioning fully.
Figure 8 presents the deviations in latitude and longitude from their respective mean values for the GNSS measurements. A striking consistency is observed across all three subplots (
Figure 8a–c): the data points are concentrated precisely at the origin of the coordinate system. This indicates that the measured values exhibit no displacement from the calculated average for latitude and longitude. In other words, within the resolution of this analysis, the GNSS positions align perfectly with the mean position over the observation period at each station.
The spatial representation of the latitude and longitude differences further underscores the stability of the GNSS measurements. The absence of a scatter or dispersion of data points around the origin suggests minimal variability in the positional data relative to the mean. This lack of fluctuation implies that the factors influencing the measurements, such as noise or short-term disturbances, do not introduce any significant deviations in the horizontal coordinates. The consistent clustering at the origin reinforces the observation that the GNSS positions are highly stable with respect to their average values.
The finding of zero deviation from the mean for both latitude and longitude has essential implications for the positional accuracy of the GNSS. It suggests that the system can provide highly consistent horizontal positioning, at least in terms of repeatability relative to the mean. However, it is crucial to recognize that this analysis focuses on deviations from the mean and does not address potential systematic errors that could shift the entire cluster of measurements away from the true position.
Figure 9 exemplifies latitude, longitude, and ellipsoid height deviations from their mean values. A consistent pattern emerges across all three subplots (
Figure 9a–c): the data points are closely grouped at the origin (0, 0, 0) of the 3D coordinate space. This indicates that the measured positions exhibit minimal displacement from the calculated average position in all three spatial dimensions. The absence of any substantial offset from the origin suggests a strong agreement between individual GNSS measurements and the mean position, reflecting a high degree of consistency in the data.
The 3D representation of the data further emphasizes the stability of the GNSS measurements. The lack of any significant spread or dispersion of data points away from the origin in any subplot signifies minimal variability in latitude, longitude, or ellipsoid height relative to their respective mean values. This absence of fluctuation in the 3D space implies that the factors influencing the measurements do not introduce substantial deviations in positional components. The consistent clustering at the origin across all three stations or measurement scenarios reinforces the conclusion that the GNSS positions are highly stable and repeatable concerning their calculated averages.
The observation of minimal deviations from the mean in latitude, longitude, and ellipsoid height has important implications for the accuracy of the 3D positioning provided by the GNSS. It suggests the system can deliver reliable and repeatable positional information in all three spatial dimensions. However, it is essential to reiterate that this analysis is based on deviations from the mean position. While it demonstrates precision (repeatability) around the average, it does not exclude the possibility of systematic errors that could bias the entire 3D position.
Figure 6,
Figure 7,
Figure 8 and
Figure 9 analyze GNSS positional discrepancies observed at the DHMS, DHMM, and DHMT stations over the seven-day period from 5 April to 12 April 2025. Their primary function is to illustrate the stability of the GNSS positioning system and to characterize systematic errors in the positional data. Specifically,
Figure 6 shows time series of latitude, longitude, and height deviations, highlighting minimal variations indicative of consistent biases.
Figure 7,
Figure 8 and
Figure 9 further demonstrate the stability of the measurements around their mean values in 2D and 3D representations. While these figures might appear to depict instantaneous measurements, they represent data collected continuously, as emphasized by the UTC time scale on the x-axes. Consequently, they do not focus on capturing the full spectrum of GNSS signal dynamics, which includes rapid fluctuations and noise, nor do they depict the system’s capabilities for detecting dynamic interference events such as jamming or spoofing. Such events are identified through separate signal processing techniques (spectral analysis and SNR monitoring), as detailed in
Section 2.3.2. The presentation of the data involves an inherent averaging effect designed to highlight persistent biases over short-term variations, and the scale of the analysis is selected to emphasize overall trends in positional deviations. Therefore,
Figure 6,
Figure 7,
Figure 8 and
Figure 9 complement the broader results by establishing a baseline of system performance, isolating systematic errors, and providing context for interpreting interference detection results. The system employs complementary analytical techniques, such as spectral analysis and SNR monitoring (detailed in
Section 2.3.2), to capture and analyze natural noise and rapid signal variations.
Figure 10 presents the frequency distribution of latitude differences from the mean for the GNSS measurements. A consistent pattern is evident across all three subplots (
Figure 10a–c): the distributions exhibit a sharp, high-frequency peak precisely at a latitude difference of 0. This indicates that most latitude measurements align exactly with the calculated mean latitude. The pronounced peak at zero signifies that the occurrence of measurements coinciding with the average value is highly frequent, demonstrating a strong central tendency in the data.
The distributions are characterized by a distinguished lack of spread or dispersion away from the zero-deviation point. This absence of variability suggests that the measured latitude values are closely concentrated around the mean, with very few instances of significant deviations. As we move away from zero, the sharp peak and the rapid decrease in frequency emphasize the stability and consistency of the latitude measurements. This pattern implies that the factors influencing the measurements do not introduce substantial fluctuations or errors in the latitude component.
The observed frequency distribution, with its sharp peak at zero deviation, has important implications for the precision of the latitude measurements. It suggests that the GNSS provides highly repeatable latitude determinations with a strong tendency to converge on the mean value. However, it is key to highlight that this analysis focuses on the distribution of differences from the mean and does not address the absolute accuracy of the measurements.
Figure 11 presents the frequency distribution of longitude deviations from the mean for the GNSS measurements. Similarly to the latitude distributions, the plots reveal a distinct, high-frequency peak at a longitude difference of 0 across all three subplots (
Figure 11a–c). This indicates a strong tendency for longitude measurements to align precisely with the calculated mean longitude.
5. Conclusions
In summary, this study demonstrates the successful implementation of a GNSS signal quality monitoring system, providing valuable insights into the challenges and potential solutions for ensuring reliable GNSS-based maritime navigation.
Permanent GNSS stations can detect and report interference events, featuring multi-band capabilities and jamming and spoofing detection without being significantly affected by interference.
By implementing software components for DHM stations (receivers), the GNSS spectra recorded at their locations can be monitored in real-time, with alarm notifications sent to the user. Thus, the GNSS signal monitoring application can detect real-time interference in the following frequency bands with a bandwidth of 60 MHz: L5—1188 MHz, L2—1231 MHz, E6—1275 MHz, L1—1584 MHz.
In the graphical interface of the software product—a core component of the monitoring system—the following elements are displayed and transmitted for each station: station location, signal spectrum for each monitored band, occurrence of interference events in each monitored band, and alerts regarding interference events.
The findings of this study significantly enhance the existing body of knowledge on GNSS signal quality monitoring and interference mitigation, particularly within the understudied Black Sea region. The system’s ability to continuously monitor GNSS signals and analyze signal power and spectral characteristics reveal a complex interplay of potential interference sources and atmospheric effects. Specifically, the data analysis indicates that while jamming events are present, their occurrence does not always correlate with heightened military activity, suggesting the influence of other localized interference sources. Additionally, the system’s measurements showed only sporadic ionospheric scintillation, which is unexpected for the region and warrants further investigation.
Implementing a real-time monitoring system at the MHD addresses a critical need for improved navigation safety and security in an area increasingly susceptible to various forms of GNSS interference. The system’s real-time capabilities enable the continuous monitoring of the GNSS spectra and prompt the detection of interference events, which allows for the quick identification and mitigation of the interference’s impact. The system’s high detection rate and low false alarm rate highlight the effectiveness of the employed signal processing and statistical analysis techniques. These results are consistent with previous studies demonstrating the capabilities of similar monitoring systems in diverse geographical contexts [
55]. The accurate localization of interference sources further enhances the system’s value, enabling targeted mitigation efforts and contributing to a more secure and reliable navigation environment [
56].
The development of the monitoring system on a Docker-based platform reflects a growing trend towards utilizing transport technologies for increased flexibility, scalability, and maintainability [
57]. This approach aligns with the broader shift towards cloud-based GNSS monitoring and analysis solutions, offering advantages. The use of such modern technologies not only streamlines operations but also facilitates the integration of additional data sources, enhancing the overall effectiveness of the monitoring system.
Implementing the GNSS signal quality monitoring system aims to extract the necessary information to support decision making. Through the real-time analysis of signal parameters received from different GNSS systems (particularly GPS and Galileo), alerts are issued in critical situations, and specific statistics and reports can be generated. This improves the quality of command, control, communications, and intelligence decisions for planning and execution.
The study’s focus on the Black Sea region highlights the unique challenges and vulnerabilities associated with this semi-enclosed body of water. The findings contribute to a better understanding of the GNSS interference landscape in the region. Moreover, the system’s potential to support decision-making processes related to command, control, communications, and intelligence underscores its broader implications for maritime safety and security [
58].
A future development stage of this system could involve installing dedicated stations for locating interference sources, given that the system already provides indications about the affected area. Future research directions may include integrating additional data sources, such as meteorological data and ship tracking information, to further enhance the system’s capabilities. Investigating the potential of artificial intelligence and machine learning for automated interference classification and source identification could also lead to significant advancements in GNSS signal monitoring.