Multi-sensor performance consistency comprehensive evaluation method

Abstract

The invention provides a multi-sensor performance consistency comprehensive evaluation method. The method comprises the following steps: determining coverage range distribution of an intermodulation product in a UWB receiving frequency band; according to the overlapping degree of the intermodulation product coverage range distribution and the UWB receiving frequency band, identifying the interference intensity level, and when the interference intensity is higher than a preset interference threshold value, identifying that the global navigation satellite system intermodulation interference exists; when the intermodulation interference of the global navigation satellite system exists, acquiring an original packet loss event record, identifying an event of which the occurrence moment of the packet loss event coincides with the amplitude sudden increase moment of the intermodulation product, and marking the event as a false packet loss event; and separating the false packet loss event from the original packet loss event record, cleaning the packet loss event record, and counting the UWB packet loss rate again.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a comprehensive evaluation method for the performance consistency of multiple sensors. Background Technology

[0002] In multi-sensor fusion positioning systems, a comprehensive evaluation of the performance consistency of each sensor is crucial, as it directly determines the positioning stability and reliability of the entire system in complex environments. Global Navigation Satellite Systems (GNSS) and Ultra-Wideband (UWB) positioning technology, as two core sensors, are often used together to achieve seamless indoor and outdoor positioning. However, when evaluating the performance of these two types of sensors simultaneously, contradictory evaluation results often emerge, making it impossible to accurately judge the overall system performance. Current evaluation methods typically test the indicators of GNSS and UWB sensors separately, such as measuring the acquisition bandwidth of GNSS and the sensitivity and packet loss rate of UWB. While these methods can independently reflect the characteristics of a single sensor, they ignore the mutual interference of radio frequency signals generated when the two sensors operate simultaneously in the same device. This interference can cause cross-interference in originally independent performance indicators, leading to distorted evaluation results. The core challenge lies in the intermodulation products that easily arise between the leakage of the GNSS local oscillator signal and the UWB carrier signal. As the acquisition bandwidth of the GNSS increases, the local oscillator leakage power changes accordingly, and the frequency position and amplitude of the intermodulation products also shift accordingly. These intermodulation products are similar to spurious signals in radio frequency circuits. When a Global Navigation Satellite System (GNSS) expands its acquisition bandwidth to improve outdoor signal acquisition capabilities, the local oscillator leakage power fluctuates, causing shifts in the frequency position and amplitude of the intermodulation products. If these products fall into the ultra-wideband (UWB) operating frequency band, such as in an indoor environment with dense multipath propagation where the UWB pulse signal intended for precise positioning is overwhelmed, demodulation will fail. This interference is particularly problematic in real-world testing scenarios. For example, in a typical indoor-outdoor transitional testing environment, expanding the GNSS bandwidth, which should have improved satellite signal acquisition efficiency and positioning update rate, unexpectedly caused a sharp increase in the UWB packet loss rate because some data packets could not be correctly decoded due to intermodulation noise. When calculating the packet loss rate, testers could not distinguish which packet losses were due to multipath fading or antenna obstruction issues inherent in the UWB itself, and which were caused by GNSS leakage interference. This confusion further amplifies the evaluation discrepancies; while the GNSS performs excellently in outdoor-dominant scenarios, UWB performance deteriorates sharply when switching to indoor environments, resulting in low consistency scores. Similar problems also occur in high-dynamic scenarios, such as navigation devices, where vibration and temperature changes can exacerbate local oscillator leakage, making intermodulation interference more unstable. Summary of the Invention

[0003] This invention provides a comprehensive evaluation method for the performance consistency of multiple sensors, mainly including: Determine the coverage distribution of intermodulation products within the UWB receiving band; Based on the overlap between the coverage distribution of intermodulation products and the UWB receiving frequency band, the interference intensity level is identified, and the presence of global navigation satellite system intermodulation interference is indicated when the interference intensity is higher than a preset interference threshold. When intermodulation interference exists in the global navigation satellite system, the original packet loss event records are acquired, and events whose occurrence time coincides with the moment of sudden increase in the amplitude of intermodulation products are identified and marked as false packet loss events. False packet loss events are separated from the original packet loss event records. After cleaning the packet loss event records, the UWB packet loss rate is recalculated. By comparing the recalculated UWB packet loss rate with the joint positioning consistency evaluation standard, when the UWB packet loss rate is lower than the upper limit allowed by the consistency evaluation standard, the packet loss judgment threshold is adjusted to match the interference intensity level, thus obtaining the optimized packet loss judgment threshold. Obtain UWB sensitivity test results when the acquisition bandwidth of the Global Navigation Satellite System is expanded, apply an optimized packet loss judgment threshold to filter the UWB sensitivity test results, evaluate the performance consistency of multiple sensors, obtain confirmation of evaluation bias elimination, and generate a multi-sensor performance consistency evaluation report.

[0004] Furthermore, determining the coverage distribution of intermodulation products within the UWB receiving band includes: Scan the local oscillator signal of the global navigation satellite system and record the leakage power value of the local oscillator signal; Monitor the carrier center frequency and received signal strength within the ultra-wideband operating frequency band to obtain the ultra-wideband carrier power spectral density distribution; Based on the local oscillator leakage power value and the ultra-wideband carrier power spectral density, the frequency position of the third-order intermodulation products is calculated, and the frequency point distribution of the intermodulation products in the ultra-wideband receiving frequency band is determined. Extract the amplitude values ​​of the intermodulation products and mark the intermodulation components that exceed a preset threshold. The frequency domain occupancy of the intermodulation components in the ultra-wideband receiving frequency band is statistically analyzed. If the center frequency point deviates from the center frequency of the ultra-wideband channel by less than a preset frequency offset threshold, it is determined to fall within the band, and the power spectral density value of the in-band intermodulation products is accumulated. The average power density is calculated by the ratio of the accumulated value to the received bandwidth, and the coverage distribution is determined.

[0005] Furthermore, the step of identifying the interference intensity level based on the overlap between the intermodulation product coverage distribution and the UWB receiving frequency band, and indicating the existence of global navigation satellite system intermodulation interference when the interference intensity is higher than a preset interference threshold, includes: Extract the start and end frequencies of the intermodulation products in the frequency domain and compare them with the boundary frequencies of the ultra-wideband receiving frequency band to determine the overlapping frequency band. Calculate the bandwidth value of the overlapping frequency band and read the power density value of each frequency point within the overlapping frequency band; The average power density within the overlapping region is obtained by summing the power density values ​​and dividing by the bandwidth value. The average power density is compared with a preset sensitivity threshold to determine the interference intensity level; If the interference intensity level is higher than the preset interference threshold, an intermodulation interference presence indicator signal is generated, and the time of interference occurrence and the range of overlapping frequency bands are recorded.

[0006] Furthermore, it also includes: acquiring the frequency position and amplitude variation trajectory of intermodulation products, collecting the frequency distribution and duration of other interference sources in the UWB working environment, analyzing the response speed of intermodulation product amplitude as the local oscillator power of the global navigation satellite system changes, evaluating the degree of agreement between the frequency position of intermodulation products and the multiple of the local oscillator frequency, and identifying the differences in spectral purity between intermodulation products and environmental noise.

[0007] Furthermore, when intermodulation interference exists in the Global Navigation Satellite System, acquiring the original packet loss event records and identifying events whose occurrence time coincides with the moment of sudden increase in the amplitude of the intermodulation product are marked as false packet loss events includes: reading the original packet loss event records from the packet loss statistics module of the UWB receiver and extracting the occurrence timestamp and received signal strength value of each packet loss event; Obtain the time sequence where the amplitude of the intermodulation product exceeds the interference threshold from the intermodulation interference determination results of the Global Navigation Satellite System; For each packet loss event's timestamp, search the time series for moments of sudden increase in intermodulation amplitude that differ from the packet loss time by less than a preset time interval; If there is a sudden increase in the intermodulation amplitude within a preset time interval, it is determined that the packet loss event and the intermodulation interference have a temporal coincidence relationship. Further verification is performed on packet loss events with a time coincidence. The changes in UWB signal reception status before and after the time point are compared. If the UWB base station signal strength remains stable before and after the packet loss and no antenna blockage occurs, the packet loss is confirmed to be caused by intermodulation interference, and the packet loss event is marked as a false packet loss event.

[0008] Furthermore, when intermodulation interference exists in the Global Navigation Satellite System, acquiring original packet loss event records and identifying events where the timing of the packet loss coincides with the timing of a sudden increase in the amplitude of the intermodulation product are marked as false packet loss events, includes: Read the packet loss event log of the ultra-wideband receiver and extract the timestamp of the occurrence and the received signal strength value; Obtain the time series where the amplitude of the intermodulation product exceeds the interference threshold; Search for the moment when the difference between the timestamp of each packet loss event and the interference threshold is less than a preset time interval, where the intermodulation amplitude suddenly increases. If there is a sudden increase in the signal strength that coincides with the time, verify the changes in the ultra-wideband signal strength before and after packet loss; If the signal strength is stable and there is no antenna obstruction, the packet loss is confirmed to be caused by intermodulation interference and marked as a false packet loss event.

[0009] Furthermore, it also includes: obtaining the antenna attitude angle and distance to the obstruction at the time of each packet loss from the original packet loss event records, collecting the UWB signal propagation path and the number of reflective surfaces at the time of packet loss, analyzing the temporal overlap between the packet loss event and the time of obstruction by the moving object, assessing the correlation between the packet loss event and the change in indoor multipath reflection intensity, and identifying the set of packet loss events caused by a sudden increase in the amplitude of intermodulation products.

[0010] Furthermore, spurious packet loss events are separated from the original packet loss event records. After cleaning the packet loss event records, the UWB packet loss rate is recalculated, including: Traverse the original packet loss event records and compare them with the marked false packet loss events; If a packet loss event is marked as false, it is removed from the record, while the real packet loss events are retained, resulting in a cleaned record. The total number of remaining packet loss events is calculated based on the cleaned records. Obtain the total number of ultra-wideband data packets sent during the test period; The recalculated ultra-wideband packet loss rate is calculated by dividing the total number of remaining packet loss events by the total number of data packets sent.

[0011] Furthermore, by comparing the recalculated UWB packet loss rate with the joint positioning consistency evaluation standard, when the UWB packet loss rate is lower than the upper limit allowed by the consistency evaluation standard, the packet loss judgment threshold is adjusted to match the interference intensity level, resulting in an optimized packet loss judgment threshold, including: Compare the relationship between the re-statistical UWB packet loss rate and the allowed upper limit; If the UWB packet loss rate is lower than the allowable upper limit, calculate the difference between the packet loss rate and the upper limit. The adjustment range of the packet loss judgment threshold is determined based on the ratio of the difference to the interference intensity level. The original packet loss threshold value is modified according to the adjustment range to obtain the optimized packet loss threshold.

[0012] Furthermore, it also includes: analyzing the phenomenon that the positioning accuracy of GNSS sensors remains within the acceptable range when the packet loss rate of UWB sensors exceeds the standard; evaluating the temporal correspondence of the synchronous increase in packet loss rate of UWB sensors after the acquisition bandwidth of GNSS sensors is expanded; determining the opposing performance of UWB sensor performance degradation and GNSS sensor performance improvement in the time dimension; and identifying the contradiction in the performance trends of the two types of sensors presented in the evaluation report.

[0013] Furthermore, the UWB sensitivity test results are obtained when the acquisition bandwidth of the Global Navigation Satellite System is expanded. The optimized packet loss judgment threshold is applied to filter the UWB sensitivity test results, the performance consistency of multiple sensors is evaluated, the elimination of evaluation bias is confirmed, and a multi-sensor performance consistency evaluation report is generated, including: Acquire UWB sensitivity test data when the acquisition bandwidth of the Global Navigation Satellite System is expanded; The optimized packet loss judgment threshold is applied to filter the UWB sensitivity test data, and abnormal data points caused by intermodulation interference are removed to obtain a filtered UWB sensitivity test result set. The true UWB sensitivity is determined based on the filtered test result set. The positioning accuracy deviation values ​​of the Global Navigation Satellite System and UWB are obtained, and the performance consistency score of the multi-sensor system is determined by comparing the positioning coordinate differences between the two types of sensors at the same time. Based on the performance consistency score and the positioning accuracy deviation of each sensor, the difference between the improvement in positioning accuracy after the expansion of the global navigation satellite system bandwidth and the UWB accuracy loss caused by intermodulation interference is calculated to obtain the net contribution value. By integrating the actual UWB sensitivity, the consistency score, the net contribution value, and the performance indicators excluding the influence of intermodulation interference, a multi-sensor performance consistency evaluation report is generated.

[0014] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a comprehensive evaluation method for the performance consistency of multiple sensors. In GNSS and UWB sensor integrated devices, GNSS local oscillator leakage and its intermodulation products easily fall into the UWB receiving frequency band, leading to an abnormally high UWB packet loss rate, which in turn affects the accuracy of the multi-sensor performance consistency evaluation. This method collects GNSS local oscillator frequency, leakage power, and the location and amplitude of intermodulation products through radio frequency spectrum monitoring. It analyzes the overlap bandwidth, power density, and time-domain response characteristics in the UWB band to identify interference intensity and confirm the existence of intermodulation interference. Furthermore, it extracts false packet losses that match sudden increases in intermodulation amplitude from packet loss events, separates true packet losses by combining antenna attitude, obstruction, and multipath factors, and re-counts the UWB packet loss rate after cleaning. It dynamically adjusts the packet loss judgment threshold to adapt to the interference level, compares the performance trend contradictions when the GNSS bandwidth is expanded, applies the optimized threshold to filter test results, and finally eliminates evaluation bias, generating an accurate multi-sensor consistency report that highlights the net improvement in positioning accuracy brought about by the expansion of GNSS bandwidth without the impact of intermodulation interference. This method effectively solves the problem of intermodulation interference being misjudged as UWB performance degradation in joint positioning, and improves the reliability of the evaluation report and the overall accuracy of multi-sensor fusion positioning. Attached Figure Description

[0015] Figure 1This is a flowchart of a multi-sensor performance consistency comprehensive evaluation method according to the present invention.

[0016] Figure 2 This is a schematic diagram of a multi-sensor performance consistency comprehensive evaluation method according to the present invention.

[0017] Figure 3 This is another schematic diagram of a comprehensive evaluation method for the performance consistency of multiple sensors according to the present invention. Detailed Implementation

[0018] To further understand the content of this invention, a detailed description of the invention is provided in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0019] like Figures 1-3 This embodiment of a multi-sensor performance consistency comprehensive evaluation method may specifically include: Step S101: Determine the coverage distribution of intermodulation products within the UWB receiving frequency band.

[0020] The local oscillator (LO) signal is scanned within the global navigation satellite system (GNSS) operating frequency band using an RF spectrum monitoring unit. The leakage power value of the LO signal near the L1 frequency point is recorded. Simultaneously, the carrier center frequency and its received signal strength are monitored within the UWB operating frequency band to obtain the power spectral density distribution of the UWB carrier in the main channel. Based on the LO leakage power value and the UWB carrier power spectral density, the frequency positions of the third-order intermodulation products (IMPs) between the LO and UWB carrier frequencies are calculated. The frequency distribution of the IIPs within the UWB receiving frequency band is determined, and the amplitude values ​​of each IIP are extracted. IIP components exceeding a preset threshold are marked. For IIP components exceeding the preset threshold, their frequency domain occupancy within the UWB receiving frequency band is statistically analyzed. If the center frequency of an IIP deviates from the UWB channel center frequency by less than a preset frequency offset threshold, the IIP is considered to fall within the UWB receiving band. The power spectral density values ​​of all IIPs falling within the band are accumulated. The average power density of the intermodulation products in the UWB receiving band is calculated by using the ratio of the accumulated power spectral density value to the UWB receiving bandwidth. The specific formula is as follows: Where APD represents the average power density, TPSD represents the total accumulated power spectral density, and BW represents the UWB receiving bandwidth. Based on the distribution of average power density in each UWB channel, the coverage distribution of intermodulation interference within the UWB receiving frequency band is determined.

[0021] In one implementation, the RF spectrum monitoring unit employs a real-time spectrum analyzer architecture, using a combination of a broadband low-noise amplifier and a mixer to synchronously monitor the local oscillator signal and the UWB carrier signal of the Global Navigation Satellite System (GNSS). When the GNSS receiver performs down-conversion processing of the satellite signal, the local oscillator signal generated by its local oscillator can leak to the RF front-end through paths such as ground plane coupling on the printed circuit board or power line conduction. This leakage is particularly noticeable in compact positioning terminals integrating multiple sensors, as the RF modules share a power supply network and grounding system.

[0022] Specifically, the local oscillator leakage power value is obtained by setting a spectrum scanning window near the L1 frequency point. When the Global Navigation Satellite System (GNSS) receiver adjusts its acquisition bandwidth to adapt to different satellite signal strengths, the leakage power of the local oscillator signal dynamically changes. Simultaneously, the UWB transceiver transmits and receives pulse signals within its operating frequency band, and the carrier center frequencies of these pulse signals are distributed across multiple channels. The RF spectrum monitoring unit processes the acquired time-domain signals using Fast Fourier Transform (FFT) to obtain the power spectral density distribution at each frequency point, which includes the frequency domain characteristics of the local oscillator leakage component and the UWB carrier component. The generation mechanism of third-order intermodulation products stems from the nonlinear characteristics of the RF circuit. When the local oscillator frequency f1 and the UWB carrier frequency f2 are simultaneously present at the input of a nonlinear device, combined frequency components such as 2f1-f2 and 2f2-f1 are generated. In the actual operating environment of a multi-sensor fusion positioning system, the nonlinear characteristics of low-noise amplifiers, mixers, and analog-to-digital converters all contribute to the intermodulation products. The frequency positions of these intermodulation products depend on the frequency relationship of the original signals, while the amplitudes are related to the power level of the original signals and the third-order cutoff point of the devices. By pre-calibrating the third-order cutoff parameters of each device, the amplitude range of intermodulation products can be predicted based on the input signal power.

[0023] In one possible implementation, the process of determining whether intermodulation products fall within the UWB receive band involves spectral overlap analysis. Each channel of a UWB receiver has its specific center frequency and bandwidth, forming a frequency window. When the calculated intermodulation product frequency falls within this window, it will affect the signal reception of that channel. The determination process first identifies the center frequency of the intermodulation product, and then calculates the deviation of that frequency from the center frequency of the nearest UWB channel. If the deviation is less than a preset frequency offset threshold, for example, less than half the channel bandwidth, then the intermodulation product is considered to potentially interfere with UWB reception.

[0024] For example, in positioning tests conducted in transitional indoor-outdoor environments, when the positioning device moves from outdoors into a building, the Global Navigation Satellite System (GNSS) receiver automatically expands its acquisition bandwidth to capture more weak satellite signals. This bandwidth expansion leads to an increase in the local oscillator (LO) frequency scan range, and correspondingly, the spectral distribution of LO leakage also broadens. Assuming a LO frequency of 1575.42 MHz at a certain moment, and UWB operating on a 6.489 GHz channel, third-order intermodulation calculations may produce intermodulation products at 4.338 GHz. Although this frequency is not within the UWB main channel, its harmonic components or second-order intermodulation with other signals may fall into the UWB receiving band, forming in-band interference. This interference manifests as narrowband noise spikes in the spectrum, raising the noise floor level of the UWB receiver and reducing its ability to detect weak reflected signals.

[0025] Preferably, the process of accumulating the power spectral density values ​​of each intermodulation product falling within the receiving band employs an energy integration method. For each intermodulation product marked as falling within the receiving band, its power spectral density exhibits a certain distribution width in the frequency domain. By integrating this distribution, the total power contributed by the intermodulation product within the UWB receiving band is obtained. When multiple intermodulation products simultaneously fall within the receiving band, the power contributions of each intermodulation product are linearly superimposed to obtain the total interference power level.

[0026] For example, in positioning applications in underground parking lots, UWB base stations are densely deployed to achieve centimeter-level positioning accuracy. If the Global Navigation Satellite System (GNSS) receiver continues to operate to maintain time synchronization, intermodulation products generated by its local oscillator leakage will interfere with multiple UWB channels. Through real-time monitoring and calculation, a frequency domain coverage map can be generated, showing the degree of intermodulation interference on each UWB channel. The coverage distribution includes not only the frequency location of the interference but also the distribution density of the interference power on each channel. By dividing the accumulated power spectral density value by the UWB receiving bandwidth, the average interference power per unit frequency is obtained. This indicator directly reflects the impact of intermodulation interference on the UWB receiver sensitivity. When the average power density exceeds the interference tolerance threshold of the UWB receiver, it leads to an increase in the bit error rate and a decrease in ranging accuracy.

[0027] In one embodiment, the coverage distribution of intermodulation interference within the UWB receiving band is presented as a two-dimensional spectrum diagram. The horizontal axis represents frequency, the vertical axis represents power density, and different colors or grayscale levels represent interference intensity levels. This visualization method allows the system to intuitively identify which UWB channels are severely interfered with and which channels are relatively clean, thus providing a basis for subsequent channel selection and interference avoidance strategies. The dynamic update frequency of the coverage distribution is synchronized with the adjustment cycle of the global navigation satellite system's acquisition bandwidth, ensuring the real-time nature and accuracy of interference assessment.

[0028] Step S102: Based on the overlap between the coverage distribution of intermodulation products and the UWB receiving frequency band, identify the interference intensity level, and indicate the existence of global navigation satellite system intermodulation interference when the interference intensity is higher than a preset interference threshold.

[0029] Based on the coverage distribution of intermodulation products, the start and end frequencies of the intermodulation products in the frequency domain are extracted and compared with the upper and lower boundary frequencies of the UWB receiving band. The intersection of these frequencies is determined as the overlapping frequency band, and the bandwidth value of the overlapping frequency band is calculated. Simultaneously, the power density values ​​of each frequency point within the overlapping frequency band are read from the coverage distribution. By summing the power density values ​​of each frequency point and dividing by the bandwidth value, the average power density within the overlapping area is obtained. The average power density is compared with a preset UWB receiver sensitivity threshold to determine the interference intensity level. If the interference intensity level is higher than a preset interference threshold, an identification signal indicating the presence of global navigation satellite system intermodulation interference is generated. The time of interference occurrence, the overlapping frequency band range, and the interference intensity level value are recorded to form an intermodulation interference determination result.

[0030] In one implementation, the intermodulation product coverage distribution is stored as a spectral density map, where each frequency point corresponds to a power density value. The extraction process determines the start and end frequencies of the intermodulation products by scanning the frequency points in the distribution map where the power density exceeds the noise floor. The boundary frequencies of the UWB receiving band are pre-set according to the current operating channel, and the intersection interval is calculated as the overlapping frequency band by comparing the frequency range of the intermodulation products with the UWB band boundary.

[0031] Specifically, the bandwidth of the overlapping frequency band is obtained by subtracting the lower boundary frequency from the upper boundary frequency of the intersection interval. When only a portion of the intermodulation products fall within the UWB receiving frequency band, the overlapping bandwidth is less than the total bandwidth of the intermodulation products. The power density values ​​at each frequency point are read point by point from the coverage distribution according to the frequency index, forming a power density sequence within the overlapping area. This sequence reflects the energy distribution characteristics of intermodulation interference within the UWB receiving frequency band. The average power density is calculated using a numerical integration method. The values ​​in the power density sequence are accumulated to obtain the total power within the overlapping frequency band, and then divided by the bandwidth value of the overlapping frequency band to obtain the average power density per unit frequency. The UWB receiver sensitivity threshold is predetermined based on the receiver's noise figure and bandwidth, representing the minimum power density requirement for the receiver to demodulate the signal normally. When the average power density approaches or exceeds this threshold, it indicates that intermodulation interference has affected the normal reception of the UWB signal.

[0032] Preferably, the interference intensity level is quantified by the difference between the average power density and the sensitivity threshold.

[0033]

[0034] L I Indicates the level of interference intensity, PD avg The average power density is represented by , and ST represents the sensitivity threshold. This formula quantifies the interference intensity by expressing the difference between the average power density and the sensitivity threshold. The preset interference threshold is set to 3 dB above the sensitivity threshold. When the interference intensity level is higher than this threshold, it indicates that intermodulation interference has seriously affected UWB performance. The generation of the identification signal includes setting the interference flag to true and writing relevant parameters into the interference log, including the precise time of interference occurrence, the start and end frequencies of the overlapping frequency bands, and the calculated interference intensity value.

[0035] For example, in a vehicle positioning scenario within a tunnel, as a vehicle moves from the tunnel entrance into the depths of the tunnel, the Global Navigation Satellite System (GNSS) signal gradually weakens, and the system automatically expands its acquisition bandwidth in an attempt to maintain satellite lock. At this point, intermodulation products may happen to fall within the UWB operating frequency band; the aforementioned determination mechanism can promptly identify this interference.

[0036] The study aims to acquire the frequency location and amplitude variation trajectory of intermodulation products, collect the frequency distribution and duration of other interference sources in the UWB working environment, analyze the response speed of intermodulation product amplitude to changes in the local oscillator power of the global navigation satellite system, evaluate the degree of agreement between the frequency location of intermodulation products and the multiple of the local oscillator frequency, and identify the differences in spectral purity between intermodulation products and environmental noise.

[0037] The frequency position sequence and corresponding amplitude value sequence of intermodulation products are continuously acquired from the radio frequency spectrum monitoring unit. Frequency and amplitude data pairs are recorded at each sampling time to form an intermodulation product time-series dataset. Simultaneously, other signal components besides intermodulation products are scanned within the UWB operating frequency band, and the center frequency, bandwidth, and start and end times of each interference source are extracted. Based on the intermodulation product time-series dataset, the change in intermodulation product amplitude between adjacent time points is calculated, and the real-time monitoring value of the local oscillator power of the global navigation satellite system is obtained. The amplitude response ratio is obtained by dividing the change in intermodulation product amplitude by the change in local oscillator power. Simultaneously, the numerical relationship between the intermodulation product frequency value and the local oscillator fundamental frequency is extracted, and it is determined whether this relationship conforms to an integer multiple or fractional multiple characteristic to determine the degree of harmonic matching. A fast Fourier transform is performed on the amplitude sequence in the intermodulation product time-series dataset to identify the main frequency components of amplitude changes. If a significant fundamental frequency and its harmonic components exist, it is determined that the intermodulation product has a periodic occurrence characteristic, and the period value is recorded. The spectral concentration is obtained by calculating the ratio of the peak power to the bottom power of the intermodulation product spectrum. The power spectral density distribution of the ambient noise is collected, and the power fluctuation range of the ambient noise in the same frequency band is calculated. The spectral concentration is compared with the power fluctuation range of the ambient noise to identify the difference in spectral purity between the intermodulation products and the ambient noise.

[0038] In one implementation, the radio frequency spectrum monitoring unit employs a high-speed analog-to-digital converter in conjunction with a digital signal processor to achieve real-time monitoring of intermodulation products. The construction of the intermodulation product time-series dataset involves setting a fixed sampling interval, recording the instantaneous frequency value and corresponding amplitude value of the intermodulation product at each sampling moment. The sampling interval is determined based on the adjustment rate of the global navigation satellite system's acquisition bandwidth to ensure the capture of the dynamic changes in the intermodulation products. Simultaneously, a wideband sweep receiver performs a full-band scan of the UWB operating frequency band to identify and record other interference signals besides intermodulation products, including spurious radiation from mobile communication base stations, electromagnetic interference from industrial equipment, and intermodulation products from other wireless devices.

[0039] Specifically, the amplitude response ratio reflects the sensitivity of intermodulation products to changes in local oscillator power. In nonlinear devices, the amplitude of a third-order intermodulation product is theoretically proportional to the cube of the input signal power. This cubic relationship can be verified by real-time monitoring of changes in the local oscillator power of the Global Navigation Satellite System (GNSS) and simultaneously recording changes in the amplitude of the intermodulation products. If the amplitude of the intermodulation product increases by approximately 3 dB when the local oscillator power increases by 1 dB, it indicates that the intermodulation product is indeed a third-order intermodulation product caused by local oscillator leakage. Confirmation of this response relationship is crucial for distinguishing genuine intermodulation interference from other types of interference. Determining the degree of harmonic agreement involves a precise analysis of the mathematical relationship between the intermodulation product frequency and the local oscillator fundamental frequency. The frequency of the intermodulation product can typically be expressed in the form mf1 ± nf2, where f1 is the local oscillator frequency, f2 is the UWB carrier frequency, m and n are positive integers, and m + n equals the intermodulation order. The degree of harmonic agreement can be quantified by calculating the deviation between the measured frequency and the theoretically predicted frequency of the intermodulation product. When the deviation is less than the preset frequency tolerance range, it is considered that the frequency component is indeed an intermodulation product generated by the interaction between the local oscillator and the UWB signal, rather than an interference signal from other sources.

[0040] For example, in train positioning, trains experience periodic signal blockages while running in tunnels. Each time a train passes a support structure within the tunnel, the Global Navigation Satellite System (GNSS) receiver automatically adjusts its local oscillator power to compensate for signal attenuation. This periodic adjustment of the local oscillator power causes the amplitude of the intermodulation products to also exhibit periodic variations. By performing a Fast Fourier Transform (FFT) on the intermodulation product amplitude sequence, characteristic frequencies related to the train speed and the spacing between tunnel structures can be identified. If the FFT results show a significant fundamental frequency component, such as 0.5 Hz, corresponding to the train passing a support structure every 2 seconds, and its harmonic components such as 1 Hz and 1.5 Hz, it can be confirmed that the intermodulation products have stable periodic characteristics. The existence of this periodic characteristic further confirms the strong correlation between the intermodulation products and the system's operating state, rather than random noise.

[0041] Preferably, the spectral concentration is calculated by extracting the 3dB and 10dB bandwidths of the intermodulation product spectrum. As narrowband interference signals, the intermodulation products' energy is mainly concentrated within a very narrow frequency range, and the ratio of peak power to bottom power in the spectrum typically exceeds 20dB. In contrast, ambient noise exhibits broadband characteristics, with its power spectral density relatively uniformly distributed over a wider frequency range, and power fluctuations generally not exceeding 6dB.

[0042] In one possible implementation, the power spectral density distribution of ambient noise is obtained by acquiring the background spectrum with the UWB transmitter off. This background spectrum includes contributions from receiver thermal noise, external electromagnetic noise, and other unrelated signals. A noise baseline model is established by statistically analyzing the power distribution characteristics of ambient noise within the UWB operating frequency band.

[0043] For example, in personnel positioning scenarios, various motors and frequency converters on the production line generate broadband electromagnetic noise. This noise exhibits a relatively flat power distribution in the spectrum, occasionally accompanied by some power frequency harmonic spikes. In contrast, the intermodulation products generated by the local oscillator leakage of the Global Navigation Satellite System (GNSS) exhibit a distinct narrowband characteristic, appearing as sharp peaks in the spectrum. By comparing the differences in their spectral characteristics, it is possible to accurately identify which interferences originate from intermodulation products and which belong to ambient background noise.

[0044] Understandably, when intermodulation interference is confirmed and its spectral characteristics are significantly different from ambient noise, a targeted narrowband notch filter can be used to suppress intermodulation products without excessively affecting the receiving bandwidth of the UWB signal.

[0045] Step S103: When intermodulation interference exists in the global navigation satellite system, acquire the original packet loss event records, identify events whose occurrence time coincides with the moment of sudden increase in the amplitude of intermodulation products, and mark them as false packet loss events.

[0046] The system reads raw packet loss event records from the internal packet loss statistics module of the UWB receiver, extracts the timestamp and received signal strength value of each packet loss event, and simultaneously obtains the time sequence where the amplitude of intermodulation products exceeds the interference threshold from the intermodulation interference determination results of the Global Navigation Satellite System (GNSS). For each packet loss event's timestamp, the system searches the time sequence for moments in the intermodulation amplitude that are less than a preset time interval from the packet loss time. If a moment in the intermodulation amplitude increases within the preset time interval is found, the packet loss event is determined to have a temporal coincidence with intermodulation interference. The packet loss events with temporal coincidence are further verified by comparing the changes in UWB signal reception status before and after the moment. If the UWB base station signal strength remains stable before and after the packet loss and no antenna obstruction occurs, the packet loss is confirmed to be caused by intermodulation interference, and the packet loss event is marked as a false packet loss event.

[0047] In one implementation, the original packet loss event records are stored in the buffer of the UWB receiver. Each record contains a precise timestamp of the packet loss, the expected sequence number of the received data packet, the actual received signal strength indication value, and bit error rate information. The Global Navigation Satellite System intermodulation interference determination result comes from the aforementioned interference identification process, recording the start time and duration of each intermodulation product amplitude exceeding the interference threshold. A timestamp alignment mechanism enables time synchronization and comparison of the two types of events.

[0048] Specifically, the determination of time correlation employs a sliding time window method. The preset time interval is determined based on the transmission cycle of UWB data packets, typically set to twice the duration of the data packets. For each packet loss event, interference events within half a time interval before and after the intermodulation interference time sequence are searched. If a sudden increase in intermodulation amplitude exists within this time interval, a preliminary determination of temporal correlation is made. This temporal correlation reflects the real-time impact of intermodulation interference on the UWB reception process. The stability verification of the UWB signal reception state is achieved by analyzing the signal characteristics before and after packet loss. Specifically, this includes checking whether the UWB base station transmit power is constant, whether there are moving obstructions between the receiving antenna and the base station, and whether there are sudden changes in the multipath propagation environment. When these conditions remain stable, and packet loss occurs precisely at the moment of intermodulation interference, the possibility of packet loss caused by other reasons can be ruled out.

[0049] Preferably, further verification also includes analyzing the regularity of packet loss patterns. Packet loss caused by real signal fading or blockage usually exhibits continuous or gradual characteristics, while packet loss caused by intermodulation interference often shows sudden and discrete characteristics. This difference in characteristics can be identified by statistically analyzing the time interval distribution of adjacent packet loss events.

[0050] For example, in forklift positioning scenarios, forklifts encounter reflections and obstructions from metal shelves when moving between racks. When the forklift passes certain specific locations, the Global Navigation Satellite System (GNSS) receiver automatically increases its acquisition bandwidth to maintain positioning, resulting in a sudden increase in intermodulation product amplitude. If packet loss occurs in UWB at this time, by checking whether the forklift's location is in an open area and whether the UWB base station signal is stable, it can be determined that the packet loss is caused by intermodulation interference rather than shelf obstruction.

[0051] The antenna attitude angle and distance to obstructions at the time of each packet loss are obtained from the original packet loss event records. The propagation path of the UWB signal and the number of reflective surfaces at the time of packet loss are collected. The temporal overlap between the packet loss event and the time of obstruction by the moving object is analyzed. The correlation between the packet loss event and the change in indoor multipath reflection intensity is evaluated. The set of packet loss events caused by the sudden increase in the amplitude of intermodulation products is identified.

[0052] The antenna attitude angle values, including pitch, roll, and yaw angles, are extracted from the original packet loss event records at the moment of each packet loss. Simultaneously, the distance to the nearest obstruction at the moment of packet loss is obtained. The degree of signal obstruction is calculated based on the antenna pattern and the location of the obstruction, constructing a spatial obstruction feature sequence at the moment of packet loss. An electromagnetic propagation path calculation method is used to determine the direct and reflected paths of the UWB signal from the transmitting and receiving antennas at the moment of packet loss. The number of each path is counted, and the material properties and reflection coefficient of the reflecting surface are extracted. The power ratio of the direct component to the reflected component in the received signal is calculated based on the propagation loss of each path, forming a multipath feature sequence. By comparing the changes in the degree of obstruction in the spatial obstruction feature sequence with the packet loss event sequence, the temporal correlation between the two sequences is calculated. If the correlation exceeds a preset threshold, the packet loss event is determined to be caused by antenna obstruction. Simultaneously, the fluctuation range of the power ratio in the multipath feature sequence is compared with the correspondence between the packet loss occurrence to determine if the packet loss event is caused by multipath fading. Based on the determination results of antenna obstruction and multipath fading, a set of packet loss events caused by the physical propagation environment is determined. By comparing the records of sudden increases in intermodulation product amplitude, packet loss events that match the time but do not belong to the physical propagation environment are identified, and a set of packet loss events caused by sudden increases in intermodulation product amplitude is determined.

[0053] In one implementation, the antenna attitude angle is acquired using a three-axis IMU sensor integrated into the positioning terminal, which fuses measurement data from a three-axis accelerometer and a three-axis gyroscope. The pitch angle reflects the antenna's tilt in the vertical plane, the roll angle characterizes the antenna's rotation around its longitudinal axis, and the yaw angle describes the antenna's pointing in the horizontal plane. These three angle values ​​collectively determine the spatial pointing of the antenna's main lobe, thus affecting the effectiveness of signal reception. Obstruction distance is acquired through UWB ranging. These sensors can detect the distribution of objects within a certain range around the antenna and calculate the straight-line distance from the nearest obstruction object to the antenna. The construction of the spatial obstruction feature sequence involves the geometrical analysis of the antenna pattern and the spatial position of the obstruction object. The antenna pattern describes the antenna's radiation or reception capability in different spatial directions, typically expressed as a gain value. Signal obstruction is most severe when the obstruction is located in the direction of the antenna's main lobe; the obstruction is relatively mild when the obstruction is deviated from the main lobe but still within the side lobe range. By calculating the angle between the direction of the obstruction and the direction of maximum antenna gain, and combining the antenna gain value in that direction with the distance to the obstruction, the degree of signal obstruction can be quantitatively assessed. The spatial obstruction feature sequence records the obstruction degree value at each packet loss moment, forming time-series data. The electromagnetic propagation path is calculated using a geometrical optics approximation method, treating the UWB pulse signal as a ray propagating in a straight line. In an indoor environment, the signal may experience multiple propagation paths from the transmitting antenna to the receiving antenna: the direct path is the shortest path where the signal reaches the receiver directly without any reflection; the single-reflection path is where the signal reaches the receiver after being reflected by a single reflective surface such as a wall, floor, or ceiling; and the multiple-reflection path involves continuous reflection from two or more reflective surfaces. The propagation loss calculation formula for each path is as follows: Where L is the total loss (dB), d is the path length (m), and ri is the reflection coefficient of the i-th reflection (0-1). The longer the path and the more reflections, the more severe the signal attenuation. The material properties of the reflecting surface determine the reflection coefficient. For example, the reflection coefficient of a metal surface is close to 1, that of a concrete wall is about 0.5-0.7, while that of wood is only 0.2-0.3.

[0054] For example, the power ratio in a multipath feature sequence reflects the relative strength of the direct and reflected signals. In open environments, the direct component dominates, resulting in a larger power ratio; however, in complex indoor environments, the superposition of signals from multiple reflection paths can cause the total power of the reflected components to approach or even exceed that of the direct component, leading to a lower power ratio. When the phase relationship of the multipath signals causes destructive interference, the received signal strength experiences deep fading, at which point the fluctuation range of the power ratio increases significantly. By monitoring the characteristics of the power ratio changes over time, packet loss events caused by multipath fading can be identified.

[0055] In one possible implementation, temporal correlation is calculated through covariance analysis within a sliding window. For the spatial occlusion feature sequence and the packet loss event sequence, the covariance of the two sequences is calculated within each time window and normalized to obtain the correlation coefficient. A preset threshold is typically set to 0.7. When the correlation coefficient exceeds this threshold, it indicates that the change in the degree of occlusion and the packet loss event are highly synchronized in time, and it can be determined that the packet loss within this window is mainly caused by antenna occlusion.

[0056] Preferably, the determination of packet loss caused by multipath fading needs to comprehensively consider the fluctuation amplitude and duration of the power ratio. When the power ratio fluctuates drastically within a short period of time, and the fluctuation amplitude exceeds 50% of the average value, it indicates severe multipath interference. If the packet loss event happens to occur during these drastic fluctuations, it is determined to be packet loss caused by multipath fading.

[0057] For example, in mobile robot localization scenarios, robots encounter different propagation environments when navigating between aisle shelves. When the robot is at the end of the shelf, the UWB signal can reach the base station via a direct path, with minimal packet loss. However, as the robot moves deeper into the aisle, the narrow space created by the metal shelves generates a strong multipath effect, causing multiple reflected signals to superimpose or cancel each other out, resulting in rapid fluctuations in the received signal strength. By analyzing the packet loss patterns in this environment, packet loss events caused by multipath fading can be clearly distinguished.

[0058] It is understandable that the set of packet loss events caused by the physical propagation environment includes all packet loss caused by antenna obstruction or multipath fading. This type of packet loss has a clear physical cause and is closely related to the spatial location of the positioning terminal and its surrounding environment. In contrast, packet loss caused by sudden increases in the amplitude of intermodulation products is unrelated to the physical environment and originates from nonlinear effects within the radio frequency circuit.

[0059] In one embodiment, packet loss events caused by intermodulation products are identified using a timestamp matching method. For each packet loss event, it is checked whether its occurrence time coincides with the moment of sudden increase in the amplitude of the intermodulation product. If the time difference is within the millisecond range and the packet loss does not belong to the set of packet losses caused by the physical propagation environment, it is classified as a packet loss caused by intermodulation products.

[0060] Step S104: Separate false packet loss events from the original packet loss event records, clean the packet loss event records, and recalculate the UWB packet loss rate.

[0061] Each packet loss event is iterated through from the original packet loss event record. Each event is compared to the previously marked false packet loss events. If an event is marked as false, the entry is removed from the original record, and the unmarked true packet loss events are retained, resulting in a cleaned packet loss event record. Based on this cleaned record, the total number of remaining packet loss events is calculated, and the total number of UWB data packets sent during the corresponding test period is obtained. The remaining total number of packet loss events is divided by the total number of data packets sent to obtain the recalculated UWB packet loss rate.

[0062] In one implementation, the separation of spurious packet loss events is achieved by establishing a tag index table. This index table records the timestamp and sequence number of each packet loss event marked as spurious, serving as a reference for the separation operation. When traversing the original packet loss event records, the identification information of each record is matched against the index table; records that match are identified as spurious packet loss and removed from the original records.

[0063] Specifically, the cleaned packet loss event records retain all genuine physical layer packet loss and packet loss caused by environmental factors, while eliminating spurious packet loss caused by intermodulation interference. This separation operation ensures the accuracy of subsequent statistics, allowing the true performance of the UWB system to be reflected. The recalculation of the packet loss rate requires a clear definition of the start and end times of the test period. The total number of data packets sent is obtained by accumulating all successfully sent and lost data packets within that period. The ratio of the remaining total number of packet loss events to the total number of data packets sent is the corrected packet loss rate, which accurately reflects the transmission reliability of the UWB system after eliminating the impact of intermodulation interference.

[0064] For example, in the performance test of an indoor positioning system, the original packet loss rate may reach 5%, but after spurious packet loss separation, the actual packet loss rate is only 2%, indicating that 3% of the packet loss is caused by intermodulation interference from the Global Navigation Satellite System, rather than a problem with the UWB system itself.

[0065] Step S105: Compare the recalculated UWB packet loss rate with the joint positioning consistency evaluation standard. If the UWB packet loss rate is lower than the upper limit allowed by the consistency evaluation standard, adjust the packet loss judgment threshold to match the interference intensity level to obtain an optimized packet loss judgment threshold.

[0066] A joint positioning consistency evaluation standard between UWB and the Global Navigation Satellite System is obtained. An allowable upper limit for UWB packet loss rate is extracted from the standard. The recalculated UWB packet loss rate is read, and its magnitude is compared with the allowable upper limit. If the UWB packet loss rate is lower than the allowable upper limit, the difference between the packet loss rate and the upper limit is calculated. Based on the ratio of this difference to the previously identified interference intensity level, the adjustment range of the packet loss judgment threshold is determined. According to the adjustment range, the original packet loss judgment threshold value is modified to match the interference intensity level. The rationality of the adjusted threshold is verified by re-testing historical packet loss data, resulting in an optimized packet loss judgment threshold.

[0067] In one implementation, the joint positioning consistency evaluation standard is determined through industry specifications or system design indicators, which specifies the upper limit of the allowable UWB packet loss rate under intermodulation interference. Indoor high-precision positioning applications typically require a packet loss rate of no more than 3%, while outdoor assisted positioning has a relatively higher tolerance, allowing a packet loss rate of 5%. The standard also includes minimum requirements for the data quality of each sensor in the sensor fusion algorithm, ensuring that other sensors can maintain positioning continuity when intermodulation interference causes a degradation in the performance of one sensor. The ratio of the difference to the interference intensity level reflects the system's interference tolerance. When the recalculated packet loss rate is 1.5%, and the allowable upper limit is 3%, there is a 1.5% margin. If the current interference intensity level is -80dBm, the sensitivity coefficient for threshold adjustment can be determined by calculating the ratio of the margin to the interference intensity. This coefficient characterizes the degree of influence of a unit change in interference intensity on the packet loss judgment threshold. The adjustment range of the packet loss judgment threshold needs to consider the system's stability requirements. An excessively large adjustment range may cause real packet loss to be misjudged as normal reception, while an excessively small adjustment range will not fully utilize the system's fault tolerance margin. Through formula The adjustment range is calculated, where A is the adjustment range, M is the margin, I is the interference intensity, and K is an empirical coefficient of 0.1. This ensures both detection sensitivity and avoids overcompensation. The adjusted threshold value enables the differentiation between signal quality degradation caused by intermodulation interference and true channel fading, improving the accuracy of packet loss detection.

[0068] Preferably, historical data verification is achieved by reprocessing received signal data from the previous time period. The adjusted threshold is applied to this historical data, the packet loss detection results under the new threshold are statistically analyzed, and compared with the actual communication performance. If the new threshold can correctly identify most real packet loss events while reducing false alarms of packet loss, the threshold adjustment is considered reasonable.

[0069] For example, in a forklift positioning system, when the Global Navigation Satellite System (GNSS) increases its acquisition bandwidth due to obstruction by metal shelving, the intermodulation interference intensity rises to -75 dBm. By dynamically adjusting the UWB packet loss detection threshold, this change in the interference environment can be adapted to, avoiding performance evaluation biases caused by fixed thresholds.

[0070] This study analyzes the phenomenon that the positioning accuracy of GNSS sensors remains within acceptable limits when the packet loss rate of UWB sensors exceeds the standard, evaluates the temporal correspondence of the synchronous increase in packet loss rate of UWB sensors after the acquisition bandwidth of GNSS sensors is expanded, determines the opposing performance of UWB sensor performance degradation and GNSS sensor performance improvement in the time dimension, and identifies the contradiction in the performance trends of the two types of sensors presented in the evaluation report.

[0071] The UWB positioning availability threshold and GNSS positioning accuracy requirements are extracted from the Joint Positioning Consistency Assessment Standard. The packet loss rate statistics of the UWB sensors during the test period are obtained, and the continuity status indicator and positioning result time interval of the GNSS sensor positioning output are collected simultaneously. Based on the comparison between the packet loss rate statistics and the availability threshold, the time period in which the UWB packet loss rate exceeds the threshold is identified. Within this time period, the GNSS positioning accuracy is checked to see if it meets the accuracy requirements. If the GNSS accuracy is still within the acceptable range, the moment when UWB packet loss exceeds the standard but GNSS accuracy is acceptable is recorded. The change records of the GNSS sensor acquisition bandwidth are extracted to obtain the starting time of bandwidth expansion. Comparing this starting time with the moment when UWB packet loss exceeds the standard, the standard deviation of the time difference sequence is calculated to determine the temporal correspondence between GNSS bandwidth adjustment and UWB packet loss change. Based on this temporal correspondence, the number of times UWB performance index decline and GNSS performance index increase occur simultaneously is counted. The proportion of opposite directions of performance change for the two types of sensors is calculated to identify the contradictory performance trends presented by the two types of sensors in the assessment report.

[0072] In one implementation, the joint positioning consistency evaluation standard establishes differentiated performance indicators based on different application scenarios. The UWB positioning availability threshold is typically set at a packet loss rate not exceeding a predetermined percentage; this threshold comprehensively considers positioning update rate requirements and ranging accuracy needs. GNSS positioning accuracy requirements include two dimensions: horizontal accuracy and vertical accuracy. Horizontal accuracy is typically required to be within the meter range, while vertical accuracy allows for relatively larger errors. These standard parameters are stored in the system configuration file as a benchmark for performance evaluation.

[0073] Specifically, the packet loss rate statistics are obtained by comparing the number of packet losses within a fixed time window with the total number of packets sent. The actual packet loss rate of the UWB sensor is recorded for each statistical period, forming time-series data. The continuous status indicators of the GNSS sensor include positioning solution status, number of visible satellites, and geometric accuracy factor. The positioning result time interval reflects the update frequency of the GNSS receiver's output positioning coordinates. Under normal circumstances, this interval remains stable, but it may lengthen or change irregularly when signal quality deteriorates. The phenomenon of excessive UWB packet loss while GNSS accuracy is acceptable reveals an inherent contradiction in the performance evaluation of the two types of sensors. This phenomenon is identified by setting dual judgment conditions: first, it checks whether the UWB packet loss rate exceeds the availability threshold; then, at the same time, it verifies whether the GNSS positioning accuracy still meets the requirements. When both conditions are met simultaneously, this moment is recorded as the point where the performance contradiction occurs. The root cause of this contradiction lies in the fact that the Global Navigation Satellite System expands its acquisition bandwidth to maintain positioning accuracy, and the resulting intermodulation interference affects the normal operation of UWB, but the performance of GNSS itself is improved due to the increased bandwidth.

[0074] For example, changes in GNSS acquisition bandwidth are recorded by monitoring receiver configuration parameters. When the bandwidth increases from the standard 2MHz to 4MHz or 8MHz, the local oscillator leakage power increases accordingly, and the probability of the resulting intermodulation products falling into the UWB operating frequency band also rises. The start time of the bandwidth increase is determined by the timestamp of the detected configuration parameter changes. Comparing this moment with the moment of a sudden increase in UWB packet loss rate, if the time difference is within milliseconds, it indicates a direct causal relationship.

[0075] In one possible implementation, the standard deviation of the time difference sequence is used to quantify the stability of the temporal correspondence. For multiple GNSS bandwidth adjustment events, their occurrence times and corresponding UWB packet loss rate changes are recorded to form a time difference sequence. The mean and standard deviation of this sequence are calculated; a smaller standard deviation indicates a more stable temporal correspondence and a stronger correlation between the performance changes of the two types of sensors. When the standard deviation is less than a preset threshold, it can be confirmed that the GNSS bandwidth adjustment is the direct cause of the UWB performance degradation.

[0076] Preferably, the statistical analysis of performance change direction is achieved by establishing a two-dimensional performance index space. The horizontal axis represents the UWB packet loss rate, and the vertical axis represents the GNSS positioning accuracy. Each sampling moment corresponds to a point in this space. When the system transitions from one state to another, a performance change vector is formed. If the vector points in the direction of performance deterioration in the UWB dimension and in the direction of performance improvement in the GNSS dimension, then the performance change directions of the two types of sensors are considered to be opposite. The proportion of opposite directions to the total number of changes during the entire test period is statistically analyzed. When this proportion exceeds a set threshold, a systematic performance trend contradiction is confirmed.

[0077] For example, in vehicle navigation applications, as a vehicle travels from an entrance to deeper underground, the GNSS signal gradually weakens. To maintain basic positioning functionality, the GNSS receiver automatically expands its acquisition bandwidth from 2MHz to 8MHz in an attempt to capture more of the weak satellite signals. This adjustment does improve the availability of GNSS in weak signal environments, keeping its positioning accuracy within 5 meters. However, the expanded bandwidth leads to a 6dB increase in local oscillator leakage power, and the resulting intermodulation products fall precisely within the 6.5GHz operating frequency band of UWB. The UWB packet loss rate jumps sharply from the normal 1% to 8%, significantly exceeding the 3% availability threshold.

[0078] Understandably, this contradictory performance trend manifests in the evaluation report as two opposing curves: GNSS performance indicators show an upward trend, while UWB performance indicators show a downward trend. Traditional independent evaluation methods would conclude that GNSS performance is excellent and UWB performance is poor, thus obscuring the true interference mechanism.

[0079] Step S106: Obtain the UWB sensitivity test results when the acquisition bandwidth of the Global Navigation Satellite System is expanded, apply the optimized packet loss judgment threshold to filter the UWB sensitivity test results, evaluate the performance consistency of multiple sensors, obtain confirmation of evaluation deviation elimination, and generate a multi-sensor performance consistency evaluation report.

[0080] UWB sensitivity test data was acquired when the bandwidth of the Global Navigation Satellite System (GNSS) was expanded. This data included received signal strength indication (RSI), bit error rate (BER), and packet reception success rate (RSR). An optimized packet loss threshold was applied to filter the test data, eliminating abnormal data points caused by intermodulation interference (IMI), resulting in a filtered UWB sensitivity test result set. The true UWB sensitivity was determined based on the minimum RSI received signal strength in the filtered result set. Simultaneously, the positioning accuracy deviation between the GNSS and UWB sensors was acquired. By comparing the positioning coordinate differences between the two types of sensors at the same time, a multi-sensor performance consistency score was determined. Based on the performance consistency score and the positioning accuracy deviation of each sensor, the improvement in positioning accuracy after the GNSS bandwidth expansion was subtracted from the UWB accuracy loss caused by IIM to obtain the net contribution value. The true UWB sensitivity, consistency score, net contribution value, and performance indicators excluding IIM's influence were integrated to generate a multi-sensor performance consistency evaluation report.

[0081] In one implementation, UWB sensitivity test data is collected under specific conditions of expanded bandwidth in global navigation satellite systems. The test data includes performance metrics across multiple dimensions, where the received signal strength indicator reflects the weakest signal level detectable by the UWB receiver, the bit error rate characterizes the reliability of data transmission, and the packet reception success rate comprehensively reflects the stability of the communication link. These raw data contain outliers caused by intermodulation interference, which need to be identified and removed using an optimized packet loss threshold.

[0082] Specifically, the filtering process for application optimization thresholds is achieved by setting a dynamic threshold range. When the signal strength value of a test data point is below the normal range but the packet loss rate is abnormally high, the system determines that the data point is affected by intermodulation interference. By removing these abnormal data points, the test data that truly reflects UWB performance is retained, forming a filtered result set. The true sensitivity of UWB is determined by statistically analyzing the minimum stable value of the received signal strength in the filtered result set, which represents the actual detection capability of the UWB receiver in the absence of intermodulation interference. The determination of the multi-sensor performance consistency score involves the spatiotemporal comparison of positioning coordinates. At the same test moment, the Global Navigation Satellite System and UWB output their respective positioning results. Spatial consistency is evaluated by calculating the Euclidean distance between the two positioning coordinates. Temporal consistency is measured by comparing the timestamp differences in positioning updates from the two types of sensors. The spatial and temporal deviations are weighted together to obtain a quantified consistency score.

[0083] Preferably, the net contribution value is calculated using a differential method. First, the overall improvement in positioning accuracy after the expansion of the Global Navigation Satellite System (GNSS) bandwidth is determined. This includes the combined effects of improved signal acquisition capabilities and optimized positioning algorithms resulting from the increased bandwidth. Then, the UWB positioning accuracy loss due to intermodulation interference is quantified, primarily manifested as increased ranging error and decreased positioning update rate. Subtracting the accuracy loss from the overall improvement yields the net contribution value after excluding the interference effects.

[0084] For example, in vehicle positioning tests, after the bandwidth of the Global Navigation Satellite System (GNSS) was increased from 2MHz to 8MHz, its positioning accuracy improved from 10 meters to 5 meters. However, UWB, due to intermodulation interference, saw its positioning accuracy deteriorate from 0.3 meters to 0.8 meters. Calculations showed a net contribution of 3.5 meters, indicating that despite the interference, the overall positioning capability was still improved.

[0085] In one embodiment, the evaluation report integrates various performance indicators and analysis results, including the true UWB sensitivity value, multi-sensor consistency score, net contribution value, and performance comparison charts of each sensor under different operating conditions.

[0086] The above are only some preferred embodiments of the present invention, but the present invention is not limited thereto, and many improvements and modifications can be made. Any improvements and modifications made based on the basic principles of the present invention should be considered to fall within the protection scope of the present invention.

Claims

1. A comprehensive evaluation method for the performance consistency of multiple sensors, characterized in that, The method includes: Determine the coverage distribution of intermodulation products within the UWB receiving band; Based on the overlap between the coverage distribution of intermodulation products and the UWB receiving frequency band, the interference intensity level is identified, and the presence of global navigation satellite system intermodulation interference is indicated when the interference intensity is higher than a preset interference threshold. When intermodulation interference exists in the global navigation satellite system, the original packet loss event records are acquired, and events whose occurrence time coincides with the moment of sudden increase in the amplitude of intermodulation products are identified and marked as false packet loss events. False packet loss events are separated from the original packet loss event records. After cleaning the packet loss event records, the UWB packet loss rate is recalculated. By comparing the recalculated UWB packet loss rate with the joint positioning consistency evaluation standard, when the UWB packet loss rate is lower than the upper limit allowed by the consistency evaluation standard, the packet loss judgment threshold is adjusted to match the interference intensity level, thus obtaining the optimized packet loss judgment threshold. Obtain UWB sensitivity test results when the acquisition bandwidth of the Global Navigation Satellite System is expanded, apply an optimized packet loss judgment threshold to filter the UWB sensitivity test results, evaluate the performance consistency of multiple sensors, obtain confirmation of evaluation bias elimination, and generate a multi-sensor performance consistency evaluation report.

2. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, Determining the coverage distribution of intermodulation products within the UWB receiving frequency band includes: Scan the local oscillator signal of the global navigation satellite system and record the leakage power value of the local oscillator signal; Monitor the carrier center frequency and received signal strength within the ultra-wideband operating frequency band to obtain the ultra-wideband carrier power spectral density distribution; Based on the local oscillator leakage power value and the ultra-wideband carrier power spectral density, the frequency position of the third-order intermodulation products is calculated, and the frequency point distribution of the intermodulation products in the ultra-wideband receiving frequency band is determined. Extract the amplitude values ​​of the intermodulation products and mark the intermodulation components that exceed a preset threshold. The frequency domain occupancy of the intermodulation components in the ultra-wideband receiving frequency band is statistically analyzed. If the center frequency point deviates from the center frequency of the ultra-wideband channel by less than a preset frequency offset threshold, it is determined to fall within the band, and the power spectral density value of the in-band intermodulation products is accumulated. The average power density is calculated by the ratio of the accumulated value to the received bandwidth, and the coverage distribution is determined.

3. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, The step of identifying the interference intensity level based on the overlap between the intermodulation product coverage distribution and the UWB receiving frequency band, and indicating the presence of global navigation satellite system intermodulation interference when the interference intensity exceeds a preset interference threshold, includes: Extract the start and end frequencies of the intermodulation products in the frequency domain and compare them with the boundary frequencies of the ultra-wideband receiving frequency band to determine the overlapping frequency band. Calculate the bandwidth value of the overlapping frequency band and read the power density value of each frequency point within the overlapping frequency band; The average power density within the overlapping region is obtained by summing the power density values ​​and dividing by the bandwidth value. The average power density is compared with a preset sensitivity threshold to determine the interference intensity level; If the interference intensity level is higher than the preset interference threshold, an intermodulation interference presence indicator signal is generated, and the time of interference occurrence and the range of overlapping frequency bands are recorded.

4. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, Also includes: The study aims to acquire the frequency location and amplitude variation trajectory of intermodulation products, collect the frequency distribution and duration of other interference sources in the UWB working environment, analyze the response speed of intermodulation product amplitude to changes in the local oscillator power of the global navigation satellite system, evaluate the degree of agreement between the frequency location of intermodulation products and the multiple of the local oscillator frequency, and identify the differences in spectral purity between intermodulation products and environmental noise.

5. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, When intermodulation interference exists in the Global Navigation Satellite System, the original packet loss event records are acquired, and events whose occurrence time coincides with the moment of sudden increase in the amplitude of intermodulation products are identified and marked as false packet loss events, including: Read the original packet loss event records from the packet loss statistics module of the UWB receiver, and extract the timestamp of each packet loss event and the received signal strength value; Obtain the time sequence where the amplitude of the intermodulation product exceeds the interference threshold from the intermodulation interference determination results of the Global Navigation Satellite System; For each packet loss event's timestamp, search the time series for moments of sudden increase in intermodulation amplitude that differ from the packet loss time by less than a preset time interval; If there is a sudden increase in the intermodulation amplitude within a preset time interval, it is determined that the packet loss event and the intermodulation interference have a temporal coincidence relationship. Further verification is performed on packet loss events with a time coincidence. The changes in UWB signal reception status before and after the time point are compared. If the UWB base station signal strength remains stable before and after the packet loss and no antenna blockage occurs, the packet loss is confirmed to be caused by intermodulation interference, and the packet loss event is marked as a false packet loss event.

6. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, Also includes: The antenna attitude angle and distance to obstructions at the time of each packet loss are obtained from the original packet loss event records. The propagation path of the UWB signal and the number of reflective surfaces at the time of packet loss are collected. The temporal overlap between the packet loss event and the time of obstruction by the moving object is analyzed. The correlation between the packet loss event and the change in indoor multipath reflection intensity is evaluated. The set of packet loss events caused by the sudden increase in the amplitude of intermodulation products is identified.

7. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, The process of separating false packet loss events from the original packet loss event records, cleaning the packet loss event records, and then recalculating the UWB packet loss rate includes: Traverse the original packet loss event records and compare them with the marked false packet loss events; If a packet loss event is marked as false, it is removed from the record, while the real packet loss events are retained, resulting in a cleaned record. The total number of remaining packet loss events is calculated based on the cleaned records. Obtain the total number of ultra-wideband data packets sent during the test period; The recalculated ultra-wideband packet loss rate is calculated by dividing the total number of remaining packet loss events by the total number of data packets sent.

8. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, The comparison and recalculation of the UWB packet loss rate with the joint positioning consistency evaluation standard, when the UWB packet loss rate is lower than the upper limit allowed by the consistency evaluation standard, adjusts the packet loss judgment threshold to be adapted to the interference intensity level, resulting in an optimized packet loss judgment threshold, including: Compare the relationship between the re-statistical UWB packet loss rate and the allowed upper limit; If the UWB packet loss rate is lower than the allowable upper limit, calculate the difference between the packet loss rate and the upper limit. The adjustment range of the packet loss judgment threshold is determined based on the ratio of the difference to the interference intensity level. The original packet loss threshold value is modified according to the adjustment range to obtain the optimized packet loss threshold.

9. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, Also includes: This study analyzes the phenomenon that the positioning accuracy of GNSS sensors remains within acceptable limits when the packet loss rate of UWB sensors exceeds the standard, evaluates the temporal correspondence of the synchronous increase in packet loss rate of UWB sensors after the acquisition bandwidth of GNSS sensors is expanded, determines the opposing performance of UWB sensor performance degradation and GNSS sensor performance improvement in the time dimension, and identifies the contradiction in the performance trends of the two types of sensors presented in the evaluation report.

10. The multi-sensor performance consistency comprehensive evaluation method according to claim 1, characterized in that, The process involves acquiring UWB sensitivity test results with expanded bandwidth from the Global Navigation Satellite System (GNSS), applying optimized packet loss threshold filtering to the UWB sensitivity test results, evaluating the performance consistency of multiple sensors, obtaining confirmation of evaluation bias elimination, and generating a multi-sensor performance consistency evaluation report, including: Acquire UWB sensitivity test data when the acquisition bandwidth of the Global Navigation Satellite System is expanded; The optimized packet loss judgment threshold is applied to filter the UWB sensitivity test data, and abnormal data points caused by intermodulation interference are removed to obtain a filtered UWB sensitivity test result set. The true UWB sensitivity is determined based on the filtered test result set. The positioning accuracy deviation values ​​of the Global Navigation Satellite System and UWB are obtained, and the performance consistency score of the multi-sensor system is determined by comparing the positioning coordinate differences between the two types of sensors at the same time. Based on the performance consistency score and the positioning accuracy deviation of each sensor, the difference between the improvement in positioning accuracy after the expansion of the global navigation satellite system bandwidth and the UWB accuracy loss caused by intermodulation interference is calculated to obtain the net contribution value. By integrating the actual UWB sensitivity, the consistency score, the net contribution value, and the performance indicators excluding the influence of intermodulation interference, a multi-sensor performance consistency evaluation report is generated.

Images