Communication quality detection method for a drone and program product

Abstract

The application discloses a communication quality detection method and program product for a UAV, and the method comprises the following steps: acquiring first actual communication signaling data and first actual flight control data of the UAV at a first time, matching the first actual flight control data with the first actual communication signaling data in the time axis of the first actual flight control data, and making the time stamp of the first actual flight control data closest to the first actual communication signaling data; determining prior state prediction data of the UAV at a second time according to the first actual communication signaling data and the first actual flight control data, and determining predicted communication signaling data of the UAV at the second time; updating the prior state prediction data according to the first actual flight control data, the first actual communication signaling data, second actual communication signaling data and the predicted communication signaling data, obtaining posterior state prediction data, and determining a communication quality index associated with the UAV. The technical scheme of the application reduces the communication quality test cost, improves the flight path accuracy of the UAV, and has high sensing and test tightness.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a communication quality testing method and program product for unmanned aerial vehicles (UAVs). Background Technology

[0002] Low-altitude wireless network testing assesses wireless signal strength, network performance, and security from multiple dimensions to ensure the reliability and stability of low-altitude networks.

[0003] Current low-altitude communication testing methods employ dedicated drones equipped with specialized testing equipment, controlled by professionals for flight testing. However, these methods require dedicated drones with high payload capacity, high stability and reliability, and long endurance, leading to limited selection and high costs. The testing equipment must integrate high-precision global navigation satellite system receivers, inertial measurement units, and altimeters, resulting in high system costs and complex maintenance. Furthermore, current drone testing methods, using fixed models and dedicated testing instruments, cannot comprehensively reflect the specific perception situation in the tested area. Summary of the Invention

[0004] This invention provides a communication quality testing method and program product for unmanned aerial vehicles (UAVs) to achieve accurate communication quality testing, reduce communication quality testing costs, improve UAV trajectory accuracy, and achieve high integration of perception and testing.

[0005] In a first aspect, embodiments of the present invention provide a communication quality detection method for unmanned aerial vehicles (UAVs), the method comprising: Acquire the first actual communication signaling data and the first actual flight control data of the UAV at multiple first moments, and use the time axis of the first actual flight control data as a reference to match the first actual communication signaling data with the closest timestamp for each first actual flight control data. Based on the matched first actual communication signaling data and the first actual flight control data, the prior state prediction data of the UAV at the second time moment is determined, and the predicted communication signaling data of the UAV at the second time moment is determined based on the prior state prediction data. The second actual communication signaling data of the UAV at the second time point is obtained. The prior state prediction data is updated according to the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data. The communication quality index associated with the UAV is determined based on the posterior state prediction data and the second actual communication signaling data.

[0006] On the other hand, embodiments of the present invention provide a communication quality detection device for unmanned aerial vehicles (UAVs), the device comprising: The data matching module is used to acquire the first actual communication signaling data and the first actual flight control data of the UAV at multiple first moments. Based on the time axis of the first actual flight control data, it matches the first actual communication signaling data with the closest timestamp for each first actual flight control data. The data determination module is used to determine the prior state prediction data of the UAV at the second time moment based on the matched first actual communication signaling data and the first actual flight control data, and to determine the predicted communication signaling data of the UAV at the second time moment based on the prior state prediction data. The communication quality index determination module is used to acquire the second actual communication signaling data of the UAV at the second time moment, update the prior state prediction data according to the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data, and determine the communication quality index associated with the UAV based on the posterior state prediction data and the second actual communication signaling data.

[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the communication quality detection method for unmanned aerial vehicles according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the communication quality detection method for a drone according to any embodiment of the present invention.

[0009] According to another aspect of the present invention, embodiments of this disclosure also provide a computer program product, including a computer program that, when executed by a processor, implements a communication quality detection method for a drone as described in any of the embodiments of this disclosure.

[0010] The technical solution of this invention acquires first actual communication signaling data and first actual flight control data of a UAV at multiple first moments. Using the time axis of the first actual flight control data as a reference, it matches the first actual communication signaling data with the closest timestamp for each first actual flight control data to achieve high-precision time synchronization of multi-source heterogeneous data, providing reliable input data for subsequent processing. Further, it determines the prior state prediction data of the UAV at a second moment based on the matched first actual communication signaling data and the first actual flight control data, and determines the predicted communication signaling data of the UAV at the second moment based on the prior state prediction data, thereby achieving advance prediction of future communication quality and improving the convergence speed of the filtering algorithm. It acquires the second actual communication signaling data of the UAV at the second moment, updates the prior state prediction data based on the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data, and the predicted communication signaling data to obtain posterior state prediction data, and determines the communication quality indicators associated with the UAV based on the posterior state prediction data and the second actual communication signaling data, improving the adaptability of communication quality testing and reducing reliance on high-precision, high-cost sensors. In summary, this technical solution reduces communication quality testing costs, improves UAV trajectory accuracy, and achieves a high degree of integration between perception and testing.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart illustrating a communication quality detection method for unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention. Figure 2 A flowchart illustrating the data preprocessing process of a communication quality detection method for unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention; Figure 3 A flowchart illustrating the data alignment process of a communication quality detection method for unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention; Figure 4 This is a flowchart illustrating a communication quality detection method for unmanned aerial vehicles (UAVs) provided in this embodiment. Figure 5A framework diagram of the overall scheme of a communication quality detection method for unmanned aerial vehicles provided in this embodiment; Figure 6 This is a flowchart illustrating the trajectory estimation process of a communication quality detection method for unmanned aerial vehicles (UAVs) provided in this embodiment. Figure 7 A schematic diagram of a communication quality detection device for a drone provided in an embodiment of the present invention; Figure 8 A schematic diagram of an electronic device that can be used to implement embodiments of the present invention is shown. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0017] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0018] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0019] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0020] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0021] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0022] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0023] Figure 1 This is a flowchart illustrating a communication quality detection method for unmanned aerial vehicles (UAVs) according to an embodiment of the present invention. This embodiment is applicable to the communication quality detection of UAVs. The method can be executed by a communication quality detection device for UAVs, which can be implemented in hardware and / or software. This communication quality detection device for UAVs can be configured in a computing device. Figure 1 As shown, the method includes: S110. Obtain the first actual communication signaling data and the first actual flight control data of the UAV at multiple first moments. Based on the time axis of the first actual flight control data, match the first actual communication signaling data with the closest timestamp for each first actual flight control data.

[0024] In this embodiment, the drone is an unmanned aircraft. This aircraft can be remotely controlled or fly freely. In the communication quality testing scenario, the drone serves as both the object of the communication quality assessment and the data source.

[0025] The first moment refers to the discrete sampling moment of existing historical data. The first actual communication signaling data is generated and reported to the network side by the UAV's onboard communication module, and is collected, stored, and provided as service signaling and measurement data by the operator's signaling platform. The first actual flight control data consists of motion state parameters recorded in real time by the UAV flight control platform. These operational state parameters are the fusion output of sensors such as the Inertial Measurement Unit (IMU), Global Positioning System (GPS), and barometer within the flight control system. It should be noted that both the first actual communication signaling data and the first actual flight control data are collected at multiple first moments.

[0026] The timeline is an axis formed by arranging data in chronological order, with each data point having a unique timestamp. In this embodiment, the timeline includes a first actual flight control data timeline and a first actual communication signaling data timeline, with the first actual flight control data timeline serving as the reference. This can be understood as follows: during the matching process, the timestamp of the first flight status data remains fixed, while the first actual communication signaling data needs to be aligned to the timestamp of the first flight status data. The timestamp is a time label attached to each data point. This label represents the time when the data occurred.

[0027] The closest first actual communication signaling data is the target first actual communication signaling data found in the set of all first communication signaling timestamps for a given first actual flight control data timestamp. This target first actual communication signaling data has the smallest difference between its timestamp and the given first actual flight control data timestamp.

[0028] It should be noted that because the sampling periods of the drone and the signaling data are inconsistent, the timeline of the drone's first actual flight control data must be used as the benchmark. Signaling data is matched where there are trajectory points, and not matched where there are no trajectories. The specific matching process is as follows: Optionally, for each main timeline trajectory point Search for the first actual communication signaling data with the closest timestamp in the communication signaling data. Furthermore, a preset maximum allowable time difference threshold is defined. For example, the maximum permissible time difference threshold can be set to half the communication signaling data sampling period. This applies if and only if... When the first actual communication signaling data is valid at the current trajectory point, the trajectory point is marked as having no valid communication data.

[0029] Specifically, multiple first-time actual communication signaling data and first-time actual flight control data of the UAV are collected at multiple first-time moments. Considering that the time base of the first-time actual communication signaling data and the first-time actual flight control data is inconsistent, the first-time actual communication signaling data with the smallest timestamp difference from each first-time actual flight control data is found based on the time axis of the first-time actual flight control data, so as to improve the accuracy of data fusion.

[0030] It should be noted that before aligning the first actual communication signaling data and the first actual flight control data in time, the collected actual communication signaling data and actual flight control data need to be preprocessed. Optionally, acquiring the first actual communication signaling data and the first actual flight control data of the UAV at multiple first moments includes: acquiring the third actual communication signaling data and the third actual flight control data of the UAV at multiple first moments, and preprocessing the third actual communication signaling data and the third actual flight control data to obtain the first actual communication signaling data and the first actual flight control data of the UAV.

[0031] The preprocessing includes at least one of the following: time standardization and outlier removal. Time standardization includes converting the signaling time and the UAV flight record time into a unified Coordinated Universal Time (UTC) format. Outlier removal includes removing records in the signaling data where the reference signal received power is less than a preset power, and removing records in the UAV data where the latitude and longitude exceed a preset range.

[0032] In this embodiment, time standardization can be the process of uniformly converting raw time fields from different data sources into the same time base, namely, the Coordinated Universal Time (UTC) format. Optionally, time standardization can convert signaling time, test record time, and UAV flight record time into UTC format to ensure time base consistency.

[0033] Signaling time is the moment recorded by the signaling platform in relation to a signaling event or measurement report. This moment can be the base station's local time, a GPS timestamp, or a Coordinated Universal Time (UTC) format. Test recording time refers to key time points manually recorded during the test. Optionally, test recording time includes the test start time and test end time. UAV flight recording time is the timestamp corresponding to the flight status data output by the UAV flight control platform. It should be noted that the UAV flight recording time can be in the form of a GPS timestamp or a UTC format, etc.

[0034] Therefore, to convert various times to a unified time standard, signaling times, test record times, and UAV flight record times can be converted to UTC format. UTC does not change with the seasons and has no time zone shift. UTC format conversion can be understood as adjusting and unifying the above times by adding or subtracting time zone shifts and correcting the difference between GPS and UTC.

[0035] Outlier removal is the process of identifying and deleting abnormal data records. Outliers can be data with physical meaning, engineering constraints, or logical anomalies. In this embodiment, optionally, outliers can be records where the reference signal reception power is less than a preset power or where the latitude and longitude of the UAV data exceed a preset range.

[0036] The reference signal received power is the linear average power of the base station reference signal measured by the UAV. The preset power is a pre-set lower threshold for the reference signal received power. It should be noted that if the reference signal received power is less than the preset power, the measurement result is unreliable.

[0037] If the latitude and longitude of the drone data exceeds the preset range, it can be understood that the drone data is not within the target testing range. Drone data refers to the status data recorded by the drone flight control platform. This status data includes, but is not limited to, timestamps, latitude and longitude, and altitude. The preset range refers to pre-defined latitude and longitude boundaries.

[0038] It should be noted that reference signal received power less than the preset power is considered outlier data to be removed from signaling data. Similarly, latitude and longitude values ​​outside the preset range in UAV data are considered outlier data to be removed from UAV data.

[0039] Furthermore, optionally, outlier data removed from signaling data also includes records corresponding to empty signaling types and missing base station identifiers.

[0040] A null signaling type occurs when the field in the signaling record used to identify the message type is null or invalid. A reference signal received power less than a preset value can be understood as the reference signal received power being less than a pre-set power threshold. The reference signal received power is the linear average power of the base station reference signal measured by the terminal. For example, the preset power threshold is -140dBm. If the reference signal received power is less than -140dBm, it indicates that the collected reference signal received power is below the lower limit defined in the specification; this situation can be considered as an unreliable measurement result.

[0041] A missing base station identifier refers to a base station identifier in the signaling record that is missing from the serving cell or target cell. The preset range is a manually defined latitude and longitude boundary. This boundary is used to determine whether the UAV data belongs to the target test area. For example, data from the UAV data between longitude 110°14′ and 114°33′ is removed.

[0042] In this embodiment, the third actual communication signaling data is communication signaling and measurement data obtained from the signaling platform, which has not been cleaned or transformed in any way. Optionally, the third actual communication signaling data is service signaling data such as X Detailed Record (XDR) and Measurement Report (MR) obtained from the signaling platform.

[0043] The signaling details are recorded as structured event logs generated by the network-side acquisition system at the interface between the core network and the base station, or within the base station. These logs are generated by parsing the signaling protocol, and each XDR record corresponds to a signaling interaction or service session during a communication process. Service signaling data consists of periodically or event-triggered wireless signal measurement results reported. These measurement results are configured by the terminal based on the base station's measurement control settings.

[0044] It should be noted that, in order to obtain the third actual communication signaling data of the UAV at the first moment, after obtaining the business signaling and measurement data from the signaling platform, the extracted fields include basic identifier, network information, time information, business information, performance, and commands.

[0045] A base identifier is an identifier used to uniquely identify a user, terminal, or session. This identifier can be used to associate all actual communication signaling data generated by the same drone at different times and locations. For example, common fields of a base identifier include, but are not limited to, International Mobile Subscriber Identity (IMSI), Mobile Subscriber Integrated Services Digital Network Number (MSISDN), International Mobile Equipment Identity (IMEI), and communication information.

[0046] Network information describes the network-side entities and their configurations that the drone accesses. This information is used to determine the base station and cell the drone is currently connected to. Optionally, common network information fields include, but are not limited to, access network type, core network element name, base station name, tracking area code, cell code, and access point name. Specifically, the access network type identifies the type of network the drone accesses; the core network element name is the name of the core network functional entity involved in the drone's communication; the base station name is the engineering or network management name of the base station the drone is connected to; and the tracking area code is a code used to identify a tracking area. The cell code is a unique identifier for the serving cell or neighboring cells, and the access point name identifies the external data network accessed by the terminal, i.e., the drone.

[0047] The time information refers to the specific moment when the signaling event or measurement report occurred. Optionally, the time information may include a start timestamp and an end timestamp to achieve time synchronization of multi-source data.

[0048] Business information describes the type of service performed by the drone, its quality of service requirements, and data volume. Performance refers to quantitative indicators related to communication quality and resource efficiency. Optionally, common fields in business information include protocol type, application category, application subcategory, call type, and Session Initiation Protocol (SIP) response code. The protocol type identifies the network protocol used by the drone during communication; the application category provides a coarse-grained classification of the application services run by the drone; the application subcategory provides a fine-grained classification of the application categories within the application category; the call type identifies the call format of a communication session; and the SIP response code is a three-digit status code returned by the server or terminal, indicating the processing result of the request. It should be noted that the application category is used to analyze the differences in user experience for different types of services in low-altitude environments, such as instant messaging, video, web browsing, and file transfer. The application subcategory is used within the application category to locate service characteristics and assist in problem reproduction and optimization; for example, the video category is divided into short video, long video, and live streaming.

[0049] Performance metrics are indicators used to quantitatively evaluate the performance of communication networks, links, or systems in terms of transmission quality, resource utilization, and user experience. Optional fields for performance metrics include: reference signal received power, signal-to-interference-plus-noise ratio (SINNR), uplink and downlink rates, uplink and downlink traffic, TCP latency, out-of-order packet count, and retransmitted packet count.

[0050] Spatial information is a set of data including geographical location, spatial distribution, and relative distance. It should be noted that this information represents the geographical coordinates of the base station, such as latitude and longitude. The latitude and longitude of the base station can be obtained through mapping from the base station identifier.

[0051] The third type of actual flight control data consists of real-time flight parameters of the UAV obtained from the UAV flight control platform. These flight parameters have not undergone any cleaning or anomaly removal process.

[0052] Optionally, the fields extracted from the third actual flight control data include at least spatial location, time information, and flight status. Spatial location refers to the UAV's geographic coordinates and altitude information in three-dimensional space. Optionally, the specific fields corresponding to the spatial information include at least longitude, latitude, and absolute altitude. The above data is used to accurately locate the UAV's position.

[0053] The time information is the timestamp of the drone's flight status data record. The specific field corresponding to this timestamp is GPS time, which is used to match the time with signaling data to achieve multi-source data synchronization.

[0054] Flight status refers to parameters describing the kinematics and attitude of the UAV. Optionally, flight status may include at least speed, yaw angle, and battery level. This parameter is used in the process of verifying the validity of the UAV's flight trajectory.

[0055] Furthermore, after preprocessing the third actual communication signaling data and the third actual flight control data, the processed data will be stored in a structured manner. That is, the preprocessed data, such as the first actual communication signaling data and the first actual flight control data, will be organized according to a unified table structure, such as by field name, type, unit, data source, and description.

[0056] This can be understood as extracting key fields and storing them in a preset data table for each first actual communication signaling data and its matching first actual flight control data. Optionally, this preset data table includes at least the following fields: "utc_time", "signaling_type", "rsrp", "sinr", "enodeb_id", "longitude", "latitude", and "absolute_height". Among these, "utc_time" represents Coordinated Universal Time (UTC), in the format "YYY-MM-DDHH:MM:SS.ms". "signaling_type" represents the signaling type; "rsrp" represents the reference signal received power; "sinr" represents the signal-to-interference-plus-noise ratio; "enodeb_id" represents the base station identifier; "longitude" represents the east longitude; "latitude" represents the north latitude; and "absolute_height" represents the absolute altitude.

[0057] For example, see Figure 2First, signaling data is acquired based on the operator's signaling platform, and then engineering parameter data is acquired based on the engineering parameter platform. The acquired signaling data and engineering parameter data are used as the first actual communication signaling data at multiple points in time for subsequent processing. Finally, UAV flight data (first actual flight control data) is acquired based on the UAV flight control platform.

[0058] Furthermore, the first actual communication signaling data and first actual flight control data of the UAV at multiple first moments are preprocessed. Specifically, using the timeline of the first actual flight control data as a benchmark, the first actual communication signaling data with the closest timestamp is matched for each first actual flight control data. After data matching, the first actual communication signaling data and first actual flight control data are cleaned. Data cleaning includes at least time standardization and outlier removal. In time standardization, the signaling time and UAV flight record time are uniformly converted to Coordinated Universal Time (UTC) format. After time standardization, outlier removal is performed. Outliers that can be removed include, but are not limited to, records where the reference signal received power is less than a preset power, and records where the latitude and longitude of the UAV data exceed a preset range. Finally, the cleaned data is organized and stored according to a unified data table structure for subsequent use.

[0059] Specifically, the system acquires the third actual communication signaling data and the third actual flight control data of the UAV at multiple first moments. Furthermore, it preprocesses the third actual communication signaling data and the third actual flight control data to obtain the UAV's first communication command and first actual flight control data, thereby improving data quality and supporting high-precision alignment of multi-source data.

[0060] S120. Determine the prior state prediction data of the UAV at the second moment based on the matched first actual communication signaling data and the first actual flight control data, and determine the predicted communication signaling data of the UAV at the second moment based on the prior state prediction data.

[0061] In this embodiment, the first actual communication signaling data after matching is valid communication signaling data that has been time-aligned. The second time point is the time point to be predicted. The prior state prediction data is the data obtained by predicting the state of the UAV at the second time point before obtaining the actual observation data at the second time point.

[0062] The predicted communication signaling data is the estimated value of the communication signaling for Lesson 20, which is derived from the prior state prediction data.

[0063] Specifically, to provide location data for communication signaling prediction, the prior state prediction data of the UAV at the second moment is determined based on the matched first actual communication signaling data and the first actual flight control data, so as to achieve advance estimation of the future state. Furthermore, based on the prior state prediction data, the predicted communication signaling data of the UAV at the second moment is determined, so as to achieve advance knowledge of communication quality change trends.

[0064] Furthermore, after determining the prior predicted state information, the predicted communication signaling data of the UAV at the second time point needs to be determined based on the prior state data. The process of determining the predicted communication signaling data will be described in detail below. Optionally, the prior predicted state information includes a first predicted position; the predicted communication signaling data may include at least one of the following: predicted reference signal received power, predicted time advance, and predicted angle of arrival.

[0065] When the predicted communication signaling data may include the predicted reference signal received power, the first predicted communication signaling data of the UAV at the second time moment is determined based on prior predicted state information. This includes: determining the predicted reference signal received power of the UAV at the second time moment using a path loss model based on the spatial distance between the first predicted position and the base station position in the prior predicted state information. The path loss model is constructed based on the reference signal power, the path loss exponent, and the spatial distance between the UAV position and the base station position. When the predicted communication signaling data may include a predicted time advance, the first predicted communication signaling data of the UAV at the second time moment is determined based on the prior predicted state information. This includes: determining the predicted time advance of the UAV at the second time moment using a light speed propagation delay model based on the spatial distance between the first predicted position and the base station position and the speed of light in the prior predicted state information. In cases where the predicted communication signaling data may include the predicted angle of arrival, the first predicted communication signaling data of the UAV at the second time moment is determined based on the prior predicted state information, including: determining the predicted angle of arrival of the UAV at the second time moment by using the four-quadrant arctangent function based on the coordinate difference between the first predicted position and the base station position in the prior predicted state information in the first direction and the second direction, wherein the first direction and the second direction are perpendicular.

[0066] In this embodiment, the prior predicted state information is a state estimate of the UAV determined based on the first actual communication signaling data and the first actual flight control data. It should be noted that the prior predicted state information includes a first predicted location. The first predicted location is a field in the prior predicted state information. This field represents the predicted geographic coordinates of the UAV at the second time point. For example, the first predicted location may include latitude, longitude, and altitude, etc.

[0067] The predicted communication signaling data consists of estimated UAV communication signaling values ​​obtained based on prior predicted state information. It should be noted that the predicted communication signaling data includes predicted reference signal received power, predicted timing advance, and predicted angle of arrival. Specifically, the predicted reference signal received power is the estimated reference signal received power of the UAV at the second moment. This estimated value reflects the downlink signal strength at the predicted location. The predicted timing advance is an estimated time advance calculated based on the distance between the first predicted location and the base station location. This estimated value is used to compensate for the time delay of signal propagation in space. The predicted angle of arrival is an estimated signal angle of arrival calculated based on the geometric relationship between the first predicted location and the base station location. This estimated value indicates the direction of arrival of the signal from the UAV to the base station.

[0068] The base station location refers to the base station's location information, such as latitude and longitude coordinates. The spatial distance between the first predicted location and the base station location is the Euclidean distance between the UAV's predicted location and the base station location.

[0069] The path loss model is a data model of how the power of a wireless signal decreases with distance as it propagates through space. This model can calculate the predicted received power of a reference signal based on spatial distance. The predicted received power of the reference signal is the power value of the reference signal transmitted by the base station at the second time point.

[0070] It should be noted that the path loss model is constructed based on the reference signal power, the path loss exponent, and the spatial distance between the UAV's location and the base station's location. The reference signal power is the power value of the reference signal transmitted by the base station. This power value is the theoretical received power at a distance of 1 meter from the base station and is a calibrable constant. The path loss exponent is an empirical coefficient reflecting the rate of signal power attenuation with distance. This coefficient can be obtained through field testing.

[0071] For example, the method for determining the predicted reference signal received power at the second time step based on the path loss model is as follows: ; in, Let k be the reference signal received power predicted at time k. The predicted state information at time k can be understood as the prediction result of the UAV's state information at time k based on information from time k-1 and earlier. The reference signal power is n, and the path loss exponent is n. For the location of the base station, Let k be the first predicted position at time k.

[0072] The speed of light is the speed at which electromagnetic waves propagate in a vacuum. The light-speed propagation delay model is used to calculate the round-trip propagation delay of a signal between a drone and a base station based on spatial distance. Based on this model, the predicted time advance of the drone at the second moment can be determined. The predicted time advance is an estimated value of the drone's time advance at the second moment. This estimated value is used to compensate for the round-trip delay of the signal propagating in space.

[0073] For example, the method for determining the prediction time lead in a light speed propagation delay model is as follows: ; in, To predict lead time, For the predicted state information at time k, The first predicted location of the drone. For the location of the base station, At the speed of light, This is the system offset. This is a discretization standard function used to map continuous time delay values ​​to discrete prediction time advances.

[0074] It should be noted that the system offset is used to compensate for fixed-time errors caused by base station and terminal processing delays. This value can be obtained through initial calibration.

[0075] The first and second directions are used to describe two mutually perpendicular coordinate axes on a horizontal plane. For example, the first direction could be east in a geographic coordinate system, and the second direction could be north. The coordinate difference is the difference between the predicted UAV position and the base station position in the first and second directions, respectively. The four-quadrant arctangent function is a function that determines the correct quadrant of the angle based on the sign of the coordinate difference. The predicted angle of arrival is calculated based on the four-quadrant arctangent function and the coordinate difference between the first and second directions.

[0076] The predicted angle of arrival (Angle of Arrival) is an estimated value of the signal angle of arrival calculated based on the geometric relationship between the first predicted location and the base station location. This estimate represents the direction of signal arrival from the drone to the base station.

[0077] For example, the method for calculating the predicted angle of arrival using the four-quadrant arctangent function is as follows: ; in, For predicting the angle of arrival prediction function, Let the predicted angle of arrival be at time k. Let k be the prior predicted state information at time k. It is the arctangent function in the fourth quadrant. The first direction component of the first predicted position. The second direction component is the first predicted position. The first directional component represents the location of the base station. The second directional component of the base station location. The coordinate difference between the first predicted location and the base station location in the first direction. This represents the coordinate difference between the first predicted location and the base station location in the second direction.

[0078] In one scenario, prior predicted state information is acquired, and the spatial distance between the first predicted position and the base station position is calculated based on the first predicted position and the base station position in the prior predicted state information. Furthermore, based on the reference signal power, path loss exponent, and the spatial distance between the UAV position and the base station position, the predicted reference signal received power of the UAV at the second time point is determined. This reduces reliance on measured signaling, lowers testing costs, and enables early prediction of low-altitude communication quality.

[0079] In one scenario, prior predicted state information is obtained, and the spatial distance between the two is calculated based on the prior predicted state information and the base station location. Furthermore, to obtain the signal propagation delay between the UAV and the base station in advance, the light speed propagation delay is used. Based on the spatial distance, the speed of light, and the system offset, the predicted time advance of the UAV at the second moment is calculated to improve communication reliability and state estimation accuracy.

[0080] In one scenario, prior predicted state information is acquired, and the coordinate differences between the first predicted position and the base station position in the prior predicted state information in the first and second directions are calculated respectively. To determine the azimuth and orientation of the UAV relative to the base station in advance, a four-quadrant arctangent function is used. Based on the differences between the first predicted position and the base station position in the first and second directions, the predicted angle of arrival of the UAV at the second time moment is determined, thereby improving the directional accuracy and positioning reliability of low-altitude communication.

[0081] S130. Obtain the second actual communication signaling data of the UAV at the second moment, update the prior state prediction data according to the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data, and determine the communication quality index associated with the UAV based on the posterior state prediction data and the second actual communication signaling data.

[0082] In this embodiment, the second actual communication signaling data is the real communication signaling data actually acquired at the second moment. This communication signaling data is actually measured by the UAV and acquired through the operator's signaling platform. The posterior state prediction data is the data after correcting the prior state prediction data after obtaining the actual observation data at the second moment. The communication quality index refers to various parameters used to quantitatively evaluate the network communication quality of the UAV during low-altitude flight. This parameter can be used to evaluate low-altitude network performance.

[0083] Specifically, after the UAV acquires the second actual communication signaling data at the second moment, in order to obtain accurate UAV state prediction data, the prior state prediction data is updated and corrected based on the second actual communication signaling data, the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data, and the predicted communication signaling data. Furthermore, based on the corrected posterior state prediction data and the second actual communication signaling data, communication quality indicators associated with the UAV are determined to suppress error accumulation and improve the accuracy of state estimation and the reliability of the indicators.

[0084] It should be noted that communication quality indicators can be determined after obtaining posterior state prediction data. Furthermore, the specific content of communication quality indicators and the process of determining communication command indicators based on posterior state prediction data and second actual communication signaling data are described in detail. Optionally, communication quality indicators include at least one of the following: effective coverage, average uplink rate, average downlink rate, and average latency.

[0085] Furthermore, when the communication quality indicators include effective coverage, the communication quality indicators associated with the UAV are determined based on the posterior state prediction data and the second communication signaling data, including: obtaining the total number of first trajectory points of the second actual communication signaling data that matches the posterior state prediction data, and the total number of second trajectory points whose reference signal received power value is greater than a preset quality threshold, and determining the effective coverage associated with the UAV based on the total number of first trajectory points and the total number of second trajectory points.

[0086] In this embodiment, the effective coverage rate is the proportion of trajectory points whose signal quality meets preset requirements out of the total number of trajectory points in the UAV's flight path. This parameter reflects the degree of coverage in low-altitude areas; a higher effective coverage rate indicates fewer areas with weak coverage. For example, the preset requirement could be that the reference signal received power is greater than a preset quality threshold. The preset quality threshold could be a pre-defined reference signal received power threshold used to distinguish whether coverage is good or not.

[0087] The total number of the first trajectory points is the total number of valid trajectories corresponding to the second actual communication signaling data that successfully matches the posterior state prediction data. Here, a successful match can be understood as the second actual communication signaling data that is time-aligned with the posterior state prediction data.

[0088] The total number of second trajectory points is the number of trajectory points among all valid trajectory points where the reference signal received power is greater than a preset quality threshold.

[0089] It should be noted that the effective coverage rate is calculated as follows: ; in, The total number of points on the second trajectory. The total number of points on the first trajectory. For effective coverage.

[0090] Optionally, an alarm can be triggered when the effective coverage rate is lower than a preset threshold to indicate a communication blind spot. For example, if the preset threshold is set to 95%, an alarm will be triggered when the calculated effective coverage rate is lower than 95%.

[0091] Specifically, to calculate the effective coverage associated with the UAV, second actual communication signaling data, matched with posterior state prediction data, is obtained to determine the total number of first trajectory points. Further, the reference signal received power is compared with a preset quality threshold to obtain the total number of second trajectory points whose reference signal received power is greater than the preset quality threshold. Based on the total number of first and second trajectory points, the effective coverage associated with the UAV is calculated to quantify low-altitude coverage quality.

[0092] Furthermore, when the communication quality indicators include uplink average rate and downlink average rate, the communication quality indicators associated with the UAV are determined based on the posterior state prediction data and the second communication signaling data. This includes: extracting the uplink instantaneous rate and downlink instantaneous rate from the second actual communication signaling data aligned with the posterior state prediction data; determining the duration and total test time corresponding to the second actual communication signaling data based on the signaling timestamps corresponding to the second actual communication signaling data at multiple trajectory points of the UAV; and determining the uplink average rate and downlink average rate associated with the UAV based on the uplink instantaneous rate, downlink instantaneous rate, duration, and total test time.

[0093] The uplink instantaneous rate is the instantaneous data transmission rate at which the UAV sends data to the base station within a specific sampling time. The downlink instantaneous rate is the instantaneous data transmission rate at which the base station sends data to the UAV within a specific sampling time.

[0094] The trajectory points are multiple discrete sampling locations on the UAV's flight path. The signaling timestamp is the time marker attached to each second actual communication signaling data. The duration is the length of the time interval represented by the second actual communication signaling data corresponding to each trajectory point. The total test time is the total duration from the first valid trajectory point to the last valid trajectory point.

[0095] The uplink average rate is the average data transmission rate at which the drone sends data to the base station during the test period. This average transmission rate reflects the average level of the drone's uplink data transmission capability. The downlink average rate is the average data transmission rate at which the base station sends data to the drone during the test period. This parameter reflects the average speed at which the drone receives data. It should be noted that the average uplink rate and average downlink speed are extracted from specified fields in the signaling data, such as uplink and downlink rates.

[0096] For example, firstly, the uplink instantaneous velocity, downlink instantaneous rate, and the number of valid signaling data points after time alignment are extracted from the second actual communication signaling data. Further, the duration of the signaling timestamp calculation in the second actual communication signaling data for each trajectory point is determined based on the signaling timestamp corresponding to the second actual communication signaling data, and the total test time is calculated accordingly. To measure communication bandwidth, the uplink average rate and downlink average rate associated with the UAV are calculated. The method for calculating the uplink average rate is as follows: ; in, Let $\frac{k}{k}$ be the instantaneous upward velocity of the $k$-th trajectory point. Let k be the duration of the trajectory point. M represents the total test time, and M represents the number of valid signaling data points.

[0097] The method for calculating the downlink average rate is as follows: ; in, Let be the downlink instantaneous velocity of the k-th trajectory point. Let k be the duration of the trajectory point. M represents the total test time, and M represents the number of valid signaling data points.

[0098] Specifically, the uplink instantaneous rate and downlink instantaneous rate are extracted from the second actual communication signaling data after alignment with the posterior state prediction data. The duration of each trajectory point and the total test time are calculated using the signaling timestamp. Then, the uplink average rate and downlink average rate are determined based on the uplink instantaneous rate, downlink instantaneous rate, duration, and total test time to eliminate the bias of non-uniform sampling and improve the reliability of the average rate index.

[0099] Furthermore, when the communication quality index includes average latency, the communication quality index associated with the UAV is determined based on the posterior state prediction data and the second communication signaling data, including: extracting instantaneous latency from the second actual communication signaling data aligned with the posterior state prediction data, and determining the average latency associated with the UAV based on the instantaneous latency of the UAV at multiple trajectory points.

[0100] In this embodiment, instantaneous latency is the time it takes for a data packet to travel from the sender to the receiver at a given sampling moment. This data can be directly extracted from the corresponding field in the second driver communication signaling data. Average latency is the average time required for a data packet to travel from the sender to the receiver within the test period. This parameter reflects the network's real-time response capability.

[0101] To determine the communication response speed, the instantaneous delay can be extracted from the second actual communication signaling data, and the average delay associated with the UAV can be calculated: ; in, For average delay, Let M be the instantaneous delay of the k-th trajectory point, and M be the number of effective delay data points.

[0102] Specifically, to analyze the impact of UAV flight altitude and speed on communication latency, instantaneous latency is extracted from the second actual communication signaling data aligned with posterior state prediction data. Based on the instantaneous latency of multiple trajectory points and the number of trajectory points, the average latency is calculated to obtain the average latency index associated with the UAV, reflecting the real-time response capability of data services in low-altitude networks, thereby improving the reliability of UAV communication and mission execution efficiency.

[0103] Furthermore, after obtaining the posterior state prediction data, target event detection can be performed on the UAV to maintain a stable system state. The target event detection process will be described in detail below. Optionally, after obtaining the posterior state prediction data, the process further includes: detecting target events on the UAV, where target events include macroscopic events, microscopic events, and communication events; macroscopic events include at least one of the following: takeoff, landing, hovering; microscopic events include at least one of the following: altitude change, horizontal acceleration, signal drop; communication events include at least one of the following: cell handover, wireless link failure, handover event; in response to the absence of a detected target event, a physically constrained cubic spline interpolation algorithm is used to upsample the second predicted position in the posterior state prediction data, ensuring that the trajectory points obtained after upsampling are consistent with the second actual communication signaling data in terms of time granularity.

[0104] In this embodiment, the target event is a specific type of event that needs to be identified, detected, or recorded during the flight of the drone.

[0105] Optionally, target events can be categorized into three types: macroscopic events, microscopic events, and communication events. When no target event is detected, the system is considered to be in a stable state.

[0106] Macro-level events describe overall phases or large-scale changes in the state of a UAV flight mission, reflecting changes in the UAV's mission status. For example, macro-level events include at least takeoff, landing, and hovering. Takeoff can be understood as the UAV's vertical ascent from the ground or takeoff platform, where its altitude increases continuously from near zero. Landing is the process of the UAV descending from the air to the ground until it stops moving, and hovering is the UAV maintaining a stationary flight state in a fixed spatial location. These situations may trigger signal fluctuations and are therefore considered target events for evaluating communication quality.

[0107] Micro-events describe events that describe rapid, localized changes in the state of a drone within a short period of time. For example, micro-events include at least altitude abrupt changes, horizontal acceleration, and signal descent. An altitude abrupt change can be understood as a significant change in the drone's vertical altitude within a very short time; horizontal acceleration is a significant change in the drone's horizontal speed; and a signal descent is a sharp drop in the received power or signal-to-interference-plus-noise ratio of the reference signal received by the drone within a very short time. These events can lead to rapid changes in signal path loss and are therefore considered target events for evaluating communication quality.

[0108] Communication events are specific events related to wireless links and signaling interactions that occur between a drone and a ground network. For example, communication events include at least cell handover, wireless link failure, and handover events. Cell handover is the process by which a drone, while moving, switches from its current serving cell to a target cell. Wireless link failure is a situation where the quality of the wireless link between the drone and the base station continuously deteriorates, causing communication to be unsustainable. These situations can lead to handover failures, and are therefore considered target events for evaluating communication quality.

[0109] In this embodiment, "no target event detected" can be understood as the system summarizing multiple predefined target events that were not detected during the UAV's flight, indicating that the current flight phase is in an abnormal state. The physically constrained cubic spline interpolation algorithm is an interpolation method based on cubic spline functions, and physical kinematic constraints are added during the interpolation process, making the generated interpolated trajectory more consistent with the actual flight characteristics of the UAV.

[0110] The second predicted position is the position component in the posterior state prediction data, which is the position estimate of the UAV after correction at the second time step. Optionally, the second predicted position can be three-dimensional coordinates including longitude, latitude, and altitude. Upsampling can be understood as increasing the density of sampling points on the time axis, that is, generating multiple intermediate trajectory points between two known trajectory points through an interpolation algorithm, thereby improving the temporal resolution of the trajectory sequence.

[0111] The time granularity is the time interval between two adjacent sampling points in the trajectory sequence. This value reflects the temporal resolution of the data.

[0112] For details, see Figure 3 After acquiring posterior state prediction data (including UAV trajectory data and communication signaling data), target events, including macroscopic events, microscopic events, and communication events, are detected. When no target event is detected, the representation is in a stable state. A physically constrained cubic spline interpolation algorithm is used to upsample the second predicted position in the posterior state prediction data, generating denser trajectory points to align its temporal granularity with the second actual communication signaling data, thus providing a synchronized data foundation for subsequent spatiotemporal correlation analysis.

[0113] For example, multiple original trajectory points are used as spline nodes, and UAV dynamic constraints are used as boundary conditions. For the stable state interval [t0, t...] n The natural boundary conditions S(t0) = 0 and S(t) are adopted. n =0, ensuring that the acceleration at the starting and ending points is zero. Further, a tridiagonal system of equations is constructed and solved using the pursuit method, calculating the spline coefficients for the longitude, latitude, and altitude components respectively.

[0114] It should be noted that after generating trajectory points using a cubic spline interpolation algorithm based on physical constraints, the velocity and acceleration corresponding to these trajectory points are calculated. Furthermore, these are compared with the physical performance limits of the UAV, such as maximum velocity and maximum acceleration. A maximum number of iterations is pre-set; for local intervals that do not meet the physical constraints, interpolation can be re-established by adding nodes. If the physical constraints are still not met after exceeding the maximum number of iterations, the iteration stops, and the interval is marked as "physical constraints not met" and recorded as a null value.

[0115] In one embodiment, the communication quality detection method for UAVs may further include: in response to the detection of a target event, aligning second actual communication signaling data with trajectory points in posterior state prediction data based on a dynamic time warping algorithm.

[0116] In this embodiment, the dynamic time programming algorithm is an algorithm used to measure the similarity between two trajectory points. This method finds the optimal alignment path by nonlinearly stretching or compressing the time axis. It can be understood as constructing a distance matrix between the trajectory points in the second actual communication signaling data and the subsequent state prediction data, and finding the path with the minimum cumulative distance through dynamic programming, ensuring a one-to-one correspondence between points on the path.

[0117] Specifically, when the system detects a target event, it uses a dynamic time warping algorithm to align the trajectory points in the second actual communication signaling data with the posterior state prediction data. When the timestamps of the two are not completely consistent, they can be matched and aligned using the dynamic time warping algorithm.

[0118] Furthermore, in order to achieve high-precision adaptive synchronization under dynamic conditions, the process of time alignment and fusion of the second actual communication signaling data and the posterior state prediction data will be further refined.

[0119] Optionally, the analysis window boundary is adaptively calculated based on the event type and duration. Trajectory points within the window are extracted from the position sequence of the posterior state prediction data, and signaling points within the window are extracted from the second actual communication signaling data. The trajectory points are converted into multi-dimensional feature vectors containing velocity, acceleration, rate of change of altitude, and rate of change of heading. The signaling points are converted into multi-dimensional feature vectors containing the rate of change of reference signal received power, rate of change of reference signal received quality, and whether a handover has occurred. Each multi-dimensional feature vector is standardized. The number of trajectory points and signaling points within the window is checked to see if they meet preset thresholds. If not, the search window is gradually expanded until the conditions are met or the maximum number of expansions is reached. Further, a dynamic time warping algorithm is executed to construct a cost matrix and a cumulative cost matrix. The optimal warping path is backtracked to obtain a one-to-one mapping relationship between trajectory point timestamps and signaling point timestamps. Based on the mapping relationship, the trajectory points in the second actual communication signaling data and the posterior state prediction data are aligned. After processing, the alignment result is output and the system automatically returns to a stable state. The alignment result is used for the calculation of communication quality indicators.

[0120] In this embodiment, the event type refers to a category of specific target events occurring during UAV flight or communication. Optionally, the event type may include at least macroscopic time, microscopic event, or communication event. The duration is the length of time the target event takes from start to finish. The window boundary is a time interval dynamically determined based on the event type and duration. For example, the duration is extended forward and backward by a certain proportion based on the time center timestamp.

[0121] The location sequence is a time-ordered sequence of trajectory point locations extracted from posterior state prediction data. The signaling point is a single record in the second actual communication signaling data. This record may contain a timestamp and a set of communication measurement parameters.

[0122] For details, see Figure 3 After detecting a target event, the timestamp of the event's center is recorded. Based on the time type and duration of the target event, the window boundary is calculated. Trajectory points within the window are obtained from the trajectory points in the posterior state prediction data location sequence based on the window boundary. And based on the trajectory points within the window, signaling points within the window are extracted from the second actual communication signaling data to focus on the target event within that interval, thereby improving subsequent time alignment and the specificity of the analysis, and increasing processing efficiency.

[0123] In this embodiment, the multidimensional feature vector is a comprehensive attribute of a data point across multiple dimensions. Optionally, the multidimensional feature vector based on trajectory point transformation includes information such as velocity, acceleration, altitude change rate, and heading change rate. The altitude change rate is the rate of change of the UAV's altitude over time. This rate of change can be calculated based on the altitude difference between adjacent trajectory points and the time interval between trajectory points. The heading change rate is the rate of change of the UAV's heading angle over time. This rate of change can be calculated based on the heading angle difference between adjacent trajectory points and the time interval.

[0124] The multidimensional feature vector based on signaling point transitions includes at least the rate of change of reference signal received power, the rate of change of reference signal received quality, and whether a handover has occurred. The rate of change of reference signal received power is the rate of change of reference signal received power over time. This rate can be calculated based on interpolation of reference signal received power between adjacent signaling points and the time interval. The rate of change of reference signal received quality is the rate of change of reference signal received quality over time. This rate can be calculated based on interpolation of reference signal received quality between adjacent signaling points and the time interval. Whether a handover has occurred can be understood as whether the signaling point is associated with a cell handover event.

[0125] Specifically, a multi-dimensional feature vector is generated based on trajectory points, including dimensions such as velocity, acceleration, rate of change of altitude, and rate of change of heading. A multi-dimensional feature vector is also generated based on signaling points, including the rate of change of reference signal received power, the rate of change of reference signal received quality, and whether a handover has occurred, providing input for dynamic time warping alignment.

[0126] In this embodiment, standardization can be understood as transforming feature values ​​of different dimensions to a uniform scale. For example, the Z-score method can be used to standardize each feature dimension. It should be noted that the Z-score method is as follows: .in, The mean, Let x be the standard deviation. x represents the original eigenvalues. These are standardized feature values.

[0127] The number of trajectory points within the window refers to the number of valid trajectory points extracted from the posterior state prediction data within the current window. The number of signaling points refers to the number of valid signaling points extracted from the second actual communication signaling data within the current window. The preset threshold is a pre-set minimum threshold for the number of trajectory points and signaling points. This threshold is used to determine whether the data in the current window is sufficient.

[0128] Gradually expanding the search window can be understood as using the event center timestamp as a baseline and expanding the time range in both directions to include more trajectory points within the window. The maximum expansion count is the maximum number of iterations allowed for gradually expanding the search window. By setting the maximum expansion count, it prevents the window from expanding indefinitely, which could introduce too much irrelevant data.

[0129] Specifically, the feature values ​​of multiple feature dimensions are independently standardized to generate standardized feature values. Further, the number of trajectory points and signaling points within the window are compared with preset thresholds. If the preset thresholds are not met, the search window is gradually expanded until the number of trajectory points and signaling points within the window meets the preset thresholds, or the number of times the search window is expanded exceeds the maximum expansion count. This ensures the minimum data requirements of the dynamic time warping algorithm and adaptively compensates for data sparsity issues.

[0130] In this embodiment, the dynamic time warping algorithm is used to measure the similarity between two time series. This algorithm allows for the search of the optimal alignment path by nonlinearly stretching or compressing the time axis.

[0131] The cost matrix represents the local matching cost between trajectory points and signaling points. It should be noted that each element C(i,j) in the cost matrix represents the Euclidean distance between the eigenvectors of trajectory point i and signaling point j. ; in, For the cost matrix, Let i be the standardized feature vector of the i-th trajectory point. Let j be the standardized feature vector of the j-th signaling point. It is the Euclidean norm.

[0132] The cumulative cost matrix is ​​filled using a dynamic programming recurrence relation. It should be noted that the elements in the cumulative cost matrix are filled using a dynamic programming recurrence relation: ; in, Let be the element at position (i,j) in the cumulative cost matrix. Let be the local matching cost at position (i,j) in the cost matrix. It is a minimum value function. The cumulative cost of the left-adjacent cell. The cumulative cost of the top-left adjacent cell. This is the cumulative cost of the adjacent cells above.

[0133] This can be understood as obtaining the cumulative cost of the current position (i,j) by adding the local matching cost corresponding to position (i,j) to the minimum cumulative cost of the three adjacent positions (left, top, and top left).

[0134] Furthermore, after determining the cumulative cost matrix, the optimal regularization path is found by backtracking. This can be understood as tracing back from the end point of the cumulative cost matrix to the starting point to determine the optimal alignment path. After obtaining the optimal alignment path, the corresponding timestamps are found based on the track point and signaling point index pairs, forming a one-to-one mapping relationship. The mapping relationship is a list of corresponding pairs formed by matching track points and signaling points according to their temporal similarity.

[0135] Furthermore, the second actual communication signaling data is aligned with the trajectory points in the posterior state prediction data based on the mapping relationship, and the alignment result is output. Optionally, an alignment strategy for the trajectory points in the second actual communication signaling data and the posterior state prediction data is determined based on the target event. For example, for macro and micro service events, the trajectory points in the second actual communication signaling data and the posterior state prediction data are mapped based on the trajectory point timestamp, and communication indicators such as the reference signal received power of multiple signaling points mapped to the same trajectory point are aggregated. The aggregation method can be selected based on the service. For communication events, the trajectory points in the second actual communication signaling data and the posterior state prediction data are mapped based on the mapping relationship, using the signaling point timestamp as a reference, and the location information of multiple trajectory points mapped to the same signaling point is aggregated according to service requirements.

[0136] Specifically, after completing window data extraction and feature standardization, the system executes a dynamic time warping algorithm. First, a cost matrix is ​​constructed, where each element represents the Euclidean distance between the standardized feature vectors of the trajectory point and the signaling point. Then, a cumulative cost matrix is ​​recursively calculated using dynamic programming to accumulate local matching costs. Finally, the system backtracks from the endpoint of the cumulative cost matrix to the starting point to obtain an optimal warping path that minimizes the total cost. This path directly establishes a one-to-one mapping between the trajectory point timestamps and the signaling point timestamps. Based on this mapping, the system performs time alignment between the data in the second actual communication signaling data and the trajectory points in the posterior state prediction data, ensuring that each trajectory point is associated with the corresponding signaling point at that time. After processing, the alignment result is output, and the system automatically returns to a stable state, awaiting the next event trigger. Ultimately, this alignment result serves as the data basis for calculating the communication quality index as described in claim 6, achieving high-precision alignment of the nonlinear time axis and ensuring the spatial accuracy of the communication quality index.

[0137] Furthermore, to ensure a close integration between the calculated indicators and the positioning results, the posterior state prediction data is evaluated based on the accuracy index of the sensing station. The evaluation process based on the accuracy index of the sensing station is then described in detail. Optionally, after obtaining the posterior state prediction data, the process also includes: using the posterior state prediction data as the baseline true value, and determining the accuracy index of the sensing station based on the measured values ​​reported by the sensing station and the baseline true value.

[0138] In this embodiment, the baseline true value is the actual value used as a reference standard in accuracy evaluation. For example, the posterior state prediction data is used as the baseline true value. The sensing station is a wireless station that simultaneously possesses communication and sensing functions. This can be understood as the sensing station not only communicating with the drone but also detecting the drone's position, speed, attitude, and other states by analyzing wireless signal information.

[0139] The measured values ​​are drone status data obtained from the sensor station. It should be noted that these measured values ​​may include measurement noise, etc.

[0140] The accuracy index of a sensor station is a numerical indicator used to quantitatively evaluate the sensing performance of a sensor station. This index reflects the degree of closeness between the reported measured value and the true reference value.

[0141] Specifically, subsequent state data is used as the baseline true value, and the measured values ​​reported by the sensing stations are obtained. The measured values ​​reported by the sensing stations are compared with the baseline true value, and the difference between the measured values ​​and the baseline true value is calculated to form the accuracy index of the sensing stations, so as to achieve a quantitative evaluation of the performance of the sensing stations.

[0142] Furthermore, the process of determining the accuracy indicators of the sensing station based on the measured values ​​reported by the sensing station and the reference true value is further refined. Optionally, the accuracy indicators of the sensing station include at least one of the following: position error, instantaneous velocity error, and height error; the position error includes at least one of the following: instantaneous position error and root mean square error of position; the height error includes at least one of the following: instantaneous height error, average height error, and standard deviation of height detection.

[0143] Further, the accuracy indicators of the sensing station are determined based on the measured values ​​reported by the sensing station and the reference true value, including at least one of the following: determining the instantaneous position error based on the second predicted position in the UAV position measurement value reported by the sensing station and the reference true value; determining the root mean square error of the position based on the instantaneous position error of the UAV at multiple trajectory points; aligning the velocity measurement modulus value reported by the sensing station with the velocity modulus value of the reference true value, and calculating the instantaneous velocity error, i.e., the absolute value of the difference between the two modulus values; determining the instantaneous altitude error based on the UAV altitude measurement value reported by the sensing station and the second predicted altitude of the UAV in the reference true value; determining the average altitude error based on the instantaneous altitude error of the UAV at multiple trajectory points; and determining the altitude detection standard deviation based on the instantaneous altitude error and the average altitude error.

[0144] In this embodiment, the UAV's position measurement value is the spatial position estimate of the UAV obtained by the sensor station through its sensing function. The instantaneous position error is the deviation between the UAV position measurement value reported by the sensor station and the true reference value at a single trajectory point. It should be noted that the instantaneous position error is calculated as follows: ; in, As the baseline truth value, The location measurement values ​​of the drone reported by the sensor station. This represents the instantaneous position error.

[0145] The root mean square error (RMSE) is the square root of the mean of the squares of the instantaneous position errors at a set of trajectory points. This data is used to evaluate the overall positioning accuracy of the sensor station. The process for determining the RMSE is as follows: ; in, This is the root mean square error of the position. The total number of trajectory points. This represents the instantaneous position error.

[0146] Optionally, after calculating the instantaneous position error between the UAV position measurement value reported by the sensor station and the reference value, the average position error can also be determined based on the instantaneous position error of the UAV at multiple trajectory points.

[0147] Specifically, after obtaining the posterior state prediction data as the baseline truth, the UAV position measurement value reported by the sensor station is time-aligned with the second predicted position in the baseline truth, and the instantaneous position error is calculated point by point, which is the three-dimensional Euclidean distance between the measured position and the baseline truth position. Furthermore, the instantaneous position errors obtained by the UAV at multiple trajectory points are statistically analyzed, and the root mean square error of the position is obtained to reflect the accuracy level of the sensor station's overall positioning of the UAV.

[0148] In this embodiment, the velocity measurement modulus is the magnitude of the UAV velocity vector measured by the sensing station through its sensing function, and the baseline true velocity modulus is the velocity estimate in the posterior state prediction data.

[0149] Instantaneous velocity error is the deviation between the velocity measurement modulus reported by the sensor station and the reference true velocity modulus at a single trajectory point. It can be understood as the absolute value of the difference between the two moduli: ; in, As the reference true value velocity vector, The velocity vector reported by the sensor station. This represents the instantaneous velocity error.

[0150] Furthermore, based on the instantaneous velocity error, the average velocity error is determined: ; in, The instantaneous velocity error is N, where N is the total number of trajectory points. This represents the average speed error.

[0151] Specifically, the velocity measurement modulus reported by the sensor station is matched with the velocity modulus in the reference true value at the same instant to ensure that both represent the instantaneous velocity of the UAV in the same flight state. That is, through a preset alignment method, a one-to-one correspondence is established between each sensor station measurement point and the corresponding posterior state predicted trajectory point, and the instantaneous velocity error is calculated, which is the absolute value of the difference between the velocity measurement modulus and the reference true velocity modulus. This error value reflects the accuracy of the sensor station's measurement of the UAV's velocity at a single instant, eliminating the interference of directional deviation on the evaluation.

[0152] In this embodiment, the altitude measurement value is the altitude estimate of the UAV at a certain moment obtained by the sensing station through its sensing function. The baseline true value is the true value used as a reference standard in the accuracy evaluation. The second predicted altitude is the altitude component in the posterior state prediction data, which is the final altitude estimate of the UAV after filtering and updating at the second moment.

[0153] Instantaneous altitude error is the deviation between the altitude measurement value reported by the sensor station and the true reference value at a single trajectory point. This error reflects the real-time error of the sensor station in measuring the altitude of the UAV. It should be noted that the process for obtaining the instantaneous altitude error is as follows: ; in, For instantaneous altitude error, As the baseline truth value, This is the measured height.

[0154] The average height error is the arithmetic mean of the instantaneous height errors at multiple trajectory points. This data characterizes the systematic bias or accuracy of the sensor station's height measurement. The method for obtaining the average height error can be understood as follows: ; in, The instantaneous altitude error is N, where N is the number of trajectory points. This represents the average height error.

[0155] For example, if the average height error is close to zero, it indicates that the sensor station has no obvious systematic deviation and the measurement accuracy is high; if the average height error is significantly positive or negative, it indicates that there is a fixed calibration problem or model bias.

[0156] It should be noted that the altitude detection standard deviation is a statistical indicator that measures the dispersion of the altitude measurements of the drone by the sensor station. The calculation process for the altitude detection standard deviation is as follows: ; in, The standard deviation of the height detection is given, and N is the total number of valid trajectory points. Let be the instantaneous altitude error at time i. It is the arithmetic mean of all N instantaneous height errors.

[0157] Specifically, the drone altitude measurement values ​​reported by the sensor station are matched point-by-point with the second predicted altitude in the baseline true value to calculate the instantaneous altitude error. Based on this, the instantaneous altitude errors obtained by the drone at multiple trajectory points are statistically analyzed to obtain the average altitude error, which reflects the systematic bias of the sensor station's altitude measurement. Furthermore, the altitude detection standard deviation is calculated; this indicator measures the dispersion of the altitude measurement error and can comprehensively evaluate the accuracy and stability of the sensor station's drone altitude measurement.

[0158] Optionally, drone behavior metrics may include at least one of the following: flight speed, flight trajectory penetration rate, etc. Flight speed is the distance the drone travels per unit time, reflecting its speed. It should be noted that the flight speed is derived from the speed field in the actual flight control data. Flight speed includes instantaneous speed and average speed. Instantaneous speed is the speed value directly read from the actual flight control data. Average speed is the average speed calculated based on the estimated trajectory, the process of which is as follows: .in, The total displacement is calculated based on the sequence of trajectory points and accumulated using Euclidean distances.

[0159] Flight trajectory penetration rate is the degree of spatial agreement between the actual flight trajectory of a UAV and the preset planned route. This data comes from the trajectory point sequence in the actual flight control data of the UAV, including valid points and excluded points. The calculation process of flight trajectory penetration rate is as follows: ;in, The number of valid trajectory points, This represents the total number of trajectory points.

[0160] The technical solution of this invention acquires first actual communication signaling data and first actual flight control data of a UAV at multiple first moments. Using the time axis of the first actual flight control data as a reference, it matches the first actual communication signaling data with the closest timestamp for each first actual flight control data to achieve high-precision time synchronization of multi-source heterogeneous data, providing reliable input data for subsequent processing. Further, it determines the prior state prediction data of the UAV at a second moment based on the matched first actual communication signaling data and the first actual flight control data, and determines the predicted communication signaling data of the UAV at the second moment based on the prior state prediction data, thereby achieving advance prediction of future communication quality and improving the convergence speed of the filtering algorithm. It acquires the second actual communication signaling data of the UAV at the second moment, updates the prior state prediction data based on the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data, and the predicted communication signaling data to obtain posterior state prediction data, and determines the communication quality indicators associated with the UAV based on the posterior state prediction data and the second actual communication signaling data, improving the adaptability of communication quality testing and reducing reliance on high-precision, high-cost sensors. In summary, this technical solution reduces communication quality testing costs, improves UAV trajectory accuracy, and achieves a high degree of integration between perception and testing.

[0161] Figure 4 This embodiment provides a flowchart illustrating a communication quality detection method for unmanned aerial vehicles (UAVs). Building upon the foregoing embodiments, this embodiment elaborates on the process of determining the prior state prediction data of the UAV at a second time point based on first actual communication signaling data and first actual flight control data. Specific implementation details can be found in the description of this embodiment. Technical features identical or similar to those in the foregoing embodiments will not be repeated here.

[0162] like Figure 4 As shown, the method includes: S210. Obtain the first actual communication signaling data and the first actual flight control data of the UAV at multiple first moments. Based on the time axis of the first actual flight control data, match the first actual communication signaling data with the closest timestamp for each first actual flight control data.

[0163] S220. Using the first actual flight control data after matching as the initial value of the state vector, and the first actual communication signaling data after matching as the reference benchmark of the observation model, a state vector containing the flight control data of the UAV is constructed based on the uniform acceleration motion model. The prior state prediction data of the UAV at the second time moment is determined by the state prediction equation of Kalman filtering.

[0164] The state prediction equation is that the prior state prediction data at the current moment is equal to the state transition matrix multiplied by the posterior state prediction data at the previous moment. The state transition matrix is ​​used to describe the relationship between the flight control data of the UAV and the evolution over time.

[0165] In this embodiment, the first actual flight control data after matching is a valid UAV flight status record that has been time-aligned. The time alignment process involves matching the signaling data with the closest timestamp based on the timeline of the flight status data.

[0166] The initial value of the state vector is the first state estimate when the Kalman filter algorithm starts. This state estimate is the relevant data of the first trajectory point in the first actual control data after matching. The first actual communication signaling data after matching is valid communication signaling data that has been time-aligned.

[0167] The observation model is a function describing the mathematical relationship between communication signaling data and flight control data. The uniformly accelerated motion model is a dynamic model assuming that the UAV moves with constant acceleration over a short period of time. This model is used to describe the changes in position, velocity, and acceleration over time.

[0168] The state vector of a UAV's flight control data is a vector used to describe the UAV's complete motion state at a given moment. For example, the state vector at least includes the UAV's position, velocity, and acceleration in three-dimensional space, i.e. .in, The location is facing east. The location is north. The celestial position, For eastward speed, For northbound speed, For the speed of the sky, For eastward acceleration, For northward acceleration, For celestial acceleration, This is the state vector.

[0169] Kalman filtering is an optimal recursive data fusion algorithm. This algorithm estimates the true state of a system from noisy measurement data through two processes: prediction and update. The state prediction equation in Kalman filtering is used to predict the prior state estimate at the current time step from the posterior state estimate of the previous time step. This equation states that the prior state prediction data at the current time step equals the state transition matrix multiplied by the posterior state prediction data from the previous time step.

[0170] For example, after constructing the state vector, based on the uniformly accelerated motion model, the state prediction equation can be expressed as: ; in, The prior state prediction data for the current moment. The posterior state prediction data from the previous time step. Let k be the state transition matrix, and k represent time step.

[0171] It should be noted that the state transition matrix is ​​a linear matrix describing the evolution of the state vector from time k-1 to time k. This matrix is ​​used to describe the evolution of the UAV's flight control data over time.

[0172] Furthermore, the method for obtaining the state transition matrix is ​​explained. The state transition matrix is ​​obtained by discretizing the uniformly accelerated kinematics formula: ; in, Here is the state transition matrix. It is a 3×3 identity matrix; The time step is the time interval between the current trajectory point and the previous valid trajectory point. This is the time step multiplied by the identity matrix; this parameter represents the contribution coefficient of velocity to position. This represents the contribution coefficient of acceleration to position.

[0173] Specifically, the first matched actual flight control data is used as the initial value of the state vector, and the first matched actual communication signaling data is used as the reference benchmark for the observation model. Based on this, a state vector incorporating UAV flight control data is constructed using a uniformly accelerated motion model. Through the state prediction equation using Kalman filtering, the posterior state estimate from the previous time step is mapped to the prior state prediction data for the current time step using the state transition matrix. This determines the prior state prediction data for the UAV at the second time step, providing a foundation for subsequent observation updates and communication quality assessment.

[0174] Furthermore, after acquiring the prior state prediction data of the UAV at the second time point, the prior state prediction data is updated based on the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data. Next, the specific determination process of the posterior state prediction data is further refined. Optionally, updating the prior state prediction data based on the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data includes: using the position and velocity in the first actual flight control data as measured values, performing a Kalman filter measurement update to perform a first correction on the prior state prediction data to obtain intermediate state prediction data; using the intermediate state prediction data as the initial state, using the first and second actual communication signaling data as measured values, and the predicted communication signaling data as a comparison benchmark, performing an extended Kalman filter measurement update to obtain the posterior state prediction data.

[0175] In this embodiment, the measured value can be understood as the actual observation data obtained by the sensor in Kalman filtering. This actual observation data is used to correct the state prediction data. For example, the measured value can be the position and velocity from the first actual flight control data obtained from the UAV flight control platform.

[0176] The Kalman filter measurement update process involves using actual measurements to correct the prior state prediction in order to obtain the posterior state estimate.

[0177] It should be noted that, to provide better initial values ​​for communication signaling fusion, the first actual flight control data acquired by the UAV flight control platform is used as the measurement value for state estimation calibration, making the state estimation closer to the true value. Therefore, Kalman filtering measurement updates are performed, and the prior state prediction data is corrected for the first time. The above process is described in the following content: First, based on the position and velocity obtained from the UAV flight control platform as measured values, the observation function... It is a linear function used to extract the predicted position and predicted velocity from the state vector.

[0178] Furthermore, by calculating the innovation, the innovation covariance matrix, and the Kalman filter gain, and utilizing the GPS prediction noise covariance matrix... To measure the reliability of measured values, and thus to predict prior state data. and its covariance matrix Weighted corrections are performed to obtain the updated sum and covariance matrix. The updated data is then used as intermediate state prediction data to effectively suppress the accumulated error in motion model predictions, providing more accurate state data for the subsequent second correction of fused communication signaling data, thereby improving overall filtering accuracy.

[0179] Furthermore, after acquiring the intermediate state prediction data, the intermediate state prediction data is used as the initial state, and the first and second actual communication signaling data are used as measured values. The predicted communication signaling data is used as the comparison benchmark, and an extended Kalman filter measurement update is performed. The extended Kalman filter measurement update process is then described. This process includes: for the current measured value, calculating the difference between the measured value and the corresponding predicted communication signaling data as innovation; determining the Kalman gain based on the observation noise covariance matrix; correcting the current state prediction data based on the Kalman gain and innovation; updating the covariance matrix; and using the output state prediction data as the input state for the next measurement value update after completing the update of the current measured value.

[0180] In this embodiment, the current measurement value can be understood as the actual observation data obtained at the current filtering time. For example, the current measurement value can be the first actual communication signaling data and the second actual communication signaling data at the current filtering time.

[0181] The corresponding predicted communication signaling data is the estimated value of communication signaling calculated by an observation model based on the current prior state prediction data. The innovation is the difference between the actual predicted value and the predicted measured value. This difference reflects the degree of inconsistency between predictions. It should be noted that, in this embodiment, the innovation is the difference between the current measured value and the corresponding predicted communication signaling data.

[0182] The observation noise covariance matrix describes the covariance of the observation noise. This structure reflects the reliability of the measurements. The Kalman gain is the weighting coefficient in the Kalman filter. This coefficient determines the weight of the predicted and measured values ​​in the final state estimation.

[0183] Specifically, in the extended Kalman filter measurement update phase, for each successfully matched measurement value, the following update steps are executed sequentially. First, the measurement value bound to the current trajectory point is obtained. ,in, For measured values, The actual measured value of the reference signal received power. This is the actual measured value of the time lead. This is the actual measured value of the angle of arrival.

[0184] Furthermore, the difference between the current measurement and the corresponding predicted communication signaling data is calculated. This difference is then used as new information. ; in, This is the updated state prediction data from the previous step. For the new interest, The measured value is then used. Subsequently, the observation noise covariance matrix is ​​adaptively adjusted by a fuzzy logic controller based on the current received reference signal power and the innovation magnitude. This is to dynamically balance the reliability of measurements under different signal qualities.

[0185] Furthermore, the Jacobian matrix is ​​calculated, where each row corresponds to a measurement value and each column corresponds to a state variable, used to linearize the nonlinear observation function. Based on this, the observation noise covariance matrix is ​​then used... and the current state covariance matrix Calculate the Kalman gain: ; in, The Kalman gain determines the weights of the predicted and measured values ​​in the final state estimation. For Jacobian matrices, To observe the noise covariance matrix.

[0186] Furthermore, the Kalman gain is used to weight the innovation and correct the current state prediction data to obtain an updated state estimate. ,in, For Kalman gain, For the new interest, Based on the state prediction data from the previous step, This provides the state prediction data for the current step. The covariance matrix is ​​updated synchronously to reduce estimation uncertainty. After updating the current measurements, the output state prediction data is provided. Covariance Matrix As the input state for the next communication observation update, all available communication observations are processed sequentially until all are processed. After all update steps are completed, the final state prediction data and covariance matrix are the optimal estimates for the current time k, awaiting the next filtering cycle. The above process, by sequentially fusing multi-source nonlinear measurements, progressively optimizes the state estimation accuracy, achieving accurate correction of the UAV's position, velocity, and acceleration.

[0187] Furthermore, in the case where the obtained state prediction data is used as posterior state prediction data, optionally, in response to the fact that both the first actual communication signaling data and the second actual communication signaling data have been processed as measurement values, the obtained state prediction data is used as posterior state prediction data.

[0188] Specifically, during the extended Kalman filter measurement update process, the system sequentially uses the first and second actual communication signaling data as measurement values, updating them one by one in order. This involves correcting the current state prediction data and updating the covariance matrix. After completion, the output state is used as the input for the next measurement update. When all measurement values ​​contained in the first and second actual communication signaling data have been processed, it indicates that all available communication observation information at the current moment has been fully integrated into the state estimation. At this point, the final obtained state prediction data is used as the posterior state prediction data to ensure sufficient fusion of multi-source data. The sequential update mechanism improves the fusion accuracy, providing the optimal state benchmark for subsequent communication quality index calculations.

[0189] It should be noted that the process of obtaining the observation noise covariance matrix before determining the Kalman gain based on the observation noise covariance matrix will be described in detail below. Optionally, before determining the Kalman gain based on the observation noise covariance matrix, the process may also include: inputting the reference signal received power value and the normalized value of the innovation from the second actual communication signaling data into the fuzzy logic controller to obtain the noise adjustment factor; and multiplying the reference observation noise covariance matrix by the noise adjustment factor to obtain the observation noise covariance matrix.

[0190] In this embodiment, the observation noise covariance matrix is ​​a square matrix describing the noise variance of the measured values ​​and the covariances between them, and this square matrix reflects the reliability of the measured values.

[0191] The normalized value of the innovation is the innovation divided by its theoretical standard deviation. It should be noted that the normalized value of the innovation measures the degree of deviation between the predicted and actual measurements. For example, a normalized value close to 0 indicates that the predicted and actual measurements are consistent, while a large normalized value indicates a significant deviation between the predicted and actual measurements.

[0192] A fuzzy logic controller is a control system based on fuzzy set theory. It can be understood as a system that outputs a continuous control quantity by fuzzifying the input variables, applying a fuzzy rule base, and then defuzzifying them. The fuzzy rule base is a set of rules that map input variables to output variables in the fuzzy logic controller. It should be noted that this rule base is based on expert experience. For example, the normalized value of the reference signal received power and the innovation are input into the fuzzy logic controller, and a noise adjustment factor is output.

[0193] The noise adjustment factor is a coefficient output by the fuzzy logic controller. This coefficient is used to scale the baseline observation noise covariance matrix. It should be noted that multiplying the determined noise adjustment factor by the baseline observation noise covariance matrix yields the adjusted observation noise covariance matrix. in, To observe the noise covariance matrix, To adjust the factor, This is the power received as a reference signal.

[0194] For example, when the received power of the reference signal is high and the innovation is low, the noise adjustment factor is less than 1, and the weight is increased; when the received power of the reference signal is weak or the innovation is large, the noise adjustment factor is greater than 1, and the weight is decreased.

[0195] Specifically, before determining the Kalman gain based on the observation noise covariance matrix, the system first performs adaptive noise adjustment, that is, inputting the reference signal received power and the normalized value of the innovation from the second actual communication signaling data into the fuzzy logic controller. Further, the fuzzy logic controller outputs a noise adjustment factor based on a preset fuzzy rule base. Subsequently, this noise adjustment factor is multiplied by the baseline observation noise covariance matrix to obtain the adjusted observation noise covariance matrix. This matrix replaces the original fixed noise matrix and is used for subsequent Kalman gain calculation, thereby achieving dynamic adjustment of the measurement weights and enhancing robustness to prediction bias.

[0196] S230. Determine the predicted communication signaling data of the UAV at the second moment based on the prior state prediction data.

[0197] S240: Obtain the second actual communication signaling data of the UAV at the second moment; update the prior state prediction data according to the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data; and determine the communication quality index associated with the UAV based on the posterior state prediction data and the second actual communication signaling data.

[0198] The technical solution provided by this invention acquires first actual communication signaling data and first actual flight control data of a UAV at multiple first moments. Using the time axis of the first actual flight control data as a reference, it matches each piece of first actual flight control data with the first actual communication signaling data whose timestamp is closest, thereby achieving data acquisition and corresponding timestamp matching. This ensures that each piece of first actual communication signaling data is associated with the closest piece of first actual flight control data. Furthermore, using the matched first actual flight control data as the initial value of the state vector and the matched first actual communication signaling data as the reference benchmark for the observation model, a state vector containing the UAV's flight control data is constructed based on a uniformly accelerated motion model. The prior state prediction data of the UAV at the second time step is determined by the state prediction equation using Kalman filtering, providing a spatial reference for signaling data prediction. Based on the prior state prediction data, the predicted communication signaling data of the UAV at the second time step is determined. The second actual communication signaling data of the UAV at the second time step is obtained. The prior state prediction data is updated based on the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data, and the predicted communication signaling data to obtain the posterior state prediction data. Based on the posterior state prediction data and the second actual communication signaling data, the communication quality indicators associated with the UAV are determined to quantify the low-altitude coverage quality, improve the evaluation accuracy and reliability, and significantly reduce the testing cost.

[0199] Figure 5 This embodiment provides an overall framework diagram of a communication quality detection method for unmanned aerial vehicles (UAVs). Based on the above embodiment, an optional example is provided, which can be used for UAV communication quality detection scenarios. Figure 5 As shown, firstly, low-altitude flight path and signaling data are acquired and processed. Based on operator signaling data, operator parameter data, and UAV flight control data, the first actual communication signaling data and first actual flight control data of the UAV at multiple first moments are obtained. The above data undergoes data cleaning and standardization. Specifically, the data is first time-standardized, converting the signaling time and UAV flight record time into a unified Coordinated Universal Time (UTC) format. After time conversion, the data undergoes data cleaning, such as removing records of airborne signaling types and reference signal received power less than a preset power from the signaling data, and removing records from the UAV data whose latitude and longitude exceed a preset range.

[0200] Furthermore, high-precision trajectory estimation is achieved based on trajectory and signaling data fusion and Kalman filtering. First, the system is initialized; the system initialization process is described below. (See [link to relevant documentation]). Figure 6 .

[0201] First, the system's state vector is defined to include components such as the UAV's position, velocity, and acceleration. When GPS achieves its first effective lock, the position and velocity data output by GPS can be directly used as the initial values ​​of the state vector, and the GPS coordinates are converted to the ENU coordinate system. When GPS is not locked, the last effective position recorded by the system is used as the initial position.

[0202] Secondly, initialize the noise covariance matrix. The process noise covariance matrix reflects the uncertainty of the motion model, such as... ;in, For noise driving matrix, For noise driving matrix, Let be the intensity of the random perturbation of acceleration in the UAV motion model. The process of determining the noise driving matrix is ​​as follows: ,in, For discrete time steps, It is an identity matrix.

[0203] Furthermore, the observation noise covariance matrix is ​​initialized. This matrix is ​​diagonal, and the position and velocity are set according to the nominal accuracy of the GPS module. The reference variance of the communication measurements is determined through experimental measurement or theoretical analysis. To achieve adaptive adjustment of the observation noise, fuzzy logic controller parameters need to be configured. Input variables are defined, such as the received power value of the reference signal and the normalized value of the innovation. The output variable is the noise adjustment factor. Further, three fuzzy sets are defined, and a fuzzy rule base is established. When the signal quality is good and the model prediction is accurate, the observation noise variance is reduced, and the weight of the measured values ​​is increased.

[0204] After the system is initialized, see [link / reference]. Figure 6Using the timeline of the first actual flight control data of the UAV as a benchmark, the first actual communication signaling data with the closest timestamp is matched for each first actual flight control data, achieving time synchronization and data matching of the signaling data. Furthermore, a state vector containing position, velocity, and acceleration is constructed based on a uniformly accelerated motion model, and the prior state prediction data of the UAV at the next moment is obtained through Kalman filter prediction equations, quantifying the uncertainty of the prediction. In the measurement update phase, the actual communication signaling is used as the measured value. A fuzzy logic controller adaptively adjusts the observation noise covariance matrix based on the current reference signal received power value and the innovation magnitude, performing extended Kalman filter updates to sequentially update the posterior state prediction data (track data and communication traffic) and the covariance matrix, obtaining high-precision posterior state prediction data. It should be noted that in the measurement update phase, based on the innovation calculation and the reference signal received power value in the communication signaling, the weighting factors of the observation noise covariance matrix are dynamically adjusted through fuzzy logic inference. When the signal quality is good and the prediction deviation is small, the observation noise is reduced; when the signal quality is poor or the prediction deviation is large, the observation noise is increased, so that the filter can adapt to the complex low-altitude channel environment and improve robustness.

[0205] Optional, see Figure 6 The filter parameters are further optimized through hyperparameter optimization. Hyperparameters mainly include the baseline values ​​of the process noise covariance matrix and the observation noise covariance matrix, as well as the output adjustment factor of the fuzzy logic controller. These can be optimized through expert instructions or by the optimization methods described below. The optimization methods include: First, deploy high-precision reference equipment. In addition to conventional sensors, the UAV should be equipped with a high-precision positioning system as a source of ground truth, such as real-time dynamic differential GPS or post-processing dynamic differential GPS. Second, conduct multi-scenario data acquisition, i.e., design and execute test routes covering typical flight conditions. Optionally, test scenarios should include at least hovering tests, uniform straight-line flight, maneuvering flight, and environmental challenge flights. Hovering tests involve hovering at a fixed point to evaluate the system's stability and convergence in a static state. Uniform straight-line flight involves flying in a straight line at different speeds to evaluate the system's tracking accuracy in a steady state. Maneuvering flight includes S-turns, figure-eight patterns, acceleration and deceleration to evaluate the system's responsiveness to dynamic changes; environmental challenge flights involve switching between signal-blocked areas and open areas to evaluate the effectiveness of the adaptive mechanism. It should be noted that during all flights, a data acquisition system should be used to synchronously record all input data, such as GPS, IMU, signaling, and high-precision position truth values. Third, define the hyperparameter search space. The process noise covariance matrix is ​​determined by the standard deviation of acceleration noise. A search range for the acceleration noise standard deviation is defined, and a one-dimensional parameter network is constructed. The observation noise covariance matrix includes GPS and communication measurement components. A GPS noise benchmark is defined using a GPS position noise benchmark standard deviation, with a search range set for it. A communication noise benchmark is defined using a communication quantity benchmark noise standard deviation, with a search range set for it. Different adjustment strategies are set for the output of the fuzzy logic controller, such as setting discrete candidate values ​​for the output values ​​of the three fuzzy sets. Furthermore, the candidate values ​​of all defined hyperparameters are combined using Cartesian products to form a complete parameter network. A complete filtering process is performed for each hyperparameter combination in the network. Further, the algorithm is run using the current parameter combination to obtain an estimated trajectory. This estimated trajectory is time-aligned with the true trajectory, the root mean square error (RMSE) of the position estimation is calculated, and the current hyperparameter combination and its corresponding RMSE value are recorded in a result table. After evaluating all parameter combinations, the result table is sorted, and the hyperparameter combination that minimizes the RMSE value is selected as the optimal configuration parameter set.

[0206] Furthermore, an adaptive event-driven process is performed to align the time granularity of the UAV trajectory and signaling data. Enhanced event detection is used to identify target events during flight, such as macroscopic events, microscopic events, and communication events. The flight state is determined based on the event type. When the flight state is stable, a physically constrained cubic spline interpolation algorithm is used to upsample the posterior state prediction data, ensuring that the trajectory points and signaling data are consistent in time granularity. In dynamic states, dynamic time warping high-precision synchronization is performed to achieve non-linear time alignment between trajectory points and signaling points, ultimately outputting a high-precision spatiotemporally synchronized trajectory.

[0207] Furthermore, a low-altitude communication quality assessment process is performed. Based on the spatiotemporal synchronized trajectory and aligned signaling data, communication quality indicators, including effective coverage, average uplink / downlink rates, and average latency, are calculated. The posterior state prediction data is used as the baseline truth value and compared with the measurements reported by the sensing stations to evaluate the accuracy indicators such as position error, velocity error, and altitude error of the sensing stations. Finally, an assessment report is generated based on the above analysis.

[0208] Figure 7 This is a schematic diagram of a communication quality detection device for a drone provided in an embodiment of the present invention. Figure 7 As shown, the communication quality detection device includes: a data matching module 310, a data determination module 320, and a communication quality index determination module 330.

[0209] The system includes a data matching module for acquiring first actual communication signaling data and first actual flight control data of the UAV at multiple first moments, and matching the first actual communication signaling data with the closest timestamp for each first actual flight control data, based on the time axis of the first actual flight control data. A data determination module is used to determine the prior state prediction data of the UAV at a second moment based on the matched first actual communication signaling data and the first actual flight control data, and to determine the predicted communication signaling data of the UAV at the second moment based on the prior state prediction data. A communication quality index determination module is used to acquire second actual communication signaling data of the UAV at the second moment, update the prior state prediction data based on the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data, and the predicted communication signaling data to obtain posterior state prediction data, and determine the communication quality index associated with the UAV based on the posterior state prediction data and the second actual communication signaling data.

[0210] The technical solution of this invention acquires first actual communication signaling data and first actual flight control data of a UAV at multiple first moments. Using the time axis of the first actual flight control data as a reference, it matches the first actual communication signaling data with the closest timestamp for each first actual flight control data to achieve high-precision time synchronization of multi-source heterogeneous data, providing reliable input data for subsequent processing. Further, it determines the prior state prediction data of the UAV at a second moment based on the matched first actual communication signaling data and the first actual flight control data, and determines the predicted communication signaling data of the UAV at the second moment based on the prior state prediction data, thereby achieving advance prediction of future communication quality and improving the convergence speed of the filtering algorithm. It acquires the second actual communication signaling data of the UAV at the second moment, updates the prior state prediction data based on the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data, and the predicted communication signaling data to obtain posterior state prediction data, and determines the communication quality indicators associated with the UAV based on the posterior state prediction data and the second actual communication signaling data, improving the adaptability of communication quality testing and reducing reliance on high-precision, high-cost sensors. In summary, this technical solution reduces communication quality testing costs, improves UAV trajectory accuracy, and achieves a high degree of integration between perception and testing.

[0211] Based on the above embodiments, the data determination module includes: The prior state prediction data determination unit is used to construct a state vector containing the flight control data of the UAV based on a uniform acceleration motion model, using the matched first actual flight control data as the initial value of the state vector and the matched first actual communication signaling data as the reference benchmark of the observation model. The unit then determines the prior state prediction data of the UAV at the second time step using a Kalman filter state prediction equation. The state prediction equation states that the prior state prediction data at the current time step is equal to the state transition matrix multiplied by the posterior state prediction data at the previous time step. The state transition matrix is ​​used to describe the relationship between the flight control data of the UAV and the evolution over time.

[0212] Based on the above embodiments, the communication quality index determination module includes: The intermediate state prediction data acquisition unit is used to take the position and velocity in the first actual flight control data as measured values, perform Kalman filter measurement update, and perform the first correction on the prior state prediction data to obtain intermediate state prediction data. The posterior state prediction data acquisition unit is used to take the intermediate state prediction data as the initial state, take the first actual communication signaling data and the second actual communication signaling data as measurement values, take the predicted communication signaling data as the comparison benchmark, and perform extended Kalman filter measurement update to obtain posterior state prediction data. The extended Kalman filter measurement update process includes: For the current measurement value, the difference between the measurement value and the corresponding predicted communication signaling data is calculated as information. The Kalman gain is determined based on the observation noise covariance matrix. The current state prediction data is corrected based on the Kalman gain and the information, and the covariance matrix is ​​updated. After the update of the current measurement value is completed, the output state prediction data is used as the input state for the next measurement value update.

[0213] Based on the above embodiments, before determining the Kalman gain according to the observation noise covariance matrix, the method further includes: The noise adjustment factor acquisition module is used to input the reference signal received power value and the normalized value of the information in the second actual communication signaling data into the fuzzy logic controller to obtain the noise adjustment factor; The covariance matrix acquisition module is used to multiply the baseline observation noise covariance matrix with the noise adjustment factor to obtain the observation noise covariance matrix.

[0214] Based on the above embodiments, the prior prediction state information may include a first predicted position; the predicted communication signaling data may include at least one of the following: predicted reference signal received power, predicted timing advance, and predicted angle of arrival; further, the data determination module may include at least one of the following: The path loss model construction unit is used to determine the predicted reference signal received power of the UAV at the second time moment based on the spatial distance between the first predicted position and the base station position in the prior predicted state information and the path loss model. The path loss model is constructed based on the reference signal power, the path loss index, and the spatial distance between the position of the UAV and the base station position. The prediction time advance determination unit is used to determine the prediction time advance of the UAV at the second moment based on the spatial distance and speed of light between the first prediction position and the base station position in the prior prediction state information, and through the speed of light propagation delay model. The predicted angle of arrival determination unit is used to determine the predicted angle of arrival of the UAV at the second time moment based on the coordinate difference between the first predicted position and the base station position in the first direction and the second direction in the prior predicted state information, by using the four-quadrant arctangent function, wherein the first direction and the second direction are perpendicular.

[0215] Based on the above embodiments, the communication quality indicators include at least one of the following: effective coverage, average uplink rate and average downlink rate, and average latency; the communication quality indicators associated with the UAV determined based on the posterior state prediction data and the second actual communication signaling data include at least one of the following: The total number of first trajectory points of the second actual communication signaling data that match the posterior state prediction data is obtained, and the total number of second trajectory points whose reference signal received power value is greater than a preset quality threshold is obtained. The effective coverage associated with the UAV is determined based on the total number of the first trajectory points and the total number of the second trajectory points. The uplink instantaneous rate and downlink instantaneous rate are extracted from the second actual communication signaling data aligned with the posterior state prediction data. The duration and total test time corresponding to the second actual communication signaling data are determined based on the signaling timestamps corresponding to the second actual communication signaling data at multiple trajectory points of the UAV. The uplink average rate and downlink average rate associated with the UAV are determined based on the uplink instantaneous rate, the downlink instantaneous rate, the duration, and the total test time. Instantaneous delays are extracted from the second actual communication signaling data aligned with the posterior state prediction data, and the average delay associated with the UAV is determined based on the instantaneous delays of the UAV at multiple trajectory points.

[0216] Based on the above embodiments, after obtaining the posterior state prediction data, the method further includes: The target event detection module is used to detect target events of the UAV, including macroscopic events, microscopic events, and communication events; the macroscopic events include at least one of the following: takeoff, landing, hovering; the microscopic events include at least one of the following: sudden altitude change, horizontal acceleration, sharp signal drop; the communication events include at least one of the following: cell handover, wireless link failure, handover event. The second predicted position upsampling module is used to upsample the second predicted position in the posterior state prediction data in response to the absence of the target event, using a cubic spline interpolation algorithm with physical constraints, so that the trajectory point obtained after upsampling is consistent with the second actual communication signaling data in terms of time granularity. The trajectory point alignment module is used to align the trajectory points in the second actual communication signaling data with the posterior state prediction data based on the dynamic time warping algorithm in response to the detection of the target event.

[0217] Based on the above embodiments, after obtaining the posterior state prediction data, the method further includes: The sensor station accuracy index determination module is used to determine the sensor station accuracy index based on the posterior state prediction data as the reference true value and the measured value reported by the sensor station and the reference true value.

[0218] Based on the above embodiments, the accuracy indicators of the sensor station include at least one of the following: position error, instantaneous velocity error, and height error; the position error includes at least one of the following: instantaneous position error and root mean square error of position; the height error includes at least one of the following: instantaneous height error, average height error, and height detection standard deviation; the determination of the sensor station accuracy indicators based on the measured values ​​reported by the sensor station and the reference true value includes at least one of the following: The instantaneous position error is determined based on the position measurement value of the UAV reported by the sensor station and the second predicted position in the reference true value. The root mean square error of the position is determined based on the instantaneous position error of the UAV at multiple trajectory points. Align the velocity measurement modulus reported by the sensor station with the reference true velocity modulus, and calculate the instantaneous velocity error, which is the absolute value of the difference between the two modulus values; The instantaneous altitude error is determined based on the altitude measurement value of the UAV reported by the sensor station and the second predicted altitude of the UAV in the reference true value. The average altitude error is determined based on the instantaneous altitude error of the UAV at multiple trajectory points. The altitude detection standard deviation is determined based on the instantaneous altitude error and the average altitude error.

[0219] The communication quality detection device for UAVs provided in this embodiment of the invention can execute the communication quality detection method for UAVs provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0220] Figure 8 A schematic diagram of an electronic device that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0221] like Figure 8As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0222] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0223] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a communication quality detection method for a drone.

[0224] In some embodiments, a communication quality detection method for a drone may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the communication quality detection method for a drone described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform a communication quality detection method for a drone by any other suitable means (e.g., by means of firmware).

[0225] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0226] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0227] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0228] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0229] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0230] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0231] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication unit 19, or installed from storage unit 18, or installed from ROM 12. When the computer program is executed by processor 11, it performs the functions defined in the methods of the embodiments of the present invention.

[0232] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0233] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A communication quality detection method for unmanned aerial vehicles (UAVs), characterized in that, include: Acquire the first actual communication signaling data and the first actual flight control data of the UAV at multiple first moments, and use the time axis of the first actual flight control data as a reference to match the first actual communication signaling data with the closest timestamp for each first actual flight control data. Based on the matched first actual communication signaling data and the first actual flight control data, the prior state prediction data of the UAV at the second time moment is determined, and the predicted communication signaling data of the UAV at the second time moment is determined based on the prior state prediction data. The second actual communication signaling data of the UAV at the second time point is obtained. The prior state prediction data is updated according to the first actual flight control data, the first actual communication signaling data, the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data. The communication quality index associated with the UAV is determined based on the posterior state prediction data and the second actual communication signaling data.

2. The communication quality detection method for unmanned aerial vehicles according to claim 1, characterized in that, The step of determining the prior state prediction data of the UAV at the second moment based on the matched first actual communication signaling data and the first actual flight control data includes: Using the matched first actual flight control data as the initial value of the state vector and the matched first actual communication signaling data as the reference benchmark of the observation model, a state vector containing the flight control data of the UAV is constructed based on the uniform acceleration motion model. The prior state prediction data of the UAV at the second time moment is determined by the state prediction equation of Kalman filtering. The state prediction equation is that the prior state prediction data at the current time moment is equal to the state transition matrix multiplied by the posterior state prediction data at the previous time moment. The state transition matrix is ​​used to describe the relationship between the flight control data of the UAV and the evolution over time.

3. The communication quality detection method for unmanned aerial vehicles according to claim 2, characterized in that, The step of updating the prior state prediction data based on the second actual communication signaling data and the predicted communication signaling data to obtain the posterior state prediction data includes: Using the position and velocity in the first actual flight control data as measured values, Kalman filtering measurement update is performed to correct the prior state prediction data for the first time, and intermediate state prediction data is obtained. Using the intermediate state prediction data as the initial state, the first actual communication signaling data and the second actual communication signaling data as measurement values, and the predicted communication signaling data as the comparison benchmark, an extended Kalman filter measurement update is performed to obtain the posterior state prediction data. The extended Kalman filter measurement update process includes: For the current measurement value, the difference between the measurement value and the corresponding predicted communication signaling data is calculated as information. The Kalman gain is determined based on the observation noise covariance matrix. The current state prediction data is corrected based on the Kalman gain and the information, and the covariance matrix is ​​updated. After the update of the current measurement value is completed, the output state prediction data is used as the input state for the next measurement value update.

4. The communication quality detection method for unmanned aerial vehicles according to claim 3, characterized in that, Before determining the Kalman gain based on the observed noise covariance matrix, the method further includes: The reference signal received power value and the normalized value of the new information in the second actual communication signaling data are input into the fuzzy logic controller to obtain the noise adjustment factor; The observation noise covariance matrix is ​​obtained by multiplying the baseline observation noise covariance matrix by the noise adjustment factor.

5. The communication quality detection method for unmanned aerial vehicles according to claim 1, characterized in that, The prior prediction state information includes a first predicted position; the predicted communication signaling data includes at least one of the following: predicted reference signal received power, predicted time advance, and predicted angle of arrival; the step of determining the first predicted communication signaling data of the UAV at the second moment based on the prediction state information includes at least one of the following: Based on the spatial distance between the first predicted position and the base station position in the prior predicted state information, the predicted reference signal received power of the UAV at the second time moment is determined by the path loss model. The path loss model is constructed based on the reference signal power, the path loss exponent, and the spatial distance between the UAV position and the base station position. Based on the spatial distance between the first predicted location and the base station location and the speed of light in the prior predicted state information, the prediction time advance of the UAV at the second moment is determined by the light speed propagation delay model; Based on the coordinate difference between the first predicted position and the base station position in the prior predicted state information in the first and second directions, the predicted angle of arrival of the UAV at the second time moment is determined by the four-quadrant arctangent function, wherein the first and second directions are perpendicular.

6. The communication quality detection method for unmanned aerial vehicles according to claim 1, characterized in that, The communication quality metrics include at least one of the following: effective coverage, average uplink rate and average downlink rate, and average latency; the communication quality metrics associated with the UAV determined based on the posterior state prediction data and the second actual communication signaling data include at least one of the following: The total number of first trajectory points of the second actual communication signaling data that match the posterior state prediction data is obtained, and the total number of second trajectory points whose reference signal received power value is greater than a preset quality threshold is obtained. The effective coverage associated with the UAV is determined based on the total number of the first trajectory points and the total number of the second trajectory points. The uplink instantaneous rate and downlink instantaneous rate are extracted from the second actual communication signaling data aligned with the posterior state prediction data. The duration and total test time corresponding to the second actual communication signaling data are determined based on the signaling timestamps corresponding to the second actual communication signaling data at multiple trajectory points of the UAV. The uplink average rate and downlink average rate associated with the UAV are determined based on the uplink instantaneous rate, the downlink instantaneous rate, the duration, and the total test time. Instantaneous delays are extracted from the second actual communication signaling data aligned with the posterior state prediction data, and the average delay associated with the UAV is determined based on the instantaneous delays of the UAV at multiple trajectory points.

7. The communication quality detection method for unmanned aerial vehicles according to claim 5, characterized in that, After obtaining the posterior state prediction data, the following is also included: The UAV is subjected to target event detection, which includes macroscopic events, microscopic events, and communication events; the macroscopic events include at least one of the following: takeoff, landing, hovering; the microscopic events include at least one of the following: sudden altitude change, horizontal acceleration, sharp signal drop; the communication events include at least one of the following: cell handover, wireless link failure, handover event; In response to the absence of the target event, a cubic spline interpolation algorithm with physical constraints is used to upsample the second predicted position in the posterior state prediction data, so that the trajectory point obtained after upsampling is consistent with the second actual communication signaling data in terms of time granularity. In response to the detection of the target event, the second actual communication signaling data is aligned with the trajectory points in the posterior state prediction data based on the dynamic time warping algorithm.

8. The communication quality detection method for unmanned aerial vehicles according to claim 1, characterized in that, After obtaining the posterior state prediction data, the following is also included: Using the posterior state prediction data as the baseline true value, the accuracy index of the sensing station is determined based on the measured values ​​reported by the sensing station and the baseline true value.

9. The communication quality detection method for unmanned aerial vehicles according to claim 8, characterized in that, The accuracy indicators of the sensor station include at least one of the following: position error, instantaneous velocity error, and height error; the position error includes at least one of the following: instantaneous position error and root mean square error of position; the height error includes at least one of the following: instantaneous height error, average height error, and height detection standard deviation; the determination of the sensor station accuracy indicators based on the measured values ​​reported by the sensor station and the reference true value includes at least one of the following: The instantaneous position error is determined based on the position measurement value of the UAV reported by the sensor station and the second predicted position in the reference true value. The root mean square error of the position is determined based on the instantaneous position error of the UAV at multiple trajectory points. Align the velocity measurement modulus reported by the sensor station with the reference true velocity modulus, and calculate the instantaneous velocity error, which is the absolute value of the difference between the two modulus values; The instantaneous altitude error is determined based on the altitude measurement value of the UAV reported by the sensor station and the second predicted altitude of the UAV in the reference true value. The average altitude error is determined based on the instantaneous altitude error of the UAV at multiple trajectory points. The altitude detection standard deviation is determined based on the instantaneous altitude error and the average altitude error.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the communication quality detection method for unmanned aerial vehicles as described in any one of claims 1-9.

Images