AI optimization indoor distribution unit electrical element integration system for 5G and Beidou fusion positioning

Abstract

The invention relates to the technical field of radio navigation, and discloses an AI optimization indoor distribution unit electrical element integration system for 5G and Beidou fusion positioning, which comprises a main working path, a static calibration reference path and a signal sampling module, according to the system, internal electron drift is calibrated by comparing signal differences between a main working path and a static calibration reference path, additional phase drift introduced by external environment coupling is compensated by analyzing reflected signals of an antenna port, and then the two drifts are superposed to generate a full-link compensation instruction to pre-adjust the signals. According to the invention, internal electronic calibration and external physical environment perception are combined, and an end-to-end calibration closed loop from an electronic link to a physical transmitting aperture is constructed, so that the system can actively adapt to dynamic change of a deployment environment, and a communication unit is converted into a navigation reference point with reliable space-time reference.

Description

TECHNICAL FIELD

[0001] The application relates to an AI-optimized room distribution unit electrical element integrated system for 5G and Beidou integrated positioning, and belongs to the technical field of radio navigation. BACKGROUND

[0002] In current radio navigation applications, high-precision position services are provided by using widely deployed terrestrial wireless communication networks, which has become an important technical direction. A large number of fifth-generation mobile communication technology (5G) room distribution units are set in indoor environments, and due to the intensive signal coverage, they provide a physical layer basis for high-precision indoor positioning. However, when these room distribution units designed to ensure communication quality are directly applied to radio navigation tasks that have high requirements for signal timing characteristics, the inherent technical constraints in their design begin to limit their application. Room distribution units are communication products produced on a large scale, and their design goal is to balance cost and performance under the premise of meeting communication protocol requirements. Therefore, during the manufacturing process, there are process tolerances in each electrical element on the internal radio frequency link, and the electrical characteristics will dynamically drift with changes in operating temperature, power supply, and component aging. For communication services, the system can compensate for this through channel coding or retransmission mechanisms. However, in high-precision positioning applications, each room distribution unit must serve as a stable reference point in space and time. Nanosecond-level time delay changes in signals on the radio frequency link can directly cause errors in positioning results, which makes the aforementioned physical layer characteristic fluctuations acceptable in the communication field become a source of error that must be addressed in positioning applications.

[0003] However, existing technologies not only have the aforementioned uncertainty in physical characteristics at the hardware level, but also have inherent limitations in signal fusion and error processing control methods, making it difficult to adapt to complex and variable actual deployment environments. For example, a Chinese invention patent with the authorization announcement number CN118131292B discloses a positioning system that fuses 5G and Beidou signals. After Beidou satellite positioning, a fixed threshold parameter based on the 5G base station signal coverage range is used to remove ambiguous solutions that do not conform to geographical common sense. The fundamental flaw of this method is that the threshold parameter used is a fixed, pre-set static value, which cannot dynamically adapt to the real-time changes in the actual distribution density of 5G base stations and signal coverage in different areas. When the positioning terminal is at the edge of the base station coverage or in a complex signal area, this one-size-fits-all filtering method based on static parameters is prone to incorrect judgments, making it difficult to effectively fuse Beidou and 5G signals, and even possibly losing 5G signals due to the incorrect removal of valid positioning points, thereby reducing positioning accuracy.

[0004] To solve this problem, the direct technical idea is to require all room division units to use high-stability precision components and to perform strict factory calibration, but this will result in a significant increase in the cost of individual room division units, making them economically unfeasible in high-density deployment scenarios, and deviating from the technical intention of using existing communication facilities for positioning. The field is faced with a difficult technical choice in engineering practice: on the one hand, high-precision positioning requires near-instrument-level electrical performance consistency from a large number of network terminal nodes, and on the other hand, large-scale deployment has a realistic need for low-cost non-precision communication components in terms of economy. In this regard, the existing technology mainly has the following shortcomings: 1. Lack of a technical method to ensure that different production batches of low-cost room division units deployed on a large scale can exhibit uniform electrical performance in operation, avoiding unpredictable fluctuations in the performance of the entire positioning system; 2. At the level of an individual room division unit, there is a lack of a system-internal mechanism that can measure and actively compensate for the dynamic drift of its own radio frequency link due to changes in working conditions in real time, so that its own stability as a measurement reference point cannot be guaranteed. Therefore, how to propose a new system integration method that does not rely on high-cost precision components, so that each independent room division unit deployed on a large scale can actively calibrate and maintain the stability of its own radio frequency link electrical characteristics in real time, and convert it from a performance-uncertain communication signal transmission system to a precise and reliable radio navigation reference point in space and time, is the technical problem to be solved by the present application. SUMMARY

[0005] The present application provides an AI-optimized room division unit electrical element integration system for 5G and Beidou integrated positioning, which mainly aims to solve the problem of how to enable large-scale deployed room division units to actively perform real-time self-calibration to maintain the stability of their radio frequency link electrical characteristics without relying on high-cost precision components, thereby addressing the problem of balancing positioning reference consistency and deployment economy in the prior art.

[0006] To achieve the above-mentioned purpose, the AI-optimized room division unit electrical element integration system for 5G and Beidou integrated positioning provided by the present application comprises: a digital baseband processing unit; a main working path connected to the digital baseband processing unit, which contains elements with dynamically changing electrical characteristics under working conditions; a static calibration reference path, whose electrical characteristics remain stable under changes in working conditions; a signal switching module configured to selectively send the probe signal output from the digital baseband processing unit to the main working path and the static calibration reference path; a signal sampling module, located before the antenna port of the main working path, configured to sample the probe signal propagating forward to the antenna port and sample the probe signal reflected back from the antenna port; The digital baseband processing unit is configured to: compare the electrical characteristic difference of the probe signal propagating forward through the main working path and the static calibration reference path to determine the real-time drift error caused by the dynamically changing element of the main working path as the first error source; analyze the characteristic of the reflected probe signal and compare it with a reference reflection characteristic predetermined under a standard environment to determine the additional phase drift caused by the coupling effect between the antenna port and the external physical environment as the second error source; superimpose the additional phase drift and the real-time drift error to generate a full-link compensation instruction, and apply the full-link compensation instruction to pre-adjust the radio frequency signal when transmitting the radio frequency signal.

[0007] Preferably, the digital baseband processing unit is further configured to perform a digital source self-calibration before determining the real-time drift error of the main working path, the digital source self-calibration including the following rules: loop back the probe signal before sending it to the digital-to-analog converter to form a pure digital loop within the digital baseband processing unit; determine the digital zero-point drift of the digital signal processing link by analyzing the probe signal passing through the pure digital loop; and apply the digital zero-point drift to pre-compensate the signal generation process when generating the probe signal.

[0008] Preferably, the digital baseband processing unit is further configured to: periodically generate a dual-tone probe signal composed of at least two single-tone signals of different frequencies and send it into the main working path via the signal switching module; perform spectral analysis on the dual-tone probe signal sampled after passing through the main working path via the signal sampling module to detect the intermodulation distortion component generated by the nonlinear characteristic of the main working path; determine the nonlinear distortion index of the main working path based on the characteristic of the intermodulation distortion component; and use the nonlinear distortion index as an input parameter for adjusting the generation algorithm of the full-link compensation instruction when generating the full-link compensation instruction.

[0009] Preferably, the digital baseband processing unit is further configured to: generate and send a probe sequence composed of at least two probe signals of different power levels into the main working path via the signal switching module when determining the real-time drift error of the main working path; measure the drift error of each probe signal in the probe sequence after passing through the main working path respectively; construct a power-drift mapping model representing the main working path based on the measured drift error related to the power level, which is used to adjust the real-time drift error of the probe signal according to the instantaneous power of the radio frequency signal outputting a power-dependent compensation value; and the digital baseband processing unit superimposes the real-time drift error and the power-dependent compensation value outputted from the power-drift mapping model according to the instantaneous power of the radio frequency signal when generating the full-link compensation instruction.

[0010] Preferably, the digital baseband processing unit is configured to analyze the characteristics of the reflected probe signal by extracting a group delay frequency response curve of the reflected probe signal as a real-time reflection characteristic of the coupling effect between the antenna port and the external physical environment; the reference reflection characteristic is a reference group delay frequency response curve calibrated under a preset standard environment; the digital baseband processing unit determines the additional phase drift by calculating a difference vector between the real-time reflection characteristic and the reference reflection characteristic, and inputting the difference vector into a model calibrated before leaving the factory to convert the difference vector into the additional phase drift.

[0011] Preferably, the digital baseband processing unit is further configured to record time series data of the historical real-time drift error and the additional phase drift; establish a time series model representing the aging trend of the element or the change trend of the external physical environment based on the time series data; calculate a quantitative index representing the change rate of the trend according to the time series model; and output predictive maintenance alarm information when the quantitative index exceeds a preset alarm threshold.

[0012] Preferably, the static calibration reference path is a microstrip line laid on a circuit board and composed of a high-stability medium, the physical length and the dielectric constant of which remain constant when the working condition changes.

[0013] Preferably, the way in which the digital baseband processing unit applies the full-link compensation instruction to pre-adjust the radio frequency signal is that, inside the digital baseband processing unit, a digital pre-distortion technology is used to apply a pre-adjustment of phase and delay to the baseband digital signal about to enter the main working path, the value of the pre-adjustment being equal in magnitude and opposite in direction to the value of the full-link compensation instruction.

[0014] Preferably, the probe signal is a broadband probe signal, the spectral width of which covers the entire frequency band of the system operation.

[0015] Preferably, the all-digital loop includes a first digital loop and a second digital loop which are physically mirror-symmetric; the digital baseband processing unit is further configured to measure the differential delay of the signal when passing through the first digital loop and the second digital loop to determine the non-uniform delay caused by the thermal gradient on the chip; determine a thermal-induced error compensation value based on the non-uniform delay; and subtract the thermal-induced error compensation value from the analysis result of the probe signal passing through the first digital loop when determining the digital zero-point drift.

[0016] Compared with the prior art, the present application has the following beneficial effects: 1. A traceable whole-link time-space reference stability mechanism is established by the cooperation of digital source self-calibration and analog link drift measurement. Before the system measures the electrical characteristics of the main working path, the digital baseband processing unit is first used to pre-calibrate the probe signal itself with a pure digital loop, which isolates the digital signal source drift caused by changes in chip operating conditions and obtains a stable digital probe signal that can be used as a reference. Then, the calibrated probe signal is sent to the main working path and the static calibration reference path. By comparing the differences in electrical characteristics of the two outputs, the real-time drift error obtained purely reflects the physical changes of the analog link. This two-step method of first calibrating the measurement tool and then performing physical measurement avoids the industry technical problem of confusing digital source errors with analog link errors and making incorrect compensation, so that the final generated compensation instructions correct the analog physical link which is accurately locked, thereby making the signal transmission timing reference of the entire chamber unit obtain double stability in logic and physics.

[0017] 2. The present application combines the calibration of the electronic characteristics inside the radio frequency link with the real-time perception of the coupling effect of the external environment on the antenna port, and realizes an end-to-end calibration closed loop from the unit interior to the signal physical emission aperture. While the system determines the real-time drift error inside the main working path using the embedded static calibration reference loop, its signal sampling module also samples the probe signal reflected back from the antenna port. The digital baseband processing unit analyzes the characteristic changes of the reflected signal to reverse the additional phase drift caused by changes in the external environment such as antenna cover aging and attachments. Finally, the compensation instructions generated by the system are the superposition of two independent measurements of internal electronic link drift and external physical aperture additional drift, which are variables of different physical causes. This method enables the system not only to correct the state changes of its internal components, but also to actively adapt to dynamic changes in its external physical environment, extending the calibration boundary from the circuit board to the actual wireless signal emission origin.

[0018] 3. In the process of linear time delay phase calibration of radio frequency link, the ability of synchronous diagnosis of link non-linear characteristics is integrated, the function of the system is expanded from single error correction to long-term monitoring of link health status, in addition to periodically generating wideband probe signal for measuring linear drift, the system further generates probe signal composed of at least two single tone signals of different frequencies and sends it into the main working path; the digital baseband processing unit detects the intermodulation distortion components generated by the non-linearity of power amplifier and other elements by spectrum analysis on the sampling signal, the characteristics of the intermodulation distortion components are used to determine a quantitative index representing the degree of link non-linearity, this index and the linear drift compensation value exist in parallel, the former is used to correct the immediate space-time reference of the signal, and the latter is used as a long-term health status indicator reflecting the aging trend of key devices, providing a new dimension for the stability and maintainability of the entire positioning system. BRIEF DESCRIPTION OF DRAWINGS

[0019] Fig. 1 The system full-link self-calibration workflow of the present application; Fig. 2 The system function and interaction use case diagram of the present application; Fig. 3 The system self-calibration hardware structure block diagram of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor belong to the scope of protection of the present application.

[0021] The application discloses an AI optimization room distribution unit electrical element integrated system for 5G and Beidou integrated positioning, which is a space-time reference point in a wireless navigation network, and integration at a system level of the system comprises a digital baseband processing unit, a main working path, a static calibration reference path, a signal switching module and a signal sampling module; the system establishes an end-to-end calibration closed loop from a digital signal source to a physical transmitting aperture through a periodic self-calibration process, compensates for signal timing errors caused by internal element dynamic drift and external environment coupling effects, and thus ensures that the transmitted navigation positioning signal has reference stability; in the wireless navigation application, the stability of each room distribution unit as a measurement reference point in a large-scale deployed indoor positioning network is difficult to guarantee, and to deal with this situation, the system is configured to first perform a digital source self-calibration to isolate digital signal source drift caused by chip working condition changes; specifically, a wideband probe signal generated by a digital oscillator in the digital baseband processing unit is copied and looped back through a pure digital loop before being sent to a digital-to-analog converter, and the pure digital loop comprises a first digital loop and a second digital loop which are physically mirror-symmetrically arranged on the chip; the digital baseband processing unit determines non-uniform delay caused by a thermal gradient generated by uneven power consumption of different functional modules on the chip by measuring nanosecond-level differential delay of a test pulse signal when passing through the first digital loop and the second digital loop; then, based on a thermal delay model calibrated through thermal simulation before leaving the factory, the differential delay is converted into a thermal error compensation value; the system determines a total digital drift containing thermal influence by analyzing the loopback signal of the first digital loop carrying the wideband probe signal, and subtracts the thermal error compensation value from the total digital drift, so as to obtain a pure digital zero-point drift which can be used as a calibration reference for subsequent analog link calibration; thereafter, the digital zero-point drift is applied to the signal generation process for pre-compensation when generating all signals for external link detection, so that a digital probe signal which can be used as a reference is obtained.

[0022] After the calibration of the measurement tool is completed, the system then measures the drift of the analog physical link in real time; since the electrical elements on the room distribution unit radio frequency link will dynamically drift with working temperature, power supply changes and element aging, the system uses the signal switching module to selectively send the wideband probe signal calibrated by the digital source into the main working path and the static calibration reference path; wherein, the main working path contains elements such as power amplifiers and filters whose electrical characteristics dynamically change with working conditions, and the static calibration reference path is a section of circuit board arranged on the circuit board with a dielectric constant temperature coefficient less than 10ppm / ℃, which is used to simulate the static state of the main working path; the system measures the differential delay of the wideband probe signal in the main working path and the static calibration reference path through the signal sampling module, and then determines the total drift of the main working path and the static calibration reference path; the system then subtracts the total drift of the static calibration reference path from the total drift of the main working path, so as to obtain the drift of the analog physical link. a microstrip line with high stable medium, whose physical length and dielectric constant remain constant when the working condition changes; a signal sampling module configured as a directional coupler, which is located before the antenna port of the main working path and samples the probe signal propagating forward to the antenna port; when the probe signal flows through the main working path, the signal sampling module samples it for the first time; then, the signal switching module sends the probe signal into the static calibration reference path, and the signal sampling module samples it for the second time; the digital baseband processing unit directly calculates the real-time drift error caused by the dynamic change of the elements in the main working path as the first error source by comparing the phase and group delay difference of the two sampling signals; to extend the calibration boundary from the circuit board to the actual wireless signal emission origin, compensate for the additional phase drift caused by the aging of the radome, attachments and other external environmental changes, the signal sampling module is also configured to sample the probe signal reflected back from the antenna port; the digital baseband processing unit extracts the group delay frequency response curve of the reflected broadband probe signal as the real-time reflection characteristic representing the coupling effect between the antenna port and the external physical environment, and compares it with a reference group delay frequency response curve calibrated in a standard environment, i.e. temperature 25 , relative humidity 50%, no reflector within 1 meter in front of the antenna in anechoic chamber; by calculating the difference vector between the real-time reflection characteristic and the reference reflection characteristic, and inputting the difference vector into a lookup table model established through multi-physical field simulation before leaving the factory, which converts the difference vector into additional phase drift, to determine the additional phase drift caused by the external physical environment coupling effect as the second error source.

[0023] Finally, the digital baseband processing unit vector-superimposes the real-time drift error (the first error source) with the additional phase drift (the second error source) to generate a full-link compensation command. When the system transmits normal 5G or BeiDou radio frequency signals, this command is applied to the digital predistortion module within the digital baseband processing unit. This module applies a pre-adjustment of phase and delay to the baseband digital signal about to enter the main working path. The magnitude of this pre-adjustment is equal to and opposite in direction to the magnitude of the full-link compensation command, thus pre-canceling the drift error that will be introduced into the full link before the signal enters the physical link. It should be noted that when the digital baseband processing unit performs the full-link self-calibration process, the transmission of its probe signal and the transmission of normal 5G service data are time-domain multiplexed. Specifically, this unit is configured to monitor downlink physical downlink sharing in real time. The channel is the resource block (RB) scheduling diagram of the PDSCH. When an idle RB not occupied by user data is identified within the current transmission time interval (TTI), the baseband data stream of the probe signal is sent only to the time-frequency positions corresponding to these idle RBs. Normal service data transmission of adjacent RBs is unaffected. The entire self-calibration measurement process is completed using one or more non-continuous TTI service idle gaps, thus periodically acquiring link status without interrupting the main communication service or reducing data throughput. To extend the system's function from simple error correction to long-term monitoring of link health, the digital baseband processing unit is also configured to periodically generate a sequence of frequencies... and The two single-tone signals constitute a dual-tone detection signal, which is then sent to the main working path; in addition, for the dual-tone detection signal used to monitor the nonlinear characteristics of the link, the frequency components of its two single-tone signals are... and The determination follows a set of engineering procedures to ensure the validity of the measurement. First, the interval between the two frequencies... It is set to be greater than the reciprocal of the signal sampling window duration so that the two signals can be clearly distinguished in the spectrum after processing by Fast Fourier Transform (FFT); secondly, The bandwidth was set to be less than one-tenth of the system channel coherence bandwidth, making the multipath fading characteristics experienced by the two signals highly correlated. This effectively eliminated the influence of channel frequency-selective fading when calculating the intermodulation distortion component. Finally, based on the technical specifications of the key components in the main working path, the third-order intermodulation distortion component was calculated. and The frequency of the signal sampling module is determined by constraints, ensuring it falls within the passband of the intermediate frequency filter. These constraints collectively define an effective frequency pair selection space, as given in the preceding embodiments. and The combination of the above is a specific technical implementation in this space. The signal sampling module samples the signal after passing through the main working path, and the digital baseband processing unit then performs spectral analysis on the sampled signal to detect the intermodulation distortion components generated by the nonlinear characteristics of the main working path at frequency points; by calculating the ratio of the power of the intermodulation distortion components to the power of the original dual-tone signal, a quantitative index representing the degree of nonlinearity of the link is determined, i.e. the third-order intermodulation suppression ratio; this index serves as a long-term health status indicator reflecting the aging trend of the key device.

[0024] At the same time, the system records the time series data of the historical real-time drift error and the additional phase drift, and based on these data, establishes an autoregressive moving average time series model for representing the trend of element aging or external environment change; when the change rate of the predicted value of the model within a preset time window exceeds the alarm threshold, for example, the phase drift rate is greater than , the system outputs predictive maintenance alarm information, providing a new dimension for the stability and maintainability of the positioning system; considering the power-dependent phase transmission characteristics of the power amplifier in the radio frequency link, i.e. the AM-PM effect, to achieve instantaneous value compensation for accurate matching with the real-time power of the signal, the digital baseband processing unit is also configured to generate and quickly send a probe sequence composed of at least two different power level probe signals, for example, one with a power of 0 dBm and the other with a power of 20 dBm; the system measures the drift error of each probe signal in the probe sequence after passing through the main working path, and based on these discrete drift error measurement points related to the power level, a power-drift mapping model representing the main working path is constructed through linear interpolation, which is used to output a power-dependent compensation value according to the instantaneous power of the radio frequency signal ; when generating the full-link compensation instruction, the digital baseband processing unit superimposes the real-time drift error, the additional phase drift, and the power-dependent compensation value output from the power-drift mapping model according to the instantaneous power of the to-be-transmitted radio frequency signal.

[0025] ​Embodiment 1: In a large intelligent warehouse center providing high-precision operation guidance for an autonomous mobile robot (AMR), the operation requirement of the scene is that the AMR needs to complete the material docking with an error of not more than 1 cm between dynamically changing shelves; and the warehouse center is deployed with hundreds of 5G room division units from different production batches, and the performance of the entire positioning system is unstable during the midday communication business peak period, the positioning results of the AMRs in some areas frequently fluctuate, and the operation requirements cannot be met, resulting in an increase in the docking failure rate; when the operation and maintenance system of the warehouse center detects that the positioning accuracy in a specific area continuously decreases and exceeds the preset threshold, the technical solution of the present application is triggered to be executed in a room division unit in the area; the system first enters the digital source self-calibration stage; since the digital baseband processing unit of the unit has formed a temperature gradient on the surface of the chip due to long-time high-load operation, the digital baseband processing unit determines the non-uniform delay caused by the thermal gradient by measuring the differential delay of the test pulse signal between the first digital loop and the second digital loop arranged in mirror symmetry inside, and generates a thermal error compensation value according to the thermal error compensation value to correct the total digital drift measured in the first digital loop subsequently, so as to obtain a pure digital zero-point drift that is not affected by the current thermal working condition of the chip; this provides a stable reference for all subsequent measurements.

[0026] Subsequently, the system enters the synchronization calibration phase of the analog link with the external environment; a broadband probe signal calibrated by the aforementioned digital source is sequentially sent into the main working path and the static calibration reference path; the digital baseband processing unit determines the real-time drift error of the main working path caused by the temperature rise of the power amplifier by comparing the difference between the two sampled signals; at the same time, the signal sampling module captures the probe signal reflected back from the antenna port, and the digital baseband processing unit finds a small difference vector between the reference reflection characteristic and the group delay frequency response curve of the reflected signal by analyzing the reflected signal; the difference vector is input into the internal model to solve the additional phase drift caused by a layer of thin dust on the surface of the radome; the cooperative measurement of the additional phase drift and the aforementioned real-time drift error reveals that the error source of the cell unit at this moment is the superposition of internal electronic drift and external physical environment pollution; considering that this moment is in the peak of communication service, the system further performs power-dependent calibration; a probe sequence composed of two different power levels of 0dBm and 20dBm probe signals is injected into the main working path; the system quickly constructs a power-drift mapping model representing the current state of the power amplifier by measuring the drift error of each probe signal; finally, a full-link compensation instruction composed of the real-time drift error, the additional phase drift, and the power-dependent compensation value obtained from the power-drift mapping model according to the instantaneous transmit power under the current high service load is generated and applied to the digital pre-distortion adjustment of all to-be-transmitted radio frequency signals; after the application of the full-link compensation instruction, the space-time reference of the positioning signal transmitted by the cell unit is restored to stability, the AMR positioning result jumping phenomenon covered by it disappears, and the material docking success rate returns to the normal level; the whole process converts a dynamic multi-factor coupled performance drift problem into a static separable path difference, environmental reflection characteristic difference and power response characteristic difference problem, and resolves the contradiction between using low-cost communication-level components and achieving high-precision navigation reference stability through the synchronous measurement and compensation of the three differences.

[0027] To further illustrate the fundamental improvement of the technical scheme of the present application relative to the conventional internal calibration technology, the following comparative example is provided.

[0028] Comparative Example 1: To verify the technical necessity of the end-to-end calibration method of the present application which combines internal electronic link calibration with external physical environment awareness, a special test platform was built, and a comparative test system was prepared which only contains a conventional internal calibration loop. The comparative system is basically consistent with the system in the embodiment of the present application in terms of hardware structure, and both include a digital baseband processing unit, a main working path, a static calibration reference path and a signal sampling module. The only essential difference between the two is that the comparative system is configured to calculate and compensate for real-time drift errors caused by internal elements by comparing the signal differences between the main working path and the static calibration reference path, but it does not have the function of analyzing antenna port reflection signals, so it does not have the ability to perceive and compensate for additional phase drift caused by external physical environment coupling effects. The test was carried out in a programmable oven, and a vector network analyzer (phase measurement resolution 0.1 picoseconds) was used to accurately monitor the signal phase changes at the radio frequency output port of the comparative system.

[0029] The test procedure is as follows, initial state calibration: under the standard ambient temperature of 25 , the comparative system is powered on and its internal calibration process is executed, and the output signal phase measured by the vector network analyzer at this moment is recorded and defined as the zero point reference of 0.0 picoseconds; internal thermal drift simulation: the environmental temperature of the programmable oven is gradually increased from 25 to 65 to simulate the thermal drift of internal elements caused by long-term high-load operation of the system, after the oven temperature stabilizes at 65 , the internal calibration process of the comparative system is executed again, and the phase drift at this moment is recorded; external environment coupling simulation: under the condition that the oven temperature is maintained at 65 , a layer of polytetrafluoroethylene dielectric plate with a thickness of 2 millimeters is tightly covered at the antenna port of the comparative system, which aims to simulate the additional coupling effect caused by changes in the external environment such as aging, dust or icing of the antenna cover, which is not caused by internal electronic links. The internal calibration process of the comparative system is executed again, and the final cumulative phase drift is recorded. The key data during the test process is recorded in Table 1.

[0030] Table 1: Test data table of Comparative Example 1.

[0031] The test data shows that under the condition of only internal temperature change (from 25 to 65 ), the comparison system can effectively suppress the phase drift caused by the internal elements to within 2.3 picoseconds through its internal static calibration reference path, showing good internal compensation capability. However, when an external medium is introduced to simulate the change in the antenna port environment, the internal calibration mechanism of the comparison system cannot compensate for the additional phase drift caused by the external introduction because the comparison system cannot perceive the change in the reflection characteristics of the antenna port in its design principle, resulting in a dramatic increase in the final cumulative phase drift to 36.5 picoseconds. The test results show that a technical solution that relies only on an internal reference path for calibration can solve the drift problem caused by the change in the working conditions of internal elements to some extent, but it is completely powerless to the signal timing error caused by the coupling effect of the antenna port and the external physical environment.

[0032] Example 2: To objectively verify the effectiveness of the technical solution of the present application in suppressing the timing error caused by the combined effects of internal element thermal drift and external antenna port environment coupling, a special test platform was built. The platform consists of a programmable oven, a vector network analyzer and a standard 5G signal generator. The programmable oven is used to simulate the working state of the room division unit at different environmental temperatures, and its temperature control accuracy is ±0.5 , the vector network analyzer is used to measure the signal phase change of the room division unit radio frequency output port, and its measurement resolution reaches 0.1 picoseconds. During the test, a sample group using the technical solution of the present application, i.e. the sample group of the present application, and three control sample groups were selected for comparative testing. Control group 1 is a standard room division unit without any self-calibration function, control group 2 is a room division unit with only internal real-time drift error compensation function, and control group 3 is a room division unit with only external additional phase drift compensation function. Before the test, all sample groups were initially calibrated at a standard environment of 25 , and the phase of their output signals was taken as the zero point reference. In the first stage of the test, all sample groups were placed in the programmable oven, and the environmental temperature was gradually increased from 25 to 65 to simulate the thermal drift caused by the change in the working conditions of the elements, and the phase drift of the output signals of each sample group was recorded at multiple temperature points. In the second stage of the test, when the temperature of the oven was stabilized at 65 , a layer of polytetrafluoroethylene medium plate with a thickness of 2 millimeters was covered on the antenna port of each sample group to simulate the coupling effect introduced by the aging or dust accumulation of the radome and other external environmental changes, and the phase drift of the output signals of each sample group was recorded again. The key data during the entire test process is recorded in Table 2.

[0033] Table 2: Phase drift test data table of different sample groups under different test conditions.

[0034] The test data of Table 2 shows that the phase shift of the control group 1 accumulates to 98.6 ps under the double influence of temperature rise and external medium coverage, and the timing reference stability decreases; the control group 2 can suppress the phase shift within 1.9 ps in the simple temperature rise process due to the internal real-time drift error compensation function, but the phase shift increases to 35.1 ps after the introduction of external medium coverage, because it cannot perceive the external environment change of the antenna port; the control group 3 has the same drift data as the control group 1 because it only has the external additional phase shift compensation function and has no correction ability for the internal element thermal drift; the maximum phase shift of the inventive sample group is only 2.1 ps in the whole test process through the synergistic work of internal electronic link drift compensation and external physical aperture additional phase shift compensation; the test results show that the AI optimized room distribution unit electrical element integrated system for 5G and Beidou integrated positioning combines the real-time calibration of internal electronic characteristics with the synchronous perception of the external environment coupling effect of the antenna port to build an end-to-end calibration closed loop.

[0035] Embodiment 3: This embodiment combines Figs. 1 to 3 to illustrate an AI optimized room distribution unit electrical element integrated system for 5G and Beidou integrated positioning, as Fig. 1 shown, the process starts from the baseband digital signal of the external service signal source, which is adjusted by the application of the digital pre-distortion link, one way as a probe signal into the main working path and interacts with the antenna and external physical environment, the other way is used to measure the internal electronic drift, the system executes two processing paths in parallel: one is to determine the additional phase shift by analyzing the physical reflection signal of the antenna port and comparing it with the reference reflection characteristics, the other is to calculate the real-time drift error by processing the measured sampling signal, finally, the real-time drift error and the additional phase shift are superimposed to generate the full-link compensation instruction, which is fed back to the application of the digital pre-distortion link on one hand, and the generation process is also supported by the calibration model library on the other hand, in addition, the system also monitors the link health status and analyzes the historical timing data to establish a trend model and output predictive maintenance alarms to the operation and maintenance system.

[0036] As Fig. 2 shown, among them, the digital baseband processing unit as the internal executor initiates the execution of the full-link self-calibration use case of the core, which includes three sub-use cases of executing digital source self-calibration, measuring internal electronic drift and compensating external environment coupling, and can be extended to apply power-dependent compensation, and the final goal is to provide high-precision positioning signals to the positioning terminal AMR; the test engineer as an external role interacts with the system to perform factory calibration and parameter configuration, and the digital baseband processing unit also executes the link health status monitoring use case, which updates the trend model adaptively and finally outputs predictive maintenance alarms to the operation and maintenance system.Fig. 3 As shown, the system mainly consists of a digital baseband processing unit, a signal switching module, a main working path, a static calibration reference path, and a signal sampling module. In operation, the probe signal generated by the digital baseband processing unit is selectively sent to the main working path or the static calibration reference path through the signal switching module. The main working path contains dynamic elements such as power amplifiers and filters, while the reference path is a high-stability medium microstrip line with constant electrical characteristics. The output signals of the two paths and the reflected signals of the antenna port are captured by the signal sampling module as a directional coupler. The sampled data collected by the signal sampling module are finally fed back to the compensation instruction generation unit of the digital baseband processing unit, thereby forming a complete signal processing and calibration closed loop.

[0037] Embodiment 4: This embodiment aims to provide a standardized pre-delivery calibration procedure for the model for converting the antenna port reflection characteristics into additional phase drift, in order to eliminate the uncertainty in the model construction process. The initial state of the calibration work is defined as follows: place a chamber unit electrical element integrated system to be calibrated, which has completed internal link calibration, in a standard electromagnetic anechoic chamber. The ambient temperature of the anechoic chamber is controlled at 25 ±1 , and the relative humidity is 50% ± 5%. In this environment, a vector network analyzer with a phase measurement accuracy better than 0.05 degrees is connected to the antenna port of the system as an external high-precision phase change observation reference. The first step of the calibration procedure is to establish a zero point reference. Under the condition that there is no reflector within 3 meters in front of the antenna, the system first performs an internal measurement, records the reflection signal group delay frequency response curve at this moment, and stores it as the reference reflection characteristics. At the same time, the transmission signal phase is measured and recorded using the vector network analyzer as the phase zero point. The second step of the calibration procedure is to introduce controllable environmental coupling disturbances. A plurality of standard dielectric plates with different dielectric constants and thicknesses, such as polytetrafluoroethylene plates with thicknesses of 1 mm, 2 mm, and 3 mm, are tightly attached to the outer surface of the antenna cover of the system to simulate different degrees of external environmental changes such as dust accumulation or icing.

[0038] For each introduced standard medium plate, the system performs a complete measurement and recording operation; first, the digital baseband processing unit inside the system captures and analyzes the probe signal reflected back from the antenna port according to the method described in the embodiment, obtains the current real-time reflection characteristics, and obtains a quantitative difference vector by calculating the difference between it and the aforementioned reference reflection characteristics; at the same time, the external vector network analyzer synchronously measures and records the change in the phase of the transmitted signal caused by the introduction of the standard medium plate, which is the true value of the additional phase drift corresponding to this disturbance; by repeating this process, a series of data pairs consisting of difference vectors and their corresponding additional phase drift true values can be obtained; finally, all the obtained data pairs are used to fill a multi-dimensional lookup table, which is the model for converting difference vectors into additional phase drift, where the input address of the lookup table is composed of the components of the difference vector, and the content stored at the address is the corresponding additional phase drift true value; between two calibrated data points, the system uses a linear interpolation algorithm to calculate the output value, and after calibration, the lookup table model is fixed in the non-volatile memory of the digital baseband processing unit, so that the system can only rely on its internal measurement of the antenna port reflection signal to query the table to determine the additional phase drift caused by the external environmental coupling effect in subsequent actual deployment, completing the deterministic construction of the core algorithm model.

[0039] Embodiment 5: This embodiment aims to provide a standardized adaptive adjustment procedure for the time series model used to output predictive maintenance alarm information and the determination of its alarm threshold. In an indoor positioning network that has been deployed and has been running for more than twelve months, one of the indoor unit electrical element integrated systems uses a set of general factory default parameters based on aging data of similar devices for its internal autoregressive moving average time series model at the time of initial deployment, and its predictive maintenance alarm threshold is set to a phase drift rate greater than 0.1 ns / month.

[0040] After the system has been running for six months, it automatically starts a model and threshold adaptive reconstruction process; first, the digital baseband processing unit retrieves the complete time series data of real-time drift error and additional phase drift recorded and stored during the past six months of operation; then, the system uses the historical data specific to this unit and its deployment environment to retrain the internal autoregressive moving average time series model, generating a new set of model parameters that can better fit the individual aging trend of the unit; based on this, the system further uses the new model to backtest the historical data of the past six months, calculates the residual sequence between the model predicted value and the actual measured value, and calculates the standard deviation of the residual sequence ; finally, the system updates the predictive maintenance alarm threshold to the standard deviation of the residual sequence Three times, that is Subsequently, the process of model reconstruction and threshold update is repeated every twelve months. Through this procedure, the predictive model and its alarm threshold used by the system are periodically recalibrated based on their own historical operating data.

[0041] Example 6: This example aims to provide a standardized offline optimization calibration procedure for determining the power level parameters of probe sequences used to construct power-drift mapping models. During the production testing phase of an indoor distributed antenna system (DAS) electrical component integration system, given the individual differences in power-dependent nonlinear phase shift characteristics of the power amplifiers used in its main operating path due to different batches, a preliminary probe sequence power level optimization is required to ensure that the power-dependent compensation value used in each independent system unit accurately matches its hardware characteristics. The calibration is performed on an RF test bench equipped with a vector network analyzer and a precision adjustable attenuator. The bit measurement resolution is better than 0.05 degrees to obtain a reference curve for the power-drift response. At the start of calibration, a system unit to be calibrated is placed on the test bench and operated in a special test mode in which its transmit power can only be controlled by an external test bench. The test bench controls the transmit power of the system unit, scanning from the lowest value of its rated operating range of -10dBm in 1dB steps to the highest value of +23dBm. At each power step point, the vector network analyzer measures and records the reference value of the phase drift generated by the main operating path at that moment, thereby obtaining a high-resolution power-drift mapping reference curve consisting of 34 data points covering the entire operating range.

[0042] Subsequently, the optimization algorithm within the digital baseband processing unit is activated. The objective function of this algorithm is to select a probe sequence consisting of detection signals from at least two different power levels from the aforementioned 34 power levels, for example, three power level points. , and Based on the phase drift values ​​corresponding to these three points, a piecewise linear interpolation model is constructed, ensuring that the maximum absolute error between this interpolation model and the reference curve over the entire working interval is minimized. To minimize; the optimization algorithm, through iterative calculation, ultimately determines the value that minimizes the objective function. three power level points, e.g. -5dBm, +15dBm and +22dBm, which correspond to the initial linear region, the non-linear inflection point and the near-saturation region of the power-drift mapping reference curve, respectively; this set of determined power level parameters is then solidified in the non-volatile memory of the system unit as its dedicated probe sequence parameters, and the system will only use these three power points for measurements in the subsequent online self-calibration procedure to construct its real-time simplified power-drift mapping model, which turns a general compensation method into a parameterized configuration for each hardware individual.

[0043] It is apparent for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, but can be implemented in other concrete forms without departing from the spirit or essential characteristics of the present application.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application, and although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. An AI-optimized room distribution unit electrical element integrated system for 5G and Beidou integrated positioning, characterized in that, The system comprises: a digital baseband processing unit; a main working path connected to the digital baseband processing unit, which contains elements with dynamic changes in electrical characteristics under different working conditions; a static calibration reference path, whose electrical characteristics remain stable under changes in working conditions; a signal switching module configured to selectively send a probe signal output from the digital baseband processing unit into the main working path and the static calibration reference path; a signal sampling module located before an antenna port of the main working path, configured to sample the probe signal propagating forward to the antenna port and sample the probe signal reflected back from the antenna port; the digital baseband processing unit is configured to: compare the electrical characteristics of the probe signal propagating forward through the main working path and the static calibration reference path to determine a real-time drift error caused by the dynamic changes of the elements in the main working path as a first error source; analyze the characteristics of the probe signal reflected back and compare them with a reference reflection characteristic determined in advance under a standard environment to determine an additional phase drift caused by the coupling effect between the antenna port and the external physical environment as a second error source; superimpose the additional phase drift and the real-time drift error to generate a full-link compensation instruction, and apply the full-link compensation instruction to pre-adjust the radio frequency signal when transmitting the radio frequency signal. 2.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein The digital baseband processing unit is further configured to perform a digital source self-calibration before determining the real-time drift error of the main working path, which includes the following rules: loop back the probe signal before sending it to the digital-to-analog converter to form a pure digital loop within the digital baseband processing unit; determine the digital zero-point drift of the digital signal processing link by analyzing the probe signal passing through the pure digital loop; and pre-compensate the signal generation process by applying the digital zero-point drift when generating the probe signal. 3.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein, The digital baseband processing unit is further configured to periodically generate a dual-tone probe signal composed of at least two single-tone signals of different frequencies and send it into the main working path through the signal switching module; perform spectral analysis on the dual-tone probe signal after passing through the main working path via the signal sampling module to detect intermodulation distortion components generated by the nonlinear characteristics of the main working path; determine the nonlinear distortion index of the main working path based on the characteristics of the intermodulation distortion components; and use the nonlinear distortion index as an input parameter for adjusting the generation algorithm of the full-link compensation instruction when generating the full-link compensation instruction. 4.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein, The digital baseband processing unit, in determining the real-time drift error of the main working path, is further configured to: generate and send, via the signal switching module, a probe sequence composed of at least two different power level probe signals into the main working path; measure the drift error of each probe signal in the probe sequence after passing through the main working path respectively; and construct a power-drift mapping model representing the main working path based on the measured drift error related to the power level, the drift mapping model being used to output a power-dependent compensation value according to the instantaneous power of the radio frequency signal The digital baseband processing unit, in determining the real-time drift error of the main working path, is further configured to: generate and send, via the signal switching module, a probe sequence composed of at least two different power level probe signals into the main working path; measure the drift error of each probe signal in the probe sequence after passing through the main working path respectively; and construct a power-drift mapping model representing the main working path based on the measured drift error related to the power level, the drift mapping model being used to output a power-dependent compensation value according to the instantaneous power of the radio frequency signal 5.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein, The digital baseband processing unit is configured to analyze the characteristics of the probe signal reflected back by extracting the group delay frequency response curve of the probe signal reflected back as the real-time reflection characteristic of the coupling effect between the antenna port and the external physical environment; the reference reflection characteristic is a reference group delay frequency response curve calibrated under a preset standard environment; the digital baseband processing unit determines the additional phase drift by calculating the difference vector between the real-time reflection characteristic and the reference reflection characteristic and inputting the difference vector into a model calibrated before leaving the factory to convert the difference vector into the additional phase drift. 6.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein, The digital baseband processing unit is further configured to record time series data of the historical real-time drift error and the additional phase drift, establish a time series model representing the aging trend of the element or the change trend of the external physical environment based on the time series data, calculate a quantitative index representing the change rate of the trend according to the time series model, and output the predictive maintenance alarm information when the quantitative index exceeds a preset alarm threshold. 7.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein, The static calibration reference path is a microstrip line made of high-stability medium and arranged on a circuit board, and the physical length and the dielectric constant of the microstrip line remain constant when the working condition changes. 8.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein, The digital baseband processing unit pre-adjusts the radio frequency signal by using a digital pre-distortion technology to apply a pre-adjustment of phase and delay to the baseband digital signal about to enter the main working path, and the pre-adjustment has the same magnitude and opposite direction as the full-link compensation instruction. 9.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 1, wherein, The probe signal is a wideband probe signal, and the spectral width covers the entire frequency band of the system. 10.The AI-optimized room unit electrical element integrated system for 5G and Beidou integrated positioning according to claim 2, wherein, The pure digital loop includes a first digital loop and a second digital loop which are physically mirror-symmetric; the digital baseband processing unit is further configured to measure the differential delay of the signal when passing through the first digital loop and the second digital loop to determine the non-uniform delay caused by the thermal gradient on the chip; and determine a thermal error compensation value based on the non-uniform delay. And when determining the digital zero-point drift, the thermal error compensation value is subtracted from the analysis result of the probe signal passing through the first digital loop.

Images