A 5G multipath signal super-resolution tracking system

Abstract

The present invention provides a 5G multipath signal super-resolution tracking system, which includes a 5G base station, a 5G signal processing module, a narrow correlation module, a maximum likelihood module, and a multipath tracking module. Without deploying additional base station upgrade equipment, the present invention can track the delay parameters of the positioning opportunity signal broadcast by the 5G base station. This system can obtain delay accuracy within the sampling interval and resolve multipath within the sampling interval, dynamically update the birth and death of some multipaths, and reduce the number of times the 5G signal delay parameters need to be recaptured. This system can meet the needs of both first-path arrival time positioning and multipath-assisted positioning, and has the characteristics of low deployment cost, low computational complexity, and high multipath resolution.

Description

Technical Field

[0001] The present invention relates to the field of 5G indoor positioning technology, and in particular to a 5G multipath signal super-resolution tracking system. Background Art

[0002] From commuting to national security, precise positioning and navigation services are essential. Existing Global Navigation Satellite Systems (GNSS), such as GPS, GLONASS, GALILEO, and BeiDou, provide excellent positioning and navigation solutions in open areas. However, GNSS often fails in harsh environments, such as indoors, in cities, and in canyons, necessitating the use of other positioning methods. Indoor positioning holds broad research potential and is a prerequisite for location-based services (LBS). Research shows that people spend 70%-90% of their time indoors, creating a substantial demand for location-based services.

[0003] As the latest generation of cellular mobile communication technology, the 5th Generation Mobile Communication Technology (5G) offers significant advantages over other indoor positioning solutions in several key areas. 5G boasts excellent device coverage density and signal characteristics in indoor and urban environments. For commercial 5G base stations, 5G downlink physical signals, such as the Primary Synchronization Signal (PSS), Secondary Synchronization Signal (SSS), Channel-State Information Reference Signal (CSI-RS), and Positioning Reference Signal (PRS), are broadcast. These signals serve not only communication functions but also serve as positioning opportunity signals for positioning services. In the upcoming 3GPP 5G Release 17, key positioning technologies will focus on line-of-sight / non-line-of-sight (LOS / NLOS) identification and multipath-assisted positioning.

[0004] The super-resolution algorithm in multipath delay estimation refers to the use of software algorithms to separate the delays of multiple signal arrival paths from a set of signal sampling values ​​under existing hardware conditions. Its delay estimation accuracy is higher than the signal sampling interval, which helps to obtain more accurate delay estimates and assist in positioning solutions in indoor multipath environments.

[0005] The delay estimation obtained based on the correlation method is limited to problems such as the sampling time interval and the multipath effect. The delay estimation accuracy does not exceed the signal sampling interval, and incorrect signal first-reach path estimation will occur in harsh multipath environments. The adaptive delay estimation method and the delay estimation method based on the maximum likelihood criterion can effectively improve the delay estimation accuracy, but their solution usually requires a large number of iterative operations, and the delay estimation calculation complexity is high. In the GNSS field, the multipath signal tracking scheme for GNSS signals has been widely used and has mature theories, but it mainly plays the role of multipath mitigation, enhances the accuracy of first-reach path tracking delay, and regards multipath signals as interference signals. For Ultra Wide-Band (UWB), Ultrawideband (UWB) and other signal systems are specially designed for delay estimation and positioning. Although delay estimation is relatively convenient, it requires additional equipment deployment or occupies communication resources, and is not suitable for large-scale promotion. The base of cellular network equipment is large, and the prospect of large-scale promotion is broad. However, there are relatively few studies on multipath super-resolution tracking of cellular network signals, especially 5G signals. Most studies only focus on tracking the first arrival path of the signal, and rarely consider the multipath birth and death phenomenon in dynamic scenarios. Changes in the number of multipaths may cause the tracking loop to lose lock, and the multipath signal parameters need to be recaptured, which brings greater computational complexity. Summary of the Invention

[0006] In response to the problems existing in the prior art, the purpose of the present invention is to provide a 5G multipath signal super-resolution tracking system, which can track multipath signals based on the positioning opportunity signals broadcast by existing 5G base stations without the need to recapture them every time they are estimated, and has the characteristic of high multipath resolution.

[0007] To achieve the above object, the technical solution adopted by the present invention is:

[0008] A 5G multipath signal super-resolution tracking system includes a 5G base station, a 5G signal processing module, a narrow correlation module, a maximum likelihood module and a multipath tracking module;

[0009] The 5G base station is used to generate 5G signals. Under multipath conditions, the 5G signal model is represented by the superposition of discrete paths:

[0010]

[0011] in, Represents the discrete sampling point index in the time domain, there are a total of signal arrival paths, 、 and Representing the The amplitude, delay and phase parameters of each path, For baseband 5G signal, represent After delay After the sampling signal, is the noise term;

[0012] The 5G signal processing module is used to receive the 5G signal, extract the signal configuration information in the 5G signal, and restore the 5G signal to a frequency domain signal; the frequency domain signal is correlated with the local signal and then output to the narrow correlation module and the maximum likelihood module; the estimated first-path delay is divided into integer multiple delay and fractional multiple delay After compensating for the fractional delay, the output signal of the 5G signal processing module is:

[0013]

[0014] in, Represents the frequency domain signal of DFT after compensating integer multiple delay, is the index of the frequency domain subcarrier, is the number of DFT points;

[0015] The narrow correlation module is used to perform narrow correlation and summation operations on the signals output by the 5G signal processing module, and pass the output to the multipath tracking module. Specifically, a set of equally spaced narrow correlators are constructed using the local signal as observations, that is, multiplying the phase of a set of e exponents. The local signal and the correlator output are:

[0016]

[0017] in, express No. The frequency domain signal of the OFDM symbol after DFT is The calculated index is The output intermediate variable of the narrow correlator, denote the lower and upper bounds of the relevant range, is the correlator spacing, and the number of narrow correlators is Here and , indicating narrow correlation; in fact It's time For units, It represents the time domain interval Continuous correlation function Medium distance discrete sampling points;

[0018] The maximum likelihood module is used to initialize the multipath tracking loop and provide initialization of some paths when the number of paths increases during multipath tracking. Specifically, the maximum likelihood module executes the MEDLL algorithm to obtain multipath parameters, initializes the tracking loop, and outputs the multipath parameter information to the multipath tracking module.

[0019] The multipath tracking module is used to track multipath parameter information and determine the number of paths in the scenario of multipath generation and extinction. The output information is used as the final result of the 5G multipath signal super-resolution tracking solution. At the same time, after the tracking process starts, the information is output to the 5G signal processing module and the maximum likelihood module as auxiliary information. Specifically, the multipath tracking module executes the EKF algorithm to track the multipath parameters and determines the number of paths at the same time. When the number of paths remains unchanged, the tracking ends and the result is output. When the number of paths increases, the tracking has not yet ended. The EKF algorithm tracks the multipath parameters and the number of paths and provides them to the maximum likelihood module. The maximum likelihood module estimates the added paths and increases the order of the EKF matrix. The tracking ends and the result is output. When the number of paths decreases, the tracking has not yet ended. The deleted path parameters are determined by the path number, the order of the EKF matrix is ​​reduced, the tracking ends, and the result is output.

[0020] The processing of the multipath tracking module is specifically as follows:

[0021] (1) Motion model: hypothetical path exist The path amplitude, delay, and delay change parameters at time are ,but All the time The state vector of each path is ; At each tracking interval, the state quantity is predicted using the following linear model and additional noise with normal distribution:

[0022]

[0023] in, 、 and Respectively All the time The path amplitude, delay, and delay change parameters of each path are in vector form. 、 and express 、 and The corresponding normal distribution noise terms are expressed in matrix form:

[0024]

[0025] in represents the process noise matrix, the number of signal path arrivals When the state transition matrix The size is a 3×3 dimensional matrix, using express When , the 3P×3P-dimensional state transition matrix is ​​expressed as the Kronecker product with the P×P-dimensional identity matrix;

[0026] The above (1) is the basis of the extended Kalman filter for the tracking process (3), explaining 、 and The meaning of

[0027] (2) Observation model: The observation model serves as the basis for the extended Kalman filter in the tracking process (3). is given by the output of a set of narrow correlators, which gives a set of discrete values ​​of the actual correlation function, The matrix calculation is as follows:

[0028] The approximate calculation of the ideal correlation function is expressed as:

[0029]

[0030] Among them, for different signal parameter configuration sets, and It is related to the scrambling sequence length and sampling frequency of the signal. is the height of the first peak of the ideal correlation function, The larger the ideal correlation function is, the closer the first zero point is. In a multipath environment, the actual correlation function is:

[0031]

[0032] Indicates the The arrival phase of the paths, the first-order Taylor expansion of a set of narrow correlator observations is: and

[0033]

[0034] Mapping relationship Expressed as:

[0035]

[0036] Among them, the observation matrix The dimension size is ; Corresponding to The time delay represented by the narrow correlator is A narrow correlator;

[0037] The above (2) is the basis of the extended Kalman filter for the tracking process (3), explaining and The meaning of

[0038] (3) The tracking process is calculated according to the following five extended Kalman filter formulas, where the bold variables represent vectors and matrices, the observation matrix H is the Jacobian matrix formed by the first-order Taylor series expansion, The observation vector representing the output of a set of narrow correlators:

[0039]

[0040] The notation rules for symbols are as follows: bold fonts represent vectors or matrices, and the index in brackets represents the filtering time; represents the covariance matrix, It can be a non-zero matrix. represents the Kalman filter gain, represents the observation noise, represents the identity matrix; represents the prior estimated covariance matrix at time t, represents the posterior estimated covariance matrix at time t, represents the prior estimated state vector at time t, represents the posterior estimated state vector at time t, Represents the prior estimated observation vector; each tracking moment t outputs As a result of tracking;

[0041] (4) Dynamically change the number of paths while tracking, and feed the information back to the tracking process (3). The idea behind the change is as follows:

[0042] Calculate the path number discrimination factor for different numbers of paths at time t:

[0043]

[0044] in, is the a priori estimated number of paths at time t, It is an integer, and its absolute value is generally not more than 3, indicating that multiple path number discrimination factors are calculated upward and downward. is the penalty factor for the change in the number of paths, is the penalty factor for too high a number of paths, and the residual signal weight is:

[0045]

[0046] The residual signal is the received signal minus the currently estimated signal:

[0047]

[0048] in, Indicates the prior condition of the variable, such as Represents the estimated value at time t and The estimated value of the real signal corresponding to the number of paths is given by The multipath parameters corresponding to the paths are calculated;

[0049] When the number of paths is higher than the number of paths at the current moment, When , the maximum likelihood module is used to re-estimate from the residual signal The most likely paths and calculate The path number discriminant factor of the path;

[0050] When the number of paths is lower than the number of paths at the current moment, From the last moment to Select multiple times without duplication in the path Calculate the sum of paths , calculate the remaining The residual signal energy of each path ,repeat times, and record them into the residual signal energy array in turn; the one with the smallest residual signal energy The sum of the paths can constitute the signal that should be estimated currently , and calculate The path number discriminant factor of the path;

[0051] After the calculation is completed, an array of path number discriminant factors is obtained; the number of paths when the path number discriminant factor is the smallest is considered to be the posterior estimated path number at the current moment: ;

[0052] Results of (4) The output is given to (3) to determine the dimension of each vector matrix in the extended Kalman filter process.

[0053] The specific processing of the 5G signal processing module after receiving the 5G signal is as follows:

[0054] The 5G signal completes carrier synchronization by searching for the global synchronization channel number and is down-converted to baseband; the synchronization signal block is used to complete symbol timing synchronization and frequency offset estimation; the configuration file of the positioning opportunity signal used for positioning is obtained, providing the necessary information for the subsequent generation of local signals; the 5G signal is restored to the frequency domain signal after removing the cyclic prefix, FFT, serial-to-parallel conversion, and demapping operations, and then correlated with the local signal and output to the narrow correlation module and maximum likelihood module.

[0055] After adopting the above scheme, since the present invention is based on the current 3GPP 5G signal system, it does not require additional base stations and upgraded equipment, and can be adopted; by tracking the positioning opportunity signal delay parameters broadcast by the 5G base station through the scheme of the present invention, due to the adoption of the two technical principles of narrow correlator and extended Kalman filter algorithm, it can obtain the delay accuracy within the sampling interval and resolve the multipath within the sampling interval, and realize data association of the multipath according to the multipath parameter estimation vectors at the previous and next moments, thereby realizing super-resolution of the multipath; based on the tracking module of the present invention, the birth and death of some multipaths are dynamically updated, which can effectively improve the stability of the tracking loop in harsh multipath scenarios, reduce the number of times the 5G signal delay parameters are recaptured, and avoid the high computing time cost brought by recapture. In summary, the present invention can achieve super-resolution tracking of multipath signals, estimate accurate multipath signal delay parameters, and can meet the needs of first-path arrival time positioning and multipath assisted positioning, with the characteristics of low deployment cost, low computational complexity, and high multipath resolution. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a system principle block diagram of the present invention;

[0057] Figure 2 The figure is a flow chart of a method according to a specific embodiment of the present invention. DETAILED DESCRIPTION

[0058] like Figure 1 As shown, the present invention discloses a 5G multipath signal super-resolution tracking system that can meet the needs of first-path arrival time positioning and multipath-assisted positioning. The system includes a 5G base station, a 5G signal processing module, a narrow correlation module, a maximum likelihood module, and a multipath tracking module.

[0059] The 5G base station is used to generate 5G signals. The 5G signal processing module is used to receive the 5G signal, extract the signal configuration information from the 5G signal, and restore the 5G signal into a frequency domain signal. The frequency domain signal is correlated with the local signal and then output to the narrow correlation module and the maximum likelihood module.

[0060] After receiving the 5G signal, the 5G signal processing module performs the following processing: the 5G signal completes carrier synchronization by searching for the global synchronization channel number and is down-converted to baseband; the synchronization signal block is used to complete symbol timing synchronization and frequency offset estimation; the configuration file of the positioning opportunity signal used for positioning is obtained, providing the necessary information for the subsequent generation of the local signal; the 5G signal is restored to the frequency domain signal after removing the cyclic prefix, FFT, serial-to-parallel conversion, and demapping operations, and then correlated with the local signal and output to the narrow correlation module and maximum likelihood module.

[0061] The narrow correlation module is used to perform narrow correlation and summation operations on the signals output by the 5G signal processing module, and pass the output to the multipath tracking module.

[0062] The maximum likelihood module is used to initialize the multipath tracking loop and provide partial path initialization when the number of paths increases during multipath tracking. Specifically, the maximum likelihood module uses the MEDLL algorithm to calculate the multipath parameters, initialize the tracking loop, and output the multipath parameter information to the tracking module.

[0063] The multipath tracking module is used to track multipath parameter information and reasonably determine the number of paths in scenarios where multipath is generated and destroyed. The information it outputs serves as the final result of the 5G multipath signal super-resolution tracking solution. At the same time, after the tracking process begins, the information is output to the 5G signal processing module and the maximum likelihood module as auxiliary information. Specifically, the multipath tracking module executes the EKF algorithm to track multipath parameters and simultaneously determines the number of paths. When the number of paths remains unchanged, the tracking ends and the result is output. When the number of paths increases, the tracking has not yet ended. The EKF algorithm tracks the multipath parameters and the number of paths and provides them to the maximum likelihood module. The maximum likelihood module estimates the added paths and increases the order of the EKF matrix. The tracking ends and the result is output. When the number of paths decreases, the tracking has not yet ended. The parameters of the deleted paths are determined by the number of paths, the order of the EKF matrix is ​​reduced, the tracking ends, and the result is output.

[0064] In order to better understand the above technical solution, the system of the present invention is applied to a multipath scenario, where the number of multipaths in the multipath scenario first remains unchanged, then increases, and finally decreases. Figure 2 As shown, the tracking process in this multipath scenario is as follows:

[0065] S1, 5G signal processing module receives 5G signal sent by 5G base station;

[0066] S2 and 5G signals search for global synchronization channel numbers, complete carrier synchronization, and are down-converted to baseband;

[0067] S3. Use the synchronization signal block to complete symbol timing synchronization and frequency offset estimation;

[0068] S4. Obtain a configuration file of a positioning opportunity signal used for positioning, providing necessary information for subsequent generation of a local signal;

[0069] The S5 and 5G signals are restored to frequency domain signals after cyclic prefix removal, FFT, serial-to-parallel conversion, and demapping operations, and then correlated with the local signal and output to the narrow correlation module;

[0070] S6: The signal is correlated through a set of narrow correlators, summed and output to the multipath tracking module;

[0071] The information output by the signal processing modules in S7 and S5 is simultaneously output to the maximum likelihood module;

[0072] S8, the maximum likelihood module executes the MEDLL algorithm to obtain the multipath parameters, initializes the tracking loop, and outputs the multipath parameter information to the multipath tracking module;

[0073] S9, the multipath tracking module executes the EKF algorithm to track the multipath parameters and determines the number of paths. If the number of paths remains unchanged, the tracking ends, the result is output, and the next tracking is performed.

[0074] S10, the multipath tracking module outputs the timing synchronization information to the 5G signal processing module;

[0075] The S11 and 5G signals are down-converted and synchronized based on the existing carrier frequency information, symbol timing synchronization, and Doppler frequency offset information. After removing the cyclic prefix, performing FFT, serial-to-parallel conversion, and demapping operations, they are restored to frequency domain signals, correlated with the local signal, and output to the narrow correlation module.

[0076] S12, the multipath tracking module executes the EKF algorithm to track the multipath parameters and simultaneously determines the number of paths; at this point, the number of paths increases, and this tracking is not yet complete;

[0077] The information provided by the 5G signal processing module in S13 and S11 is output to the maximum likelihood module;

[0078] S14, the multipath tracking module executes the EKF algorithm to track the multipath parameters and the number of paths and provides them to the maximum likelihood module;

[0079] S15, the maximum likelihood module executes the MEDLL algorithm to estimate the added path from the residual signal;

[0080] S16, increase the order of the EKF matrix; this tracking is completed, the results are output, and the next tracking is carried out;

[0081] S17. The tracking module outputs the timing synchronization information to the 5G signal processing module;

[0082] The S18 and 5G signals are down-converted and synchronized based on the existing carrier frequency information, symbol timing synchronization, and Doppler frequency offset information. After removing the cyclic prefix, performing FFT, serial-to-parallel conversion, and demapping operations, they are restored to frequency domain signals, correlated with the local signal, and output to the narrow correlation module.

[0083] S19: The multipath tracking module executes the EKF algorithm to track multipath parameters and simultaneously determines the number of paths. At this point, the number of paths decreases, but the tracking is not yet complete.

[0084] S20: The parameters of the deleted paths are determined by the number of paths, and the order of the EKF matrix is ​​reduced. The current tracking is completed, the results are output, and the next tracking is carried out.

[0085] This method avoids re-estimating all multipath delay parameter components each time, reducing computational complexity. Furthermore, the super-resolution tracking concept achieves delay accuracy within the sampling interval and resolves multipath components within the sampling interval. The tracking algorithm effectively utilizes prior information about multipath delay parameters while accounting for multipath generation and extinction in dynamic scenarios. It can effectively adapt to switching between line-of-sight and non-line-of-sight scenarios, enhancing the robustness of the tracking loop.

[0086] In the present invention, the 5G signal model sent by the 5G base station is: taking the positioning reference signal mapping of one time slot as an example:

[0087]

[0088] in, Indicates the The PRS scrambling sequence of OFDM symbols, represents the PRS scrambling sequence index, Indicates an index of RE, time domain OFDM symbol index , frequency domain subcarrier index , Represents frequency domain density, offset ; Indicates the number of subcarriers mapped to the frequency domain subcarriers. represents the number of DFT points, After zero padding, the number of subcarriers is equal to , Represents the index of discrete sampling points in the time domain, Represents the cyclic prefix length. Under multipath conditions, the 5G signal model can be expressed as the superposition of discrete paths:

[0089]

[0090] in, Represents the discrete sampling point index in the time domain, there are a total of signal arrival paths, 、 and Representing the The amplitude, delay and phase parameters of each path, For baseband 5G signal, represent After delay After the sampling signal, is the noise term;

[0091] The specific processing flow of the 5G signal processing module is as follows:

[0092] Down-convert to baseband, use the synchronization signal block to complete symbol timing synchronization and frequency offset estimation, obtain the necessary configuration file, generate a local reference signal to complete the estimation of the received signal code phase, and the estimated first-path delay can be divided into integer multiple delay and fractional multiple delay. After compensating for the fractional delay, the output signal of the 5G signal processing module is:

[0093]

[0094] in, Represents the frequency domain signal of DFT after compensating integer multiple delay, is the index of the frequency domain subcarrier, is the number of DFT points;

[0095] The narrow correlation module processes the following: using the local signal to construct a set of equally spaced narrow correlators as observations, that is, multiplying the phase of a set of e exponentials. The local signal and the correlator output are:

[0096]

[0097] in, express No. The frequency domain signal of the OFDM symbol after DFT is The calculated index is The output intermediate variable of the narrow correlator, denote the lower and upper bounds of the relevant range, is the correlator spacing, and the number of narrow correlators is Here and , indicating narrow correlation; in fact It's time For units, It represents the time domain interval Continuous correlation function Medium distance discrete sampling points;

[0098] The process of initializing the tracking loop by the maximum likelihood module is:

[0099] The multipath signal model contains amplitude , delay and phase , so from the perspective of the correlation function, these are the multipath parameters to be solved. The estimation is obtained by minimizing the mean square error according to the MEDLL calculation.

[0100] The processing principle of the multipath tracking module is as follows:

[0101] (1) Motion model: hypothetical path exist The path amplitude, delay, and delay change parameters at time are ,but All the time The state vector of each path is ; At each tracking interval, the state quantity is predicted using the following linear model and additional noise with normal distribution:

[0102]

[0103] in, 、 and Respectively All the time The path amplitude, delay, and delay change parameters of each path are in vector form. 、 and express 、 and The corresponding normal distribution noise terms are expressed in matrix form:

[0104]

[0105] in represents the process noise matrix, the number of signal path arrivals When the state transition matrix The size is a 3×3 dimensional matrix, which can be used express When , the 3P×3P-dimensional state transition matrix is ​​expressed as the Kronecker product with the P×P-dimensional identity matrix;

[0106] (1) As the basis for the extended Kalman filter in the tracking process (3), explain 、 and The meaning of

[0107] (2) Observation model: The observation model serves as the basis for the extended Kalman filter in the tracking process (3). is given by the output of a set of narrow correlators, which gives a set of discrete values ​​of the actual correlation function, The matrix calculation is as follows:

[0108] The approximate calculation of the ideal correlation function can be expressed as:

[0109]

[0110] Among them, for different signal parameter configuration sets, and It is related to the scrambling sequence length and sampling frequency of the signal. is the height of the first peak of the ideal correlation function, The larger the ideal correlation function is, the closer the first zero point is. In a multipath environment, the actual correlation function is:

[0111]

[0112] Indicates the The arrival phase of the paths, the first-order Taylor expansion of a set of narrow correlator observations is: and

[0113]

[0114] Mapping relationship It can be expressed as:

[0115]

[0116] Among them, the observation matrix The dimension size is ; Corresponding to The time delay represented by the narrow correlator is A narrow correlator.

[0117] (2) As the basis for the extended Kalman filter in the tracking process (3), explain and The meaning of

[0118] (3) The tracking process is calculated according to the following five extended Kalman filter formulas, where the bold variables represent vectors and matrices, the observation matrix H is the Jacobian matrix formed by the first-order Taylor series expansion, The observation vector representing the output of a set of narrow correlators:

[0119]

[0120] The notation rules for symbols are as follows: bold fonts represent vectors or matrices, and the index in brackets represents the filtering time; represents the covariance matrix, It can be a non-zero matrix. represents the Kalman filter gain, represents the observation noise, represents the identity matrix; represents the prior estimated covariance matrix at time t, represents the posterior estimated covariance matrix at time t, represents the prior estimated state vector at time t, represents the posterior estimated state vector at time t, Represents the prior estimated observation vector; each tracking moment t outputs As a result of tracking;

[0121] (4) Dynamically change the number of paths while tracking, and feed the information back to the tracking process (3). The idea behind the change is as follows:

[0122] Calculate the path number discrimination factor for different numbers of paths at time t:

[0123]

[0124] in, is the a priori estimated number of paths at time t, It is an integer, and its absolute value is generally not more than 3, indicating that multiple path number discrimination factors are calculated upward and downward. is the penalty factor for the change in the number of paths, is the penalty factor for too high a number of paths, and the residual signal weight is:

[0125]

[0126] The residual signal is the received signal minus the currently estimated signal:

[0127]

[0128] in, Indicates the prior condition of the variable, such as Represents the estimated value at time t and The estimated value of the real signal corresponding to the number of paths is given by The multipath parameters corresponding to the paths are calculated.

[0129] When the number of paths is higher than the number of paths at the current moment, When , the maximum likelihood module is used to re-estimate from the residual signal The most likely paths and calculate The path number discriminant factor of the path;

[0130] When the number of paths is lower than the number of paths at the current moment, From the last moment to Select multiple times without duplication in the path Calculate the sum of paths , calculate the remaining The residual signal energy of each path ,repeat times, and record them into the residual signal energy array in turn; the one with the smallest residual signal energy The sum of the paths can constitute the signal that should be estimated currently , and calculate The path number discriminant factor of the path;

[0131] After the calculation is completed, an array of path number discriminant factors is obtained. The number of paths when the path number discriminant factor is the smallest is considered to be the posterior estimated path number at the current moment:

[0132] .

[0133] Results of (4) The output is given to (3) to determine the dimension of each vector matrix in the extended Kalman filter process.

[0134] In summary, this invention can track the delay parameters of the positioning opportunity signal broadcast by 5G base stations without deploying additional base station upgrade equipment. This can achieve delay accuracy within the sampling interval and resolve multipath within the sampling interval, dynamically update the birth and death of some multipaths, and reduce the number of times the 5G signal delay parameters need to be recaptured. This invention can address the needs of both first-path arrival time positioning and multipath-assisted positioning, and has the characteristics of low deployment cost, low computational complexity, and high multipath resolution.

[0135] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0137] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0138] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0139] It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claim. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, third etc. does not indicate any order. These words may be interpreted as names.

[0140] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

Claims

1. A 5G multipath signal super-resolution tracking system, characterized by: Includes 5G base station, 5G signal processing module, narrow correlation module, maximum likelihood module and multipath tracking module; The 5G base station is used to generate 5G signals. Under multipath conditions, the 5G signal model is represented by the superposition of discrete paths: in, Represents the discrete sampling point index in the time domain, there are a total of signal arrival paths, 、 and Representing the The amplitude, delay and phase parameters of each path, For baseband 5G signal, represent After delay After the sampling signal, is the noise term; The 5G signal processing module is used to receive the 5G signal, extract the signal configuration information in the 5G signal, and restore the 5G signal to a frequency domain signal; the frequency domain signal is correlated with the local signal and then output to the narrow correlation module and the maximum likelihood module; the estimated first-path delay is divided into integer multiple delay and fractional multiple delay After compensating for the fractional delay, the output signal of the 5G signal processing module is: in, Represents the frequency domain signal of DFT after compensating integer multiple delay, is the index of the frequency domain subcarrier, is the number of DFT points; The narrow correlation module is used to perform narrow correlation and summation operations on the signals output by the 5G signal processing module, and pass the output to the multipath tracking module. Specifically, a set of equally spaced narrow correlators are constructed using the local signal as observations, that is, multiplying the phase of a set of e exponents. The local signal and the correlator output are: in, express No. The frequency domain signal of the OFDM symbol after DFT is The calculated index is The output intermediate variable of the narrow correlator, denote the lower and upper bounds of the relevant range, is the correlator spacing, and the number of narrow correlators is Here and , indicating narrow correlation; in fact It's time For units, It represents the time domain interval Continuous correlation function Medium distance discrete sampling points; The maximum likelihood module is used to initialize the multipath tracking loop and provide initialization of some paths when the number of paths increases during multipath tracking. Specifically, the maximum likelihood module executes the MEDLL algorithm to obtain multipath parameters, initializes the tracking loop, and outputs the multipath parameter information to the multipath tracking module. The multipath tracking module is used to track multipath parameter information and determine the number of paths in the scenario of multipath generation and death. The information it outputs serves as the final result of the 5G multipath signal super-resolution tracking solution. At the same time, after the tracking process starts, the information is output to the 5G signal processing module and the maximum likelihood module as auxiliary information. Specifically, the multipath tracking module executes the EKF algorithm to track multipath parameters and determines the number of paths at the same time. When the number of paths remains unchanged, the tracking ends and the result is output. When the number of paths increases, the tracking has not yet ended. The EKF algorithm tracks the multipath parameters and the number of paths and provides them to the maximum likelihood module. The maximum likelihood module estimates the added paths and increases the order of the EKF matrix. The tracking ends and the result is output. When the number of paths decreases, the tracking has not yet ended. The deleted path parameters are determined by the path number, the order of the EKF matrix is ​​reduced, the tracking ends, and the result is output.

2. A 5G multipath signal super-resolution tracking system according to claim 1, characterized in that: The processing of the multipath tracking module is specifically as follows: (1) Motion model: hypothetical path exist The path amplitude, delay, and delay change parameters at time are ,but All the time The state vector of each path is ; At each tracking interval, the state quantity is predicted using the following linear model and additional noise with normal distribution: in, 、 and Respectively All the time The path amplitude, delay, and delay change parameters of each path are in vector form. 、 and express 、 and The corresponding normal distribution noise terms are expressed in matrix form: in represents the process noise matrix, the number of signal path arrivals When the state transition matrix The size is a 3×3 dimensional matrix, using express When , the 3P×3P-dimensional state transition matrix is ​​expressed as the Kronecker product with the P×P-dimensional identity matrix; The above (1) is the basis of the extended Kalman filter for the tracking process (3), explaining 、 and The meaning of (2) Observation model: The observation model serves as the basis for the extended Kalman filter in the tracking process (3). is given by the output of a set of narrow correlators, which gives a set of discrete values ​​of the actual correlation function, The matrix calculation is as follows: The approximate calculation of the ideal correlation function is expressed as: Among them, for different signal parameter configuration sets, and It is related to the scrambling sequence length and sampling frequency of the signal. is the height of the first peak of the ideal correlation function, The larger the ideal correlation function is, the closer the first zero point is. In a multipath environment, the actual correlation function is: Indicates the The arrival phase of the paths, the first-order Taylor expansion of a set of narrow correlator observations is: Mapping relationship Expressed as: Among them, the observation matrix The dimension size is ; Corresponding to The time delay represented by the narrow correlator is A narrow correlator; The above (2) is the basis of the extended Kalman filter for the tracking process (3), explaining and The meaning of (3) The tracking process is calculated according to the following five extended Kalman filter formulas, where the bold variables represent vectors and matrices, the observation matrix H is the Jacobian matrix formed by the first-order Taylor series expansion, The observation vector representing the output of a set of narrow correlators: The notation rules for symbols are as follows: bold fonts represent vectors or matrices, and the index in brackets represents the filtering time; represents the covariance matrix, It can be a non-zero matrix. represents the Kalman filter gain, represents the observation noise, represents the identity matrix; represents the prior estimated covariance matrix at time t, represents the posterior estimated covariance matrix at time t, represents the prior estimated state vector at time t, represents the posterior estimated state vector at time t, Represents the prior estimated observation vector; each tracking moment t outputs As a result of tracking; (4) Dynamically change the number of paths while tracking, and feed the information back to the tracking process (3). The idea behind the change is as follows: Calculate the path number discrimination factor for different numbers of paths at time t: in, is the a priori estimated number of paths at time t, It is an integer whose absolute value does not exceed 3, indicating that multiple path number discrimination factors are calculated upward and downward. is the penalty factor for the change in the number of paths, is the penalty factor for too high a number of paths, and the residual signal weight is: The residual signal is the received signal minus the currently estimated signal: in, represents the prior condition of the variable, Represents the estimated value at time t and The estimated value of the real signal corresponding to the number of paths is given by The multipath parameters corresponding to the paths are calculated; When the number of paths is higher than the number of paths at the current moment, When , the maximum likelihood module is used to re-estimate from the residual signal The most likely paths and calculate The path number discriminant factor of the path; When the number of paths is lower than the number of paths at the current moment, From the last moment to Select multiple times without duplication in the path Calculate the sum of paths , calculate the remaining The residual signal energy of each path ,repeat times, and record them into the residual signal energy array in turn; the one with the smallest residual signal energy The sum of the paths can constitute the signal that should be estimated currently , and calculate The path number discriminant factor of the path; After the calculation is completed, an array of path number discriminant factors is obtained; the number of paths when the path number discriminant factor is the smallest is considered to be the posterior estimated path number at the current moment: ; The result of (4) above The output is given to (3) to determine the dimension of each vector matrix in the extended Kalman filter process.

3. The 5G multipath signal super-resolution tracking system according to claim 1, characterized in that: The specific processing of the 5G signal processing module after receiving the 5G signal is as follows: The 5G signal completes carrier synchronization by searching for the global synchronization channel number and is down-converted to baseband; the synchronization signal block is used to complete symbol timing synchronization and frequency offset estimation; the configuration file of the positioning opportunity signal used for positioning is obtained, providing the necessary information for the subsequent generation of local signals; the 5G signal is restored to the frequency domain signal after removing the cyclic prefix, FFT, serial-to-parallel conversion, and demapping operations, and then correlated with the local signal and output to the narrow correlation module and maximum likelihood module.

Images