Space-time two-dimensional adaptive anti-interference system of GNSS receiver

Abstract

The invention discloses a space-time two-dimensional adaptive anti-interference system of a GNSS receiver. The space-time two-dimensional adaptive anti-interference system comprises a space-time conversion module, a digital oscillator, a direct digital frequency synthesizer, a low-pass filter, a sampling extraction module and an adaptive algorithm module. According to the invention, a plurality of time domain registers are connected behind each array element of a receiver, the plurality of array elements can realize spatial domain filtering, and the plurality of time domain registers and taps connected to a single array element can realize time domain filtering, that is, a multi-channel signal is converted into a space-time signal through a space-time conversion module; the space-time two-dimensional adaptive anti-interference system not only can utilize the signal difference in the spatial dimension, but also can process the signal change in the time dimension. Orthogonal down-conversion, low-pass filtering and extraction sampling are carried out on the input signals, and preprocessing of the signals is completed; and then the power inversion algorithm of the adaptive algorithm module is used, and the adaptive weight of the receiver antenna array is adjusted, so that the output power in the interference direction is minimized, and null is formed.

Description

Technical Field

[0001] The invention relates to the field of satellite navigation technology, in particular to a space-time two-dimensional adaptive anti-interference system of a GNSS receiver. Background Art

[0002] The Global Satellite Navigation System (GNSS) is an air-based radio navigation and positioning system that can provide users with all-weather three-dimensional coordinates, speed and time information at any location on the Earth's surface or in near-Earth space.

[0003] A GNSS receiver (Global Navigation Satellite System Receiver) is an electronic device used to receive, decode, process and utilize radio signals transmitted by the Global Navigation Satellite System (GNSS). Its core function is to accurately determine and output information such as geographic location, speed, and time by analyzing received satellite signals. A compatible GNSS receiver can simultaneously receive and process satellite signals from different systems (GPS, GLONASS, GALILEO, BDS) to achieve more stable, reliable and accurate positioning services.

[0004] In the information age, interference and anti-interference technology are a pair of mutually opposing technologies, and the two are almost undergoing technological iterations simultaneously. The greater the intensity of the interference signal, the more we need to continuously improve the anti-interference technology to deal with it. Since GNSS signals will experience attenuation when they pass through the atmosphere and reach the ground GNSS receiver after being sent from the satellite, this means that their power level at the receiving end is very low, which makes them particularly susceptible to interference. At present, interference to GNSS satellite signals is divided into artificial interference and potential interference; potential interference mainly includes: in-band RF interference, out-of-band RF interference, and environmental interference; artificial interference includes: suppression interference and deceptive interference. To completely solve this problem, in theory, it can be achieved by increasing the transmission signal power of the navigation satellite, but this requires huge costs, and it is impossible to replace all navigation satellites at one time. Therefore, in order to ensure that the receiver can efficiently and correctly receive GNSS signals, the currently commonly used anti-interference methods are mainly based on the anti-interference technology of the receiver. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a space-time two-dimensional adaptive anti-interference system for a GNSS receiver, which introduces a space-time two-dimensional structure based on spatial domain filtering, and enhances the anti-interference capability of the GNSS receiver by applying a power inversion algorithm to the space-time two-dimensional structure.

[0006] The technical solution of the present invention is:

[0007] A space-time two-dimensional adaptive anti-interference system for a GNSS receiver includes a space-time conversion module, a digital oscillator, a direct digital frequency synthesizer, a low-pass filter, a sampling and extraction module, and an adaptive algorithm module; the space-time conversion module is to add a time domain register after each antenna array element of the GNSS receiver, the antenna array element samples the analog radio frequency signal in space, the analog-to-digital converter performs analog-to-digital conversion on the analog radio frequency signal collected by the antenna array element, and the time domain register stores these signals in sequence, and the antenna array element and the time domain register are sampled at the same sampling frequency, so as to combine the processing of the time domain and the space domain. In summary, the digital oscillator is used to generate sine waves and cosine waves as carriers for coherent demodulation, and then send them to the direct digital frequency synthesizer. The direct digital frequency synthesizer mixes the coherently demodulated carrier with the intermediate frequency signal after analog-to-digital conversion, and then filters out the high-frequency components in the mixed signal through a low-pass filter to obtain a baseband signal. The sampling and extraction module reduces the sampling rate by extracting samples from the baseband signal. The adaptive algorithm module applies a power inversion algorithm to adjust the optimal weights of the antenna array to minimize the output power in the interference direction, forming a zero sink, thereby increasing the degree of freedom and enhancing the anti-interference capability.

[0008] The specific steps of the power inversion algorithm are:

[0009] The output Y of the GNSS receiver is assumed to be the following formula (1):

[0010] Y=W H X(1);

[0011] In formula (1), W = [w1, ..., w M ] T is the adaptive weight of the antenna array, M represents the number of GNSS receiver antenna arrays, H is the conjugate transpose operation, X = [x1(t), …, x M (t)] T is the signal received by the GNSS receiver, T is the matrix transpose;

[0012] Set W1 = 1 and minimize the power of output y(n). Then, according to the following formula (2), there exists a unique set of weights that satisfies the output power P when w1 = 1. out Minimum, and can produce a zero point in the direction of the interference signal,

[0013]

[0014] In formula (2), f(θ M-1 ) is a function expression, -j represents the imaginary unit, θ M-1 is the direction angle of the interference signal, is the phase of the interfering signal;

[0015] A set of optimal weights is obtained to satisfy equation (2). When the adaptive weights W = [w1, ..., wM] T Not zero and the output power P out =|y(n)| 2 In the smallest case, a zero point will be generated in the direction of the interference signal;

[0016] Define the first array element as the reference array element, the adaptive weight is 1, set d = x1 as the reference signal, and the output Y obtained by multiplying the input signals of the remaining M-1 array elements with the corresponding adaptive weights M-1 It is the following formula (3):

[0017]

[0018] In formula (3), W M-1 =[w2, w3, ..., w M ] T , X M-1 =[x2, x3, ..., x M ] T ;

[0019] The calculation of the mean square error ξ of the M array element outputs is shown in the following formula (4):

[0020]

[0021] In formula (4), represents the autocorrelation matrix of the M-1 array element input signals other than the reference signal x1; represents the cross-correlation matrix between the first array element and the input signals of M-1 array elements;

[0022] Then ξ is W M-1 Find the gradient and set it to zero, as shown in the following formula (5):

[0023]

[0024] From formula (5), we can get Optimal weight W opt2 It is expressed as the following formula (6):

[0025]

[0026] Formula (6) is the optimal weight W opt2 The first element of a matrix is ​​1, and the other elements are Since the goal of the least squares method is to minimize the residual sum of squares, the adaptive weight W is inversely proportional to the autocorrelation matrix, and the other elements all have negative signs to ensure that the final optimal weight minimizes the residual sum of squares.

[0027] The specific steps of applying the power inversion algorithm to adjust the weight of the antenna array are:

[0028] It is defined that N far-field narrowband signals are transmitted to a GNSS receiver with M array elements. The signal received by the GNSS receiver is shown in the following equation (7):

[0029]

[0030] In formula (7), u i (t) represents the complex envelope of the received signal; φ(t) represents the phase of the received signal; w0 is the carrier frequency of the signal; s i (t) represents the i-th narrowband signal; j represents the imaginary unit; τ represents the delay, which means the time difference of the signal reaching different array elements;

[0031] Due to the relative characteristics of the time domain and the frequency domain, a narrowband signal means that the time domain changes slowly, so it satisfies the following equation (8):

[0032]

[0033] Substituting formula (8) into formula (7), we get: Therefore, the signal received by the kth array element is expressed as follows (9):

[0034]

[0035] In formula (9), g ki is the gain of the kth array element for the ith narrowband signal, n k (t) is the noise of the kth array element at time t, τ ki is the time delay of the i-th narrowband signal reaching the k-th array element relative to the reference array element;

[0036] Therefore, the vector set of signals received by M array elements at time t is expressed as follows (10):

[0037]

[0038] Assuming that the direction of each array element is consistent and there is no mutual coupling between the array elements, and each channel is consistent, equation (10) is simplified to the following equation (11):

[0039]

[0040] Formula (11) can be expressed as a matrix:

[0041] The steering matrix in formula (11) It is only related to the direction of the signal and the arrangement of the array elements;

[0042] The signal received by the array element, i.e. the input signal, is expressed as the following formula (12):

[0043] X(t)=X S (t)+X n (t) (12);

[0044] In formula (12), X S (t) is the signal matrix received by the antenna array, i.e., M array elements, X n (t) is the noise matrix received by the antenna array. According to the input signal of equation (12) and equation (1), the output Y(t) of the GNSS receiver at time t is obtained as follows (13):

[0045] Y(t)=W H X(t)=W H X s (t)+W H X n (t) (13);

[0046] It is assumed that each array element is connected to N time domain registers, that is, the number of taps of each array element is N, and the delay interval is T, where T≤1 / B, B is the processing bandwidth inside the GNSS receiver, and the signal x of the mth tap on the i-th array element is i,m It is expressed as the following formula (14):

[0047] x i,m =x i,1 (t-(m-1)T)=x 1,1 (t-τ i -(m-1)T) (14);

[0048] In formula (14), x 1，1 is the signal received by the reference element, τ i is the time delay between the i-th array and the set reference array element, and the output Y is expressed as the following formula (15):

[0049]

[0050] In formula (15), The vector representing the signal received by the array element, i.e. the input signal;

[0051] The power of equation (15) is obtained as follows (16):

[0052] P out =E{|Y| 2}=E{(W H X)(W H X) H =W H R xx W (16);

[0053] In formula (16), P out Represents output power, which measures the average energy of the processed signal; R XX =E{XX H} is the autocorrelation matrix of the space-time array input; E represents the expected value, which is used to describe the average value of the random variable;

[0054] The following formula (17) is the optimal weight W when the adaptive weight of the first array element is constrained to 1. opt Solution:

[0055]

[0056] Formula (17) is an optimization problem, the goal is to minimize the output power P out , while satisfying the constraint that the adaptive weight of the first array element is 1; argmin represents the minimum output power P out , subject to means optimization under given constraints, that is, when the constraint of the adaptive weight of the first array element is 1; a means the constraint vector of MN×1 dimension, and the steering vector of the input signal in the definition space is a s , the time direction vector is a t =[1,e -jwT ,…,e -j(N-1)T ], so the space-time two-dimensional direction vector is represents the Kronecker product;

[0057] Finally, W is obtained by the following formula (18): opt Minimum value of:

[0058]

[0059] The digital oscillator, direct digital frequency synthesizer, low-pass filter, sampling and extraction module and adaptive algorithm module are all arranged in a field programmable gate array FPGA. The intermediate frequency signal received by the FPGA is a digital signal in the form of a floating point number, which needs to be first converted into a digital signal in the form of a fixed point number and then mixed. The conversion formula is shown in the following formula (19):

[0060] fixed=x×2 Q (19);

[0061] In formula (19), x is a digital signal in floating point form, fixed is a digital signal in fixed point form, and Q represents Q format.

[0062] The analog-to-digital converter is a 16-bit analog-to-digital converter with a bit width of 12 bits. The sampling rate of the analog-to-digital converter is f sSatisfies the following formula (20):

[0063]

[0064] In formula (20), n is a set positive integer, and f0 is the frequency of the intermediate frequency signal.

[0065] The adaptive algorithm module updates the optimal weights in an iterative manner. First, the optimal weights are obtained according to the power inversion algorithm. -W M-1 is the optimal weight of the input signal of M-1 array elements;

[0066] The weight iteration formula is as follows (21):

[0067]

[0068] The operation steps of the adaptive algorithm module are shown in the following formula (22):

[0069]

[0070] In formula (22), Y(t) represents the output signal at the current time t; e(t) represents the error signal at the current time t; W(t+1) represents the weight vector at the next time; Indicates the weight update amount.

[0071] Advantages of the present invention:

[0072] The present invention connects multiple time domain registers after each array element of the receiver, and the multiple array elements can realize spatial domain filtering. The multiple time domain registers and taps connected to a single array element realize time domain filtering, that is, through the space-time conversion module, the multi-channel signal is converted into a space-time signal, so that the space-time two-dimensional adaptive anti-interference system can not only utilize the signal difference in the spatial dimension, but also process the signal change in the time dimension; then the input signal is orthogonally down-converted, low-pass filtered and sampled to complete the signal preprocessing; and then the power inversion algorithm is used to adjust the adaptive weights of the GNSS receiver antenna array to minimize the output power in the interference direction and form a null.

[0073] The present invention introduces a space-time two-dimensional structure based on spatial domain filtering, enhances the anti-interference capability of the GNSS receiver by applying a power inversion algorithm to the space-time two-dimensional structure, and reduces the complexity of calculation by iteratively approximating the optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is a principle block diagram of the space-time two-dimensional adaptive anti-interference system of the present invention. DETAILED DESCRIPTION

[0075] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0076] See Figure 1 A space-time two-dimensional adaptive anti-interference system for a GNSS receiver includes a space-time conversion module 1, a digital oscillator 2, a direct digital frequency synthesizer (DDS module) 3, a low-pass filter 4, a sampling and extraction module 5, and an adaptive algorithm module 6; the space-time conversion module 1 is to add a time domain register after each antenna array element of the GNSS receiver, the antenna array element samples the analog radio frequency signal in space, the analog-to-digital converter performs analog-to-digital conversion on the analog radio frequency signal collected by the antenna array element, and the time domain register stores these signals in sequence, and the antenna array element and the time domain register are sampled at the same sampling frequency, so as to convert The processing in the time domain and the space domain are combined. The digital oscillator 2 is used to generate sine waves and cosine waves as carriers for coherent demodulation, and then sent to the DDS module 3. The DDS module 3 mixes the coherently demodulated carrier with the intermediate frequency signal after analog-to-digital conversion, and then filters out the high-frequency components in the mixed signal through the low-pass filter 4 to obtain the baseband signal. The sampling and extraction module 5 reduces the sampling rate by extracting samples from the baseband signal. The adaptive algorithm module 6 applies the power inversion algorithm to adjust the optimal weight of the antenna array to minimize the output power in the interference direction, form a null, improve the degree of freedom, and enhance the anti-interference ability.

[0077] Since GNSS signals are sent by satellites that are constantly moving in the sky, it is impossible to know the specific direction of the GNSS signals. Therefore, based on the fact that the power of satellite signals is about 20dB lower than thermal noise, it can be defined that signals greater than thermal noise are intentional or unintentional interference to satellite signals; using this feature, when the power is not zero, the lowest power must form a null in the direction of the interference signal. Since the number of nulls is related to the degrees of freedom of the system. When the degrees of freedom are limited, the output power needs to be minimum and the output is not zero, then the null in the direction of the antenna array will preferentially align with the direction of the signal with greater interference power. Since the power of satellite signals is lower than thermal noise, this method will not affect satellite signals anyway. Due to its characteristics, this method is very suitable for suppressing interference in weak signals, because it does not require advance knowledge of the signal's direction information and reference signal, and can adaptively perform spatial domain filtering on signals with greater power than thermal noise.

[0078] S11. The specific steps of the power inversion algorithm are:

[0079] The output Y of the GNSS receiver is assumed to be the following formula (1):

[0080] Y=W H X (1);

[0081] In formula (1), W = [w1, ..., w M ] T is the adaptive weight of the antenna array, M represents the number of GNSS receiver antenna arrays, H is the conjugate transpose operation, X = [x1(t), …, x M (t)] T is the signal received by the GNSS receiver, T is the matrix transpose;

[0082] Set w1 = 1 and minimize the power of output y(n). Then, according to the following formula (2), there exists a unique set of weights that satisfies W1 = 1, so that the output power P out Minimum, and can produce a zero point in the direction of the interference signal,

[0083]

[0084] In formula (2), f(θ M-1 ) is a function expression, -j represents the imaginary unit, θ M-1 is the direction angle of the interference signal, is the phase of the interfering signal;

[0085] A set of optimal weights is obtained to satisfy equation (2). When the adaptive weights W = [w1, ..., w M ] T Not zero and the output power P out =|y(n)| 2 In the smallest case, a zero point will be generated in the direction of the interference signal;

[0086] Define the first array element as the reference array element, the adaptive weight is 1, set d = x1 as the reference signal, and the output Y obtained by multiplying the input signals of the remaining M-1 array elements with the corresponding adaptive weights M-1 It is the following formula (3):

[0087]

[0088] In formula (3), W M-1 =[w2, w3, ..., w M ] T , X M-1 =[x2, x3, ..., x M ] T ;

[0089] The calculation of the mean square error ξ of the M array element outputs is shown in the following formula (4):

[0090]

[0091] In formula (4), represents the autocorrelation matrix of the M-1 array element input signals other than the reference signal x1; represents the cross-correlation matrix between the first array element and the input signals of M-1 array elements;

[0092] Then ξ is W M-1 Find the gradient and set it to zero, as shown in the following formula (5):

[0093]

[0094] From formula (5), we can get Optimal weight W opt2 It is expressed as the following formula (6):

[0095]

[0096] Formula (6) is the optimal weight W opt2 The first element of a matrix is ​​1, and the other elements are Since the goal of the least squares method is to minimize the residual sum of squares, the adaptive weight W is inversely proportional to the autocorrelation matrix, and the other elements are all negative, ensuring that the final optimal weight minimizes the residual sum of squares;

[0097] S12, spatial domain filtering is to compensate the phase by weights after the antenna array element according to the phase difference caused by the different time when the signal arrives at different array elements, so as to control the constructive interference or destructive interference of the signal in the same direction;

[0098] It is defined that N far-field narrowband signals are transmitted to a GNSS receiver with M array elements. The signal received by the GNSS receiver is shown in the following equation (7):

[0099]

[0100] In formula (7), u i (t) represents the complex envelope of the received signal; φ(t) represents the phase of the received signal; w0 is the carrier frequency of the signal; s i (t) represents the i-th narrowband signal; j represents the imaginary unit; τ represents the delay, which means the time difference of the signal reaching different array elements;

[0101] Due to the relative characteristics of the time domain and the frequency domain, a narrowband signal means that the time domain changes slowly, so it satisfies the following equation (8):

[0102]

[0103] Substituting formula (8) into formula (7), we get: Therefore, the signal received by the kth array element is expressed as follows (9):

[0104]

[0105] In formula (9), g ki is the gain of the kth array element for the ith narrowband signal, n k (t) is the noise of the kth array element at time t, τ ki is the time delay of the i-th narrowband signal reaching the k-th array element relative to the reference array element;

[0106] Therefore, the vector set of signals received by M array elements at time t is expressed as follows (10):

[0107]

[0108] Assuming that the direction of each array element is consistent and there is no mutual coupling between the array elements, and each channel is consistent, equation (10) is simplified to the following equation (11):

[0109]

[0110] Formula (11) can be expressed as a matrix:

[0111] The steering matrix in formula (11) It is only related to the direction of the signal and the arrangement of the array elements;

[0112] The signal received by the array element, i.e. the input signal, is expressed as the following formula (12):

[0113] X(t)=X S (t)+X n (t) (12);

[0114] In formula (12), X S (t) is the signal matrix received by the antenna array, i.e., M array elements, X n (t) is the noise matrix received by the antenna array. According to the input signal of equation (12) and equation (1), the output Y(t) of the GNSS receiver at time t is obtained as follows (13):

[0115] Y(t)=W H X(t)=W H X s (t)+W H X n (t) (13);

[0116] S13, spatial domain filtering is to use the taps for phase compensation after the array element to adjust the phase of the received signal, so as to constructively interfere with the direction of the satellite signal, thereby forming an enhanced main lobe in the direction of the navigation signal, and destructively interfere with the interference signal, forming a null in the direction of the interference signal, thereby achieving optimization and interference of the directional pattern;

[0117] It is assumed that each array element is connected to N time domain registers, that is, the number of taps of each array element is N, and the delay interval is T, where T≤1 / B, B is the processing bandwidth inside the GNSS receiver, and the signal x of the mth tap on the i-th array element is i,m It is expressed as the following formula (14):

[0118] x i,m =x i,1 (t-(m-1)T)=x 1,1 (t-τ i -(m-1)T) (14);

[0119] In formula (14), x 1，1 is the signal received by the reference element, τ i is the time delay between the i-th array and the set reference array element, and the output Y is expressed as the following formula (15):

[0120]

[0121] In formula (15), The vector representing the signal received by the array element, i.e. the input signal;

[0122] The power of equation (15) is obtained as follows (16):

[0123] P out =E{|Y| 2}=E{(W H X)(W H X) H =W H R xx W (16);

[0124] In formula (16), P out Represents output power, which measures the average energy of the processed signal; R XX =E(XX H} is the autocorrelation matrix of the space-time array input; E represents the expected value, which is used to describe the average value of the random variable;

[0125] The following formula (17) is the optimal weight W when the adaptive weight of the first array element is constrained to 1. opt Solution:

[0126]

[0127] Formula (17) is an optimization problem, the goal is to minimize the output power Po u t, and the constraint of the adaptive weight of the first array element is 1; argmin represents the minimum output power Po u t, subject to means optimization under given constraints, that is, when the constraint of the adaptive weight of the first array element is 1; a means the constraint vector of MN×1 dimension, and the steering vector of the input signal in the definition space is a s , the time direction vector is a t =[1,e -jw ,…,e -jw(N-1)T ], so the space-time two-dimensional direction vector is represents the Kronecker product;

[0128] Finally, W is obtained by the following formula (18): opt Minimum value of:

[0129]

[0130] S14, the adaptive algorithm module updates the optimal weights in an iterative manner. First, the optimal weights are obtained according to the power inversion algorithm. -W M-1 is the optimal weight of the input signal of M-1 array elements;

[0131] The weight iteration formula is as follows (21):

[0132]

[0133] The operation steps of the adaptive algorithm module are shown in the following formula (22):

[0134]

[0135] In formula (22), Y(t) represents the output signal at the current time t; e(t) represents the error signal at the current time t; W(t+1) represents the weight vector at the next time; Indicates the weight update amount.

[0136] Among them, the digital oscillator, direct digital frequency synthesizer, low-pass filter, sampling and extraction module and adaptive algorithm module are all set in the field programmable gate array FPGA;

[0137] Since binary fixed-point numbers do not have a clear decimal point representation, a decimal point symbol cannot be used inside the FPGA to distinguish the integer part from the decimal part like floating-point numbers; however, decimal data is inevitable in many applications. To solve this problem, the Q format representation is used to divide the total bit width into three fixed parts, namely the sign bit, the integer part, and the decimal part; when the bit width of one part increases, the bit width of the other part will inevitably decrease. Therefore, if you need to improve the accuracy, you need to increase the bit width of the decimal part and reduce the bit width of the integer part; if you need to represent a larger numerical range, increase the bit width of the integer part and reduce the bit width of the decimal part accordingly.

[0138] When using the Q0 format, all bits are used to represent integers, ranging from -32768 to 32767. For applications that require higher precision, the Q15 format can be used, in which the decimal part has the largest bit width and the value range is -1 to 0.999695;

[0139] That is, the intermediate frequency signal received by the FPGA is a digital signal in floating point form, which needs to be converted into a digital signal in fixed point form first, and then mixed. The conversion formula is shown in the following formula (19):

[0140] fixed=x × 2 Q (19);

[0141] In formula (19), x is a digital signal in floating point form, fixed is a digital signal in fixed point form, and Q represents Q format.

[0142] The signal received by the antenna array is an analog signal, which needs to be converted into a digital signal by an analog-to-digital converter (ADC) before it can be input into the FPGA for processing and analysis. Generally, the power of the navigation signal when it reaches the ground receiver is about -130dBm, while the power of the common interference signal is -60dBm. Therefore, the dynamic range of the ADC needs to be at least 70dB to ensure that the navigation signal can be distinguished. According to the formula It can be concluded that a dynamic range of 70dB can be achieved when the bit width is 12bit, so the analog-to-digital converter uses a 16-bit analog-to-digital converter with a bit width of 12bit;

[0143] Since sampling in the time domain corresponds to shifting in the frequency domain, the sampling rate needs to be set so that the spectrum does not overlap after the shift. The sampling rate of the analog-to-digital converter is f s Satisfies the following formula (20):

[0144]

[0145] In formula (20), n is a set positive integer, and f0 is the frequency of the intermediate frequency signal.

[0146] The present invention stores the sine wave and cosine wave generated by the digital oscillator 2 in the ROM IP core of the FPGA, samples the sine wave and cosine wave using Matlab and quantizes them into 16-bit signed numbers, then writes a script in Matlab to generate the .coe file required by Xilinx, and finally loads the storage files of the two orthogonal signals through the IP core of the ROM.

[0147] The time domain register of the present invention adopts 4 time domain levels, so the data input to the space-time two-dimensional adaptive anti-interference system is cached in four levels. A time domain register with the same bit width as the data is set after each array element receiving signal port, and the data is similar to a shift register through the sampling clock. When each clock cycle arrives, the data is delayed by one beat.

[0148] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A space-time two-dimensional adaptive anti-interference system for a GNSS receiver, characterized in that: It includes a space-time conversion module, a digital oscillator, a direct digital frequency synthesizer, a low-pass filter, a sampling and extraction module, and an adaptive algorithm module; the space-time conversion module is to add a time domain register after each antenna array element of the GNSS receiver, the antenna array element samples the analog RF signal in space, and the analog-to-digital converter performs analog-to-digital conversion on the analog RF signal collected by the antenna array element. The time domain register stores these signals in sequence, and the antenna array element and the time domain register are sampled at the same sampling frequency, thereby combining the time domain and space domain processing. The digital oscillator is used to produce The generated sine wave and cosine wave are used as the carrier of coherent demodulation, and then sent to the direct digital frequency synthesizer. The direct digital frequency synthesizer mixes the coherently demodulated carrier with the intermediate frequency signal after analog-to-digital conversion, and then filters out the high-frequency components in the mixed signal through a low-pass filter to obtain a baseband signal. The sampling and extraction module reduces the sampling rate by extracting samples from the baseband signal. The adaptive algorithm module applies the power inversion algorithm to adjust the optimal weight of the antenna array to minimize the output power in the interference direction, form a zero sink, increase the degree of freedom, and enhance the anti-interference ability.

2. The space-time two-dimensional adaptive anti-interference system for a GNSS receiver according to claim 1, characterized in that: The specific steps of the power inversion algorithm are: The output Y of the GNSS receiver is assumed to be the following formula (1): Y=W H X (1); In formula (1), W = [w1,…,w M ] T is the adaptive weight of the antenna array, M represents the number of GNSS receiver antenna arrays, H is the conjugate transpose operation, X = [x1(t),…,x M (t)] T is the signal received by the GNSS receiver, T is the matrix transpose; Set w1 = 1 and minimize the power of output y(n). Then, according to the following formula (2), there exists a unique set of weights that satisfies the output power P when w1 = 1. out Minimum, and can produce a zero point in the direction of the interference signal, In formula (2), f(θ M-1 ) is a function expression, -j represents the imaginary unit, θ M-1 is the direction angle of the interference signal, is the phase of the interfering signal; A set of optimal weights is obtained to satisfy equation (2). When the adaptive weights W = [w1,…,w M ] T Not zero and the output power P out =|y(n)| 2 In the smallest case, a zero point will be generated in the direction of the interference signal; Define the first array element as the reference array element, the adaptive weight is 1, set d = x1 as the reference signal, and the output Y obtained by multiplying the input signals of the remaining M-1 array elements with the corresponding adaptive weights M-1 It is the following formula (3): In formula (3), W M-1 =[w2,w3,…,w M ] T , X M-1 =[x2,x3,…,x M ] T ; The calculation of the mean square error ξ of the M array element outputs is shown in the following formula (4): In formula (4), represents the autocorrelation matrix of the M-1 array element input signals other than the reference signal x1; represents the cross-correlation matrix between the first array element and the input signals of M-1 array elements; Then ξ is W M-1 Find the gradient and set it to zero, as shown in the following formula (5): From formula (5), we can get Optimal weight W opt2 It is expressed as the following formula (6): Formula (6) is the optimal weight W opt2 The first element of a matrix is ​​1, and the other elements are Since the goal of the least squares method is to minimize the residual sum of squares, the adaptive weight W is inversely proportional to the autocorrelation matrix, and the other elements all have negative signs to ensure that the final optimal weight minimizes the residual sum of squares.

3. The space-time two-dimensional adaptive anti-interference system for a GNSS receiver according to claim 2, characterized in that: The specific steps of applying the power inversion algorithm to adjust the weight of the antenna array are: It is defined that N far-field narrowband signals are transmitted to a GNSS receiver with M array elements. The signal received by the GNSS receiver is shown in the following equation (7): In formula (7), u i (t) represents the complex envelope of the received signal; φ(t) represents the phase of the received signal; w0 is the carrier frequency of the signal; s i (t) represents the i-th narrowband signal; j represents the imaginary unit; τ represents the delay, which means the time difference of the signal reaching different array elements; Due to the relative characteristics of the time domain and the frequency domain, a narrowband signal means that the time domain changes slowly, so it satisfies the following equation (8): Substituting formula (8) into formula (7), we get: Therefore, the signal received by the kth array element is expressed as follows (9): In formula (9), g ki is the gain of the kth array element for the ith narrowband signal, n k (t) is the noise of the kth array element at time t, τ ki is the time delay of the i-th narrowband signal reaching the k-th array element relative to the reference array element; Therefore, the vector set of signals received by M array elements at time t is expressed as follows (10): Assuming that the direction of each array element is consistent and there is no mutual coupling between the array elements, and each channel is consistent, equation (10) is simplified to the following equation (11): Formula (11) can be expressed as a matrix: The steering matrix in formula (11) It is only related to the direction of the signal and the arrangement of the array elements; The signal received by the array element, i.e. the input signal, is expressed as the following formula (12): X(t)=X S (t)+X n (t) (12); In formula (12), X S (t) is the signal matrix received by the antenna array, i.e., M array elements, X n (t) is the noise matrix received by the antenna array. According to the input signal of equation (12) and equation (1), the output Y(t) of the GNSS receiver at time t is obtained as follows (13): Y(t)=W H X(t)=W H X s (t)+W H X n (t) (13); It is assumed that each array element is connected to N time domain registers, that is, the number of taps of each array element is N, and the delay interval is T, where T≤1 / B, B is the processing bandwidth inside the GNSS receiver, and the signal x of the mth tap on the i-th array element is i,m It is expressed as the following formula (14): x i,m =x i,1 (t-(m-1)T)=x 1,1 (t-τ i -(m-1)T) (14); In formula (14), x 1,1 is the signal received by the reference element, τ i is the time delay between the i-th array and the set reference array element, and the output Y is expressed as the following formula (15): In formula (15), The vector representing the signal received by the array element, i.e. the input signal; The power of equation (15) is obtained as follows (16): P out =E{|Y| 2 }=E{(W H X)(W H X) H }=W H R xx W (16); In formula (16), P out Represents output power, which measures the average energy of the processed signal; R XX =E{XX H } is the autocorrelation matrix of the space-time array input; E represents the expected value, which is used to describe the average value of the random variable; The following formula (17) is the optimal weight W when the adaptive weight of the first array element is constrained to 1. opt Solution: Formula (17) is an optimization problem, the goal is to minimize the output power P out , while satisfying the constraint that the adaptive weight of the first array element is 1; argmin represents the minimum output power P out , Subject to means optimization is performed under given constraints, i.e., when the constraint of the adaptive weight of the first array element is 1; a represents the MN×1-dimensional constraint vector, and the steering vector of the input signal in the definition space is a s , the time direction vector is a t =[1,e -jwT ,…,e -j(N-1)T ], so the space-time two-dimensional direction vector is represents the Kronecker product; Finally, W is obtained by the following formula (18): opt Minimum value of:

4. The space-time two-dimensional adaptive anti-interference system for a GNSS receiver according to claim 1, characterized in that: The digital oscillator, direct digital frequency synthesizer, low-pass filter, sampling and extraction module and adaptive algorithm module are all arranged in a field programmable gate array FPGA. The intermediate frequency signal received by the FPGA is a digital signal in the form of a floating point number, which needs to be first converted into a digital signal in the form of a fixed point number and then mixed. The conversion formula is shown in the following formula (19): fixed=x×2 Q (19); In formula (19), x is a digital signal in floating point form, fixed is a digital signal in fixed point form, and Q represents Q format.

5. The space-time two-dimensional adaptive anti-interference system for a GNSS receiver according to claim 1, characterized in that: The analog-to-digital converter is a 16-bit analog-to-digital converter with a bit width of 12 bits. The sampling rate of the analog-to-digital converter is f s Satisfies the following formula (20): In formula (20), n is a set positive integer, and f0 is the frequency of the intermediate frequency signal.

6. The space-time two-dimensional adaptive anti-interference system for a GNSS receiver according to claim 4, characterized in that: The adaptive algorithm module updates the optimal weights in an iterative manner. First, the optimal weights are obtained according to the power inversion algorithm. -W M-1 is the optimal weight of the input signal of M-1 array elements; The weight iteration formula is as follows (21): The operation steps of the adaptive algorithm module are shown in the following formula (22): In formula (22), Y(t) represents the output signal at the current time t; e(t) represents the error signal at the current time t; W(t+1) represents the weight vector at the next time; Indicates the weight update amount.

Images