A high-efficiency computing device and implementation method based on navigation communication anti-interference system

Abstract

The present invention belongs to the technical field of navigation communication anti-interference, and particularly relates to an efficient computing device and implementation method based on a navigation communication anti-interference system, wherein the computing module comprises a slope lookup table unit, an intersection lookup table unit, a binary conversion unit, a normalization unit, a sequence left shift unit, an address calculation unit, etc.; the implementation method comprises pre-creating a slope lookup table and an intersection lookup table, calculating addresses, address lookup slope parameters and intersection parameters, multiplier calculation, etc.; the present invention realizes fast and accurate calculation of complex formulas through innovative computing modules and efficient lookup table technology, significantly improves the computing efficiency of the anti-interference system, avoids complex square root and reciprocal calculations by pre-storing fitting parameters, greatly reduces computing delay and resource consumption, and the accuracy and efficiency of the lookup table ensure the accuracy of the calculation results, fully meeting the urgent needs of the navigation system for high-precision and low-latency information processing.

Description

Technical Field

[0001] The present invention belongs to the technical field of navigation communication anti-interference, and in particular relates to a high-efficiency computing device based on a navigation communication anti-interference system and an implementation method thereof. Background Art

[0002] In the current field of navigation communication anti-interference technology, the signal processing process involves several key steps, including:

[0003] Signal reception: Receive signals through antenna array;

[0004] Signal conversion: using down converter and analog-to-digital converter ADC to convert the received signal into universal digital data;

[0005] Anti-interference processing: The digital data is fed into the anti-interference system, which contains a core computing module to perform key computing tasks to obtain the output value;

[0006] In the above signal processing flow, the core computing module of the anti-interference system faces an important challenge, that is, it is impossible to directly calculate the inverse square root However, the inverse square root has important applications in signal processing, filtering, positioning algorithms, etc., especially in navigation and communication anti-interference systems, where its calculation accuracy and speed are crucial to achieving high-precision and low-latency navigation information:

[0007] To solve this problem, traditional methods usually use a two-step calculation process: first calculate the square root and then calculate its reciprocal. However, both calculations have significant challenges:

[0008] Square root calculation: It relies on complex numerical analysis methods, such as long division and Newton iteration. Although these methods can produce accurate results, they are computationally complex and require the anti-interference system to have a large amount of hardware resources to support them.

[0009] Reciprocal calculation: Traditional division methods, such as phase shift subtraction method, remainder recovery method, constant division method, and Newton iteration method, also have the problem of high resource consumption, requiring complex circuit design and a large number of computing units, and high calculation latency;

[0010] In addition, in order to calculate The value is converted to a fixed-point number for subsequent processing and needs to be multiplied by , and round down to the nearest integer, , A is a natural number, x is a general digital data, and floor is a floor function;

[0011] To sum up this series of processing steps, in the navigation communication anti-interference system, high latency and resource consumption rate will directly affect the performance and stability of the system, especially in a complex and changeable communication environment. The system needs to process a large amount of data quickly and accurately to ensure the accuracy and real-time nature of the navigation information. Therefore, the existing method of calculating the inverse square root and the application of the core calculation module in the field of navigation communication anti-interference are greatly limited. Summary of the invention

[0012] 1. Technical issues to be resolved

[0013] The present invention aims to solve the key problems of high complexity, large resource consumption, and high calculation delay in the square root and reciprocal calculation module in the traditional navigation communication anti-interference system. In particular, in view of the cumbersome process of step-by-step calculation of square roots and reciprocals in the current calculation module, as well as the resource consumption and delay problems caused by reliance on a large amount of hardware resources and complex circuit design, the present invention proposes an efficient computing device and implementation method based on the navigation communication anti-interference system, aiming to reduce resource consumption, improve computing efficiency, and shorten computing delay, thereby meeting the urgent needs of the navigation system for high-precision and low-latency information processing.

[0014] (II) Technical solution

[0015] The present invention is implemented by the following technical scheme: The present invention proposes an efficient computing device based on a navigation communication anti-interference system. The navigation communication anti-interference system is provided with an anti-interference module for processing received general digital data and converting it into an output value after the interference is effectively suppressed or eliminated. The anti-interference module is embedded with a computing module, which is used to calculate the formula The approximate value of , A and B are both natural numbers, x is general digital data, floor is the floor function, and the calculation modules include:

[0016] The slope lookup table unit is used to store and quickly retrieve the slope lookup table. The slope lookup table is a pre-calculated data set, which contains the slope parameters k corresponding to each address interval and the specific storage addresses of these slope parameters k in the lookup table; the storage address is determined based on the identifier or index corresponding to each address interval formed by equally dividing the interval [1, 2); the slope parameter k is determined: for each address interval, the formula is used To calculate a series of points, and use a linear function y=kx+b to fit these points, so as to determine the slope parameter k of the interval;

[0017] The intersection lookup table unit is used to store and quickly retrieve the intersection lookup table. The intersection lookup table is also a pre-calculated data set, which contains the intersection parameters b corresponding to each address interval and the specific storage addresses of these intersection parameters in the lookup table; the storage address and the intersection parameter b are determined in the same way as the slope lookup table;

[0018] A binary conversion unit, used for converting general digital data into binary numbers and recording the number of digits N of the binary number;

[0019] The normalization unit is connected to the binary conversion unit and is used to perform normalization processing on the binary number. The processing method is specifically to start from the highest index N-1 in the binary number, scan to the lower bits, find the first bit position with a value of 1, record the index of the position as C-1, and retain all bits from the lowest bit to the bit to form a normalized subsequence;

[0020] The sequence left shift unit is connected to the normalization unit and is used to shift the normalized subsequence to the left by the number of bits N-C, thereby generating a left-shifted binary number;

[0021] The address calculation unit is connected to the sequence left shift unit, the slope lookup table unit and the intersection lookup table unit respectively, and is used to convert the left-shifted binary number into an address; the specific method of converting to the address is: the address calculation unit determines whether the value of C-1 is an odd number, if C-1 is an even number, then extracts all subsequences of bit positions from N-1 to N-1-log2 (M) from the left-shifted binary number, and converts them into a decimal number h, and the address is equal to h+1; if C-1 is an odd number, the address is equal to h+1+ log2 (M); wherein M is the number of address interval divisions;

[0022] A multiplier, connected to the sequence left shift unit and the slope lookup table unit, respectively, for extracting all subsequences of bit positions from N-1 to N-1-(A-B) in the left-shifted binary number, converting them into a decimal number j, and multiplying them with the slope parameter k to obtain a product result;

[0023] The intersection left shift unit is connected to the intersection lookup table unit and is used to shift the intersection parameter b left by AB bits, which is ;

[0024] The adder is connected to the multiplier and the intersection left shift unit respectively, and is used to Add the multiplication result to obtain the addition result;

[0025] The denormalization unit is connected to the adder and is used to denormalize the addition result. , and obtain the output value.

[0026] Preferably, the navigation communication anti-interference system includes:

[0027] At least one antenna element for receiving signals;

[0028] Each antenna element is respectively connected to a down-converter, which is used to convert the received signal into a low-frequency signal;

[0029] Each down-converter is respectively connected to an analog-to-digital converter ADC, which is used to convert the intermediate frequency signal into universal digital data;

[0030] The outputs of all analog-to-digital converters are summed up and connected to an anti-interference module, which is used to convert the received digital data into output values ​​after the interference is effectively suppressed or eliminated;

[0031] A calculation module is provided in the anti-interference module for performing relevant calculations;

[0032] A digital-to-analog converter DAC is connected to the anti-interference module, and the digital-to-analog converter DAC is used to convert digital data back into analog signals;

[0033] The digital-to-analog converter is connected to an up-converter, which is used to convert the analog signal into a radio frequency signal.

[0034] The present invention also provides a method for realizing an efficient computing device based on a navigation communication anti-interference module, wherein the computing formula is An approximate value of , A is a natural number, x is a general digital data, floor is a floor function, and the implementation method is implemented based on any one of the above-mentioned efficient computing devices;

[0035] The implementation method comprises the following steps:

[0036] a. Pre-create a slope lookup table and an intersection lookup table, and store them in the slope lookup table unit and the intersection lookup table unit respectively. The specific steps are as follows:

[0037] I. Interval division:

[0038] Given an interval [1, 2), plan it as Divide equally, there will be M+1 endpoints and M address intervals, and the length of each address interval is , where m is a positive integer;

[0039] II. Endpoint calculation:

[0040] For address intervals from 1 to M, calculate their corresponding endpoints ( ),in, , ;

[0041] i=1, 2, …, M; A>B, B is a natural number;

[0042] III. Slope parameter calculation;

[0043] For each address interval 1 to M, calculate the slope parameter of the fitting function once ;

[0044] Slope parameter The calculation formula is:

[0045] ;

[0046] Where, i=1, 2…, M-1;

[0047] IV. Calculation of intersection parameters:

[0048] For each address range 1 to M, calculate the y-axis intersection point of the fitting function once ;

[0049] Y-axis intersection The calculation formula is:

[0050] ;

[0051] Where, i=1, 2…, M-1;

[0052] V. Constructing the lookup table unit:

[0053] Create a slope lookup table to store each address range 1 to M and its corresponding slope parameter ;

[0054] Create another intersection lookup table to store each address range 1 to M and its corresponding y-axis intersection point ;

[0055] The slope lookup table and the intersection lookup table are stored in the slope lookup table unit and the intersection lookup table unit respectively;

[0056] b. Calculate the address. The specific steps are as follows:

[0057] 1st) The binary conversion unit receives the general digital data, converts it into a binary number, and records the number of bits of the binary number as N;

[0058] 2nd) The normalization unit normalizes the binary number, that is, starting from the highest index N-1 of the binary number, scanning to the lower bits, finding the first bit position with a value of 1, recording the index of this position as C-1, and retaining all bits from the lowest bit to this bit to form a normalized subsequence, represented by x[C-1:0];

[0059] 3rd) The sequence left shift unit shifts x[C-1:0] to the left by N-C digits, thereby generating a left-shifted binary number z;

[0060] 4th) The address calculation unit converts the left-shifted binary number z into an address f, wherein the address calculation unit determines the parity of C-1 to calculate the address f; the specific method of converting to the address is: the address calculation unit determines whether the value of C-1 is an odd number, and if C-1 is an even number, all subsequences of bit positions from N-1 to N-1-log2 (M) are extracted from the left-shifted binary number z, and converted into a decimal number h, where f is equal to h+1; if C-1 is an odd number, f is equal to h+1+ log2 (M); wherein M is I in step a and M in the interval division;

[0061] c. The slope lookup table unit and the intersection lookup table unit directly find the slope parameter corresponding to the address f according to the calculated address f and intersection parameters , slope parameter Input to the multiplier, intersection parameter Shift the unit input leftward toward the intersection;

[0062] d. Multiplier calculation: The multiplier extracts all subsequences from N-1 to N-1-(A-B) in the left-shifted binary number z and converts them into decimal representation, denoted as j. One end of the multiplier inputs j and the other end inputs the slope parameter , multiply the two and get the product j∙k f ;

[0063] e. Move the intersection point left by one unit and change the intersection point parameter Move leftward by the bit position A-B, denoted as ;

[0064] f. The adder input has the product result j∙k f as well as , add the two together to get the result;

[0065] g. The denormalization unit divides the result obtained in step f by , and obtain the output value.

[0066] Preferably, when the interval is extended to M+1 to 2M, the formula is used To calculate a series of points, and use a linear function y=kx+b to fit these points to determine the slope parameter k and the intersection parameter b of the interval.

[0067] (III) Beneficial effects

[0068] The present invention realizes the calculation of complex formulas through innovative calculation modules and efficient lookup table technology. The fast and accurate calculation significantly improves the computing efficiency of the anti-interference system. By pre-storing the fitting parameters, it avoids complex square root and reciprocal calculations, greatly reducing the computing delay and resource consumption. The accuracy and efficiency of the lookup table ensure the accuracy of the calculation results, fully meeting the urgent needs of the navigation system for high-precision and low-latency information processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0070] Figure 1 It is a system block diagram of the computing module of the present invention.

[0071] Figure 2 It is a system block diagram of the anti-interference system of the present invention. DETAILED DESCRIPTION

[0072] Reference Figure 1 As shown, the present invention proposes an efficient computing device based on a navigation communication anti-interference system. The navigation communication anti-interference system is provided with an anti-interference module for processing received general digital data and converting it into an output value after the interference is effectively suppressed or eliminated. The anti-interference module is embedded with a computing module, which is used to calculate the formula The approximate value of , A and B are both natural numbers, x is general digital data, floor is the floor rounding function, and the calculation modules include:

[0073] The slope lookup table unit is used to store and quickly retrieve the slope lookup table. The slope lookup table is a pre-calculated data set, which contains the slope parameters k corresponding to each address interval and the specific storage addresses of these slope parameters k in the lookup table; the storage address is determined based on the identifier or index corresponding to each address interval formed by equally dividing the interval [1, 2); the slope parameter k is determined: for each address interval, the formula is used To calculate a series of points, and use a linear function y=kx+b to fit these points, so as to determine the slope parameter k of the interval;

[0074] The intersection lookup table unit is used to store and quickly retrieve the intersection lookup table. The intersection lookup table is also a pre-calculated data set, which contains the intersection parameters b corresponding to each address interval and the specific storage addresses of these intersection parameters in the lookup table; the storage address and the intersection parameter b are determined in the same way as the slope lookup table;

[0075] A binary conversion unit, used for converting general digital data into binary numbers and recording the number of digits N of the binary number;

[0076] The normalization unit is connected to the binary conversion unit and is used to perform normalization processing on the binary number. The processing method is specifically to start from the highest index N-1 in the binary number, scan to the lower bits, find the first bit position with a value of 1, record the index of the position as C-1, and retain all bits from the lowest bit to the bit to form a normalized subsequence;

[0077] The sequence left shift unit is connected to the normalization unit and is used to shift the normalized subsequence to the left by the number of bits N-C, thereby generating a left-shifted binary number;

[0078] The address calculation unit is connected to the sequence left shift unit, the slope lookup table unit and the intersection lookup table unit respectively, and is used to convert the left-shifted binary number into an address; the specific method of converting to the address is: the address calculation unit determines whether the value of C-1 is an odd number, if C-1 is an even number, then extracts all subsequences of bit positions from N-1 to N-1-log2 (M) from the left-shifted binary number, and converts them into a decimal number h, and the address is equal to h+1; if C-1 is an odd number, the address is equal to h+1+ log2 (M); wherein M is the number of address interval divisions;

[0079] A multiplier, connected to the sequence left shift unit and the slope lookup table unit, respectively, for extracting all subsequences of bit positions from N-1 to N-1-(A-B) in the left-shifted binary number, converting them into a decimal number j, and multiplying them with the slope parameter k to obtain a product result;

[0080] The intersection left shift unit is connected to the intersection lookup table unit and is used to shift the intersection parameter b left by AB bits, which is ;

[0081] The adder is connected to the multiplier and the intersection left shift unit respectively, and is used to Add the multiplication result to obtain the addition result;

[0082] The denormalization unit is connected to the adder and is used to denormalize the addition result. , and obtain the output value.

[0083] Reference Figure 2 As shown, the navigation communication anti-interference system includes:

[0084] At least one antenna element for receiving signals;

[0085] Each antenna element is respectively connected to a down-converter, which is used to convert the received signal into a low-frequency signal;

[0086] Each down-converter is respectively connected to an analog-to-digital converter ADC, which is used to convert the intermediate frequency signal into universal digital data;

[0087] The outputs of all analog-to-digital converters are summed up and connected to an anti-interference module, which is used to convert the received digital data into output values ​​after the interference is effectively suppressed or eliminated;

[0088] A calculation module is provided in the anti-interference module for performing relevant calculations;

[0089] A digital-to-analog converter DAC is connected to the anti-interference module, and the digital-to-analog converter DAC is used to convert digital data back into analog signals;

[0090] The digital-to-analog converter is connected to an up-converter, which is used to convert the analog signal into a radio frequency signal.

[0091] The present invention also provides a method for realizing an efficient computing device based on a navigation communication anti-interference module, wherein the computing formula is An approximate value of , A is a natural number, x is a general digital data, floor is a floor function, and the implementation method is implemented based on any one of the above-mentioned efficient computing devices;

[0092] The implementation method comprises the following steps:

[0093] a. Pre-create a slope lookup table and an intersection lookup table, and store them in the slope lookup table unit and the intersection lookup table unit respectively. The specific steps are as follows:

[0094] I. Interval division:

[0095] Given an interval [1, 2), plan it as Divide equally, there will be M+1 endpoints and M address intervals, the length of each address interval is , where m is a positive integer;

[0096] II. Endpoint calculation:

[0097] For address intervals from 1 to M, calculate their corresponding endpoints ( ),in, , ;

[0098] i=1, 2, …, M; A>B, B is a natural number;

[0099] III. Slope parameter calculation;

[0100] For each address interval 1 to M, calculate the slope parameter of the fitting function once ;

[0101] Slope parameter The calculation formula is:

[0102] ;

[0103] Where, i=1, 2…, M-1;

[0104] IV. Calculation of intersection parameters:

[0105] For each address range 1 to M, calculate the y-axis intersection point of the fitting function once ;

[0106] Y-axis intersection The calculation formula is:

[0107] ;

[0108] Where, i=1, 2…, M-1;

[0109] V. Constructing the lookup table unit:

[0110] Create a slope lookup table to store each address range 1 to M and its corresponding slope parameter ;

[0111] Create another intersection lookup table to store each address range 1 to M and its corresponding y-axis intersection point ;

[0112] The slope lookup table and the intersection lookup table are stored in the slope lookup table unit and the intersection lookup table unit respectively;

[0113] b. Calculate the address. The specific steps are as follows:

[0114] 1st) The binary conversion unit receives the general digital data, converts it into a binary number, and records the number of bits of the binary number as N;

[0115] 2nd) The normalization unit normalizes the binary number, that is, starting from the highest index N-1 of the binary number, scanning to the lower bits, finding the first bit position with a value of 1, recording the index of this position as C-1, and retaining all bits from the lowest bit to this bit to form a normalized subsequence, represented by x[C-1:0];

[0116] 3rd) The sequence left shift unit shifts x[C-1:0] to the left by N-C digits, thereby generating a left-shifted binary number z;

[0117] 4th) The address calculation unit converts the left-shifted binary number z into an address f, wherein the address calculation unit determines the parity of C-1 to calculate the address f; the specific method of converting to the address is: the address calculation unit determines whether the value of C-1 is an odd number, and if C-1 is an even number, all subsequences of bit positions from N-1 to N-1-log2 (M) are extracted from the left-shifted binary number z, and converted into a decimal number h, where f is equal to h+1; if C-1 is an odd number, f is equal to h+1+ log2 (M); wherein M is I in step a and M in the interval division;

[0118] c. The slope lookup table unit and the intersection lookup table unit directly find the slope parameter corresponding to the address f according to the calculated address f and intersection parameters , slope parameter Input to the multiplier, intersection parameter Shift the unit input leftward toward the intersection;

[0119] d. Multiplier calculation: The multiplier extracts all subsequences from N-1 to N-1-(A-B) in the left-shifted binary number z and converts them into decimal representation, denoted as j. One end of the multiplier inputs j and the other end inputs the slope parameter , multiply the two and get the product j∙k f ;

[0120] e. Move the intersection point left by one unit and change the intersection point parameter Move leftward by the bit position A-B, denoted as ;

[0121] f. The adder input has the product result j∙k f as well as , add the two together to get the result;

[0122] g. The denormalization unit divides the result obtained in step f by , and get the output value;

[0123] Among them, when the processing address range is extended from M+1 to 2M, the formula is used To calculate a series of points, and use a linear function y=kx+b to fit these points, respectively determine the slope parameter k and intersection parameter b of the interval.

[0124] Among them, the calculation module and implementation method are applied in the field of navigation communication anti-interference technology.

[0125] The efficient computing device of the present invention is a key component in the navigation communication anti-interference system. Its working principle is based on a pre-calculated lookup table and a series of efficient digital processing steps to achieve fixed-point number in a very short time (within 3 system clocks). Therefore, the present invention proposes an efficient computing device and implementation method based on the navigation communication anti-interference system, aiming to reduce resource consumption, improve computing efficiency, and shorten computing delay, thereby meeting the urgent needs of the navigation system for high-precision and low-latency information processing:

[0126] The present invention significantly improves the computational efficiency and performance of the anti-interference system by:

[0127] Significantly improve computing efficiency:

[0128] The present invention creates two lookup table units (slope lookup table and intersection lookup table) to pre-store the fitting parameters (slope k and y-axis intersection b) obtained based on linear function fitting. This design allows the required fitting parameters to be quickly obtained directly through the lookup table during operation, avoiding complex square root and reciprocal calculations, thereby significantly improving the calculation efficiency;

[0129] The introduction of efficient digital processing steps, such as normalization, sequence left shift, and address calculation, further reduces the computing latency and improves the overall computing efficiency;

[0130] Significantly reduce resource consumption:

[0131] The use of lookup tables significantly reduces the need for hardware resources, because the complex calculation process has been completed and stored in advance and does not need to be performed at runtime;

[0132] Efficient algorithm design reduces the number and complexity of computing units, thereby further reducing resource consumption;

[0133] Meeting high precision and low latency requirements:

[0134] The accuracy and efficiency of the lookup table ensure the accuracy of the calculation results and meet the navigation system's demand for high-precision information;

[0135] The fast calculation process ensures low-latency information processing, meeting the urgent real-time requirements of the navigation system;

[0136] Have good scalability;

[0137] For the extended interval, the present invention provides similar formulas and fitting methods for calculating the slope parameters and the intersection parameters, thereby conveniently expanding the application range of the lookup table to meet different calculation requirements;

[0138] In summary, the present invention realizes the calculation of complex formulas through innovative calculation modules and implementation methods. The fast and accurate calculation significantly improves computing efficiency, reduces resource consumption, and meets the urgent needs of the navigation system for high-precision and low-latency information processing, while also having good scalability.

[0139] Example:

[0140] The navigation communication anti-interference system receives the signal, and after a series of processing, it is converted into universal digital data by the analog-to-digital converter ADC. The universal digital data is 304240287024. The anti-interference module is used to facilitate Calculate and convert it into a fixed-point number, which is ;

[0141] The present invention is based on , pre-create the slope lookup table and the intersection lookup table, and store them in the slope lookup table unit and the intersection lookup table unit respectively. The specific steps are as follows:

[0142] Interval partitioning: Given an interval [1, 2), plan it as Divide equally, there will be 512+1=513 endpoints and 512 address intervals, and the length of each address interval is ;

[0143] Endpoint calculation: For the address range from 1 to 512, according to the formula , , calculate its corresponding endpoint ( ); 44 is greater than B;

[0144] That is , ; i = 1, 2, …, 512;

[0145] Slope parameter calculation: For each address range 1 to 512, calculate the slope parameter of the fitting function once ;

[0146] Slope parameter The calculation formula is:

[0147] ;

[0148] Where, i=1, 2…, 512-1;

[0149] Intersection parameter calculation: For each address range 1 to 512, calculate the y-axis intersection of the fitting function once ;

[0150] Y-axis intersection The calculation formula is:

[0151] ;

[0152] Where, i=1, 2…, 512-1;

[0153] Create a slope lookup table to store each address range 1 to 512 and its corresponding slope parameter ; As shown in the following table:

[0154]

[0155] Create another intersection lookup table to store each address range 1 to 512 and its corresponding y-axis intersection point ; As shown in the following table:

[0156]

[0157] The slope lookup table and the intersection lookup table are stored in the slope lookup table unit and the intersection lookup table unit respectively;

[0158] Calculate the address. The specific steps are as follows:

[0159] The binary conversion unit receives the general digital data 304240287024 and converts it into a binary number, which is represented as:

[0160] (100011011010110001000100101110100110000) 2; record the number of digits as N=50;

[0161] The normalization unit normalizes the binary number, that is, starting from the highest index 50-1 of the binary number (100011011010110001000100101110100110000)2, scans to the lower bits, finds the first bit position with a value of 1, records the index of this position as 39-1, and retains all bits from the lowest bit to this bit to form a normalized subsequence, represented by x[C-1:0], which is x [39-1:0];

[0162] Sequence left shift unit pair x [39-1:0] is shifted left by the digits 50-39, thus generating a left-shifted binary number z, which is:

[0163] (100011011010110001000100101110100110000000000000000)2

[0164] The address calculation unit determines that 39-1 is an even number, that is, it extracts the bit subsequence from 50-1 to 50-1-log2 (512) and converts it into a decimal number h, h=54, and f is equal to 54+1=55;

[0165] The slope lookup table unit and the intersection lookup table unit directly find out the slope parameter -450480 and the intersection parameter 1495294 corresponding to the address 55 according to the calculated address 55, -450480 is input to the multiplier, and 1495294 is input to the intersection left shift unit;

[0166] Multiplier calculation: The multiplier extracts all subsequences of the bit positions starting from 50-1 to 50-1-(43-20) in the left-shifted binary number z (10001101101011000100010010111010011000000000000000)2, and converts it into a decimal number of 9284676. The multiplier inputs 9284676 at one end and -450480 at the other end. The two are multiplied together to obtain the product result -4182560844480.

[0167] The intersection left shift unit shifts 1495294 to the left by 43-20 bits, which is 1495294∙2 43-20 =12543435210752;

[0168] The adder adds -4182560844480 and 12543435210752 to get 8360874366272.

[0169] To normalize the unit, 8360874366272÷ , we get 15947102, which is The results of the approximation;

[0170] Verification: First use the computer to verify The calculated result is 15947090. By comparison, we found that there is only a slight difference between the result obtained by the method of the present invention and the result directly calculated by the computer, and the error rate is 1-15947090÷15947102=7.52488×10 -7 It can be seen that the calculation result of the reciprocal square root of the present invention is extremely close to the real result. This extremely small error rate fully demonstrates the excellent accuracy and high reliability of the present invention in calculating the reciprocal square root.

[0171] In summary, the square root reciprocal calculation method of the present invention not only has an efficient calculation process, but also obtains a result that is extremely close to the true value, and the error is far below the acceptable range, thereby verifying the accuracy and reliability of the method of the present invention.

Claims

1. An efficient computing device based on a navigation communication anti-interference system, wherein the navigation communication anti-interference system is provided with an anti-interference module for processing received general digital data and converting it into an output value after the interference is effectively suppressed or eliminated, and the anti-interference module is embedded with a computing module for calculating the formula The approximate value of , A is a natural number, x is a general digital data, floor is a floor function, which is characterized by: The calculation modules include: The slope lookup table unit is used to store and quickly retrieve the slope lookup table. The slope lookup table is a pre-calculated data set, which contains the slope parameters k corresponding to each address interval and the specific storage addresses of these slope parameters k in the lookup table; the storage address is determined based on the identifier or index corresponding to each address interval formed by equally dividing the interval [1, 2); the slope parameter k is determined: for each address interval, the formula is used To calculate a series of points, and use a linear function y=kx+b to fit these points, so as to determine the slope parameter k of the interval, B is a natural number; The intersection lookup table unit is used to store and quickly retrieve the intersection lookup table. The intersection lookup table is also a pre-calculated data set, which contains the intersection parameters b corresponding to each address interval and the specific storage addresses of these intersection parameters in the lookup table; the storage address and the intersection parameter b are determined in the same way as the slope lookup table; A binary conversion unit, used for converting general digital data into binary numbers and recording the number of digits N of the binary number; The normalization unit is connected to the binary conversion unit and is used to perform normalization processing on the binary number. The processing method is specifically to start from the highest index N-1 in the binary number, scan to the lower bits, find the first bit position with a value of 1, record the index of the position as C-1, and retain all bits from the lowest bit to the bit to form a normalized subsequence; The sequence left shift unit is connected to the normalization unit and is used to shift the normalized subsequence to the left by the number of bits N-C, thereby generating a left-shifted binary number; The address calculation unit is connected to the sequence left shift unit, the slope lookup table unit and the intersection lookup table unit respectively, and is used to convert the left-shifted binary number into an address; the specific method of converting to the address is: the address calculation unit determines whether the value of C-1 is an odd number, if C-1 is an even number, then all subsequences of bit positions from N-1 to N-1-log2 (M) are extracted from the left-shifted binary number, and converted into a decimal number h, and the address is equal to h+1; if C-1 is an odd number, the address is equal to h+1+ log2 (M); wherein M is the number of address interval divisions; A multiplier, connected to the sequence left shift unit and the slope lookup table unit, respectively, for extracting all subsequences of bit positions from N-1 to N-1-(A-B) in the left-shifted binary number, converting them into a decimal number j, and multiplying them with the slope parameter k to obtain a product result; The intersection left shift unit is connected to the intersection lookup table unit and is used to shift the intersection parameter b left by AB bits, which is ; The adder is connected to the multiplier and the intersection left shift unit respectively, and is used to Add the multiplication result to obtain the addition result; The denormalization unit is connected to the adder and is used to denormalize the addition result. , and obtain the output value.

2. According to claim 1, a high-efficiency computing device based on a navigation communication anti-interference system is characterized in that: The navigation communication anti-interference system includes: At least one antenna element for receiving signals; Each antenna element is respectively connected to a down-converter, which is used to convert the received signal into a low-frequency signal; Each down-converter is respectively connected to an analog-to-digital converter ADC, which is used to convert the intermediate frequency signal into universal digital data; The outputs of all analog-to-digital converters are summed up and connected to an anti-interference module, which is used to convert the received digital data into output values ​​after the interference is effectively suppressed or eliminated; A calculation module is provided in the anti-interference module for performing relevant calculations; A digital-to-analog converter DAC is connected to the anti-interference module, and the digital-to-analog converter DAC is used to convert digital data back into analog signals; The digital-to-analog converter is connected to an up-converter, which is used to convert the analog signal into a radio frequency signal.

3. A method for implementing an efficient computing device based on a navigation communication anti-interference module, characterized in that: The implementation method is implemented based on the high-efficiency computing device described in any one of claims 1-2; The implementation method comprises the following steps: a. Pre-create a slope lookup table and an intersection lookup table, and store them in the slope lookup table unit and the intersection lookup table unit respectively. The specific steps are as follows: I. Interval division: Given an interval [1, 2), plan it as Divide equally, there will be M+1 endpoints and M address intervals, the length of each address interval is , where m is a positive integer; II. Endpoint calculation: For address intervals from 1 to M, calculate their corresponding endpoints ( ),in, , ; i=1, 2, …, M; A>B, B is a natural number; III. Slope parameter calculation; For each address interval 1 to M, calculate the slope parameter of the fitting function once ; Slope parameter The calculation formula is: ； Where i=1, 2…, M; IV. Calculation of intersection parameters: For each address range 1 to M, calculate the y-axis intersection point of the fitting function once ; Y-axis intersection The calculation formula is: ； Where i=1, 2…, M; V. Constructing the lookup table unit: Create a slope lookup table to store each address range 1 to M and its corresponding slope parameter ; Create another intersection lookup table to store each address range 1 to M and its corresponding y-axis intersection point ; The slope lookup table and the intersection lookup table are stored in the slope lookup table unit and the intersection lookup table unit respectively; b. Calculate the address. The specific steps are as follows: 1st) The binary conversion unit receives the general digital data, converts it into a binary number, and records the number of bits of the binary number as N; 2nd) The normalization unit normalizes the binary number, that is, starting from the highest index N-1 of the binary number, scanning to the lower bits, finding the first bit position with a value of 1, recording the index of this position as C-1, and retaining all bits from the lowest bit to this bit to form a normalized subsequence, represented by x[C-1:0]; 3rd) The sequence left shift unit shifts x[C-1:0] to the left by N-C digits, thereby generating a left-shifted binary number z; 4th) The address calculation unit converts the left-shifted binary number z into an address f, wherein the address calculation unit determines the parity of C-1 to calculate the address f; the specific method of converting to the address is: the address calculation unit determines whether the value of C-1 is an odd number, and if C-1 is an even number, all subsequences of bit positions from N-1 to N-1-log2 (M) are extracted from the left-shifted binary number z, and converted into a decimal number h, where f is equal to h+1; if C-1 is an odd number, f is equal to h+1+ log2 (M); wherein M is I in step a and M in the interval division; c. The slope lookup table unit and the intersection lookup table unit directly find the slope parameter corresponding to the address f according to the calculated address f and intersection parameters , slope parameter Input to the multiplier, intersection parameter Shift the unit input leftward toward the intersection; d. Multiplier calculation: The multiplier extracts all subsequences from N-1 to N-1-(A-B) in the left-shifted binary number z and converts them into decimal representation, denoted as j. One end of the multiplier inputs j and the other end inputs the slope parameter , multiply the two and get the product j∙k f ; e. Move the intersection point left by one unit and change the intersection point parameter Move leftward by the bit position A-B, denoted as ; f. The adder input has the product result j∙k f as well as , add the two together to get the result; g. The denormalization unit divides the result obtained in step f by , and obtain the output value.

4. The method for implementing a high-efficiency computing device based on a navigation communication anti-interference module according to claim 3 is characterized in that: When the interval is extended to M+1 to 2M, use the formula To calculate a series of points, and use a linear function y=kx+b to fit these points to determine the slope parameter k and the intersection parameter b of the interval.

Images