# The device for calculating the azimuthal correlation function

(57) Abstract:

The invention is intended to perform the function of correlation of the input signal data with the reference function in the satellites with synthetic aperture radar. This device has associated with the control unit and between the unit two-dimensional memory, driver support functions, signal processor, the output unit of the buffer memory. The two-dimensional block of memory contains a circular driven parallel shift register and fragments of memory, each of which has one data input, two address input, the address selection and one output, connected with the respective input cyclic managed parallel shift register, the driver support functions performed in the form of a matrix of interpolators, each of which consists of a shaper is a common part of the support functions and the shapers of individual parts of each support function whose inputs are connected to the output of the shaper is a common part of the support functions, signal processor consists of parallel controlled shift registers and the matrix of processor elements, the rows of which are located along the azimuthal axis, and the columns along the axis palestinianism registers, and vertically and horizontally - on pieces and uses corresponding to the interpolators of driver support functions, with each output parallel controlled shift register connected to information inputs of processor elements of the corresponding row of column groups, and the output of each driver's individual part of the support functions with the inputs of the support functions of processor elements of the corresponding column pieces and uses and outputs of the processor elements of the matrix connected to the input of the output buffer memory, and inputs the parallel controlled shift registers corresponding to one row of the matrix of processor elements, are connected with the corresponding output cyclic managed parallel shift register block two-dimensional memory. 9 Il. The invention relates to computer technology and can be used for the calculation of the correlation function of the input signal data reference function (PF) in the satellites of the type "DIAMOND" (Russia), ERS-1 (Europe), JERS-1 (Japan), RADARSAT (Canada), SIR-A (USA), for aircraft type AWACS and JSTARS, and in medicine in tomography and ultrasonic sensing, and Geology.A device for calculating the azimuth azimuthally correlation function. A disadvantage of this device is that it does not take into account the effect of migration on the cruise, which does not allow to obtain high resolution. The device also does not allow parallel processing of the input information. In addition, the reference function is stored in the memory unit, and this leads to the need to transfer large amounts of information when changing the set of reference functions.It is also known a device for calculating the azimuthal correlation function that contains the control unit, the host of the two-dimensional memory, the reference features and the compute node point correlation functions. This device takes into account migration range, but does not allow you to choose the samples in accordance with various curves migration range for different axis range of points in the image. This circumstance does not allow to obtain high resolution. In addition, the use of block memory for storing reference functions leads to the need to transfer large amounts of information when changing the set of reference functions.The proposed device for calculating the azimuthal correlation function allows to eliminate these drawbacks and to obtain a technical result, warehouses eligibility parameters ceteris paribus.This is achieved by a device for calculating the azimuthal correlation function includes the control unit, the host of the two-dimensional memory, the reference features and the compute node point correlation functions, and the node of the two-dimensional memory consists of a group of (M+N) memory blocks, where M is the azimuthal number of the sequences of samples that form a two-dimensional grid of samples group adjacent axis range broken lines, with the bend at the same azimuthal coordinate axis; N the number of shifts on the reference axis range group azimuthal sequence of samples corresponding to the change in distance when changing azimuthal coordinates in the results matrix, and block cyclic shift register, reference functions consists of a matrix (g,h)-x generators of support functions, where g 1.G, G the number of support functions required to generate the results, having the same azimuthal coordinate; h 1.N, N the number of generators of the support functions necessary to calculate the reference functions in the formation of results having the same coordinate in range, and the matrix (g,h)-h groups, arithmetic units, and the compute node point correlating functions consists of matradee and the same group of neighboring axis range of the azimuthal sequences of samples, and groups of blocks shift registers, the data inputs of memory blocks of the group connected to the information input device, the first and second address outputs of the control unit are connected with the same address inputs of memory blocks of the group, selects the addresses of the memory blocks of the group connected to the group of outputs of the address selection control unit, respectively, the outputs of the memory blocks of the group are connected with the corresponding information input unit cyclic shift registers, the input offset control which is connected to the output of the offset control of the control unit, the output of each (g,h)-th generator of the support functions connected with the inputs of the l-th arithmetic units each (g, h)-th group, where l 1.L, L the number of support functions, calculated using additives to the same General part of the f-th information input (where f 1.M+N) of each unit shift register group is connected with f-m output block cyclic shift registers host a two-dimensional memory, the control inputs of the shift block shift register group are connected with a group of control outputs of the shift control unit, the i-th output of the a-th block shift registers, where: a 1.N, connected to the first input (i,b)-x correlators, where b (a-1)xK+1.a x K, K A/(N+1) with atropinisation, where (g-1)xF+1. gxF, F the number of adjacent axis range correlators receiving the same values of the support function, d (h-1)x L. hxL, the outputs of correlators connected to the output device.In Fig. 1 schematically presents "broken band" selection of samples of the received signals in accordance with the effect of migration distance in the coordinate system of the azimuth/range; Fig. 2 scheme of the ideal sequence of samples of the received signals for forming an adjacent axis azimuth of the points in the image of Fig.3 scheme of a single sequence of samples of the received signals for forming an adjacent axis azimuth of the points in the image of Fig. 4 scheme of the perfect blend of sequences of samples of the received signals for forming an adjacent axis azimuth of the points in the image of Fig.5 is a structural diagram of a device for calculating the azimuthal correlation function of Fig.6 is a structural diagram of a node of a two-dimensional memory of Fig.7 is a structural diagram of the (1,1)-th generator of the reference functions (1,1)-th group arithmetic units and connected with her part of the matrix of correlators of Fig.8 block diagram of the compute node point correlation functions of Fig. 9 schema matching sequence counts biocoenose use multiprocessor matrix structure to reduce the processing time of the input information. Secondly, it is the possibility of accounting for the effect of migration on the cruise, allowing you to increase resolution of the resulting image. Thirdly, it is the ability to create on-Board variant of the device for calculation of the azimuthal correlation function due to the reduction of hardware costs. This is based on the following principles:the selection of samples of the received signal from the input two-dimensional array within the "broken band", has constant width and taking into account the migration range; using the same sample sequence for the formation of several neighboring axis azimuth points in the image; using the same support function for the formation of several neighboring axis range of points in the image; calculating a reference functions several neighboring axis azimuth points in the image using the correction addition to the basic functions one Central point.For this purpose, in particular, used in the proposed device multiprocessor matrix structure should have a maximum size in range and azimuth. This raises the need to submit samples of the received signal in a multiprocessor matrix p is the property and described in the coordinate system of the azimuth/range migration curve (line migration distance), presented on Fig.1, which shows the position of the samples of the received signal on a two-dimensional grid (matrix samples), formed by lines of constant distance and constant azimuth, separated from each other by the amount of the incremental step. The samples of the received signal are located at the points of intersection of these lines. For processing select count received signal that is closest to the line migration distance. For each point in the image has its own line migration in range (curves in Fig.1) and a sequence of samples ("polyline" in Fig.1). In accordance with the invention, for points of the respective curves migration range, "polyline" selection of samples of the received signals have "kinks" in the same azimuthal coordinate axis, and splitting the received signals are within the "broken band" (Fig.1 she shaded), with constant width formed in each azimuthal position of the M neighboring axis range of the samples of the received signals.Optimal for calculations at the stage of processing mutually correlation function (Y) the received and reference signals PCA for the l-th point in the line image is the ratio of:

Y

_{l}=Z

^{(}

_{mtm(l)counting the reference signal of the l-th point of the image.In this case, when forming mutually correlation function for each point in the line image from a matrix of counts (see Fig.1) choose a sequence of samples of the received signals. So for the formation of mutually correlation function of the point b (see Fig.2) use the sequence of counts for points, and for the formation of mutually correlation function of the point From the sequence of samples for point C. it should be borne in mind that the line migration distance, is shown in Fig.1, represents the beginning (left end) of the sequence to point b (see Fig.2).For the formation of mutually correlation functions for several pixels in the line image (see Fig.3), such as points b, D, C, in the invention using a single (the same) a sequence of samples of the received signals, i.e., corresponding to the Central point in the image line (point D). In the present invention this is achieved by using multi-processor matrix structure, using for the formation of several neighboring axis azimuth image point one "broken band" selection of samples received signals>)where Zmm-th count of the received signal during the processing of each point in the line image;Each "broken line", forming collectively the "broken band" selection of samples of the received signals is used to receive the adjacent axis azimuth of image points with respect to the Central one in this line.When calculating mutually correlation function in equation (2) gain error ( ) compared with the same calculation function from (4) that does not exceed the values for this kind of errors identified in the TOR for a specific task:(3) where EI, JP timing generating the error (see Fig.4);EJ sequence of samples to obtain point B;The error for the prototype, created on the basis of the present invention, the technical specification set no more than 5% When the aperture size of 2000 and the number of neighboring processed simultaneously on a single sequence of points 64 this error does not exceed 3.2% i.e. tolerable.The same sequence of received signals has a usage limit, beyond which the increase in multiprocessor matrix patterns in the azimuth direction leads on the edges of the aperture nd patterns in the azimuth direction shift value increases. To compensate for this effect, the width of the broken strip" increase the value of this shift (N) in the invention is administered shear structure.The device for calculating the azimuthal correlation function contains the control unit 1 (see Fig.5), which, in particular, may be made in the form of ROM, site two-dimensional memory 2, the reference functions 3 and compute node point correlation functions 4.Node two-dimensional memory 2 (see Fig.6) consists of a group of f-x memory blocks 5-f, where f 1.M+N, M is the azimuthal number of the sequences of samples that form a two-dimensional grid of samples group adjacent axis range broken lines, with the bend at the same azimuthal coordinate axis, N is the number of shifts on the reference axis range group azimuthal sequence of samples corresponding to the change in distance when changing azimuthal coordinates in the results matrix, and block cyclic shift registers 6.Reference functions 3 (see Fig.7) consists of a matrix (g,h)-x generators support functions 7-g-h where g 1.G, G the number of support functions required to generate the results, having the same azimuthal coordinate, h 1.H, N the number of generators of support functions, range, and the matrix (g,h)-x groups l-x arithmetic units 8-l-(g-h), where l 1.L, L the number of support functions, calculated using additives to the same General part.Each arithmetic unit performs the operation: f(X) KO(X) + K1 X + K2 X, where KO(X) is the initial value set by the reference generator functions, K1, K2 are the coefficients of the polynomial (Fig. 7 shows the (1,1)-group arithmetic units and (1,1)-th generator), and the compute node point correlation functions [4] (see Fig.8) consists of a matrix (i,j)-th correlators 9-i-j, where i is 1. M, j 1.A, A number of adjacent axis azimuth of the results generated using the same group of neighboring axis range of the azimuthal sequences of samples, and a-x block shift registers 10 a, where a 1.N+1.The data inputs 11-f memory blocks 5-f group is attached to informacionnomu input device 11, the first address output 12 of the control unit 1 is connected with the first address inputs 13-f memory blocks 5-f group, the second address output 14 of the control unit 1 is connected with the second address inputs 15-f memory blocks 5-f group, the inputs of the address selection 16-f memory blocks 5-f group connected to the group of f-x outputs the address selection 17-f control unit 1, respectively, the outputs 18-f units Puma, input offset control 20 which is connected to the output 21 of the control by the shift control unit 1, the output 22-g-h each (g,h)-th generator of the support functions 7-g-h is connected to the inputs 23-1-g-h).23-L-(g-h) l-x arithmetic units 8-l-(g-h) for each (g,h)-th group, f-th information input 24-a-f of each unit shift registers 10-a group connected with f-m output 25-f of the block cyclic shift registers 6 node two-dimensional memory 2, the control inputs of the shift 26-a a-x block shift registers 10-a group connected with a group a-x 27 outputs a control by the shift control unit 1, the i-th output 28-a-i a-th block of the shift registers 10 and connected to the first inputs 29-i-b (i, b)-x correlators 9-i-b, where b (a-1)x K+1.a x K, K A/(N+1), rounded in the direction of a greater whole, output 30-l-(g-h) l-th arithmetic unit 8-l-(g-h)-th group is connected to the second inputs 31-c-d (c,d)-x correlators 9-C-d, where C (g-1) x F+1.g x F, F number of adjacent axis range correlators receiving the same values of the support function, d (h-1) x L. h L x outputs 32-i-j correlators 9-i-j is connected to the output 33 of the device.The proposed device for calculating the azimuthal correlation function is as follows.The recording process, the incoming samples of the received signal are served in ATI 5-1 (Fig.6), the second count on the information input 11-2 of the second memory block 5-2, etc. Sequence of memory blocks 5-1.5-(M+N) when recording samples is cylindrical in nature: ((M+N)+1)-th reference record again in the first memory block 5-1, but on the second address ((M+N)+2-th reference record in the second memory block 5-2, also on the second address, and so on, whenever the sequence reaches (M+N)-th memory block 5-(M+N), the loop is repeated, starting with the first memory block 5-1. During the write process, the active may be any of the two addresses supplied to the address inputs 13-f or 15-f memory blocks 5-1. 5-(M+N) with the address outputs 12 and 14 of the control unit 1. When the record sequence reaches the highest address of the memory blocks 5-1.5-(M+N), the recording start producing again with the first address, with the previously recorded data will be overwritten.The reading process: when the process of recording in the memory blocks 5-1.5-(M+N) accumulates information about the received signals, the number of which is equal to the aperture size, then simultaneously with the recording process of beginning to make the reading process. To ensure the selection of samples within the "broken bands of constant width in node two-dimensional memory 2 put the memory blocks 5-f, imetec, any (M+N) neighboring-axis range of the samples of the received signal are recorded in two adjacent addresses of memory blocks 5-f so that they occupy in the General case, an intermediate position between the sets of readings recorded at the lower and larger of the addresses. For the formation of a "broken band", taking into account the migration range of the first few blocks of memory 5-1. 5-X samples are choosing more of the addresses, and the remaining memory blocks 5-(X+1).5-(M+N) at the lower of addresses For this purpose with the second address output 14 of the control unit 1 more address served on the second address inputs 15-f two-dimensional blocks of memory 5-f, a lower address from the first address output 12 on the first address inputs 13-f two-dimensional blocks of memory 5-f, and the inputs of the address selection 16-1.16-X is available from the outputs of the address selection 17-1.17-f control unit, a value corresponding to the choice of the second address inputs 15-1.15-X, and inputs the address selection 16-(X+1).16-(M+N) value corresponding to the first address inputs 13-(X+1).13-(M+N).For example, if M 5, N 2 (see Fig.9) and, therefore, when the number of memory blocks, equal to seven, the neighboring samples of the first received signal with sequence numbers from the 31st to the 37th belong to the sets of readings recorded at "5" and "6". When atlake memory 5-7 at the address "5", 36th count in the memory block 5-1 at the address "6", and the 37-th sample in the memory block 5-2 at the address "6". To select 31 37th counts of the memory blocks on the second address inputs 15-f serves the address "6", the first address inputs 13-f serves the address "5". The inputs of the address selection 16-3.16-7 served with output address selection 17-3.17-7 value corresponding to the selection address inputs 13-3.13-7 first address and the input address selection 16-1 and 16-2 from the outputs of the address selection 17-1 and 17-7 serves a value corresponding to the selection address inputs 15-1 and 15-2 of the second address. As a result, the output 18-1 of the first memory block 5-1 receive the 36th reference output 18-2 memory block 5-2 receive 37th reference output 18-3 memory block 5-3, respectively, of the 31st count, etc. on the output 18-7 memory block 5-7 35th countdown.To restore the natural order of samples within a set timing with outputs 18-1.18-7 act on the information inputs 19-1.19-7 block cyclic shift register 6. Input offset control 20 output 21 of the control by the shift control unit 1 serves the value corresponding to the cyclic shift to "2". In the cyclic shift on the "2" output 25-1 receive 31st count, the output 25-2 32 count, etc. on the output 25-7 respectively 37th count in estestvennosti 31.37-th receive information inputs 24-a-1. 24-a-7 a-h blocks shift registers 10 and the compute node point correlation functions [4] (see Fig.5).Thus, for M=5 and N=2 the 31st countdown arrives at the inputs of 24-1-1.24-3-1 blocks shift registers 10-1.10-3 (see Fig.8); the 32nd count arrives at the inputs of 24-1-2. 24-3-2 blocks shift registers 10-1,10-3,37th countdown arrives at the inputs of 24-1-7.24-3-7 blocks shift registers 10-1.10-3. In the case of slope curves migration in range, when each group b-th column in the matrix of correlators required to issue its own set of indications on the control input shift 26-1 output offset control 27-1 serves a value corresponding to zero shift, resulting in outputs 28-1-1.28-1-5 unit shift registers 10-1 receives the samples from the 31st through 35th on the input offset control 26-2 output offset control 27-2 serves a value corresponding to the shift to "1", resulting in outputs 28-2-1.28-2-5 unit shift registers 10-2 receives the samples from 32 St to 36th, and to the input of the offset control 26-3 output offset control 27-3 serves a value corresponding to the shift to "2", resulting in outputs 28-3-1.28-3-5 unit shift registers 10-3 receives the samples from 33rd on the 37th. The sample outputs 28-a-1. 28-a-5 unit shift registers 10 and are received at the first inputs 29-i-b to the Loki 8-l-(g-h) for each (g,h)-th group of reference functions 3 (see Fig.1), getting the original values from the outputs 22-g-h, calculate the values of the support functions that come with l-x outputs 30-l-(g-h) arithmetic units 9-l-(g-h) (see Fig.7) of the second input 31-c-d correlators 9-c-d (see 7,8). Each correlator 9-c-d (see Fig.8) with the samples of the received signal, the number of which is equal to the aperture size, the first inputs 29-i-j and the reference function by the second inputs 31-i-j generates its value azimuthal correlation functions that come through outputs 32-i-j correlators 9-i-j to the output 33 of the device.The described device (prototype) from January 1992 working in NGOs engineering in the framework of the project "DIAMOND" and is intended to perform the function of correlation of the input signal data with the reference function. The invention is implemented on the basis of specially designed for this task, three types of matrix VLSI: N-HM-036.2; N-GM-011; N-GM-012 integration 13000 transistors on the chip. At first VLSI correlator implemented on the basis of the second reference generator functions and the arithmetic unit, the third VLSI is used as a cyclic shift register and shift registers. Memory blocks are executed on the memory chips 565 B RU 5. All chips in die spook, mounted on the cell structurally collected into blocks. The blocks are made in the on-Board fan cooled. The proposed device is controlled by a universal machine type IBM PC and has the following specifications: input stream 10 Mbit/s speed 10 billion op/c, dimensions: 23 x 30 x 35 cm weight 35 kg, power consumption 200 W, the azimuthal resolution up to 1 m The device for calculating the azimuthal correlation function that contains the control unit, the host of the two-dimensional memory, the reference features and the compute node point correlation functions, wherein the node of the two-dimensional memory consists of a group of M+N memory blocks (where M is the azimuthal number of the sequences of samples that form a two-dimensional grid of samples group adjacent axis range broken lines, with the bend at the same azimuthal coordinate axis, N is the number of shifts on the reference axis range group azimuthal sequence of samples corresponding to the change in distance when changing azimuthal coordinates in the matrix results) and block cyclic shift register, reference functions consists of a matrix (g, h)-x generators reference foilow coordinate, h= 1,H, H the number of generators of the support functions necessary to calculate the reference functions in the formation of results having the same coordinate in range) and the matrix (g, h)-h groups of the arithmetic units, the compute node point correlation function consists of a matrix (i, j)-th correlators (where i=1,M, j=1,A, A number of adjacent axis azimuth of the results generated using the same group of neighboring axis range of the azimuthal sequences of samples) and group N+1 blocks of shift registers, the data inputs of memory blocks of the group connected to the information input device, the first and second address outputs of the control unit are connected with the same address inputs of memory blocks of the group, selects the addresses of the memory blocks of the group connected to the group of outputs of the address selection control unit, respectively, the outputs of the memory blocks of the group are connected with the corresponding information input unit cyclic shift registers, the input offset control which is connected to the output of the offset control of the control unit, the output of each (g, h)-th generator of the support functions of the matrix connected to inputs l-x arithmetic blocks (where l=1,L, L the number of support functions, wycislo the f= 1,M+N) of each unit shift register group is connected with f-m output block cyclic shift registers host a two-dimensional memory, the control inputs of the shift block shift register group are connected with a group of control outputs of the shift control unit, the i-th output of the a-th block shift registers (where a=1,N) connected to the first input (i, b)-x correlators matrix (where b=(a-1) K+1,a K, K=A/(N+1), rounded in the direction of a greater whole), the output of the l-th arithmetic unit (g, h)-th group in the matrix is connected to the second inputs (c, d)-x correlators (where c=(g-1) F+1,g F, F number of adjacent axis range correlators receiving the same values of the support function, d=(h-1) L,h L) matrix, the outputs of the correlators of the matrix connected to the output device. }

**Same patents:**

FIELD: specialized information extracting means.

SUBSTANCE: device has displacement registers, comparator block, XOR element, multiplexer, triggers, counter, AND elements.

EFFECT: simplified construction.

1 dwg

FIELD: computer science.

SUBSTANCE: device has first and second regenerators of random evenly spaced signals, second and first comparators, generator of short pulses, second and first binary counters, decoder, D-trigger, first and second RS-triggers, AND element, XOR element, reverse counter, clock pulse generator, divider with rebuilt division coefficient , pulse distributor, group of M synchronization blocks, group of M pulse distributors, N (M-10) - input elements R and N blocks for calculating ordinates of correlation function.

EFFECT: simplified construction and higher reliability .

2 cl, 4 dwg, 1 tbl

FIELD: the invention refers to the technique of detection of a target and determination of the direction at a target.

SUBSTANCE: the mode is realized by way of receiving of ultra wideband impulses reflected from the target, of delaying them on various time multitude in various channels of surveillance and multi channel processing. In the first variant of the current mode variation of the form of receiving impulses on a great number of discrete time positions are carried out by way of averaging-out by channels of surveillance at known direction of incoming reflected impulses in a beforehand designed control sector and then found valuation of the form of receiving impulse is used as a base signal in multi channel correlation processing. In the second variant valuation of magnitude of receiving impulse is formed in concrete moment of time for each base direction in beforehand given angular sector of control, valuation of the form of the receiving impulse is found according to formed valuations of magnitude 0f the receiving signal for various discrete moments of time; found valuation of receiving impulse is used as a base signal in multi channel correlation processing; out of multitude of results of correlation processing correlation maximum is chosen. This maximum is used as preliminary threshold decision statistics in the procedure of detection of the target; the direction of incoming reflected impulses with the help of interpolating valuation of the position of the correlation maximum in the environs of that base direction for which the largest result of multi channel correlation processing.

EFFECT: the use of this invention at location of a target with the help of ultra wideband impulses allows to receive signals incoming not only from in advance chosen base directions.

6 cl,9 dwg

FIELD: data processing in broadband radio communications and radio navigation.

SUBSTANCE: proposed method intended for use where reception of extended-spectrum data signals keyed by simulation-resistant pseudorandom nonlinear derivative sequences is always preceded by synchronization includes concurrent accumulation of periodic mutually correlated function values of signal segments arriving from output of dynamically matched adjustable filters with two standard sampling lines affording generation of random derivative, as well as determination of time step numbers of their mutual shift corresponding to delay synchronism. Then current delay of entire signal being received is found from combination of these time step numbers. Used as dynamically matched adjustable filters in search channels are acousto-electronic convolvers.

EFFECT: reduced time and hardware requirement for searching broadband delay signals characterized in high simulation resistance.

2 cl, 9 dwg

FIELD: engineering of specialized devices, meant for determining correlation functions of random processes.

SUBSTANCE: correlator consists of circular distributor (1), containing generator of clock impulses and serially coupled triggers (2), two limiter amplifiers (3), key circuits (4), logical elements "AND" (5,9), "OR" (6) and "NOT" (7), memorizing element (8) and impulse counters (10). In circular distributor (1) between "n" and "1" triggers (2), a "n+1" trigger (2) is additionally coupled, which makes it possible to introduce additional preparation operation for memorizing element (8), in other words, its nullification. Second input of "n+1" trigger (2) by preparation signal is connected to output of "n"-trigger (2), and its first output is connected to controlling input of additional "n+1" key circuit (4), input of which is connected to output of memorizing element (8), and output - to ground connection. Input clamp by second input signal through second limiter amplifier (3) is connected to second input contact of mode switch (12), its first input contact is connected to input of first key circuit (4) and to output of first limiter amplifier (3), input of which is connected to input clamp of first input signal; output contact of mode switch (12) is connected to second input of "AND" element (9).

EFFECT: increased precision of multi-channel polar correlator, expanded functional capabilities of correlator due to introduction of mode for determining mutual correlation function of two signals.

1 dwg

FIELD: navigation techniques.

SUBSTANCE: when forming an image of a surface, on which a sensor is moving, components of the image with spatial wavelength less than twice the size of the optical cell are destroyed, before calculation of shift between the reference and the current frames on the discrete set of numbers Arefij and Acurij, stored in the first and second memory buffers. Initial continuous functions Aref(x,y), Acur(x,y) are restored, and then the shift between the frames is calculated by comparing the restored continuous functions.

EFFECT: increased accuracy and resolution capacity of calculating shift.

2 cl, 10 dwg

FIELD: physics; radio-technology.

SUBSTANCE: invention pertains to radio-technology, and particularly to optimum receiving of pseudonoise signals. The technical outcome is the increased resistance to interference of the output signal. According to the method, the radio frequency oscillation is converted to the video frequency range. The signal envelope is separated, sampled on time and amplitude on two levels, "1" and "0". The obtained code is recorded in an n-bit shift register of a discrete matched filter, matched by a direct code, and a discrete matched filter, matched by an inverse code. The automatic correlation function of the received signal is generated by removing the constant component from the result of adding output signals of the indicated discrete matched filters. The device which implements the method consists of a multiplier (1), low-pass filter (2), bidirectional limiter (3), cascade for coinciding with "1" (4), cascade for coinciding with "0" (8), inverter (7), n-bit shift registers (5,9), n-input adders (6,10), dual input adder (11), device for removing constant component (12), polling clock pulse source (T). At the first n-bit shift register and the first n-input adder there is discrete filter, matched by a direct signal code, and at the second n-bit shift register and the second n-input adder there is a discrete filter, matched by an inverse signal code.

EFFECT: increased resistance to interference of the output signal.

2 cl, 4 dwg

FIELD: electricity.

SUBSTANCE: invention can be used for detection of complex signals in those radio-technical systems which are not capable to change signal phase quickly. Method for detection of Barker signal in matched filter which signal is modified by substitution "-1" elements for "0" elements consists in following. Received signal is filtered in auxiliary filter matched relatively to one signal position, delayed in time-delay circuit where delay between it's branches is equal to duration of one signal position. Corresponding elements of received signal code and elements of reference code are added one-by-one by modulo 2. Results of additions are inverted, summed, and in accordance with summing results the decision about detection is made.

EFFECT: possibility to process complex signals with constant phase.

2 dwg

FIELD: education.

SUBSTANCE: method consists in the following: placing on the monitor of a control question and versions of the answer to the given question, a choice of one, preferable trained, a version of the answer by means of moving of the manipulator cursor to the location of its indicator and the subsequent definition of correctness of the answer on a final cursor position; after placing of indicators of versions of answers in the range from the moment of the cursor movement beginning till the moment of fixing of its final position form a file trajectory parametres of its movement on the monitor, determine autocorrelation function of the obtained array of trajectory parametres and determine level of confidence of the trained person in the obtained knowledge by position of a maximum of function at which arrangement within the set error of decision-making in the beginning of co-ordinates gives the conclusion about the confident or uncertain motivation of the answer.

EFFECT: increase of reliability and accuracy of the control of obtained knowledge level.

5 dwg

FIELD: physics; computer facilities.

SUBSTANCE: invention concerns computer facilities and can be used for algorithmic diagnosing and compensation of failure of process of automatic support of the object by the television tracer of correlative type. For this purpose the known television tracer of correlative type is equipped with the analyser of failure of process of support and the additional commutator.

EFFECT: provision of fault-free object following at occurrence of sudden failures in the correlator at the expense of identification of failure of tracing by the correlator and automatic transmission to the given situation of priority of a number system of coordinates.

1 dwg