SU1688185A1 - Adaptive analizer of spectrum with linear prediction - Google Patents

Abstract

The invention relates to a radio metering technique and can be used to analyze spectral characteristics under conditions of small a priori information about the class of random processes under study. The purpose of the invention is to improve the accuracy of spectrum measurements. It is achieved by introducing the coefficient 6 into the device, switching unit 7, dividing unit 8 and forming new functional connections. In addition, the device contains an analog-to-digital converter 1. a lattice adaptive filter 2, a dispersion meter 3, a Fourier transducer 4, a micro-command formation unit 5. 1 hp f-ly, 2 ill.

Description

The invention relates to a radio metering technique and can be used to analyze spectral characteristics under conditions of small a priori information about the class of random processes under study.

The aim of the invention is to improve the accuracy of the spectrum measurement.

FIG. 1 shows a block diagram of a spectrum analyzer: FIG. 2 is a block diagram of a lattice adaptive filter.

The adaptive spectrum analyzer (Fig. 1) contains an input analog-to-digital converter 1. a trellis adaptive filter 2, a dispersion meter 3, a Fourier transducer 4, a micro-command generation unit 5, a coefficient 6 register, a switching unit 7, a division unit 8. The outputs of the analog-to-digital converter 1 and the coefficient b register are connected to the first

and the second inputs of the switching unit 7, the output of which is connected to the input of the trellis filter 2. The first output of the filter 2 is connected to the input of the dispersion meter 3, and the second to the input of the Fourier transducer 4. The outputs of the dispersion meter 3 and the Fourier transducer 4 are connected to the block inputs 8, the output of which is the output of the spectrum analyzer. The first output of the micro-command generating unit 5 is connected to the control input of the analog-digital converter 1. The second output of the block 5 is connected to the control inputs of the switching unit 7 and the coefficient register 6. The third, fourth, fifth, sixth and seventh outputs of block 5 are connected to the address input, the first, second, third and fourth control inputs of the grating filter 2. The eighth, ninth and tenth outputs of block 5 are connected respectively to the control

About 00 00

00 SP

to the inputs of the dispersion meter 3, the Fourier transducer 4, and the dividing unit 8.

Lattice adaptive filter 2 (Fig. 2) contains M identical series-connected links 9-1-9-M. Each link contains blocks 10-12 of delay, an element of OR 13, blocks 14-17 of multiplication, adders 18-21, a scale amplifier 22. The input of the adder 18, the first inputs of blocks 14 and 15 of multiplication are connected and are the first information input of the link, the second information the input of which are the combined inputs of the delay blocks 10 and 12. The first and second information inputs of the link 9-1 are combined and are the information input of the lattice filter 2. The outputs of the delay blocks JO and 12 are connected to the first inputs of the adder 19 of the multiplication blocks 16 and 17 through the OR 13 element. The outputs of multipliers 15 and 16 are connected to the first and second inputs of the adder 20, the output of which is connected to the first input of the adder 21 through the scale amplifier 22, the output of the latter is connected to the second inputs of the multiplication blocks 14 and 17 and the delay block 11.

The output of the delay unit 11 is connected to the second input of the adder 21. The output of the multiplication unit 17 is connected to the second input of the adder 18, the output of which is connected to the second input of the multiplication unit 15 and is the first output of the link 9. The output of the unit

The multiplier 14 is connected to the second input of the adder 19, the output of which is connected to the second input of the multiplication unit 15 and is the second output of the link 9. The first output of the 9-M link connected to the input of the quad 23 is the second output of the trellis filter. The second output of the 9-M link is connected to the input of the quadrant 24. The outputs of the quadrants 23 and 24 are connected to the inputs of the adder 25, the output of which is the first output of the grating adaptive filter 2. The address input of the filter 2 is connected in each link to the control inputs of the blocks 14 and 17 multiply. The first control input of the lattice filter 2 is connected to the control inputs of the blocks

15 and 16 multiply. The second, third and fourth control inputs of the lattice filter 2 are connected respectively to the control inputs of the delay units 11-12.

The analyzer performs the measurement in three steps: the calculation of the partial correlation and dispersion coefficients, the calculation of the linear prediction coefficients. calculation of the square of the amplitude-frequency response of the lattice filter 2 and the spectrum.

Adaptive spectrum analyzer with linear prediction works as follows.

In the initial state, in the glint 5 of the formation of microinstructions (Fig. 1), the initial conditions are established. The coefficient is also preliminarily recorded in register b.

So - 1, and in blocks of multiplication 14-17, blocks 10, 12 of the delay of the adaptive lattice filter 2 (FIG. 2) are written zeros.

At the stage of spectrum measurement, the signal under study is fed to the analyzer input.

x (t), and in block 5 of the formation of micro-commands, at the first output, a clock signal is generated, which is fed to the control input of the analog-digital converter 1. The received XL sample from the output of the converter 1 is fed to the first input of the switching unit 7, Simultaneously to the control input the latter is fed a zero potential C1 from the second output of block 5, allowing the transfer of the reference XL to the output

block 7 and further to the information input of the lattice adaptive filter 2. In filter 2, XL is fed to the first link 9-1 to the first adder 18 and the first block 10 of the delay, as well as to the first inputs of the multiplication blocks 14 and 15.

In accordance with the address А 1, coming from the third output of the block 5 of the formation of micro-commands, at the first and second outputs of the first link 9-1 are formed

bc and ri.L errors predict back and forth by the ratio

(bi, i bo.L qi i TO, v

4 Cn, L r0. L-1 - QU br.L

Bm, l prediction errors. The rm, L, m 2, M links of 3-2-9-M are formed similarly to the corresponding addresses A 2, 3M.

At the same time, the partial correlation coefficient qm.L is calculated equally in all links 9 m (m - 1 -M) in the second, third currents 15, 16 multiplication, the third, fourth adders 20, 21, the scale multiplier

22 and the second block 11 of the delay in the ratio

Qm.C CJm, L-1 + / (rm.L-1 bm-I.L-1 + + bm. L-rrm-1, L-2)

when sequentially coming from the outputs of the micro-command formation unit 5 to the control inputs of the lattice filter 2

impulses C2, SZ, C4. In practice, / is selected from 0.1 to 1.

Bm.L. prediction errors rm.L from the outputs of the last link the 9-M grating of the filter is squared and summed in quadrants 23, 24 and the fifth adder 25. From the first output of filter 2, the error L is (Lm. and + rm,) input to measure Ate 3 variances, which implements the function Aai-12 + eat L2. At stage II, using the partial correction coefficients qm, i, m - 1, M, the linear prediction coefficient Am.i, m1, M is calculated.

The linear prediction coefficients for the trellis filter 2 are the counts of its impulse response with the 9-T values of Hg, T 1, M. That is, at stage II, the values of the coefficient qm, L do not change. Also, in each delay block 10, the error value rms is saved for the previous sample xi, and the delay block 12 controlled by C5 pulses from block 5 is used to operate filter 2 at the 2nd stage.

The calculation of the impulse response of the lattice filter 2 is carried out using a single pulse C1, coming from the second output of the micro-command generating unit 5 to the control inputs of the coefficient register 6 and the switching unit 7

From the output of the register b, the code of the coefficient So is fed to the second input of block 7 and from the output of the latter to the input B of the lattice filter 2 in which, in accordance with the addresses A 1, 2 ... M. similarly to the work at the first stage, the values are calculated at the first and second outputs of the links 9-1-9-M. Unlike the first stage, the control pulses C2, C3, C4 are not output by the block 5, which allows not changing at the second stage values qm, l. GTD-1, t 1, M, In total, from output 2 of block 5 in the second stage, the inputs of register 6 and block 7 receive (M + 1) the controlling unit pulse C1. The input of filter 2 from the output of block 7 is received in series () coefficients, the first being equal to So 1, and the remaining M being equal to O .. At the output of the lattice filter 2, the value of the linear prediction coefficients Am, L, m 0 is formed M, where AO, L 1, which are fed to the input of the Fourier transducer 4 and recorded into it in terms of pulses from the eighth output of block 5 to the control input of the transducer 4.

The third stage of spectrum calculation starts at the moment when the formation of the (M + 1) -th AM.L. In this case, the formation of codes A and control pulses C1 and C5 of the Fourier transducer 4, which is rtbote him, for example, according to the well-known subroutine fast Fourier transform (FFT) or hardware-oriented principle (similar to the serial Fourier transformer X6-8), performs the calculation p yes squares of Fourier coefficients

are the square of the amplitude-h with the total characteristics of the M-ro lattice filter of the order of KL2 (f).

From the output of the Fourier transducer, 4 KL (f) values are fed to the first input

block 8 division, the second input of which

the value of the dispersion ОЈ comes from the output of the meter 3. By the pulses coming from the 9 output of block 5 to the control input

block 8. is the calculation of the estimate 2

spectrum by the ratio Gift) ot / KL (f)

After the arrival of the next reference XL + I, the calculation of the spectrum starts from stage I

Claims (2)

1. An adaptive spectrum analyzer with linear prediction, containing a series-connected lattice adaptive filter, a dispersion meter, The second input of which is connected to the first output of the micro-command formation unit, the second output of which is connected to the control input of the analog-digital converter, and the third, fourth, fifth, sixth and seventh outputs to the first, second, third, fourth and fifth entrances of the lattice adaptive filter, the eighth output - with the first input of the Fourier transducer, the second input of which is connected to the second output of the trellis filter, the information input of the analog to digital converter is connected to the input of the device, characterized in that to increase spectrum accuracy, factor register, switching unit and dividing unit are entered in hero, while the output of the analog-digital converter is connected to the first input of the switching unit, with the second the input of which is connected to the output of the coefficient register, the output of the switching unit is connected to the sixth input of the lattice adaptive filter, the outputs of the dispersion meter and the Fourier transducer are connected respectively to the first and second inputs of the division unit, the third input of the switching unit and the input of the command register are combined and connected to the second output of the forming unit microinstructions, the tenth output of which is connected to the third input dividing unit, and its output is connected to the output device. 2. The analyzer according to claim 1, characterized in that the lattice adaptive filter is made in the form of M identical consecutively connected links, first and second quadrants and an adder, the inputs connected to the outputs of quadrants, and the output to the first output of the filter, each The M link consists of the first, second and third delay blocks, the OR element, the first, second, third and fourth adders, a scale amplifier; the first input of the grating adaptive filter in each link is connected to the control inputs of the first and the fourth multiplication unit, the output of the fourth multiplication unit through the first adder connected in series, the third multiplication unit, the third adder, scale amplifier and the fourth adder are connected to the second inputs of the first and fourth multiplication units and the first input of the third delay unit, the second input of the lattice adaptive filter in each link connect to the first input of the second delay unit, the output of which through the OR element is connected respectively to the first input of the second adder, the second input of the third the multiplier unit and the third flow path of the fourth multiplication unit, the third input of the lattice adaptive filter in each link is connected to the third input of the third multiplication unit and the first input of the second multiplication unit, the output of the latter is connected to the second input of the second adder, the fourth input of the grating adaptive filter in each link through the third the delay unit is connected to the second input / Ω of the fourth adder, the fifth input of the lattice adaptive filter in each link through the first delay unit is connected to the second input of the OR element, the output The second adder is connected to the second input of the second multiplication unit, the sixth input of the lattice adaptive filter in each link is connected to the second inputs of the first adder, the first and second delay blocks and to the third inputs of the first and second multiplication blocks, the first and second outputs of each link are connected to the outputs of the first and second adders, the first output of the M-th link is connected to the second output of the lattice adaptive filter and to the input of the first quad, the second output of the M-ro link is connected to the input of the second quad a, and the second output of the device is connected to the input of the first quad. "M go : uh Q