DE2654135C1 - Circuit arrangement for determining the modulation of input pulses - Google Patents

Description

The invention relates to a circuit arrangement for determining the modulation of input pulses. in particular of radar pulses, by means of a monitoring receiver using a filter bank with M bandpass filters connected in parallel with respect to the input signal, the rectified output signals of which are transmitted in M channels.

If received pulses, in particular radar pulses, are analyzed with a filter bank of bandpass filters arranged next to one another in terms of frequency, the respective output signals of the filters give time responses that depend on the shape and bandwidth of the signal spectrum. With the help of detectors at the exits the filter can roughly determine the frequency position and the effective bandwidth to a certain extent will.

A more detailed analysis of the composition of the spectrum of the received pulse (s) is available with known means possible; However, the effort required for this in the construction of such spectrum analyzers is very high. The invention is based on the object of showing a simple way which makes it possible to obtain the most important spectral information about the pulses to be analyzed, in order to draw conclusions about the modulation or the composition of the received signal can. According to the invention, which relates to a circuit arrangement of the type mentioned at the beginning.

this is achieved in that the duration of an input pulse in a first time measuring device is determined and this time period is shown as the first reference value that in each channel a first and a second delay device is provided, which are connected in parallel on the input side, that for each Channel the delay time of the first together with the delay time of the second delay device gives the same value that the respective first delay devices one from channel to channel 5 | f continuously increasing by the same value, the respective second delay devices one of the channels

{§ to channel continuous at the same value decreasing delay time that the outputs of all

H first delay devices are connected together and deliver a first comparison pulse whose

H duration is measured and represented as a second reference variable that the outputs of all second delay inputs

p directions are interconnected and deliver a second comparison pulse, the duration of which is measured and

fj; is represented as the third reference variable, and that from the three measured reference variables the number of through

II signals occupied the bandpass filter conclusions on a possible frequency and / or phase modulation of the

-K input pulses are drawn.

| ö In this way, the only way to analyze the spectral composition of the input pulses is practically

ί§- application of two delay arrangements that are complementary to their delay time required and

i | a subsequent time measurement. The information obtained in this way about the occupancy of the individual channels and

| i over the duration of the different comparison pulses makes it possible to make relatively precise statements about the spectra

vi? 3 to make the impulses to be assessed. The solution is particularly beneficial when thrown over The filter bank is available anyway for other reasons

If this prerequisite does not apply, then the solution principle on which the invention is based can advantageously

can also be applied in a hard way by adding two more signal paths to the direct signal transmission path.

Viji transmission paths are connected in parallel, so that in each of the further signal transmission paths, frequency-dispersive

Sive delay devices are provided, the frequency-time functions of which are complementary to one another run that rectifiers are provided in each of the three signal transmission paths that for each of the

■ '-. -

I; three signal transmission paths a measurement of the pulse duration is carried out in three timing circuits,

frv where in the two other signal transmission paths this measurement at the output of the delay device

;; i processing is carried out and the three time values obtained in this way are displayed for further evaluation.

ψ In this case, two oppositely dispersed

S; ' ve, each complementary delay device, z. B. in the form of dispersive high-frequency lines.

K The invention and its developments are explained in more detail below with reference to drawings

% shows

: 4 F i g. 1 in a block diagram the structure of the circuit arrangement according to the invention with the filter bank

: forming bandpass filters,

F i g. 2 shows a circuit arrangement with the first row of delay devices,

F i g. 3 shows a circuit arrangement with the second row of delay devices,

~ ^: F i g. 4 shows the course of the delay time as a function of the frequency for the two arrangements

; ■ ·. ■ Fig. 2 and Fig. 3,

F i g. 5 a schematic representation of the display device for the evaluation,

F i g. * a modification of the invention in which only two dispersive delay devices are provided are.

In Fig. 1, the input signal x (t) This input signal a filter bank is supplied, consisting of M frequency-lined up band filters BFi to BMF is the pass band width of these filters BFi to BFM is ^ / "refers to the respective center frequency is indicated as fm i to FMM 1 to GMm the respective channels K 1 to KM provided. at the output of band filter 5FL to BEMsmd rectifier G. by this rectifier G 1 to GM the envelope of the high frequency time response is formed, the thus obtained signals of the different channels K 1 to KM to be in a Summing stage ASO summarized This produces an output pulse whose time function is denoted by B 0 (t) . The duration TO of this pulse is determined in a time measuring circuit IDO , for example by starting a counter with the rising edge, the count of which is stopped when the falling edge occurs The time value TO of the output pulse B0 (t ) is used as the first reference variable for further signal evaluation in FIG. 5 supplied to the analyzer shown and shown there for the evaluation.

It is also possible, instead of or in addition to the measurement in IDO , to carry out a measurement of the duration of the input pulses in the high-frequency position, i.e. before the band filters 5F1 to BFM. This is shown by the time measuring circuit / Of, the associated time period being denoted by τ

In the individual channels K 1 to KM sensors SS 1 to SSM are provided, which are located between the band filters ÖF1 and the rectifiers G 1 to GM . These sensors determine which of the channels K 1 to KM is occupied by signals Mark occupied by the useful signal. In addition, the respective amplitude value itself, which has been reached in the occupied channel, can also be scanned and made available for the evaluation. At F i g. 5, this information about the respectively occupied channel in the analyzer is denoted by k.

In each of the channels K 1 to KM , branching points V1 to VM are provided after the corresponding rectifier G 1 to GM. The two outputs of these branch points are marked VIl, V12; V21; V22; ... VMl to VM2 denotes The junction VIl, V21, ... VMl is connected to the arrangement J, welcb? in Fig. 2 is shown, while at the branches V12, V22, ... VM 2 the arrangement II according to FIG. 3 is switched on.
The arrangement according to FIG. 2 is a bank of deceleration devices VFIl bis

VEM I ₁ which are inserted into the respective transmission channels Af 1 to KM with different delay times. The delay device for the channel K 1, that is to say at the connection VX 1, has the delay time 0. The delay device for the other transmission channels up to the transmission channel M are each stepped, namely integer multiples of ΔΤ. Accordingly, the arrangement I results in the time shift JT for the channel K 2 and the delay time (M- 1) · ΔΤ for the channel M. The signals thus obtained at the output of the delay devices VEH to VEM 1 are combined in an addition circuit AS I and result in an output signal of the form B \ (t). The time duration TX is determined from this output signal in the time measuring circuit IDX (second reference value) and the second value of the pulses B \ (t) obtained in this way is also used for the analysis of the spectral composition in the analyzer according to FIG. 5

ίο shown. The time measurement can be carried out analogously to the evaluation circuit ID 0 according to FIG. 1 take place.

3 shows the arrangement II with the delay devices VE \ 2 to VEM2, which are each to be connected to the outputs V12 to VM2 of the channels KX to KM according to FIG. The delay time of successive stages is in turn different by the time value ΔΤ. However, the delay time is complementary to the delay time of the arrangement I according to FIG. 2 elected. This means that the Channel K \ connected delay device VE 12 the delay time (M- 1) · ^ That, while the delay time for the delay device VEM2 is selected to be zero

The output signals obtained in this way due to the different delays of the individual signals from the various channels K 1 to KM ?!? the delay device! VE ! 2 to VEM2 wsrdsü Id of one addition stage -452 and result in an output pulse which is designated B2 (t). the The duration of this pulse is in turn determined in a timing circuit ID2 and is referred to below as 7 * 2 (third reference value). The course of the delay times in the two arrangements I (according to FIG. 2) and II (according to FIG. 3) is complementary, because the running times of the two delay devices connected in parallel of the same channel are independent of the number of the respective channel to (MX) ■ ^ complete.

For clarification, reference is also made to FIG. 4 pointed out, where the respective transit time and that normalized to multiples of z / Tauf are plotted on the ordinate. The number of the respective band filter BFX to BFM is shown on the abscissa. The stepped line beginning at zero from the left and rising to the right shows the course of the delay as it occurs from the arrangement I to FIG. 2 is achieved. The stepped line falling from the left shows the course of the delay unit as a function of the frequency or the filter more for the arrangement according to Fig. 3.

It is useful if the delay time JT, which is the multiple of the delay times for the individual delay devices VE according to FIG. 2 and 3, according to which relationship is chosen

^ = If

The spectral sensitivity of the arrangement constructed in this way decreases with increasing steepness of the transit time. It could therefore make sense to provide several devices of the form shown in Figure 1 with different delay times in order to describe unknown signals even more precisely.

As when looking at FIG. 4 as is readily apparent, another solution to the principle according to the invention can be achieved in that instead of individual delay devices separated by channels a total of only two complementary, frequency-dispersive and thus inversely frequency-dependent delay devices in their running time are used. This corresponds to the two that delimit the stair lines Straight line according to FIG. 4. Fig. 6 shows the basic structure of this circuit arrangement. About a first Channel K 0 is used to determine the duration of the first reference variable Γ0 or r from the input pulse x (t) in the timing circuit ID 0. In the second channel KA connected in parallel to this, there is a dispersive delay device VEA, the delay time of which increases with the frequency (corresponding, for example, to the line rising to the right in FIG. 4). After the rectifier GRA , the time TX is determined as a second reference variable in the time measuring circuit ID 1. In the parallel channel KB , the complementary channel is used in the delay device VEB Generates delay means for delaying VEA and after rectification in GRB in the Zeitmeßschaitung ID 2 measured the pulse duration and provided as the third reference value T2, the delay means VEB has, for the course. B. the line sloping to the right in FIG. 4. The time values TO, Ti and T2 of this circuit arrangement thus correspond to the time values of FIGS. 1, 2 and 3, which are also designated, so that the representation and evaluation can also be performed with the arrangement according to FIG. 5 can be done. If necessary, you can do this in parallel for the channels K 0, KA and KB by means of the filter bank BF 1 to BFM analogously to FIG. 1 the occupied channels K 1 to KM are determined and via the sensors SS 1 to SSM for display and evaluation in FIG. 5 can be provided.

If the dispersive delay devices VEA and VES according to FIG. 6 are frequency-dependent have changing damping characteristics, appropriate compensation must be used in a known manner. Circuits are provided which make it possible to compensate for this frequency-dependent attenuation curve if the errors that occur as a result would be too large. Delay lines with the desired complementary frequency response can preferably be used for the dispersive delay devices VEA and VES. The analysis of the pulse durations or TO, Ti and TZ measured at the respective timing devices ie

the functions B0 (t) as the first reference quantity, the function B i (t) as the second and B2 (t) as the third reference quantity can be carried out according to the following considerations:

The broader the signal spectrum of the received pulse BO (t) , the more bandpass filters deliver an output signal. Accordingly, the greater is the number k that is occupied by the sensors SS1 to SSM

marked channels. The broader the signal spectrum, the more the two reference quantities B \ (t) and B2 (t) \ on of the reference quantity B0 (t) differ, because the delay elements make the dispersion more and more pronounced generated output pulses become longer and longer.

In detail, the following procedure is used to describe the properties of unknown received pulses: At the input of the circuit, ie before the band filters BF 1 to BFM , the duration r of the incoming pulse is measured in a pulse duration measuring stage IDE. Then, in the further course of the analysis of jcü, '.. n pulse, the time TO (in ID 1), Ti (in ID 1) and T2 (in / Z? 2) is determined. In addition, during the time in which a reference signal B0 (t) is present, the individual detector displays of the individual filter outputs are stored via the sensors 551 to SSM and made available for later processing. The data obtained in this way allow, overall, a good description of the modulation of a received pulse if the envelopes of the time responses of the bandpass filters in the filter bank are appropriately designed for the received pulses to be analyzed.

Some characteristic cases are considered below in order to illustrate the effects on the comparison signals B \ (t) and B 2 (t) . They have been mathematically tested for a filter bank of grade 10 Bessel filters. The following definition applies:

Af = 3dB bandwidth of the bandpass filters BFX to BFM according to FIG. 1

k = the number of responding sensors lying next to each other at the respective filter outputs BF I to BFM.

20 Example 1

Unmodulated carrier-frequency pulse of duration r

a) "narrow" range of services, Af> -

r «TO a Ti - T2

b) "broad" range of services, Af < -

ί-4-Ι

ro < Ti = T2 ^ jf

The reference signals B 1, 2 (t) are identical except for a time shift.

Example 2
Carrier-frequency pulse of duration r with phase coding (TV sub-pulses of duration ro)

The power spectrum is similar to that of the subpulse and with a suitable choice of Af generally applies Af S ^ -

O-r + ^

Ti «Γ2« γ ^

The reference signals B i £ (t) are approximately the same apart from the time shift and a fine structure resulting from the coding. This means that phase-coded pulses can also be recognized from the time values.

Example 3
Frequency-modulated pulse (chirp pulse) of duration r with linear FM (frequency deviation A)

With a usual "signal dispersion" (frequency deviation A χ pulse duration τ) > 50, A is equal to the effective width of the power spectrum. In the following, Af <A is assumed, which only applies to "high-resolution" chirp pulses

it filters respond (kAf ^ A)

(The characteristic is the linear graduation ffAO = start of the initiation pulse, f £ 0 = end of the pulse) of the response times inM0 <f <f £ 0): r « TO

The reference signals B l, 2ffj can take two different forms:

a) 1 pulse with Tl, 2 »r

b) k pulses with 7 "1,2 <r each and linear time graduation.

The reference signals BX (t) and B2 (t) differ fundamentally! (In contrast to example 2), 3 cases are possible

a) BX (t) and B2 (t) have the form a), but TX <T2 or TX> T2 applies. (In addition, the amplitudes ίο of the signals are fundamentally different);

ß) BX (t) is of the form a), B2 (t) of the form b) or vice versa.

B Iff) and B2 (t) have the form b), but differ in terms of the different sizes of the time graduation. In addition, Γ1> T2orri <Γ2 applies to all k single impulses.

Further examples with frequency-keyed pulses or chirp pulses with non-linear FM show nothing fundamentally new. The first case is similar to Example 2 with the difference that TX and T2 differ if the difference in the carrier frequencies before and after a sampling time is in the order of magnitude of Jf . The second case can be explained directly with example 3) if one takes into account that a non-linear FM has the effect of a non-linear time graduation of the filter responses. The reference signals can then also represent mixed forms of 3a) and 3b).

The previous considerations can be used to set up a scheme in order to draw conclusions about unknown radar pulses from the measured temporal variables and then responding detectors. It seems particularly important that an easy decision can be made as to whether a radar pulse of relative long duration

Af

carries an additional modulation or not. A comparison of example 1 with the other examples shows: only Signals without pulse compression provide for large r:

r «TO =» TX = T2

The method described in the previous section leads to problems if several radar pulses arrive either simultaneously or with a time difference smaller than approximately (M + 2) z /

I AT » - provided].

If pulses are received at the same time or if they overlap in time, there is a real separation in the evaluation with the arrangement according to FIG. 1 is not easily possible if the spectra of the signals overlap. That at least one detection of two impulses of this type can be carried out is evident from the others Considerations clear.

Before the signals hit the filter bank, a preliminary decision can be made whether there are temporally separated radar pulses by means of suitable measurements. A by-product of this measurement is the determination of the pulse duration r of the undistorted high-frequency input signal in the time measuring device IDH.

Only a special comparison of the signals Bi (t) \ i = 0,1,2 enables the temporal separation of pulses by separating the information belonging to different pulses in the Bi (t). To do this, it is necessary to consider the times M / and tEi (i - 0,1,2) in connection with the recorded displays of the detectors at the filter outputs. The evaluation method is to be explained using an example. A narrow band pulse

hits the filter bank and runs through exactly one filter with the number k (1 <k <M). The envelope curve of the time response of the filter k then simultaneously represents the sum signal £ Oft). The reference signals BX (t) \ md BX (t) do not differ in form from B Oft), since the dispersion of the delay lines traversed for

_M ΔΙ> - is still negligible. Due to the complementarity of the running times in orders I and II, according to Fig. 2 and 3

(1) (tA2- e> 5 Since the detector with the number k responds, the relationships also apply

U) (k - \) AT <lAl - iAO <kAT 1 \ _{<k <M}

(3) (M -k) AT <tA2 - tAO <(M -k + \) AT} -kM

f (M-2k) AT <tA2 - tA \ <(M-2k + 2) AT \ <k <- ~

[(-M + 2k-2) AT <tA2 - lA 1 < (2k -M) AT -ψ <k <M

what one of the F i g. 4 can be easily removed. All four conditions also apply in the event that you replace the starting times tA through the corresponding end times tE. The complementarity of the delay lines thus enables an exact assignment of the different time scales for the Bi (t).

For signals with a larger bandwidth, but which still satisfy Δ (> - and pass through at most one filter,

the dispersion of the delay lines comes into play and condition (1) is no longer met exactly. Instead, the following applies:

(5) ΚΊΔΤ '< (iA 2 - iA 0) + (iA i - iA G) < (M + i) JT

or an analogous inequality for the end times tE.

All previous considerations can also be extended to broadband signals that pass through several filters lying next to one another. Conditions for tAi and tEi (i = 0, \ 2) result from the numbers of the appropriate detectors in the filter bank and the complementary transit time characteristics of the delay lines. which enable a clear assignment of the Bi (t) to a single received pulse. An extension of this idea is the possibility of reading from the continuously recorded signals fi / ftf whether there are different impulses that are either temporally or spectrally separated. Violations of complementarity basically indicate different impulses, a precise analysis of the Bi (t) enables the separation of the associated ·. «! Information. With a corresponding accuracy of the time measurements, it is thus possible to recognize the case that two different pulses are present which overlap in time and in their spectrum. ₃₀ I.

With the aid of the arrangement described, it is thus possible to receive spectral information from unknown receivers

capture signals, in particular radar signals, to map information in the time domain and based on this Figure to carry out the evaluation. A continuous display of the sum signals and a Subsequent evaluation in the sense of the above explanations offer the possibility of unknown received pulses to describe in a simple way and to draw conclusions about the signal modulation.

For this purpose 3 sheets of drawings

Claims (7)

Patent claims: 1. Circuit arrangement for determining the modulation of input pulses, in particular radar pulses, by means of a monitoring receiver using a filter bank with M with respect to the Bandpass filters connected in parallel to the input signal, the rectified output signals of which are transmitted in M channels, characterized in that the duration of an input pulse is determined in a first time measuring device (IDE and / or / DO) and this duration (TO or r) is represented as the first reference value, that in each channel (K 1 to KM) a first and a second delay device (z. B. VEi 1, VE 12 for KI) are provided, which are connected in parallel on the input side, that for each Channel (K 1 to KM) the delay time of the first (e.g. VE 11) together with the delay time of the second delay device (e.g. VE 12) results in the same value that the respective first delay devices (VEM ii to VEMi) one from channel to channel continuously increasing by the same value (ΔΤ) , the respective second delay devices (VE 12 to VEM 2) have a delay time continuously decreasing from channel to channel by the same value (ΔΤ) , that the outputs of all first delay devices (VEii to VEMi) are interconnected and deliver a first comparison pulse, the duration (Ti) of which is measured and represented as a second reference value that the outputs of all second delay devices (VE 12 to VEM2) are interconnected and provide a second comparison pulse whose Dausr (T2) is measured and represented as the third reference variable, and that 3US the three measured reference variables (TQ, Ti, T2) and the number of occupied by signals Bandpafifiaer (BFi to BFM) draw conclusions about a possible frequency and / or phase modulation of the Input pulses are drawn. 2. Circuit arrangement according to claim 1, characterized in that the undelayed signals of the M channels are interconnected and the timing circuit (ID 0) is connected for the duration (TO) of the first reference variable after the interconnection point (ASO) 3. Circuit arrangement according to one of the preceding claims, characterized in that before the input of the BandpaßfiKer (BFi to BFM) one for the determination dt; first or an additional reference variable to this serving time measuring circuit (IDE) is provided 4. Circuit arrangement according to one of the preceding claims, characterized in that band-pass filters (BFi to BFM), the passbands of which abut, change in their center frequencies each differentiate a constant value (ΔΓ). 5. Circuit arrangement according to one of the preceding claims, characterized in that the differences in the delay times ΔΤ of successive channels is selected according to the relationship ΔΤ - -jjf, where J / is the bandwidth of a bandpass filter (BFi to BFM) \ st 6. Circuit arrangement according to one of the preceding claims, characterized in that the Channel occupied by a signal (K 1 to KM) is detected by a sensor (SS 1 to SSM) and is also shown for evaluation 7. Circuit arrangement for determining the modulation of input pulses, in particular radar pulses, by means of a monitoring receiver, using a filter bank with M with respect to the Bandpass filters connected in parallel to the input signal, the rectified output signals of which are transmitted in M channels, characterized in that two further signal transmission paths (KA, KB) are connected in parallel to the direct signal transmission path (K 0), that in each of the further signal transmission paths (KA, BK) frequency-dispersive delay devices (VEA , VEB) are provided, the frequency-time functions of which are complementary to each other, so that in each of the three signal transmissions supply paths (K 0, KA, KB) rectifiers (GR 0, GRA, GRB) are provided that a measurement of the pulse duration is made in three timing circuits (IDO, JDi, ID 2) for each of the three signal transmission paths (ATO, KA, KB) with the two further signal transmission paths (KA, KB) this measurement is carried out at the output of the delay device (VEA, VEB) and the three time values (TO, Ti, T2) obtained in this way are displayed for further evaluation.