RU2803198C1 - Method for generating and detecting a sync pulse of a phase-shift keyed signal - Google Patents

Abstract

FIELD: communications.

SUBSTANCE: on the transmitting side, the formation of the DPSI sequence, with which the phase of the carrier frequency of the clock pulse is modulated, is carried out by bitwise multiplication of one of the Walsh sequences Wk with one of the sequences of the full DPM_SI code, and on the receiving side, the sequence DPSI, the correlation function of the DPM_SI with the sequence Wk is calculated, then the sum of the moduli of the correlation functions of the DPH sequence with all sequences Wj is calculated, for j≠ k, the obtained sum of modules of correlation functions is subtracted from the first calculated correlation function, and if the resulting difference is greater than zero, then a detection sign is formed.

EFFECT: ensuring the absence of false detections of a noise-like signal sync pulse associated with the correlation properties of the modulating sequences.

1 cl, 1 dwg, 3 tbl

Description

The invention relates to the field of communications and can find application in systems that use noise-like phase-shift keyed signals.

In communication systems with noise-like signals, information is transmitted using service and information elements. The service element (sync pulse) of the noise-like signal is used to detect it. Information is transmitted using information elements. Modulation of the carrier frequency phase of service and information elements is carried out by binary sequences.

Detection of the sync pulse of a phase-shift keyed signal is carried out by a detector consisting of a Matched Filter (MF), an Adaptive Signal Level Determinator and a Threshold Device.

The operation of such a detector is based on the calculation of the correlation function (CF) of the signal observed in a given processing cycle with the sequence by which the phase of the synchronizing pulse (CP _SI ) was modulated. The adaptive signal strength detector calculates the threshold. When the value of the correlation function exceeds the value of the calculated adaptive threshold, the Threshold Device records the detection of the signal.

The disadvantage of such a detector is the occurrence of “false alarms” and “false detections”.

The occurrence of “false alarms” is associated with internal noise emissions exceeding the established noise threshold. The higher the noise threshold, the fewer “false alarms”, but the worse the sensitivity of the receiver, and, consequently, the shorter the signal detection range.

To combat false alarms, a mechanism is used to set a noise threshold at the output of the detector, which provides a specified probability of false alarms.

The occurrence of “false detections” is associated with:

- poor correlation properties of the binary sequences used (ACF of the modulating sequence of the synchronizing pulse and VCF of this sequence with modulating sequences of information elements);

- increase in side peaks of modulating sequences due to external interference;

- method for calculating the adaptive threshold;

- the presence of internal receiver noise at the input of the receiving device.

The number of “false detections” can be minimized only by selecting sequences in which the ACFs have a given value of side peaks, and the values of the VCFs do not exceed this value.

It is still possible to select reference sequences for synchronizing pulse modulation that have side ACF peaks less than a specified value. But to select for the selected reference synchronization pulse sequence a set of sequences for modulating information elements of the signal, for which the maximum peak of the VCF is equal to a given value, is a very difficult, almost impossible computational task (with a base of the signal element n = 256, it is necessary to calculate 2 ²⁵⁶ ≈ 10 ⁷⁷ VCF and select from them those for which the maximum peak of the CCF does not exceed the specified value). In addition, the selection of binary sequences that have good correlation properties significantly reduces their number.

The choice of one or another method for calculating the adaptive threshold depends on the nature of external interference, and is therefore unpredictable. For the same synchronization pulse, due to the randomness of the characteristics of external interference, the calculated adaptive threshold may be underestimated, which can lead to a “false detection,” or overestimated, which can lead to a missed signal.

The occurrence of “false detections” reduces the throughput of the communication system and requires additional computing resources to process “false detections,” which reduces the reliability of the hardware.

In patent No. 2608769 it was proposed to form the modulating sequence of the synchronization pulse as:

DP _SI = W _K *DP _M _ _SI , (1)

where W _K is the Walsh sequence that acts as a sign for detecting DP _SI by the “main peak of the CF”, DP _M _ _SI is the generating sequence that makes the Walsh sequence W _K noise-like, and the symbol “*” means bitwise multiplication.

The proposed detector reduces the number of “false detections”, but has the following disadvantages:

- selection of binary sequences with good ACF and VCF is necessary;

- implementation of the device “Determinant of the feature “main peak of CF” is a rather complex task.

Patent No. 2760567 proposes a detector that is independent of the correlation properties of sequences modulating the phase of the signal. The phase of the sync pulse is modulated as in the patent described above (No. 2608769).

Since from (1) it follows that

W _K = DP _SI *DP _M _ _SI ,

then it is proposed to replace the detection of DP _SI with the detection of W _K.

The operation of such a detector is based on the orthogonality of Walsh sequences W _j , where j = 1, 2, ..., N. In each cycle of processing the digitized signal (sequence of measurements), DP _X = DP _VX * DP _M _ _SI is calculated. The DP _X processing block consists of N channels, in each of which the CF values of the DPx sequence with all sequences W _j are calculated.

If the sequence DP _VX = DP _SI is observed at the input of such a multichannel detector, then DP _X = W _K . In this case, in accordance with the proposed criterion (orthogonality of Walsh sequences), the value of CF(W _K ,W _K )=B, and the value

( + ) < Δ.

In any other case, i.e. if DP _VH ≠DP _SI , then DP _X ≠W _K , the criterion will not be met.

Such a multi-channel detector significantly reduces the number of “false detections” and does not require the selection of binary sequences with good ACF and VCF.

However, it has a drawback associated with the presence of internal noise in the receiver. For small signals, random noise emissions will interfere with determining the orthogonality of the DP _X =W _K with other Walsh sequences. In this case, orthogonality can only be detected if all channels do not exceed the noise threshold.

A significant drawback of the existing and described above detectors is the issuance of “false alarms”.

The proposed detector avoids “false alarms” and “false detections” associated with the correlation properties of modulating sequences. Any sequence of the full code can be used as signal phase modulating sequences.

Before formulating the criterion for the proposed DP _SI detector, let’s consider one interesting property of complete code sequences. This property will be demonstrated using sequences with base B=8 as an example.

Let us calculate the correlation functions (CF) of each sequence of the complete DP _X code with the Walsh sequences W. We obtain Table 1.

Table 1 DP _x KF (DP _x ,W ₁ ) (DP _x ,W ₂ ) (DP _x ,W ₃ ) (DP _x ,W ₄ ) (DP _x ,W ₅ ) (DP _x ,W ₆ ) (DP _x ,W ₇ ) (DP _x ,W ₈ ) 1 2 3 4 5 6 7 8 0 -8 0 0 0 0 0 0 0 1 -6 -2 -2 2 -2 2 2 -2 2 -6 2 -2 -2 -2 -2 2 2 ……. ………. ………. ……….. ………….. ………….. ………….. ………….. ………….. 127 6 -2 -2 -2 -2 -2 -2 -2 128 -6 2 2 2 2 2 2 2 129 -4 0 0 4 0 4 4 0 ……. ………. ………. ……….. ………….. ………….. ………….. ………….. ………….. 152 -2 2 2 6 2 -2 -2 2 153 0 0 0 8 0 0 0 0 154 0 4 0 4 0 -4 0 4 ……. ………. ………. ……….. ………….. ………….. ………….. ………….. ………….. 253 6 -2 2 2 2 2 -2 -2 254 6 2 2 -2 2 -2 -2 2 255 8 0 0 0 0 0 0 0

For the sequence W _K we take, for example, the Walsh sequence W ₄ .

In table 1 in columns 1,2,3,5,6,7,8 in all rows we replace the CF values with mod CF. We get table 2 (in table 2 in column 4 the CF values remain).

For each row, we sum the values of modKF _j in columns 1,2,3,5,6,7,8 (except for column 4). Let us denote the resulting sum as ∑modKF (in Table 3 this is column 9).

For each line we calculate the value {KF₄(DP_X,W₄) -∑modКФ}, i.e. from the values of column 4 we subtract the corresponding values of column 9. The results are presented in column 10 of table 3.

In Table 3, all values of column 10 are negative, with the exception of row number 153, the value of which is positive and equal to base B=8. Sequence 153 received after removal from DP_VX masking sequence DP_{M_SI}. DP sequence₁₅₃ = W₄.

The same picture is obtained with W _K =W ₁ , W _K =W ₂ , ..., W _K =W ₈ .

A similar picture is observed for base B=16.

Thus, we can conclude that for any sequence of the complete DP code_VX≠DP_SI (DP_X≠W_K) value (KF_K(DP_X,W_K) -∑modKF) will be negative and only with DP_VX=DP_SI (DP_X=W_K) value (KF_K(DP_X,W_K) -∑modКФ) will be equal to B, i.e. >0.

The proposed detector uses precisely this property of binary sequences.

The diagram of the proposed multichannel SI detector is presented in Fig. 1.

Taking into account internal noise, the signal, in the general case, is determined by the expression:

DP _VX ^s+sh = DP _VX ^s + DP _VX ^w .

In Block 1 , the masking DP _{M_SI} is “removed” from the input signal DP _VX ^s+sh :

DP _X ^s+sh = DP _VH ^s+sh *DP _M _ _SI = (DP _VH ^s + DP _VH ^w )*DP _M _ _SI =

=DP _VX ^s * DP _M _ _SI + DP _VX ^w *DP _M _ _SI =DP _X ^s +DP _X ^w .

In Block 2 the following are calculated:

modKF ₁ (DP _X ^s+sh ,W ₁ )=modKF ₁ (DP _X ^s *W ₁ )+ modKF ₁ (DP _X ^w *W ₁ ),

modKF ₂ (DP _X ^s+w ,W ₂ )=modKF ₂ (DP _X ^s *W ₂ )+ modKF ₂ (DP _X ^w *W ₂ ),

…………………………………………………………………….

modKF _K-1 (DP _X ^s+sh ,W _K-1 )=modKF _K-1 (DP _X ^s *W _K-1 )+ modKF _K-1 (DP _X ^w *W _K-1 ),

modKF _K+1 (DP _X ^s+sh ,W _K+1 )=modKF _K+1 (DP _X ^s *W _K+1 )+ odKF _K+1 (DP _X ^w *W _K+1 )

……………………………………………………………………………………………….

modKF _N (DP _X ^s+sh ,W _N )= modKF _N (DP _X ^s *W _N )+ modKF _N (DP _X ^w *W _N ).

In Block 3 we calculate:

∑mod KF = modKF ₁ (DP _X ^s+sh ,W ₁ )+ modKF ₂ (DP _X ^s+sh ,W ₂ )+…+

+modKF _K-1 (DP _X ^s+sh ,W _K-1 )+ modKF _K+1 (DP _X ^s+sh ,W _K+1 )+ …+

+modKF _N (DP _X ^s+sh ,W _N )=

=modKF ₁ (DP _X ^with *W ₁ )+ modKF ₁ (DP _X ^w *W ₁ )+

+modKF ₂ (DP _X ^with *W ₂ )+ modKF ₂ (DP _X ^w *W ₂ )+…+

+modKF _K-1 (DP _X ^with *W _K-1 )+ modKF _K-1 (DP _X ^w *W _K-1 )+

+modKF _K+1 (DP _X ^with *W _K+1 )+ modKF _K+1 (DP _X ^w *W _K+1 )+…+

+modKF _N (DP _X ^with *W _N )+ modKF _N (DP _X ^w *W _N )=

={ DP _X ^s ,W _j )+ KF(DP _X ^c ,W _j )}+

+{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}.

In Block 4 the expression is calculated:

KF _K -∑modKF=KF _K (DP _X ^s+sh ,W _K )-∑modKF=

={KF _K (DP _X ^with ,W _K )+ KF _K (DP _X ^w ,W _K )}-

-{ DP _X ^s ,W _j )+ KF(DP _X ^c ,W _j )}-

-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}=

=[KF _K (DP _X ^s ,W _K )-{ DP _X ^s ,W _j )+ KF(DP _X ^c ,W _j )}]+

+[KF _K (DP _X ^w ,W _K )-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}]. (2)

Based on the properties of binary sequences described above and expression (2), the detection criterion W _K will look like this:

[KF _K (DP _X ^s ,W _K )-{ DP _X ^s ,W _j )+ KF(DP _X ^c ,W _j )}]+

+[KF _K (DP _X ^w ,W _K )-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}]>0 (3)

Let's consider possible situations at the input of the SI detector.

1. There is no signal at the detector input.

In this case, only the noise component will be at the detector input, i.e. criterion (3) will look like:

KF _K -∑modKF =

=KF _K (DP _X ^w ,W _K )-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}>0. (4)

Expression (4) will be greater than 0 only if at the detector input the noise randomly forms DP _VХ ^w = DP _SI (i.e. at the input of the channels there will be DP _X ^w = W _K ). The probability of such an event is 1/2 ^V. Since when the signal base is B = 256, the probability of formation of a sequence of DP _SI by noise will be equal to 1/2 ²⁵⁶ ≈1/10 ⁷⁷ , then the probability of fulfilling expression (4) (the probability of a false alarm) will be practically equal to 0.

If DP_VX ^w≠DP_SI (DP_X ^w≠W_K) then for any noise emissions expression (4) practically will always be negative.

2. At the detector input there is a signal DP _VX ^with ≠DP _SI (after “removing” the masking, an arbitrary sequence of DP _X ^with ≠W _K ) and noise will be obtained. In this case, in expression (3) the components

[KF _K (DP _X ^s ,W _K )-{ DP _X ^s ,W _j )+ KF(DP _X ^c ,W _j )}] and

[KF _K (DP _X ^w ,W _K )-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}]

will be negative and, therefore, expression (3) as a whole will be negative.

3. At the detector input, the signal is DP _VX ^with = DP _SI .

In this case, DP _X ^with =W _K and in expression (3) the component

{ DP _X ^s ,W _j )+ KF(DP _X ^c ,W _j )} will be equal to 0.

The detection criterion will look like:

KF _K (DP _X ^s ,W _K )+

+[KF _K (DP _X ^w ,W _K )-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}]>0

The component CF _K (DP _X ^s ,W _K ) will be positive, and the component:

[KF _K (DP _X ^w ,W _K )-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}] will be negative.

For large signals DP _X ^with =W _K expression (3) will be satisfied. When the signal decreases (range increases), criterion (3) will be fulfilled as long as the module of the positive value of the CF component _K (DP _X ^w , W _K ) is greater than the module of the noise component

[KF _K (DP _X ^w ,W _K )-{ DP _X ^w ,W _j )+ KF(DP _X ^w ,W _j )}].

Calculations show that the detection range of the proposed detector is comparable to the detection range of a classic detector. At the same time, in the proposed detector there are no “false alarms”, “false detections” and there is no need to select sequences with “good” ACF and VCF.

Claims (1)

A method for generating and detecting a synchronizing pulse of a noise-like signal, in which on the transmitting side the formation of a DP _SI sequence, with the help of which the phase of the carrier frequency of the synchronizing pulse is modulated, is carried out by bitwise multiplication of one of the Walsh sequences W _k with one of the sequences of the full DP _SI code, and on the receiving side by bitwise by multiplying the sequences DP _SI and DP _{M_SI} , the sequence DP _X is calculated, the correlation function of DP _X with the sequence Wk is calculated, then the sum of the modules of the correlation functions of the sequence DP _X with all sequences W _j is calculated, for j≠k, the resulting sum of modules is subtracted from the first calculated correlation function correlation functions, and if the resulting difference is greater than zero, then a detection sign is formed.