RU2264043C1 - Digital detector of complicated signals - Google Patents

Abstract

FIELD: radio engineering.

SUBSTANCE: device implements algorithms of discontinuous Fourier transformation and fast folding of received and bearing signals, to provide search-less detection of complicated signals with sizeable bases. Digital synchronized filter processes at video-frequency with use of standard digital assemblies and elements. Device has multipliers 3,4,14,15, phase changer 2 for π/2, circuit for delay for length of signal element 1, low frequency filters 5,6, analog-digital converters 7,8,12, microprocessor systems of discontinuous Fourier transformation 9,10, microprocessor systems of reversed discontinuous Fourier transformation 16,17, generator of bearing pseudo-random series 11, microprocessor systems of discontinuous Fourier transformation of bearing pseudo-random series 13, square-ware generators 18,19, adder 20, arithmetic device for taking square root 21, threshold device 22, discontinuous process pulse generator 23, device for splitting frequency in two 24, clock generator 25, interconnected by appropriate functional connections.

EFFECT: broader functional capabilities, higher efficiency.

2 dwg

Description

The invention relates to the field of radio communications, discrete information transmission systems using complex broadband signals based on pseudorandom sequences of the maximum period and Gold signals with binary phase shift keying (0, π) and is intended for constructing digital detectors of complex signals.

The closest in technical features to this device is the optimal detector of complex signals with unknown amplitude and random initial phase [1, p. 350]. This device has maximum noise immunity in normal noise with known parameters of the received signal, in addition to the amplitude and initial phase of high-frequency oscillations. The prototype detector contains two multipliers, a carrier frequency harmonic oscillator, a π / 2 phase shifter, two identical matched filters, two quadrants and an adder. A phase-shifting harmonic reference oscillator generates two quadrature harmonic oscillations Cos (ω ₀ t) and Sin (ω ₀ t), where ω ₀ is the carrier frequency of the received signal. The use of quadrature channels eliminates the influence of a random initial phase of the carrier oscillation. At the outputs of the multipliers, the input radio signal is translated into the video frequency region and the elements of the complex signal are selected, which are processed by two identical matched filters.

The disadvantage of the prototype is that for normal operation of the detector requires a priori information about the frequency of the received signal to translate the radio signal into the video area and the complexity of manufacturing matched filters, especially with large bases of signals used.

To eliminate the noted disadvantages of the analog detector of the prototype, it is proposed to use a digital detector of complex signals. Due to the fact that the carrier frequency of the received signal is not known for various reasons, it is proposed to use the autocorrelation signal processing method as the optimal method for receiving signals of an unknown shape to isolate the modulating function from a mixture of radio signal and interference. The autocorrelation method as applied to the task of isolating the elements of a complex signal consists in multiplying the received radio signal and delayed by the duration of the element of the complex signal.

The received radio signal phase-manipulated by a pseudo-random sequence can be represented as

s ₁ (t) = Ad _i [rect (t-iτ)] Cos (ω ₀ t + φ ₀ ), 0≤t≤T _c , i = 0,1 ... N-1

where A is the signal amplitude;

ω ₀ , φ ₀ - carrier frequency and the initial phase of harmonic oscillation;

T _c - signal duration;

τ is the duration of the signal element;

d _i = (1, -1) - elements of a pseudo-random sequence;

On the interval of the duration of the signal element, the result of multiplication will be

y ₁ (t) = s ₁ (t) s ₁ (t-τ) = Ad _i Cos (ω ₀ t + φ ₀ ) Ad _i-1 Cos (ω ₀ (t-τ) + φ ₀ ) =

= 1 / 2A ² d _i d _i-1 [Cos (ω ₀ τ) + Cos (2ω ₀ t-ω ₀ τ + 2φ ₀ )].

After filtering the high-frequency component using a low-pass filter, we obtain

Z ₁ (t) = 1 / 2A ² d _i d _i-1 Cos (ω ₀ τ).

As can be seen, during autocorrelation processing, the dependence on the initial phase φ _{0 of} harmonic oscillation is eliminated, but the coefficient Cos (ω ₀ τ) appears, depending on the value of the carrier frequency and the duration of the signal element. To eliminate this dependence, you must enter a quadrature channel. Accepted. the radio signal is passed through the phase rotation circuit at π / 2 and we obtain

S ₂ (t) = Ad _i [rect (t-iτ)] Cos (ω ₀ t + φ ₀ -π / 2) =

= Ad _i [rect (t-iτ)] Sin (ω ₀ t + φ ₀ ), 0≤t≤T _c , i = 0,1 ... N-1.

After multiplication in the quadrature channel, we obtain

y ₂ (t) = s ₂ (t) s ₁ (t-τ) = Ad _i Sin (ω ₀ t + φ ₀ ) Ad _i-1 Cos (ω ₀ (t-τ) + φ ₀ ) =

= 1 / 2A ² d _i d _i-1 [Sin (ω ₀ τ) + Sin (2ω ₀ t-ω ₀ τ + 2φ ₀ )].

At the output of the low-pass filter of the second channel we get

z ₂ (t) = 1 / 2A ² d _i d _i-1 Sin (ω ₀ τ).

Note that the product of elements of the sequence d _i d _i-1 from elements (1, -1) is equivalent to modulo addition of two elements d _i d _i-1 composed of (1, 0). One of the properties of linear recurrence sequences is that if a sequence is added modulo two with the same sequence but shifted by a number of elements shifted by a certain unequal period of the signal [2, p. 58], the same sequence is obtained, but with a different shift . The same is true for Gold sequences, since they are obtained by modulo summation of two two different recurrent sequences of maximum period. So the sequence

d _k = d _i d _i-1 , k = 0, 1 ... N-1

will be the same sequence as accepted, but with a different delay.

Therefore, at the outputs of the low-pass filters in two channels, a modulating function d _k with some unknown constant coefficients Cos (ω ₀ τ) and Sin (ω ₀ τ) is allocated. Then, in both channels, identical digital processing of the selected modulating function of the received signal is performed.

The analog signals from the outputs of the low-pass filters in both channels are sampled. The sampling rate is related to the maximum frequency of the spectrum of the modulating function. Usually the emitted signal on the transmitting side is limited in spectrum to the main lobe of the spectrum of the modulating function. The width of the main lobe is determined by the duration of the elementary symbol and is equal to the clock frequency f _τ = 1 / τ, which should be taken as the maximum frequency of the spectrum of the sampled signal, that is, f _max = f _τ . Then, according to the Kotelnikov theorem, the sampling rate is chosen equal to

f _d ≥2f _max = 2f _τ .

To isolate the signal, we use a discrete convolution of two time functions - the signal from the output of the quadrature channel and the reference signal generated in the generator of the copy of the detector signal. It is advisable to calculate the discrete convolution using the algorithm not in the time but in the spectral region, which requires fewer operations, especially when using the "fast Fourier transform".

Since the digital detector must produce an output signal that matches in shape with the cross-correlation function of the received and reference signals, the signal can be processed using the "speed" convolution algorithm. After analog-to-digital conversion, a sequence of encoded into digital samples

x ₁ (k) = z ₁ (kT _d ), k = 0, 1 ... N _d -1,

where T _d is the sampling interval equal to T _d = 1 / f _d ;

N _d = E _c [T _c / T _d ] is the number of samples per signal duration;

E _c [x] is the integer part of the number x,

fed to a microprocessor system that performs discrete Fourier transform (DFT). From its output, a sequence of spectral coefficients

S ₁ (n) = DFT [x ₁ (k)], n = 0, 1 ... N _d -1

multiplied with a sequence of spectral coefficients

S0 (n) = DFT [x ₀ (k)], n = 0, 1 ... N _d -1,

received from the reference sequence of the signal copy generator. The sequence thus obtained

S10 (n) = S1 (n) S0 (n), n = 0, 1 ... N _d -1

undergoes inverse discrete Fourier transform (ODPF) in a microprocessor system

H1 (k) = ODPF [S10 (n)], k = 0.1 ... N _d -1.

At the output of the second quadrature channel, we similarly obtain

S2 (n) = DFT [x ₂ (k)], S20 (n) = S2 (n) S0 (n) and H2 (k) = DFT [S20 (n)].

In modern practice, DFT and DFT are carried out in the same device. To eliminate the influence of the coefficients Cos (ω ₀ τ) Sin (ω ₀ τ), the outputs of both channels are squared and summed. Finally, the output signal of the digital detector has the form

H (k) = [H1 ² (k) + H2 ² (k)] ^{l / 2} .

The claimed device with digital processing of a complex phase-shift (0, π) signal is shown in FIG. It contains:

1 - delay circuit for the duration of the signal element τ;

2 - phase shifter on π / 2;

3, 4 - the first and second multipliers;

5, 6 - the first and second low-pass filters;

7, 8 - the first and second analog-to-digital converters;

9, 10 - the first and second microprocessor systems DFT;

11 - generator reference pseudo-random sequence;

12 - the third analog-to-digital Converter;

13 - microprocessor system DFT reference sequence;

14, 15 - the third and fourth multipliers;

16, 17 - the first and second microprocessor-based ODF systems;

18, 19 - the first and second quadrators;

20 - adder;

21 is an arithmetic device for curing a square root of a number;

22 - threshold device;

23 - sampling pulse generator;

24 - frequency divider into two;

25 - clock generator;

The device operates as follows. The received radio signal from the detector input goes directly to the first input of the first multiplier (3), and to the first input of the second multiplier (4) through the phase shifter to π / 2 (2). At the second inputs of the multipliers (3) and (4), the radio signal enters through the delay element (1) for the duration of the signal element τ. Diagrams of the signals at the outputs of various devices of the digital detector are shown in figure 2. The first graph shows the pseudo-random sequence (modulating function) u (t), which must be selected for processing, and the reference pseudo-random sequence uo (t), which is generated by the signal copy generator. The reference sequence is a mirror image of the received sequence, i.e. uo (t) = u (T _c -t), where T _c is the signal duration. Here are the recurrence sequences of a maximum period of 15 elements. The second graph shows the function s1 (t) - a harmonic signal modulated in phase by the sequence u (t) with some delay. In the drawings, where two dependencies are shown, one of them rises (or falls) for clarity. The third graph shows the dependences y1 (t) and y2 (t) at the outputs of the first and second multipliers. The fourth graph shows the signals (x1 _k and x2 _k ) from the outputs of the low-pass filters after sampling them in analog-to-digital converters. Signals x1 _k and x2 _k are sequences of digital samples following through sampling intervals T _d . The last graph shows the inverse discrete Fourier transforms for the discrete convolution of the signal of the first quadrature channel and the reference sequence H1 _{k of the} discrete convolution of the signal of the second channel and the reference sequence H2 _k and the resulting function H _k = (H1 _k ² + H2 _k ² ) ^1/2

which is a function of the mutual correlation of the received signal and the reference sequence of the signal copy generator. The signal H _k and is the output of a digital detector, through which the signal is detected. After choosing the maximum reference H _{k max} and comparing it with a threshold, a decision is made to detect a signal if it is exceeded.

Thus, the totality of the introduced devices and their relationships eliminates the influence of a priori uncertainty about the carrier frequency and digitally processes the received signal to detect it, which was not in the prototype.

Therefore, the technical solution meets the criterion of "novelty." In addition, since the required technical result is achieved by the entire newly introduced set of essential features, which are not found in the well-known patent and scientific literature on the filing date of the application, the invention meets the criterion of "inventive" level.

Sources of information

1. Varakin L.E. Communication systems with noise-like signals. - M .: Radio and communications, 1985 .-- 384 p., Ill.

2. Dyadyunov N.G. and Senin A.I. Orthogonal and quasi-orthogonal signals. Ed. E.M. Tarasenko. M.: Communication, 1977 .-- 224 p., Ill.

Claims (1)

A digital detector of complex signals containing the first and second multipliers, a phase shifter by π / 2, characterized in that an additional delay circuit for the duration of the signal element, the first and second low-pass filters, the first and second analog-to-digital converters, the first and second microprocessor systems are introduced discrete Fourier transform, reference pseudorandom sequence generator, third analog-to-digital converter, microprocessor system of discrete Fourier transform of the reference pseudo random sequence, the third and fourth multipliers, the first and second microprocessor systems of the inverse discrete Fourier transform, the first and second quadrators, the adder, the arithmetic device for extracting the square root of the number, the threshold device, the sampling pulse generator, the frequency divider by two, the clock generator, and the first input of the first multiplier is connected directly to the detector input, the first input of the second multiplier is connected to the detector input through a phase shifter π / 2, the second inputs of the first and second multipliers are connected to the detector input through a delay circuit for the duration of the signal element, the output of the first multiplier is connected through a series-connected first low-pass filter, the first analog-to-digital converter, the first microprocessor discrete Fourier transform, and the third multiplier , the first microprocessor system of the inverse discrete Fourier transform and the first quadrator is connected to the first input of the adder, the output of the second multiplier through interconnected second low-pass filter, second analog-to-digital converter, second microprocessor-based discrete Fourier transform system, fourth multiplier, second microprocessor-based inverse discrete Fourier transform and second quadrator connected to the second input of the adder, one output of the sampling pulse generator through a series-connected frequency divider to two, a clock generator, a reference pseudorandom sequence generator, a third analog-to-digital the converter and the microprocessor system of the discrete Fourier transform of the reference pseudo-random sequence are connected to the second inputs of the third and fourth multipliers, the other output of the sampling pulse generator is connected to the second inputs of the first, second and third analog-to-digital converters, the adder output is connected to the square root arithmetic extractor from the input of the threshold device, the output of the threshold device is the output of a digital detector of complex s catch.

Images