RU2349049C2 - Method and device of demodulation of galilei signals from variable binary offset carrier (altboc) - Google Patents

Abstract

FIELD: physics, communication.

SUBSTANCE: invention concerns receivers of global navigating satellite system (GNSS) which work with satellite signals of Galilei from a variable binary offset carrier (AltBOC). The method of demodulation of a signal from a variable binary offset carrier, containing, at least, two subcarriers (E5a, E5b), each of which has cophased and a quarter-phase component, promodulated by pseudorandom codes, and quarter-phase components (E5aQ, E5bQ) modulated pilots-signals without data transmission, and cophased components (E5aI, E5bI) modulated data transmission signals, contains following stages: transformation of signals from a variable binary displaced bearing in the intermediate frequency; a filtration of the conversed signals by means of the band-pass filter and digitisation of the filtered signals; Formation of a phase bearing and rotational displacement of a phase bearing lumped signals according to the specified phase bearing; correlation turned lumped signals; formation for every subcarrier (E5a, E5b) pseudorandom binary codes and a phase of subcarrier which use for correlation turned lumped signals.

EFFECT: simplification of demodulation of Galilei signals.

11 cl, 13 dwg

Description

FIELD OF THE INVENTION

The present invention relates in a broad sense to receivers of a global navigation satellite system (GNSS), in particular to receivers that operate with Galilean satellite signals with a variable binary offset carrier (AltBOC).

State of the art

Global navigation satellite system (GNSS) receivers, such as Galileo receivers, determine location based on signals received from the orbit of GNSS satellites. GNSS satellites transmit signals using at least one carrier, with each carrier being modulated by at least a binary pseudo-random (PRN) code consisting of a random sequence of ones and zeros that are periodically repeated. Units and zeros in the PRN code are referred to as “code pulses”, and transitions in the code from one to zero and from zero to one that occur during “code pulse times” are referred to as “code pulse transitions”. Each satellite with a GNSS system uses an individual PRN code, and therefore the GNSS receiver can link the received signal to a separate satellite by determining the content of the PRN code in the signal.

The GNSS receiver calculates the difference between the signal transmission time and the signal receiving time by the receiver. The receiver then calculates the distance, or “pseudo distance” from the satellite, based on the corresponding time difference. Using “pseudo-distances” from at least four satellites, the receiver determines their global location.

To determine the time difference, the GNSS receiver synchronizes the locally generated PRN code with the PRN code in the received signal by aligning the pulses of the code in each of the codes. The GNSS receiver then determines how much the locally generated PRN code is time-shifted relative to the given synchronized satellite PRN code during transmission, and calculates the corresponding “pseudo-distance”. The closer the GNSS receiver combines the locally generated PRN code with the PRN code in the received signal, the more accurately the GNSS receiver can determine the corresponding time difference and "pseudo-distance", as well as, in turn, its global location.

Code synchronization operations include satellite PNR detection and code tracking. To detect the PRN code, the GNSS receiver makes a series of correlation measurements, which are time-separated by a code element. After detecting GNSS, the receiver tracks the received code. This is usually done using “Early-Minus-Late” correlation measurements, that is, measurements of the difference between (i) the correlation measurement associated with the PRN code in the received signal and the leading variant of the locally generated PRN code , and (ii) a correlation measurement associated with the PRN code in the received signal and a lagging version of the local PRN code. Then the GNSS receiver uses leading-minus-delayed measurements in the DLL (delay lock loop-automatic delay adjustment), and an error signal is generated that is proportional to the mismatch between the local and the received PRN codes. The error signal is used, in turn, to control the PRN code generator, which shifts the local PRN code in order to significantly minimize the DLL error signal.

The GNSS receiver also aligns the satellite carrier with the local carrier using correlation measurements associated with the instantaneous version of the local PRN code. To accomplish this, the receiver uses phase-locked loop (PLL) when tracking the carrier.

The European Commission and the European Space Agency (ESA) are improving the GNSS system known as the Galileo systems. Galilean satellites transmit two signals in the E5a frequency band (1176.45 MHz) and two signals in the E5b frequency band (1207.14 MHz) as a combined signal with a center frequency of 1191.795 MHz and a bandwidth of at least 70 MHz using AltBOC modulation. AltBOC signal generation is described in the European Commission's document “Status of Galileo Frequency and Signal Design”, by GWHein, J. Godet, JLIssler, JCMartin, P. Erhard, R .Lucas-Rodriguez and T.Pratt, published on September 25, 2002 at the following address: http://europa.eu.int/comm/energy_transport/galileo/documents/technical_en.htm. Like GPS (Global Positioning System) satellites, Galileo satellites transmit an individual PRN code, and the Galileo receiver therefore associates the received signal with a separate satellite. Accordingly, the Galileo receiver determines the corresponding "pseudo-distances" based on the difference between the satellite signal transmission time and the AltBOC signal reception time by the receiver.

A standard linear offset carrier (LOC) modulates the time domain of the signal using the sinusoid sin (ω ₀ t), which shifts the signal frequency to the upper sideband and the corresponding lower sideband. BOC modulation performs a frequency shift using a square waveform, or sign (sin (ω ₀ t)), and is usually denoted as BOC (f _s , f _c ), where f _s is the subcarrier frequency (square waveform) and f _c is this is the frequency distribution of the pulses of the code. For clarity, 1.023 MHz coefficients are usually excluded from the designation, and BOC (15.345 MHz, 10.23 MHz) modulation is denoted by BOC (15, 10).

Modulation of the time domain of the signal by the complex exponent e ^jω0 shifts the frequency of the signal only to the upper side frequency band. The goal of AltBOC modulation is to generate in a coherent form E5a and E5b bands that are modulated by complex exponents, or subcarriers, so that the signals can be received as a broadband "BOC-like signal." Each band E5a and E5b has a corresponding in-phase (I) and quadrature (Q) distribution, or PRN codes, with E5a codes shifted to the lower frequency band and E5b codes shifted to the upper frequency band. The corresponding E5a and E5b quadrature carriers are modulated by the pilot signal without data transmission, and the corresponding common-mode carriers are modulated by both PRN codes and data transmission signals.

AltBOC modulation has the advantage that E5a (I and Q) and E5b (I and Q) signals can be processed independently of each other as traditional BPSK (10) (Binary Phase-Shift Keying) signals or together, which leads to great advantages in terms of tracking noise and multipath.

Regarding the principle of demodulation in AltBOC modulation, it is enough to approximate the AltBOC signal in the main frequency band with its analogue AltLOC (Alternate Linear Offset Carrier - variable linear offset carrier):

Where

c ₁ (t) is the PRN code of the component (E5bI) of the E5b data and d ₁ (t) is the corresponding modulation bit;

c ₂ (t) is the PRN code of the component (E5aI) of the E5a data and d ₂ (t) is the corresponding modulation bit;

with ₃ (t) - PRN code of the component (E5bQ) of the E5b pilot;

c ₄ (t) - PRN code of the component (E5aQ) of the E5a pilot;

exponents representing the modulation subcarrier E5a and E5b;

ω _s is the pulsation of the lateral displacement frequency: ω _s = 2 πf _s , where f _s = 15,345 MHz.

In fact, s (t) contains additional product components, and the subcarrier exponents are quantized. This effect is not explicitly included in the formula for simplicity. S (t) is modulated on a carrier E5 at a frequency of 1191.795 MHz.

Most of the previous publications describe AltBOC from the point of view of satellite satellite equipment, i.e. from the point of view of the transmitter. But the issue of admission is given very little attention.

The publication "Comparison of the tracking error for AWGN code for Alternative-BOC, Complex-LOC and Complex-BOC modulation options in the E5 band of the Galilean signal", by M. Soellner and Ph. Erhard, GNSS 2003, April 2003, reveals the principle of constructing an AltBOC receiver to track the AltBOC pilot component as shown in FIG. 1.

Figure 1 shows that the AltBOC receiver receives a signal from antenna 1, which includes joint codes transmitted by all satellites within sight. The received signal is fed to RF / IF stage 2, which traditionally converts the received RF signal to an intermediate frequency IF signal, the frequency of which is compatible with other components of the receiver, then filters the IF signal through an IF bandpass filter that has a passband at the desired average frequency carrier, and samples the filtered IF signal with a frequency that satisfies the Nyquist theorem in order to generate the corresponding digital in-phase (I) and quadrature (Q) signal samples in a known manner. The filter bandwidth should be large enough to allow the passage of the primary harmonic of the AltBOC joint pilot code, which is approximately 51 MHz. A wide bandwidth leads to a relatively steep edge of the code pulse in the received code and, accordingly, to fairly well-defined correlation peaks.

The AltBOC receiver contains a local carrier oscillator 4, for example, of the NCO type (numerically controlled oscillator), synchronized with the IF frequency, to generate a phase angle of rotation of M bits supplied to phase shifter 3, which receives IF signal samples of N bits. After conversion in the phase shifter, the signal samples into N bits are fed to three complex correlators, each of which contains a signal multiplier 10, 11, 12 and an integrator 13, 14, 15. Integrators summarize the signal samples for a predetermined integration time T _int .

The AltBOC receiver also contains another local NCO type generator 5, synchronized with a frequency f _c of the code element, which controls the AltBBOC complex code generator 6 to locally generate complex PRN pilot codes for this satellite. The generated pilot codes pass through a multi-bit delay line 7 containing three cells E, P, L, which form, respectively, leading, instant and delayed reference signals of local PRN codes, which are fed to the input of multipliers 10, 11, 12.

The signals C _E , C _P , C _L supplied from the integrators 13, 14, 15 are then used to generate the carrier phase and code error signals that control the NCO generators 4, 5.

The disadvantage of the generator 6 AltBOC code is its complexity and multi-bit. I.e,

it produces a quantized version of the Alt-LOC signal during transmission in the main band (it is assumed that only the pilot component is tracked) in the following form:

The generation of such a complex signal transmitted in the main band causes difficulties. The construction scheme presented in figure 1 also assumes that all components (delay line, multipliers and integrators) operate on complex multi-bit numbers.

Disclosure of invention

The purpose of the present invention is to provide a simplified method and device for the demodulation of Galileo signals.

This goal is achieved by a method of demodulating signals with a variable binary biased carrier, containing at least two subcarriers, each having in-phase and quadrature components modulated by pseudo-random codes. In this case, the quadrature components are modulated by the pilot signals without data transmission, and the in-phase components are modulated by the data transmission signals. The specified method contains the following operations:

- conversion of signals with a variable binary biased carrier to an intermediate frequency, filtering using a band-pass filter of the converted signals and sampling the filtered signals,

- the formation of the phase of the carrier and the rotation of the phase of the carrier of the sampled signals in accordance with the specified phase of the carrier,

- correlation of signals obtained as a result of rotation of the carrier phase of the discretized signals.

In accordance with the invention, this method also comprises generating pseudo-random binary codes and a subcarrier phase for each subcarrier, which are used to correlate signals obtained as a result of the rotation of the carrier phase of the sampled signals.

According to a preferred embodiment of the invention, the method also comprises the step of converting said pseudo-random codes of said subcarriers into phase angles, which are combined respectively with the phases of the subcarriers so as to obtain the resulting phase angles for each subcarrier; wherein the resulting phase angles are phase shifted so as to obtain at least one leading, instantaneous and at least one delayed phase angle for each subcarrier. Said correlation operation comprises the steps of phase rotation of said rotated sampled signals by said leading, instantaneous and delayed phase angles of each subcarrier in order to obtain a leading, instantaneous and delayed reference signal of said sampled signals for each subcarrier, as well as integrating respectively leading, instantaneous and delayed reference signals for each subcarrier for a given time.

According to a preferred embodiment of the invention, the method also comprises a phase rotation operation of said rotated sampled signals in accordance with said phases of the subcarriers so as to obtain phase-rotated sampled signals for each subcarrier before correlation of said rotated sampled signals.

In accordance with a preferred embodiment of the invention, the method also includes a discharge shift operation of said pseudo-random codes in order to obtain at least one leading, instantaneous and at least one delayed pseudo-random code, wherein said correlation operation comprises the steps of combining said phase-rotated discretized signals for each subcarrier with the indicated leading, instant and lagging pseudo-random codes and integrating the resulting signals in for a predetermined time so as to obtain leading, instantaneous and delayed correlation signals for each subcarrier. Also, this method contains a low-frequency post-correlation phase, which includes the phase rotation of the leading and lagging correlation signals of each subcarrier, respectively, by opposite constant phase angles, as well as the addition of the thus obtained leading correlation signals of these subcarriers, instantaneous correlation signals of these subcarriers and the received delayed correlation signals of these subcarriers in order to obtain correspondingly leading leading th, instantaneous and delayed correlation signals.

According to a preferred embodiment of the invention, the method further comprises the step of determining the combined carrier and subcarrier frequencies for each subcarrier, wherein the phase rotation operations of said carrier and said subcarriers are combined into one phase rotation operation for each subcarrier using said combined carrier and subcarrier frequencies.

According to a preferred embodiment of the invention, said correlation operation comprises the operation of combining the phase-rotated sampled signals for each subcarrier, respectively, with the pseudo-random codes of the specified subcarrier, and the operation of integrating the resulting signals for a predetermined time to obtain a correlation signal for each subcarrier.

According to a preferred embodiment of the invention, the method comprises a low-frequency post-correlation phase comprising combining the correlation signals for said subcarriers to obtain an instantaneous correlation signal used as a PLL (PLL) input of a discriminator, an excitation generator that controls the specified operation of the carrier rotation, and leading-minus -delayed correlation signal used as an input of the discriminator DLL, an exciting generator controlling this formation code and the specified phase formation of the subcarrier.

In accordance with a preferred embodiment of the invention, the leading-minus-delayed correlation signal is obtained by correlation signals of said subcarriers E5a and E5b using the following formula:

C _{E5, EmL} = j (C _{E5a, 0} -C _{E5b, 0} ).

In accordance with a preferred embodiment of the invention, the DLL discriminator is a “Dot-product power” type discriminator and performs the following operation:

D = Real [C _{E5, EmL} · C * _{E5, 0} ],

where Real () is a function that reflects the real part of a complex number,

the signal D serves to excite the generator that controls the formation of the code and the formation of the phase of the subcarrier.

In accordance with a preferred embodiment of the invention, the DLL discriminator performs the following operation:

C _{E5, EmL} = j (C _{E5a, 0} -C _{E5b, 0} ),

where Imag () is a function that reflects the imaginary part of a complex number.

The invention also relates to a device for demodulating signals with a variable binary biased carrier, containing at least two subcarriers, each of which has a common-mode and quadrature component, modulated by pseudorandom codes, while the quadrature components are modulated by pilot signals without data transmission, and the common-mode components are modulated data transmission signals. In accordance with the invention, this device comprises means for implementing the aforementioned method.

Brief Description of the Drawings

The invention, as well as its features and advantages will be more apparent from the following description, presented with reference to the attached drawings.

Figure 1 - functional block diagram of the channel AltBOC demodulator according to the prototype.

Figure 2 is a graph of the correlation function of a single component when demodulating each component of the AltBOC signal.

Figure 3 is a Fresnel diagram of the correlation functions of the individual components E5aQ and E5bQ.

Figure 4 represents the peak values of the combined correlation function combining the correlation functions of the individual components E5aQ and E5bQ.

5 is a functional block diagram of an AltBOC channel demodulator in accordance with a first embodiment of the present invention.

6 is a functional block diagram of an AltBOC channel demodulator in accordance with a second embodiment of the present invention.

7 is a functional block diagram of the AltBOC demodulator with two channels shown in Fig.6.

Fig. 8 is a functional block diagram of a two-channel AltBOC demodulator in accordance with a third embodiment of the present invention.

Fig.9 is a Fresnel diagram of the correlation functions of the individual components E5aQ and E5bQ obtained with the AltBOC demodulator shown in Fig.9.

FIG. 10 is a functional block diagram of a receiver comprising an AltBOC demodulator shown in FIG. 8 according to a first embodiment.

11 is a functional block diagram of a receiver comprising an AltBOC demodulator shown in FIG. 8 according to a second embodiment.

12 and 13 are functional block diagrams of an AltBOC receiver obtained from the receiver of FIG. 11 according to the third and fourth embodiments.

The implementation of the invention

The main features of the invention are described in detail below. According to the AltBOC demodulation principle, a pilot channel is formed by combining the E5aQ and E5bQ signals. AltBOC pilot signal consists of components c3 and c4:

where ω _s is the ripple of the displacement frequency in the lateral frequency band: ω _s = 2πf _s , with f _s = 15.345 MHz.

. Соответствующая корреляционная функция С_Е5bQ(τ) может быть легко выведена (принимая полосу пропускания бесконечной):In principle, each component can be demodulated by correlating s _p (t) with a sequence of code pulses, with an _i- sequence multiplied by the complex conjugate value of the corresponding subcarrier exponent, that is, to track the ₃ (t) component, the receiver must correlate with

. The corresponding correlation function C _E5bQ (τ) can be easily derived (assuming an infinite bandwidth):

Where

the sign "∝" means "proportionally";

τ is the delay between the incoming signal and the reference signals with a local code and a subcarrier;

T _int - integration time;

T _c - pulse duration in units of time.

Changes in signal C _E5bQ (τ) as a function of code tracking error are shown in FIG. Graphs 17 and 18 are respectively the real (I) and imaginary (Q) components of this function, while graph 16 is their amplitude. It can be seen that it is a complex function of τ: if the reference signals with a local code and a subcarrier do not coincide, the energy passes from I to the Q branch. Such a correlation peak cannot be tracked, since the deviations of the code and the carrier are not explicitly separated: any code offset leads to a carrier phase tracking error. Since the carrier cycle is basically much faster than the code cycle, the energy in the Q-branch will tend to zero, which leads to the apparent pure correlation peak of the BPSK in the code cycle.

Additional information necessary to use the BOC principle is that the other sideband is coherently transmitted in the frequency range 2f _s = ω _s / π. The correlation function C _Е5аQ (τ) is obtained by correlating s _p (t) with

The Fresnel diagram in FIG. 3 gives a visual representation of the complex correlations of C _E5aQ (τ) and C _E5aQ (τ). In this diagram, both correlations are represented as vectors in the I, Q plane. As the code delay increases, τ C _Е5bQ and C _Е5аQ rotate by an angle + ω _s τ and -ω _s τ, respectively, and their amplitude decreases in accordance with the triangle function, forming two spirals, as shown in FIG.

The combined function of the correlation peak can be calculated by summing the correlations C _E5aQ (τ) and C _E5bQ (τ); which corresponds to the addition of vectors in figure 3:

As shown in FIG. 4, the function C _E5Q (τ), which corresponds to the AlktBOC peak correlation function, is valid (graph 36) for all code delays, the imaginary part (graph 37) is zero and, therefore, can be used to track code.

As for the pilot channel, the combined E5a / E5b correlation peak is obtained by simple summation of the individual peaks E5a and E5b. As for the data channel, the same principle can be used, but the data bits must be eliminated before combining: the correlation peak of the E5 data is represented as

Such a process of estimating a bit makes the tracking channel less coarse, especially when the signal-to-noise ratio is low (C / N ₀ ), where the probability of bit error is high.

Based on this principle, five preferred embodiments of the AltBOC demodulator will be considered in accordance with the invention. With the skillful separation of pre- and post-correlation processing, AltBOC processing during transmission in the main band can be performed with low losses with respect to traditional BPSK signals (BPSK - coherent reception of opposite phase modulation signals).

AltBOC demodulators presented below are built on the assumption that the pilot channel is being monitored, but the same approach applies to data channel tracking.

и

и суммирование этих двух комплексных корреляций. В приемнике это производится в двух идентичных каналах, совместно использующих один и тот же локальный код и генераторы несущей.It has been shown that building AltBOC correlation peaks involves correlating the input signal with

and

and summing up these two complex correlations. At the receiver, this is done in two identical channels sharing the same local code and carrier generators.

Эта операция эквивалентна повороту входного сигнала на угол -ω_st-π/2 с последующим умножением на с₃ PRN импульсов и интегрирование. Умножение на импульсы кода может рассматриваться как дополнительный поворот на 0°, если импульс является +1, или на 180°, если импульс является -1. Этот процесс ведет к первому построению канала AltBOC демодулятора, как представлено на фиг.5.As explained above, the demodulation of component c ₃ includes a correlation of the input signal with

This operation is equivalent to rotating the input signal by the angle -ω _s t-π / 2, followed by multiplication by ₃ PRN pulses and integration. Multiplication by code pulses can be considered as an additional rotation by 0 ° if the pulse is +1, or by 180 ° if the pulse is -1. This process leads to the first construction of the AltBOC channel demodulator, as shown in Fig.5.

In Fig. 5, the AltBOC channel of the demodulator receives a signal through antenna 1, which includes AltBOC composite codes transmitted by all satellites in the field of view. The received signal is fed to RF / IF stage 2, which converts the received RF signal to an IF signal of intermediate frequency, frequency-matched with other components of the receiver, then filters the IF signal through an IF bandpass filter with a passband at a given average carrier frequency and samples the filtered IF signal with a frequency that meets the requirements of the Nyquist theorem so as to form the corresponding digital in-phase (I) and quadrature (Q) samples of the signal for N bits in a known manner. The filter bandwidth is large enough to go through the fundamental of the composite pilot code, which is about 51 MHz. A wide bandwidth gives relatively sharp transitions of the code pulses in the received code, and therefore, correlation peaks are pretty well defined.

The AltBOC demodulator contains a local oscillator 4, for example, of the NCO type (numerically controlled oscillator), synchronized with the IF frequency, to form a phase angle of M bits, which is supplied to phase shifter 3, which receives IF signal samples of N bits. The signal samples after processing in the phase shifter 3 are fed in parallel to the three phase shifters 25, 26, 27 before being integrated into the three corresponding integrators 28, 29, 30, which are summed at the input of the signal sample during the integration time T _int .

Further, the AltBOC demodulator contains another local NCO type generator 5, synchronized with the code element frequency f _c and generating the code element frequency and the subcarrier frequency f _s = 1.5 f _c , to excite the subcarrier phase generator 20 and the E5b code generator 21. The output of E5b code generator 21 is coupled to a PRN phase detector 22. Subcarrier phase generator 20 generates a subcarrier phase of M bits at a frequency f _s supplied from the NCO of code generator 5. The E5bQ code generator 21 generates E5bQ code elements (0 or 1) at a frequency f _c obtained from the NCO of the code generator 5. The PRN phase detector translates the code elements (0 or 1) into the rotation angle of phase 0 or π.

The corresponding output signals of the phase generator of the subcarrier and the phase detector PRN are added by the adder 23, while the output signal of the adder is a phase-shifted signal (real number encoded in M bits) that controls the multi-bit delay line 24 with three cells E, P, L generating respectively, leading, instantaneous and delayed reference signals of the received PRN codes, which are supplied in the form of phase shifts to the corresponding phase shifters 25, 26, 27.

The correlation signals C _{E5b, -1} , C _{E5b, 0} and C _{E5b, 1} , coming from the integrators 28, 29, 30, are then used as an input of discriminators, which determine the phase mismatch between the code and the carrier, which are used to control NCO 4, 5 generators.

The demodulator channel in FIG. 5 has two main differences from the traditional AltBOC channel of the demodulator shown in FIG. 1:

- the input of the delay line 7 is the phase shift in the form of the actual value of the signal;

- multiplication by the momentum of the code preceding the integration is replaced by phase rotation.

Although the strobe pulse for this architecture is smaller than in the traditional one in FIG. 1, it is still quite large compared to the traditional 1-bit delay line.

The architecture described in FIG. 5 can be significantly improved due to the fact that E, P, L rotators 25, 26, 27 operate at the same frequency, but with a fixed phase difference. Namely, if P rotator 26 provides a phase shift of -ω _s t-π / 2, E rotator 25 provides a phase shift of -ω _s (t + dT _c / 2) -π / 2 and L rotator 27 -ω _s (t- dT _c / 2) -π / 2, where d is the Early-Late (leading-delayed) displacement in units of pulses and T _c is the pulse duration. This constant phase difference ± ω _s dT _c / 2 can be obtained from integration and is performed at a low frequency during post-correlation (after integration).

This leads to optimization of the architecture, as shown in Fig.6. Compared to the architecture of FIG. 5:

- each of the three rotators 25, 26, 27 is replaced by a signal multiplier 33, 34, 35, respectively;

;- the rotator 31 of the subcarrier E5bQ is placed between the output of the carrier rotator 3 and the corresponding inputs of the signal multipliers 33, 34, 35 and performs phase rotation on

;

- the multi-bit delay line 24 is replaced by a delay line 32 of the code for one bit (the PRN phase detector is deleted) and is controlled directly by the generator 21 of the E5b code and

- two multipliers 36, 37 of the signal respectively on e- ^jα and e ^{jα are} inserted respectively at the outputs E and L of the integrators 28 and 30.

Two signal multipliers 36, 37 relate to the low-frequency post-correlation stage (after integration), while another part of this architecture refers to the high-frequency pre-correlation stage.

According to this architecture, only one additional block is added with respect to the traditional BPSK demodulator - this is the subcarrier rotator 31, the phase of which is controlled by the NCO code generator 5. Such an architecture is mathematically equivalent to the architecture of FIG. 5 if α is set to ω _s dT _c / 2. However, other α values may be chosen to obtain virtually any other phase shift between the leading and retarded reference signals.

For clarity, the AltBOC demodulator architecture in FIG. 5 and FIG. 6 shows only three complex correlators (leading, instant, and lagging). In fact, tracking a side lobe may require at least two additional correlators (super-forward and super-retarded), but this entails a direct expansion of the system.

Therefore, the architecture shown in FIG. 5 or FIG. 6 can be expanded to any number of correlators. For example, n leading and m lagging correlators can be used, each powered by a corresponding delay line cell. With _{E5b, 0} corresponds to instantaneous correlation. Usually, the leading and lagging correlations are calculated using the delay of one cell relative to the instantaneous correlation, that is, they correspond to C _{E5b, 1} and C _{E5b, -1} . However, they can be set using any other delay. Typically, the purpose of additional correlations is to determine lateral amplitude tracking.

5 and 6 illustrate the architecture of one separate channel. In the AltBOC receiver, these two channels for the E5 signal (one for E5a and one for E5b) are connected and the correlations are summed to form the AltBOC correlation signal. Such a combined channel obtained from the architecture of FIG. 6 is shown in FIG. 7.

7, the architecture comprises a common RF / IF stage 2, NCO carrier rotator 3, generator 4, and NCO code generator 5.

, в то время как фазовращатель 31а поднесущей выполняет поворот фазы на

.Each channel E5a, E5b contains a subcarrier phase shifter 31a, 31b, an E5a / E5b code generator 21a, 21b, powered by a delay line 32a, 32b, respectively, three respective correlators E, P, L, each of which includes a multiplier 33a, 34a, 35a, 33b 34b 35b of the signal and integrator 28a, 29a, 30a, 28b, 29b, 30b. The leading and retarded branches of each channel E5a and E5b contain two respective signal multipliers 36a, 37a, 36b, 37b by a coefficient equal to e- ^jα and e ^jα , respectively. Subcarrier phase shifter 31b rotates phase by

while the subcarrier phase shifter 31a rotates the phase by

.

Channel E5a also contains an additional signal multiplier 41a by a factor of -1 included between the NCO code generator 5 and the subcarrier rotator E5aQ 31a. The outputs of these two channels are summed up by three adders 42, 43, 44, at the output of which the correlation signals C _{E5, 1} , C _{E5, 0} and C _{E5, -1} are received, respectively

Expanding formulas (4) and (5), we can obtain the following formulas that determine the correlations C _{E5b, k} and C _{E5a, k} :

where α = ω _s bT _c / 2 = 2πf _s bT _c / 2. The Early-Late interval d is determined by the clock frequency of the delay line 32. Typically, d is in the range from 0.1 to 1.

As for tracking, the receiver uses C _{E5, k} correlations to construct phase code and carrier discriminators whose output is proportional to the tracking error of the code phase and the carrier, respectively.

The main value used in the PLL discriminator is the instantaneous correlation C _{E5, 0} . The primary value used in the discriminator DLL is the difference between the Early (leading) and Late (lagging) correlations, also referred to as the leading-minus-delayed correlation and denoted by C _{E5, Eml} . This difference is defined as follows:

In the special case, d = 1 / (2f _s T _c ) = 1 / (2 * 15.345 / 10.23) = 1/3, α is π / 2, and it can be shown that C _{E5, Eml is} proportional to j (C _{E5a, 0} -C _{E5b, 0} ) for small tracking errors τ. This fact leads to a sharp decrease in channel complexity, since only instantaneous correlations (C _{E5a, 0} and C _{E5b, 0} ) are necessary to calculate the tracking of both the code and the carrier.

This property can be demonstrated by reprocessing the expression for C _{E5, Eml} as follows, taking into account that α = π / 2. Substituting formulas (8) - (13) into formula (14), we obtain the following:

On the other hand, with small code tracking errors (τ << 1), j (C _{E5a, 0} -C _{E5b, 0} ) it is simplified:

This relation shows that C _{E5, Eml is} proportional to j (C _{E5a, 0} -C _{E5b, 0} ). The coefficient (2-d) does not matter, since it is solely a gain compensated for by the normalization of the discriminator.

This leads to the architecture shown in FIG. 8, equivalent to the architecture in FIG. 7 in the case where d = 1/3, but is much simpler.

Compared to the architectures in FIGS. 6 and 7, this architecture does not contain a code delay line 32a, 32b and has a separate correlator for each E5a and E5b code. Each correlator contains a separate signal multiplier 51a, 51b, receiving a signal from the output of the corresponding subcarrier rotator E5a E5b 31a, 31b and codes from the corresponding code generator E5a and E5b 21a, 21b and a separate integrator 52a, 52b. The output signals C _{E5a, 0} and C _{E5b, 0 of the} integrators 52a, 52b are fed to the adder 63 in order to obtain an instantaneous correlation signal C _{E5, 0} , and to the comparator 64, and the multiplier 65 so as to obtain an Early-minus-Late ( leading-minus-lag) correlation signal C _{E5, Eml} = j (C _{E5a, 0} -C _{E5b, 0} ).

You can see that the last version of the architecture is extremely simple, because it contains only one correlator needed in the channel. Surprisingly, it should be concluded that the AltBOC demodulator can be performed very effectively in terms of strobe pulses, despite the apparent complexity.

The latest architecture shows that AltBOC signal tracking can be performed without leading or lagging correlator. Such an amazing result can be understood from another Fresnel diagram shown in Fig.9. As it was established above, the code shift τ is proportional to the angle φ between the correlation vectors С _{E5а, 0} and С _{E5b, 0} : φ = 2ω _s τ. From the diagram it can be seen that the vector j (C _{E5a, 0} -C _{E5b, 0} ), indicated by “EL corr” in the diagram, obtained by subtracting the vector C _{E5b, 0} from the vector C _{E5a, 0} and by rotating the resulting vector 90 degrees, is valid and has an amplitude proportional to the angle φ. This is the main reason why tracking AltBOC code does not require leading and lagging code reference signals: the code offset can be obtained exclusively from instant correlators.

FIG. 10 represents a receiver comprising the AltBOC demodulator of FIG. 8, PLL (phase locked loop) and a DLL (delay locked loop) controlling the NCO carrier 4 generator and the NCO code generator 5. The PLL comprises a demodulator 71, the output of which P is filtered by a PLL filter 72 before applying a carrier generator 4 to the NCO control input. PLL discriminator 71 is a discriminator that implements the characteristic of arctangent, which calculates the angle of the complex number C _{E5, 0} :

The DLL (delay auto-tuning) contains a discriminator DLL that receives the correlation signal C _{E5, eml} , and a DLL filter 76 associated with the NCO control input of the code generator 5. DLL discriminator is a discriminator that performs the function of a scalar product that calculates the signal D = Real (C _{E5, Eml} · C * _{E5, 0} ). Thus, the DLL discriminator contains a complex conjugate function 73, to which a signal C _{E5, 0} and a signal from a multiplier 74 are supplied to multiply the signals obtained after multiplying by j in the multiplier 65, by a complex conjugate function 73. Then, the signal D obtained in accordance with with a function 75 representing the real part of the complex signal generated by the signal multiplier 74.

After some algebraic operations, the simpler architecture of FIG. 11 can be obtained from the architecture of FIG. 10, which requires fewer operations to calculate the same discriminator DLL. According to the discriminator in figure 10:

Thus, the DLL discriminator in FIG. 11 comprises a complex conjugate function 81, to which a correlation signal C _{E5a, 0} and a signal from a multiplier 82 are supplied to multiply the complex conjugate function signal and the correlation signal C _{E5b, 0} . Then, signal D is obtained using the function Imag () 83, which extracts the imaginary part of the complex signal generated by the signal multiplier 82.

A further modification of the architecture of FIG. 11 consists in replacing the Imag () operator with the Angle () operator (that is, a block of the same functionality as arctan discriminator 71).

In addition, the architecture of FIG. 11 can be optimized as shown in FIG. 12, taking into account that the phase rotation in the carrier rotator 3, followed by the phase rotation in the subcarrier rotators 31a, 31b, can be combined into a single phase shifter with a phase corresponding to the sum of the phases of the carrier and subcarrier.

Thus, in FIG. 12, the carrier rotator 3, the two subcarrier rotators 31a, 31b, and the multiplier 41a of FIG. 11 are replaced by two phase shifters 92a and 92b (one for each channel E5a and E5b) receiving a down-frequency signal with RF / IF Scheme 2. In addition, the subcarrier phase generated by the NCO code generator 4 is added by the adder 93a to the phase generated by the NCO generator 3 of the carrier and subtracted from it by the adder 93b, the addition results being fed to the channel phase shifters 92a, 92b, respectively E5a, E5b.

The architecture shown in FIG. 13 can be obtained from the previous one by replacing the NCO code generator with a simpler NCO 95, generating only the code pulse frequency f _c , and the code pulse frequency multiplier f _c 1.5 by 1.5, so as to obtain the subcarrier frequency f _s , which is the input of adders 93a, 93b. This requires duplication of the NCO carrier generator 4 to each channel E5a, E5b. The carrier frequency monitored by the PLL is supplied to the adders 93a, 93b, the respective outputs of which drive the NCO carriers 91a, 91b of the carrier of both channels E5a, E5b so as to control the combined carrier + subcarrier frequencies of both channels E5a, E5b.

In this architecture, the high-frequency processes of preliminary correlation of the channels E5a, E5b remain identical. They both contain phase shifters 92a, 92b, two NCO generators 91a, 91b, code generators 21a, 21b and a correlator. Moreover, if the NCO generator is duplicated and introduced one per channel, then each of the high-frequency processes of preliminary correlation of the E5a and E5b channels is identical to the traditional BPSK (binary phase modulation) channel, which gives huge advantages when designing a combined AltBOC / BPSK receiver.

Of course, the architecture optimization performed in FIGS. 12 and 13 can also be applied to the architectures in FIGS. 5, 6 or 7.

Claims (11)

1. A method of demodulating signals with a variable binary biased carrier, containing at least two subcarriers (E5a, E5b), each of which has a common-mode and quadrature component modulated by pseudorandom codes, and the quadrature components (E5aQ, E5bQ) are modulated by pilot signals without data transmission, and the common-mode components (E5aI, E5bI) are modulated by data transmission signals, the method comprising the following operations: converting signals from a variable binary offset carrier to an intermediate frequency, filtering using a bandpass filter of the converted signals and sampling the filtered signals, generating a carrier phase and rotating the carrier phase of the sampled signals in accordance with the specified carrier phase and correlating the sampled rotated signals, characterized in that it also contains pseudorandom binary binary subcarriers (E5a, E5b) codes and phase subcarrier, which are used to correlate the sampled rotated signals. 2. The method according to claim 1, further comprising the step of converting said pseudo-random codes of said subcarriers into phase angles, which are combined respectively with the phases of the subcarriers in such a way as to obtain the resulting phase angles for each subcarrier; moreover, the resulting phase angles are phase shifted so as to obtain at least one leading, instantaneous and at least one retarded phase angle for each subcarrier, wherein said correlation operation comprises phase rotation operations of said rotated sampled signals in accordance with the indicated leading, instantaneous and delayed phase angles of each subcarrier in order to obtain the leading, instantaneous and delayed reference signal of the indicated sampled signals for each subcarrier, as well as integration of a leading, instantaneous, and delayed reference signal for each subcarrier for a given time (T _int ), respectively. 3. The method according to claim 1, further comprising a phase rotation operation of said rotated sampled signals in accordance with said phases of the subcarriers so as to obtain phase-rotated sampled signals for each subcarrier (E5a, E5b) before correlating said rotated sampled signals. 4. The method according to claim 3, further comprising a discharge shift operation of said pseudo-random codes so as to obtain at least one leading, instantaneous and at least one delayed pseudo-random code, wherein said correlation operation comprises combining operations of said phase-rotated sampled signals for each subcarrier with the specified leading, instantaneous and lagging pseudo-random codes and integrating the resulting signals over a given time (T _int ) so that the floor to learn leading, instantaneous and delayed correlation signals (C _{E5a, -1} , C _{E5a, 0} , C _{E5a, 1} ; C _{E5b, -1} , C _{E5b, 0} , C _{E5b, 1} ) for each subcarrier (E5a, E5b), the method also includes a low-frequency post-correlation phase, which includes the rotation of the leading and retarded correlation signals of each subcarrier respectively by opposite constant phase angles (jα, -jα), as well as the addition of the thus obtained leading correlation signals of these subcarriers, instantaneous correlation signal in said subcarriers and the thus obtained retarded correlation signals of said subcarriers so as to obtain respectively resultant advancing, immediate and delayed correlation signals (C _{E5a, -1,} C _{E5a, 0;} C _{E5a, 1} ). 5. The method according to claim 3, further comprising the step of determining the combined carrier and subcarrier frequencies for each subcarrier, wherein the phase rotation operations in accordance with the indicated carrier phase and the indicated subcarrier phases are combined into one phase rotation operation for each subcarrier using the combined frequencies carrier and subcarrier. 6. The method according to claim 3 or 5, characterized in that said correlation operation comprises the operation of combining said phase-rotated sampled signals for each subcarrier (E5a, E5b), respectively, with pseudorandom codes of the specified subcarrier and the integration operation for a given time (T _int ) the resulting signals to obtain a correlation signal (C _{E5a, 0} , C _{E5b, 0} ) for each subcarrier. 7. The method according to claim 6, additionally containing a low-frequency post-correlation phase containing the operation of combining the correlation signals (C _{E5a, 0} , C _{E5b, 0} ) for these subcarriers (E5a, E5b) to obtain an instantaneous correlation signal (C _{E5, 0} ) serving PLL discriminator input, the excitation generator (4) controlling said carrier rotation operation, and advancing-negative delayed correlation signal (C _{E5, EmL,} serving DLL discriminator input, the excitation oscillator (5) controlling said code generation and said ormirovaniem subcarrier phase. 8. The method according to claim 7, characterized in that the leading-minus-delayed correlation signal (C _E5 , _EmL ) is obtained from the correlation signals (C _{E5a, 0} ; C _{E5b, 0} ) for the specified subcarriers (E5a, E5b) using following formula:
C _E5 , _EmL = j (C _{E5a, 0} -C _{E5b, 0} ). 9. The method according to any one of claims 7 and 8, characterized in that the DLL discriminator is a discriminator of the type "degree of scalar product" and performs the following operation:
D = Real [C _E5 , _EmL · C * _{E5, 0} ],
where Real () is a function that reflects the real part of a complex number,
the signal D serves to excite the generator (5), which controls the specified formation of the code and the specified phase formation of the subcarrier. 10. The method according to any one of claims 7 and 8, characterized in that the DLL discriminator performs the following operation:
D = Imag (C _{E5b, 0} -C * _{E5a, 0} ),
where Imag () is a function that reflects the imaginary part of a complex number. 11. A device for demodulating signals with a variable binary biased carrier containing at least two subcarriers (E5a E5b), each of which has in-phase and quadrature components modulated by pseudorandom codes, while the quadrature component (E5aQ E5bQ) is modulated by pilot signals without data transmission, and the in-phase component (E5aI E5bI) is modulated by data transmission signals, characterized in that it contains means for implementing the method according to any one of claims 1 to 10.

Images