RU2278470C2 - Method for synchronization of radio signal - Google Patents

Abstract

FIELD: radio engineering, possible use for engineering radio-receiving devices.

SUBSTANCE: method includes, under conditions of multi-beam transmission of radio signal with determining of broadband signal receipt time, accompanied by a set of interfering reflected signals, vector q is generated representing correlation of input signal with a set of bearing transmitted signals, overlapping the area of possible moments of appearance of direct and reflected signals with discontinuousness, providing for clear contour of signal front, while digitization frequency exceeds capacity of pass band of filter utilized in RF Front-End more than four times; generation of vectors representing estimates of complex amplitudes, including amplitudes and phases of direct and reflected signals for each one of delay values of direct signal as weighed totals of generated correlations

generation of likelihood functional of type

for each one of direct signal delay values ; determining moment of receipt of direct signal on basis of searching for maximum of functional

EFFECT: increased precision of synchronization.

4 dwg

Description

The claimed invention relates to radio engineering and can be used in the design of radio receivers operating in the conditions of multipath propagation of radio signals.

There is a known solution to the problem of confident reception of signals due to synchronization (determining the moment of arrival of a radio signal) under multipath conditions, based on the selection of the leading edge of the direct signal. Technically, it is based on the use of narrow discriminator gates, the width of which should be less than the delay Δτ of the reflected beam relative to the direct one. Currently, quite a lot of such methods are known, in particular, one of the methods is known as the Strobe and Edge Correlator and is described in the report by Garin L., Van Diggelen F. and Rousseau J.M. "Strobe and Edge Correlator Multipath Mitigation for Code", proceedings of ION GPS-96, Kansas City, September 17-20 [1].

More advanced methods appeared a little later. Among them, it should be noted the development of the same authors, known as the Enhanced Strobe Correlator, described in Garin L. and Rousseau J.M. "Enhanced Strobe Correlator Multipath Rejection for Code and Carrier", proceedings of ION GPS-97, Kansas City, September 16-19 [2], as well as the MMT solution by Leica Hatch R., Keegan R. and Stansell T . "Leica's Code and Phase Multipath Mitigation Techniques", proceedings of ION National Technical Meeting, Santa Monica, January 14-16 [3]. Similar solutions from other developers are known under the abbreviations HRC (McGraw G. and Braasch M. "GNSS Multipath Mitigation Using Gated and High Resolution Correlator Concepts", proceedings of ION National Technical Meeting, San Diego, January 25-27 [4].) And MEDLL (Townsend B., Van Nee R., Fenton P. and Van Dierendonck K. "Performance Evaluation of the Multipath Estimating Delay Lock Loop", proceedings of ION National Technical Meeting, Anaheim, January 18-20 [5]).

These methods of combating multipath due to synchronization were embodied in the navigation receivers of such manufacturers as Ashtech (Magellan), JNS, Leica, NovAtel, Topcon, Trimble, etc.

Closest to the claimed invention is a synchronization method using a narrow band correlator (Narrow Correlator), published by American experts Van Dierendonck AJ, Fenton P. and Ford T. "Theory and Performance of Narrow Correlator Spacing in a GPS Receiver", proceedings of ION National Technical Meeting, San Diego, January 27-29 [6].

A disadvantage common to all known solutions is the fact that to ensure beam separation, it is necessary to use broadband RF Front-End, so that the front width of the direct signal τ _f does not exceed the delay of the reflected signal Δτ. At the same time, the use of broadband filters and narrow gates leads to a number of problems:

- the frequency selection of narrow-band industrial noise in broadband RF Front-End filters is deteriorating;

- it becomes impossible to use industrial RF Front-Ends with a standard bandwidth (such as RF Front-Ends, developed in accordance with the standard for wireless communications in the 2.4 GHz frequency range), which leads to higher cost of end devices;

- known methods do not provide a correct determination of the moment of arrival of a direct signal if the delay of the reflected signals is a multiple of the duration of the element of the broadband phase-shifted signals, as well as if the delay of the reflected signals is less than the width of the gates used;

- the use of known methods is characterized by the rejection of signals with delays greater than the width of the strobe. In addition to suppressing reflected signals, this also leads to suppression of a part of the power of the direct signal, thereby reducing the likelihood of correctly extracting bits of digital information contained in the received radio signal.

The present invention solves the problem of more efficient separation of the direct signal from the input mixture, consisting of direct and reflected signals. In this case, the initial prerequisite is the provision of high-precision synchronization (determining the moment of arrival of a complex phase-shift signal) under conditions when the delay of the interfering reflected signals Δτ can be less than the signal front width τ _f .

To achieve this technical result, it is proposed first of all to provide compensation for the reflected signals from the received mixture of direct and reflected signals. The main difficulty of such compensation is the difficulty of forming an estimate of the reflected signal for its subtraction from the input mixture. This is primarily due to the possible complex composition of the interfering delayed signal, which in most cases is a mixture of many individual reflected signals with different random delays, amplitudes and phases. In addition, the presence of a direct signal that also has random parameters interferes with an accurate estimate of the reflected signal.

Overcoming this difficulty is based on a joint assessment of the totality of random parameters of the entire input signal, presented in the form

Here A is the complex amplitude of the direct signal,

τ is the moment of arrival of the signal (delay of the direct signal),

α _i are the complex amplitudes of the individual components of the reflected signal having delays i · Δ, and Δ is the minimum distinguishable delay between the components, taken equal to the sampling period of the signal (Δ = l / F _cl ),

n (t) is the internal noise of the receiver.

According to the synthesis of the optimal algorithm, the best root-mean-square estimate of the vector α composed of the weights of the reflected signals α _i , i = 1 ... L, obtained for each value τ = k · Δ from the uncertainty range τ for a fixed value of A, is

where K is a matrix of coefficients of size L × L, determined by the waveform, the level of internal noise and a priori information about the magnitude of the complex amplitudes of the direct and reflected signals,

q _τ is a vector of length L containing the signal samples at the output of the correlator calculated for the corresponding signal delay values τ + Δ, ..., τ + L · Δ.

Thus, the vector of the complex amplitudes of the reflected signals is obtained in the form of a linear combination of L successive correlator outputs, and the algorithm for estimating the signal delay is reduced to finding the maximum functional

where q ₀ - reference signal at the output of the correlator for the delay τ,

and - a vector of length L, taking into account the mutual correlation of the direct and reflected signals with delays Δ, ..., L · Δ relative to the direct,

To ₀ is the weight coefficient.

The first term in the expression determines the contribution of the direct signal, the second - the contribution of the reflected signals, while the maximum functional is reached at the point of the true value of the signal delay.

In practical implementation, the algorithm for estimating the signal delay is reduced to the initial search for the signal by frequency and delay (Preliminary Signal Search Engine) with subsequent calculation of the signal values at the output of the correlator for a set of values of τ in the vicinity of the initial estimate of the delay τ _pre (Multi-channel Correlator), by calculating the values functional L (τ) and search for its maximum (Signal Delay Estimator). After evaluating the true signal delay τ _est and the complex coefficients of the reflected signals α, the reflected signals in the received mixture ξ (τ) (Reflected Signal Compensator) are compensated. The output signal s (t), free from distortion caused by the influence of reflected signals, is fed to the Digital Information Extractor in order to extract useful digital information contained in the signal Θ.

Figure 1 presents the execution algorithm of the proposed method, where:

1 - "Preliminary signal search engine" performs an initial search for a signal by frequency and delay (signal detection and preliminary estimation of its parameters).

2 - "Multi-channel correlator" correlates the input signal with a set of reference signals (signal replica) in the range of possible values of the delay of the direct signal.

3 - "Signal delay estimator" determines the delay of the direct signal τ _est , as well as evaluates the complex amplitudes α of the reflected signals.

4 - "Reflected signal compensator" compensates for interfering reflections in the input signal mixture.

5 - "Digital information extractor" extracts useful digital information contained in the signal.

The computational complexity of the algorithm shown in FIG. 1 is of the order of 2 · N · L · (L + 2) operations of multiplication and addition, where N is the number of analyzed positions for the signal delay.

The performance of the considered algorithm and evaluation of its characteristics was carried out on the basis of modeling. The input signal mixture was modeled as the sum of the direct and reflected signals, then the signal was passed through a filter with a bandwidth inversely proportional to the duration of the elementary pulse of the pseudorandom sequence, and then the received signal was processed in accordance with the algorithm to estimate the delay of the direct signal.

The simulation results are presented in the graphs below (figure 2-4).

Figure 2 shows the signal module at the output of a multi-channel correlator (a direct signal corresponds to a zero position in the delay). As can be seen, the presence of reflected signals completely distorts the shape of the direct signal, making it virtually impossible to determine the true signal delay.

для τ=0. Видно, что оценки комплексных амплитуд близки к истинным значениям.Figure 3 presents the modules of complex coefficients for direct and reflected signals. Zero delay corresponds to a direct signal. The curve with round markers corresponds to the true value of the complex coefficients specified in the simulation, the curve with square markers corresponds to the module for estimating complex coefficients obtained as

for τ = 0. It can be seen that the estimates of complex amplitudes are close to the true values.

Figure 4 shows the response of the received signal at the output of the correlator (curve with round markers) and the response of the signal at the output of the correlator after compensating for the reflected signals (curve with square markers). It is seen that the shape of the signal response obtained after compensation of the reflected signals is close in shape to the response in the case of receiving a single (direct) signal. Thus, the assessment of the complex coefficients α _τ and their subsequent use to compensate for the reflected signals allows us to restore the response form of the direct signal at the output of the correlator and thereby ensure high accuracy in estimating the true delay of the direct signal. The accuracy of the delay estimate is determined by the filter bandwidth used in the analog part of the receiver (i.e., the width of the autocorrelation function of the signal passed through the filter) and the signal-to-noise ratio. Under the conditions considered during modeling, the accuracy of the delay estimate is about 1 ... 2 ns (or 0.3-0.6 m).

Thus, the considered method allows

- estimate with high accuracy the moment of arrival (true delay) of the radio signal;

- evaluate the amplitudes and phases of the interfering reflected signals;

- compensate for the reflected signals in the input mixture by preparing a signal for extracting bits of digital information.

It should be noted that the method allows you to compensate for the reflected signals in the signal, while the direct signal is not distorted.

Claims (1)

A method for synchronizing a radio signal in conditions of multipath propagation of a radio signal with determining the moment of arrival of a broadband signal, accompanied by a set of interfering reflected signals, including generating a vector q of correlations of the input signal with a set of reference transmitted signals that overlap the range of possible moments of the appearance of the direct and reflected signals with discreteness, which provides a clear front signal, while the sampling frequency exceeds the bandwidth This filter used in the RF filter tract, more than 4 times; formation of vectors of estimates of complex amplitudes, including the amplitudes and phases of the direct and reflected signals for each of the delay values of the direct signal as weighted sums of the formed correlations

where K is a matrix of coefficients of size L × L, determined by the waveform, the level of internal noise and a priori information about the magnitude of the complex amplitudes of the direct and reflected signals; τ is the moment of arrival of the signal (delay of the direct signal); q _τ is a vector of length L containing the samples of the signal at the output of the correlator calculated for the corresponding signal delay values τ + Δ, ..., τ + L · Δ, where Δ is the sampling frequency; formation of the likelihood functional of a species

for each of the direct signal delay values, where q ₀ is the reference signal at the output of the correlator for the delay τ; and - a vector of length L, taking into account the mutual correlation of the direct and reflected signals with delays Δ, ..., L · Δ relative to the direct; To ₀ - weight coefficient; determination of the moment of arrival of a direct signal based on the search for the maximum functional

.

Images