CN107493117B - The two-dimentional joint acquisition method of direct expansion msk signal under a kind of high dynamic - Google Patents

Abstract

The present invention provides a two-dimensional joint acquisition method of high dynamic down direct spread MSK signal, comprising the following steps: the intermediate frequency direct spread MSK signal is processed by receiving conversion filter, frequency down conversion, low pass filtering and extraction to obtain approximate direct spread BPSK signal; the approximate direct spread BPSK signal and the local code are jointly captured to obtain the estimated value of the code phase error and the estimated value of the Doppler frequency offset; after Doppler compensation is performed on the approximate direct spread BPSK signal based on the estimated value, the subsequent Accurate code tracking and despreading demodulation can be performed to obtain information. The invention has the advantages of small algorithmic complexity, simultaneous acquisition of pseudocode phase and Doppler frequency offset, small estimation error under high dynamics, and the like, and is very suitable for the application environment of high dynamics and low signal-to-noise ratio.

Description

A two-dimensional joint acquisition method of direct spread MSK signal under high dynamics

technical field

The invention relates to a digital communication technology, in particular to a two-dimensional joint acquisition method of direct-spread MSK signals under high dynamics.

Background technique

At present, spread spectrum systems mostly use BPSK and QPSK modulation methods, but these two modulation methods cannot be applied to the application environment with severe nonlinear distortion, Doppler frequency shift and multipath fading. The direct spread MSK communication system has the advantages of low interception of the spread spectrum system, multi-user random site selection capability, strong anti-interference performance, constant envelope of the MSK signal, high spectrum utilization, energy concentration, fast sidelobe attenuation, and band With the advantages of low external radiation power and insensitivity to nonlinear distortion, it has been widely used in tactical data link, civil aviation ground-air data link, missile guidance command transmission, satellite communication and other fields. Therefore, the direct expansion MSK signal still has a good application prospect in the field where the direct expansion BPSK/QPSK signal cannot be applied.

There are two classic MSK signal capture methods, one is code capture based on sliding correlation, which is simple to implement but takes a long time to capture; the other is code capture based on matched filtering. The capture correlation peaks of these two methods are easily affected by Doppler frequency offset, which is obviously not suitable for capture under high dynamic conditions.

In a highly dynamic environment, the most classic estimation method for Doppler frequency offset is nonlinear transformation combined with FFT. The specific method is to perform square or multiple nonlinear transformation on the received signal, and then process the peak value of the spectrum according to the signal FFT The index of the frequency offset estimate is obtained, but this method cannot work under low signal-to-noise ratio. G.J.R.Povey and others first proposed a capture model based on the combination of digital partial matched filter and FFT (PMF-FFT). Two-dimensional acquisition of pseudo-code phase and carrier frequency offset, but this method is mainly suitable for MPSK signals, and the search range of Doppler frequency offset is small, and it is still not applicable in high dynamic environment. In recent years, some scholars have successively proposed code acquisition methods under high dynamics such as differential acquisition and two-step acquisition, but it is difficult to strike a balance between low signal-to-noise ratio and estimation accuracy. Therefore, how to accurately and quickly complete the joint estimation of the pseudo code phase and Doppler frequency offset of the spread spectrum signal in a high dynamic environment has become a key technology of the direct spread MSK all-digital receiver.

Contents of the invention

The purpose of the present invention is to provide a two-dimensional joint acquisition method of direct spread MSK signals under high dynamics, to realize the capture of direct spread MSK signals under large Doppler frequency deviation, so that direct spread MSK signals can first realize multiple Puller frequency offset compensation, in the case that the residual frequency offset does not affect the correlation peak, the code capture of the signal is realized. And under low signal-to-noise ratio, higher detection probability can be obtained, and higher precision can be obtained at the same time, so as to ensure that the received signal has obtained better coarse synchronization before entering the signal tracking (fine synchronization) processing module.

The present invention comprises the following steps:

The intermediate frequency direct spread MSK signal is processed by receiving conversion filter, down-conversion, low-pass filtering, and extraction to obtain an approximate direct spread BPSK signal.

The approximate direct spread BPSK signal and the local code are jointly captured to obtain the estimated value of the code phase error and the estimated value of the Doppler frequency offset.

After performing Doppler compensation and despreading on the approximate direct spread BPSK signal based on the above estimated value, the information is acquired and transmitted after being modulated by MSK; where

Joint capture includes:

The approximate direct spread BPSK signal and the local code are convolved after delay processing. When the local code is aligned with the approximate direct spread BPSK signal, the estimated value of the code phase error is obtained at the position of the maximum correlation peak obtained by convolution. The correlation peak An estimate of the Doppler offset is obtained from the phase of the signal at .

As an improvement of the present invention, change the delay sample point, obtain some delayed approximate direct spread BPSK signals and local codes, and convolve the two signals under the same delay sample point, and obtain all Doppler frequency The estimated value of bias is averaged as the estimated value of Doppler compensation and despreading.

As an improvement of the present invention, the estimated value of Doppler frequency offset obtained after averaging the phases of multiple correlation peaks obtained after convolution is used as the estimated value of Doppler compensation and despreading.

As an improvement of the present invention, all correlation operations in the algorithm can be replaced by FFT in time domain convolution operation.

Compared with the traditional direct spread MSK signal acquisition method, the present invention has the following advantages: (1) all correlation operations in the algorithm can be replaced by FFT in the time domain convolution operation, which greatly reduces the complexity of the algorithm; The code phase and carrier frequency offset are searched at the same time, and the acquisition time is M spread spectrum code periods, and the parameter M can be flexibly selected according to the acquisition time and acquisition probability performance; (3) by delay multiplying the received signal and the local code, and then Correlation is used to obtain the correlation peak and estimate the code phase difference and Doppler frequency offset. By adjusting the delay amount and delay number, the Doppler frequency search range can be controlled to achieve a balance with the estimation accuracy, which is suitable for high dynamic environments; (4) use The receiving conversion filter processes the direct spread MSK signal and converts it into direct spread BPSK signal, which solves the problem of complex form of MSK signal, and can apply the fast acquisition method suitable for direct spread BPSK signal to direct spread MSK signal.

The present invention will be further described below in conjunction with the accompanying drawings.

Description of drawings

Fig. 1 is a system block diagram of an embodiment of the present invention.

Fig. 2 is a flow chart of the signal processing of the joint estimation module according to the embodiment of the present invention.

Fig. 3 is a schematic diagram of the variation curves of the detection probability and the false alarm probability with the input signal-to-noise ratio according to the embodiment of the present invention.

Detailed ways

The condition for accurate despreading and demodulation of direct spread MSK signal is signal synchronization, including pseudocode and carrier synchronization. Signal synchronization is divided into coarse synchronization and fine synchronization, and the present invention focuses on signal coarse synchronization, that is, acquisition of pseudocode and Doppler frequency offset.

The transmission signal adopts the form of "pilot code + transmission data". M data symbols with all 1s are inserted in front of the transmission data to synchronize the spreading code and carrier at the receiver. After the pilot code compensates the Doppler frequency offset and code phase, the data signal part can be accurately despread and demodulated.

The direct spread MSK signal can be generated in a serial manner, that is, the spread signal and the carrier cos(2πf ₁ t) are BPSK modulated to obtain the direct spread BPSK signal, and then the direct spread MSK signal can be generated through the conversion filter g(t) . The impulse response of the conversion filter is

in, f _c is the carrier frequency, and T is the chip width of the spreading code.

For the serial generation method of direct spread MSK signal, the filter h(t) matching the conversion filter g(t) is used at the receiving end to receive, and the impulse response of the corresponding receiving conversion filter is

The output signal of the conversion filter is down _- converted and low-pass filtered by f1, and then extracted by K times to obtain an approximate direct expansion BPSK baseband signal. This decimation multiple is consistent with the oversampling multiple, and the output baseband signal can be expressed as

Among them, τ is the code phase error; is the initial phase, which is set to 0 for the convenience of analysis; R _c is the spreading code rate; f _d is the Doppler frequency offset; ω _n is Gaussian complex white noise; since the pilot code part data is 1, formula (3 ) can be rewritten as:

The joint acquisition of the approximate direct spread BPSK signal and the local code includes the following steps:

In step S101, the received signal y(n) is delayed, that is, the received signal is delayed by D samples and then multiplied by the conjugate of the undelayed signal. The process is as follows:

Among them, ω'(n) is the noise after delay processing.

Step S102, modify the local code c(n), that is, do the same delay processing:

c _D (n) = c ^* (n)c(n+D) (6)

Step S103, convolving the delayed signal y _D (n) with the local correction code c _D (n) to obtain:

Among them, N is the period of the spreading code, and the convolution can be performed by the correlator.

In step S104, it can be seen from formula (7) that when the local code is aligned with the received signal, the correlation peak obtained by convolution is the largest:

So far, the estimated value of the code phase error It can be obtained from the position of the relevant peak.

Compared with the traditional code acquisition algorithm, it is obvious that this method eliminates the influence of Doppler frequency on the acquisition output correlation peak-to-peak value.

Step S105, the estimated Doppler frequency offset can be obtained from the phase of the signal at the correlation peak as

Among them, arg[z] is the angle of the complex number z; Indicates the correlation peak output result.

The estimated value represented by formula (9) is only based on one delay, in order to improve the accuracy, the concept of multiple delays is introduced. Therefore, the joint capture process changes to:

Step S201, initializing the delay amount D=D _max -d, where D _max is the maximum delay amount, and d is the delay number, where the maximum delay amount D _max and the delay number d are controllable to obtain a higher detection probability.

Step S202: Delay the received signal y(n) and the local code c(n) at the same time, the principle of delay is the same as the above steps S102 and S103, and obtain y _D (n) and c _D (n).

Step S203, input y _D (n) and c _D (n) into the correlator for convolution, search for the correlation peak position in the correlation output, and M correlation peaks can be obtained, where the position of the correlation peak is the estimated value of the code phase error

In step S204, the estimated value of the Doppler frequency offset can be obtained from the phase of the signal at the correlation peak, the same as (9).

Step S205, judge whether D>D _max , if the judgment condition is no, change the value of D, D=D+1, return to step S202; if the judgment condition is yes, enter step S206.

Step S206, averaging all acquired Doppler frequency offset estimates as the final Doppler frequency offset estimate

Wherein, d is the delay number, and when d=0, it is a single delay; D _max is the maximum delay.

The above process is based on the Doppler frequency offset estimation made by the correlation peak of a data symbol. Here, the average value of the correlation peak phase of the pilot code of M symbols can be used to replace the correlation peak phase of one symbol to improve the accuracy. Therefore, The capture process is modified to:

Step S301, initializing the delay amount D=D _max -d, where D _max is the maximum delay amount and d is the delay number, where the maximum delay amount D _max and the delay number d are controllable to obtain a higher detection probability.

Step S302: Simultaneously delay the received signal y(n) and the local code c(n). The delay principle is the same as the above steps S102 and S103 to obtain y _D (n) and c _D (n).

Step S303, input y _D (n) and c _D (n) into the correlator for convolution, and search for the correlation peak position in the correlation output, and M correlation peaks can be obtained, where the position of the correlation peak is the estimated value of the code phase error

In step S304, the estimated value of the Doppler frequency offset can be obtained from the phase of the signal at the correlation peak according to equation (9).

Step S305, averaging the phases of the obtained M correlation peaks:

in, is the resulting average phase value of delay D; is the output value of the correlation peak.

Step S306, judge whether D>D _max , if the judgment condition is no, change the value of D, D=D+1, return to step S302; if the judgment condition is yes, enter step S307.

Step S307, averaging all acquired Doppler frequency offset estimates as the final Doppler frequency offset estimate

Wherein, d is the delay number, and when d=0, it is a single delay; D _max is the maximum delay.

According to formula (12), it can be seen that the Doppler frequency offset search range of the present invention is:

Then, the maximum Doppler estimated frequency offset is:

The innovation point of the present invention is that the estimated value of pseudocode phase and Doppler frequency offset can be obtained at the same time by using the matching output correlation peak value of the delayed multiplication signal, so as to realize the two-dimensional joint capture of pseudocode and Doppler frequency offset, and the delay number d and the delay maximum value D _max are both controllable to achieve a balance between the detection probability and the Doppler estimation range.

Example

The invention is a two-dimensional combined capture method of direct-spread MSK signals under high dynamics. As shown in Figure 1, firstly, the received intermediate frequency direct spread MSK signal is processed by receiving conversion filter, down-conversion, low-pass filtering, and extraction to obtain an approximate direct spread BPSK signal, and then the approximate direct spread BPSK signal is sent to the joint estimation module to obtain The code phase error and the estimated value of the Doppler frequency offset are used to complete the work of the code capture link. After the received signal passes through the Doppler compensation/despreading module, the MSK modulated signal can be obtained.

System sampling frequency f _s =245.52MHz, intermediate frequency f _c =76.725MHz, oversampling multiple K=12, spreading code rate is R _c =20.46Mchip/s, data rate is 20kbps, spreading code adopts Gold sequence, code The length N=1023, and the number of pilot code symbols M=20.

The received intermediate frequency direct spread MSK signal is passed through the receiving conversion filter with an impulse response of h(t), and an approximate direct spread BPSK signal is obtained. The impulse response of the receiving conversion filter is

in, is the spreading code period. The frequency response of the conversion filter is

The design of the filter of the present invention adopts the convex optimization technology. At first, the filter design problem needs to be converted into a convex optimization problem, and the Chebyshev approximation of the converted filter can be established as a convex optimization model:

Among them, sup is the infimum; ω=2πf is the angular frequency; D(ω) is a given frequency response function; H(ω) is the frequency response of the designed filter, h(n) is the filter coefficient, and N ₀ is the filter order.

The practically applied filter coefficient h(n) is obtained through the cvx toolbox of Matlab software.

The signal after receiving the conversion filter is down-converted by f ₁ , where After down-conversion, low-pass filtering, and then K-fold decimation, the decimation multiple is consistent with the over-sampling multiple. Since the data of the pilot code part is 1, the baseband signal obtained at this time is

Among them, τ is the code phase error; is the initial phase; R _c is the spreading code rate; f _d is the Doppler frequency offset; ω _n is Gaussian complex white noise; for the convenience of analysis, set the initial phase

As shown in Figure 2, the steps of the joint capture module are as follows:

Step 1: Initialize the delay amount D=D _max -d.

D _max is the maximum delay amount, and d is the delay number, where the maximum delay amount D _max and the delay number d are controllable to obtain a greater detection probability.

Step 2: Simultaneously delay the received signal y(n) and the local code c(n).

The delay processing process is: delaying the signal by D samples, multiplying it with the undelayed signal after taking the conjugate, and obtaining the delayed signal y _D (n) and the local correction code c _D (n).

Step 3: Input y _D (n) and c _D (n) into the correlator, search for the correlation peak position in the correlation output, and M correlation peaks can be obtained. The position of the correlation peak here is the estimated value of the code phase error

Step 4: Average the phases of the obtained M correlation peaks:

in, is the resulting average phase value of delay D; is the output value of the correlation peak.

Step 5: Determine whether D>D _max .

If the judging condition is no, change the value of D, D=D+1, return to step 2; if the judging condition is yes, go to step 6.

Step 6: Average the (d+1) phase calculation results and calculate the Doppler frequency offset estimate:

According to formula (20), it can be seen that the Doppler frequency offset search range of the present invention is:

Then, the maximum Doppler estimated frequency offset is:

Fig. 3 _shows the _{detection probability} and The change curve of false alarm probability with signal-to-noise ratio. Calculation formula (21) can be obtained: Doppler frequency offset estimation range is [-284.17kHz, 284.17kHz].

It can be seen from the figure that when the signal-to-noise ratio reaches -20dB, the detection probability reaches 0.9155. Due to the constant false alarm threshold setting criterion, the false alarm probability is hardly affected by the signal-to-noise ratio. It can be seen that the pseudo-code-Doppler joint acquisition method designed in the present invention can accurately capture the pseudo-code phase and carrier frequency offset in a high dynamic environment.

It can be seen from this that the two-dimensional joint estimation method of direct spread MSK signal under high dynamics designed by the present invention has the advantages of small algorithm complexity and simultaneous completion of pseudo code phase and Doppler frequency offset compared with the existing acquisition method It has the advantages of capture and small estimation error under high dynamics, which is very suitable for the application environment of high dynamics and low signal-to-noise ratio, and has strong practical value.

Claims (5)

1. a two-dimensional joint acquisition method of direct-spread MSK signal under high dynamics, is characterized in that, comprises the following steps: The intermediate frequency direct spread MSK signal is processed by receiving conversion filter, down-conversion, low-pass filtering, and extraction to obtain an approximate direct spread BPSK signal. The approximate direct spread BPSK signal and the local code are jointly captured to obtain the estimated value of the code phase error and the estimated value of the Doppler frequency offset. After performing Doppler compensation on the approximate DS BPSK signal based on the above two estimated values, accurate code tracking and despreading and demodulation are performed to obtain information; where Joint capture includes: The approximate direct spread BPSK signal and the local code are respectively convolved after delay processing of D samples, and when the local code is aligned with the approximate direct spread BPSK signal, the estimation of the code phase error is obtained at the position of the maximum correlation peak obtained by convolution value, an estimate of the Doppler offset is obtained from the phase of the signal at the correlation peak. 2. method according to claim 1, it is characterized in that, change delay sample point D, obtain the approximate direct spread BPSK delay processing signal and local correction code after some delays, and two signal rolls under the same delay sample point Take the average value of all the estimated values of Doppler frequency offset obtained as the estimated value of Doppler compensation and despreading, where the estimated value of Doppler frequency offset after taking the average value is Among them, the phase of the correlation peak D is the delay sample point, D _max is the maximum delay amount, d is the delay number, R _D (*) is the approximate direct spread BPSK signal and the local code after the delay processing and the convolved signal, Is the estimated value of the code phase error, and R _c is the rate of the spreading code. 3. The method according to claim 2, wherein the Doppler frequency offset estimate is obtained after averaging the phases of a plurality of correlation peaks obtained after convolution, wherein the phases of a plurality of correlation peaks are averaged value is The Doppler frequency offset estimate obtained by averaging the phases of multiple correlation peaks is where M is the number of correlation peaks. 4. according to the described method of claim 1,2 or 3, it is characterized in that, intermediate frequency direct spread MSK signal adopts the serial mode to produce, specifically: the signal after the spread spectrum and carrier cos (2πf ₁ t) carry out BPSK modulation and obtain The BPSK signal is directly expanded, and then the MSK signal is generated by the conversion filter. The impulse response of the conversion filter is in, f _c is the carrier frequency, and T is the chip width of the spreading code. 5. method according to claim 4, is characterized in that, adopts the filter h (t) that is matched with conversion filter g (t) to receive intermediate frequency direct spread MSK signal, the impulse response of corresponding reception conversion filter for in, f _c is the carrier frequency, and T is the chip width of the spreading code.