CN1202633C - Method and device for quick capturing pseudo random codes in dynamic mass signal condition - Google Patents

Abstract

The invention relates to a fast acquisition method and device for a pseudo-random code under the dynamic condition of a large signal. The entire dynamic range of the signal is divided into two areas according to the signal-to-noise ratio, and two different PN code acquisition strategies are set. The acquisition strategy suitable for areas with high SNR is called fast search, and the acquisition strategy suitable for areas with low SNR is called conventional search. At the beginning of acquisition, it is assumed that the SNR is low and conventional search is used. Low, use low threshold to capture PN code. If the signal-to-noise ratio is high, it is judged to be at a high signal-to-noise ratio in the normal search process, and it is transferred to the fast search. Under the condition of introducing the automatic gain control circuit device, the PN code acquisition scheme of the present invention can normally work under the bigger signal dynamic condition with respect to common fixed threshold receiver, and ideal average acquisition time is arranged, especially in large frequency The effect is more obvious under partial conditions. The invention does not need additional equipment specially used for measuring the signal-to-noise ratio, and can be widely used in the fields of code division multiple access satellite communication and wireless communication.

Description

Pseudo-random code acquisition method and device under large-signal dynamic conditions

Technical field:

The invention relates to a receiving method and device of a direct sequence spread spectrum/code division multiple access (DS/CDMA) communication system, which can realize a pseudo-random sequence (PN code) under large signal dynamic conditions and large carrier frequency offset conditions synchronization.

Background technique:

In the direct sequence spread spectrum communication system, in order to correctly receive the demodulated signal, it is necessary to synchronize the local pseudo-random sequence (PN code) with the pseudo-random sequence in the received signal first. The process of PN code synchronization is generally divided into two stages: code capture and code tracking. This patent relates to the process of PN code capture.

Usually, the PN code acquisition method utilizes the autocorrelation characteristics of the pseudo-random code, that is, the local PN code is used to correlate with the received PN code. When the two are completely aligned, the autocorrelation value is the largest, and the phase difference between the two is one code. When there are more than one slice, its autocorrelation value is relatively small. Because in the direct sequence spread spectrum communication system, it is difficult to realize the synchronization of the carrier before the synchronization of the PN code, so the non-coherent receiver is often used to realize the acquisition of the PN code. The principle of the non-coherent receiver is to perform envelope detection on the output of the correlator to obtain a code synchronization test statistic. When the test statistic is greater than the preset threshold, it is considered that the received signal is synchronized with the local PN code.

The structure of the acquisition circuit of the PN code can be mainly divided into two types: a sliding correlator method and a matched filter method. The matched filter method can realize the rapid capture of PN codes in a short period of time, but in an all-digital system, the hardware complexity required to realize the digital matched filter is high, thus limiting its application range. The sliding correlator method is widely used in engineering practice because of its simple implementation, but its acquisition time is relatively long. In order to increase the acquisition speed of the sliding correlator method, generally sliding correlators using different phase local PN codes can be used to form a parallel correlator group for code acquisition.

In a communication system, the channel changes, and the power and signal-to-noise ratio of the receiver input signal change. For a digital receiver with a limited word length, the dynamic range of the input signal is too large or too small will cause the reception Therefore, the front end of the receiver often needs to use automatic gain control technology (AGC) to control the amplitude of the input signal. In the spread spectrum system, non-coherent AGC is often used to control the power of the input signal, that is, to keep the total power of the input useful signal and noise constant, but when the signal-to-noise ratio of the input signal changes greatly, the non-coherent ACC The amplitude of the useful signal in the output signal will also have relatively large changes, so the amplitude of the autocorrelation main peak output by the correlator will also have relatively large changes. For some pseudo-random codes with large autocorrelation sidelobes, such as GOLD codes, the value of the code synchronization test statistic in the case of non-synchronization at high SNR is often greater than that of the code synchronization test at low SNR Statistics. In addition, in some occasions, there is a large deviation between the local carrier frequency of the receiver and the carrier frequency of the received signal. Before the PN code synchronization is realized, the frequency difference is difficult to be compensated, and the unknown frequency difference will lead to a larger observation space. Therefore, for a receiver using a fixed decision threshold, it is difficult to adapt to a larger signal dynamic range, and it is difficult to realize fast synchronization of PN codes.

Invention content:

The object of the present invention is to provide a method and circuit device for PN code acquisition in a direct sequence spread spectrum communication system. Under the condition of introducing non-coherent automatic gain controller (AGC), it can achieve reliable acquisition of PN code with shorter acquisition time and simpler all-digital circuit in a large signal dynamic range, especially at carrier frequency. The effect is more pronounced when the difference is much larger than the information rate.

The fast acquisition method of pseudo-random code under the dynamic condition of large signal among the present invention, its step comprises:

1) According to the statistical characteristics of the code synchronization test statistic and the possible frequency difference range, the entire dynamic range of the signal is divided into a low SNR area and a high SNR area according to the SNR level;

2) Set the low threshold TL adapted to the low SNR area and the high threshold TH adapted to the high SNR area respectively according to the required capture performance; When searching for the PN code phase, the false alarm probability tends to 0 and the detection probability tends to 1, and the probability of false lock tends to 0; the setting of the high threshold satisfies the false alarm when searching the PN code phase with a high threshold in the high signal-to-noise ratio area When the probability tends to 0 and the detection probability tends to 1, and the signal-to-noise ratio is in the low SNR area, the probability that the code synchronization test statistic exceeds the threshold tends to 0;

3) Conventional search is applied in areas with low SNR, low-threshold TL is used, the dwell time is long, and multi-sample detection is adopted for each dwell; fast search is applied in high SNR areas, dwell time is short, and each dwell time is short. The number of samples of is less than that of conventional search;

4) After the conventional search is completed, perform a full-range fast search to detect all PN code phases once. If the code synchronization test statistic corresponding to a certain phase exceeds TH, then it is determined that the phase is the correct PN code phase, otherwise It is determined that the captured PN code phase is the correct PN code phase.

The capture method of the present invention also includes a method for further accelerating the search process, the steps of which include:

1) After setting the first dwell probability P, when a high signal-to-noise ratio is satisfied, the probability of the code synchronization test statistic obtained by the conventional search passing the first dwell is greater than P;

2) First take TL as the threshold value, use conventional search to capture PN codes, and record the number of first dwell times and the number of code phase adjustments at the same time;

3) When the number of code phase adjustments has reached a fixed value N, check whether the probability that the code synchronization test statistic has passed the first dwell is greater than P, and if it is greater than P, it is determined that the signal-to-noise ratio is high at this time, and directly Go to quick search;

Multi-dwell, multi-sample detection was employed in the conventional search. The number of samples used for the next dwell test of the same observation point is greater than the number of samples used for the last dwell test.

In the fast search, the residence time is short and the number of detection samples is small.

The device for acquiring pseudo-random codes under large-signal dynamic conditions of the present invention includes a correlator or correlator group for obtaining code synchronization test statistics, and also includes a separate correlator and its corresponding phase of a tracking captured PN code phase The PN code generating circuit; also includes a counting device that has counted the number of times of the first dwell and the number of code phase adjustments.

The present invention provides a PN code acquisition scheme for DS/CDMA systems in an additive Gaussian channel. Set two thresholds, high and low, where the high threshold is used for correlation peak capture under high SNR, and the low threshold is used for correlation peak capture under low SNR. The search strategy can also be divided into search strategies under low SNR and high SNR according to the difference of SNR.

In the case of a low signal-to-noise ratio, since the amplitude of the useful signal in the output signal of the AGC amplifier is small, a lower threshold needs to be used for capture. Moreover, under low signal-to-noise ratio, the method of comparing a single sample with the threshold is likely to lead to a higher false alarm probability and a lower detection probability. Therefore, a multi-sample test is adopted at this time, that is, multiple code synchronization test statistics are taken. The average value is compared with the threshold to reduce the time required for synchronization and the possibility of locking on the wrong PN code phase. In order to further reduce the average synchronization time, reduce the false alarm probability of code capture and increase the detection probability, it is necessary to perform multiple dwell checks when the signal-to-noise ratio is low, and often the number of samples used for each dwell check is greater than the previous one. . This search strategy under low signal-to-noise ratio is called conventional search in the present invention. However, if the conventional search is used in a higher SNR, since the autocorrelation characteristics of PN codes are often not ideal, the amplitude of its autocorrelation sidelobes is often large, so the probability of false alarms may be high, resulting in the average acquisition time It is even possible to lock on the wrong PN code phase, resulting in "false lock".

In the case of high signal-to-noise ratio, since the amplitude of the useful signal in the signal output by the AGC amplifier is relatively large, it is necessary to use a higher threshold for code capture, so as to ensure a lower false alarm probability so that the search results will not be Falsely locked on the autocorrelation sidelobe of the PN code. In addition, due to the high signal-to-noise ratio, a small number of samples or even a single sample can be used for each PN code phase for inspection to obtain a high detection probability, thereby reducing the time required for acquisition. This search strategy under high SNR is called fast search in this scheme. However, if the fast search strategy is used in a lower SNR, since the amplitude of the synchronous correlation peak of the PN code is relatively small, the probability of exceeding the high threshold is extremely small, so that the phase capture of the PN code cannot be realized correctly.

In the present invention, at the beginning of the acquisition process, it is assumed that the signal-to-noise ratio is low at this time, so a conventional search is used. If the signal-to-noise ratio is indeed relatively low at this time, it is appropriate to use a low threshold for PN code acquisition. The problem to be solved by the present invention is how to judge that the current SNR is in the high SNR area in the conventional search process, so as to switch to the fast search strategy as quickly as possible, thereby reducing the total average capture time.

If the SNR is actually high, since the autocorrelation sidelobe of the PN code is large, the probability of the code synchronization test statistic in the case of non-synchronization at this time is greater than the low threshold is very high, so the first time through the conventional search The probability of staying is also relatively high, so this feature can be used to judge the signal-to-noise ratio. During the capture process, record the number of first dwell times in the detection of a certain number of continuous phases. If it is greater than a preset value, it is considered that the signal-to-noise ratio at this time is high, so you can use high signal instead. Synchronization strategy under noise ratio. When the signal-to-noise ratio is quite high, some of the code synchronization test statistics on the PN code phase may pass through multiple dwells in the conventional search, resulting in a "false lock" situation. When a false lock occurs, regular searches cannot pass multiple residency checks to exit the false lock. In the present invention, after the conventional search is completed, the fast search strategy is used to check all phases. Under low signal-to-noise ratio, the autocorrelation sidelobe and main peak of the code synchronization test statistic often cannot pass the high threshold of the fast search. , so if none of the code synchronization test statistics on the pseudo-random code phase can exceed the high threshold, the PN code phase captured by the previous conventional strategy is considered to be the correct synchronization phase. Conversely, if the signal-to-noise ratio is high at this time, the main peak of the code synchronization test statistic will exceed the high threshold with a high probability, while the autocorrelation sidelobe has only a small probability to pass the high threshold, so if in the fast search strategy if capture If the code synchronization test statistic on a phase exceeds the high threshold, the PN code phase is considered to be the correct synchronization phase. The fast search performed after the conventional search completes the capture is called full-scale fast search.

In general, advantages and positive effects of the present invention can be summarized as follows:

1. Under the conditions of using a non-coherent automatic gain controller and frequency offset, the PN code acquisition scheme provided by the present invention can work reliably in a larger signal dynamic range than ordinary single-threshold receivers.

2. Under the conditions in 1 above, a relatively ideal average capture time can be obtained.

3. The PN code acquisition circuit provided by the present invention can realize the acquisition schemes described in 1 and 2 above with a small hardware cost, and does not need additional equipment specially used for measuring the signal-to-noise ratio.

Description of drawings:

Fig. 1 is a functional block diagram of a digital signal receiver used in the present invention.

Fig. 2 is the structure of the non-coherent receiver used in the embodiment of the present invention

Fig. 3 is a block diagram of a code capture circuit used in an embodiment of the present invention.

Fig. 4 is a flowchart of the PN code acquisition strategy used in the embodiment of the present invention.

Figure 5 is a flowchart of a conventional search in a PN code acquisition strategy used in an embodiment of the present invention.

FIG. 6 is a flow chart of fast search in the PN code acquisition strategy used in the embodiment of the present invention.

implementation plan:

The invention has a specific embodiment. The receiver input signal involved is a direct sequence spread spectrum signal. The spread spectrum sequence is obtained by padding the GOLD sequence with a period of 1023, and the period is 1024. The information rate is 2.4kbps, and the spreading ratio is 1024. The channel is an additive Gaussian white noise channel, the range of E _b /N ₀ is 7.8dB to 60dB, and the frequency difference is less than 600Hz. The communication is a burst mode, and it is required to realize the fast capture of the PN code in a relatively short period of time.

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

Fig. 1 is a system structure diagram of the baseband signal processing unit in the DS/CDMA all-digital receiver. The input signal of the baseband signal processing unit is the baseband analog signal obtained by down-conversion after intermediate frequency processing. The baseband signal output by the intermediate frequency unit is controlled by the automatic gain control module (AGC) 100 and 101 before its amplitude is sent to the A/D converter 110 and 111. The reason for doing this is: because the dynamic range of the signal in the embodiment is very large , in order to ensure the normal operation of the A/D converter, AGC is added before it so that the level of the signal plus noise entering the A/D converter remains within a constant range (that is, the linear working area of the A/D converter). Since the analog signal after the IF down-conversion still has a large frequency difference, it needs to go through the quadrature digital down-conversion 120 to correct the frequency deviation, and the residual frequency deviation after correction is within 600 Hz. The function of the digital matched filter 130 is to maximize the signal-to-noise ratio at the point sampled by the digital correlator bank. The resulting I-channel and Q-channel baseband digital signals are sent to the pseudo-random code acquisition circuit 131 described in the present invention to complete the phase synchronization process of the PN code.

The AGC used in this embodiment is a non-coherent AGC, which differs from the coherent AGC in that its work does not depend on the despread signal. The non-coherent AGC makes the ratio of the amplitude of the useful signal and the amplitude of the noise signal in the A/D input signal change, and the total output signal strength (average power) remains constant. In this embodiment, the output total signal strength is normalized to 1.

In this embodiment, the code acquisition algorithm adopts a serial search scheme, and the generation circuit of a single code acquisition test statistic is shown in FIG. 3 . The I-way digital correlator group 200 and the Q-way digital correlator group 201 perform correlation integration operations on the outputs of the I and Q-way matched filters and the output signal of the local PN code generator 230 respectively, and the integral cleaning control signal is generated by 240 . The outputs of the I and Q correlator groups are added through the squarers 210 and 211 and then added at 220 to obtain a code synchronization test statistic.

In this embodiment, the detection of the PN code phase is performed with a step size of 1/2 chip, and each correlator group is composed of 20 correlators, and the received signals of two adjacent correlators are The PN code phases differ by 1/2 chip, and this correlator group can simultaneously detect 20 PN code phases within the range of 10 PN code chips.

Under the condition of using non-coherent AGC, the autocorrelation sidelobe synchronous recognition detection amount at high SNR (such as E _b /N ₀ =40dB) has a considerable probability to be greater than the mean value of the main peak at low SNR. On the premise of no prior knowledge of SNR, if the threshold is designed according to the statistical characteristics of the main peak under low SNR, when the average value of PN code synchronous recognition detection is greater than the judgment threshold, it cannot be judged that it is under high SNR. The autocorrelation side lobe is still the autocorrelation main lobe under low signal-to-noise ratio, and the amplitude information on other observation points must be used to assist the decision.

In the conventional search, since it is assumed that the signal-to-noise ratio at this time is low, in order to reduce the average capture time, a three-dwell detection method is used in this embodiment, and the number of samples for each dwell is 4, 8, and 16 respectively. . During the three dwells, the accumulation of the code synchronization test statistics is completed by the multi-sample accumulator 300 . Get the average value of each PN code phase synchronization test statistic collected in this dwelling in each resident, and select the maximum value in a plurality of average values by 320 and compare with the low threshold by 320, if this value is greater than the low threshold Threshold, after this dwell, the control module 330 performs the next dwell on the PN code phase. If three dwells are passed, the PN code phase is considered to be the correct PN code phase captured by the conventional search.

Through the simulation and analysis of the statistical characteristics of the code synchronization test quantity under different SNRs, it is found that the power (variance) of the white noise remains approximately constant in the range of E _b /N ₀ ≤ 20dB due to the effect of AGC. The white noise in the input signal of the despreading unit in the interval occupies most of the power. If the threshold is designed according to the situation that E _b /N ₀ is equal to 10dB, the threshold can work normally within this range. When E _b /N ₀ increases further, the variance of white noise begins to decrease, and the influence of sidelobe of signal autocorrelation increases gradually, thus the possibility of "false lock" of the detector increases. When 20dB<E _b /N ₀ <27dB, the probability of the conventional search passing the first stop increases obviously, but "false lock" rarely occurs. When E _b /N ₀ >27dB, "false lock" is likely to occur, and the larger the signal-to-noise ratio, the easier it is to occur. Therefore, define E _b /N ₀ ≤20dB as the low signal-to-noise ratio area and design a threshold that can satisfy the average acquisition time as the low threshold in the conventional search process. The high SNR area is taken as above 20dB, and the high threshold set according to this SNR area satisfies that the false alarm probability is close to 0 and the detection probability is close to 1 within the range of 20dB≤E _b /N ₀ ≤60dB. And in the low SNR area, the probability that the code synchronization test statistic exceeds the threshold is close to 0, so such a high threshold setting is suitable for fast search and full-range fast search under high SNR.

After setting the high and low thresholds, the normal search and the full fast search can cover the dynamic range of the entire signal. The next question is how to distinguish between low SNR and high SNR situations. At the beginning of the code acquisition, since there is no prior knowledge about the signal-to-noise ratio, the code acquisition can only be performed from the conventional search. Conventional search works well when E _b /N ₀ ≤ 20dB`. When 20dB<E _b /N ₀ <27dB, due to the high probability of false alarm, the probability of conventional search passing the first stop increases, so the average acquisition time increases, resulting in a decline in the performance of the receiver, but at this time we The signal-to-noise ratio at this time can be judged by the probability of the receiver staying for the first time, and a more appropriate fast search strategy can be adopted instead. The probability P is set by the probability that the receiver passes the first dwell when 20dB<E _b /N ₀ <27dB. When E _b /N ₀ >27dB, the probability that the code synchronization test statistic under non-synchronous conditions passes through three dwells increases, which is likely to cause false locks. Therefore, after the regular search is completed, it is necessary to join a full-range fast search using a high threshold search. In the case of high SNR, the probability of the code synchronization test statistic in the non-synchronous case exceeding the high threshold is very small, while the probability of the correlation peak in the synchronous case exceeding this threshold is very high, so the whole fast search can easily make the receiver The computer exits the false lock caused by the high signal-to-noise ratio and quickly finds the correct PN code phase.

The hardware circuit resources used in this embodiment are: AGC, a correlator group that includes 20 correlators, a counting device that has counted the number of dwell times and code phase adjustment times for the first time, a free correlator and the PN of the corresponding phase A code generation circuit, a squarer, and a corresponding code capture control circuit.

Below in conjunction with Figure 4, the process of PN code acquisition is briefly described as follows:

As shown in Figure 5, steps 1, 2, and 3 are the normal search process:

1. Add 1 to the code phase adjustment counter. Use the correlator group to observe 4 times at 20 continuous PN code phase observation points respectively, obtain 4 sample values for each observation point, take the mean value as the code synchronization test statistic, and take the maximum value of the test statistic in this group (record its The corresponding correlator serial number is L, L<=20) is compared with the low threshold. If it is less than the low threshold, adjust the phase of the local PN code to the next group of phases and repeat this step. If it is greater than the low threshold, add 1 to the resident counter for the first time and then operate according to the counter value as follows: if the resident counter value for the first time is less than 15, then directly jump to step 2; if the resident counter value for the first time is exceeded If it is greater than 15, the first dwell probability has been calculated (the first dwell counter value is divided by the code phase adjustment counter value), if the probability is greater than the preset threshold P, then jump to step 4 for quick search Otherwise, skip to step 2 after resetting (clearing) the dwell counter and the code phase adjustment times counter for the first time.

2. The correlator group continues to observe 8 times at the same observation point for verification, take the average value of 8 sample values as the judgment value, take the judgment value of the Lth correlator and compare it with the low threshold, if it is less than the threshold, adjust the local PN code From the phase to the next group of phases, repeat step 1.

3. The correlator group continues to observe 16 times at the same observation point for confirmation, take the average value of 16 sample values as the judgment value, compare the judgment value of the Lth correlator with the low threshold, and adjust the local PN code if it is less than the threshold From the phase to the next group of phases, repeat step 1; if it is greater than the threshold, set the PN code phase information of the Lth correlator to the free correlator, and the free correlator enters the code tracking process and turns to step 5 at the same time.

As shown in Figure 6, steps 4 and 5 are the quick search process:

4. The correlator group uses the fast search strategy to continue searching on all phases, using a high threshold when searching, and only observes one sample at each observation point and takes the mean value as the code synchronization test statistic until all phases are traversed. In this process, if the judgment amount exceeds the threshold, observe another sample at the corresponding observation point position, if it still exceeds the threshold, consider the real main peak you are looking for, and skip to step 11. On the contrary, if no threshold point is found after traversing a week, return to step 1.

5. The correlator group uses the fast search strategy to continue searching on all phases, using a high threshold when searching, and only observes one sample at each observation point and takes its mean value as the code synchronization test statistic until all phases are traversed. In this process, if the test statistic exceeds the threshold, observe another sample at the position of the corresponding observation point, and if the judgment value still exceeds the threshold, then the phase of this point is considered to be the true main peak that is sought. On the contrary, if no observation point exceeding the high threshold is found after traversing a week, it is determined that the current position of the free correlator is the real main peak.

Claims (4)

1. A pseudo-random code acquisition method under a large signal dynamic condition, the steps comprising: 1) According to the statistical characteristics of the code synchronization test statistic and the possible frequency difference range, the entire dynamic range of the signal is divided into a low SNR area and a high SNR area according to the SNR level; 2) Set the low threshold TL adapted to the low SNR area and the high threshold TH adapted to the high SNR area respectively according to the required capture performance; When searching for the PN code phase, the false alarm probability tends to 0 and the detection probability tends to 1, and the probability of false lock tends to 0; the setting of the high threshold satisfies the false alarm when searching the PN code phase with a high threshold in the high signal-to-noise ratio area When the probability tends to 0 and the detection probability tends to 1, and the signal-to-noise ratio is in the low SNR area, the probability that the code synchronization test statistic exceeds the threshold tends to 0; 3) Conventional search is applied in areas with low SNR, low-threshold TL is used, the dwell time is long, and multi-sample detection is adopted for each dwell; fast search is applied in high SNR areas, dwell time is short, and each dwell time is short. The number of samples of is less than that of conventional search; 4) After the conventional search is completed, perform a full-range fast search to detect all PN code phases once. If the code synchronization test statistic corresponding to a certain phase exceeds TH, then it is determined that the phase is the correct PN code phase, otherwise It is determined that the captured PN code phase is the correct PN code phase. 2. The pseudo-random code acquisition method under the dynamic condition of large signal as claimed in claim 1, characterized in that the method also includes a method for further accelerating the search process, and its steps include: 1) The threshold probability P of the first stay is set, and the probability of the code synchronization test statistic obtained by the conventional search when the high signal-to-noise ratio is satisfied is greater than P; 2) With TL as the threshold value, the conventional search is used to capture the PN code, and the probability of the first stay is detected. If the probability is greater than P, it is determined that the signal-to-noise ratio is high at this time, and it is directly transferred to the whole fast search. 3. The pseudo-random code acquisition method under the dynamic condition of large signal as claimed in claim 1, characterized in that in the conventional search, multiple dwell and multiple sample detections are adopted; the sample used for the next dwell check of the same observation point The number is greater than the sample size used in the last dwell test. 4. The method for acquiring pseudo-random codes under large-signal dynamic conditions as claimed in claim 1, characterized in that the dwell time in the fast search is short and the number of detection samples is small.

Images