A tracking method for variable symbol rate spread spectrum signal

Abstract

The present invention discloses a tracking method for variable symbol rate spread spectrum signals, wherein a seamless switching coefficient calculation module and a loop bandwidth adjustment coefficient calculation module are added to a carrier tracking loop, wherein the seamless switching coefficient corrects the predicted value of the change amount of the loop filtering result when the symbol rate is switched, and the loop bandwidth adjustment coefficient corrects the original bandwidth after the symbol rate is switched to obtain a new bandwidth, and then the filter parameters are updated with the new bandwidth and applied to the carrier loop. The present invention provides a tracking method for variable symbol rate spread spectrum signals based on a second-order frequency-locked loop and a third-order phase-locked loop, respectively, which reduces loop jitter while keeping the loop at the optimal bandwidth, reduces the difficulty and cost of upgrading the existing loop, and improves the stability of the receiving loop at the symbol rate switching moment, and the method is applicable to a variety of carrier loop structures, including a pure frequency-locked loop structure, a pure phase-locked loop structure, a frequency-locked loop series phase-locked loop structure, and a frequency-locked loop parallel phase-locked loop structure.

Description

Technical Field

[0001] The present invention relates to the technical field of spread spectrum signal processing, and in particular to a tracking method for a variable symbol rate spread spectrum signal. Background Art

[0002] Compared with the original signal, the spread spectrum communication signal has a wider bandwidth, better anti-interference and anti-noise performance, and can also achieve multiple access, making the communication process more confidential. At present, the spread spectrum communication technology has developed to a relatively mature level and has been widely used in military, civil and commercial fields. In the spread spectrum signal receiver, tracking is the key technology to achieve precise synchronization. Tracking is mainly divided into two parts: carrier tracking and pseudo code tracking. The carrier loop usually adopts a structure combining a frequency locked loop and a phase locked loop, and the pseudo code loop generally uses a delay locked loop structure.

[0003] At present, the communication rate is getting higher and higher. However, if the channel conditions deteriorate, the quality of communication will be greatly reduced. Therefore, scholars have proposed to solve the above problem by changing the communication rate, that is, using a lower communication rate when the signal-to-noise ratio is low, and using a higher communication rate when the signal-to-noise ratio is high, so as to achieve better communication performance.

[0004] In the article "Research on Multipath Reception Technology of Variable Rate Spread Spectrum Communication System" published by Yang Jingfan in 2018, a simplified construction scheme for variable rate spread spectrum communication signals was proposed, that is, the fixed multiple relationship between symbol period and code period is used to control the change of code chip rate. After simplification, the actual independent variable rate control parameter is only the code chip rate. In the variable rate spread spectrum communication system, due to the time-varying code chip rate, the cross-correlation performance of the spread spectrum sequence of the previous and next symbols is seriously deteriorated, and the problem of inter-symbol interference is introduced in multipath interference. An interference cancellation RAKE reception scheme based on FFT is proposed. In the article "Adaptive Variable Rate Data Transmission Model of Parallel Combination Spread Spectrum Communication System" published by Gao Weimin et al. in Volume 43, Issue 13 of "Modern Electronic Technology" in 2020, the working principle of the parallel combination spread spectrum system is introduced. The transmitter selects r of the M PN code sequences and transmits them after being modulated by the carrier at the same time. Therefore, the larger the value of r, the more bits can be transmitted in parallel, and the higher the data transmission rate. In order to solve the problem of low transmission efficiency of parallel combined spread spectrum system at fixed speed, an adaptive variable rate data transmission model combined with channel estimation technology is proposed. When the channel condition changes during the communication process, the signal-to-noise ratio fed back by the channel estimation changes the selected PN code sequence number r, thereby controlling the data transmission rate and improving the data transmission efficiency. In the paper "Research on Variable Rate Communication of Satellite Internet of Things Based on Signal-to-Noise Ratio Estimation" published by Mao Xiwen et al. at the 17th Annual Conference of Satellite Communications in 2021, in order to solve the problem of satellite communication efficiency caused by channel quality changes, an adaptive variable rate spread spectrum communication scheme under signal-to-noise ratio estimation is proposed. The feedback of the current transmission environment is obtained by signal-to-noise ratio estimation, and then the current spread spectrum communication rate is adjusted by adaptive adjustment system. The two adjustment schemes mentioned in the paper include changing the spread spectrum code rate and changing the number of bits of parallel transmission information, so as to maximize the transmission efficiency under the premise that the bit error rate meets the expected requirements.

[0005] The variable rate spread spectrum communication system discussed above, variable rate refers to changing the spread spectrum code rate, the number of PN code sequences r or the number of parallel transmission information bits, and the focus is mostly on the adaptive variable rate transmission problem at the transmitting end, without considering the impact of the symbol rate change on the tracking loop at the receiving end. Therefore, the present invention proposes a method that enables the tracking loop to achieve seamless switching when the symbol rate changes. Summary of the invention

[0006] The purpose of the present invention is to provide a tracking method for a variable symbol rate spread spectrum signal, in which a seamless switching coefficient calculation module and a loop bandwidth adjustment coefficient calculation module are added to the original traditional carrier tracking loop, and the seamless switching coefficient is used to correct the predicted value of the change in the loop filtering result when the symbol rate is switched, and the bandwidth is adjusted by the loop bandwidth adjustment coefficient after the symbol rate is switched, so that the loop jitter can be reduced while the loop always maintains the optimal bandwidth, thereby reducing the difficulty and cost of upgrading the existing loop and improving the stability of the receiving loop at the moment of symbol rate switching.

[0007] To achieve the above object, the present invention provides a tracking method for a variable symbol rate spread spectrum signal, the steps of which are as follows:

[0008] S1, using the sine and cosine signals generated by the local carrier NCO and the leading, immediate and lagging three-way pseudo codes generated by the local pseudo code generator, the received signal is stripped of the carrier and the pseudo code;

[0009] S2, performing coherent integration on the signal after the pseudo code stripping obtained in step S1, the coherent integration time will change with the change of the symbol rate, and obtain the three-way coherent integration results of leading, immediate and lagging;

[0010] S3, input the coherent integration results of the leading and lagging branches into the pseudo code phase detector to obtain the pseudo code phase detection result, input the pseudo code loop filter, and obtain the filtering result and input it into the local pseudo code generator to generate the local pseudo code of the next cycle;

[0011] Input the coherent integration result of the instantaneous branch into the carrier discriminator, obtain the discriminant result and input it into the carrier loop filter, input the obtained filtering result into the local carrier NCO, and generate the local carrier of the next cycle;

[0012] S4, adding a seamless switching coefficient calculation module to the carrier tracking loop in step S3, when the symbol rate changes:

[0013] In the second-order frequency-locked loop, the seamless switching coefficient calculation module calculates the frequency-locked loop seamless switching coefficient according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, and uses the frequency-locked loop seamless switching coefficient to correct the Doppler frequency change prediction item, and adds the frequency discrimination error adjustment value to obtain the second-order frequency-locked loop filter expression suitable for variable symbol rate;

[0014] In a third-order phase-locked loop, a seamless switching coefficient calculation module calculates a phase-locked loop seamless switching coefficient according to a proportional relationship between a symbol rate before switching and a symbol rate after switching, and uses the phase-locked loop seamless switching coefficient to correct a Doppler angular frequency variation prediction item, and adds a phase-locked error adjustment value, thereby obtaining a third-order phase-locked loop filter expression suitable for a variable symbol rate;

[0015] S5. The carrier loop bandwidth adjustment module calculates the loop bandwidth adjustment coefficients of the second-order frequency-locked loop and the third-order phase-locked loop according to the change in the coherent integration time before and after the symbol rate switching in step S4, uses the loop bandwidth adjustment coefficient to correct the bandwidth before switching to obtain a new bandwidth, and then updates the filter parameters according to the new bandwidth and applies it to the carrier loop.

[0016] Preferably, in step S4, the seamless switching coefficient of the second-order frequency locked loop is calculated according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, including the following steps:

[0017] The digital discrete transfer function of the loop filter of the second-order frequency-locked loop is:

[0018]

[0019] Where ξ is the damping coefficient, the coherent integration time and carrier loop bandwidth corresponding to the Mth symbol rate are T M and B L-M , loop natural frequency K=K d K v is the total loop gain, where K d and K v They are the discriminator gain and NCO gain respectively;

[0020] The corresponding time domain expression is:

[0021]

[0022] Where y(k) is the Doppler frequency filtered value output by the k-th cycle filter, e f (k) is the frequency discrimination result of the kth period;

[0023] According to the acceleration model, the filtering formula of the switching point is analyzed, and the Doppler frequency change rate is assumed to be f a , then formula (2) can be rewritten as:

[0024]

[0025] Formula (3) means that the filter value of the kth cycle of the second-order frequency-locked loop filter is equal to the filter value of the k-1th cycle plus the difference between the filter values ​​of the k-1th and k-2th cycles, plus the frequency discrimination error adjustment value. When the frequency-locked loop is locked, [y(k-1)-y(k-2)] can be regarded as the predicted value of the Doppler frequency change within a coherent integration cycle.

[0026] Assume that in the first M cycles, the symbol rate changes for the Mth time, that is, the coherent integration time changes from T M-1 Change to T M; Since the coherent integration time is usually in the order of ms, the Doppler frequency change rate before and after the switch can be approximately considered unchanged, so in T M-1 The change of Doppler frequency over time is:

[0027] Δf M-1 =f a ·T M-1 (4)

[0028] In T M The change of Doppler frequency over time is:

[0029] Δf M =f a ·T M (5)

[0030] From equations (4) and (5), we can see that after the symbol rate is switched, the Doppler frequency variation within a coherent integration period changes with the coherent integration time. Therefore, the calculation of y(l M ) needs to correct the predicted value of the Doppler frequency change [y(k-1)-y(k-2)] according to the proportional relationship between the Doppler frequency changes before and after the switching. This correction coefficient is the frequency-locked loop seamless switching coefficient.

[0031] The output of each cycle loop filter is the Doppler frequency value at the middle point of the coherent integration cycle. M -1 cycle midpoint to the lth M The Doppler frequency change at the middle point of a cycle is half of T M-1 The Doppler frequency change over time and half of T M The Doppler frequency variation over time is composed of Using the first M -1 to lth M The Doppler frequency variation between the midpoints of a cycle is greater than that of T M-1 The change of Doppler frequency over time, k = l M Time-locked loop seamless switching coefficient α f-M for:

[0032]

[0033] Similarly, using T M The change of Doppler frequency in the time period is greater than that in the previous M -1 to lth M The Doppler frequency change between the middle points of the cycle is calculated to get k = l M +1 when frequency-locked loop seamless switching coefficient α f-M The value of is:

[0034]

[0035] Therefore, the filter expression is written as:

[0036]

[0037] in,

[0038]

[0039]

[0040] Preferably, in step S4, the seamless switching coefficient of the third-order phase-locked loop is calculated according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, including the following steps:

[0041] The digital discrete transfer function of the loop filter of the third-order phase-locked loop is:

[0042]

[0043] Among them, ω np-M is the loop natural frequency corresponding to the Mth symbol rate, K = K d K v is the total loop gain, K d and K v They are phase detector gain and NCO gain respectively;

[0044] The corresponding time domain expression is:

[0045]

[0046] in, is the Doppler angular frequency filtered value output by the kth cycle filter, e p (k) is the phase detection result of the kth cycle;

[0047] The filtering formula of the switching point is also analyzed using the acceleration model. Assuming the Doppler angular frequency change rate is ω a , then formula (10) can be rewritten as:

[0048]

[0049] Formula (11) means that the filter value of the kth cycle of the third-order phase-locked loop filter is equal to the filter value of the k-1th cycle plus the difference between the filter values ​​of the k-1th and k-2th cycles, plus the phase error adjustment value. When the phase-locked loop is locked, It can be regarded as the predicted value of the change of Doppler angular frequency within a coherent integration period;

[0050] Assume that in the first M cycles, the symbol rate changes for the Mth time, that is, the coherent integration time changes from TM-1 Change to T M ; Since the coherent integration time is usually in the order of ms, the Doppler angular frequency change rate before and after the switch can be approximately considered unchanged, so in T M-1 The change of Doppler angular frequency over time is:

[0051] Δω M-1 =ω a ·T M-1 (12)

[0052] In T M The change of Doppler angular frequency over time is:

[0053] Δω M =ω a ·T M (13)

[0054] From equations (12) and (13), it can be seen that after the symbol rate is switched, the Doppler angular frequency variation within a coherent integration period changes with the coherent integration time. Therefore, the calculation It is necessary to correct the predicted value of the Doppler angular frequency change according to the proportional relationship between the Doppler angular frequency change before and after the switch. This correction factor is the phase-locked loop seamless switching factor;

[0055] The output of each cycle loop filter is the Doppler angular frequency value at the middle point of the coherent integration cycle. M -1 cycle midpoint to the lth M The Doppler angular frequency change at the middle point of a cycle is half of T M-1 The change in Doppler angular frequency over time and half of T M The Doppler angular frequency variation over time is composed of Using the first M -1 to lth M The Doppler angular frequency variation between the midpoints of a cycle is greater than that of T M-1 The change of Doppler angular frequency over time is calculated to get k = l M Phase-locked loop seamless switching coefficient α p-M for:

[0056]

[0057] Similarly, using T M The change of Doppler angular frequency in the time period is greater than that in the previous M -1 to lth M The Doppler angular frequency change between the middle points of the cycle is calculated to obtain k = l M +1 when the phase-locked loop seamless switching coefficient α p-M The value of is:

[0058]

[0059] Therefore, the filter expression is written as:

[0060]

[0061] in,

[0062]

[0063]

[0064]

[0065] Preferably, in step S5, the carrier loop bandwidth adjustment module calculates the loop bandwidth adjustment coefficient of the second-order frequency locked loop according to the change of the coherent integration time before and after the symbol rate switching in step S4, including the following steps:

[0066] The optimal loop bandwidth formula for the second-order frequency-locked loop is:

[0067]

[0068] in, is the second-order derivative of the Doppler frequency, C / N0 is the carrier-to-noise ratio, F=1 at high carrier-to-noise ratio, and F=2 at low carrier-to-noise ratio;

[0069] In the formula for the optimal loop bandwidth of the second-order frequency-locked loop, It is approximately considered to be equal to 1, so equation (17) can be approximately expressed as:

[0070]

[0071] In the above formula, in addition to the coherent integration time T M , assuming that other terms do not change before and after the switching, therefore, when the coherent integration time is changed from T M-1 Change to T M When the optimal loop bandwidth changes from B F-(M-1) Change to B F-M , the corresponding proportionality coefficient β M for:

[0072]

[0073] Preferably, in step S5, the carrier loop bandwidth adjustment module calculates the loop bandwidth adjustment coefficient of the third-order phase-locked loop according to the change of the coherent integration time before and after the symbol rate switching in step S4, including the following steps:

[0074] The optimal loop bandwidth formula for a third-order phase-locked loop is:

[0075]

[0076] in, is the second derivative of the Doppler angular frequency;

[0077] From the optimal loop bandwidth formula of the third-order phase-locked loop, we can see that due to the coherent integration time T M Item It is approximately equal to 1, so the effect of changing the coherent integration time on the optimal loop bandwidth of the phase-locked loop is negligible.

[0078] Preferably, in step S1, the received digitized intermediate frequency signal is expressed as:

[0079]

[0080] Where P is the received signal power; D M (·) represents the data code of the Mth symbol rate switched between multiple symbol rates, D M (iT s )=±1;C(·) is the code rate f c C / A code sequence; f d is the Doppler frequency; i is the number of sampling points; T s is the sampling time interval; τ is the received signal code phase delay; f IF is the intermediate frequency of the signal; is the initial phase of the received signal; n0(i) is a Gaussian white noise with a unilateral power spectrum density of N0 and a mean of 0;

[0081] The in-phase and quadrature signals generated by the local carrier NCO are expressed as:

[0082]

[0083]

[0084] in, is the Doppler frequency estimate; is the initial phase estimate of the received signal.

[0085] Preferably, in step S2, after the received signal is subjected to carrier stripping, pseudo code stripping, and integral clearing operations, the co-phase and quadrature coherent integration results are obtained and written as:

[0086]

[0087]

[0088] Where R(·) is the ideal autocorrelation function of the pseudo-random signal; Δτ e is the code phase estimation error; Δf d is the difference between the true value and the estimated value of the Doppler frequency; is the difference between the true phase value and the estimated phase value; k represents the kth coherent integration; T M is the coherent integration time corresponding to the Mth symbol rate, which changes with the switching of the symbol rate; n I (k) and n Q (k) is Gaussian noise, and the noise variance is

[0089] The present invention also provides an overall structure of a tracking method for a variable symbol rate spread spectrum signal, which mainly includes down-conversion, pseudo-code stripping, integral clearing, pseudo-code identification, pseudo-code filtering, carrier identification and carrier filtering, wherein the carrier filtering adds a seamless switching coefficient calculation module and a loop bandwidth adjustment coefficient calculation module.

[0090] Therefore, the present invention adopts the tracking method of the variable symbol rate spread spectrum signal, and its technical effects are as follows:

[0091] (1) The method of the present invention adds a seamless switching coefficient to the original carrier loop filter, so that the jitter caused by the symbol rate change to the loop is greatly reduced.

[0092] (2) The use of the loop bandwidth adjustment coefficient in the present invention ensures that the loop bandwidth is always at the optimal bandwidth after the coherent integration time is changed, thereby improving the steady-state tracking performance of the loop.

[0093] (3) The method of the present invention only adds a seamless switching coefficient calculation module and a loop bandwidth adjustment coefficient calculation module to the original traditional tracking loop, which can improve the problems caused by the symbol rate change, thereby reducing the difficulty and cost of upgrading the existing loop.

[0094] (4) The method described in the present invention is applicable to a variety of carrier loop structures, including a pure frequency-locked loop structure, a pure phase-locked loop structure, a frequency-locked loop-series phase-locked loop structure, and a frequency-locked loop-parallel phase-locked loop structure.

[0095] (5) The method provided by the present invention can be used to update the existing loop after limited improvement on the basis of the existing technology, thereby reducing the upgrade cost, saving expenses, and expanding the application scope of the method.

[0096] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 It is a structural schematic diagram of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention;

[0098] Figure 2 It is a digital implementation block diagram of a loop filter of a second-order frequency-locked loop in an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention;

[0099] Figure 3 It is a schematic diagram of the Doppler frequency variation before and after symbol rate switching of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention;

[0100] Figure 4 It is a digital implementation block diagram of a loop filter of a third-order phase-locked loop in an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention;

[0101] Figure 5 It is a schematic diagram of the Doppler angular frequency variation before and after symbol rate switching in an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention;

[0102] Figure 6 It is a schematic diagram of a frequency tracking curve when a carrier loop of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention does not adopt a seamless switching coefficient;

[0103] Figure 7 It is a schematic diagram of a phase detection result when a carrier loop of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention does not adopt a seamless switching coefficient;

[0104] Figure 8 It is a schematic diagram of a phase filtering result when a carrier loop of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention does not adopt a seamless switching coefficient;

[0105] Fig. 9 It is a schematic diagram of IQ two-way integration clearing results when the carrier loop of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention does not adopt seamless switching coefficients;

[0106] Fig.10 It is a schematic diagram of a frequency tracking curve when a carrier loop of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention adopts a seamless switching coefficient;

[0107] Fig.11 It is a schematic diagram of a phase detection result when a carrier loop of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention adopts a seamless switching coefficient;

[0108] Fig.12 It is a schematic diagram of a phase filtering result when a carrier loop of an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention adopts a seamless switching coefficient;

[0109] Fig.13It is a schematic diagram of IQ two-way integration clearing results when a carrier loop adopts seamless switching coefficients in an embodiment of a tracking method for a variable symbol rate spread spectrum signal of the present invention. DETAILED DESCRIPTION

[0110] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0111] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0112] Embodiment 1

[0113] like Figure 1 As shown, the present invention provides an overall structure of a tracking method for a variable symbol rate spread spectrum signal, which mainly includes down-conversion, pseudo-code stripping, integral clearing, pseudo-code identification, pseudo-code filtering, carrier identification and carrier filtering, wherein the carrier filtering adds a seamless switching coefficient calculation module and a loop bandwidth adjustment coefficient calculation module.

[0114] The present invention provides a tracking method for a variable symbol rate spread spectrum signal, the steps of which are as follows:

[0115] S1, using the sine and cosine signals generated by the local carrier NCO and the leading, immediate and lagging three-way pseudo codes generated by the local pseudo code generator, the received signal is stripped of the carrier and the pseudo code;

[0116] The expression of the received digitized intermediate frequency signal is:

[0117]

[0118] Where P is the received signal power; D M (·) represents the data code of the Mth symbol rate switched between multiple symbol rates, D M (iT s )=±1; C(·) is the C / A code sequence with code rate fc; fd is the Doppler frequency; i is the number of sampling points; T s is the sampling time interval; τ is the received signal code phase delay; f IF is the signal intermediate frequency; is the initial phase of the received signal; n0(i) is a Gaussian white noise with a unilateral power spectrum density of N0 and a mean of 0;

[0119] The in-phase and quadrature signals generated by the local carrier NCO are expressed as:

[0120]

[0121]

[0122] in, is the Doppler frequency estimate; is the initial phase estimate of the received signal.

[0123] S2, performing coherent integration on the signal after the pseudo code stripping obtained in step S1, the coherent integration time will change with the change of the symbol rate, and obtain the three-way coherent integration results of leading, immediate and lagging;

[0124] After the received signal is stripped of carrier, stripped of pseudo code, and cleared of integration, the co-phase and quadrature coherent integration results are written as:

[0125]

[0126]

[0127] Where R(·) is the ideal autocorrelation function of the pseudo-random signal; Δτ e is the code phase estimation error; Δf d is the difference between the true value and the estimated value of the Doppler frequency; is the difference between the true phase value and the estimated phase value; k represents the kth coherent integration; T M is the coherent integration time corresponding to the Mth symbol rate, which changes with the switching of the symbol rate; n I (k) and n Q (k) is Gaussian noise, and the noise variance is

[0128] S3, input the coherent integration results of the leading and lagging branches into the pseudo code phase detector to obtain the pseudo code phase detection result, input the pseudo code loop filter, and obtain the filtering result and input it into the local pseudo code generator to generate the local pseudo code of the next cycle;

[0129] Input the coherent integration result of the instantaneous branch into the carrier discriminator, obtain the discriminant result and input it into the carrier loop filter, input the obtained filtering result into the local carrier NCO, and generate the local carrier of the next cycle;

[0130] S4, adding a seamless switching coefficient calculation module to the carrier tracking loop in step S3, when the symbol rate changes:

[0131] In the second-order frequency-locked loop, the seamless switching coefficient calculation module calculates the frequency-locked loop seamless switching coefficient according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, and uses the frequency-locked loop seamless switching coefficient to correct the Doppler frequency change prediction item, and adds the frequency discrimination error adjustment value to obtain the second-order frequency-locked loop filter expression suitable for variable symbol rate;

[0132] The seamless switching coefficient of the second-order frequency locked loop is calculated according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, including the following steps:

[0133] The digital implementation block diagram of the loop filter of the second-order frequency-locked loop is as follows: Figure 2 As shown, its discrete transfer function is:

[0134]

[0135] Where ξ is the damping coefficient, the coherent integration time and carrier loop bandwidth corresponding to the Mth symbol rate are T M and B L-M , loop natural frequency K=K d K v is the total loop gain, where K d and K v They are the discriminator gain and NCO gain respectively;

[0136] The corresponding time domain expression is:

[0137]

[0138] Where y(k) is the Doppler frequency filtered value output by the k-th cycle filter, e f (k) is the frequency discrimination result of the kth period;

[0139] According to the acceleration model, the filtering formula of the switching point is analyzed. Assuming the Doppler frequency change rate is f a , then formula (7) can be rewritten as:

[0140]

[0141] Formula (8) means that the filter value of the kth cycle of the second-order frequency-locked loop filter is equal to the filter value of the k-1th cycle plus the difference between the filter values ​​of the k-1th and k-2th cycles, plus the frequency discrimination error adjustment value. When the frequency-locked loop is locked, [y(k-1)-y(k-2)] can be regarded as the predicted value of the Doppler frequency change within a coherent integration cycle.

[0142] Assume that in the first M cycles, the symbol rate changes for the Mth time, that is, the coherent integration time is changed from T M-1 Change to T M ; Since the coherent integration time is usually in the order of ms, the Doppler frequency change rate before and after the switch can be approximately considered unchanged, so in T M-1 The change of Doppler frequency over time is:

[0143] Δf M-1 =f a ·TM-1 (9)

[0144] In T M The change of Doppler frequency over time is:

[0145] Δf M =f a ·T M (10)

[0146] From equations (9) and (10), we can see that after the symbol rate is switched, the Doppler frequency variation within a coherent integration period changes with the coherent integration time. Therefore, the calculation of y(l M ) needs to correct the predicted value of the Doppler frequency change [y(k-1)-y(k-2)] according to the proportional relationship between the Doppler frequency changes before and after the switching. This correction coefficient is the frequency-locked loop seamless switching coefficient.

[0147] The output of each cycle loop filter is the Doppler frequency value at the middle point of the coherent integration cycle. M -1 cycle midpoint to the lth M The Doppler frequency change at the middle point of a cycle is half of T M-1 The Doppler frequency change over time and half of T M The Doppler frequency variation over time is composed of Using the first M -1 to lth M The Doppler frequency variation between the midpoints of a cycle is greater than that of T M-1 The change of Doppler frequency over time, k = l M Time-locked loop seamless switching coefficient α f-M for:

[0148]

[0149] Similarly, using T M The change of Doppler frequency in the time period is greater than that in the previous M -1 to lth M The Doppler frequency change between the middle points of the cycle is calculated to get k = l M +1 when frequency-locked loop seamless switching coefficient α f-M The value of is:

[0150]

[0151] Therefore, the filter expression is written as:

[0152]

[0153] in,

[0154]

[0155]

[0156] In a third-order phase-locked loop, a seamless switching coefficient calculation module calculates a phase-locked loop seamless switching coefficient according to a proportional relationship between a symbol rate before switching and a symbol rate after switching, and uses the phase-locked loop seamless switching coefficient to correct a Doppler angular frequency variation prediction item, and adds a phase-locked error adjustment value, thereby obtaining a third-order phase-locked loop filter expression suitable for a variable symbol rate;

[0157] The seamless switching coefficient of the third-order phase-locked loop is calculated according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, including the following steps:

[0158] The digital implementation block diagram of the loop filter of the third-order phase-locked loop is as follows Figure 4 As shown, its discrete transfer function is:

[0159]

[0160] Among them, ω np-M is the loop natural frequency corresponding to the Mth symbol rate, K = K d K v is the total loop gain, K d and K v They are phase detector gain and NCO gain respectively;

[0161] The corresponding time domain expression is:

[0162]

[0163] in, is the Doppler angular frequency filtered value output by the kth cycle filter, e p (k) is the phase detection result of the kth cycle;

[0164] The filtering formula of the switching point is also analyzed using the acceleration model. Assuming the Doppler angular frequency change rate is ω a , then formula (15) can be rewritten as:

[0165]

[0166] Formula (16) means that the filter value of the kth cycle of the third-order phase-locked loop filter is equal to the filter value of the k-1th cycle plus the difference between the filter values ​​of the k-1th and k-2th cycles, plus the phase error adjustment value. When the phase-locked loop is locked, It can be regarded as the predicted value of the change of Doppler angular frequency within a coherent integration period;

[0167] Assume that in the firstM cycles, the symbol rate changes for the Mth time, that is, the coherent integration time is changed from T M-1 Change to T M ; Since the coherent integration time is usually in the order of ms, the Doppler angular frequency change rate before and after the switch can be approximately considered unchanged, so in T M-1 The change of Doppler angular frequency over time is:

[0168] Δω M-1 =ω a ·T M-1 (17)

[0169] In T M The change of Doppler angular frequency over time is:

[0170] Δω M =ω a ·T M (18)

[0171] From equations (17) and (18), it can be seen that after the symbol rate is switched, the Doppler angular frequency variation within a coherent integration period changes with the coherent integration time. Therefore, the calculation It is necessary to correct the predicted value of the Doppler angular frequency change according to the proportional relationship between the Doppler angular frequency change before and after the switch. This correction factor is the phase-locked loop seamless switching factor;

[0172] The output of each cycle loop filter is the Doppler angular frequency value at the middle point of the coherent integration cycle. M -1 cycle midpoint to the lth M The Doppler angular frequency change at the middle point of a cycle is half of T M-1 The change in Doppler angular frequency over time and half of T M The Doppler angular frequency variation over time is composed of Using the first M -1 to lth M The Doppler angular frequency variation between the midpoints of a cycle is greater than that of T M-1 The change of Doppler angular frequency over time is calculated to get k = l M Phase-locked loop seamless switching coefficient α p-M for:

[0173]

[0174] Similarly, using T M The change of Doppler angular frequency in the time period is greater than that in the previous M -1 to lth M The Doppler angular frequency change between the middle points of the cycle is calculated to obtain k = l M+1 when the phase-locked loop seamless switching coefficient α p-M The value of is:

[0175]

[0176] Therefore, the filter expression is written as:

[0177]

[0178] in,

[0179]

[0180]

[0181]

[0182] The following simulation analysis of the carrier loop structure of the frequency-locked loop in series with the phase-locked loop shows how the use of the seamless switching coefficient can improve the loop jitter at the switching moment. In the loop structure of the frequency-locked loop in series with the phase-locked loop, most of the acceleration and speed dynamics are eliminated by the frequency-locked loop. After the frequency-locked loop is locked, the remaining phase error is eliminated by the phase-locked loop. However, this part of the phase difference fluctuates around a certain constant and does not conform to the acceleration motion model. Therefore, the carrier loop only needs to use the frequency-locked loop seamless switching coefficient.

[0183] Set the simulation signal-to-noise ratio to -20 dB, the Doppler frequency change rate to 2.5 kHz / s, and assume that the coherent integration time changes from 5 ms to 1.25 ms in the 1000th cycle. According to equations (11) and (12), the frequency-locked loop seamless switching coefficient at this time is calculated as α f-M =0.625(k=l M ) and α f-M =0.4(k=l M +1). If the frequency-locked loop seamless switching coefficient is not used, the loop will lose lock directly, such as Figure 6 , 7 , 8, and 9; after the frequency-locked loop seamlessly switches the coefficient, the loop can basically achieve a smooth transition, as shown in Fig.10 , 11 , 12, and 13.

[0184] S5. The carrier loop bandwidth adjustment module calculates the loop bandwidth adjustment coefficients of the second-order frequency-locked loop and the third-order phase-locked loop according to the change in the coherent integration time before and after the symbol rate switching in step S4, uses the loop bandwidth adjustment coefficient to correct the bandwidth before switching to obtain a new bandwidth, and then updates the filter parameters according to the new bandwidth and applies it to the carrier loop.

[0185] The optimal loop bandwidth formula for the second-order frequency-locked loop is:

[0186]

[0187] in, is the second-order derivative of the Doppler frequency, C / N0 is the carrier-to-noise ratio, F=1 at high carrier-to-noise ratio, and F=2 at low carrier-to-noise ratio;

[0188] The optimal loop bandwidth formula for a third-order phase-locked loop is:

[0189]

[0190] in, is the second derivative of the Doppler angular frequency;

[0191] According to the optimal loop bandwidth formula of the second-order frequency-locked loop and the third-order phase-locked loop, it can be seen that the optimal loop bandwidth is related to the carrier-to-noise ratio C / N0 and the coherent integration time T M And the second-order rate of change of Doppler frequency.

[0192] In the expression of the optimal loop bandwidth of the second-order frequency-locked loop mentioned in equation (22), due to T M C / N0>>1, It is approximately considered to be equal to 1, so formula (22) can be approximately expressed as:

[0193]

[0194] In formula (24), in addition to the coherent integration time T M , assuming that other terms do not change before and after the switching, therefore, when the coherent integration time is changed from T M-1 Change to T M When the optimal loop bandwidth changes from B F-(M-1) Change to B F-M , the corresponding proportionality coefficient β M for:

[0195]

[0196] From the optimal loop bandwidth expression of the third-order phase-locked loop in equation (23), it can be seen that as the coherent integration time T M As the coherent integration time T increases, the optimal loop bandwidth of the phase-locked loop increases accordingly. M As the coherent integration time T decreases, the optimal loop bandwidth of the phase-locked loop will also decrease. M Item 2T M ·C / N0>>1, thus making It is approximately equal to 1, so the effect of changing the coherent integration time on the optimal loop bandwidth of the phase-locked loop can be ignored.

[0197] Similarly, the carrier loop structure of the frequency-locked loop in series with the phase-locked loop is simulated to analyze the improvement of the steady-state tracking performance by using the loop bandwidth adjustment coefficient. Taking the coherent integration time of 5ms and 1.25ms as an example, the simulation signal-to-noise ratio is set to -20dB, and the loop bandwidth adjustment coefficient β at this time is calculated according to formula (25): M is 0.574, when the coherent integration time is 5ms, the FLL bandwidth is set to 3Hz; when the coherent integration time becomes 1.25ms, the FLL bandwidth is adjusted to 1.7Hz according to the loop bandwidth adjustment coefficient, and the PLL bandwidth does not change. The statistical results of the variance of the steady-state frequency tracking error in the two cases of not adjusting the FLL bandwidth and adjusting the FLL bandwidth according to the loop bandwidth adjustment coefficient are shown in Table 1 below:

[0198] Table 1 Comparison of the variance statistics of the steady-state frequency tracking error under two conditions

[0199] <![CDATA[Variance of steady-state frequency tracking error / Hz 2 > The frequency locked loop FLL bandwidth is not adjusted 0.0372073 Frequency Locked Loop (FLL) Bandwidth Adjustment 0.0277623

[0200] It can be seen from the table that adjusting the frequency locked loop FLL bandwidth according to the loop bandwidth adjustment coefficient reduces the variance of the steady-state frequency tracking error and improves the steady-state tracking performance of the carrier loop.

[0201] Therefore, the present invention adopts the above-mentioned tracking method of a variable symbol rate spread spectrum signal, which not only reduces the jitter of the loop when the symbol rate is switched, but also adjusts the loop bandwidth according to the change of the coherent integration time, thereby improving the overall performance of the tracking loop.

[0202] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for tracking a variable symbol rate spread spectrum signal, characterized in that: Here are the steps: S1, using the sine and cosine signals generated by the local carrier NCO and the leading, immediate and lagging three-way pseudo codes generated by the local pseudo code generator, the received signal is stripped of the carrier and the pseudo code; S2, performing coherent integration on the signal after the pseudo code stripping obtained in step S1, the coherent integration time will change with the change of the symbol rate, and obtain the three-way coherent integration results of leading, immediate and lagging; S3, input the coherent integration results of the leading and lagging branches into the pseudo code phase detector to obtain the pseudo code phase detection result, input the pseudo code loop filter, and obtain the filtering result and input it into the local pseudo code generator to generate the local pseudo code of the next cycle; Input the coherent integration result of the instantaneous branch into the carrier discriminator, obtain the discriminant result and input it into the carrier loop filter, input the obtained filtering result into the local carrier NCO, and generate the local carrier of the next cycle; S4, adding a seamless switching coefficient calculation module to the carrier tracking loop in step S3, when the symbol rate changes: In the second-order frequency-locked loop, the seamless switching coefficient calculation module calculates the frequency-locked loop seamless switching coefficient according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, and uses the frequency-locked loop seamless switching coefficient to correct the Doppler frequency change prediction item, and adds the frequency discrimination error adjustment value to obtain the second-order frequency-locked loop filter expression suitable for variable symbol rate; In a third-order phase-locked loop, a seamless switching coefficient calculation module calculates a phase-locked loop seamless switching coefficient according to a proportional relationship between a symbol rate before switching and a symbol rate after switching, and uses the phase-locked loop seamless switching coefficient to correct a Doppler angular frequency variation prediction item, and adds a phase-locked error adjustment value, thereby obtaining a third-order phase-locked loop filter expression suitable for a variable symbol rate; S5. The carrier loop bandwidth adjustment module calculates the loop bandwidth adjustment coefficients of the second-order frequency-locked loop and the third-order phase-locked loop according to the change in the coherent integration time before and after the symbol rate switching in step S4, uses the loop bandwidth adjustment coefficient to correct the bandwidth before switching to obtain a new bandwidth, and then updates the filter parameters according to the new bandwidth and applies it to the carrier loop.

2. The tracking method of a variable symbol rate spread spectrum signal according to claim 1, characterized in that: In step S4, the seamless switching coefficient of the second-order frequency locked loop is calculated according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, including the following steps: The digital discrete transfer function of the loop filter of the second-order frequency-locked loop is: Where ξ is the damping coefficient, the coherent integration time and carrier loop bandwidth corresponding to the Mth symbol rate are T M and B L-M , loop natural frequency K=K d K v is the total loop gain, where K d and K v They are the discriminator gain and NCO gain respectively; The corresponding time domain expression is: Where y(k) is the Doppler frequency filtered value output by the k-th cycle filter, e f (k) is the frequency discrimination result of the kth period; According to the acceleration model, the filtering formula of the switching point is analyzed, and the Doppler frequency change rate is assumed to be f a , then formula (2) can be rewritten as: Formula (3) means that the filter value of the kth cycle of the second-order frequency-locked loop filter is equal to the filter value of the k-1th cycle plus the difference between the filter values ​​of the k-1th and k-2th cycles, plus the frequency discrimination error adjustment value. When the frequency-locked loop is locked, [y(k-1)-y(k-2)] can be regarded as the predicted value of the Doppler frequency change within a coherent integration cycle. Assume that in the first M cycles, the symbol rate changes for the Mth time, that is, the coherent integration time is changed from T M-1 Change to T M ; Since the coherent integration time is usually in the order of ms, the Doppler frequency change rate before and after the switch can be approximately considered unchanged, so in T M-1 The change of Doppler frequency over time is: Δf M-1 =f a ·T M-1 (4) In T M The change of Doppler frequency over time is: Δf M =f a ·T M (5) From equations (4) and (5), it can be seen that after the symbol rate is switched, the Doppler frequency variation within a coherent integration period changes with the coherent integration time. Therefore, the calculation of y(l M ) needs to correct the predicted value of the Doppler frequency change [y(k-1)-y(k-2)] according to the proportional relationship between the Doppler frequency changes before and after the switching. This correction coefficient is the frequency-locked loop seamless switching coefficient. The output of each cycle loop filter is the Doppler frequency value at the middle point of the coherent integration cycle. M -1 cycle midpoint to the lth M The Doppler frequency change at the middle point of a cycle is half of T M-1 The Doppler frequency change over time and half of T M The Doppler frequency variation over time is composed of Using the first M -1 to lth M The Doppler frequency variation between the midpoints of a cycle is greater than that of T M-1 The change of Doppler frequency over time is calculated as k = l M Time-locked loop seamless switching coefficient α f-M for: Similarly, using T M The change of Doppler frequency in the time period is greater than that in the previous M -1 to lth M The Doppler frequency change between the middle points of the cycle is calculated to get k = l M +1 when frequency-locked loop seamless switching coefficient α f-M The value of is: Therefore, the filter expression is written as: in, 3. The tracking method of a variable symbol rate spread spectrum signal according to claim 1, characterized in that: In step S4, the seamless switching coefficient of the third-order phase-locked loop is calculated according to the proportional relationship between the symbol rate before switching and the symbol rate after switching, including the following steps: The digital discrete transfer function of the loop filter of the third-order phase-locked loop is: Among them, ω np-M is the loop natural frequency corresponding to the Mth symbol rate, K = K d K v is the total loop gain, K d and K v They are phase detector gain and NCO gain respectively; The corresponding time domain expression is: in, is the Doppler angular frequency filtered value output by the kth cycle filter, e p (k) is the phase detection result of the kth cycle; The filtering formula of the switching point is also analyzed using the acceleration model. Assuming the Doppler angular frequency change rate is ω a , then formula (10) can be rewritten as: Formula (11) means that the filter value of the kth cycle of the third-order phase-locked loop filter is equal to the filter value of the k-1th cycle plus the difference between the filter values ​​of the k-1th and k-2th cycles, plus the phase error adjustment value. When the phase-locked loop is locked, It can be regarded as the predicted value of the change of Doppler angular frequency within a coherent integration period; Assume that in the first M cycles, the symbol rate changes for the Mth time, that is, the coherent integration time is changed from T M-1 Change to T M ; Since the coherent integration time is usually in the order of ms, the Doppler angular frequency change rate before and after the switch can be approximately considered unchanged, so in T M-1 The change of Doppler angular frequency over time is: Give M-1 =ω a ·T M-1 (12) In T M The change of Doppler angular frequency over time is: Give M =ω a ·T M (13) From equations (12) and (13), it can be seen that after the symbol rate is switched, the Doppler angular frequency variation within a coherent integration period changes with the coherent integration time. Therefore, the calculation It is necessary to correct the predicted value of the Doppler angular frequency change according to the proportional relationship between the Doppler angular frequency change before and after the switch. This correction factor is the phase-locked loop seamless switching factor; The output of each cycle loop filter is the Doppler angular frequency value at the middle point of the coherent integration cycle. M -1 cycle midpoint to the lth M The Doppler angular frequency change at the middle point of a cycle is half of T M-1 The change in Doppler angular frequency over time and half of T M The Doppler angular frequency variation over time is composed of Using the first M -1 to lth M The Doppler angular frequency variation between the midpoints of a cycle is greater than that of T M-1 The change of Doppler angular frequency over time is calculated to get k = l M Phase-locked loop seamless switching coefficient α p-M for: Similarly, using T M The change of Doppler angular frequency in the time period is greater than that in the previous M -1 to lth M The Doppler angular frequency change between the middle points of the cycle is calculated to obtain k = l M +1 when the phase-locked loop seamless switching coefficient α p-M The value of is: Therefore, the filter expression is written as: in, 4. The tracking method of a variable symbol rate spread spectrum signal according to claim 1, characterized in that: In step S5, the carrier loop bandwidth adjustment module calculates the loop bandwidth adjustment coefficient of the second-order frequency-locked loop according to the change of the coherent integration time before and after the symbol rate switching in step S4, including the following steps: The optimal loop bandwidth formula for the second-order frequency-locked loop is: in, is the second-order derivative of the Doppler frequency, C / N0 is the carrier-to-noise ratio, F=1 at high carrier-to-noise ratio, and F=2 at low carrier-to-noise ratio; In the formula for the optimal loop bandwidth of the second-order frequency-locked loop, due to T M C / N0>>1, It is approximately considered to be equal to 1, so equation (17) can be approximately expressed as: In the above formula, in addition to the coherent integration time T M , assuming that other terms do not change before and after the switching, therefore, when the coherent integration time is changed from T M-1 Change to T M When the optimal loop bandwidth changes from B F-(M-1) Change to B F-M , the corresponding proportionality coefficient β M for:

5. The tracking method of a variable symbol rate spread spectrum signal according to claim 1, characterized in that: In step S5, the carrier loop bandwidth adjustment module calculates the loop bandwidth adjustment coefficient of the third-order phase-locked loop according to the change of the coherent integration time before and after the symbol rate switching in step S4, including the following steps: The optimal loop bandwidth formula for a third-order phase-locked loop is: in, is the second derivative of the Doppler angular frequency; From the optimal loop bandwidth formula of the third-order phase-locked loop, we can see that due to the coherent integration time T M Item 2T M C / N0>>1, It is approximately equal to 1, so the effect of changing the coherent integration time on the optimal loop bandwidth of the phase-locked loop is negligible.

6. The tracking method of a variable symbol rate spread spectrum signal according to claim 1, characterized in that: In step S1, the received digitized intermediate frequency signal is expressed as: Where P is the received signal power; D M (·) represents the data code of the Mth symbol rate switched between multiple symbol rates, D M (iT s )=±1;C(·) is the code rate f c C / A code sequence; f d is the Doppler frequency; i is the number of sampling points; T s is the sampling time interval; τ is the received signal code phase delay; f IF is the signal intermediate frequency; is the initial phase of the received signal; n0(i) is a Gaussian white noise with a unilateral power spectrum density of N0 and a mean of 0; The in-phase and quadrature signals generated by the local carrier NCO are expressed as: in, is the Doppler frequency estimate; is the initial phase estimate of the received signal.

7. The method for tracking a variable symbol rate spread spectrum signal according to claim 1, characterized in that: In step S2, after the received signal is subjected to carrier stripping, pseudo code stripping, and integral clearing operations, the co-phase and quadrature coherent integration results are obtained and written as: Where R(·) is the ideal autocorrelation function of the pseudo-random signal; Δτ e is the code phase estimation error; Δf d is the difference between the true value and the estimated value of the Doppler frequency; is the difference between the true phase value and the estimated phase value; k represents the kth coherent integration; T M is the coherent integration time corresponding to the Mth symbol rate, which changes with the switching of the symbol rate; n I (k) and n Q (k) is Gaussian noise, and the noise variance is

Images