Extremely simple synchronization scheme based on double correlation operators
By adopting a minimalist synchronization scheme of dual correlation operators in short burst communication, autocorrelation frequency deviation estimation and block-related data compensation are realized, which solves the problem of limited pilot resources and storage resources, and reduces pilot overhead and Storage space occupies, improving system performance.
Patent Information
- Application Number
- CN202411277680.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-12
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-09-12
AI Technical Summary
When pilot resources and storage resources are limited, traditional synchronization solutions are difficult to achieve effective frequency-phase decoupling and data compensation in short burst communication, resulting in low carrier frequency deviation and phase deviation estimation accuracy, and data compensation occupies storage The space is large, which increases the difficulty of hardware implementation.
A minimalist synchronization scheme based on dual correlation operators is proposed, including autocorrelation frequency deviation estimation and block-related data compensation. Through frequency-phase decoupling estimation and block-related data compensation, it reduces pilot overhead and storage space usage and reduces The dependence of frequency deviation on phase deviation estimation.
This solution achieves a code error performance close to the traditional synchronization solution under limited pilot overhead, reducing storage space usage and reducing the maximum accumulated phase caused by residual frequency bias in each data block.
Smart Images

Figure CN119071131A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of short burst communications such as narrowband Internet of Things, radar, cellular / satellite Internet of Things, and more particularly to a minimalist synchronization scheme based on a dual correlation operator. Background Art
[0002] At present, short burst communication has been widely used in modern communication systems, such as narrowband Internet of Things, radar, cellular / satellite Internet of Things, etc. However, limited pilot resources and storage resources will bring great challenges to traditional synchronization schemes and their hardware implementation. Therefore, it is particularly important to design a minimalist synchronization scheme to support reliable short burst communication under resource-constrained conditions.
[0003] For short burst communications using low-cost oscillators, the corresponding transceivers will be affected by large carrier frequency and phase offsets, which results in that the frequency offset estimation of the traditional synchronization scheme with the assistance of a small number of pilots will seriously affect the phase offset estimation and the data compensation occupies a large storage space. Hosseini and Perrins proposed a pilot-assisted carrier synchronization scheme for short burst shaping offset quadrature phase shift keying transmission systems in the paper "Burst-mode synchronization for SOQPSK". Considering the preamble structure in the LoRa (Long Range) protocol, Tapparal, Afisiadis, Mayoraz et al. proposed a joint delay, carrier frequency offset and phase offset synchronization scheme in the paper "An open-source LoRa physical layer prototype on GNU radio". For wideband linear frequency modulation radar systems, Yuan, He, Zhang et al. designed an effective joint frequency and phase offset estimation in the paper "Instantaneous velocity estimation based on envelope cross correlation andjoint frequency-phase estimation for wideband LFM radar" using the frequency peak of envelope cross correlation as input. Considering the orthogonal time-frequency modulation system of low-orbit satellites, Li, Zhang, Ju, et al. used pilot assistance in the article "Downlink carrier frequency offset estimation for OTFS-based LEO satellite communication system" to accurately estimate the carrier frequency and phase offset. In addition, in response to the large Doppler frequency shift and phase offset problems in low-orbit satellite communication systems, Zhang Jiayi, Yu Zhongyang, Zhu Min, et al. designed a pilot-assisted carrier synchronization scheme in the article "Design of carrier synchronization algorithm for SCMA system in low-orbit satellite communication". However, these synchronization schemes all consider the so-called frequency-phase coupling estimation, that is, the frequency offset estimator is executed before the phase offset estimator, then the performance of the phase offset estimator depends on the accuracy of the frequency offset estimator, which is the coupling effect of frequency offset on phase offset. In order to weaken this coupling effect, frequency-phase coupling estimation often requires the use of a large pilot overhead, but this is not conducive to short burst communication.Rice pointed out in the article "Carrier-phase and frequency-estimation bounds for transmissions with embedded reference symbols" that in frequency-phase coupling estimation, as long as the initial sampling moment is set at the middle position of the preamble structure, the Cramer-Rao bounds of the frequency offset estimation will be decoupled from the Cramer-Rao bounds of the phase offset estimation. However, since the receiving end does not know the position of the initial sampling moment, it is impossible to set the initial position artificially. On the other hand, after the frequency-phase coupling estimation, most of them will use continuous moment data compensation, that is, the received data is compensated for the frequency and phase offset moment by moment. This data compensation method can obtain good demodulation performance, but its disadvantage is that it requires a large storage space (proportional to the number of transmitted data symbols), which is bound to increase the difficulty of hardware implementation of burst communication systems. Summary of the invention
[0004] In view of the problem that the prior art has poor performance under the dual limitations of pilot resources and storage resources, the present invention proposes a minimalist synchronization scheme based on a dual correlation operator, which includes two parts: frequency-phase decoupling estimation and block-related data compensation. The first part includes an autocorrelation frequency offset estimation that is insensitive to fading and a maximum likelihood phase offset estimation that is low in complexity and resistant to large frequency offsets. The second part can significantly reduce the storage space occupancy and greatly reduce the maximum accumulated phase caused by the residual frequency offset in each data block.
[0005] In order to achieve the above object, the technical solution of the present invention is achieved as follows:
[0006] A minimalist synchronization scheme based on a dual correlation operator includes the following steps:
[0007] S1: Get the received pilot sequence r p And receive the pilot sequence r p Perform a demodulation operation to obtain a demodulated sequence v;
[0008] S2: Perform a generalized correlation operation on the obtained demodulated sequence v to obtain the autocorrelation operator R(β 11 |v1) and the cross-correlation operator R(β|v);
[0009] S3: For the obtained autocorrelation operator R(β 11 |v1) Perform frequency-phase decoupling estimation to obtain the frequency offset estimate and phase bias estimates
[0010] S4: The obtained frequency offset estimate and phase bias estimates Compensate to the cross-correlation operator R(β|v) and the received data sequence r dThe corrected cross-correlation operator R(β|v′) and the corrected received data sequence are obtained ;
[0011] S5: The obtained corrected cross-correlation operator R(β|v′) and the corrected received data sequence separating into a plurality of corrected cross-correlation values and a plurality of corrected received data blocks;
[0012] S6: performing block correlation data compensation on the obtained multiple correction mutual correlation values and the multiple correction received data blocks to obtain multiple secondary correction received data blocks;
[0013] S7: Combine the obtained multiple secondary correction receiving data blocks into a complete receiving data sequence r d ′.
[0014] Step S1 of obtaining the received pilot sequence r p The method is:
[0015] S1.1: Consider a Rice flat fading single-carrier communication system, the corresponding received signal r = {r(k), k∈κ p ∪κ d} and the received signal r(k) in the kth symbol period is expressed as:
[0016]
[0017] Where h is the channel gain, whose amplitude follows the Rice distribution determined by the Rice factor, T s is the symbol period, f d and θ are the carrier frequency deviation and phase deviation respectively, s(k) is the energy-normalized modulation signal, n(k) is a complex Gaussian random variable, is an imaginary unit, There are m pilot blocks p1~p m The sampling index set of is m-1 data blocks d1~d m-1 The sampling index set, and the i-th pilot block p i The sampling index of and the lth data block d l The sampling index of The specific forms are:
[0018]
[0019] in, Respectively represent the lengths of the lth, ith, lth, and tth pilot blocks; Respectively represent the length of the lth and tth data blocks; according to the m pilot blocks p1~p m The sampling index set κ pGet the received pilot sequence r p .
[0020] According to the m pilot blocks p1~p m The sampling index set κ p Get the received pilot sequence r p The method is:
[0021] S1.2: Traverse m pilot blocks p1~p at time k m The sampling index set κ p , the corresponding received pilot sequence r p = {r p (k),k∈κ p}, and the received pilot signal r in the kth symbol period p (k) is expressed as:
[0022]
[0023] In the formula, s p (k) is the modulated pilot signal in the kth symbol period.
[0024] The method of the demodulation operation described in step S1 is:
[0025] S1.3: Receive the pilot signal r p (k) and the modulated pilot signal s p The demodulated signal v(k) is obtained by multiplying the conjugate form of (k):
[0026] v(k)=r p (k) s p (k) *
[0027] In the formula, * is the conjugate operation;
[0028] S1.4: When time k traverses the pilot block sampling index set κ p When the demodulated signal v(k) forms the demodulated sequence
[0029]
[0030] The method of generalizing the related operations described in step S2 is:
[0031] S2.1: Demodulation sequence using the first pilot block p1 The conjugate form of the demodulated signal v(k) corresponding to the pilot block p1 is conjugated with its delayed signal v(k+β 11 ) is multiplied to obtain the autocorrelation operator R(β 11 |v1):
[0032]
[0033] In the formula, is the length of pilot block p1, represents the autocorrelation delay length;
[0034] S2.2: Using the demodulation sequence of the first pilot block p1 and other m-1 pilot blocks p2~p m Demodulation sequence The pilot blocks p2~p m The corresponding conjugate form of the demodulated signal v(k) and its delayed signal v(k+β 1i ) are multiplied to obtain the cross-correlation operator R(β|v)={R(β 12 |v 12 ),…,R(β 1i |v 1i ),...,R(β 1m |v 1m )}, the i-th (i=2,3,...,m) cross-correlation value R(β 1i |v 1i )for:
[0035]
[0036] In the formula, the modulation sequence Demodulation Sequence v 1i is the demodulation sequence v1 of pilot block p1 and the i The demodulation sequence v i The union of is the cross-correlation delay length.
[0037] The method for frequency-phase decoupling estimation described in step S3 is:
[0038] S3.1: Using the autocorrelation operator R(β 11 |v1), the frequency offset estimate is obtained by taking the phase angle operation
[0039]
[0040] Where arg{·} is the argument operation;
[0041] S3.2: Using the autocorrelation operator By comparing the demodulated signal v(k) corresponding to the pilot block p1 with the autocorrelation operator The conjugate form of is summed and then the phase angle is calculated to obtain the phase deviation estimate represents a special autocorrelation delay length and Biased estimate The calculation formula is:
[0042]
[0043] Wherein, λ=1,2,... is the decoupling coefficient.
[0044] The method for obtaining the corrected cross-correlation operator R(β|v′) in step S4 is:
[0045] S4.1: The frequency offset estimate and phase bias estimates Compensation is added to the cross-correlation operator R(β|v) to obtain the corrected cross-correlation operator R(β|v′)={R(β 12 |v′ 12 ),...,R(β 1i |v′ 1i ),...,R(β 1m |v′ 1m )}, and the i-th corrected cross-correlation value R(β 1i |v′ 1i )for:
[0046]
[0047] In the formula, the modulation sequence is the result of compensating the frequency and phase deviation estimates of v(k), and the demodulation sequence For v(k+β 1i )v(k) is the result after frequency offset and phase offset estimation are compensated, v′ 1i is the demodulation sequence v1′ of pilot block p1 and the i The demodulation sequence v i ′, the union of;
[0048] The corrected received data sequence obtained in step S4 The method is:
[0049] S4.2: The frequency offset estimate and phase bias estimates Compensation to the received data sequence r d The corrected received data sequence is obtained from in, And the corrected received data signal r′(k) in the kth symbol period is:
[0050]
[0051] Step S5 is to obtain the corrected cross-correlation operator R(β|v′) and the corrected received data sequence. The method of separating into a plurality of correction cross-correlation values and a plurality of correction received data blocks is:
[0052] S5.1: Sampling index sets by data block Separate the corrected cross-correlation operator R(β|v′) into m-1 corrected cross-correlation values R(β 12 |v′ 12 ),R(β 13 |v′ 13 ),...,R(β 1m |v′ 1m );
[0053] S5.2: Sampling index sets by data block The received data sequence will be corrected Split into m-1 correction receiving data blocks
[0054] The method for compensating the block-related data in step S6 is:
[0055] The m-1 corrected cross-correlation values R(β 12 |v′ 12 ),R(β 13 |v′ 13 ),...,R(β 1m |v′ 1m ) and m-1 corrected received data blocks Multiply them respectively to get m-1 secondary correction receiving data blocks And the lth secondary correction received data block The secondary correction received data signal in the kth symbol period Has the following form:
[0056]
[0057] Step S7 combines the obtained multiple secondary corrected received data blocks into a complete received data sequence r d The method of ' is:
[0058] Sampling index collections by data block m-1 secondary correction received data blocks Synthesize a complete receive data sequence
[0059] The beneficial effects of the present invention are:
[0060] 1. Reducing pilot overhead: Compared with the frequency-phase coupling estimation of the traditional synchronization scheme, the frequency-phase decoupling estimation of the present invention can weaken the dependence of the phase offset estimation on the frequency offset estimation performance, thereby reducing the pilot overhead.
[0061] 2. Reduce storage space usage: Compared with the continuous moment data compensation of the traditional synchronization scheme, the block-related data compensation of the present invention can reduce the number of data compensation values, thereby reducing storage space usage.
[0062] 3. Guaranteeing system performance: Under limited pilot overhead, the scheme of the present invention can achieve error performance close to that of the traditional synchronization scheme. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0064] Figure 1 This is a flowchart of the implementation of the extremely simple synchronization solution based on the dual correlation operator proposed in the present invention.
[0065] Figure 2 It is the estimation performance of the autocorrelation frequency offset estimator in the present invention.
[0066] Figure 3 It is the frequency offset resistance capability of the maximum likelihood phase offset estimator under the decoupling coefficient λ=1 in the present invention.
[0067] Figure 4 is the frequency offset resistance of the maximum likelihood phase offset estimator under the decoupling coefficient λ=2 in the present invention.
[0068] Figure 5 The performance comparison of block-related data compensation and continuous-time data compensation in the present invention is shown in FIG. DETAILED DESCRIPTION
[0069] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0070] like Figure 1 As shown, a minimalist synchronization scheme based on a dual correlation operator includes the following steps:
[0071] S1: Get the received pilot sequence r p And receive the pilot sequence r p Perform a demodulation operation to obtain a demodulated sequence v.
[0072] Get the received pilot sequence r p The method is: traverse the pilot block sampling index set at the sampling time of the received signal to obtain a received pilot sequence.
[0073] Step S1.1: For short burst communication systems such as narrowband IoT, radar, cellular / satellite IoT, etc., consider the Rice flat fading transmission channel, and the corresponding received signal r = {r(k), k∈κ p ∪κ d} and the received signal r(k) in the kth symbol period is expressed as:
[0074]
[0075] Where h is the channel gain, whose amplitude follows the Rice distribution determined by the Rice factor, T s is the symbol period, f d and θ are the carrier frequency deviation and phase deviation respectively, s(k) is the energy-normalized modulation signal, n(k) is a complex Gaussian random variable, is an imaginary unit, There are m pilot blocks p1~p m The sampling index set of Data blocks d1~d m-1 The sampling index set, and the i-th pilot block p i The sampling index of and the lth data block d l The sampling index of The specific forms are:
[0076]
[0077]
[0078] In the formula, Respectively represent the lengths of the lth, ith, lth, and tth pilot blocks; Respectively represent the length of the lth and tth data blocks.
[0079] Step S1.2: Consider time k and traverse the pilot block sampling index set κ p , the corresponding received pilot sequence r p = {r p (k),k∈κ p}, and the received pilot signal r in the kth symbol period p (k) is expressed as:
[0080]
[0081] In the formula, sp (k) is the modulated pilot signal in the kth symbol period.
[0082] The method of the demodulation operation described in step S1 is:
[0083] Step S1.3: Receive the pilot signal r p (k) and the modulated pilot signal s p The demodulated signal v(k) is obtained by multiplying the conjugate form of (k):
[0084]
[0085] In the formula, * is the conjugate operation, n′(k)=n(k)·s p (k) * is Gaussian noise, because s p (k) is a known modulated pilot signal.
[0086] Step S1.4: At time k, traverse the pilot block sampling index set κ p When the demodulated signal v(k) forms the demodulated sequence
[0087]
[0088] S2: Perform a generalized correlation operation on the obtained demodulated sequence v to obtain the autocorrelation operator R(β 11 |v1) and the cross-correlation operator R(β|v), where the autocorrelation operator is used to construct the frequency-phase decoupling estimation in the minimalist synchronization scheme, and the cross-correlation operator is used to construct the block-correlation data compensation in the minimalist synchronization scheme.
[0089] The method of generalizing the related operations described in step S2 is:
[0090] Step S2.1: Use the demodulation sequence of the first pilot block p1 The conjugate form of the demodulated signal v(k) corresponding to the pilot block p1 is conjugated with its delayed signal v(k+β 11 ) is multiplied to obtain the autocorrelation operator R(β 11 |v1):
[0091]
[0092] In the formula, is the length of pilot block p1, represents the autocorrelation delay length, ψ=E{|h| 2} is the equivalent average energy of the received pilot signal, E represents the expected operation, h is the channel gain, ξ(β 11 ) is the noise term and can be expressed as:
[0093]
[0094] In the formula, n′(k+β 11 ) is the delayed signal of Gaussian noise n′(k),
[0095] Step S2.2: Use the demodulation sequence of the first pilot block p1 and other m-1 pilot blocks p2~p m Demodulation sequence The pilot blocks p2~p m The corresponding conjugate form of the demodulated signal v(k) and its delayed signal v(k+β 1i ) are multiplied to obtain the cross-correlation operator R(β|v)={R(β 12 |v 12 ),...,R(β 1i |v 1i ),...,R(β 1m |v 1m )}, and the i-th (i=2,3,...,m) cross-correlation value R(β 1i |v 1i )for:
[0096]
[0097] In the formula, the modulation sequence Demodulation Sequence v 1i is the demodulation sequence v1 of pilot block p1 and the i The demodulation sequence v i The union of is the cross-correlation delay length, ξ(β 1i ) is the noise term, which can be expressed as:
[0098]
[0099] In the formula, n′(k+β 1i ) is the delayed signal of Gaussian noise n′(k).
[0100] S3: For the obtained autocorrelation operator R(β 11 |v1) Perform frequency-phase decoupling estimation to obtain the frequency offset estimate and phase bias estimates
[0101] The method for frequency-phase decoupling estimation described in step S3 is:
[0102] Step S3.1: Using the autocorrelation operator R(β 11|v1), the frequency offset estimate is obtained by taking the phase angle operation
[0103]
[0104] Where arg{·} is the argument operation;
[0105] Step S3.2: Using the autocorrelation operator By comparing the demodulated signal v(k) corresponding to the pilot block p1 with the autocorrelation operator The conjugate form of is summed and then the phase angle is calculated to obtain the phase deviation estimate represents a special autocorrelation delay length and
[0106]
[0107] Where λ=1,2,... is the decoupling coefficient and ξ is the noise term, and has the following form:
[0108]
[0109] S4: The obtained frequency offset estimate and phase bias estimates Compensate to the cross-correlation operator R(β|v) and the received data sequence r d The corrected cross-correlation operator R(β|v′) and the corrected received data sequence are obtained The purpose of the compensation operation is to reduce the accumulated phase caused by frequency offset and phase offset in the cross-correlation operator and the received data sequence, thereby facilitating the subsequent block-related data compensation.
[0110] The specific method is:
[0111] Step S4.1: Set the frequency offset estimate and phase bias estimates Compensation is added to the cross-correlation operator R(β|v) to obtain the corrected cross-correlation operator R(β|v′)={R(β 12 |v′ 12 ),...,R(β 1i |v′ 1i ),...,R(β 1m |v′ 1m )}, and the i-th (i=2,3,...,m) corrected cross-correlation value R(β 1i |v′ 1i )for:
[0112]
[0113] In the formula, the modulation sequence is the result of compensating the frequency and phase deviation estimates of v(k), and the demodulation sequence For v(k+β 1i )v(k) is the result after frequency offset and phase offset estimation are compensated, v′ 1i is the demodulation sequence v1′ of pilot block p1 and the i The demodulation sequence v i ′, To estimate the residual frequency offset after compensation, ξ(β 1i ) is the noise term.
[0114] Step S4.2: Set the frequency offset estimate and phase bias estimates Compensation to the received data sequence r d The corrected received data sequence is obtained from in, And the corrected received data signal r′(k) in the kth symbol period has the following form:
[0115]
[0116] In the formula, is the residual phase deviation, It is still a Gaussian noise.
[0117] S5: The obtained corrected cross-correlation operator R(β|v′) and the corrected received data sequence Separated into multiple corrected cross-correlation values and multiple corrected received data blocks.
[0118] The specific method is:
[0119] Step S5.1: Sampling index sets according to data blocks Separate the corrected cross-correlation operator R(β|v′) into m-1 corrected cross-correlation values
[0120] Step S5.2: Sampling index sets according to data blocks The received data sequence will be corrected Split into m-1 correction receiving data blocks
[0121] S6: performing block correlation data compensation on the obtained multiple corrected mutual correlation values and corrected received data blocks to obtain multiple secondary corrected received data blocks.
[0122] The method for compensating the block-related data in step S6 is:
[0123] The m-1 corrected cross-correlation values R(β 12 |v′ 12)~R(β 1m |v′ 1m ) and m-1 corrected received data blocks Multiply them respectively to get m-1 secondary correction receiving data blocks And the lth secondary correction received data block The secondary correction received data signal in the kth symbol period Has the following form:
[0124]
[0125] In the formula, is the new data sampling time index, is a superimposed noise term, which will have a significant impact on the proposed block-related data compensation. This impact needs to be reduced in the following two ways: The first way is to keep the pilot block length small. The second method is to consider the smaller number of pilot blocks m and increase the length of each pilot block.
[0126] S7: Combine the obtained multiple secondary correction receiving data blocks into a complete receiving data sequence r d ′.
[0127] The specific method is:
[0128] Sampling index collections by data block m-1 secondary correction received data blocks Synthesize a complete receive data sequence
[0129] From the above steps, it can be found that the proposed block (cross) correlation data compensation only requires compensation values of m-1 data blocks in length, while the continuous time data compensation of the traditional synchronization scheme requires Compensation value (i.e. ), so the proposed block (mutual) correlation data compensation occupies less (compensation value) storage space. In addition, combined with the definition of each data block sampling index set, it can be found that the proposed block (mutual) correlation data compensation can significantly reduce the storage space of each data block. The maximum accumulated phase (j = 1, 2, ..., m-1) is due to the fact that the range of the new data sampling time index k' has been changed from the original Reduce to
[0130] In order to further illustrate the beneficial effects of the present invention, a comparative description is given through simulation experiments in this embodiment, as follows:
[0131] Simulation condition 1: Quadrature phase shift keying modulation, symbol period T s =50μs, length of the first pilot block Signal-to-noise ratio SNR = 0:5:25dB, Rice factor K = 0,5,10dB.
[0132] Simulation condition 2: QPSK (Quadrature phase shift keying) modulation, symbol period T s =50μs, the residual frequency offset Δf (after performing frequency-phase decoupling estimation or frequency-phase coupling estimation) d =200Hz, Rice factor K=10dB, number of pilot blocks m=3 (then number of data blocks m-1=2), data symbol length is 150, and pilot symbol lengths are 3, 6, and 9 respectively. It should be noted that when the pilot length is 3, it is necessary to use 1 pilot symbol in the first pilot block, and set the lengths of the second and third pilot blocks to 1. Similarly, when the pilot lengths are 6 and 9, it is necessary to use 2 and 3 pilot symbols in the first pilot block, and set the lengths of the second and third pilot blocks to 2 and 3.
[0133] Simulation 1:
[0134] Considering simulation condition 1, the following two cases are discussed: For the maximum likelihood phase deviation estimator under frequency-phase decoupling estimation, when the decoupling coefficient λ = 1, it can be used in |f d |≤1800Hz to capture the phase deviation in the range of [-π,π); when the decoupling coefficient λ>1, taking the decoupling coefficient λ=2 as an example, it can capture the phase deviation in the range of [-π,π) in |f d |≤3600Hz, the phase deviation in the range of [-10π / 11,10π / 11) is estimated.
[0135] Figure 2 The mean square error (MSE) performance of the autocorrelation frequency offset estimator is given. Here, it is assumed that the frequency offset f d =1600Hz (within the frequency deviation resistance range of the maximum likelihood phase deviation estimator corresponding to the frequency-phase decoupling estimation when the decoupling coefficient λ=1).
[0136] from Figure 2 The simulation results show that for Rice factor K = 0dB and Rice factor K = 10dB, the autocorrelation frequency offset estimator can achieve good mean square error MSE performance. When the signal-to-noise ratio SNR ≥ 25dB, the mean square error MSE of the autocorrelation frequency offset estimator will be less than 1×10 -4 From the statistical average point of view, the residual frequency deviation This value is the residual frequency deviation Δf in simulation condition 2 d =200Hz source.
[0137] Simulation 2:
[0138] Consider simulation condition 1, Figure 3 , Figure 4 The MSE performance of the maximum likelihood phase deviation estimator under frequency-phase decoupled estimation and frequency-phase coupled estimation is shown. Here, the signal-to-noise ratio SNR = 20dB is considered.
[0139] from Figure 3 , Figure 4 The simulation results show that as long as the frequency deviation value is within the agreed range, Regardless of the decoupling coefficient λ and the Ricean factor K, the maximum likelihood phase deviation estimator under frequency-phase decoupling estimation can achieve better performance than the maximum likelihood phase deviation estimator under frequency-phase coupling estimation. This is because a specific delay length is selected in the autocorrelation operator. It not only reduces the number of complex multiplication operations of phase deviation estimation, but also reduces the impact of large frequency deviation on phase deviation estimation. In addition, since the increase of Ricean factor K makes the imaginary part of channel gain h approach 0, the corresponding maximum likelihood phase deviation estimator performance is better, so the phase deviation estimation performance of Ricean factor K = 10dB is better than that of Ricean factor K = 5dB.
[0140] Simulation 3:
[0141] Considering simulation condition 2, from the perspective of short burst communication applications (such as narrowband IoT), the storage resources of its hardware implementation are quite valuable. Therefore, it is necessary to evaluate the storage space occupied by these two types of data compensation. Based on the above conditions, it can be seen that the storage space occupied by continuous moment data compensation can be regarded as the length of the transmitted data, that is, 150; while the storage space occupied by the proposed block-related data compensation is only related to the number of data blocks used, that is, 2. Obviously, in the process of data compensation, the latter occupies very little storage space for the compensation value.
[0142] Figure 5 The bit error rate (BER) performance of these two data compensations is given.
[0143] from Figure 5It can be seen from the simulation results that when residual frequency offset exists, the QPSK system without frequency offset compensation cannot work properly; but after adopting block-related data compensation and continuous moment data compensation, the corresponding system performance will be improved to varying degrees, and the larger the pilot block length, the more obvious the corresponding performance improvement. In general, the performance of the proposed data compensation is similar to that of the continuous moment data compensation. However, in the high signal-to-noise ratio area (for example, signal-to-noise ratio SNR>21dB) and when few pilots are used (for example, less than 6), the proposed data compensation is more than 1 dB worse than the continuous moment data compensation. The reason is that the noise variance of the latter will be smaller under high signal-to-noise ratio.
[0144] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A minimalist synchronization scheme based on a dual-correlation operator, characterized in that: The following steps are involved: S1: Get the received pilot sequence r p And receive the pilot sequence r p Perform a demodulation operation to obtain a demodulated sequence v; S2: Perform a generalized correlation operation on the obtained demodulated sequence v to obtain the autocorrelation operator R(β 11 |v1) and the cross-correlation operator R(β|v); S3: The obtained autocorrelation operator R(β 11 |v1) Perform frequency-phase decoupling estimation to obtain the frequency offset estimate and phase bias estimates S4: The obtained frequency offset estimate and phase bias estimates Compensate to the cross-correlation operator R(β|v) and the received data sequence r d The corrected cross-correlation operator R(β|v′) and the corrected received data sequence are obtained S5: The obtained correction cross-correlation operator R(β|v′) and the correction received data sequence Separating into a plurality of corrected cross-correlation values and a plurality of corrected received data blocks; S6: performing block correlation data compensation on the obtained multiple corrected mutual correlation values and the multiple corrected received data blocks to obtain multiple secondary corrected received data blocks; S7: Combine the obtained multiple secondary correction received data blocks into a complete received data sequence r d ′.
2. The minimalist synchronization scheme based on dual correlation operators according to claim 1, characterized in that: The step S1 described in obtaining the received pilot sequence r p The method is: S1.1: Consider a Ricean flat fading single-carrier communication system, the corresponding received signal r = {r(k), k∈κ p ∪κ d } and the received signal r(k) in the kth symbol period is expressed as: Where h is the channel gain, whose amplitude obeys the Rice distribution determined by the Rice factor, T s is the symbol period, f d and θ are the carrier frequency deviation and phase deviation respectively, s(k) is the energy-normalized modulation signal, n(k) is a complex Gaussian random variable, is the imaginary unit, There are m pilot blocks p1~p m The sampling index set of There are m-1 data blocks d1~d m-1 The sampling index set, and the i-th pilot block p i The sampling index of and the lth data block d l The sampling index of The specific forms are: in, Respectively represent the lengths of the lth, i-th, lth, and tth pilot blocks; Respectively represent the length of the lth and tth data blocks; according to the m pilot blocks p1~p m The sampling index set κ p Get the received pilot sequence r p .
3. The minimalist synchronization scheme based on dual correlation operators according to claim 2, characterized in that: According to the m pilot blocks p1~p m The sampling index set κ p Get the received pilot sequence r p The method is: S1.2: At time k, traverse m pilot blocks p1~p m The sampling index set κ p , the corresponding received pilot sequence r p ={r p (k),k∈κ p }, and the received pilot signal r in the kth symbol period p (k) is expressed as: Where s p (k) is the modulated pilot signal in the kth symbol period.
4. The minimalist synchronization scheme based on dual correlation operators according to claim 3, characterized in that: The method of the demodulation operation described in step S1 is: S1.3: Receive the pilot signal r p (k) and the modulated pilot signal s p The demodulated signal v(k) is obtained by multiplying the conjugate form of (k): v(k)=r p (k)·s p (k) * Where * is the conjugate operation; S1.4: When time k traverses the pilot block sampling index set κ p When the demodulation signal v(k) forms the demodulation sequence 5. The minimalist synchronization solution based on dual correlation operators according to claim 4, characterized in that: The method of generalizing the related operations described in step S2 is: S2.1: Demodulation sequence using the first pilot block p1 The conjugate form of the demodulated signal v(k) corresponding to the pilot block p1 is combined with its delayed signal v(k+β 11 ) is multiplied to obtain the autocorrelation operator R(β 11 |v1): Where, is the length of the pilot block p1, represents the autocorrelation delay length; S2.2: Use the demodulation sequence of the first pilot block p1 and other m-1 pilot blocks p2~p m Demodulation sequence The pilot blocks p2~p m The conjugate form of the corresponding demodulated signal v(k) and its delayed signal v(k+β 1i ) are multiplied to obtain the cross-correlation operator R(β|v)={R(β 12 |v 12 ),...,R(β 1i |v 1i ),...,R(β 1m |v 1m )}, the i-th (i=2,3,...,m) cross-correlation value R(β 1i |v 1i )for: In the formula, the modulation sequence Demodulation sequence v i ={v(β 1i ),v(1+β 1i ),..., v 1i The demodulation sequence v1 of the pilot block p1 and the pilot block p i The demodulation sequence v i The union of is the cross-correlation delay length.
6. The minimalist synchronization scheme based on dual correlation operators according to claim 5, characterized in that: The method for frequency-phase decoupling estimation described in step S3 is: S3.1: Using the autocorrelation operator R(β 11 |v1), and obtain the frequency offset estimate by taking the argument operation Where arg{·} is the argument operation; S3.2: Using the autocorrelation operator By comparing the demodulated signal v(k) corresponding to the pilot block p1 with the autocorrelation operator The conjugate form of is summed and the phase angle is obtained to obtain the phase deviation estimate represents a special autocorrelation delay length and Phase bias estimate The calculation formula is: Where λ=1,2,... is the decoupling coefficient.
7. The minimalist synchronization solution based on dual correlation operators according to claim 6, characterized in that: The method for obtaining the corrected cross-correlation operator R(β|v′) in step S4 is: S4.1: The frequency offset estimate and phase bias estimates Compensate to the cross-correlation operator R(β|v) to obtain the corrected cross-correlation operator R(β|v′)={R(β 12 |v′ 12 ),...,R(β 1i |v′ 1i ),...,R(β 1m |v′ 1m )}, and the i-th corrected cross-correlation value R(β 1i |v′ 1i )for: In the formula, the modulation sequence The result after compensating the frequency offset and phase offset estimation value of v(k) is the demodulation sequence For v(k+β 1i )v(k) is the result after frequency offset and phase offset estimation are compensated, v′ 1i The demodulation sequence v1′ of the pilot block p1 and the pilot block p i The demodulation sequence v i ′’s union, The corrected received data sequence obtained in step S4 The method is: S4.2: The frequency offset estimate and phase bias estimates Compensate to the received data sequence r d The corrected received data sequence is obtained from in, And the corrected received data signal r′(k) in the kth symbol period is:
8. The minimalist synchronization scheme based on dual correlation operators according to claim 7, characterized in that: Step S5 is to obtain the corrected cross-correlation operator R(β|v′) and the corrected received data sequence. The method of separating into multiple correction mutual correlation values and multiple correction received data blocks is: S5.1: Sampling index sets by data block Separate the corrected cross-correlation operator R(β|v′) into m-1 corrected cross-correlation values R(β 12 |v′ 12 ),R(β 13 |v′ 13 ),...,R(β 1m |v′ 1m ); S5.2: Sampling index sets by data block The received data sequence will be corrected Split into m-1 correction receiving data blocks 9. The minimalist synchronization scheme based on dual correlation operators according to claim 8, characterized in that: The method for compensating the block-related data in step S6 is: The m-1 corrected cross-correlation values R(β 12 |v′ 12 ),R(β 13 |v1′3),...,R(β 1m |v1′ m ) and m-1 corrected received data blocks Multiply them respectively to get m-1 secondary correction receiving data blocks And the lth secondary correction received data block The secondary correction received data signal in the kth symbol period Has the following form:
10. The minimalist synchronization solution based on dual correlation operators according to claim 9, characterized in that: Step S7 combines the obtained multiple secondary corrected received data blocks into a complete received data sequence r d 'The method is: Sampling index sets according to data blocks, m-1 secondary correction received data blocks Synthesize a complete receive data sequence
Citation Information
Patent Citations
Method and device for frequency deviation correction based on data field partitioning
CN102647377A
Joint frequency phase estimation method based on partial decoupling
CN107911323A
Spread spectrum receiver with multi-bit correlator
WO1997040398A2
