ZC sequence-based autocorrelation detection method

Abstract

The invention belongs to the field of signal detection, and particularly relates to correlation detection based on ZC sequence detection signals. According to the method, parameters of the received ZC sequences are firstly determined, part of the ZC sequences are taken out to be combined according to the arrangement mode of the ZC sequences, and compared with direct correlation of original signals, correlation detection of the combined ZC sequences can improve the detection rate of the received signals.

Description

Technical Field

[0001] The invention belongs to the field of signal detection, in particular to correlation detection based on ZC sequence detection signals. Background Art

[0002] In communication systems, especially digital communication systems, synchronization is a very important issue.

[0003] The ZC sequence (Zadoff-Chu sequence) is a discrete sequence with good properties and a special linear frequency modulation pulse compression sequence. The ZC sequence is widely used in synchronization and channel estimation in communication systems.

[0004] The generation formula of ZC sequence is shown in formula (1):

[0005]

[0006] The two most important parameters in the ZC sequence are the root number u in formula (1) and the length L of the ZC sequence. RA In the application process, the length of the ZC sequence L RA is an odd number, usually a prime number. The length of the ZC sequence is L RA , the size of the sequence space of ZC sequences is L RA -1, so there are L RA -1 ZC sequence. What needs to be determined during the detection process is the root sequence number u and the length L of the ZC sequence. RA .

[0007] The autocorrelation of a ZC sequence refers to the correlation operation between the sequence and itself. The autocorrelation can reflect the periodicity and repeatability of the sequence, and is very important in the synchronization of communication systems and channel estimation.

[0008] The value generated by the ZC sequence is a complex number. In this case, we can use the Euler formula e jα = cos(α) + jsin(α) to get the real and imaginary parts of the complex number, whose amplitude is always 1. At the same time, the sequence itself has its own autocorrelation and good cross-correlation, and after Fourier transform, it also meets its own properties and is suitable for transmitting time and frequency as a synchronization signal, especially in OFDM (Orthogonal Frequency Division Multiplexing) signals.

[0009] The autocorrelation of the ZC sequence is:

[0010]

[0011] Wherein: c(k) is the ZC sequence, M is the period of the ZC sequence, mod is the modulus, τ is the time shift length of the ZC sequence itself, and the cyclic autocorrelation of the ZC sequence satisfies the characteristic that the sidelobe is 0.

[0012] After the ZC sequence undergoes fast Fourier transform, its autocorrelation and periodicity exist, and it is often used as a synchronization sequence in OFDM signals.

[0013] The OFDM system transmitter performs channel coding and interleaving on the information bits, and modulates them to each subcarrier after serial-to-parallel conversion. In order to reduce inter-symbol interference, a cyclic prefix should be added. If the cyclic prefix is ​​taken into account, the frequency domain signal is transformed into the time domain expression after fast Fourier transform:

[0014]

[0015] In formula (3), x(n) is the time domain expression of the OFDM signal, X(k) is the data modulated on the kth subcarrier, N is the total number of system subcarriers, and Ncp is the length of the cyclic prefix.

[0016] Assuming that the system has no carrier frequency deviation and the sampling signals at both ends of the transmitter and receiver are synchronized, after passing through the multipath channel and being affected by noise, and considering that the system has a timing deviation d, the received signal after sampling can be expressed as:

[0017]

[0018] Where L is the number of multipath components; h(l) is the channel impulse response; w(n) is the additive white Gaussian noise, and d is the timing deviation.

[0019] The ZC sequence in the synchronization sequence of the OFDM signal usually has different arrangements. In different ZC arrangements, when the local ZC sequence is used for autocorrelation, the detection coefficient is not high due to the different arrangements. Summary of the invention

[0020] In view of the above problems, the present invention proposes a detection method of a ZC sequence based on an even-odd arrangement, thereby greatly improving the detection rate of the ZC sequence. The present invention is used for a detection method for the synchronization head of an OFDM signal, which obtains a corresponding ZC sequence by synchronizing and demodulating the received signal, and correlates the ZC sequence with the local ZC sequence through different combinations according to the arrangement mode and even-odd arrangement of the ZC sequence, thereby improving the detection coefficient of the signal.

[0021] The technical solution of the present invention is:

[0022] A method for autocorrelation detection based on ZC sequence parity arrangement, the scheme comprises the following steps:

[0023] S1. Extract the ZC sequence from the received OFDM signal;

[0024] S2. Determine the length and root sequence number of the retrieved ZC sequence to generate a local ZC sequence;

[0025] S3, firstly correlate the extracted ZC sequence with the local ZC sequence to determine the correlation coefficient of the detection, and at the same time separate the extracted ZC sequence and the local ZC sequence into four signals in odd and even order;

[0026] S4. Arrange and combine the taken ZC sequence in odd-even order and perform autocorrelation with the local ZC sequence again to determine the correlation coefficient of the detection.

[0027] The beneficial effect of the present invention is that the correlation after the odd-even arrangement and combination can greatly improve the detection rate of the ZC sequence. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Schematic diagram of the distribution of ideal ZC sequence.

[0029] Figure 2 FIG. 4 is a schematic diagram of the frequency domain distribution of the ZC sequence to be detected.

[0030] Figure 3 Schematic diagram of the root number of the ZC sequence, (a) is a schematic diagram of the ZC sequence with a root number of 600, and (b) is a schematic diagram of the ZC sequence with a root number of 1.

[0031] Figure 4 Schematic diagram of detection coefficients of unprocessed ZC sequence.

[0032] Figure 5 It is a detection coefficient diagram after combining ZC sequences of different odd and even positions. (a) is a detection schematic diagram after combining S1 and S2, and (b) is a detection schematic diagram after combining S3 and S4.

[0033] Figure 6 It is a detection coefficient diagram of a separate part of the ZC sequence, where (a) is a detection schematic diagram after S1 is correlated with the ZC sequence, and (b) is a detection schematic diagram after S2 is correlated with the ZC sequence. DETAILED DESCRIPTION

[0034] The technical principle and scheme of the present invention are described in detail below with reference to the accompanying drawings and simulation examples:

[0035] The present invention is divided into four parts, namely, taking out the ZC sequence, determining the root sequence number and length, separating the odd and even ZC sequence, and performing autocorrelation with the local ZC sequence.

[0036] S1. Perform time and frequency domain synchronization on the received OFDM signal, and use the cyclic prefix of the OFDM signal to synchronize and extract the ZC sequence.

[0037]

[0038] Among them, N G is the length of the cyclic prefix, y l is the input OFDM signal, δ is the starting position of the signal, and N is the total number of subcarriers in the OFDM signal.

[0039] After synchronizing the OFDM signal, the ZC sequence A to be detected can be obtained by demodulation.

[0040] S2: Based on the length of the ZC sequence obtained in S1, the autocorrelation of the ZC sequence is used to traverse the root sequence number of the sequence to prepare for the subsequent parity separation.

[0041]

[0042] Where: c(k) is the ZC sequence, M is the period of the ZC sequence, mod is the modulus, τ is the time shift length of the ZC sequence itself, and its size ranges from 0 to M. The cyclic autocorrelation of the ZC sequence satisfies the characteristic that the sidelobe is 0.

[0043] The root sequence number and length are determined by using autocorrelation to generate a local ZC sequence B.

[0044] S3. Separate the obtained ZC sequence A using the constant mode characteristic of the ZC sequence.

[0045] First, ZC sequence A and ZC sequence B are directly correlated to obtain the correlation detection coefficient. Then, ZC sequence A is processed and separated according to the odd and even order according to different module values ​​to obtain signal 1, signal 2, signal 3, and signal 4. At the same time, the locally generated ZC sequence B is also processed in the same way.

[0046] S4. Perform different permutations and combinations of ZC sequence A and related operations with ZC sequence B

[0047] The signal obtained by odd-even separation of ZC sequence A is retained with two signals, signal 1 and signal 2, and the rest are assigned to 0 to obtain sequence C. At the same time, the local ZC sequence B is processed in the same way to obtain sequence D. Sequence C and sequence D are correlated and normalized to obtain the coefficients to be detected.

[0048] Simulation Example

[0049] In OFDM signals, the ZC sequence is used as a synchronization sequence for simulation and detection to verify the detection coefficient under this condition.

[0050] Figure 1 It is the graph generated by the ZC sequence under ideal conditions. The real and imaginary parts are distributed on the unit circle, and its modulus is always 1. In practical applications, there will be multipath effects, which will affect the amplitude of the ZC sequence, such as Figure 2 shown.

[0051] Figure 2 It is the ZC sequence A in the frequency domain obtained after the OFDM signal is synchronized and demodulated. As can be seen from the figure, the ZC sequence A is divided into 4 parts, namely S1, S2, S3, and S4. The modulus values ​​of S1 and S2 are the same, and the modulus values ​​of S3 and S4 are the same. The fluctuations are due to the effects of the receiver and multipath effects in the actual reception process.

[0052] Then, the autocorrelation of the ZC sequence is used to perform correlation operations on S1, S2, S3, and S4, thereby obtaining the root number of the ZC sequence, and thus obtaining Figure 3 The normalized autocorrelation coefficients are shown, where (a) is the root number 600 of S1 and S2, (b) is the root number 1 of S3 and S4, and the total length of the four signals is 601.

[0053] After determining the root sequence numbers of the four parts, ZC sequence B with root sequence numbers 1 and 600 is generated locally and correlated with ZC sequence A. However, due to the phase difference between S2 and S1, and between S4 and S3, the correlation coefficient obtained after correlating the locally generated ZC sequence B with the received ZC sequence A is as follows: Figure 4 As shown, the correlation coefficient is not high at this time.

[0054] Combine S1, S2, S3, and S4 of ZC sequence A, such as Figure 5 As shown, S1 and S2 are retained, and the remaining positions are assigned 0. At the same time, the corresponding S1 and S2 of the ZC sequence B are retained, and the remaining positions are also assigned 0, and then autocorrelation is performed. At this time, the detection coefficient is significantly higher than the correlation coefficient of the original unprocessed signal, thereby improving the detection rate of the signal. Figure 5 Combining S3 and S4 can also improve the detection rate of the ZC sequence.

[0055] Figure 6 To keep S1 and S2 of ZC sequence A separately and assign the rest to 0, ZC sequence B is processed in the same way, and then the detection coefficients are obtained after correlation processing.

[0056] Depend on Figure 5 , Figure 6 It can be seen that the detection coefficients after combining and arranging the odd-even arranged ZC sequences are higher than the detection coefficients directly correlating the original ZC sequences.

Claims

1. A ZC sequence-based autocorrelation detection method, characterized in that: The following steps are involved: S1. Perform time and frequency domain synchronization on the received OFDM signal, and use the cyclic prefix of the OFDM signal to synchronize and extract the ZC sequence. Among them, N G is the length of the cyclic prefix, y l is the input OFDM signal, δ is the starting position of the signal, and N is the total number of subcarriers of the OFDM signal; After synchronizing the OFDM signal, the ZC sequence A required for detection is obtained through demodulation; S2. Based on the length of the ZC sequence obtained in S1, use the autocorrelation of the ZC sequence to traverse the root number of the sequence: Wherein, c(k) is the ZC sequence, M is the period of the ZC sequence, mod is the modulus, τ is the time shift length of the ZC sequence itself, and its size ranges from 0 to M. The cyclic autocorrelation of the ZC sequence satisfies the characteristic that the sidelobe is 0; The root sequence number and length are determined by using autocorrelation to generate a local ZC sequence B; S3. Separate the obtained ZC sequence A using the constant mode property of the ZC sequence: The ZC sequence A obtained in S1 is processed and separated in odd-even order according to different module values ​​to obtain signal 1, signal 2, signal 3, and signal 4. Meanwhile, the locally generated ZC sequence B is processed in the same way. S4, performing different permutations and combinations of ZC sequence A and performing related operations with ZC sequence B; The signal obtained by odd-even separation of ZC sequence A is retained, and two signals, signal 1 and signal 2, are assigned to 0, to obtain sequence C; at the same time, the local ZC sequence B is processed in the same way to obtain sequence D. Finally, sequence C and sequence D are correlated and normalized to obtain the coefficients to be detected.

Images