Two-dimensional synchronization method based on OFDM system

Abstract

The invention discloses a two-dimensional synchronization method based on an OFDM (Orthogonal Frequency Division Multiplexing) system, which comprises the following steps of: carrying out baseband processing on a received signal to obtain a baseband signal; performing frequency offset compensation and cross-correlation operation on the baseband signal; performing threshold judgment on a correlation value obtained by the cross-correlation operation, and outputting a correct synchronization position; and according to the correct synchronization position, carrying out accurate frequency offset estimation and frequency offset compensation on the received signal. In the invention, in order to reduce the influence of the frequency offset on the cross-correlation operation, the frequency offset is compensated firstly and then the cross-correlation operation is carried out, so that two-dimensional search on time and frequency is formed, and the two-dimensional search is also called time-frequency two-dimensional synchronization; most of the frequency offset is compensated firstly, so that the influence caused by the frequency offset during cross-correlation operation is reduced as much as possible.

Description

Technical Field

[0001] The present invention relates to the technical field of OFDM, and particularly to a two-dimensional synchronization method based on an OFDM system. Background Art

[0002] OFDM (Orthogonal Frequency Division Multiplexing) is an orthogonal frequency division multiplexing technology. OFDM technology is a multi-carrier modulation technology that realizes parallel transmission of high-speed serial data through frequency division multiplexing. It has good anti-fading and anti-interference performance and can support multi-user access. However, this technology is susceptible to carrier frequency offset and timing synchronization offset. Carrier frequency offset will cause interference between sub-carriers, and timing synchronization error will cause interference between symbols. Therefore, the research and implementation of synchronization algorithms for these two synchronization errors are of great significance.

[0003] Defects and deficiencies of the prior art: In the existing synchronization algorithms, there are time synchronization and frequency synchronization. Time synchronization is mainly to determine the position of the OFDM symbol, and frequency synchronization is mainly to estimate the Doppler frequency shift suffered by the signal when passing through the channel and compensate it. Time synchronization can be further divided into autocorrelation operation and cross-correlation operation. However, the autocorrelation operation is susceptible to noise, and the cross-correlation algorithm is susceptible to carrier frequency offset. Frequency synchronization mostly uses autocorrelation operation, and uses the correlation between the front and back signals to calculate their phase difference, thereby estimating the frequency offset of the system. The autocorrelation operation in time synchronization is susceptible to noise, the cross-correlation operation is susceptible to carrier frequency offset, and frequency synchronization is affected by the result of time synchronization. Most synchronization schemes first perform autocorrelation operation to obtain the rough time synchronization position, then estimate and compensate the frequency offset according to the rough synchronization position, and finally perform cross-correlation synchronization to achieve fine time synchronization. However, this method is still affected by the first rough synchronization position. If the estimation error is large, the subsequent synchronization will be incorrect.

[0004] For example: In the prior art 1 (a patent application for an invention named "Frequency Offset Estimation and Compensation Method for MIMO High-Speed Receiver" with the application number CN202111012843.X), through the frame start detection module, the cross-correlation between the known frame header information zt(m) stored locally and the input signal S(k) before frequency offset correction is calculated, and the cross-correlation is directly performed. However, when the frequency offset of the signal is large, the peak value after the cross-correlation operation will be misjudged, that is, the wrong synchronization position will be found.

[0005] In the prior art 2 (a patent application for an invention named "A Time-Frequency Synchronization Method and Device" with the application number CN201710217689.7), time synchronization is performed first and then frequency synchronization. Similarly, when the frequency offset is too large, the synchronization position will be incorrect. Summary of the Invention

[0006] The object of the present invention is to overcome the deficiencies of the prior art and provide a two-dimensional synchronization method based on an OFDM system.

[0007] The object of the present invention is achieved by the following technical solutions:

[0008] In a first aspect of the present invention, there is provided a two-dimensional synchronization method based on an OFDM system, including the following steps:

[0009] Perform baseband processing on the received signal to obtain a baseband signal;

[0010] Perform frequency offset compensation and cross-correlation operation on the baseband signal;

[0011] Perform threshold judgment on the correlation values obtained from the cross-correlation operation and output the correct synchronization position;

[0012] According to the correct synchronization position, perform accurate frequency offset estimation and frequency offset compensation on the received signal.

[0013] Further, the frequency offset compensation for the baseband signal is achieved by generating a complex signal through a DDS IP core and multiplying it with the received signal;

[0014] The complex signal generated by the DDS IP core changes its frequency through an NCO control word, that is, different frequency complex signals are generated by changing the NCO control word, thereby achieving multiplex frequency offset compensation.

[0015] Further, the frequency offset compensation for the baseband signal has a compensation frequency ranging from -f to f, and the compensation step is Δf.

[0016] Further, the cross-correlation operation is implemented through matched filtering, and in an FPGA, the matched filtering is implemented through a FIR IP core, and the filter coefficients are the corresponding local sequences; that is, the cross-correlation operation is to perform a cross-correlation operation between the signals after frequency offset compensation for each path and the local sequences;

[0017] The local sequences are generated by Matlab, and then they need to be fixed-point processed and stored in the RAM of the FPGA as filter coefficients.

[0018] Further, the threshold judgment on the correlation values obtained from the cross-correlation operation and outputting the correct synchronization position includes:

[0019] For the correlation values obtained from the cross-correlation operation, take the maximum value, average value, and the position corresponding to the maximum value within each frame of data;

[0020] Next, threshold determination is performed on each frame of data. When the maximum value is greater than N times the average value, the peak corresponding to the maximum value is determined to be a valid peak.

[0021] The synchronization position is found according to the peak position, that is, the position corresponding to the maximum value.

[0022] Further, the operations of taking the maximum value and the average value within each frame of data for the correlation values obtained by the cross-correlation operation, and the position corresponding to the maximum value, include:

[0023] According to the frame end flag, the correlation values of each frame of data are accumulated, and an averaging operation is performed to obtain the average value.

[0024] The correlation values of each frame of data are read in sequence. If the correlation value is greater than the maximum value register, the maximum value register is updated; otherwise, the maximum value in the maximum value register remains unchanged until the frame end flag is read; the maximum value and the position corresponding to the maximum value are obtained.

[0025] Further, the received signal uses a CAZAC sequence with a repetitive structure as the synchronization sequence. The first part and the third part of the CAZAC sequence are the same data to achieve frequency synchronization, and the second and fourth parts of the CAZAC sequence are opposite to each other.

[0026] Further, the operations of performing accurate frequency offset estimation and frequency offset compensation on the received signal according to the correct synchronization position, include:

[0027] The signal after coarse synchronization is divided into two parts and stored in two RAMs respectively. Only the parts with repetitive structures are calculated, and the parts with opposite numbers are not calculated.

[0028] The two segments of signals stored in the two RAMs are subjected to conjugate multiplication and accumulation respectively. The value obtained by performing phase calculation on the accumulated value is the frequency offset value.

[0029] According to the estimated frequency offset value, the CORDIC IP core is used to perform phase rotation on the signal to achieve frequency offset compensation.

[0030] The beneficial effects of the present invention are:

[0031] In an exemplary embodiment of the present invention, to reduce the influence of frequency offset on the cross-correlation operation, frequency offset is compensated first and then the cross-correlation operation is performed, thereby forming a two-dimensional search in time and frequency, which is also called time-frequency two-dimensional synchronization. By compensating most of the frequency offset first, the influence of frequency offset during the cross-correlation operation is minimized as much as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of a two-dimensional synchronization method based on an OFDM system provided in an exemplary embodiment of the present invention;

[0033] Figure 2 Schematic diagram of baseband signal frequency offset compensation provided in an exemplary embodiment of the present invention;

[0034] Figure 3 Schematic diagram of cross-correlation operation provided in an exemplary embodiment of the present invention;

[0035] Figure 4 Schematic diagram of threshold judgment provided in an exemplary embodiment of the present invention;

[0036] Figure 5 Schematic diagram of CAZAC sequence provided in an exemplary embodiment of the present invention;

[0037] Figure 6 Schematic diagram of accurate frequency offset estimation and frequency offset compensation provided in an exemplary embodiment of the present invention. Detailed implementation manners

[0038] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0039] In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation of the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0040] In the description of the present invention, it should be noted that unless otherwise clearly defined and limited, the terms "installation", "connection", and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0041] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0042] See Figure 1 , Figure 1An OFDM system two-dimensional synchronization method provided in an exemplary embodiment of the present invention is shown, including the following steps:

[0043] Perform baseband processing on the received signal to obtain a baseband signal;

[0044] Perform frequency offset compensation and cross-correlation operation on the baseband signal;

[0045] Perform threshold judgment on the correlation values obtained by the cross-correlation operation, and output the correct synchronization position;

[0046] According to the correct synchronization position, perform accurate frequency offset estimation and frequency offset compensation on the received signal.

[0047] Specifically, in this exemplary embodiment, first, perform baseband processing on the received signal to obtain a baseband signal, and perform multi-path frequency offset compensation on the baseband signal; then perform cross-correlation operation with the local sequence on each compensated signal. Only the peak value of the correlation peak corresponding to the correctly compensated branch is the highest. Perform threshold determination on the correlation values to output the correctly synchronized signal, and then perform accurate frequency synchronization on the synchronized signal again.

[0048] Therefore, in this exemplary embodiment, to reduce the influence of frequency offset on the cross-correlation operation, first compensate the frequency offset and then perform the cross-correlation operation, thereby constituting a two-dimensional search in time and frequency, which is also called time-frequency two-dimensional synchronization. By compensating most of the frequency offset first, the influence of frequency offset during the cross-correlation operation is minimized as much as possible.

[0049] The following content will describe the preferred implementation manners of each step:

[0050] More preferably, in an exemplary embodiment, as Figure 2 shown, the frequency offset compensation for the baseband signal is realized by generating a complex signal through a DDS IP core and multiplying it with the received signal;

[0051] The complex signal generated by the DDS IP core changes the frequency through the NCO control word, that is, different frequency complex signals are generated by changing the NCO control word, thereby realizing multi-path frequency offset compensation.

[0052] Specifically, in this exemplary embodiment, a multi-path frequency offset compensation + sliding correlation method is adopted to realize two-dimensional synchronization. Finally, accurate frequency synchronization can be achieved according to the synchronization position, and the frequency offset compensation:

[0053] In the FPGA, this part can generate complex signals through the DDS IP core and multiply them with the received signals to achieve frequency offset compensation. The complex signals generated by the DDS IP core can be frequency-changed through the NCO frequency control word. That is to say, by changing the NCO control word, complex signals with different frequencies can be generated, and thus multi-channel frequency offset compensation can be achieved.

[0054] In other exemplary embodiments, phase rotation can also be performed through the Cordic IP core to complete frequency offset compensation, but more resources are consumed. As for using the Cordic IP core in subsequent frequency offset estimation, it is because it can also calculate the phase of the signal. The NCO control word can be used to generate complex signals with multiple frequencies to complete multiple frequency compensations, and the implementation of this method is relatively simple.

[0055] More preferably, in an exemplary embodiment, for the frequency offset compensation of the baseband signal, the compensation frequency ranges from -f to f, and the compensation step is Δf.

[0056] Specifically, in this exemplary embodiment, by presetting the range and step size of the compensation frequency, when the actual frequency offset is within the compensation range, the impact brought by it can be reduced.

[0057] More preferably, in an exemplary embodiment, as Figure 3 shown, the cross-correlation operation is implemented through matched filtering, and in the FPGA, the matched filtering is implemented through the FIR IP core, and the filter coefficients are the corresponding local sequences; that is, the cross-correlation operation is to perform cross-correlation operations on the signals after frequency offset compensation for each path with the local sequences;

[0058] The local sequences are generated by Matlab, and then they need to be fixed-point processed and stored in the RAM of the FPGA as the filter coefficients.

[0059] Specifically, in this exemplary embodiment, matched filtering can be used to replace sliding correlation. In the FPGA, the matched filtering can be implemented through the FIR IP core, and its filter coefficients are the corresponding local sequences. At the same time, the method of multiplexing multipliers is used to reduce the consumption of DSP.

[0060] Implementing with matched filtering is relatively simple. In the FPGA, the Fir IP core can be directly called to implement filtering, which is relatively easy to implement. The filter here is to perform correlation operations on the received signals and the filter coefficients (originally, correlation operations on the received signals and the local sequences), so its coefficients are also the corresponding local sequences. The local sequences can be generated by Matlab, and then they need to be fixed-point processed and stored in the RAM of the FPGA as the filter coefficients.

[0061] More preferably, in an exemplary embodiment, asFigure 4 As shown, the threshold judgment is performed on the correlation value obtained by the cross-correlation operation, and the correct synchronization position is output, including:

[0062] Taking the maximum value, average value, and the position corresponding to the maximum value within each frame of data for the correlation value obtained by the cross-correlation operation;

[0063] Then, the threshold judgment is performed on each frame of data. When the maximum value is greater than N times the average value, the peak corresponding to the maximum value is determined to be a valid peak;

[0064] The synchronization position is found according to the peak position, that is, the position corresponding to the maximum value.

[0065] Specifically, in this exemplary embodiment, the threshold judgment is performed on each frame of data. When the maximum value is greater than N times the average value, the peak corresponding to the maximum value is determined to be a valid peak, reducing the false alarm probability to ensure that the peak is the correct peak.

[0066] In a specific exemplary embodiment, the value of N is 8 times.

[0067] More preferably, in an exemplary embodiment, as Figure 4 shown, taking the maximum value, average value, and the position corresponding to the maximum value within each frame of data for the correlation value obtained by the cross-correlation operation includes:

[0068] According to the frame end flag, the correlation values of each frame of data are accumulated, and the average operation is performed to obtain the average value;

[0069] The correlation values of each frame of data are read in sequence. If the correlation value is greater than the maximum value register, the maximum value register is updated; otherwise, the maximum value in the maximum value register remains unchanged until the frame end flag is read; the maximum value and the position corresponding to the maximum value are obtained.

[0070] More preferably, in an exemplary embodiment, as Figure 5 shown, the received signal uses a CAZAC sequence with a repetitive structure as the synchronization sequence. The first part and the third part of the CAZAC sequence are the same data to achieve frequency synchronization, and the second and fourth parts of the CAZAC sequence are opposite to each other.

[0071] Specifically, there are many algorithms based on the special cyclic prefix structure of OFDM. Since its prefix is directly copied from the tail data, the autocorrelation operation can be performed on the prefix to achieve synchronization. However, the synchronization performance achieved by using this structure is poor, so it is generally used for coarse time synchronization. This structure can also be used for frequency synchronization, and its synchronization performance is affected by the prefix length. The longer the cyclic prefix length, the better the synchronization performance, but too long a cyclic prefix will cause resource waste.

[0072] In this exemplary embodiment, for the above synchronization scheme, a CAZAC sequence is used as the synchronization sequence. The CAZAC sequence has good autocorrelation and cross-correlation characteristics. At the same time, when designing the synchronization sequence, a certain repetitive structure is adopted to facilitate more accurate frequency synchronization in the subsequent process.

[0073] More specifically, the CAZAC sequence with a repetitive structure is used as the synchronization sequence, and its structure is as Figure 5 shown. The first part and the third part are the same data for realizing frequency synchronization, and the second and fourth parts are opposite numbers to each other. CP is the cyclic prefix of the OFDM symbol, which is copied from the tail data of the OFDM symbol. Directly using the repetitive structure for cross-correlation operation will generate side peaks, which will interfere with the determination of the timing synchronization position. Adopting a structure with partial opposite numbers can reduce the side peaks and thus reduce the influence brought by them.

[0074] More preferably, in an exemplary embodiment, as Figure 6 shown, according to the correct synchronization position, performing accurate frequency offset estimation and frequency offset compensation on the received signal includes:

[0075] Dividing the roughly synchronized signal into two parts and storing them in two RAMs respectively, only calculating the part with repetitive structure, and not calculating the part with opposite numbers;

[0076] Performing conjugate multiplication and accumulation on the two segments of signals stored in the two RAMs respectively, and the value obtained by performing phase calculation on the accumulated value is the frequency offset value;

[0077] According to the estimated frequency offset value, using the CORDIC IP core to perform phase rotation on the signal to achieve frequency offset compensation.

[0078] Specifically, in this exemplary embodiment, this module step can use the CORDIC IP core to achieve phase calculation and phase rotation, so as to complete frequency synchronization. First, divide the roughly synchronized signal into two parts and store them in two RAMs respectively (only calculate the part with repetitive structure, and do not calculate the part with opposite numbers). Subsequently, perform conjugate multiplication and accumulation on the two segments of signals, and the value obtained by performing phase calculation on the accumulated value is the frequency offset value. According to the estimated frequency offset value, use the CORDIC IP core to perform phase rotation on the signal to achieve frequency offset compensation.

[0079] Obviously, the above embodiments are only examples clearly described and not limitations on the implementation manners. For those of ordinary skill in the art, based on the above description, other different forms of changes or variations can be made. It is not necessary and impossible to list all the implementation manners here. And the obvious changes or variations derived therefrom are still within the protection scope of the present invention.

Claims

1. A two-dimensional synchronization method based on an OFDM system, characterized in that: The following steps are involved: Perform baseband processing on the received signal to obtain a baseband signal; Performing frequency offset compensation and cross-correlation operations on the baseband signal; Perform threshold judgment on the correlation value obtained by the cross-correlation operation and output the correct synchronization position; According to the correct synchronization position, the frequency offset of the received signal is accurately estimated and compensated.

2. A two-dimensional synchronization method based on OFDM system according to claim 1, characterized in that: The frequency deviation compensation of the baseband signal is achieved by generating a complex signal through the DDS IP core and multiplying the complex signal with the received signal; The complex signal generated by the DDS IP core changes its frequency through the NCO control word. That is to say, by changing the NCO control word, complex signals of different frequencies are generated, thereby achieving multi-channel frequency deviation compensation.

3. A two-dimensional synchronization method based on OFDM system according to claim 2, characterized in that: The baseband signal frequency offset is compensated, the compensation frequency is from -f to f, and the compensation step is Δf.

4. The two-dimensional synchronization method based on OFDM system according to claim 1, characterized in that: The cross-correlation operation is implemented by matched filtering, and matched filtering is implemented by FIR IP core in FPGA, and the filter coefficient is the corresponding local sequence; that is, the cross-correlation operation is to perform cross-correlation operation on each frequency offset compensated signal with the local sequence; The local sequence is generated by Matlab, and then needs to be processed into a fixed point and stored in the RAM of the FPGA as the coefficients of the filter.

5. The two-dimensional synchronization method based on OFDM system according to claim 1, characterized in that: The step of performing threshold judgment on the correlation value obtained by the cross-correlation operation and outputting the correct synchronization position comprises: For the correlation values ​​obtained by the cross-correlation operation, the maximum value and the average value in each frame of data are taken, as well as the corresponding position of the maximum value; Then, a threshold is determined for each frame of data. When the maximum value is greater than N times the average value, the peak value corresponding to the maximum value is determined to be a valid peak value. The synchronization position is found according to the peak position, i.e. the corresponding position of the maximum value.

6. A two-dimensional synchronization method based on OFDM system according to claim 5, characterized in that: The method of taking the maximum value and the average value of the correlation values ​​obtained by the cross-correlation operation in each frame of data, as well as the position corresponding to the maximum value, comprises: According to the frame end mark, the relevant values ​​of each frame data are accumulated, and the average operation is performed to obtain the average value; The relevant value of each frame of data is read in sequence. If the relevant value is greater than the maximum value register, the maximum value register is updated, otherwise the maximum value in the maximum value register is maintained until the frame end mark is read; the maximum value and the corresponding position of the maximum value are obtained.

7. The two-dimensional synchronization method based on OFDM system according to claim 1, characterized in that: The received signal uses a CAZAC sequence with a repetitive structure as a synchronization sequence, the first part and the third part of the CAZAC sequence are the same data for achieving frequency synchronization, and the second and fourth parts of the CAZAC sequence are opposite numbers to each other.

8. A two-dimensional synchronization method based on OFDM system according to claim 7, characterized in that: The method of accurately estimating and compensating the frequency offset of the received signal according to the correct synchronization position includes: The roughly synchronized signal is divided into two parts and stored in two RAMs respectively. Only the part with repeated structure is calculated, and the opposite part is not calculated. The two signals stored in two RAMs are conjugate-multiplied and accumulated, and the phase of the accumulated value is calculated to obtain the frequency deviation value. The CORDIC IP core is then used to perform phase rotation on the signal based on the estimated frequency offset value to achieve frequency offset compensation.

Images