A 5G self-organizing network coarse synchronization segmentation design method based on domestic chips

Abstract

This invention provides a coarse synchronization segmentation design method for 5G self-organizing networks based on domestically produced chips, relating to the field of mobile communication technology. The method includes the following steps: S1, capturing air interface signals according to the sampling rate issued by OAM and the common NRSSB cycle; S2, calculating the frequency difference and performing zero-frequency shifting based on the center frequency of the receiving bandwidth; S3, segmenting and merging the zero-frequency shifted signal into windows, and performing downsampling processing on the signal; S4, performing AGC gain processing on the signal; S5, performing in-window decision on the signal; S6, after all in-window signals have completed in-window decision, performing inter-window decision; S7, setting a timeout threshold and dynamically adjusting the window length. The beneficial effects of this invention are: reduced computational complexity, efficient utilization of computational resources, improved computational efficiency, and high accuracy and reliability of signal detection.

Description

Technical Field

[0001] This invention relates to the field of mobile communication technology, and in particular to a coarse synchronization segmentation design method for 5G self-organizing networks based on domestically produced chips. Background Technology

[0002] 5G communication system is short for fifth-generation mobile communication technology, which is a new generation of cellular mobile communication technology. Building upon the advantages of its predecessors, it incorporates numerous innovations and upgrades, aiming to provide higher data transmission rates, lower latency, more connected devices, and more reliable communication services.

[0003] 5G coarse synchronization technology plays a crucial and fundamental role in the entire 5G communication system. It is the first step in establishing an effective communication link between communication devices (such as base stations and mobile terminals) and a prerequisite for subsequent operations such as fine synchronization, channel estimation, and data demodulation. With the development of communication technology, communication data is growing exponentially. Existing 5G coarse synchronization technologies often face problems such as massive data volume, computational complexity, low data processing efficiency, and low signal detection accuracy when performing coarse synchronization processing.

[0004] How to solve the above-mentioned technical problems is the challenge facing this invention. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a coarse synchronization segmentation design method for 5G self-organizing networks based on domestically produced chips, which reduces computational complexity, efficiently utilizes computing resources, improves data processing efficiency, and enhances the accuracy and reliability of signal detection.

[0006] The technical solution adopted by this invention to solve its technical problem is as follows: This invention provides a coarse synchronization segmentation design method for 5G self-organizing networks based on domestically produced chips, including the following steps:

[0007] S1. Based on the sampling rate issued by OAM, capture the air interface signal according to the common NRSSB cycle;

[0008] Based on the common 20ms NRSSB period, air interface data is captured at 25ms intervals to ensure complete NRSSB data is captured. If the NRSSB data period is longer than 20ms, the capture start point can be offset from the basic NRSSB period, and multiple captures can be performed until complete NRSSB data is captured.

[0009] S2. Calculate the frequency difference based on the center frequency of the receiving bandwidth and perform zero-frequency shift;

[0010] S3. The signal after zero-frequency shift is segmented and merged into a window, and the signal is downsampled.

[0011] S4. Perform AGC gain processing on the signal;

[0012] S5. Set the sliding-related sequence, perform related calculations, and make in-window decisions;

[0013] S6. After all in-window signals have completed in-window decision-making, perform inter-window decision-making.

[0014] S7. Set the timeout threshold and dynamically adjust the window length.

[0015] Preferably, step S1 further includes ping-pong design for the data buffer.

[0016] While processing ping, data can continue to be moved to pong, improving data processing efficiency, allowing data reading and writing to be performed in parallel, reducing data processing waiting time, and ensuring that data can continuously and stably flow into subsequent processing stages.

[0017] Preferably, step S2 specifically comprises:

[0018] The frequency difference between the current NRSSB center frequency and the receiving bandwidth center frequency is calculated based on the receiving bandwidth center frequency. The data at the current NRSSB center frequency is then moved to the receiving bandwidth center frequency position by taking the modulus of the subcarrier bandwidth.

[0019] Preferably, step S3 specifically comprises:

[0020] The signal after zero-frequency shift is segmented, and overlapping data is compensated for each segment. The segmented signals are then merged into a window. The signal is downsampled by a four-stage CIC filter and a CIC compensation filter. The four-stage CIC filter downsamples by 16 times, and the CIC compensation filter downsamples by 2 times.

[0021] The four-stage CIC filter sequentially performs two moving average filters, one 4x downsampling, two moving average filters, and one 4x downsampling to achieve a 16x downsampling.

[0022] The signal is divided into 250 segments and merged into windows. Every 4 segments are merged into one window. The length of one segment is 25089+513. The data overlapping at the two adjacent ends is 513 points. The first segment is padded with 526 zeros forward and the last segment is padded with 513 zeros backward. Data that is not enough for one window is padded with zeros. The data overlapping between adjacent windows is 273.

[0023] By compensating for overlapping data, the impact of signal abrupt changes at segment boundaries is reduced, thereby improving signal quality and accuracy.

[0024] Downsampling can reduce the amount of data, lower the computational complexity of subsequent processing, and at the same time preserve the key features of the signal to a certain extent, thereby improving processing efficiency.

[0025] Preferably, step S4 specifically comprises:

[0026] S40. Calculate the average power of the complex signal based on the sliding window, and set a target power;

[0027] S41. Calculate the AGC factor and perform a shift operation based on the final calculation result.

[0028] AGC gain processing can automatically adjust the gain according to the signal strength, ensuring that the signal is processed within an appropriate amplitude range, avoiding processing errors or information loss caused by signals that are too strong or too weak, and improving the accuracy of signal processing.

[0029] Preferably, step S5 specifically comprises:

[0030] S50. According to the 3GPP protocol, three sets of sliding correlation sequences and three sets of local sequences are set to perform sliding correlation operations. The three sets of local sequences are improved to 256 points through inverse fast Fourier transform and 18 CP lengths are added.

[0031] S51. Read the maximum correlation value of the sliding correlation value within the window, the position of the maximum correlation value within the window (corresponding column number) and the group number (corresponding row number minus 1), calculate the mean of the correlation values ​​corresponding to the group numbers of the other two cell groups where the maximum correlation value is located within the window, and obtain the noise value.

[0032] S52. Set a threshold value, calculate the product of the threshold value and the noise value, and then compare it with the maximum correlation value:

[0033] If the maximum correlation value is greater than the product of the threshold value and the noise value, calculate the synchronization position, and add the maximum correlation value of the decision window, the group number, and the synchronization position to the candidate set.

[0034] Synchronization position = position of the maximum correlation value within the window + (k-1)*3072, where k is 1 when calculating in segments.

[0035] If the maximum correlation value is less than or equal to the product of the threshold value and the noise value, the decision window is ignored.

[0036] Preferably, step S6 specifically comprises:

[0037] Find the maximum value of the sliding correlation value in the candidate set, perform a shift operation based on the AGC factor, obtain the window index and the in-window index, and calculate the starting position of the PSS sample point. That is, window index (63) * window length (3072) + in-window index + appropriate inter-window estimation position (generally 1 / 10 of the window length).

[0038] Since S4 boosts the signal, the candidate data for the current segment needs to be restored based on the segmented AGC factor calculated by the AGC module. This means shifting the peak data to the right according to the AGC factor to avoid data distortion caused by signal boosting and affecting the detection results.

[0039] Preferably, step S7 specifically comprises:

[0040] Set a timeout threshold, initial window length, and window length adjustment range. A timer records the time taken to process signals within the window and compares it to the timeout threshold. If no valid signal is detected within the timeout threshold, the window length is increased; if a signal is detected within the timeout threshold but exhibits large fluctuations or frequent false positives, the window length is decreased. The adjusted window length takes effect the next time new data is processed.

[0041] The initial window length is determined based on prior knowledge of the signal, such as signal bandwidth and transmission mode. For example, for a 100MHz bandwidth signal in 5G NR, the initial window length can be set to include a certain number of symbol periods, such as the time length corresponding to 10 symbol periods. A minimum and maximum window length are then set. The minimum value must ensure sufficient signal features are included for detection, while the maximum value should not be too long, leading to excessive processing latency.

[0042] If no valid signal is detected within the timeout threshold, consider increasing the window length to increase signal accumulation time and improve the ability to detect weak signals. For example, each increase can be 20%-50% of the current window length, but it cannot exceed the set maximum window length. If a signal is detected within the timeout threshold, but the signal fluctuates greatly or false positives occur frequently, consider decreasing the window length to help track signal changes more quickly and improve the timeliness of signal detection. Each decrease can be 10%-30% of the current window length, but it cannot be less than the set minimum window length.

[0043] The beneficial effects of this invention are as follows: reduced computational complexity, efficient utilization of computing resources, improved data processing efficiency, and high accuracy and reliability of signal detection. When capturing signals, a ping-pong design is implemented in the data buffer, allowing data reading and writing to be performed in parallel, rationally allocating resources, and improving data processing efficiency. Overlapping data is compensated for after signal segmentation, reducing the impact of signal abrupt changes at segment boundaries and improving signal quality and accuracy. Downsampling of the data effectively reduces the amount of data, lowers the complexity of subsequent calculations, and improves computational efficiency. AGC gain processing automatically adjusts the gain according to the signal strength, ensuring that the signal is processed within a suitable amplitude range, avoiding processing errors or information loss due to excessively strong or weak signals, and improving the accuracy of signal processing. Setting a timeout threshold enables dynamic adjustment of the window length, rationally allocating computing resources, and effectively improving data processing efficiency and the timeliness of signal detection. Attached Figure Description

[0044] Figure 1 This is a diagram illustrating the method steps of the present invention. Detailed Implementation

[0045] To clearly illustrate the technical features of this solution, the following detailed implementation method will be used to explain the solution.

[0046] See Figure 1 As shown, this embodiment is a coarse synchronization segmentation design method for 5G self-organizing networks based on domestically produced chips, including the following steps:

[0047] S1. Based on the sampling rate issued by OAM, capture the air interface signal according to the common NRSSB cycle;

[0048] Based on the common 20ms NRSSB period, air interface data is captured at 25ms intervals to ensure complete NRSSB data is captured. If the NRSSB data period is longer than 20ms, the capture start point can be offset from the basic NRSSB period, and multiple captures can be performed until complete NRSSB data is captured.

[0049] S2. Calculate the frequency difference based on the center frequency of the receiving bandwidth and perform zero-frequency shift;

[0050] S3. The signal after zero-frequency shift is segmented and merged into a window, and the signal is downsampled.

[0051] S4. Perform AGC gain processing on the signal;

[0052] S5. Set the sliding-related sequence, perform related calculations, and make in-window decisions;

[0053] S6. After all in-window signals have completed in-window decision-making, perform inter-window decision-making.

[0054] S7. Set the timeout threshold and dynamically adjust the window length.

[0055] Step S1 also includes ping-pong design for the data buffer.

[0056] While processing ping, data can continue to be moved to pong, improving data processing efficiency, allowing data reading and writing to be performed in parallel, reducing data processing waiting time, and ensuring that data can continuously and stably flow into subsequent processing stages.

[0057] Step S2 is as follows:

[0058] Calculate the frequency difference between the NRSSB center frequency and the receiving bandwidth center frequency within 50M above and below the receiving bandwidth center frequency, take the modulus of the subcarrier bandwidth, and move the data of the current NRSSB center frequency to the receiving bandwidth center frequency position.

[0059] Step S3 is as follows:

[0060] The signal after zero-frequency shift is segmented, and overlapping data is compensated for each segment. The segmented signals are then merged into a window. The signal is downsampled by a four-stage CIC filter and a CIC compensation filter. The four-stage CIC filter downsamples by 16 times, and the CIC compensation filter downsamples by 2 times.

[0061] The four-stage CIC filter sequentially performs two moving average filters, one 4x downsampling, two moving average filters, and one 4x downsampling to achieve a 16x downsampling.

[0062] The signal is divided into 250 segments and merged into windows. Every 4 segments are merged into one window. The length of one segment is 25089+513. The data overlapping at the two adjacent ends is 513 points. The first segment is padded with 526 zeros forward and the last segment is padded with 513 zeros backward. Data that is not enough for one window is padded with zeros. The data overlapping between adjacent windows is 273.

[0063] By compensating for overlapping data, the impact of signal abrupt changes at segment boundaries is reduced, thereby improving signal quality and accuracy.

[0064] Downsampling can reduce the amount of data, lower the computational complexity of subsequent processing, and at the same time preserve the key features of the signal to a certain extent, thereby improving processing efficiency.

[0065] Step S4 is as follows:

[0066] S40. Calculate the average power of the complex signal based on the sliding window, and set a target power;

[0067] S41. Calculate the AGC factor and perform a shift operation based on the final calculation result.

[0068] AGC gain processing can automatically adjust the gain according to the signal strength, ensuring that the signal is processed within an appropriate amplitude range, avoiding processing errors or information loss caused by signals that are too strong or too weak, and improving the accuracy of signal processing.

[0069] Step S5 is as follows:

[0070] S50. According to the 3GPP protocol, three sets of sliding correlation sequences and three sets of local sequences are set to perform sliding correlation operations. The three sets of local sequences are improved to 256 points through inverse fast Fourier transform and 18 CP lengths are added.

[0071] S51. Read the maximum correlation value of the sliding correlation value within the window, the position of the maximum correlation value within the window (corresponding column number) and the group number (corresponding row number minus 1), calculate the mean of the correlation values ​​corresponding to the group numbers of the other two cell groups where the maximum correlation value is located within the window, and obtain the noise value.

[0072] S52. Set a threshold value, calculate the product of the threshold value and the noise value, and then compare it with the maximum correlation value:

[0073] If the maximum correlation value is greater than the product of the threshold value and the noise value, calculate the synchronization position, and add the maximum correlation value of the decision window, the group number, and the synchronization position to the candidate set.

[0074] Synchronization position = position of the maximum correlation value within the window + (k-1)*3072, where k is 1 when calculating in segments.

[0075] If the maximum correlation value is less than or equal to the product of the threshold value and the noise value, the decision window is ignored.

[0076] Step S6 is as follows:

[0077] Find the maximum value of the sliding correlation value in the candidate set, perform a shift operation based on the AGC factor, obtain the window index and the in-window index, and calculate the starting position of the PSS sample point. That is, window index (63) * window length (3072) + in-window index + appropriate inter-window estimation position (generally 1 / 10 of the window length).

[0078] Since S4 boosts the signal, the candidate data for the current segment needs to be restored based on the segmented AGC factor calculated by the AGC module. This means shifting the peak data to the right according to the AGC factor to avoid data distortion caused by signal boosting and affecting the detection results.

[0079] Step S7 is as follows:

[0080] Set a timeout threshold, initial window length, and window length adjustment range. A timer records the time taken to process signals within the window and compares it to the timeout threshold. If no valid signal is detected within the timeout threshold, the window length is increased; if a signal is detected within the timeout threshold but exhibits large fluctuations or frequent false positives, the window length is decreased. The adjusted window length takes effect the next time new data is processed.

[0081] In this embodiment, the timeout threshold is set to the length of one time slot.

[0082] The initial window length is determined based on prior knowledge of the signal, such as signal bandwidth and transmission mode. For example, for a 100MHz bandwidth signal in 5G NR, the initial window length can be set to include a certain number of symbol periods, such as the time length corresponding to 10 symbol periods. A minimum and maximum window length are set; the minimum value must ensure sufficient signal features are included for detection, while the maximum value cannot be too long, leading to excessive processing latency.

[0083] If no valid signal is detected within the timeout threshold, consider increasing the window length to increase signal accumulation time and improve the ability to detect weak signals. For example, each increase can be 20%-50% of the current window length, but it cannot exceed the set maximum window length. If a signal is detected within the timeout threshold, but the signal fluctuates greatly or false positives occur frequently, consider decreasing the window length to help track signal changes more quickly and improve the timeliness of signal detection. Each decrease can be 10%-30% of the current window length, but it cannot be less than the set minimum window length.

[0084] The technical features of this invention not described can be implemented by or using existing technology, and will not be repeated here. Of course, the above description is not a limitation of this invention, and this invention is not limited to the examples above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of this invention should also be within the protection scope of this invention.

Claims

1. A method for designing a coarse synchronization segment of a 5G self-organizing network based on a domestic chip, characterized in that, Includes the following steps: S1. Capture the air interface signal according to the sampling rate issued by OAM and the common NRSSB cycle; S2. Calculate the frequency difference based on the center frequency of the receiving bandwidth and perform zero-frequency shift; S3. The signal after zero-frequency shift is segmented and merged into a window, and the signal is downsampled. S4. Perform AGC gain processing on the signal; S5. Set the sliding-related sequence, perform related calculations, and make in-window decisions; S6. After all in-window signals have completed in-window decision-making, perform inter-window decision-making. S7. Set the timeout threshold and dynamically adjust the window length; Step S3 specifically involves: The signal after zero-frequency shift is segmented, and overlapping data is compensated for each segment. The segmented signals are then merged into a window. The signal is downsampled by a four-stage CIC filter and a CIC compensation filter. The four-stage CIC filter downsamples by 16 times, and the CIC compensation filter downsamples by 2 times. Step S5 specifically involves: S50. According to the 3GPP protocol, three sets of sliding correlation sequences and three sets of local sequences are set to perform sliding correlation operations. The three sets of local sequences are improved to 256 points through inverse fast Fourier transform and 18 CP lengths are added. S51. Read the maximum correlation value, the position of the maximum correlation value within the window, and the group number of the sliding correlation value within the window. Take the correlation values ​​corresponding to the group numbers of the other two cell groups where the maximum correlation value is located within the window as the mean value to obtain the noise value. S52. Set a threshold value, calculate the product of the threshold value and the noise value, and then compare it with the maximum correlation value: If the maximum correlation value is greater than the product of the threshold value and the noise value, calculate the synchronization position, and add the maximum correlation value of the decision window, the group number, and the synchronization position to the candidate set. If the maximum correlation value is less than or equal to the product of the threshold value and the noise value, the decision window is ignored. Step S6 specifically involves: Find the maximum value of the sliding correlation value in the candidate set, perform a shift operation according to the AGC factor, obtain the window index and the in-window index, and calculate the starting position of the PSS sample point; Step S7 specifically involves: Set the timeout threshold, initial window length, and window length adjustment range. Record the time taken to process the signal within the window using a timer and compare it with the timeout threshold. If no valid signal is detected within the timeout threshold, the window length is increased. If a signal is detected within the timeout threshold, but the signal fluctuates greatly or misjudgments are frequent, the window length is decreased.

2. The method of claim 1, wherein the method is based on a homegrown chip-based 5G self-organizing network coarse synchronization segmentation design. Step S1 also includes ping-pong design for the data buffer.

3. The method of claim 1, wherein the method is based on a homegrown chip-based 5G self-organizing network coarse synchronization segmentation design. Step S2 specifically involves: The frequency difference between the current NRSSB center frequency and the receiving bandwidth center frequency is calculated based on the receiving bandwidth center frequency. The data at the current NRSSB center frequency is then moved to the receiving bandwidth center frequency position by taking the modulus of the subcarrier bandwidth.

4. The method of claim 1, wherein the method is a design method of a 5G self-organizing network coarse synchronization segmentation based on a homegrown chip. Step S4 specifically involves: S40. Calculate the average power of the complex signal based on the sliding window, and set a target power; S41. Calculate the AGC factor and perform a shift operation based on the final calculation result.

Images