Optical fiber bending loss test system and method based on spectrum analysis

Abstract

The present disclosure provides a fiber bending loss test system and method based on spectral analysis, which relates to the field of fiber detection technology. This application uses multi-wavelength scanning to obtain power values ​​and interference signals, and transmits at least two measurement data in parallel within one transmission cycle; writes to the storage array according to the wavelength index and batch identifier, pauses and fills in points when insufficient alignment is detected; after filling in points, transmission is resumed and distributed loss information is calculated. In this way, data integrity, transmission throughput and real-time loss calculation accuracy are taken into account in a large-scale measurement environment, effectively avoiding the extension of measurement cycles and index confusion in conventional linear detection or fixed offset methods, and significantly improving the efficiency and reliability of fiber loss testing.

Description

Technical Field

[0001] The present disclosure relates to the field of optical fiber detection technology, and in particular to an optical fiber bending loss testing system and method based on spectral analysis. Background Art

[0002] Fiber bend loss testing, especially distributed fiber measurement based on spectral analysis, often requires detecting attenuation changes caused by fiber bending at multiple wavelengths while simultaneously collecting large amounts of data, such as power values ​​and interference signals. Because bending increases light leakage outside the fiber core, the associated loss distribution requires real-time monitoring at high resolution and multiple wavelengths. However, traditional measurement systems, which mostly use serial transmission or simple batch storage, struggle to timely process the large amounts of data generated by multi-wavelength scanning and are prone to alignment errors or index conflicts when writing data to memory. This not only prolongs the measurement cycle and increases the risk of data loss, but also directly impacts the accuracy of bend loss and the phase resolution of interference signals, making it difficult to meet the growing demand for high-precision, multi-wavelength, real-time monitoring during fiber deployment. Therefore, there is an urgent need for a highly efficient and reliable fiber bend loss testing method based on spectral analysis to address the shortcomings of existing technologies in data acquisition, transmission, and alignment storage, and to improve the real-time and accuracy of fiber bend loss measurements. Summary of the Invention

[0003] In response to the deficiencies of the existing technology, the present application provides a fiber bending loss testing system and method based on spectral analysis.

[0004] In a first aspect, the present application provides a method for testing optical fiber bending loss based on spectral analysis, comprising:

[0005] Perform multi-wavelength scanning on the optical fiber under test to obtain measurement data at each wavelength; wherein the measurement data includes: power value and interference signal;

[0006] Packing the measurement data according to a predefined wavelength index rule, and transmitting at least two measurement data in parallel within one transmission cycle;

[0007] Receive the parallel transmitted measurement data, and write each measurement data into a corresponding address of the storage array by mapping the index to the storage address according to the wavelength index and batch identifier carried by each measurement data;

[0008] In response to detecting that the number of measurement data in the current batch does not meet the preset alignment length, triggering a transmission pause and performing a point filling operation;

[0009] In response to the completion of the patching operation, transmission is resumed, and optical fiber loss calculation is performed on the multi-wavelength data written in the storage array to generate distributed loss information of the tested optical fiber.

[0010] Optionally, packaging the measurement data according to a predefined wavelength index rule and transmitting at least two measurement data in parallel within one transmission cycle includes:

[0011] Adding header information to the measurement data of each wavelength based on the wavelength index rule and the scanning order, wherein the header information includes a wavelength index, a batch identifier, a serial number, and a timestamp;

[0012] Sending the measurement data of each wavelength in parallel based on the packet header information;

[0013] The header information of the parallel transmission is parsed at the receiving end, the corresponding measurement data is mapped to the target address in the storage array according to the wavelength index and batch identifier, and the arrival order is rearranged and / or supplemented according to the sequence number and timestamp.

[0014] Optionally, the wavelength index rule includes:

[0015] Before the scan begins, a wavelength list is created that is arranged in descending order according to the working wavelength range of the optical fiber under test, and a basic index corresponding to its order is assigned to each wavelength;

[0016] In each scanning cycle of a transmission batch, a dynamic offset related to the batch identifier is applied to the basic index to generate a wavelength index actually used;

[0017] detecting whether the actually used wavelength index conflicts with a previously allocated wavelength index;

[0018] In response to detecting a conflict, adaptively adjusting the dynamic offset until there is no conflict and then packaging the corresponding wavelength;

[0019] Record the correspondence between the established wavelength index and wavelength and batch identifier, and keep it consistent during data packaging and parallel transmission.

[0020] Optionally, in response to detecting a conflict, adaptively adjusting the dynamic offset until there is no conflict and then packaging the corresponding wavelengths includes:

[0021] Read the allocated wavelength index set in the current batch from the conflict detection record;

[0022] Incrementing the dynamic offset, and recalculating the actually used wavelength index based on the updated dynamic offset;

[0023] Determining whether the recalculated wavelength index still conflicts with an existing wavelength index in the wavelength index set;

[0024] In response to there still being a conflict, the increment of the state offset and the conflict detection are repeatedly performed until there is no conflict;

[0025] The finally determined actually used wavelength index is written into the wavelength index set, and the measurement data of the corresponding wavelength is continuously packaged.

[0026] Optionally, increasing the dynamic offset includes:

[0027] When performing multi-wavelength scanning, local extreme value moments or several phase trigger points are extracted from the real-time interference signal and combined with the current count value of the system clock to generate a pseudo-random seeding value;

[0028] The pseudo-random seed value is used for initial and iterative random increments of the dynamic offset.

[0029] Optionally, increasing the dynamic offset further includes:

[0030] When a batch filling operation is detected to be triggered and / or completed, the pseudo-random seeding value is recalculated based on the latest local extreme value moment and the system clock;

[0031] After the patching is completed and the transmission is resumed, the updated dynamic offset is synchronized with the allocated index and batch identification mapping table.

[0032] Optionally, the triggering of pausing transmission and performing a point-filling operation includes:

[0033] Lock the wavelength index allocation status of the current batch and suspend sending subsequent wavelength measurement data at the transmission end;

[0034] During the pause, identifying a set of wavelength indices that have not yet reached the alignment length, and performing a local rescan or interpolation calculation based on the set to generate missing measurement data;

[0035] Fill the measurement data obtained by repeated scanning or interpolation calculation into the corresponding position of the storage array according to the established wavelength index and batch identification;

[0036] Update the wavelength index-dynamic offset mapping table within the batch;

[0037] After confirming that the data volume of the current batch has reached the alignment length, parallel transmission is resumed, and multi-wavelength scanning and data packaging are continued based on the updated mapping table.

[0038] Optionally, generating missing measurement data includes:

[0039] Counting the distribution of missing measurement data in the wavelength index set, and determining whether it exceeds a preset local scan threshold;

[0040] In response to the distribution of the missing measurement data being greater than the local scanning threshold, activating a local re-scanning process to re-send measurement light pulses or interference signal samples only to the optical fiber section corresponding to the missing wavelength index to obtain corresponding power values ​​and / or interference signals;

[0041] In response to the distribution of the missing measurement data being less than or equal to the local scanning threshold, or the scanning resources entering a preset tight state, performing interpolation calculation based on distributed phase information of adjacent sampling points or adjacent wavelengths to generate the missing measurement data;

[0042] The missing measurement data obtained by local repeated scanning or interpolation calculation are filled into the corresponding position of the storage array according to the established wavelength index and batch identification.

[0043] Optionally, in response to the completion of the patching operation, resuming transmission, and performing optical fiber loss calculation on the multi-wavelength data written in the storage array to generate distributed loss information of the tested optical fiber includes:

[0044] Reading and registering the multi-wavelength data, and aligning the batch identifiers and timestamps;

[0045] performing a ratio operation and / or phase demodulation on the power values ​​and / or interference signals included in the multi-wavelength data;

[0046] Based on the ratio calculation result and / or phase distribution, it is mapped to the axial position of the optical fiber to generate distributed loss information.

[0047] In a second aspect, the present application provides an optical fiber bending loss testing system based on spectral analysis, comprising:

[0048] An acquisition unit is used to perform multi-wavelength scanning on the optical fiber under test and obtain measurement data at each wavelength; wherein the measurement data includes: power value and interference signal;

[0049] A transmission unit, configured to package the measurement data according to a predefined wavelength index rule, and transmit at least two measurement data in parallel within one transmission cycle;

[0050] a mapping unit, configured to receive the parallel transmitted measurement data and write each measurement data into a corresponding address of the storage array by mapping the index to the storage address according to the wavelength index and batch identifier carried by each measurement data;

[0051] a first processing unit, configured to trigger a transmission suspension and perform a point filling operation in response to detecting that the number of measurement data in a current batch does not satisfy a preset alignment length;

[0052] The second processing unit is configured to resume transmission in response to completion of the patching operation, and perform optical fiber loss calculation on the multi-wavelength data written in the storage array to generate distributed loss information of the tested optical fiber.

[0053] Compared with the existing technology, the present application realizes efficient management and alignment of massive distributed optical fiber measurement data by combining dynamic offset and wavelength index rules on the basis of multi-wavelength scanning and real-time parallel transmission; when insufficient alignment length is detected, local repeated scanning or interpolation is used to quickly supplement the missing data, and index conflict problems are resolved through adaptive random seeding values, thereby taking into account data integrity, transmission throughput and real-time loss calculation accuracy in a large-scale measurement environment, effectively avoiding the extension of measurement cycle and index confusion in conventional linear detection or fixed offset methods, and significantly improving the efficiency and reliability of optical fiber loss testing. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A flowchart of a method for testing optical fiber bending loss based on spectral analysis provided in an embodiment of the present application;

[0055] Figure 2 A flowchart of a method for increasing the dynamic offset provided in an embodiment of the present application;

[0056] Figure 3 Schematic diagram of an optical fiber bending loss test system based on spectral analysis provided in an embodiment of the present application.

[0057] Explanation of reference numerals: 10, acquisition unit; 20, transmission unit; 30, mapping unit; 40, first processing unit; 50, second processing unit. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0059] See also Figure 1 FIG. 1 is a flow chart of a method for testing optical fiber bending loss based on spectral analysis according to an embodiment of the present application. The method includes steps S101 to S105, wherein:

[0060] S101: Perform multi-wavelength scanning on the optical fiber under test to obtain measurement data at each wavelength; wherein the measurement data includes: power value and interference signal;

[0061] S102: Packing the measurement data according to a predefined wavelength index rule, and transmitting at least two measurement data in parallel within one transmission cycle;

[0062] S103: receiving the parallel transmitted measurement data, and writing each measurement data into a corresponding address of a storage array by mapping the index to a storage address according to a wavelength index and a batch identifier carried by each measurement data;

[0063] S104: In response to detecting that the number of measurement data in the current batch does not meet the preset alignment length, triggering a transmission suspension and performing a point filling operation;

[0064] S105: In response to the completion of the point-filling operation, transmission is resumed, and optical fiber loss calculation is performed on the multi-wavelength data written into the storage array to generate distributed loss information of the measured optical fiber.

[0065] In practice, a multi-wavelength scanning light source and corresponding fiber measurement unit are configured to illuminate the fiber under test at multiple wavelengths and collect echoes. Using a pre-set scanning cycle and spectral allocation strategy, multiple probe light pulses of different wavelengths are generated sequentially or simultaneously within a specific time window. Excitation at these wavelengths generates scattered or interference signals from the fiber under test. After photoelectric detection and conversion, a multi-channel output containing power values ​​and interference signals is obtained.

[0066] Next, the multi-channel output is digitally sampled and packaged for each measurement data item according to a predefined wavelength indexing rule. In this embodiment, this indexing rule can be based on the wavelength configuration of the light source or combined with the order information of each scan, so that each measurement data item carries the identifier corresponding to its wavelength and batch. Because at least two measurement data items must be transmitted in parallel within a transmission cycle, after packaging, these data items are sent in parallel to the receiver via a high-speed communication interface (such as PCIe or Ethernet), reducing the latency and error accumulation that may be associated with serial transmission.

[0067] At the receiving end, based on the wavelength index and batch identifier carried by each piece of measurement data, a mapping relationship is established from index to storage address, accurately writing the data to the corresponding address in the storage array. This way, even when the data volume is huge, the data of each wavelength and batch can still be stored in an orderly manner, preventing confusion in subsequent processing.

[0068] In some cases, if the system detects that the current batch of data has not yet met the pre-set alignment length (e.g., due to missing data at certain wavelengths, reception delays, or interruptions caused by interference), transmission is paused and a point-fill operation is performed. At this point, the system can stop receiving further batches of data to wait for late packets to arrive, or perform local rescanning or interpolation calculations to fill in the missing measurement data. Once the point-fill operation is complete and alignment requirements are confirmed, the system resumes transmission and continues processing subsequent batches of data, ensuring the integrity and timing stability of subsequent wavelength data.

[0069] Finally, the aligned multi-wavelength data is fed into the fiber loss calculation module. By comparing power values ​​or performing phase demodulation on the interference signal, loss information at different locations on the tested fiber is obtained. This embodiment typically uses phase demodulation to process the interference signal to more accurately detect subtle changes along the fiber. However, the power ratio method or other known algorithms can also be used to calculate insertion loss or backscatter loss. Ultimately, a distributed loss curve or chart is generated, visually displaying the attenuation distribution and potential fault points of the tested fiber.

[0070] For example, in a large city's backbone fiber network, distributed loss testing is required to quickly locate and monitor the loss distribution of fiber lines, often tens to hundreds of kilometers long. Because these fibers carry multiple services concurrently within the carrier's network, the testing system must complete multi-wavelength scanning and data acquisition without interrupting existing communications, and promptly identify any fiber microbends, abnormal connector loss, or localized damage.

[0071] For example, a fiber optic test instrument is equipped with laser sources at wavelengths of 1550 nm and 1625 nm, and an additional 1310 nm wavelength is used to separate or compare with existing signals. These three wavelengths of probe light pulses are emitted along the fiber under test in a time-sharing or parallel manner.

[0072] The signals scattered or reflected in the optical fiber are split and combined and then sent to the high-speed detection module to obtain the power value corresponding to each wavelength and possible interference signals (such as using a coherent detection unit to collect the phase of the echo), ultimately forming multi-channel measurement data.

[0073] At the data acquisition end, the measurement data of the three wavelengths are packaged together using pre-set wavelength indexing rules. Each data packet carries information such as "wavelength index" and "batch identification" to facilitate differentiation by downstream modules.

[0074] In one transmission cycle, the measurement data of at least two wavelengths are transmitted in parallel through the PCIe or 10G Ethernet interface, thereby shortening the waiting time window of a single wavelength and achieving higher throughput.

[0075] The receiver writes the measurement data for each wavelength and each batch to the target address of the high-speed storage array through a mapping relationship (index-storage address). Due to the dual marking of batch identification and wavelength index, correct alignment and storage can be ensured even if there are short delays in the network.

[0076] If it detects that the echo data of certain wavelengths does not arrive within the expected time (for example, due to network jitter or local packet loss), the system triggers a warning that "the current batch does not meet the alignment length", suspends subsequent data transmission, and waits for or supplements the missing data.

[0077] Because this trunk fiber occasionally experiences interference due to external construction, which can cause some measurement data to be delayed or frames to be lost, the system detects insufficient alignment length and pauses sending new batches.

[0078] During this period, index matching is performed on the late data or a "local rescan" is performed on the instrument side to obtain the missing measurement data. If the amount of missing data is small and rescanning is not required, an interpolation algorithm can be used to approximate compensation for the interference signal.

[0079] When the filling is completed, the system confirms that the amount of data in the current batch has reached the pre-set alignment standard, and then resumes transmission and processes the next batch of data.

[0080] After transmission is restored, the aligned and written multi-wavelength data (including data obtained by timely supplementation or interpolation) is read out and demodulated using the power ratio method or the interference signal phase.

[0081] For power value measurement, the system can calculate the insertion loss of the optical fiber at different positions based on the echo intensity ratio of each wavelength; for interference signal measurement, the system uses a phase demodulation algorithm to perform more precise loss monitoring at each point along the optical fiber.

[0082] Combined with fiber length calibration, distributed loss information along the entire fiber can be obtained. If the loss is too high at a certain location, the system will issue an immediate alarm, allowing operators to quickly locate the fault point and arrange for repairs.

[0083] Optionally, packaging the measurement data according to a predefined wavelength index rule and transmitting at least two measurement data in parallel within one transmission cycle includes:

[0084] Adding header information to the measurement data of each wavelength based on the wavelength index rule and the scanning order, wherein the header information includes a wavelength index, a batch identifier, a serial number, and a timestamp;

[0085] Sending the measurement data of each wavelength in parallel based on the packet header information;

[0086] The header information of the parallel transmission is parsed at the receiving end, the corresponding measurement data is mapped to the target address in the storage array according to the wavelength index and batch identifier, and the arrival order is rearranged and / or supplemented according to the sequence number and timestamp.

[0087] In the specific implementation, the scanning end (such as FPGA or high-speed data acquisition card) first adds a header information to each measurement data according to the wavelength index rule and scanning order. The header contains the following four fields:

[0088] Wavelength index: used to indicate the wavelength channel to which the measurement data belongs. It can be dynamically generated based on a pre-established wavelength list or the switching order of a multi-wavelength light source.

[0089] Batch identification: used to distinguish adjacent transmission cycles or adjacent scanning batches to prevent data confusion when switching batches;

[0090] Sequence number: reflects the position of the data in the same batch in terms of time or logical sequence, which facilitates the receiving end to correct the arrival order;

[0091] Timestamp: records the time when the measurement data is generated or packaged, and is used to supplement or reorder when the network is out of order or delayed.

[0092] After the packet header is added, the system sends measurement data of at least two wavelengths in parallel within one transmission cycle.

[0093] For example, within the FPGA, DMA or multi-channel transmission operations can be triggered simultaneously for multiple wavelength channels, and data frames can be sent to the receiver in batches or streams via high-speed buses such as PCIe or Ethernet. This effectively improves throughput and reduces the waiting time and data queue accumulation caused by traditional serial transmission.

[0094] The receiving end can also be deployed in FPGA or host computer software to quickly classify each data block according to the wavelength index and batch identifier. In this case, a "wavelength index-storage address" mapping table can be used to directly write the data to the corresponding address range in the storage array. If the serial number or timestamp is found to be inconsistent with the expected order, the data is first stored in a temporary buffer, waiting for the previous data packet to arrive, and then reordered according to the serial number or timestamp order. If data at certain wavelengths is delayed, lost, or severely out of order, the serial number and timestamp can be used to determine the missing part, triggering a subsequent point filling mechanism (including waiting for late data or partial rescanning) to complete the data required for the current batch alignment length.

[0095] In this way, in scenarios involving large-scale parallel transmission of multi-wavelength scanned data, the receiver no longer relies on general buffering and linear indexing. Instead, it directly locates, aligns, and processes data at each wavelength and batch based on the packet header field, avoiding index confusion and data loss under high-throughput conditions. Furthermore, the combination of sequence numbers and timestamps enables the system to rearrange and fill in points in an out-of-order manner, ensuring the integrity of critical data and measurement accuracy even in the presence of network jitter or sudden delays, thereby improving the efficiency and robustness of multi-wavelength fiber loss testing.

[0096] Optionally, the wavelength index rule includes:

[0097] Before the scan begins, a wavelength list is created that is arranged in descending order according to the working wavelength range of the optical fiber under test, and a basic index corresponding to its order is assigned to each wavelength;

[0098] In each scanning cycle of a transmission batch, a dynamic offset related to the batch identifier is applied to the basic index to generate a wavelength index actually used;

[0099] detecting whether the actually used wavelength index conflicts with a previously allocated wavelength index;

[0100] In response to detecting a conflict, adaptively adjusting the dynamic offset until there is no conflict and then packaging the corresponding wavelength;

[0101] Record the correspondence between the established wavelength index and the wavelength and batch identifier, and keep it consistent during subsequent data packaging and parallel transmission.

[0102] In practice, before the scan officially begins, a wavelength list is created based on the operating wavelength range of the fiber under test. For example, if the operating band is from 1500 nm to 1600 nm, the wavelengths in this range can be sorted into a list, either at equal intervals or according to the available wavelengths configured in the device. These wavelengths are numbered 1, 2, 3, …, n, forming a basic index corresponding to each wavelength. This step aims to provide an ordered initial wavelength index, allowing for quick identification of the wavelength sequence during subsequent multi-wavelength parallel measurements.

[0103] At the start of each transmission batch or the beginning of the next scanning cycle, the system applies a dynamic offset to the aforementioned base index based on the batch identifier (e.g., batch_id) to generate the actual wavelength index. The dynamic offset can be determined by a combination of batch_id and a preset constant, or calculated in conjunction with other real-time parameters (e.g., random seeding value).

[0104] In multi-wavelength scanning, using only a fixed base index can easily lead to data identification confusion due to repeated use of the same index. By applying a different offset to each batch, new unique identifiers can be assigned to each wavelength in a "staggered" manner within the index space, effectively "dispersing conflicts and staggering occupancy" during large-scale batch measurements. This is similar to performing multiple probes in a hash data structure, except that this embodiment combines batch identifiers with base indexes to form a controllable and hierarchical index generation method, taking into account the needs of both batch and wavelength management.

[0105] After the system calculates the actual wavelength index to be used, it checks whether this index conflicts with a previously assigned wavelength index. A conflict often manifests itself as a wavelength already occupying the same index value within the current batch or time window. If this is detected, adaptive adjustment is triggered, whereby the dynamic offset is incremented or updated according to a specific algorithm or rule (such as incremental detection or random seeding) until a non-conflicting, usable value is found in the index space. This ensures that, in a highly parallel multi-wavelength environment, any scan batch receives an independent and unambiguous wavelength index.

[0106] The finalized wavelength index, along with the corresponding wavelength and batch identifier, is then recorded in an index mapping table maintained by the system. This mapping table is continuously referenced during subsequent data packaging and transmission to ensure that the sender and receiver use the same index identifier for the same wavelength and batch. This mapping table ensures that even in the event of network delays or out-of-order data, the "wavelength index + batch identifier" combination accurately assigns data to the correct storage address, ensuring the stability of multi-wavelength measurement data alignment.

[0107] By assigning basic indexes to the wavelength list and adding the dynamic offset introduced by the batch identifier, the wavelength indexes for each batch are initially dispersed, significantly reducing the probability of "index collisions" during large-scale batch parallel measurements. Adaptive adjustments are then made to remedy any minor collisions. This indexing rule is not only applicable to measurements of a small number of wavelengths, but can also be extended to distributed fiber optic testing scenarios that simultaneously manage dozens or even hundreds of wavelength channels.

[0108] Optionally, in response to detecting a conflict, adaptively adjusting the dynamic offset until there is no conflict and then packaging the corresponding wavelengths includes:

[0109] Read the allocated wavelength index set in the current batch from the conflict detection record;

[0110] Incrementing the dynamic offset, and recalculating the actually used wavelength index based on the updated dynamic offset;

[0111] Determining whether the recalculated wavelength index still conflicts with an existing wavelength index in the wavelength index set;

[0112] In response to there still being a conflict, the increment of the state offset and the conflict detection are repeatedly performed until there is no conflict;

[0113] The finally determined actually used wavelength index is written into the wavelength index set, and the measurement data of the corresponding wavelength is continuously packaged.

[0114] In large-scale multi-wavelength parallel transmission, simply overlaying the "base index + batch identifier" can easily lead to conflicts due to duplicate index occupancy within the same batch. Specifically, when there are many wavelengths and scanning batches frequently, different wavelengths may obtain the same index based on the same offset, resulting in data mismatches and batch confusion. While traditional static allocation or linear probing can alleviate this problem, in the real world of high-speed data writing and periodic index updates, it is still prone to numerous conflicts or duplicate probing fallbacks, resulting in reduced system throughput.

[0115] Therefore, in order to further improve the efficiency and dynamic adaptability of index allocation, after detecting a conflict, the present application uses an incremental or adaptive adjustment method to perform multiple attempts on the dynamic offset until an unoccupied index value is found.

[0116] In practice, before each batch scan or when switching between batches, a "wavelength index set" is maintained, recording the actual wavelength indexes that have been successfully assigned and are currently in use by the current batch. This set can be stored in high-speed registers within the FPGA or in a hash table within the master CPU or software to facilitate conflict detection.

[0117] For example, if the index set occupied by the first four wavelengths is {10, 12, 15, 18}, when the system needs to generate an index for a new wavelength, it will first compare it with this set.

[0118] Furthermore, if a new actually used wavelength index X is calculated and it is found that X ∈ {allocated index set}, it means that a conflict occurs.

[0119] Since the data is in the packaging state, in order to prevent further disorder in the batch, the packaging process will be temporarily locked and the adaptive adjustment process will be entered instead.

[0120] Furthermore, the current dynamic offset is incremented once or multiple times according to a custom algorithm (which can be configured as a fixed step size + 1, incremented by 1 each time a detection is made; or combined with a random seeding value for random increment).

[0121] For example, a formula example for recalculating an index is provided: index = base index + batch identifier × M + dynamic offset (new), where M is a constant or a variable parameter;

[0122] For example, if the current base index is 5, the batch identifier is 2, M=8, the initial dynamic offset is 3 and a conflict occurs, the system sets the offset + 1 to 4 and recalculates the index value: 5 + 2×8 + 4 = 5 + 16 + 4 = 25.

[0123] Furthermore, by comparing the new index value with the existing set, it is determined whether the conflict still exists; if no conflict is detected, it means that the current index can be used safely; if there is still a conflict, the dynamic offset will continue to be incremented or jumped in the next loop until the index value has no overlap with the occupied index set.

[0124] For example, if the new index 25 is not in {10, 12, 15, 18}, it can be considered a success, otherwise continue to increment the offset to 5 and so on until a free value is found.

[0125] When the system confirms that there is no conflict in the current index, it writes this index into the "wavelength index set" to prevent it from being reused by other wavelengths in the future; it releases the temporary lock state, resumes the packaging and parallel transmission of the wavelength data, and continues to accept index applications for the next wavelength.

[0126] Through the above adaptive adjustment, the current wavelength can always find an unoccupied index within this batch, thereby completing subsequent data packaging; the downstream (receiving end) can accurately classify the wavelength data based on the "wavelength index + batch identifier" to avoid disorder or overlap issues.

[0127] In this way, not only can index conflicts be effectively reduced, but adaptive adjustments can also be quickly completed and packaging can be continued when conflicts occur. This is extremely practical for multi-wavelength real-time measurement systems in application environments with high batch concurrency and high data integrity requirements.

[0128] See also Figure 2 FIG. 2 is a flowchart of a method for increasing the dynamic offset provided in an embodiment of the present application, comprising steps S201 to S202, wherein:

[0129] S201: when performing multi-wavelength scanning, extracting local extreme value moments or several phase trigger points from the real-time interference signal, combining them with the current count value of the system clock, and generating a pseudo-random seeding value;

[0130] S202: Using the pseudo-random seed value to perform initial and iterative random increments on the dynamic offset.

[0131] Among them, although index conflicts can be resolved to a certain extent by increasing the index by linear or fixed steps, when the index space is limited or the number of wavelengths is large, problems such as "periodic conflicts" or "low detection efficiency" may still occur.

[0132] To this end, this application combines the real-time information in the interference signal with the system clock to generate a high-entropy pseudo-random seeding value, which is introduced into the incremental calculation of the dynamic offset to form a randomized conflict resolution strategy, further improving the adaptability in multi-wavelength high-concurrency scenarios.

[0133] In a specific implementation, when performing multi-wavelength scanning, not only the power value of each wavelength is obtained, but also the phase or intensity change of the echo signal is obtained through coherent detection or other interferometric measurement means;

[0134] When there are micro-perturbations, scattering, attenuation and other phenomena inside the optical fiber, the interference signal will have a local extreme value (such as a peak or valley) at a specific time point, or a significant phase jump.

[0135] Furthermore, a continuous monitoring logic is set up (which can be deployed on an FPGA or high-speed processor) to perform threshold detection or extreme value tracking on the real-time interference signal. When an extreme value or a significant phase transition is detected, the timestamp T0 at that moment is recorded, or several phase trigger points (such as multiple peak and valley moments appearing in a single scan) are extracted.

[0136] For example, in a periodic scan, the peak moment of the interference signal corresponding to a certain wavelength may be detected to be T0 = 12345 (the internal count value of the system).

[0137] The current count value ClockCount of the system clock is usually continuously accumulated or counted in a loop. If the count value is combined with the local extreme value time T0 or the phase trigger point, a more random or unpredictable value R can be generated.

[0138] For example, the formula of R can be:

[0139]

[0140] in, Indicates bitwise XOR or other random mixing operations. HashFunction can be lightweight hashes such as MD5, CRC16, and mixed linear congruential hashes.

[0141] The R obtained through this process is the pseudo-random seeding value, which includes the transient characteristics of the interference signal and combines the dynamics of the system clock to ensure a higher overall entropy value.

[0142] When incremental detection is required for a certain wavelength index, the system uses the above pseudo-random seed value R as the initial increment or iterative increment;

[0143] For example, if the initial dynamic offset Offset_Init = 10 and the pseudo-random seed value R = 7, the current increment can be set to Offset_Next = Offset_Init + R = 17. If a conflict still exists, the increment can be continued by combining more random operations (such as updating R again), where Offset_Next is used as the incremental probe value for the next index allocation. If the conflict is still not resolved, R can be updated again or Offset_Next can be incremented using a linear or random step size until a conflict-free actual index is found.

[0144] In this way, the increment of each detection is no longer a fixed or predictable constant, which greatly reduces the possibility of repeated collisions of index detection in a periodic or local range.

[0145] Furthermore, local extrema extraction of the interference signal can be performed continuously during each scan, or it can be centrally refreshed during batch switching or point filling operations. This pseudo-random seeding strategy can be used in parallel with fixed-step detection: after several linear detections, a random update is triggered to balance computational overhead and random effects.

[0146] This comprehensive scheme enables fast detection of free indexes in a massively parallel data environment, while also enabling higher uncertainty to avoid duplicate conflicts when needed.

[0147] In this way, by combining the local extreme value of the interference signal and the current count value of the system clock to construct a pseudo-random seeding value, it is possible to achieve irregular and unpredictable dynamic offset increments when index conflicts occur, effectively avoiding the drawbacks of index normalization or periodic conflicts under large-scale parallel measurements, thereby helping to achieve more efficient and scalable fiber distributed loss measurements.

[0148] Optionally, increasing the dynamic offset further includes:

[0149] When a batch filling operation is detected to be triggered and / or completed, the pseudo-random seeding value is recalculated based on the latest local extreme value moment and the system clock;

[0150] After the patching is completed and the transmission is resumed, the updated dynamic offset is synchronized with the allocated index and batch identification mapping table.

[0151] In multi-wavelength, highly parallel testing environments, point-padding operations are often used to compensate for missing or delayed batch data. When the system detects that the current batch has not yet reached the preset alignment length, it may pause new data transmission, perform local rescans, or perform interpolation calculations. Once the point-padding operation is triggered or completed, the index allocation state within the batch will also change. If the original random seeding value or dynamic offset is still used, index conflicts or duplication may occur, making it difficult to ensure consistent data writes when subsequent batches are restored.

[0152] To this end, the present application proposes to recalculate the pseudo-random seeding value based on the new local extreme value of the interference signal and the system clock information when the supplement point trigger and / or completion is detected to ensure the synchronous update of the dynamic offset and the index mapping table.

[0153] In practice, an event monitoring mechanism can be set up in the batch management module. When it detects that the alignment length of the current batch is insufficient and the decision is made to perform point filling, a "point filling trigger" signal is issued. Once the point filling is complete and the missing data has been confirmed to be filled or interpolated, a "point filling completed" signal is generated. Both types of signals can be used to trigger the recalculation of the random seed value below.

[0154] Similar to the method described above, when the supplement point is triggered or completed, a new timestamp T_new is extracted from the local extreme moment or phase trigger point of the latest interference signal, and the current count value ClockCount_new of the system clock is read at the same time;

[0155] T_new and ClockCount_new are combined through a hybrid algorithm such as XOR or hash to form a new pseudo-random seed value R_new, which serves as the basis for updating the subsequent dynamic offset.

[0156] This process can be considered as a "random source reset", ensuring that the random seeding values ​​that may have been locally repeated due to long-term use are replaced by new extreme value moments.

[0157] After generating R_new, the system will perform one or more iterative updates based on R_new according to the conflict resolution strategy.

[0158] If some wavelengths or batch indexes have changed during the point filling process, this new dynamic offset needs to be synchronized with the "allocated index and batch identification mapping table" to ensure that when transmission is subsequently restored, it can correctly determine which indexes have been used and which can still be allocated.

[0159] For example, if the system assigns a temporary index to wavelength A to quickly fill in data when filling in points, a new dynamic offset must be registered in the mapping table after recovery to avoid conflict with wavelength B again.

[0160] When the filling point is completed and the dynamic offset and mapping table are successfully updated, the system sends a "resume transmission" signal, allowing subsequent batches or new wavelength data to continue to be packaged and transmitted;

[0161] After this, the new index allocation will be based on the reset random seed value R_new, thereby improving the ability to resist conflicts and avoiding confusion in subsequent measurements caused by index "holes" or "overlaps" left during the filling period.

[0162] It should be noted that if high system reliability is required, a verification process can also be set up: before resuming transmission, the key wavelength index of the current batch or the next batch is pre-checked to confirm that there is no situation where multiple wavelengths compete for the same index due to temporary allocation of supplementary points;

[0163] Once a potential conflict or anomaly is discovered, the pseudo-random seed value update will be triggered again to eliminate the hidden dangers before formal recovery.

[0164] In this way, when the batch patching trigger and / or completion is detected, the pseudo-random seeding value is recalculated based on the new interference signal extreme value moment and the system clock, and the updated dynamic offset is synchronized with the index mapping table. This not only ensures the consistency of patching points and index allocation in multi-wavelength, high-concurrency measurement environments, but also provides a solid technical guarantee for subsequent continuous scanning and data recovery.

[0165] Optionally, the triggering of pausing transmission and performing a point-filling operation includes:

[0166] Lock the wavelength index allocation status of the current batch and suspend sending subsequent wavelength measurement data at the transmission end;

[0167] During the pause, identifying a set of wavelength indices that have not yet reached the alignment length, and performing a local rescan or interpolation calculation based on the set to generate missing measurement data;

[0168] Fill the measurement data obtained by repeated scanning or interpolation calculation into the corresponding position of the storage array according to the established wavelength index and batch identification;

[0169] Update the wavelength index-dynamic offset mapping table within the batch;

[0170] After confirming that the data volume of the current batch has reached the alignment length, parallel transmission is resumed, and multi-wavelength scanning and data packaging are continued based on the updated mapping table.

[0171] In high-throughput, multi-wavelength measurement scenarios, to ensure data integrity, the system typically requires a set "alignment length." If the data volume within a batch does not reach this alignment length (for example, due to wavelength data delays, network packet loss, or measurement anomalies), continuing to send subsequent batches of data can easily lead to index confusion or measurement omissions, making accurate distributed loss analysis difficult later.

[0172] To this end, this application proposes a method based on "locking the current batch index status, pausing transmission and performing point replenishment operations" to ensure that the parallel transmission process of the next batch is entered after the data is restored or replenished.

[0173] In specific implementations, when it is detected that the measurement data of the current batch has not yet met the alignment length, a "pause transmission" command is immediately sent to the transmission end; at this time, the index allocation table of the current batch is temporarily frozen, prohibiting the writing of new wavelengths or new dynamic offsets to prevent index confusion during the point filling operation;

[0174] For example, if the current batch is expected to collect 100 wavelength data, but only 90 have been collected, then after the lock operation begins, no new wavelengths (101st, 102nd, etc.) will be allowed to enter this batch.

[0175] Furthermore, the existing wavelength index set is read to find a list of wavelengths that have not yet reached the alignment length (or missing measurement data entries). For missing data points, if it is possible to obtain them through a short "local repeated scan" (for example, only a small number of wavelengths or key wavelengths are missing), the system will send a signal to the light source end or the interferometer detection end to perform local emission and acquisition again for these wavelengths. If the missing amount is small and detection resources are limited, interpolation calculations can be performed based on the acquired adjacent wavelength information or distributed phase information at adjacent times to generate the corresponding measurement data.

[0176] For example, when a wavelength has normal measurement results at both the previous moment and the next moment, the intermediate missing points can be estimated by interpolation.

[0177] When missing data is obtained through local repeated scanning or interpolation calculations, it is filled back into the corresponding address of the storage array according to the defined position of its "wavelength index + batch identifier". In this way, the current batch of data in the storage array is completed and realigned with other wavelength data at the timing and batch levels, ensuring the integrity of subsequent loss calculations or phase demodulation.

[0178] After the point filling is completed, if new dynamic offsets or index allocation changes are generated during the local rescan, these changes need to be written to the mapping table. For example, if a wavelength in the batch temporarily uses an additional dynamic offset during the rescan, this update will still be effective for subsequent batches to avoid index overlap or loss. The mapping table can be kept up to date by merging records or inserting operations to ensure that the filled and subsequently added data can be correctly distinguished when transmission resumes.

[0179] When it is confirmed that the data volume of the current batch has reached the pre-aligned length (for example, all 100 wavelength data have been written), a "resume transmission" signal is issued; the transmission end immediately releases the pause state and packages and sends the new wavelength or next batch of data in parallel; during the parallel scanning process of subsequent batches, the updated index-offset mapping table and corresponding point filling strategy are still used to maintain efficient data management.

[0180] By locking the index status, identifying missing data and interpolating or rescanning, updating the index mapping table, and finally resuming transmission, the system is able to balance data integrity and parallel efficiency in a high-speed multi-wavelength measurement environment, and respond efficiently and accurately to situations where batch data is insufficient.

[0181] Optionally, generating missing measurement data includes:

[0182] Counting the distribution of missing measurement data in the wavelength index set, and determining whether it exceeds a preset local scan threshold;

[0183] In response to the distribution of the missing measurement data being greater than the local scanning threshold, activating a local re-scanning process to re-send measurement light pulses or interference signal samples only to the optical fiber section corresponding to the missing wavelength index to obtain corresponding power values ​​and / or interference signals;

[0184] In response to the distribution of the missing measurement data being less than or equal to the local scanning threshold, or the scanning resources entering a preset tight state, performing interpolation calculation based on distributed phase information of adjacent sampling points or adjacent wavelengths to generate the missing measurement data;

[0185] The missing measurement data obtained by local repeated scanning or interpolation calculation are filled into the corresponding position of the storage array according to the established wavelength index and batch identification.

[0186] In real-time multi-wavelength scanning scenarios, some wavelength data may be missing due to network latency, fiber disturbances, or unstable interference signals. If the missing data is too large and interpolation is performed, it can cause significant errors. However, frequently triggering full, repeated scans consumes significant measurement resources and prolongs the measurement cycle.

[0187] To this end, this application introduces the "local scan threshold" and "resource shortage" judgment logic to distinguish the processing methods of large-scale missing and small-scale missing, and try to use interpolation when resources are limited to balance measurement accuracy and overall efficiency.

[0188] In practice, during pauses or routine monitoring, the system scans the existing wavelength index set and counts the number and distribution of missing measurement data. For example, are the missing data concentrated in a specific wavelength segment or scattered across various wavelength channels? Based on a preset "local scan threshold," the system determines whether the missing data distribution exceeds this threshold. This threshold can be set by the operator based on actual measurement scenarios and is typically closely related to factors such as the allowable interpolation error and scanning resource utilization.

[0189] For example, if the total number or concentration of missing data is greater than or equal to 20% (threshold), it is judged as "large-scale missing", which may easily lead to inaccurate interpolation; if it is lower than this ratio, it can be regarded as "small-scale" or sporadic missing, which is suitable for quick filling by interpolation.

[0190] When the missing distribution is greater than or equal to the threshold, the system activates local rescanning:

[0191] The system sends a command to the corresponding light source module to re-emit the detection light pulse only to the optical fiber section where the missing wavelength index is located;

[0192] The receiving end collects the echo signal again to obtain a new power value or interference signal;

[0193] This eliminates the need to perform a complete scan of the entire wavelength range, reducing the impact on measurement resources and timing.

[0194] For example, if it is found that wavelength index 1015 is missing in a large area, only wavelength 1015 is scanned again.

[0195] In a specific implementation, if the missing distribution is less than or equal to the local scanning threshold, or the system detects that the scanning resources have entered a "tight state" (e.g., the laser is performing other priority tasks), the system performs interpolation based on the distributed phase information of adjacent sampling points or adjacent wavelengths;

[0196] Among them, for wavelengths with missing power values, linear or polynomial interpolation can be performed using the power values ​​at adjacent moments; for wavelengths with missing interference signals, phase interpolation or approximate compensation can be performed based on the phase distribution of adjacent wavelengths, local extreme value characteristics, etc.

[0197] In this way, it avoids occupying valuable hardware channels and extra scanning time, while ensuring data continuity without reducing overall real-time performance.

[0198] For example, if only 2 to 3 points of wavelength index 12 are missing, and the data signals of adjacent wavelengths 11 and 13 are intact, an interpolation algorithm is used to generate approximate values ​​for those missing points of index 12.

[0199] Regardless of whether local rescanning or interpolation algorithms are used, the missing measurement data finally obtained must be written back to the storage array according to the location of its wavelength index and batch identifier;

[0200] This operation can be performed by the system calling a mapping function (index-address) to ensure that the supplemented data is aligned with the existing measurement results in terms of physical address, facilitating subsequent loss calculations;

[0201] If necessary, the supplementation method (scanning or interpolation) can also be recorded for subsequent traceability or quality assessment.

[0202] In this way, this application sets the local scanning threshold and resource shortage judgment mechanism, and flexibly selects "local repeated scanning" or "interpolation calculation" in the point filling link, which can not only improve data quality and avoid the serious impact of large-area missing, but also avoid the waste of time and resources caused by frequent scanning, and ultimately achieve more reliable and efficient multi-wavelength distributed measurement.

[0203] Optionally, in response to the completion of the patching operation, resuming transmission, and performing optical fiber loss calculation on the multi-wavelength data written in the storage array to generate distributed loss information of the tested optical fiber includes:

[0204] Reading and registering the multi-wavelength data, and aligning the batch identifiers and timestamps;

[0205] performing a ratio operation and / or phase demodulation on the power values ​​and / or interference signals included in the multi-wavelength data;

[0206] Based on the ratio calculation result and / or phase distribution, it is mapped to the axial position of the optical fiber to generate distributed loss information.

[0207] After completing local rescanning or interpolation calculations, the point-filling operation can bring the data within the current batch to the preset alignment length. However, for subsequent fiber loss analysis, complete multi-wavelength data alone is not enough; in multi-wavelength distributed measurements, data from different batches, different timestamps, and different wavelengths need to be aligned to ensure that they correspond to the same fiber position or the same measurement window in the time or spatial domain. Without correct ratio calculations or phase demodulation, accurate loss information cannot be obtained, resulting in inaccurate judgment of the fault point or attenuation distribution.

[0208] To solve this problem, the present application proposes to immediately perform multi-wavelength data reading and alignment, as well as subsequent loss calculation processes after the filling point is completed, so as to quickly and accurately obtain the distributed loss information of the optical fiber.

[0209] In practice, after the point-filling operation is completed and a "resume transmission" signal is sent to the system, the current batch of multi-wavelength data is read from the storage array. Since the data for each wavelength has previously been aligned at the batch ID and timestamp level (the point-filling operation may involve interpolation or local repeated scanning), a secondary registration is performed using the batch ID and timestamp to eliminate residual deviations caused by data disorder or delay.

[0210] For example, the wavelength data collected at different times are matched according to the latest or same reference time to form a multi-wavelength data group under "the same spatial position and the same sampling window".

[0211] Furthermore, for measurement data containing power values, the system typically uses ratio calculations to evaluate insertion loss, backscatter loss, etc.; that is, a reference wavelength or reference power is compared with the current wavelength to obtain the relative attenuation. For measurement data containing interference signals, a phase demodulation or phase unwrapping algorithm is required:

[0212] First, the interference signal is demodulated to extract the phase variation with fiber length or time;

[0213] If the signal spans multiple 2π periods, phase unwrapping technology can be used to ensure the continuity of the phase curve;

[0214] At this point, the phase distribution of each wavelength at different spatial positions can be obtained.

[0215] Specifically, for scattering or coherent measurements, ratio operation or phase demodulation can be selected according to actual needs, or they can be performed simultaneously to compare different results and improve measurement robustness.

[0216] In specific implementation, after obtaining the attenuation or phase distribution of each wavelength, it is necessary to perform spatial coordinate mapping in combination with the optical fiber length calibration or echo delay information.

[0217] For example, if time domain reflectometry is used (OTDR-like principle), the waveform sampling points can be mapped to the actual length positions of the fiber;

[0218] For another example, if the frequency domain or phase modulation principle is used, the spatial distribution can also be obtained through time delay-distance conversion or frequency domain-position mapping.

[0219] Ultimately, the system obtains the corresponding loss value (or phase value) at each spatial sampling point, thereby forming distributed loss information. If required, this information can be visualized as a "distance-loss" curve, showing the attenuation inside the fiber in real time and enabling rapid detection of potential anomalies or faults.

[0220] In this way, in multi-wavelength distributed measurement, the present application successfully converts the multi-wavelength data after filling in points into clear and intuitive distributed loss information by using ratio calculation or phase demodulation plus spatial coordinate mapping, providing a high-efficiency and high-precision solution for the status evaluation and fault location of optical fiber lines.

[0221] Based on the same inventive concept, the embodiments of the present disclosure also provide an optical fiber bending loss test system based on spectral analysis corresponding to the optical fiber bending loss test method based on spectral analysis. Since the principle of solving the problem by the system in the embodiments of the present disclosure is similar to the above-mentioned optical fiber bending loss test method based on spectral analysis in the embodiments of the present disclosure, the implementation of the system can refer to the implementation of the method, and the repeated parts will not be repeated.

[0222] Reference Figure 3 FIG. 1 is a schematic diagram of an optical fiber bending loss test system based on spectral analysis provided in an embodiment of the present application, wherein the system includes:

[0223] The acquisition unit 10 is used to perform multi-wavelength scanning on the optical fiber under test to obtain measurement data at each wavelength; wherein the measurement data includes: power value and interference signal;

[0224] The transmission unit 20 is configured to package the measurement data according to a predefined wavelength index rule and transmit at least two measurement data in parallel within one transmission cycle;

[0225] A mapping unit 30 is configured to receive the parallel transmitted measurement data and write each measurement data into a corresponding address of the storage array by mapping the index to the storage address according to the wavelength index and batch identifier carried by each measurement data;

[0226] The first processing unit 40 is configured to trigger a transmission suspension and perform a point filling operation in response to detecting that the number of measurement data in the current batch does not meet a preset alignment length;

[0227] The second processing unit 50 is configured to resume transmission in response to completion of the patching operation, and perform fiber loss calculation on the multi-wavelength data written in the storage array to generate distributed loss information of the tested optical fiber.

[0228] Those skilled in the art will understand that in the above-described methods of specific embodiments, the order in which the steps are presented does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible inherent logic. It should be understood that determining B based on A does not mean determining B based solely on A; B can also be determined based on A and / or other information.

[0229] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0230] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Claims

1. A method for testing optical fiber bending loss based on spectral analysis, characterized in that: include: Perform multi-wavelength scanning on the optical fiber under test to obtain measurement data at each wavelength; wherein the measurement data includes: power value and interference signal; Packing the measurement data according to a predefined wavelength index rule, and transmitting at least two measurement data in parallel within one transmission cycle; Receive the parallel transmitted measurement data, and write each measurement data into a corresponding address of the storage array by mapping the index to the storage address according to the wavelength index and batch identifier carried by each measurement data; In response to detecting that the number of measurement data in the current batch does not meet the preset alignment length, triggering a transmission pause and performing a point filling operation; The triggering of pausing transmission and performing the filling operation includes: locking the wavelength index allocation status of the current batch and pausing the transmission of subsequent wavelength measurement data at the transmission end; during the pause period, identifying a set of wavelength indexes that have not yet reached the alignment length, and performing local repeated scanning or interpolation calculation based on the set to generate missing measurement data; and filling the measurement data obtained by the repeated scanning or interpolation calculation into corresponding positions in the storage array according to the established wavelength index and batch identifier; In response to the completion of the point-filling operation, transmission is resumed, and optical fiber loss calculation is performed on the multi-wavelength data written in the storage array to generate distributed loss information of the tested optical fiber.

2. The method according to claim 1, characterized in that Packing the measurement data according to a predefined wavelength index rule and transmitting at least two measurement data in parallel within one transmission cycle includes: Adding header information to the measurement data of each wavelength based on the wavelength index rule and the scanning order, wherein the header information includes a wavelength index, a batch identifier, a serial number, and a timestamp; Sending the measurement data of each wavelength in parallel based on the packet header information; The header information of the parallel transmission is parsed at the receiving end, the corresponding measurement data is mapped to the target address in the storage array according to the wavelength index and batch identifier, and the arrival order is rearranged and / or supplemented according to the sequence number and timestamp.

3. The method according to claim 2, characterized in that The wavelength indexing rules include: Before the scan begins, a list of wavelengths arranged in descending order is established based on the working wavelength range of the optical fiber under test, and a basic index corresponding to its order is assigned to each wavelength; In each scanning cycle of a transmission batch, a dynamic offset related to the batch identifier is applied to the basic index to generate a wavelength index actually used; Detecting whether the actually used wavelength index conflicts with a previously allocated wavelength index; In response to detecting a conflict, adaptively adjusting the dynamic offset until there is no conflict and then packaging the corresponding wavelengths; Record the correspondence between the established wavelength index and wavelength and batch identifier, and keep it consistent during data packaging and parallel transmission.

4. The method according to claim 3, characterized in that In response to detecting a conflict, adaptively adjusting the dynamic offset until there is no conflict and then packaging the corresponding wavelengths includes: Read the allocated wavelength index set in the current batch from the conflict detection record; Incrementing the dynamic offset, and recalculating the actually used wavelength index based on the updated dynamic offset; Determining whether the recalculated wavelength index still conflicts with an existing wavelength index in the wavelength index set; In response to there still being a conflict, the increment of the dynamic offset and the conflict detection are repeatedly performed until there is no conflict; The finally determined actually used wavelength index is written into the wavelength index set, and the measurement data of the corresponding wavelength is continuously packaged.

5. The method according to claim 4, characterized in that Incrementing the dynamic offset includes: When performing multi-wavelength scanning, local extreme value moments or several phase trigger points are extracted from the real-time interference signal and combined with the current count value of the system clock to generate a pseudo-random seeding value; The pseudo-random seed value is used for initial and iterative random increments of the dynamic offset.

6. The method according to claim 5, characterized in that Incrementing the dynamic offset further includes: When a batch filling operation is detected to be triggered and / or completed, the pseudo-random seeding value is recalculated based on the latest local extreme value moment and the system clock; After the patching is completed and the transmission is resumed, the updated dynamic offset is synchronized with the allocated index and batch identification mapping table.

7. The method according to claim 6, characterized in that The triggering of suspending transmission and performing the filling operation further includes: Update the wavelength index-dynamic offset mapping table within the batch; After confirming that the data volume of the current batch has reached the alignment length, parallel transmission is resumed, and multi-wavelength scanning and data packaging are continued based on the updated mapping table.

8. The method according to claim 7, characterized in that Generating missing measurement data includes: Counting the distribution of missing measurement data in the wavelength index set, and determining whether it exceeds a preset local scan threshold; In response to the distribution of the missing measurement data being greater than the local scanning threshold, activating a local re-scanning process to re-send measurement light pulses or interference signal samples only to the optical fiber section corresponding to the missing wavelength index to obtain corresponding power values ​​and / or interference signals; In response to the distribution of the missing measurement data being less than or equal to the local scanning threshold, or the scanning resources entering a preset tight state, performing interpolation calculation based on distributed phase information of adjacent sampling points or adjacent wavelengths to generate the missing measurement data; The missing measurement data obtained by local repeated scanning or interpolation calculation are filled into the corresponding position of the storage array according to the established wavelength index and batch identification.

9. The method according to claim 8, characterized in that In response to the completion of the patching operation, the transmission is resumed, and the optical fiber loss calculation is performed on the multi-wavelength data written in the storage array to generate the distributed loss information of the tested optical fiber, including: Reading and registering the multi-wavelength data, and aligning the batch identifiers and timestamps; performing a ratio operation and / or phase demodulation on the power values ​​and / or interference signals included in the multi-wavelength data; Based on the ratio calculation result and / or phase distribution, it is mapped to the axial position of the optical fiber to generate distributed loss information.

10. The optical fiber bending loss test system based on spectrum analysis is characterized by: include: An acquisition unit is used to perform multi-wavelength scanning on the optical fiber under test and obtain measurement data at each wavelength; wherein the measurement data includes: power value and interference signal; A transmission unit, configured to package the measurement data according to a predefined wavelength index rule, and transmit at least two measurement data in parallel within one transmission cycle; a mapping unit, configured to receive the parallel transmitted measurement data and write each measurement data into a corresponding address of the storage array by mapping the index to the storage address according to the wavelength index and batch identifier carried by each measurement data; The first processing unit is configured to trigger a transmission pause and perform a point filling operation in response to detecting that the number of measurement data in a current batch does not meet a preset alignment length; the triggering of the transmission pause and the point filling operation includes: locking the wavelength index allocation state of the current batch and pausing the transmission of subsequent wavelength measurement data at the transmission end; during the pause period, identifying a set of wavelength indexes that have not yet reached the alignment length, and performing a local rescan or interpolation calculation based on the set to generate missing measurement data; and filling the measurement data obtained by the rescan or interpolation calculation into corresponding positions in the storage array according to the established wavelength index and batch identifier; The second processing unit is configured to resume transmission in response to completion of the patching operation, and perform optical fiber loss calculation on the multi-wavelength data written in the storage array to generate distributed loss information of the tested optical fiber.

Images