Earthquake data real-time monitoring and backhaul method and system based on Beidou short message

Abstract

The application relates to the technical field of earthquake monitoring and discloses a method and system for real-time monitoring and back transmission of earthquake data based on a Beidou short message. A sending end organizes continuously collected seismic waveforms in a minute-level time window, performs down-sampling on seismic waveforms of a target channel, and adopts adaptive compression facing fixed load constraints to compress the seismic waveforms, so that conventional real-time monitoring data are adapted to the upper limit of a single effective payload of a Beidou short message. After the conventional real-time monitoring data are encapsulated into messages through a general communication layer, the messages are sent to a receiving end. The receiving end analyzes the messages, performs sorting, splicing and loss detection, and generates a retransmission or high-precision back transmission request. The sending end responds to the request, extracts corresponding data from a cache as retransmission supplementary delivery or back transmission data, and sends the data according to priority scheduling. The application can realize real-time data back transmission and subsequent fine analysis support of a remote earthquake monitoring station under the condition of low power consumption and long-term unattended operation.

Description

Technical Field

[0001] This invention relates to the field of earthquake monitoring technology, specifically to a method and system for real-time monitoring and transmission of earthquake data based on BeiDou short message service. Background Technology

[0002] Earthquake monitoring is a crucial foundation for earthquake early warning, seismic situation assessment, geodynamic process research, and disaster emergency response. Current earthquake monitoring systems have evolved from early mechanical and analog recording methods to comprehensive monitoring systems characterized by digitalization, networking, and real-time operation. Broadband digital seismographs, strong-motion meters, MEMS sensors, and distributed fiber optic sensing technologies are increasingly being applied in monitoring scenarios. Simultaneously, seismic networks are continuously extending into mountainous areas, islands, border regions, deserts, and other areas with weak infrastructure, making data transmission and long-term maintenance at remote stations increasingly critical issues. Existing methods mainly include local storage followed by manual retrieval, real-time transmission via public networks such as 4G / 5G / Wi-Fi, fixed link transmission via dedicated lines or local area networks, and low-power wireless transmission methods such as data radios and LoRa. However, in remote areas, these methods often suffer from insufficient real-time performance, limited coverage, unstable links, high construction and maintenance costs, or difficulty in long-term unattended operation.

[0003] Against this backdrop, satellite communication has become an important alternative link for remote monitoring scenarios. In addition to providing positioning, navigation, and timing services, the BeiDou system also possesses short message communication capabilities, characterized by wide coverage, independence from terrestrial public networks, and bidirectional interaction. Existing research indicates that the BeiDou-3 global short message communication service already has the systemic foundation to support services such as positioning reports, emergency search and rescue, and routine short messages. In the fields of disaster monitoring and telemetry feedback, BeiDou short messages have also demonstrated good engineering application potential; for example, existing research has proposed a GNSS observation data encoding and transmission scheme based on BeiDou short messages for landslide monitoring scenarios. However, BeiDou short messages themselves still have constraints such as small single message payload, limited transmission cycle time, fluctuating link latency, and the possibility of packet loss and out-of-order delivery. Therefore, this type of link is more suitable for small data services such as status variables, alarm quantities, and positioning information, while existing technologies are not yet directly applicable to data objects with high sampling rates, multiple channels, and strong time series, such as continuous seismic waveforms.

[0004] On the other hand, seismic data compression technology has accumulated considerable research experience over a long period. Classical studies have shown that seismic or geophysical data compression typically employs a combination of transform, quantization, and coding techniques to reduce data redundancy and improve compression efficiency. In subsequent research, wavelet transform, due to its excellent multi-resolution representation capabilities, has been widely used in seismic data compression and is often combined with quantization, entropy coding, and other methods. In recent years, related research has further focused on optimizing the balance between compression ratio and reconstruction accuracy under given error constraints. However, most existing compression methods are geared towards storage optimization, general link transmission, or error-constrained compression, and are not specifically designed for the continuous waveform backhaul requirements under fixed payload and fixed transmission cycle conditions, such as BeiDou short messages. For this type of scenario, simply increasing the average compression ratio is insufficient; it is also necessary to ensure that the output of each processing window can stably adapt to the short message payload budget, while simultaneously considering waveform reconstruction quality, protocol overhead control, and subsequent transmission organization requirements. Furthermore, since these compression algorithms are typically deployed on low-power edge devices or embedded transmitters at remote stations, the computing power, storage resources, and energy supply of these devices are relatively limited, thus significantly constraining algorithm complexity, storage overhead, and runtime latency. While some compression methods can achieve better compression results in high-performance computing environments, they have heavy computational burdens, large parameter scales, or high implementation complexity, making it difficult to directly meet the engineering application requirements under low power consumption and long-term unattended operation conditions.

[0005] Besides compression, data organization and reliable transmission under constrained links are also critical issues. Existing research on field monitoring and satellite telemetry systems shows that under low-bandwidth and unstable link conditions, it is usually necessary to design supporting mechanisms such as data caching, packet encapsulation, missing data detection, retransmission recovery, and power management; otherwise, it is difficult to form a truly usable end-to-end closed loop. This problem is even more pronounced for BeiDou short message links. Although existing research has verified the engineering feasibility of BeiDou short message in various monitoring services, for continuous seismic waveforms, there is still a lack of an integrated technical solution that can simultaneously handle minute-level real-time monitoring, cross-message organization, missing data detection, retransmission recovery, and high-precision on-demand backhaul.

[0006] In summary, current related technologies have made significant progress in earthquake monitoring, satellite short message communication, disaster telemetry, and data compression. However, they have not yet effectively solved the problems of real-time representation, reliable transmission, and on-demand high-precision retransmission of continuous earthquake waveforms in remote areas under the limited link conditions of BeiDou short message communication. Therefore, it is still necessary to propose a method for real-time monitoring and high-precision backhaul of earthquake data based on BeiDou short message communication to meet the engineering application needs of earthquake stations in areas with no public network, weak public network, or unstable public network under long-term, low-power, and unattended operation conditions. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention provides a method and system for real-time monitoring and transmission of seismic data based on BeiDou short message service. This system solves the problem of seismic stations in areas without public networks, with weak public networks, or with unstable public networks struggling to achieve stable remote transmission of continuous waveform data. It enables continuously acquired seismic data to achieve efficient compression, reliable transmission, missing data recovery, and high-precision on-demand retransmission for low-power edge devices, even under conditions where the single-transmission payload of BeiDou short message service is limited, the transmission cycle is fixed, and the link may experience delay fluctuations, packet loss, and out-of-order delivery. This balances real-time monitoring capabilities, waveform recoverability, and the need for subsequent detailed analysis.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for real-time monitoring and transmission of earthquake data based on BeiDou short message service, comprising: The transmitting end organizes continuously acquired seismic waveforms in minute-level time windows, performs downsampling on the seismic waveforms of the target channel, and uses adaptive compression oriented to fixed load constraints to compress the conventional real-time monitoring data to match the upper limit of the single effective payload of Beidou short messages; after encapsulating the conventional real-time monitoring data into messages through the general communication layer, it is sent to the receiving end. The receiving end parses the received messages, sorts, reassembles, and detects missing messages based on the message sequence number, block location field, and block length field. When a missing message is detected, a retransmission request is generated. When detailed analysis of a specific time period is required, a return transmission request is generated. The sending end responds to the retransmission request by retrieving the corresponding message from the sent message buffer as retransmission data; the sending end responds to the return request by retrieving the seismic waveforms of the specified time period and channel from the local raw waveform buffer and organizing them into return data; after encapsulating the retransmission data and return data into messages through the general communication layer, they are scheduled to be sent to the receiving end with a higher priority than regular real-time monitoring data.

[0009] In one embodiment, the compression using adaptive compression oriented towards fixed load constraints specifically includes: The downsampled seismic waveform is subjected to a three-level wavelet decomposition, and multiple sub-band coefficients are obtained using the coif5 wavelet basis. Each sub-band coefficient is clustered and quantized, and the resulting symbol sequence is sequentially subjected to run-length coding and Huffman coding. Based on the comparison between the code length of the compressed code stream obtained by encoding and the upper limit of the target payload of Beidou short message, the number of clusters of at least one sub-band coefficient is dynamically adjusted, and the compressed code stream is output under the constraint that the output code length does not exceed the upper limit of the target payload, with the goal of minimizing the normalized mean square error of the compressed seismic waveform relative to the downsampled seismic waveform.

[0010] In one embodiment, the general communication layer abstracts the data to be transmitted into data blocks and a single BeiDou short message into a message. Each message includes a message-level header and a payload area. The message-level header includes a message sequence number and a block location field. A block length field is set at the beginning of each new data block in the payload area. If a data block can be completely placed into the current message, it is written in its entirety; otherwise, only the portion that can be accommodated is written. The remaining portion of the data block is continued in subsequent messages, and the continued portion is not written to the block length field again.

[0011] In one embodiment, the stitching restoration specifically includes: When a message contains both the continuation of the previous data block and the start of a new data block, the block position field P identifies the starting offset of the block length field of the first complete data block in the current message. The receiving end determines the length of the continuation part and the starting position of the new data block based on P, and identifies the data block boundary in combination with the block length field L to complete the cross-message splicing and recovery.

[0012] In one embodiment, the priorities from high to low are: backhaul data, retransmission and resend data, and regular real-time monitoring data.

[0013] In one embodiment, the sending end responds to the transmission request by extracting seismic waveforms of a specified time period and channel from the local raw waveform buffer and organizing them into transmission data, specifically including: In response to the return request, the sending end extracts 1 second of the original seismic waveform from a single channel each time, downsamples the frequency of the original seismic waveform from 250Hz to 100Hz, and converts the data format of the sampling points from float32 to float16, so that the main data size of a single time slice of the returned data is 200 bytes. The returned data uses a 4-byte Unix timestamp for absolute time identification, a channel identifier field to identify the target channel corresponding to the returned data, and a 1-byte reserved value is set in the original minute offset field position as a return type identifier. The receiving end identifies the transmitted data based on the reserved value, recovers the absolute time based on the Unix timestamp, determines the target channel based on the channel identifier, and completes the seismic waveform stitching according to the Unix timestamp order.

[0014] Secondly, the present invention provides a real-time monitoring and transmission system for earthquake data based on BeiDou short message service, used to implement the method of any embodiment of the first aspect, including: The acquisition end is used for continuous acquisition of seismic waveforms; At the transmitting end, continuously acquired seismic waveforms are organized in minute-level time windows. The seismic waveforms of the target channel are downsampled and compressed using adaptive compression oriented towards fixed load constraints, ensuring that the conventional real-time monitoring data matches the upper limit of the single effective payload of the BeiDou short message service. The conventional real-time monitoring data is encapsulated into messages through the general communication layer and sent to the receiving end. In response to retransmission requests, the corresponding message is retrieved from the sent message buffer as retransmission data. In response to feedback requests, the transmitting end retrieves seismic waveforms of a specified time period and channel from its local raw waveform buffer and organizes them into feedback data. The retransmission data and feedback data are encapsulated into messages through the general communication layer and then scheduled and sent to the receiving end with a higher priority than the conventional real-time monitoring data. At the receiving end, the received messages are parsed, and sorted, reassembled, and missing messages are detected based on the message sequence number, block location field, and block length field. When a missing message is detected, a retransmission request is generated, and when detailed analysis of a specific time period is required, a return request is generated.

[0015] In one embodiment, the compression using adaptive compression oriented towards fixed load constraints specifically includes: The downsampled seismic waveform is subjected to a three-level wavelet decomposition, and multiple sub-band coefficients are obtained using the coif5 wavelet basis. Each sub-band coefficient is clustered and quantized, and the resulting symbol sequence is sequentially subjected to run-length coding and Huffman coding. Based on the comparison between the code length of the compressed code stream obtained by encoding and the upper limit of the target payload of Beidou short message, the number of clusters of at least one sub-band coefficient is dynamically adjusted, and the compressed code stream is output under the constraint that the output code length does not exceed the upper limit of the target payload, with the goal of minimizing the normalized mean square error of the compressed seismic waveform relative to the downsampled seismic waveform.

[0016] In one embodiment, the general communication layer abstracts the data to be transmitted into data blocks and a single BeiDou short message into a message. Each message includes a message-level header and a payload area. The message-level header includes a message sequence number and a block location field. A block length field is set at the beginning of each new data block in the payload area. If a data block can be completely placed into the current message, it is written in its entirety; otherwise, only the portion that can be accommodated is written. The remaining portion of the data block is continued in subsequent messages, and the continued portion is not written to the block length field again.

[0017] In one embodiment, the stitching restoration specifically includes: When a message contains both the continuation of the previous data block and the start of a new data block, the block position field P identifies the starting offset of the block length field of the first complete data block in the current message. The receiving end determines the length of the continuation part and the starting position of the new data block based on P, and identifies the data block boundary in combination with the block length field L to complete the cross-message splicing and recovery.

[0018] The system and method in this invention correspond to each other; the specific technical solutions applicable to the method are also applicable to the system.

[0019] Compared with the prior art, the beneficial technical effects of the present invention are: This invention addresses the problem of unstable backhaul of continuous seismic waveforms in areas with no, weak, or unstable public networks by proposing a real-time monitoring and high-precision backhaul method for seismic data based on BeiDou short message service. Compared with existing technologies, this invention incorporates the link constraints of approximately 229 bytes of effective payload and a minimum transmission interval of approximately 60 seconds for civilian BeiDou short message service into the data processing flow. This allows minute-level window construction, downsampling, compression encoding, and message encapsulation to be designed around fixed payload and fixed transmission intervals, thereby improving the transmittability and real-time performance of continuous seismic waveforms in constrained satellite short message links.

[0020] This invention transforms original high-sampling-rate, multi-channel seismic waveforms into a data representation suitable for short message payloads through minute-level window organization, target channel selection, downsampling processing with anti-aliasing filtering, and adaptive compression oriented towards fixed load constraints. Taking a 60-second original waveform in 250Hz, 3-channel, float32 format as an example, the data size of a single original window is approximately 180,000 bytes; even after downsampling to 10Hz and single-channel data, it is still approximately 2,400 bytes, significantly exceeding the upper limit of short message payloads. Therefore, this invention further utilizes wavelet decomposition, cluster quantization, run-length encoding, Huffman coding, and output code length feedback adjustment to ensure that the compression result satisfies fixed load constraints while preserving as many main waveform features as possible. Figure 13 The compression performance results obtained from testing 1000 data points selected from the STEAD dataset are presented to illustrate that the present invention can still achieve good compression results under conditions of small waveform reconstruction error.

[0021] This invention also achieves unified encapsulation, priority scheduling, cross-message continuation, missing data detection, and retransmission recovery of conventional real-time monitoring data, retransmitted data, and high-precision backhaul data through a general communication layer, improving data integrity and reliability under constrained link conditions. Conventional real-time monitoring data uses a 1-byte minute offset instead of a 4-byte Unix timestamp, reducing time stamp overhead by 75%. High-precision backhaul data uses a 1-second, 100Hz, single-channel, float16 organization method, making the main data volume of a single time slice approximately 200 bytes, which can well adapt to a single 229-byte payload. Therefore, this invention, while ensuring minute-level real-time monitoring capabilities, also addresses the need for high-precision waveform retransmission during critical periods, demonstrating good engineering feasibility and long-term unattended application value. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the overall system flow.

[0023] Figure 2 This is a flowchart of adaptive compression for fixed load constraints.

[0024] Figure 3 This is the flowchart for adaptive iterative control.

[0025] Figure 4 This is a comparison of the time-domain waveforms of earthquake event samples before and after compression.

[0026] Figure 5 This is a comparison chart of the spectral characteristics of earthquake event samples before and after compression.

[0027] Figure 6 This is a comparison of the time-domain waveforms of the noise sample before and after compression.

[0028] Figure 7 This is a comparison of the spectral characteristics of the noise sample before and after compression.

[0029] Figure 8 This is a diagram of the overall architecture of the communication protocol for the BeiDou short message link.

[0030] Figure 9 This is a schematic diagram of a multi-data-block splicing and encapsulation method.

[0031] Figure 10 This is a timing diagram for the message missing detection and retransmission recovery mechanism.

[0032] Figure 11 This is a schematic diagram of the receiving station status monitoring page.

[0033] Figure 12 This is a schematic diagram of the receiving end waveform viewing and SAC format file export page.

[0034] Figure 13 This is a graph showing the results of the compression performance test.

[0035] Figure 14 This is an overview of the real-time waveform comparison between a teleseismic event in a certain sea area and the BeiDou station on the selected island in the example, as well as the 4G reference station at the same location.

[0036] Figure 15 This is a partial comparison of the waveforms of BeiDou real-time transmission, high-precision transmission, and real-time transmission from a 4G station at the same location for a teleseismic event in a certain sea area under single-station conditions.

[0037] Figure 16 This is a flowchart of the method of the present invention. Detailed Implementation

[0038] A preferred embodiment of the present invention will now be described in detail with reference to the accompanying drawings.

[0039] like Figure 16 As shown, a real-time monitoring and transmission method for earthquake data based on BeiDou short message service in this invention includes: S1, the transmitting end organizes continuously acquired seismic waveforms in minute-level time windows, performs downsampling on the seismic waveforms of the target channel, and uses adaptive compression oriented to fixed load constraints to compress the conventional real-time monitoring data to match the upper limit of the single effective payload of Beidou short message; after encapsulating the conventional real-time monitoring data into messages through the general communication layer, it is sent to the receiving end; S2, the receiving end parses the received message, sorts, reassembles and restores it according to the message sequence number, block position and block length fields, and performs missing detection. When a message is missing, a retransmission request is generated. When fine analysis of a specific time period is required, a return transmission request is generated. S3, the sending end responds to the retransmission request and extracts the corresponding message from the sent message buffer as retransmission data; the sending end responds to the return request and extracts the seismic waveforms of the specified time period and channel from the local original waveform buffer and organizes them into return data; after encapsulating the retransmission data and return data into messages through the general communication layer, they are scheduled to be sent to the receiving end with a higher priority than regular real-time monitoring data.

[0040] This invention proposes a real-time monitoring and high-precision backhaul method for seismic data based on BeiDou short message service. It is designed for seismic monitoring stations in areas with no, weak, or unstable public networks. Addressing constraints such as limited payload per BeiDou short message link, fixed transmission interval, potential latency fluctuations, message loss, and out-of-order delivery, a seismic data processing and transmission scheme is constructed, collaboratively completed by the sender and receiver. Under the civilian BeiDou short message application conditions addressed in this invention, the capabilities of a Level 2 civilian BeiDou short message communication card are preferably used as a design constraint. The maximum payload per short message is designed to be approximately 229 bytes, and the minimum transmission interval is designed to be approximately 60 seconds. Due to the high sampling rate, multi-channel, and strong temporal characteristics of continuous seismic waveforms, without specialized processing for payload and transmission interval, it is difficult to achieve minute-level continuous backhaul under the aforementioned short message link conditions. Therefore, this invention does not simply use BeiDou short message service to send compressed files, but rather designs an end-to-end method based on fixed payload, fixed transmission interval and reliable recovery requirements, from waveform organization, downsampling, compression coding, message encapsulation, missing data detection, retransmission recovery to high-precision on-demand backhaul.

[0041] This invention starts by receiving the original seismic waveform, and proceeds through local caching, minute-level organization, downsampling, adaptive compression, unified message encapsulation, and BeiDou short message transmission, before finally reaching the receiving end to complete parsing, splicing, restoration, storage, and display, forming a complete closed loop. Figure 1This diagram illustrates the overall flow of the present invention, from data acquisition to processing and transmission, and then to backend reception and display. It can be used to explain the overall architecture and data flow of the present invention. The acquisition end includes a vibration acquisition module, a data storage module, and a data forwarding module. The transmission end includes an RDSS communication module, a general communication layer, a data preprocessing and adaptive compression algorithm, and a frame parsing module. The reception end includes a general communication layer, a decompression module, a database module, and a visualization module.

[0042] It should be noted that the acquisition end uses an existing mature solution, which is not the main content of this invention. The focus of this invention is: under the limited conditions of 229 bytes per BeiDou short message packet and a transmission interval of about 60 seconds, to convert continuous seismic waveforms into a data expression form that can stably adapt to short message transmission under acceptable error conditions, and to still achieve data block recovery, missing data retransmission, and high-precision backhaul during critical periods through a unified communication layer when the link is unstable.

[0043] In one implementation, the transmitting end continuously receives raw seismic waveforms from the acquiring end and writes them to a local SD card or other non-volatile storage medium in chronological order. For routine real-time monitoring services, a real-time processing unit is constructed with minute-level time windows to downsample the target channel data and perform adaptive compression for fixed load constraints, ensuring that the output meets the effective payload limit of a single BeiDou short message transmission. Subsequently, the routine real-time monitoring data, retransmission data, and high-precision backhaul data are uniformly organized into messages through a general communication layer and sent via BeiDou short messages according to a priority scheduling strategy. After receiving the messages, the receiving end sorts, splices, restores, and stores them. When missing data is detected, a retransmission request is initiated, and when further analysis is needed, a high-precision backhaul request is initiated, thereby achieving real-time monitoring, reliable transmission, and high-precision on-demand retransmission of continuous seismic data under constrained satellite short message link conditions. The following is a detailed description of the technical details of the method.

[0044] 1. An adaptive compression method for fixed load constraints.

[0045] The transmitting end continuously receives raw seismic waveforms from the acquiring end and writes the raw 250Hz sampling rate, 3-channel waveform data into a local buffer in chronological order. This buffer is used to extract raw data for a specified time period and channel when a high-precision data transmission request is received. To quantitatively illustrate the difference in data scale between the raw continuous waveforms and the BeiDou short message link, the raw sampling rate is assumed to be... The time window length is The number of channels is The number of bytes occupied by a single sampling point is The amount of raw data within a time window It can be represented as: .

[0046] In a preferred embodiment, , , If a single sampling point is stored in float32 format, that is The raw data volume of the three-channel waveform per minute is then... .

[0047] Therefore, the raw minute-level three-channel waveform data volume is far greater than the 229-byte payload capacity of a single BeiDou short message, and cannot be directly transmitted in real time via the BeiDou short message link. For routine real-time monitoring services, this invention uses minute-level time windows to segment and organize continuous waveforms. Each time a new whole-minute window is formed, a real-time processing flow is triggered. This whole-minute window organization method is beneficial for establishing a clear correspondence between data blocks and minute-level time markers, and also for matching the approximately 60-second transmission cycle of BeiDou short messages.

[0048] In a preferred embodiment, in the conventional real-time monitoring mode, a target channel is selected from the current whole-minute window as the real-time transmission object, preferably the vertical component Z channel. Considering the limited effective payload of the BeiDou short message link, and the fact that this invention is deployed on low-power edge devices or embedded transmitters at remote stations, where computing power, storage resources, and energy supply are all limited, the target channel data is resampled before compression processing to reduce the data size of a single window and the burden of subsequent processing. This resampling is preferably implemented using a finite impulse response filter with anti-aliasing filtering, so as to downsample the original 250Hz sampling rate of the waveform to a target sampling rate of 10Hz while controlling the aliasing effect and maintaining the main waveform characteristics. Let the original sequence be... The resampled sequence is The resampling process can then be represented as: ; in, For the coefficients of the anti-aliasing finite impulse response filter, The length of the filter kernel. Indicates the output index The corresponding starting position in the input sequence. Through this processing, the original high-sampling-rate waveform can be transformed into a low-sampling-rate waveform that is more suitable for transmission on constrained links, while controlling the effects of aliasing and preserving the main waveform characteristics.

[0049] To further quantify the effect of downsampling on reducing data size, let the sampling rate after downsampling be... Real-time transmission channel count The number of bytes occupied by a single sampling point is The amount of data after channel selection and downsampling in a single minute window. It can be represented as: .

[0050] In a preferred embodiment, , , If a single sampling point is stored in float32 format, that is The amount of data in a single-channel minute window after downsampling is: .

[0051] Therefore, even after target channel selection and downsampling, the data volume of a single minute window is still significantly greater than the upper limit of the effective payload of a single BeiDou short message. Thus, this invention further performs adaptive compression on the downsampled waveform to meet the fixed payload constraints, ensuring that the final output consistently meets the requirements for single short message transmission.

[0052] For the downsampled 10Hz single-channel minute window waveform, this invention further employs a combination of transform domain compression and output code length constraint control. Specifically, firstly, a three-level wavelet decomposition is performed on the downsampled waveform sequence, preferably using the coif5 wavelet basis, to obtain multiple sub-band coefficients; then, cluster quantization is performed on each sub-band coefficient, preferably using K-means++ cluster initialization, with each sub-band coefficient initially using the same cluster number K=4; furthermore, run-length encoding is performed on the quantized symbol sequence first, followed by Huffman encoding, to reduce the code length overhead caused by repetitive patterns and symbol redundancy. To reduce the additional overhead caused by explicitly transmitting encoding rules during each encoding, this invention introduces a static prefix code table to save the payload space occupied by the encoding rule description; finally, the encoding results and necessary descriptive information are organized into a compressed bitstream.

[0053] In the above cluster quantization process, let the first... The first wavelet subband The coefficients are The set of cluster centers corresponding to this sub-band is Then the quantization mapping of this coefficient can be expressed as: , ; in, For the first The number of clusters used for each wavelet subband This represents the clustering quantization mapping function for this subband. In this method, each wavelet coefficient is mapped to its nearest cluster center, and the original continuous wavelet coefficient sequence is converted into a finite symbol set, facilitating subsequent run-length coding and Huffman coding. Compared to the fixed quantization step size method, this clustering quantization method can adaptively form quantization centers according to the different subband coefficient distributions, which is beneficial for preserving the main waveform features with a smaller number of symbols.

[0054] The core of this invention is not simply pursuing a higher average compression ratio, but rather ensuring that the compression result can stably adapt to the fixed effective payload constraints of BeiDou short messages. To this end, this invention introduces an adaptive control mechanism oriented towards the output code length, dynamically adjusting the number of clusters used for cluster quantization in each subband based on the output length of the current compression result. This ensures that the requirements for single short message transmission can be stably met even with varying waveform complexity across different minute windows. In one implementation, the number of clusters for each wavelet subband is set... The constructed parameter vector for: ; in, Let be the number of subbands after wavelet decomposition; let the compressed output code length be . The maximum target load per window is The adaptive compression process of the present invention can then be expressed as an optimization problem under a fixed load constraint as follows: , ; in, This represents the number of clusters in a given cluster parameter vector. The normalized mean square error of the compressed reconstructed waveform relative to the downsampled original waveform under certain conditions is used. Through this constraint optimization expression, this invention directly incorporates the fixed payload limit of BeiDou short messages into the compression parameter selection process, minimizing waveform reconstruction error while ensuring that the compressed bitstream does not exceed the upper limit that a single short message can carry.

[0055] Figure 2 The adaptive compression process for fixed load constraints is shown, fully demonstrating the closed-loop process from minute window input, downsampling, transform domain compression, code length detection, parameter adjustment, and compression result output. Figure 3The adaptive iterative control flow is further illustrated to demonstrate how the invention repeatedly adjusts compression parameters based on the current code length feedback until the target payload requirements are met. The transmitting end, based on the relationship between the current compression result and the upper limit of the target payload, adaptively adjusts the compression parameters and repeatedly executes the control logic of the compression process. In the specific iteration process, the transmitting end performs wavelet decomposition, cluster quantization, run-length encoding, and Huffman coding on the current minute window, and then calculates the current output code length. And determine whether it satisfies If the condition is not met, then adjust the clustering number of at least one wavelet subband. The quantization and encoding processes are re-executed; if satisfied, the current compression result is output. Thus, this invention uses the short message payload limit as a hard constraint in the compression process, rather than passively determining whether transmission is possible after compression is complete.

[0056] In one implementation, let the original sequence after downsampling be... The decompressed and restored sequence is The compression distortion is preferably expressed using the normalized mean square error (NMSE) as follows: ; in, The number of waveform sampling points before and after compression is given. By adaptively adjusting the compression parameters, this invention minimizes waveform reconstruction error while ensuring the output code length does not exceed the upper limit of the target payload. Compared with fixed-parameter compression methods, this method can better adapt to changes in waveform complexity within different minute windows, avoiding situations where the output code length is too long to be transmitted in some windows, and also avoiding excessive information loss due to over-compression in some windows.

[0057] In a further embodiment, the target payload upper limit The available payload, message-level fields, and necessary data block description fields of a single BeiDou short message are jointly determined. For routine real-time monitoring data, the total available length of a single BeiDou short message is preferably 229 bytes, with message-level fields and necessary data fields occupying some space. Therefore, the target upper limit for compressing the main waveform data is preferably no more than 225 bytes. Through this design, the transmitting end can stably organize the routine real-time monitoring data corresponding to a whole-minute window into a single short message transmission opportunity, so that the whole-minute window organization corresponds to the approximately 60-second transmission cycle, thereby improving the continuity and determinism of real-time monitoring data feedback.

[0058] To illustrate the applicability of the compression method of the present invention to different data samples, Figure 4 and Figure 5 The comparison results of a specific earthquake event sample in the STEAD dataset before and after compression in the time and frequency domains are presented; among them, Figure 4This is used to illustrate the preservation of the main waveform shape (original signal, decoded signal) in the time domain before and after compression. Figure 5 This is used to illustrate the consistency of the main spectral characteristics (original signal spectrum, decoded signal spectrum) before and after compression. Furthermore, Figure 6 The comparison results in the time domain of a certain noise sample before and after compression (original signal and decoded signal) are given. Figure 7 The comparison results in the frequency domain of a noise sample before and after compression (original signal spectrum and decoded signal spectrum) are presented to illustrate that the present invention has good representation and recovery capabilities for waveforms without event background. As can be seen from the above figures, the present invention can still retain the main waveform features required for earthquake monitoring and analysis while adapting to the fixed payload requirements of BeiDou short message service.

[0059] In a further embodiment, the sending end appends time identifier information to the header of the compressed minute-level real-time monitoring data block. The time identifier information is preferably the minute offset of the current whole-minute data block relative to the start of its 4-hour time period, where the minute offset is represented by 1 byte and ranges from 0 to 239. In this way, the receiving end can accurately correlate the minute offset with the actual time of the whole-minute data block based on its local time. As long as the data is successfully transmitted within 4 hours, the actual time corresponding to the data block can be correctly recovered.

[0060] 2. General communication layer, priority scheduling and high-precision backhaul method.

[0061] To adapt to the characteristics of BeiDou short message service—small payload, low-frequency transmission, and uncertain link—this invention further designs a general communication layer in addition to the compression method. This communication layer does not directly rely on the specific semantics of the service data, but is responsible for message-level identification, data block boundary description, cross-message continuation, missing data detection, retransmission recovery, and high-precision backhaul support, thereby uniformly carrying conventional real-time monitoring data, retransmitted data, and high-precision backhaul data. Through this general communication layer, this invention decouples service data organization from the physical transmission of short messages, enabling different types of data to be encapsulated, scheduled, and recovered using a unified data block and message structure.

[0062] For ease of description, this invention abstracts the logically transmitted service content into data blocks (Data), and the actual message sent via BeiDou short message service into a message (Message). A single message includes a message-level header and a payload area. The message-level header includes at least a message sequence number N and a block position field P; a block length field L is set at the start position of each new data block in the payload area. The message sequence number N identifies the position of the current message in the transmission sequence, providing a basis for the receiver's sorting, missing data detection, and retransmission index; the block position field P identifies the starting offset of the length field of the first new data block in the current message payload area; if the starting part of the current message payload is a continuation of the previous data block, the receiver can determine the length of the continuation part and the starting position of the new data block based on P; the block length field L represents the logical total length of the subsequent corresponding data blocks, allowing the receiver to identify data block boundaries; simultaneously, P and L can also form a mutual verification. Figure 8 The overall structure of the communication protocol for the BeiDou short message link is shown, which can be used to illustrate the overall composition of the general communication layer of this invention.

[0063] During message organization, if a data block can be completely placed into the current message, it is written in its entirety; if a data block cannot be completely placed into the current message, only the portion that can be accommodated is written (e.g., ...). Figure 9 Data3(1) in the middle), the rest (such as ... Figure 9 Data3(2) in the message continues to be sent in subsequent messages, and the length field corresponding to the data block is not written again for the continuation part. When a message contains both the continuation part of the previous data block and the start part of a new complete data block, the position of the first complete data block length field in the current message is identified by the block position field P, so as to ensure that the receiver can correctly splice and recover cross-message data blocks. Figure 9 A schematic diagram of a multi-data-block concatenation and encapsulation method is provided to illustrate the organization method when a message simultaneously accommodates the continuation portion and the start content of a new data block. Through this organization method, the present invention can fully utilize the remaining space in the current message, continuing to fill in the start portion of the next data block after sending the tail of one data block, thereby reducing short message payload waste and improving link utilization. Figure 9 In this table, N1, N2, and N3 are message sequence numbers; P1, P2, and P3 are block location fields; L1, L2, L3, and L4 are block length fields; and Data1, Data2, and Data4 are data blocks. M1, M2, and M3 represent three messages.

[0064] When the conditions for sending BeiDou short messages are met, the sending end does not simply send newly generated data sequentially, but performs unified scheduling for different service types. This invention supports at least three types of services to be sent: high-precision backhaul data, retransmission and resend data, and regular real-time monitoring data. In a preferred embodiment, their priorities, from highest to lowest, are: high-precision backhaul data, retransmission and resend data, and regular real-time monitoring data. When a sending opportunity arises, the sending end first checks whether a high-precision backhaul task exists; if it does, it prioritizes organizing the high-precision backhaul message; if it does not exist, it checks whether a retransmission task awaiting retransmission exists; only if neither exists does it send regular real-time monitoring data. Through this priority scheduling mechanism, the transmission of critical and missing data can be prioritized even when BeiDou short message sending opportunities are limited.

[0065] Upon receiving a message, the receiving end first sorts the messages according to their sequence number N. Then, it determines whether the beginning of the message is a continuation of the previous data block based on the block position field P, and identifies the boundaries of the complete data block based on the block length field L. For data blocks sent across multiple messages, the receiving end reassembles them. When the receiving end detects discontinuous message sequence numbers, and these discontinuities persist after a certain period, it determines that a message is missing and records the missing message sequence number. It then generates a retransmission request based on the missing message sequence number and sends it to the sending end. Upon receiving the retransmission request, the sending end retrieves the corresponding message from the sent message buffer, sends it first according to a priority scheduling strategy, and does not retransmit it with other data blocks. Instead, it exclusively receives the entire message to maintain the consistency of the retransmitted message structure with the original message, reducing the complexity of the receiving end's reassembly and deduplication processing. Figure 10 The timing diagram of the message loss detection and retransmission recovery mechanism is shown to illustrate the data recovery process of the present invention under the conditions of packet loss and out-of-order delivery in the link.

[0066] To balance real-time accessibility with post-event detailed analysis capabilities, this invention further incorporates a high-precision on-demand backhaul mechanism in addition to the conventional real-time monitoring mode. The sending end continuously stores raw, high-sampling-rate, multi-channel waveform data in a local cache. When the receiving end needs to review and analyze a specific time period corresponding to an earthquake event, it sends a high-precision backhaul request to the sending end, including at least the start time, target channel, and target duration. Based on the request parameters, the sending end extracts the high-sampling-rate data for the corresponding time range and channel from its local cache and organizes it into segments according to the short message payload capacity. Unlike conventional real-time monitoring data, high-precision backhaul data is geared towards event review and manual interpretation. Its priority is higher than retransmission and supplementary data and conventional real-time monitoring data. It enters a high-priority backhaul queue at the sending end and is not concatenated with other data blocks.

[0067] In a preferred embodiment, the transmitting end extracts 1 second of single-channel raw waveform data for each high-precision return request, downsamples the original 250Hz waveform to 100Hz, and simultaneously converts the sampled data format from float32 to float16 to reduce the data volume while preserving waveform details during critical periods as much as possible. Let the high-precision return sampling rate be... The length of a single return time slice is The number of return channels is The number of bytes occupied by a single sampling point is The amount of data in a single high-precision return time slice It can be represented as: .

[0068] In a preferred embodiment, , , And float16 single sampling points occupy The data volume of a single high-precision return time slice is: .

[0069] Therefore, the main body of the high-precision backhaul data (1 second, 100Hz, single channel, float16 format) is 200 bytes. Even after adding a 1-byte type identifier, a 4-byte timestamp field, and necessary message-level fields, it can still be controlled within a single effective payload of 229 bytes. Through this design, high-precision backhaul does not involve unconstrained transmission of the original continuous data. Instead, it reorganizes the waveform of the critical event period into transmittable data segments according to the short message payload boundaries, thereby achieving on-demand retransmission with higher time resolution than conventional 10Hz real-time monitoring under constrained link conditions.

[0070] For high-precision backhaul data, since its request time may exceed the current 4-hour timeframe, it is not feasible to rely solely on the 1-byte minute offset used for regular real-time monitoring data for time identification. Therefore, for high-precision backhaul data, this invention uses a 4-byte Unix timestamp to identify the data time and sets a channel identifier field to identify the target channel corresponding to the backhaul data; simultaneously, a 1-byte reserved value is set in the original minute offset field position as a high-precision backhaul type identifier, preferably 0xFF. After receiving this type of data block, the receiving end enters the high-precision backhaul parsing process based on the reserved value 0xFF, recovers its absolute time position based on the Unix timestamp, determines the target channel based on the channel identifier, and completes waveform splicing according to the Unix timestamp order, thereby obtaining high-precision waveform data for the specified time period and channel.

[0071] Because this invention simultaneously supports three types of services—high-precision backhaul, retransmission, and routine real-time monitoring—routine real-time monitoring data may experience a certain transmission lag within a short period. To address this, this invention employs a data block organization method within a general communication layer. After sending a complete data block in a message, if there is still remaining payload space, the beginning portion of the next data block is written to improve payload utilization. Simultaneously, since high-precision backhaul and retransmission are typically event-triggered or exception-triggered services, after these tasks are completed, the sending end continues to retrieve the delayed minute-window data from the routine real-time monitoring queue. When multiple data blocks to be sent exist in the routine real-time monitoring queue, the sending end preferably continuously fills in subsequent data blocks or their initial fragments, provided the remaining payload of a single message allows, to improve payload utilization. This gradually eliminates transmission lag and restores a stable real-time state over a period of time. Through this mechanism, this invention can simultaneously meet the three requirements of real-time monitoring, missing data recovery, and high-precision retransmission of critical events within a limited transmission cycle.

[0072] 3. System physical object, receiving end page and overall implementation method.

[0073] In one embodiment, the system of the present invention comprises three parts: a data acquisition end, a transmitting end, and a receiving end. The data acquisition end is used to continuously acquire raw seismic waveforms. The transmitting end is used to receive data output from the data acquisition end and perform local caching, minute-level organization, target channel selection, downsampling, adaptive compression, message encapsulation, priority scheduling, retransmission response, high-precision backhaul processing, and BeiDou short message transmission. The receiving end is used to receive BeiDou short messages and perform message parsing, sequence number sorting, missing data detection, data block splicing and recovery, decompression and reconstruction, result storage, station status monitoring, and waveform display. Through the coordinated operation of the above three parts, the present invention forms a complete closed loop from front-end acquisition, end-side processing, satellite short message transmission to back-end reception and display.

[0074] In a preferred embodiment, the transmitter is implemented based on a low-power embedded platform, preferably the STM32H743 platform. Besides handling minute-level processing of routine real-time monitoring data and sending BeiDou short messages, this transmitter is also responsible for maintaining a raw high-sampling-rate waveform buffer, a sent message buffer, a routine real-time monitoring queue, a retransmission queue, and a high-precision backhaul queue. The raw high-sampling-rate waveform buffer is used to extract data for a specified time period upon receiving a high-precision backhaul request; the sent message buffer is used to quickly locate and resend the corresponding message after the receiver reports a missing message; the three types of transmission queues are used to implement priority scheduling among high-precision backhaul, retransmission, and routine real-time monitoring data.

[0075] Figure 11This is a schematic diagram of the receiving station status monitoring page, illustrating a centralized monitoring method for the message reception status and link operation status of each front-end station. In one implementation, this page displays the message reception status of different stations within a certain time range through a message status graph, and displays information such as station number, operation status, message quantity, packet loss quantity, and success rate through a station status summary area. Through this page, maintenance personnel can intuitively grasp the message continuity, link stability, and abnormal packet loss of each station, thereby promptly detecting problems such as station offline, link anomalies, message loss, or data transmission interruption.

[0076] Figure 12 This diagram illustrates the waveform viewing and SAC export page at the receiving end, demonstrating how the receiving end can query, display, and export the recovered seismic waveforms. In one implementation, the page allows filtering of target waveforms by station number, start and end time, and data type, displaying the recovered seismic waveforms in the waveform display area. When further manual verification, phase analysis, or professional software processing is required, the retrieved waveform data can be exported as an SAC format file. Through this page, the receiving end can not only visualize waveform data but also provide data support for seismic event verification and subsequent detailed analysis.

[0077] Example The real-time earthquake data monitoring and high-precision backhaul method based on BeiDou short messages proposed in this invention was applied to the real-time earthquake monitoring work on a certain island.

[0078] To address the challenges of insufficient public network access, dispersed station distribution, and high requirements for long-term continuous operation in earthquake monitoring scenarios on a certain island, a system architecture consisting of a data acquisition unit, a transmitter, and a receiver was adopted for practical deployment. The data acquisition unit continuously acquires raw seismic waveforms; the transmitter performs local caching, minute-level organization, downsampling, adaptive compression, message encapsulation, and BeiDou short message transmission on the acquired high-sampling-rate, multi-channel seismic waveforms; and the receiver receives BeiDou short messages, performs message parsing, data recovery, result storage, and visualization, thus forming a complete closed loop from front-end acquisition, edge processing, satellite short message transmission to back-end reception and display.

[0079] In this embodiment, 11 BeiDou seismometers are deployed on the island for real-time monitoring, with some stations also equipped with 4G-based seismometers for waveform comparison. The front-end equipment at each station continuously acquires seismic waveforms on-site, while the transmitting end is implemented on an embedded platform, preferably using the STM32H743 platform. During operation, the transmitting end writes the original high-sampling-rate, multi-channel seismic data to local storage for extraction of the original waveforms when a high-precision backhaul request is triggered. Simultaneously, using minute windows as the basic processing unit, it performs preprocessing, resampling, wavelet decomposition, quantization encoding, and adaptive iterative control oriented towards output code length constraints on the target channel data, ensuring that the compression result stably meets the BeiDou short message payload limitations. The compressed real-time monitoring data, along with retransmission and high-precision backhaul data in the system, is organized and scheduled through the general communication layer and sent to the receiving end via the BeiDou short message link.

[0080] In this embodiment, after receiving a message, the receiving end performs message sorting, cross-message splicing, data recovery, and result storage based on fields such as message sequence number, block position, and block length. When the receiving end detects missing messages or incomplete data block recovery, it automatically generates a retransmission request and sends it to the sending end. When further analysis of an abnormal period or specific event is required, the receiving end can also initiate a high-precision backhaul request to the sending end to obtain raw high-sampling-rate waveform data for a specified time range and channel. Thus, this invention achieves coordinated operation of conventional real-time monitoring, missing message detection and retransmission recovery, and high-precision on-demand retransmission in actual operation.

[0081] To verify the application effect of the present invention under real earthquake event conditions, this embodiment further incorporates verification with an actual recorded earthquake event. This event was a strong earthquake occurring in a certain sea area, with a magnitude of 7.6 MW and a focal depth of approximately 230 km. Due to the large magnitude and deep focal depth of this event, its teleseismic waveform could be recorded by the island station, making it suitable as a verification example for the real-time transmission capability and waveform recovery effect of the present invention's BeiDou short message service. Figure 14 The diagram shows the waveform of the teleseismic event transmitted in real time via the BeiDou short message link at the BeiDou seismic monitoring station on the island and recovered at the receiving end, as well as a comparative overview of the waveform transmitted in real time by the 4G station at the same location. Figure 14 The vertical curves correspond to different station or terminal numbers. Among them, Beidou_4270580, Beidou_4270579, Beidou_4270578 and Beidou_4270576 represent waveforms obtained by real-time transmission and recovery from different Beidou seismic monitoring stations via Beidou short messages, while 4G_EB007441, 4G_EB007440, 4G_EB007437 and 4G_EB007436 represent waveforms obtained by real-time transmission from 4G reference stations in the same area or near the same location. Figure 14 The upper part of the figure shows the waveform recovered from real-time transmission via BeiDou short message service, while the lower part shows the reference waveform from 4G real-time transmission. This illustrates the invention's ability to express the main waveform characteristics of distant earthquakes in real-time under limited BeiDou short message link conditions. As shown in the figure, around 12:49:40 to 12:50:00, the recovered waveforms from multiple BeiDou stations exhibited significant distant earthquake arrival responses, and their main waveform morphology, vibration duration, and subsequent attenuation trends corresponded to the 4G reference waveform. This demonstrates that the invention can express the main earthquake waveform characteristics through the BeiDou short message link in real distant earthquake event scenarios. Furthermore, Figure 15 The figure shows a local comparison of the 4G real-time transmission waveform, the BeiDou real-time backhaul waveform, and the high-precision backhaul waveform obtained from the teleseismic event record of the same location in this sea area. In Figure 15, 4G-EB007440-Z-250Hz represents the reference waveform obtained from the real-time transmission of the Z channel of the 4G reference station at the same location with a sampling rate of 250Hz; Beidou-4270578-Z-10Hz represents the 10Hz waveform obtained by downsampling, adaptive compression, and real-time backhaul of the Z channel of BeiDou station 4270578 through BeiDou short message; Remote Command-4270578-Z-100Hz represents the 100Hz Z channel waveform obtained by the sending end after the receiving end initiates a high-precision backhaul request, which is extracted from the local original waveform buffer and organized for backhaul. This comparison demonstrates that, for waveforms after a remote seismic event reaches the island station, this invention can still achieve real-time expression and recovery of the main waveform features under the limited BeiDou short message link conditions; when higher fidelity is required, it can also obtain waveform data with higher time resolution for key periods through a high-precision backhaul mechanism, thereby meeting the needs of event review and fine analysis.

[0082] It should be noted that, Figure 14 and Figure 15 The 4G waveform in the figure is not a necessary transmission link upon which this invention relies, but is used as reference data at the same location to verify the consistency between the BeiDou short message real-time return waveform and the high-precision return waveform and the conventional public network real-time transmission waveform. Through this comparison, it can be more intuitively illustrated that this invention can still retain the main waveform characteristics of teleseismic events in scenarios without a public network or with a weak public network, and supplement the higher time resolution waveforms for critical periods through the high-precision return mechanism when needed.

[0083] Furthermore, in the actual operation of this embodiment, the system and method proposed in this invention demonstrate good engineering adaptability. On the one hand, the transmitting end can run for extended periods on a low-power embedded platform, adapting to the unattended application requirements of stations in remote areas; on the other hand, the receiving end can perform station status monitoring, waveform display, and data management, facilitating remote monitoring of the operating status and data reception of each station by maintenance personnel. Therefore, this embodiment shows that this invention can not only achieve real-time monitoring and reliable transmission of continuous seismic data under limited satellite short message links, but also meet the high-precision retransmission requirements of data during critical periods, possessing good practical deployment value and promising prospects for widespread application.

[0084] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0085] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the accompanying drawings may include multiple steps or stages, which are not necessarily completed at the same time, but may be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but may be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0086] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0087] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention, and no reference numerals in the claims should be construed as limiting the scope of the claims.

[0088] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for real-time monitoring and transmission of seismic data based on BeiDou short message service, characterized in that, include: The transmitting end organizes continuously acquired seismic waveforms in minute-level time windows, performs downsampling on the seismic waveforms of the target channel, and compresses them using adaptive compression oriented towards fixed load constraints: the downsampled seismic waveforms are subjected to three-level wavelet decomposition, and multiple sub-band coefficients are obtained using the coif5 wavelet basis; each sub-band coefficient is clustered and quantized, and the resulting symbol sequences are sequentially subjected to run-length encoding and Huffman encoding; based on the comparison between the code length of the compressed code stream obtained by encoding and the upper limit of the target load of BeiDou short messages, the number of clusters of at least one sub-band coefficient is dynamically adjusted, and the compressed code stream is output with the goal of minimizing the normalized mean square error of the compressed seismic waveform relative to the downsampled seismic waveform, while satisfying the constraint that the output code length does not exceed the upper limit of the target load. This enables regular real-time monitoring data to be adapted to the upper limit of the single effective payload of BeiDou short messages; after the regular real-time monitoring data is encapsulated into a message through the general communication layer, it is sent to the receiving end; The receiving end parses the received messages, sorts them according to the message sequence number, block position field, and block length field, performs concatenation and reconstruction, and performs missing message detection. When a missing message is detected, a retransmission request is generated. When fine analysis of a specific time period is required, a return transmission request is generated. Concatenation and reconstruction includes: when a message contains both the continuation part of the previous data block and the start part of the new data block, the block position field P indicates the starting offset of the block length field of the first complete data block in the current message. The receiving end determines the length of the continuation part and the starting position of the new data block according to P, and identifies the data block boundary in combination with the block length field L to complete cross-message concatenation and reconstruction. The sending end responds to the retransmission request by retrieving the corresponding message from the sent message buffer as retransmission data; the sending end responds to the return request by retrieving the seismic waveforms of the specified time period and channel from the local raw waveform buffer and organizing them into return data; after encapsulating the retransmission data and return data into messages through the general communication layer, the retransmission data and return data are scheduled to be sent to the receiving end with a higher priority than regular real-time monitoring data. In the general communication layer, the data to be transmitted is abstracted into data blocks, and a single BeiDou short message is abstracted into a message. Each message includes a message-level header and a payload area. The message-level header includes a message sequence number and a block location field. The payload area sets a block length field at the beginning of each new data block. If a data block can be completely placed into the current message, it is written in its entirety; otherwise, only the portion that can be accommodated is written. The remaining portion of the data block is continued in subsequent messages, and the continued portion is not written to the block length field again.

2. The method for real-time monitoring and transmission of seismic data based on BeiDou short message service according to claim 1, characterized in that, The priorities, from highest to lowest, are: backhaul data, retransmitted and supplemented data, and routine real-time monitoring data.

3. The method for real-time monitoring and transmission of seismic data based on BeiDou short message service according to claim 1, characterized in that, The sending end responds to the transmission request by extracting seismic waveforms of a specified time period and channel from its local raw waveform cache and organizing them into transmission data, specifically including: In response to the return request, the sending end extracts 1 second of the original seismic waveform from a single channel each time, downsamples the frequency of the original seismic waveform from 250Hz to 100Hz, and converts the data format of the sampling points from float32 to float16, so that the main data size of a single time slice of the returned data is 200 bytes. The returned data uses a 4-byte Unix timestamp for absolute time identification, a channel identifier field to identify the target channel corresponding to the returned data, and a 1-byte reserved value is set in the original minute offset field position as a return type identifier. The receiving end identifies the transmitted data based on the reserved value, recovers the absolute time based on the Unix timestamp, determines the target channel based on the channel identifier, and completes the seismic waveform stitching according to the Unix timestamp order.

4. A real-time monitoring and transmission system for seismic data based on BeiDou short message service, used to implement the method described in any one of claims 1 to 3, comprising: The acquisition end is used for continuous acquisition of seismic waveforms; At the transmitting end, continuously acquired seismic waveforms are organized in minute-level time windows. The seismic waveforms of the target channel are downsampled and compressed using adaptive compression oriented to fixed load constraints, so that the conventional real-time monitoring data can be adapted to the upper limit of the single effective payload of Beidou short message. After the conventional real-time monitoring data is encapsulated into a message through the general communication layer, it is sent to the receiving end. In response to a retransmission request, the sender retrieves the corresponding message from the sent message buffer as retransmission data; in response to a return request, the sender retrieves the seismic waveforms of the specified time period and channel from the local raw waveform buffer and organizes them into return data; after encapsulating the retransmission data and return data into messages through the general communication layer, the sender schedules and sends them to the receiver with a higher priority than regular real-time monitoring data. At the receiving end, the received messages are parsed, sorted, reassembled, and missing messages are detected based on the message sequence number, block location field, and block length field. When a missing message is detected, a retransmission request is generated, and when detailed analysis of a specific time period is required, a return request is generated.

5. A real-time seismic data monitoring and transmission system based on BeiDou short message service as described in claim 4, characterized in that, The compression method employs adaptive compression oriented towards fixed load constraints, specifically including: The downsampled seismic waveform is subjected to a three-level wavelet decomposition, and multiple sub-band coefficients are obtained using the coif5 wavelet basis. Each sub-band coefficient is clustered and quantized, and the resulting symbol sequence is sequentially subjected to run-length coding and Huffman coding. Based on the comparison between the code length of the compressed code stream obtained by encoding and the upper limit of the target payload of Beidou short message, the number of clusters of at least one sub-band coefficient is dynamically adjusted, and the compressed code stream is output under the constraint that the output code length does not exceed the upper limit of the target payload, with the goal of minimizing the normalized mean square error of the compressed seismic waveform relative to the downsampled seismic waveform.

6. A real-time seismic data monitoring and transmission system based on BeiDou short message service as described in claim 4, characterized in that, In the general communication layer, the data to be transmitted is abstracted into data blocks, and a single BeiDou short message is abstracted into a message. Each message includes a message-level header and a payload area. The message-level header includes a message sequence number and a block location field. The payload area sets a block length field at the beginning of each new data block. If a data block can be completely placed into the current message, it is written in its entirety; otherwise, only the portion that can be accommodated is written. The remaining portion of the data block is continued in subsequent messages, and the continued portion is not written to the block length field again.

7. A real-time seismic data monitoring and transmission system based on BeiDou short message service as described in claim 4, characterized in that, The splicing and restoration specifically includes: When a message contains both the continuation of the previous data block and the start of a new data block, the block position field P identifies the starting offset of the block length field of the first complete data block in the current message. The receiving end determines the length of the continuation part and the starting position of the new data block based on P, and identifies the data block boundary in combination with the block length field L to complete the cross-message splicing and recovery.

Images