High-density interconnection signal transmission method and system

Abstract

The invention relates to the field of radar signal transmission, in particular to a high-density interconnection signal transmission method and system. The method comprises the following steps: acquiring an original point cloud frame signal; performing dynamic adaptive segmentation on the obtained original point cloud frame signal to generate a plurality of signal data blocks; identifying a vehicle-mounted physical transmission channel; performing channel real-time quality quantification on the vehicle-mounted physical transmission channel, and generating a transmission quality coefficient of each channel; performing transmission matching on the plurality of signal data blocks based on the transmission quality coefficient, and sending the plurality of signal data blocks to a signal receiving end; the received data transmission information is collected based on a signal receiving end; performing delay compensation on the data transmission information to obtain a multi-channel delay compensation parameter; and performing signal transmission regulation and control based on the multi-channel delay compensation parameter. According to the invention, efficient, stable and synchronous radar signal transmission is realized.

Description

Technical Field

[0001] This invention relates to the field of radar signal transmission, and more particularly to a high-density interconnected signal transmission method and system. Background Technology

[0002] During the long-term operation of vehicle-mounted radar, with the continuous increase in sensor data volume and high-frequency sampling of radar signals, traditional radar signal transmission methods have gradually revealed problems such as insufficient bandwidth, high transmission latency, and rising data packet loss rates. Simultaneously, external interference, electromagnetic noise, and vehicle network security threats can also affect the transmission quality and reliability of radar signals. These factors may lead to delays and decreased accuracy in target identification information, and even cause misjudgments or safety accidents in autonomous driving systems. Therefore, ensuring the integrity, reliability, and real-time performance of vehicle-mounted radar signals during high-density, high-speed transmission has become an urgent technical challenge.

[0003] Traditional radar signal transmission methods primarily rely on standard communication interfaces or low-density data channels, such as CAN bus, Ethernet, or simple fiber optic links. While these methods can meet basic requirements in low-speed or medium-volume data transmission scenarios, they often suffer from bandwidth bottlenecks, inconsistent latency, and signal conflicts when dealing with multiple radars, high resolution, and parallel transmission across multiple channels. Furthermore, existing transmission schemes largely depend on static configuration or timed scheduling, lacking specificity and adaptability, making it difficult to achieve real-time optimization and high-density transmission management of radar signals. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a high-density interconnect signal transmission method and system, thereby resolving at least one of the aforementioned technical problems.

[0005] To achieve the above objectives, the present invention provides a high-density interconnect signal transmission method, comprising the following steps: Step S1: Acquire the original point cloud frame signal; dynamically adapt and segment the acquired original point cloud frame signal to generate multiple signal data blocks; Step S2: Identify the vehicle-mounted physical transmission channels; perform real-time quality quantization on the vehicle-mounted physical transmission channels to generate a transmission quality coefficient for each channel; Step S3: Based on the transmission quality coefficient, perform transmission matching on multiple signal data blocks and send the multiple signal data blocks to the signal receiving end; Step S4: Based on the data transmission information received by the signal receiver, perform delay compensation on the data transmission information to obtain multi-channel delay compensation parameters; Step S5: Perform signal transmission control based on the multi-channel delay compensation parameters.

[0006] This specification provides a high-density interconnect signal transmission system for performing the high-density interconnect signal transmission method described above, comprising: The data segmentation unit is used to acquire the original point cloud frame signal; it dynamically adapts and segments the acquired original point cloud frame signal to generate multiple signal data blocks. A quality quantization unit is used to identify the vehicle-mounted physical transmission channel; to perform real-time quality quantization on the vehicle-mounted physical transmission channel and generate a transmission quality coefficient for each channel; A transmission matching unit is used to perform transmission matching on multiple signal data blocks based on the transmission quality coefficient, and send the multiple signal data blocks to the signal receiving end; The delay compensation unit is used to collect and receive data transmission information from the signal receiver and perform delay compensation on the data transmission information to obtain multi-channel delay compensation parameters. The transmission control unit is used to control signal transmission based on the multi-channel delay compensation parameters.

[0007] The beneficial effects of this invention are as follows: By dynamically adapting and segmenting the original point cloud frame signal, the system can flexibly segment the signal according to the point cloud density, data bandwidth, and current transmission channel status, avoiding buffering delays caused by excessively large data blocks. After being segmented into multiple independent data blocks, multi-channel parallel transmission can be achieved, thereby significantly improving the real-time transmission rate of LiDAR data. Dynamic segmentation can automatically adjust the data block size according to the network status, reducing the pressure of packet loss or retransmission caused by bandwidth fluctuations. Through block transmission and block processing, the receiving end can decode while receiving, improving the overall system response speed. By identifying each physical channel (such as Ethernet, CAN-FD, fiber optic channel, etc.) and quantifying its real-time transmission quality, the link health status can be dynamically monitored. The transmission quality coefficient provides a basis for subsequent data block scheduling, enabling the system to automatically optimize data allocation strategies based on channel performance. Real-time quantization can trigger automatic switching or degradation strategies when channel quality deteriorates, ensuring signal continuity and robustness. This step provides a unified quality assessment basis for different protocol channels (such as Ethernet, USB, PCIe), improving system compatibility. By dynamically allocating data based on transmission quality coefficients, high-priority or large-volume data blocks can be preferentially assigned to high-quality channels, improving overall throughput. This avoids over-occupancy of high-load channels and enhances the real-time performance and balance of data transmission. When the quality of a channel degrades, the system can quickly re-allocate data blocks to other channels, ensuring continuous data transmission. In multi-radar systems, channel resources can be dynamically scheduled based on the importance of each radar signal, achieving more efficient bandwidth utilization. By calculating delay compensation parameters, time alignment of point cloud data transmitted across multiple channels can be performed at the receiving end, ensuring the integrity and synchronization of point cloud frames. Delay compensation reduces time offset in multi-channel transmission, avoiding spatial errors or ghosting in point clouds caused by delay differences. When channel performance fluctuations lead to inconsistencies in delay, the compensation mechanism can correct this in real time, maintaining stable system output. The compensated synchronization signal can be directly used in high-speed sensing algorithms (such as SLAM and target recognition), ensuring the temporal consistency of high-frame-rate radar data. By applying delay compensation parameters inversely to the channel scheduling logic, dynamic control and adaptive optimization are achieved. The system can automatically adjust data block allocation, transmission rate, or channel weights, achieving self-learning optimization. Maintain stable transmission performance in complex in-vehicle environments (such as temperature changes, interference, and electromagnetic noise). The optimized signal has higher synchronization accuracy and time consistency, providing reliable data support for the perception layer of autonomous driving. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the steps of a high-density interconnect signal transmission method according to the present invention; Figure 2 This is a detailed flowchart illustrating the implementation steps of step S1. Figure 3This is a flowchart illustrating the detailed implementation steps of step S2. Detailed Implementation

[0009] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0010] This application provides a high-density interconnect signal transmission method and system. The execution entities of the high-density interconnect signal transmission method and system include, but are not limited to, mechanical equipment, data processing platforms, cloud server nodes, network upload devices, etc., which can be considered as general computing nodes in this application. The data processing platform includes, but is not limited to, at least one of an audio / image management system, an information management system, and a cloud data management system.

[0011] Please see Figures 1 to 3 This invention provides a high-density interconnect signal transmission method, comprising the following steps: Step S1: Acquire the original point cloud frame signal; dynamically adapt and segment the acquired original point cloud frame signal to generate multiple signal data blocks; Step S2: Identify the vehicle-mounted physical transmission channels; perform real-time quality quantization on the vehicle-mounted physical transmission channels to generate a transmission quality coefficient for each channel; Step S3: Based on the transmission quality coefficient, perform transmission matching on multiple signal data blocks and send the multiple signal data blocks to the signal receiving end; Step S4: Based on the data transmission information received by the signal receiver, perform delay compensation on the data transmission information to obtain multi-channel delay compensation parameters; Step S5: Perform signal transmission control based on the multi-channel delay compensation parameters.

[0012] In this embodiment, see Figure 1 The diagram below illustrates the steps of a high-density interconnect signal transmission method according to the present invention. In this example, the steps of the high-density interconnect signal transmission method include: Step S1: Acquire the original point cloud frame signal; dynamically adapt and segment the acquired original point cloud frame signal to generate multiple signal data blocks; In this embodiment, the original point cloud frame signal is acquired through a multi-line lidar scanning unit. This point cloud frame signal is formed by multiple beams of light emitted by the laser and reflected signals from environmental objects, containing information such as distance, reflection intensity, and angle. It is typically generated by sampling at a frequency of millions of points per second. To achieve efficient subsequent data transmission, the system dynamically adapts and segments the original point cloud frame signal. The segmentation process first calculates the local aggregation degree of the point cloud through spatial density analysis, classifying high-density areas (such as areas in front of vehicles or areas with concentrated obstacles) and low-redundancy areas (such as flat roads or open areas). Subsequently, based on the point cloud distribution characteristics and intra-frame timestamp differences, the point cloud frame is divided into multiple signal data blocks, each corresponding to a specific spatial range and sampling time period. To ensure the transmissibility of the data structure, each data block header embeds index information, priority tags, and frame synchronization identifiers for subsequent channel matching and transmission scheduling. Dynamic adaptation segmentation not only achieves hierarchical management of point cloud data in spatial and temporal dimensions but also lays the foundation for multi-channel parallel transmission, enabling the system to adaptively allocate signals according to the importance and bandwidth conditions of different areas, thereby improving overall transmission efficiency and data reconstruction accuracy.

[0013] Step S2: Identify the vehicle-mounted physical transmission channels; perform real-time quality quantization on the vehicle-mounted physical transmission channels to generate a transmission quality coefficient for each channel; In this embodiment, during the signal transmission phase, it is necessary to first identify the physical transmission channels in the vehicle's internal and external communication systems. Typical vehicle communication links include Ethernet, CAN-FD, fiber optic channels, and wireless transmission modules. The system uses a channel identification module to read channel bandwidth, link status, and transmission delay information to establish a multi-channel transmission topology. Subsequently, real-time quality quantization is performed on each physical channel to generate a transmission quality coefficient. The quality quantization process is based on multiple performance parameters, including bandwidth utilization, latency jitter, packet loss rate, and link stability. Bandwidth utilization reflects the channel load; latency jitter indicates the instability of data packet transmission time; packet loss rate measures data reliability; and link stability is used to assess long-term communication continuity. By continuously monitoring these indicators and performing weighted calculations, a real-time transmission quality coefficient for each channel can be obtained. This coefficient dynamically reflects changes in channel availability and transmission reliability. When a channel experiences bandwidth fluctuations or an increase in packet loss rate, its quality coefficient will decrease, thus being automatically downweighted in subsequent scheduling. This mechanism ensures that high-density point cloud data can be preferentially transmitted in stable, low-latency channels, reducing signal loss and retransmission overhead.

[0014] Step S3: Based on the transmission quality coefficient, perform transmission matching on multiple signal data blocks and send the multiple signal data blocks to the signal receiving end; In this embodiment, after channel quality quantization, the system executes an adaptive transmission matching strategy based on the importance, size, and priority of each signal data block, combined with the quality coefficients of each transmission channel. The matching process is implemented through a multi-dimensional decision model, allocating high-priority, high-density point cloud data blocks to channels with higher transmission quality coefficients, while allocating low-priority or low-redundancy data blocks to secondary channels to achieve optimal bandwidth resource allocation. After matching, the system generates a channel tag for each data block and embeds the corresponding channel identifier and time synchronization information in the data block header to ensure accurate signal reassembly at the receiving end. To further optimize transmission efficiency, the system introduces a dynamic scheduling mechanism during the transmission phase, adjusting the data block transmission order based on real-time channel status. When an increase in latency or packet loss is detected in a certain channel, the data block is immediately reallocated to a backup channel. Through this dynamic transmission matching method, the system can maximize bandwidth utilization and reduce transmission latency fluctuations under multi-channel parallel transmission conditions, ensuring stable, low-latency signal interconnection of high-density radar point cloud data in complex network environments.

[0015] Step S4: Based on the data transmission information received by the signal receiver, perform delay compensation on the data transmission information to obtain multi-channel delay compensation parameters; In this embodiment, at the signal receiving end, the system collects and records the transmission information of all incoming data blocks for channel delay analysis and compensation calculation. Each data block is accompanied by a transmission timestamp, channel identifier, and arrival time information upon reception. The system calculates the single-channel delay by comparing the transmission and reception time differences. To accurately reflect the dynamic characteristics of the delay, the system continuously samples the delay value within multiple time slices to generate a delay fluctuation curve. Subsequently, a sliding window filtering method is used to eliminate random fluctuation interference and extract periodic delay changes and sudden delay anomalies. Based on these characteristics, the system calculates the average delay, delay variance, and compensation offset for each channel, generating multi-channel delay compensation parameters. The compensation parameters are used to correct the time differences between channels in subsequent transmission control, ensuring that the point cloud data can be accurately reassembled in chronological order at the receiving end. Especially in high-speed driving or complex communication environments, the delay compensation mechanism can effectively reduce the problem of data block arrival time misalignment and ensure the time consistency of the point cloud frame structure. The final delay compensation parameters will be continuously updated to achieve adaptive correction of dynamic channel delay.

[0016] Step S5: Perform signal transmission control based on the multi-channel delay compensation parameters.

[0017] In this embodiment, after calculating the multi-channel delay compensation parameters, the system enters the signal transmission control stage to achieve adaptive optimization of the overall point cloud transmission process. The control process is completed collaboratively through a two-layer feedforward and feedback mechanism. First, the system fine-tunes the transmission timing of each channel using the delay compensation parameters, ensuring that high-priority data blocks are transmitted first and correctly assembled in chronological order at the receiving end. Subsequently, combining real-time channel quality coefficients and historical delay fluctuation trends, the system dynamically adjusts the data transmission rate and buffer depth of each channel to prevent channel congestion or uneven load. Based on this, the system can optimize the compression ratio and encoding strategy according to the delay compensation results, achieving a dynamic balance between delay and bandwidth. For example, when the delay compensation is large, the system reduces the data compression ratio in high-density areas to reduce the decoding burden; when the channel delay is stable, the compression ratio is appropriately increased to improve transmission efficiency. Through this collaborative control method, the system achieves delay consistency control and data throughput optimization of point cloud signals in a multi-channel environment, ensuring stable, low-latency, and high-fidelity real-time transmission of high-density vehicle-mounted LiDAR signals in high-speed interconnect scenarios.

[0018] In this embodiment, see Figure 2 The diagram below illustrates the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Original point cloud frame signals are acquired based on a multi-line lidar scanning unit; Spatial density analysis is performed on the original point cloud frame signal to divide the point cloud signal into high-density regions and low-redundancy regions. The scene information of the high-density region and the low-redundancy region are analyzed to generate scene information complexity. Adaptive compression coding is performed based on the complexity of scene information to generate a hierarchical coding sequence; The hierarchical coded sequence is dynamically adapted and segmented to generate multiple signal data blocks.

[0019] In this embodiment, the multi-line lidar scanning unit periodically scans the vehicle's driving environment with a fixed angular resolution and rotation speed, acquiring three-dimensional distance information of the target space by emitting laser pulses and receiving reflected signals. Each laser line is emitted at different elevation angles, achieving 360° panoramic coverage in conjunction with a rotation mechanism. Laser time-of-flight (ToF) is used to calculate the precise distance from the point to the sensor center, while simultaneously recording reflection intensity, echo count, and timestamp information to form a complete raw point cloud frame. A single frame of point cloud typically contains millions of spatial points; for example, a 128-line lidar can reach approximately 4 million points per frame, each point containing data dimensions such as (x, y, z, intensity, timestamp). To ensure spatial accuracy, geometric calibration and time synchronization are required, including angle compensation of the optical system, delay correction of each laser emitting unit, and IMU attitude synchronization alignment, ensuring accurate correspondence of point cloud spatial coordinates in a unified reference frame. A scanning frequency of 10 Hz or 20 Hz is used during acquisition to ensure temporal continuity of continuous frame data. The raw point cloud frame is cached in a binary structure in a high-speed cache, providing the basic data for subsequent spatial density analysis and encoded transmission. The raw point cloud frame is first mapped to a three-dimensional Cartesian coordinate system and then spatially partitioned using voxelization. The voxel size is set according to the LiDAR resolution, typically using a 0.1 m × 0.1 m × 0.1 m cube as the basis for spatial partitioning. The number of points in each voxel is counted, and the point density ρ = N / V is calculated to quantify the point distribution within that spatial region. Based on the density threshold ρ... t The settings define areas with density above a threshold as high-density areas and areas below the threshold as low-redundancy areas. The density threshold is determined based on road environment characteristics and the overall point cloud distribution; for example, in an urban road scene, ρ... t Approximately 100 points per cubic decimeter can be used. During density analysis, the reflection intensity and neighborhood structure characteristics of the point cloud are weighted to avoid errors caused by simply dividing the data by quantity. Finally, the analysis results are organized using an octree structure to achieve rapid indexing and spatial mapping of high and low density regions. The result of this step is a generated spatial density distribution map, which characterizes the degree of aggregation of point cloud information in different spatial regions, providing a structural basis for complexity calculation.

[0020] Based on the spatial density partitioning results, scene feature analysis is performed on different regions to establish an information complexity model. High-density regions primarily involve geometric feature extraction, including curvature calculation, normal vector estimation, edge detection, and plane fitting, to identify the geometric features of building edges, vehicle shapes, and road obstacles. Low-redundancy regions focus on evaluating spatial smoothness and point distribution uniformity, quantifying the structural simplicity of the region through information entropy and variance calculations. A complexity function C = α·H + β·σ + γ·κ is constructed by integrating multiple indicators, where H represents the point cloud information entropy, σ is the spatial density variance, κ is the rate of curvature change, and α, β, and γ are weighting coefficients, which can be set to 0.4, 0.3, and 0.3 respectively to balance the influence of geometric and statistical features. The complexity value reflects the structural information content and change intensity of each region, thus forming a complexity distribution map of the entire frame point cloud. This distribution guides subsequent encoding and compression strategies, enabling high-complexity regions to receive high-fidelity encoding, while low-complexity regions can be processed with higher compression ratios. Based on the distribution of scene complexity, adaptive compression encoding is implemented for point cloud data. First, the point cloud is divided into three levels—low, medium, and high—according to different encoding strategies. High-complexity regions employ high-fidelity encoding, using octree-based spatial hierarchical encoding combined with a local geometric prediction model. Difference residual compensation reduces redundancy while maintaining geometric accuracy. Medium-complexity regions apply a hybrid compression strategy, balancing accuracy and compression ratio. Low-complexity regions use voxel downsampling or lightweight encoding methods based on occupancy bits to significantly reduce data volume. During encoding, a dynamic quantization step size and region-adaptive bit allocation mechanism are used to adjust the encoding parameters of each region in real time according to complexity, thereby maintaining overall perception quality under bandwidth constraints. The encoded output is organized in a hierarchical sequence. Each encoding unit contains a region index, complexity identifier, and compression parameters, forming a hierarchical point cloud data stream, providing the input data foundation for subsequent dynamic block partitioning and multi-channel scheduling transmission.

[0021] The hierarchical coded sequence needs to be dynamically segmented before transmission to match different data link conditions. By monitoring real-time bandwidth and buffer usage, the size and allocation order of data blocks are adaptively determined based on delay constraints and priority rules. The coded sequence is first divided into Basic Data Units (BDUs), each containing a fixed number of points or a fixed data byte capacity. Then, through a priority scheduling algorithm, multiple basic units are combined into Signal Data Blocks (SDBs) based on region complexity, timestamps, and spatial locations. Data blocks in high-complexity regions are transmitted first to ensure the timeliness and integrity of critical environmental information; data blocks in low-complexity regions can be transmitted later or in batches, thereby achieving efficient bandwidth utilization. Each data block is accompanied by a region identifier, timestamp, and redundancy check code to ensure data synchronization and complete parsing at the receiving end. After dynamic adaptation and segmentation, the point cloud signal is transmitted in parallel in multiple blocks, significantly improving the transmission rate and real-time performance without affecting the overall data continuity, and realizing rapid interconnection and efficient transmission of high-density lidar signals.

[0022] In this embodiment, see Figure 3 The diagram below illustrates the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Identify the vehicle's physical transmission channels and monitor real-time channel status information; Based on the real-time channel status information, bandwidth utilization, latency jitter, packet loss rate, and link stability are calculated to generate channel transmission status performance indicators. Calculate the real-time available capacity and transmission reliability coefficient of the vehicle-mounted physical transmission channel; Based on the channel transmission status performance indicators, the real-time available capacity, and the transmission reliability coefficient, the channel real-time quality is quantified to generate the transmission quality coefficient for each channel.

[0023] In this embodiment, the vehicle-mounted system contains multiple physical transmission channels, including Ethernet channels (such as 10GBASE-T and 1000BASE-X), PCIe channels, fiber optic channels, and high-speed CAN buses. First, each physical link needs to be identified and classified to establish a unified channel identification table. During identification, channel identification information, such as port number, link speed, transmission medium type, and connection topology, is obtained through hardware interface scanning and link-layer handshake detection. After identification, a link status monitoring mechanism is used to sample the real-time status of each channel. The sampling frequency is typically set between 10 Hz and 100 Hz to balance accuracy and computational overhead. Monitoring parameters include link bandwidth utilization, instantaneous throughput, packet transmission rate, latency, and bit error rate. By statistically analyzing the messages within each sampling period, the actual data traffic of the current channel can be calculated, and the instantaneous occupancy rate can be obtained by combining this with the link's maximum rated speed. The monitoring module also needs to record abnormal events such as sudden congestion and link interruptions to reflect the transient fluctuation characteristics of the link. All collected status information is recorded in time-series format, providing data support for subsequent performance indicator calculations and channel quality quantification. After real-time status acquisition is completed, statistical analysis is performed on the channel data to generate transmission performance indicators. Bandwidth utilization is obtained by calculating the ratio of the actual data volume per unit time to the channel's theoretical maximum bandwidth. For example, in a 10 Gbps Ethernet channel, if 8 Gbit of data is transmitted within the sampling period, the bandwidth utilization is 80%. Delay jitter is characterized by the variance or range of the round-trip time (RTT), calculating the change in transmission delay for multiple consecutive data packets to reflect link delay fluctuations. Packet loss rate is calculated by statistically analyzing the difference in the number of data packets between the sending and receiving ends: Packet loss rate = (Ns - Nr) / Ns × 100%, where Ns is the number of packets sent and Nr is the number of packets received. Link stability is calculated by statistically analyzing the fluctuations in bandwidth and delay within a set time window, and its stability coefficient S = 1 - (σ / μ), where σ is the standard deviation of bandwidth or delay, and μ is its mean. A higher value indicates a more stable link. These indicators form a channel transmission performance description vector, including four dimensions: {bandwidth utilization, delay jitter, packet loss rate, and stability coefficient}. This performance metric can be updated in real time to reflect the dynamic characteristics of the channel under different load and environmental conditions, providing a quantitative basis for subsequent channel quality assessment.

[0024] Channel available capacity refers to the effective bandwidth actually available for data transmission under the current state. It is calculated based on the maximum link bandwidth minus bandwidth occupancy and packet loss compensation. The specific formula is: C_avail = B_max × (1 - U - L), where B_max is the maximum channel bandwidth, U is the bandwidth occupancy rate, and L is the packet loss rate correction coefficient. Available capacity reflects the remaining transmission capacity of the current link and is used for dynamic allocation of point cloud data streams. The transmission reliability coefficient describes the stability and reliability of the link under continuous transmission conditions. Considering packet loss rate, latency jitter, and stability coefficient, it can be defined as: R = (1 - L) × S × exp(-J), where L is the packet loss rate, S is the stability coefficient, and J is the normalized latency jitter index. This coefficient varies between 0 and 1; a higher value indicates stronger reliability. To improve calculation accuracy, a moving average can be applied to multiple consecutive sampling periods to filter out the impact of short-term fluctuations. The calculation results of available capacity and reliability coefficient can dynamically reflect the transmission load status of the channel at different operating stages, providing a foundation for priority scheduling of data blocks and multi-channel parallel transmission strategies.

[0025] After obtaining multiple performance indicators for each channel, a comprehensive evaluation is required to generate transmission quality coefficients that can be directly used for decision-making and scheduling. During quantization, factors such as bandwidth utilization, latency jitter, packet loss rate, link stability, available capacity, and reliability are normalized, and weights are assigned to each parameter according to the importance of the transmission task. Channels with ample bandwidth, stable latency, and low packet loss rate will be assigned higher quality coefficient values. The calculation results reflect not only the instantaneous performance of the channel but also its transmission stability over a continuous time period. To ensure the reliability of the results, quality quantization needs to be performed periodically, typically with an update interval of 100 to 200 milliseconds, to adapt to the high-frequency changes in radar point cloud data streams. The final generated transmission quality coefficients serve as input to the channel scheduling module, enabling adaptive allocation of data streams based on the real-time quality differences of each channel. This allows high-quality channels to prioritize critical data transmission tasks, thereby improving the overall efficiency and transmission security of LiDAR signal interconnection in a multi-channel environment.

[0026] In this embodiment, step S3 includes the following steps: Calculate the size and priority of the plurality of signal data blocks; Based on the transmission quality coefficient, adaptive transmission matching is performed on the size and priority to obtain the optimal transmission channel; Based on the optimal transmission channel, a channel tag is generated, and information is embedded in the data block header to send multiple signal data blocks to the signal receiving end.

[0027] In this embodiment, after hierarchical encoding and segmentation, the size and priority of each signal data block need to be calculated to achieve efficient transmission scheduling. The size of each signal data block is determined by the number of point clouds it contains, the encoding compression ratio, and the amount of auxiliary information contained in the data block. First, the effective payload length of each data block is statistically analyzed, and its total data volume in the binary sequence is calculated, while the timestamp and spatial location index are recorded. Then, the importance level of the data blocks is divided according to the complexity of scene information and the encoding level. For example, point clouds in high-complexity areas (such as vehicle edges, pedestrian or obstacle boundaries) are given high priority, medium-complexity areas are given secondary priority, and planar areas or spatial point clouds are given low priority. The priority calculation also considers real-time performance and task relevance; for example, data blocks related to emergency obstacle avoidance detection have higher priority. To prevent excessive congestion, the system will balance the data block size so that each block is as close as possible to the optimal payload threshold (usually set between 256 KB and 512 KB) during transmission, which can ensure data integrity and improve link utilization. The final result is a scheduling table containing data block size, timestamp, and priority identifier, providing input for subsequent channel matching. After obtaining the size and priority information of each signal data block, adaptive matching needs to be performed in conjunction with the real-time quality coefficient of each physical transmission channel. The matching process is implemented through a multi-dimensional trade-off strategy, comprehensively considering the available bandwidth, transmission delay, reliability, and current load of the channel. First, the priority of the data block is compared with the channel quality coefficient to establish a matching matrix, so that high-priority data blocks tend to be allocated to high-quality, high-stability channels. For large data blocks, links with sufficient bandwidth and low packet loss rate are selected first to ensure integrity; while small or low-priority data blocks can be allocated to links with lower load or suboptimal links to improve overall transmission efficiency. A dynamic adjustment mechanism is introduced during the matching process. When the quality coefficient of a certain channel fluctuates, the system can reallocate some data blocks in real time to maintain stable transmission. In addition, time constraints must be considered, and data blocks close to the delay limit are prioritized for transmission to prevent task delays. This adaptive matching strategy enables optimal utilization of multi-channel resources in complex vehicle communication environments, allowing high-density point cloud data to be rationally allocated among different links, thus balancing transmission timeliness and stability, and obtaining the best transmission channel combination scheme at the current moment.

[0028] After determining the optimal transmission channel, the data blocks to be sent need to be identified and encapsulated to ensure that the receiving end can correctly identify and reassemble them. First, a channel label is assigned to each data block. This label contains the channel number, transmission timestamp, priority identifier, and redundancy check information. The label generation process is automatically completed by the scheduling module based on the matching results, ensuring that data blocks can be accurately routed during multi-channel parallel transmission. Subsequently, the channel label and metadata information are embedded into the data block header to form a complete transmission frame structure. In addition to the channel label, the header also carries a synchronization flag, encoding parameters, and a data block sequence number, used for order recovery and error checking at the receiving end. After encapsulation, the data blocks are sent to different physical interfaces according to the assigned channel. Common transmission methods include 10G Ethernet channels, fiber optic links, and PCIe interconnects. Each channel maintains time synchronization during transmission to ensure consistent timing of point cloud data reassembly at the receiving end. After transmission, the receiving end decapsulates and reassembles the data blocks based on the label information to recover a complete hierarchical point cloud frame. Through this process, multi-channel transmission achieves parallelization and adaptability, enabling high-density lidar point cloud signals to achieve low-latency, high-reliability, fast interconnection and efficient transmission in vehicle systems.

[0029] In this embodiment, step S4 includes the following steps: Based on the data transmission information received by the signal receiver; Calculate the reception timestamp of each signal block based on the data transmission information; Extract the transmission timestamp of the signal data block; The transmission delay of the signal block is calculated based on the sending timestamp and receiving timestamp, and the transmission delay value of each signal block is extracted. Extract the header information of the signal block, trace the transmission channel, and mark the channel delay based on the transmission delay value to obtain the delay value of each channel; Adaptive delay compensation is performed based on the delay values ​​of each channel to obtain multi-channel delay compensation parameters.

[0030] In this embodiment, during multi-channel point cloud data transmission, the signal receiver undertakes the tasks of data aggregation and timing recovery. First, a reception monitoring mechanism for each physical channel needs to be established to capture and register signal data blocks arriving from different channels in real time. The receiver continuously monitors the input buffer queue of each channel through a high-speed buffer interface (such as a PCIe or 10G Ethernet receiver module), recording basic information such as the arrival time of the data block, channel identifier, data length, and verification status. To ensure acquisition accuracy, the receiver employs a high-precision clock synchronization mechanism (such as IEEE 1588 PTP or GPS clock synchronization) to ensure that the timestamps of all channels remain consistent within a microsecond error range. During acquisition, each data block undergoes validity verification, including CRC check, frame header integrity detection, and tag consistency comparison, to prevent bit errors or out-of-order data. For data blocks that pass verification, the receiver stores their metadata in a status buffer, forming a reception information record table containing time, channel, and data attributes. These records form the raw input for subsequent delay calculations and channel compensation analysis, ensuring that the multi-channel transmission status is fully and in real-time reflected at the receiving end, providing high-precision data support for subsequent delay assessment and compensation. When the frame header of a data block is recognized by the receiving module, a timestamp recording is immediately triggered and bound to the unique number of the data block. To avoid time errors caused by multi-threaded concurrency, the receiving module uses time latching technology to ensure that the timestamp record corresponds to the arrival time of the first byte of the data block. After recording, the received timestamp information is stored in the received information table and archived together with the channel number and data block identifier. If a data block fragment arrives or there is a retransmission, the arrival time of the last frame of the complete data block is used as the received timestamp. To further improve time consistency, all timestamps undergo synchronization correction in a unified clock domain, enabling direct comparison of time information from different channels. In this way, each data block has an independent and comparable received timestamp, providing the basis for subsequent matching with the sending timestamp and calculating transmission delay.

[0031] During delay analysis, it is necessary to extract the transmission timestamp recorded at the transmitting end when the corresponding signal data block is generated. The transmission timestamp is usually embedded in the frame header information of the data block and is generated by the scheduling module based on the master clock during the transmission phase. When parsing the data block, the receiving end reads this timestamp from the frame header and matches it with its local received timestamp. During extraction, it is necessary to verify whether the timestamp format and channel label are consistent, ensuring a one-to-one correspondence with the identifier of the currently received data block. If an abnormal timestamp is detected (such as exceeding the synchronization range or a mismatched format), the data block is marked as an abnormal packet to avoid introducing errors into delay statistics. The extracted transmission timestamp and received timestamp are stored together in the delay calculation buffer for subsequent calculations. To prevent errors caused by clock drift, the receiving end synchronizes and corrects the clocks of the transmitting and receiving ends at fixed intervals, thereby ensuring the consistency of the time base. Through this process, each data block forms a pair of valid timestamp records, accurately reflecting the complete transmission path and time span experienced by the signal from transmission to reception. After obtaining the transmission and received timestamps, the transmission delay of each signal data block can be calculated. Delay calculation is based on the time difference between two timestamps, reflecting the total transmission time from signal transmission at the sender to successful reception at the receiver. During calculation, delays caused by non-transmission factors, such as receiver buffer write delays or data parsing time, must be excluded through synchronization control logic. The resulting delay values ​​are initially grouped according to channel type for subsequent channel-level delay analysis. To identify abnormal fluctuations in transmission, the delay values ​​of multiple consecutive data blocks are statistically analyzed to form delay variation curves. If a channel's delay shows a continuous increase or a sudden peak, it is marked as a potentially congested or interfering channel. The delay calculation results are stored in time-series format, with each data block corresponding to an independent delay record. This calculation process provides a clear picture of the transmission status of the multi-channel system at different time periods, offering accurate quantitative data for subsequent channel delay tracking and compensation.

[0032] After calculating the latency value for each data block, its transmission path needs to be traced based on the frame header information to distinguish the latency performance of different channels. The frame header contains channel number, priority identifier, and channel label information. By parsing these fields, the source channel of the data block can be determined. After tracing, a mapping relationship is established between the latency value corresponding to the data block and its channel. Subsequently, statistical analysis is performed on the latency of multiple data blocks within the same channel to calculate the average, maximum, and latency fluctuation range, thereby obtaining the overall latency characteristics of the channel in the current time period. Based on the latency characteristics, each channel is marked with latency; high-latency channels are marked as restricted channels, and low-latency channels are marked as priority channels. This marking information is updated in real time, providing a reference for dynamic scheduling and adaptive compensation. The establishment of latency marking enables the system to form a multi-channel latency state diagram, thereby identifying transmission bottlenecks or unstable links and providing a quantitative basis for subsequent latency balancing and compensation strategies. First, the current baseline latency of the system is determined, i.e., the minimum latency among all channels. Then, the latency difference of other channels relative to this baseline is calculated to obtain the required compensation duration. Compensation can be achieved by adjusting the data transmission timing, buffer release time, or the unpacking rhythm of the receiver. For channels with higher latency, the system prioritizes optimizing their data streams; for channels with lower latency, a microsecond-level delay is introduced at the buffer layer to align the final arrival time with other channels. The adaptive mechanism continuously monitors channel latency changes and automatically updates compensation parameters when environmental or link conditions fluctuate. The resulting multi-channel latency compensation parameters can be applied in subsequent transmission cycles, enabling seamless splicing of point cloud data blocks at the receiving end in spatial and temporal order. This achieves time alignment and synchronous reconstruction of multi-channel radar signals, ensuring the timeliness and consistency of the vehicle-mounted LiDAR system in high-speed transmission scenarios.

[0033] In this embodiment, the specific steps for obtaining multi-channel delay compensation parameters by performing adaptive delay compensation based on the delay values ​​of each channel are as follows: Delay fluctuations at multiple time points are calculated for the delay values ​​of each channel to obtain the delay fluctuation curves for each channel; Sliding window filtering and deep analysis are performed on the delay fluctuation curves of each channel to generate periodic delay variation characteristics and sudden delay anomaly characteristics; Based on the periodic delay variation characteristics and sudden delay anomaly characteristics, the delay variation trend is predicted to obtain the delay prediction sequence for future time periods; The latest delay values ​​of each channel are collected in real time; the delay prediction error of the delay prediction sequence is calculated based on the latest delay values ​​of each channel, and the delay prediction error of multiple channels is extracted; Based on the delay prediction error, the error offset amplitude is inverted and dynamic compensation is performed to obtain multi-channel delay compensation parameters.

[0034] In this embodiment, after obtaining the delay value of each transmission channel, it is necessary to statistically analyze and model its changes at different time points to obtain delay fluctuation characteristics. First, the delay values ​​are sampled and recorded in chronological order to form a time-series dataset. The sampling period is typically set between milliseconds and hundreds of milliseconds to ensure the accuracy of capturing short-term fluctuations. Subsequently, the average delay and fluctuation amplitude are calculated within each time slice to describe the transient trend of delay changes. By fitting delay samples from multiple consecutive time periods, the delay fluctuation curve of each channel can be obtained, which reflects the delay change pattern of the channel during operation. For channels with high stability, the fluctuation curve is relatively flat with a small range of change; while for channels that are greatly affected by external interference or load changes, the curve exhibits obvious fluctuation characteristics. Delay fluctuation calculation is not only used to identify channels with poor stability, but also to reveal potential congestion or jitter sources in the transmission path. Each curve is stored and timestamped for subsequent filtering and trend analysis, thereby realizing continuous monitoring and dynamic modeling of channel delay characteristics. After obtaining the delay fluctuation curves for each channel, smoothing and deep feature extraction are required to identify periodic patterns and sudden anomalies. First, a sliding window filtering method is applied to the delay curves. By setting a fixed-width window on the time series, the delay values ​​within the window are weighted and averaged to reduce the impact of random noise and transient interference. The window width is set based on the channel sampling frequency and transmission stability, typically ranging from 10 to 50 sampling points. The filtered curve more clearly shows the regular changes in delay over time. Subsequently, periodic and anomaly features are extracted through a combined time-domain and frequency-domain analysis. Periodic features manifest as regular fluctuations in delay values ​​over time, usually related to changes in communication load or periodic interference sources; sudden delay anomalies are characterized by sharp fluctuations in the curve within a short period, often caused by momentary link congestion or hardware buffer conflicts. By analyzing the amplitude, duration, and frequency of occurrence of the delay curves, a dynamic feature model of delay changes can be established, providing a crucial reference for subsequent delay prediction and compensation mechanisms.

[0035] After extracting the periodic delay variation characteristics and sudden anomaly characteristics, trend prediction is needed to anticipate the channel delay change trend in the future. The prediction process is based on time series analysis of historical delay data, implemented by establishing a delay variation model. First, pattern recognition and trend fitting are performed on the periodic features to extract the period length and rate of change of delay rise and fall, thus establishing a periodic prediction model. For the sudden anomaly, anomaly weight analysis based on probability statistics is used to assess its probability of occurrence and impact range. By superimposing the periodic trend and the anomaly disturbance, a delay prediction sequence for the future period can be generated. This sequence outputs the possible delay level and fluctuation range of each channel in the next few sampling periods, based on time. The prediction sequence not only reflects the overall delay trend but also marks the potential delay peak positions in advance, enabling the system to pre-adjust before delay changes occur, providing a basis for real-time delay compensation, thereby improving the feedforward control capability of multi-channel data transmission. To verify the accuracy of the prediction results, the latest delay values ​​of each channel need to be collected in real time and compared with the predicted values. The collection process maintains the same sampling frequency as the aforementioned delay monitoring to ensure time alignment. The difference between the predicted and actual measured values ​​at each moment is calculated to obtain the delay prediction error. This error reflects the accuracy of the prediction model's response to changes in current channel delay. A small error indicates that the model reflects the channel characteristics well; a persistently large error indicates a nonlinear abrupt change in the channel state, requiring adjustment of the model parameters. Error analysis is not limited to single-point comparisons but also includes mean square error and offset direction evaluation within a time window to distinguish between systematic biases and random errors. By statistically analyzing the prediction errors of multiple channels, it is possible to determine which channels are in a stable transmission state and which channels exhibit delay drift or abnormal trends. The extracted delay prediction error results will serve as an important input for the next step of dynamic compensation, enabling the system to make targeted real-time adjustments based on the delay characteristics of different channels.

[0036] After obtaining the delay prediction error for each channel, error offset inversion and dynamic compensation are required to correct the deviation of the prediction model and maintain time synchronization between multiple channels. First, the magnitude and direction of the error are analyzed to determine whether it is a positive offset (delay exceeding prediction) or a negative offset (delay ahead of prediction). Then, the dynamic delay offset of each channel is deduced based on the offset magnitude, and corresponding compensation parameters are generated. Compensation is achieved by adjusting the data buffer release timing of each channel or the unpacking delay at the receiver, realigning the arrival times of all channels. For channels with large delay offsets, the system can appropriately reduce data scheduling priority or increase transmission advance; while for channels with small offsets, microsecond-level timing fine-tuning is performed to achieve precise synchronization. The dynamic compensation mechanism is automatically executed in each update cycle, updating the compensation parameters immediately when delay characteristics change, thereby ensuring the time consistency and transmission stability of the entire system. The final multi-channel delay compensation parameters can be applied to the transmission control module in real time, enabling low-latency, low-jitter synchronous transmission of multi-channel point cloud signals from the LiDAR in a high-speed interconnect environment.

[0037] In this embodiment, the specific steps of step S5 are as follows: The data transmission information is decoded to extract the decoded electronic cloud data; Calculate the point cloud missing rate, temporal misalignment rate, and geometric distortion of the decoded electrical cloud data; The data block transmission integrity is evaluated based on the point cloud missing rate, temporal misalignment rate, and geometric distortion, resulting in a transmission integrity index. Information loss analysis is performed based on the transmission integrity index, and the encoding parameters are iteratively adjusted to obtain the compression adjustment parameters; The compression adjustment parameters and multi-channel delay compensation parameters are coordinated and controlled to make decisions, and feedforward optimization control is performed.

[0038] In this embodiment, after the multi-channel vehicle-mounted LiDAR data transmission is completed, the receiving end first performs a decoding operation on the received data transmission information to recover the original point cloud signal. The core objective of decoding is to restore the point cloud data that conforms to the spatial topology and temporal order from the hierarchical compression coding structure. First, by reading the channel tags and coding level identifiers embedded in the data block header, the data blocks of each channel are recombined and sorted according to the timestamp information. Subsequently, a corresponding decoding scheme is used to restore the data for each data block, with different levels of coding structures corresponding to different decompression ratios. For example, low-compression ratio inverse quantization and decoding processing is used for data in high-density areas to preserve detail accuracy; while a high-compression ratio fast decoding algorithm is used for low-redundancy areas. During the decoding process, data verification and redundancy repair are required. Data integrity is confirmed by CRC checksum, and forward error correction (FEC) is used to repair slightly lost data segments. The decoded point cloud data is output in frames and temporarily stored in a high-speed cache for subsequent quality evaluation. This process can restore the original coded signal into a point cloud dataset that can be directly used for spatial reconstruction and temporal analysis, providing a basic input for subsequent transmission integrity calculation and compression adjustment.

[0039] After decoding, a quantitative analysis of the integrity and consistency of the point cloud data is required. First, the point cloud missing rate is calculated to measure the proportion of data lost during transmission and decoding. The missing rate is statistically determined by comparing the number of points in the decoded point cloud frame with that in the original sampled frame. Second, the temporal misalignment rate is calculated to reflect the synchronization deviation of the point cloud in the time dimension. This metric is achieved by comparing the consistency between the timestamp sequence of the point cloud frames and the sampling period. If the interval between adjacent frames exceeds a set threshold (e.g., 1 millisecond), it is counted as a temporal misalignment event. Finally, the geometric distortion is evaluated to characterize the deviation of the point cloud in its spatial structure. By calculating the deformation ratio of key geometric features (such as planes, edges, and curved surfaces) in the point cloud, the degree of geometric deviation caused by transmission errors or compression distortion can be quantitatively reflected. These three metrics together constitute the core measurement system for point cloud data quality, comprehensively revealing data distortion, time delay mismatch, or spatial offset problems caused in the signal transmission and reconstruction stages, providing fundamental data support for subsequent transmission integrity assessment.

[0040] The three indicators are standardized to ensure comparability on a unified scale. They are then weighted and aggregated based on their impact on overall data quality. Geometric distortion typically has the highest weight because it directly affects the accuracy of point cloud spatial reconstruction; temporal misalignment rate has the second highest weight, reflecting the synchronization of data frames; and point cloud missing rate reflects signal loss. The weighted combination yields a transmission integrity index, which reflects the proportion of effective information retained and the degree of spatial reconstruction for each data block during transmission. A higher index indicates less difference between the reconstructed point cloud data and the original signal, indicating better transmission quality. The evaluation results are recorded and associated with channel numbers for continuous monitoring of the performance of different transmission paths. This comprehensive evaluation process enables dynamic quantification of multi-channel data quality, serving as a key input for subsequent information loss analysis and compression adjustments.

[0041] After obtaining the transmission integrity index, it is necessary to further analyze the sources of information loss and adaptively adjust the compression coding strategy. The core of information loss analysis is to identify the specific factors that cause the point cloud quality degradation, such as over-compression, signal interference, channel jitter, or delay offset. By comparing the integrity index of each data block with its corresponding coding parameters (including quantization step size, coding level, and data block priority), the impact of compression parameters on data quality can be determined. When the system detects a continuous decline in the integrity index or abnormal fluctuations in local channels, it will initiate a parameter iteration mechanism to adjust the compression ratio, quantization accuracy, and block partitioning strategy. For example, the compression ratio is reduced and the fidelity is increased in point cloud regions of high-complexity scenarios; the compression ratio is appropriately increased in low-redundancy regions to improve bandwidth utilization. After each round of iteration, the system recalculates the transmission integrity to verify the effect of parameter optimization. Through this cyclical feedback method, the coding strategy can be dynamically optimized, enabling the system to achieve a balance between transmission quality and resource efficiency, thereby generating compression adjustment parameters suitable for the current transmission state. After obtaining the compression adjustment parameters and multi-channel delay compensation parameters, coordinated regulation is required to achieve feedforward optimization control of the overall system performance. The core objective of coordinated control is to manage data compression, transmission delay, and channel load status in a unified manner, maintaining an optimal balance of point cloud signals across spatial, temporal, and bandwidth dimensions. First, by comprehensively analyzing the current channel delay compensation parameters and transmission integrity trends, the system's critical load threshold is determined. Then, compression adjustment parameters are input into the decision module and matched with channel delay characteristics to ensure the compression ratio is compatible with the channel delay characteristics. For example, when channel delay increases, the system can reduce the compression ratio in advance to reduce retransmission overhead; when delay is stable, the compression ratio is appropriately increased to improve bandwidth utilization. The feedforward control mechanism optimizes transmission parameters in advance based on predicted delay trends and compression feedback results, avoiding passive responses to network fluctuations. Finally, coordinated control generates global optimization instructions to achieve synchronous transmission and dynamic balance of data quality across multiple channels, ensuring low-distortion, high-consistency real-time transmission of high-density automotive LiDAR signals in high-speed interconnection scenarios.

[0042] In this embodiment, the specific steps for performing information loss analysis based on the transmission integrity index and iteratively adjusting the encoding parameters to obtain the compression adjustment parameters are as follows: Information loss analysis is performed based on the transmission integrity index to identify lost information and the reasons for loss. Based on the lost information, the distribution of information loss areas is classified to obtain high loss rate areas, medium loss rate areas and low loss rate areas; Adjust the compression ratio of the high loss rate region to improve coding redundancy; A balance analysis of compression efficiency and integrity is performed on the region with the highest loss rate, and iterative calculations are performed to output the efficiency-integrity balance point. The latest compression adjustment parameters are determined based on the rate-integrity balance point. Improve the compression ratio of the low loss rate region; Iterate through the above encoding parameter adjustments to output the compression adjustment parameters.

[0043] In this embodiment, after obtaining the transmission integrity index of each data block, a systematic analysis of the specific content and causes of information loss is required. First, the integrity index is compared with the original point cloud reference model to identify point cloud regions exhibiting anomalies in spatial, temporal, or intensity dimensions. These anomalies mainly manifest as sudden changes in point cloud density, spatial discontinuities, temporal jumps, or discontinuous light intensity. Through comparative analysis, the type of information loss can be preliminarily determined, including point cloud data packet loss, frame overlap due to temporal misalignment, and detail loss due to compression distortion. Subsequently, based on the data transmission path and channel delay records, the possible causes of information loss are traced and analyzed. For example, if the lost regions are concentrated within data blocks of a certain channel and accompanied by high latency fluctuations, it can be determined that it is caused by link jitter or buffer congestion; if a high proportion of geometric distortion appears in high-compression ratio data blocks, it indicates that excessive compression parameters are causing distortion. Through this process, not only can the specific location of lost data be identified, but the proportion of various loss causes in the overall transmission can also be quantified, thus providing a targeted basis for subsequent partitioning adjustments and encoding optimization. After identifying information loss, the lost regions need to be classified spatially and temporally to distinguish point cloud regions with different loss rates. First, each frame of point cloud is divided into several small units according to a spatial grid. Each unit contains a fixed number of points (e.g., 500 to 1000 points), and its point cloud loss rate is calculated. By statistically analyzing the proportion of lost points in each unit, a regional loss distribution map can be formed. Then, based on the statistical results, grading thresholds are set. For example, regions with a loss rate exceeding 10% are defined as high loss rate regions, those between 3% and 10% as medium loss rate regions, and those below 3% as low loss rate regions. After classification, the system labels and visualizes each region to intuitively reflect the differences in the spatial distribution of lost information. High loss rate regions are typically concentrated at the edge of the field of view, in areas with severe occlusion, or in areas with unstable transmission channels; medium loss rate regions are more dispersed and are often related to critical bandwidth areas or coding fluctuations; low loss rate regions are mostly located within the main field of view. Through this distribution classification, differentiated compression strategies can be formulated for different regions, achieving precise optimization and targeted signal enhancement.

[0044] For identified high-loss-rate regions, strategies to reduce compression ratio and increase coding redundancy are needed to enhance data resilience and recovery capabilities during transmission. Specifically, the point cloud data of these regions is first extracted separately from the original hierarchical coding structure, and its compression parameters are adjusted. By reducing the quantization step size, increasing sampling accuracy, and adding more inter-frame prediction reference points, the signal fidelity of these regions is improved. For example, reducing the original compression ratio from 1:10 to 1:6 allows each point cloud block to contain more original information, thereby enhancing subsequent error recovery capabilities. Simultaneously, a forward error correction (FEC) mechanism can be introduced during coding to add redundant check codes to high-loss regions, preventing irrecoverable information loss due to single-frame packet loss. After adjustment, the encoded stream for these regions is regenerated, and its priority is updated in the compression parameter table to ensure higher bandwidth allocation in subsequent transmission scheduling. In this way, high-loss-rate regions achieve stronger data robustness and error resistance, significantly reducing point cloud missing data and geometric distortion. The medium loss rate region typically lies within a dynamic balance between compression ratio and data integrity, thus requiring iterative calculations to determine the optimal equilibrium point. First, multiple compression experiments are conducted on the point cloud data in this region, each using different compression ratio parameters, recording the corresponding transmission integrity index and compression efficiency. By comparing multiple results, a functional relationship curve between compression ratio and integrity can be obtained. Subsequently, numerical iterative methods (such as multi-point fitting search or interval convergence methods) are used to determine the optimal intersection point, i.e., the efficiency-integrity balance point. This balance point represents the highest achievable compression ratio while maintaining the clarity of the point cloud structure. For example, if the integrity index decreases by less than 5% at a compression ratio of 1:8, then this is the optimal point. This process can be repeated multiple times until the balance point stabilizes and no longer fluctuates. The obtained balance parameters will serve as a new coding benchmark for the medium loss rate region, enabling the system to maintain the spatial continuity and reconstruction accuracy of the point cloud without significantly increasing bandwidth consumption, achieving the optimal match between transmission performance and signal quality.

[0045] After determining the efficiency-integrity balance point in the medium loss rate region, an updated compression adjustment parameter set needs to be generated accordingly. This parameter set will comprehensively consider the optimization needs of high, medium, and low loss rate regions, forming a hierarchical coding strategy. First, the low compression ratio parameter of the high loss rate region is merged with the balanced compression ratio of the medium loss rate region, and a dynamic adjustment range is set. For the medium loss rate region, the balanced point compression ratio is used as the center value, allowing automatic fine-tuning within a certain range based on channel status to adapt to different network conditions. Then, these parameters are stored in the coding control table along with the region identifiers for the coding module to call during real-time processing. By establishing parameter mapping relationships, the system can automatically apply different compression configurations according to the region type of the point cloud during the coding stage, achieving flexible adaptive compression. The new compression adjustment parameters not only reduce the loss of detail caused by high compression but also maintain transmission stability under limited bandwidth conditions, providing a basis for subsequent overall compression iterations. In the low loss rate region, due to the stable transmission of point cloud signals, small latency fluctuations, and low bit error rate, the compression ratio can be appropriately increased to release more bandwidth resources. Specifically, a higher quantization step size and data block aggregation strategy are adopted in this region to improve compression efficiency by reducing redundant information. For example, the compression ratio is increased from 1:10 to 1:14 to reduce the transmission bandwidth occupied by this part of the data. At the same time, the temporal consistency of the inter-frame prediction model is maintained to prevent temporal drift or distortion caused by high compression ratio. For channels with stable signal quality, a variable bit rate (VBR) strategy can be further introduced to dynamically adjust the coding rate according to the transmission status, thereby automatically increasing the compression ratio when network resources are scarce and restoring the original accuracy when the channel is idle. This method can improve data throughput and bandwidth utilization while ensuring overall system stability. By optimizing the compression ratio in low-loss-rate regions, dynamic reallocation of system resources can be achieved, reserving higher-quality data bandwidth space for high-loss and medium-loss regions. After completing the hierarchical compression strategy adjustment for high, medium, and low-loss-rate regions, the system enters the iterative optimization phase. The core of the iterative process is to achieve continuous adaptive optimization of the coding configuration through cyclic evaluation and parameter updates. First, in each iteration, the transmission integrity index of each region is recalculated and compared with the results of the previous round to determine whether the adjustment effect has met expectations. If the integrity of high-loss regions significantly improves and the total system bandwidth does not exceed the limit, the current redundancy is maintained; if the integrity of medium-loss regions changes only slightly, their balance point parameters are further fine-tuned; if the compression ratio improvement of low-loss regions does not significantly affect the overall quality, the compression ratio is further increased. Through this cyclical process, the system can gradually approach the globally optimal compression configuration state. After each iteration, a new set of compression adjustment parameters is generated and input into the encoding module for execution, ensuring that the entire vehicle-mounted LiDAR signal transmission process is always in a dynamically optimal compression-quality balance state. The final output compression adjustment parameters not only improve transmission efficiency but also maximize the spatial accuracy and temporal consistency of point cloud data.

[0046] In this embodiment, the specific steps for coordinating and controlling the compression adjustment parameters and multi-channel delay compensation parameters to perform feedforward optimization control are as follows: Identify vehicle motion status and environmental perception results; The vehicle motion state includes real-time speed, acceleration, and steering angle rate; The environmental perception results include road density ahead, obstacle distribution, and traffic flow characteristics; Based on the environmental perception results, point cloud density change analysis is performed to extract the road point cloud density trend; Based on the vehicle motion state, the data transmission demand is predicted according to the road point cloud density trend, and a data transmission demand prediction value is generated. Coordinated control decisions are made on compression adjustment parameters and multi-channel delay compensation parameters to generate a coordinated transmission control strategy.

[0047] In this embodiment, the high-density vehicle-mounted LiDAR signal transmission system first requires accurate identification of the vehicle's motion state and the perception results of its surrounding environment, providing input for subsequent data acquisition density analysis and transmission prediction. The vehicle's motion state is jointly described by its attitude and dynamic characteristics, mainly including real-time speed, longitudinal acceleration, and steering angular rate. Through joint sampling by the vehicle-mounted inertial measurement unit (IMU), wheel speed sensors, and steering encoder, continuous motion parameters can be acquired with millisecond-level time accuracy. Simultaneously, the environmental perception results are jointly output by forward, lateral, and circumferential LiDAR, millimeter-wave radar, and cameras, including information such as road geometry, road density ahead, obstacle distribution, and traffic flow characteristics. For example, when the road density ahead is high and obstacles are concentrated, the point cloud acquisition system needs to increase the sampling frequency and spatial resolution to ensure the integrity of scene reconstruction. The system uses a multi-sensor fusion algorithm to synchronize the vehicle's motion state and environmental information in a unified coordinate system in both time and space, thereby forming a motion-environment joint description corresponding to each moment, providing complete input for dynamic point cloud sampling and data transmission strategy adjustment. During vehicle operation, real-time motion directly affects the spatiotemporal sampling characteristics of point cloud data. Therefore, high-precision monitoring and dynamic updating of speed, acceleration, and steering angular rate are required. Real-time speed is typically obtained through the fusion of wheel speed sensors and GNSS speed information to eliminate biases caused by a single sensor source. Longitudinal and lateral accelerations are measured in real-time by the IMU's accelerometer, with sampling frequencies exceeding 200Hz to meet the dynamic response requirements in high-speed driving scenarios. Steering angular rate is acquired through steering system angle sensors to characterize the vehicle's turning radius and attitude change trends. These three parameters, after time synchronization and filtering correction, are input into the vehicle state estimation module to describe the vehicle's transient dynamic behavior. When vehicle speed changes or acceleration fluctuates significantly, the system identifies it as a dynamic state enhancement phase, requiring an increase in point cloud sampling density to cope with rapid changes in the field of view. When the steering angular rate increases significantly, a lateral point cloud sampling optimization mechanism is triggered to ensure the spatial continuity of the scene in front of the vehicle. In this way, motion state parameters not only reflect the vehicle's motion trend but also provide prior information for point cloud data acquisition and transmission scheduling.

[0048] Environmental perception results are crucial for describing the dynamic characteristics of the external world surrounding the vehicle, directly impacting the sampling strategy and transmission priority of point cloud data. The road density ahead is calculated using the density of ground echo points in the LiDAR scan point cloud, reflecting the complexity of the road structure. Obstacle distribution is identified by spatial mapping of point cloud clustering and target recognition results, determining the position and quantity of objects such as vehicles, pedestrians, and guardrails in three-dimensional space. Traffic flow characteristics are derived from the temporal continuity analysis of multi-frame point clouds, including the relative speed, direction of movement, and consistency of flow direction of targets. The system projects these three types of environmental indicators onto the vehicle coordinate system and dynamically fuses them along the time axis to generate real-time environmental complexity evaluation results. When road density increases or obstacle distribution becomes concentrated, it means the point cloud data will be denser and more frequently changing, requiring more bandwidth resources to be allocated during encoding and transmission. Conversely, when traffic flow characteristics are stable and environmental changes are minimal, the sampling rate can be appropriately reduced and redundancy compressed, thus saving transmission overhead. Through dynamic identification of environmental features, the system achieves feedforward adjustment of data flow, laying the foundation for subsequent transmission demand prediction. After obtaining the environmental perception results, the system needs to analyze the trend of point cloud acquisition density changing with the scene. First, by statistically analyzing the changes in the number of point clouds in multiple consecutive sampling frames, the point cloud density distribution per unit time and per unit space is calculated. For example, in complex traffic scenes, the point cloud density may reach more than 250,000 points per second, while in open roads it drops to less than 100,000 points per second. By comparing the density change rates between different frames, the system identifies trend areas of increasing, decreasing, or stable density. Subsequently, combining the dynamic characteristics of road structure and obstacles, a road point cloud density trend curve is generated to reflect the temporal variation characteristics of the point cloud generation rate. When a continuous increase in density trend is detected, it means that the amount of sampled data will increase significantly, and the requirements for transmission channel bandwidth and latency control will increase accordingly. Through this trend analysis, the system can predict the point cloud data growth rate in the near future and provide a predictive basis for transmission load scheduling. Finally, the point cloud density trend is passed as an input parameter to the transmission demand prediction module to establish a dynamic mapping relationship between data volume and vehicle motion state.

[0049] By combining vehicle motion state with road point cloud density trends, the system can predict point cloud data transmission demand over a future period. First, based on the vehicle's speed change rate and acceleration curve, the system calculates the field-of-view change rate within the future sampling period. Then, it couples this rate with the current road point cloud density trend to predict the amount of data generated per unit time. For example, when a vehicle is traveling at high speed and its acceleration increases, the forward field of view changes faster, the point cloud update rate increases, and the predicted transmission demand rises accordingly. When the vehicle's steering angle rate increases, the lateral field of view expands, requiring additional data bandwidth to support synchronous sampling of the lateral point cloud. Through multidimensional fitting of historical sampling trends and motion states, the system generates a continuous sequence of predicted transmission demand, indicating future transmission bandwidth, latency tolerance, and buffer allocation strategies. The predicted values ​​not only reflect data volume trends but also include estimates of system load and communication link capabilities, enabling feedforward adjustments to the transmission strategy to avoid sudden data congestion or latency accumulation. The system compares the predicted transmission demand with the current channel bandwidth status to determine if the transmission pressure exceeds a threshold. If the predicted load increases, the system prioritizes adjusting compression parameters, reducing the encoding precision of non-critical areas, and optimizing the channel scheduling order. Subsequently, by combining delay compensation parameters, the system balances the time alignment between channels, prioritizing the transmission of high-priority data blocks (such as high-density forward point clouds) while delaying the scheduling of low-priority data blocks. When environmental complexity decreases or the vehicle enters a stable driving phase, the system can automatically restore the original compression ratio and reduce the delay compensation amount to save resource consumption. The entire collaborative decision-making process operates in a closed-loop manner: transmission demand prediction drives feedforward control, and real-time feedback corrects the compensation parameters, enabling the system to achieve an optimal balance between bandwidth utilization, transmission delay, and data integrity. The final generated collaborative transmission control strategy can adapt to different driving and environmental scenarios, achieving low-latency and highly robust interconnected transmission of high-density LiDAR signals.

[0050] In this embodiment, a high-density interconnect signal transmission system is provided for performing the high-density interconnect signal transmission method described above, including: The data segmentation unit is used to acquire the original point cloud frame signal; it dynamically adapts and segments the acquired original point cloud frame signal to generate multiple signal data blocks. A quality quantization unit is used to identify the vehicle-mounted physical transmission channel; to perform real-time quality quantization on the vehicle-mounted physical transmission channel and generate a transmission quality coefficient for each channel; A transmission matching unit is used to perform transmission matching on multiple signal data blocks based on the transmission quality coefficient, and send the multiple signal data blocks to the signal receiving end; The delay compensation unit is used to collect and receive data transmission information from the signal receiver and perform delay compensation on the data transmission information to obtain multi-channel delay compensation parameters. The transmission control unit is used to control signal transmission based on the multi-channel delay compensation parameters.

[0051] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, 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 the equivalents of the application are intended to be included within the invention.

[0052] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein are implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A high-density interconnect signal transmission method, characterized in that, Includes the following steps: Step S1: Acquire the original point cloud frame signal; dynamically adapt and segment the acquired original point cloud frame signal to generate multiple signal data blocks; Step S2: Identify the vehicle's physical transmission channel; The vehicle-mounted physical transmission channels are subjected to real-time quality quantization to generate a transmission quality coefficient for each channel. Step S3: Based on the transmission quality coefficient, perform transmission matching on multiple signal data blocks and send the multiple signal data blocks to the signal receiving end; Step S4: Collect the received data transmission information based on the signal receiver; Delay compensation is performed on the data transmission information to obtain multi-channel delay compensation parameters; Step S5: Perform signal transmission control based on the multi-channel delay compensation parameters.

2. The high-density interconnect signal transmission method according to claim 1, characterized in that, The specific steps of step S1 are as follows: Original point cloud frame signals are acquired based on a multi-line lidar scanning unit; Spatial density analysis is performed on the original point cloud frame signal to divide the point cloud signal into high-density regions and low-redundancy regions. The scene information of the high-density region and the low-redundancy region are analyzed to generate scene information complexity. Adaptive compression coding is performed based on the complexity of scene information to generate a hierarchical coding sequence; The hierarchical coded sequence is dynamically adapted and segmented to generate multiple signal data blocks.

3. The high-density interconnect signal transmission method according to claim 1, characterized in that, The specific steps of step S2 are as follows: Identify the vehicle's physical transmission channels and monitor real-time channel status information; Based on the real-time channel status information, bandwidth utilization, latency jitter, packet loss rate, and link stability are calculated to generate channel transmission status performance indicators. Calculate the real-time available capacity and transmission reliability coefficient of the vehicle-mounted physical transmission channel; Based on the channel transmission status performance indicators, the real-time available capacity, and the transmission reliability coefficient, the channel real-time quality is quantified to generate the transmission quality coefficient for each channel.

4. The high-density interconnect signal transmission method according to claim 1, characterized in that, Step S3 is as follows: Calculate the size and priority of the plurality of signal data blocks; Based on the transmission quality coefficient, adaptive transmission matching is performed on the size and priority to obtain the optimal transmission channel; Based on the optimal transmission channel, a channel tag is generated, and information is embedded in the data block header to send multiple signal data blocks to the signal receiving end.

5. The high-density interconnect signal transmission method according to claim 1, characterized in that, The specific steps of step S4 are as follows: Based on the data transmission information received by the signal receiver; Calculate the reception timestamp of each signal block based on the data transmission information; Extract the transmission timestamp of the signal data block; The transmission delay of the signal block is calculated based on the sending timestamp and receiving timestamp, and the transmission delay value of each signal block is extracted. Extract the header information of the signal block, trace the transmission channel, and mark the channel delay based on the transmission delay value to obtain the delay value of each channel; Adaptive delay compensation is performed based on the delay values ​​of each channel to obtain multi-channel delay compensation parameters.

6. The high-density interconnect signal transmission method according to claim 5, characterized in that, The specific steps for obtaining multi-channel delay compensation parameters through adaptive delay compensation based on the delay values ​​of each channel are as follows: Delay fluctuations at multiple time points are calculated for the delay values ​​of each channel to obtain the delay fluctuation curves for each channel; Sliding window filtering and deep analysis are performed on the delay fluctuation curves of each channel to generate periodic delay variation characteristics and sudden delay anomaly characteristics; Based on the periodic delay variation characteristics and sudden delay anomaly characteristics, the delay variation trend is predicted to obtain the delay prediction sequence for future time periods; The latest delay values ​​of each channel are collected in real time; the delay prediction error of the delay prediction sequence is calculated based on the latest delay values ​​of each channel, and the delay prediction error of multiple channels is extracted; Based on the delay prediction error, the error offset amplitude is inverted and dynamic compensation is performed to obtain multi-channel delay compensation parameters.

7. The high-density interconnect signal transmission method according to claim 1, characterized in that, The specific steps of step S5 are as follows: The data transmission information is decoded to extract the decoded electronic cloud data; Calculate the point cloud missing rate, temporal misalignment rate, and geometric distortion of the decoded electrical cloud data; The data block transmission integrity is evaluated based on the point cloud missing rate, temporal misalignment rate, and geometric distortion, resulting in a transmission integrity index. Information loss analysis is performed based on the transmission integrity index, and the encoding parameters are iteratively adjusted to obtain the compression adjustment parameters; The compression adjustment parameters and multi-channel delay compensation parameters are coordinated and controlled to make decisions, and feedforward optimization control is performed.

8. The high-density interconnect signal transmission method according to claim 7, characterized in that, The specific steps for performing information loss analysis based on the transmission integrity index and iteratively adjusting the encoding parameters to obtain the compression adjustment parameters are as follows: Information loss analysis is performed based on the transmission integrity index to identify lost information and the reasons for loss. Based on the lost information, the distribution of information loss areas is classified to obtain high loss rate areas, medium loss rate areas and low loss rate areas; Adjust the compression ratio of the high loss rate region to improve coding redundancy; A balance analysis of compression efficiency and integrity is performed on the region with the highest loss rate, and iterative calculations are performed to output the efficiency-integrity balance point. The latest compression adjustment parameters are determined based on the rate-integrity balance point. Improve the compression ratio of the low loss rate region; Iterate through the above encoding parameter adjustments to output the compression adjustment parameters.

9. The high-density interconnect signal transmission method according to claim 7, characterized in that, The specific steps for coordinating and controlling the compression adjustment parameters and multi-channel delay compensation parameters to perform feedforward optimization control are as follows: Identify vehicle motion status and environmental perception results; The vehicle motion state includes real-time speed, acceleration, and steering angle rate; The environmental perception results include road density ahead, obstacle distribution, and traffic flow characteristics; Based on the environmental perception results, point cloud density change analysis is performed to extract the road point cloud density trend; Based on the vehicle motion state, the data transmission demand is predicted according to the road point cloud density trend, and a data transmission demand prediction value is generated. Coordinated control decisions are made on compression adjustment parameters and multi-channel delay compensation parameters to generate a coordinated transmission control strategy.

10. A high-density interconnect signal transmission system, characterized in that, For performing the high-density interconnect signal transmission method as described in claim 1, comprising: The data segmentation unit is used to acquire the original point cloud frame signal; it dynamically adapts and segments the acquired original point cloud frame signal to generate multiple signal data blocks. A quality quantization unit is used to identify the vehicle-mounted physical transmission channel; to perform real-time quality quantization on the vehicle-mounted physical transmission channel and generate a transmission quality coefficient for each channel; A transmission matching unit is used to perform transmission matching on multiple signal data blocks based on the transmission quality coefficient, and send the multiple signal data blocks to the signal receiving end; The delay compensation unit is used to collect and receive data transmission information from the signal receiver and perform delay compensation on the data transmission information to obtain multi-channel delay compensation parameters. The transmission control unit is used to control signal transmission based on the multi-channel delay compensation parameters.

Images