An intelligent warehouse heterogeneous data communication system based on edge cloud cooperation

Abstract

The application discloses a kind of wisdom warehousing heterogeneous data communication systems based on edge cloud cooperation, it is related to wisdom warehousing communication technical field, specifically includes: data access module, feature extraction module, classification coding module, preemption sending module, receiving buffer module and window reorganization module;Real-time acquisition is sent in the original message of heterogeneous equipment in warehousing field, and carries out protocol-independent stripping and analysis, generates and embeds timing semantic label head;The original message is divided into very short control frame and long message data, the long message data is based on wireless channel coherence time Dynamic slicing is carried out, and forward error correction erasure code is applied to carry out redundancy coding, and generate fragment microframe;Very short control frame and fragment microframe are carried out time domain interleaving and space scattering transmission, and preemption mechanism is triggered when very short control frame arrives;The short control frame and fragment microframe are received;The short control frame is directly transmitted, and the fragment microframe is self-healing reorganization, restores original message.

Description

Technical Field

[0001] This invention relates to the field of smart warehouse communication technology, specifically to a smart warehouse heterogeneous data communication system based on edge-cloud collaboration. Background Technology

[0002] Modern smart warehousing systems deploy a large number of automated guided vehicles (AGVs), stacker cranes, automated shuttles, industrial robots, and various sensors and image acquisition devices. These devices often interact with each other based on different underlying communication protocols, forming a typical heterogeneous data communication environment. Meanwhile, to achieve global scheduling and real-time control of the warehousing system, an edge-cloud collaborative architecture consisting of edge computing gateways and cloud servers is typically employed.

[0003] In actual warehousing environments, especially in high-density automated high-bay warehouses, communication systems face the following severe engineering challenges:

[0004] (1) Densely arranged steel shelves create severe electromagnetic shielding and multipath reflection for radio signals, resulting in drastic fluctuations in wireless channel quality, extremely short coherence time, and high packet loss rate.

[0005] (2) The warehouse contains both critical control instructions with very small data volume but extremely high real-time requirements, and long message data with large data volume.

[0006] (3) Traditional edge gateways use first-in-first-out or simple priority queues. When large image frames fill the channel for transmission, if the channel fading causes transmission failure, the global retransmission of large packets will occupy the channel for a long time, causing subsequent very short emergency control commands to be severely delayed or even lost; while if large packets are directly discarded, it will cause cloud artificial intelligence recognition to be interrupted.

[0007] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0008] The purpose of this invention is to provide a smart warehouse heterogeneous data communication system based on edge-cloud collaboration, which enables millisecond-level preemptive transmission of short control frames in warehouse environments with high-density metal interference and drastic changes in channel quality, and ensures that long message data can be losslessly reassembled at the receiving end without global retransmission, thus solving the problems mentioned in the background technology.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a smart warehouse heterogeneous data communication system based on edge-cloud collaboration, comprising an edge data access gateway located at the warehouse site and a cloud front-end receiving node located at the upper-level control center, specifically including: a data access module, a feature extraction module, a classification and encoding module, a preemptive sending module, a receiving buffer module, and a window reassembly module;

[0010] The edge data access gateway includes:

[0011] Data access module: Real-time collection of raw messages sent by heterogeneous devices in the warehouse;

[0012] Feature extraction module: performs protocol-independent stripping and analysis on the original message, and generates and embeds a time-series semantic tag header;

[0013] Classification and coding module: Divides the original message into very short control frames and long message data, dynamically slices the long message data based on the coherence time of the wireless channel, and applies forward error correction erasure code for redundant coding to generate fragmented micro-frames;

[0014] Preemptive transmission module: The extremely short control frame and fragmented microframes are transmitted through multiple physical channels with temporal interleaving and spatial fragmentation, and a preemptive mechanism is triggered when the extremely short control frame arrives.

[0015] The cloud-based front-end receiving node includes:

[0016] Receive buffer module: Receives the short control frame and fragmented microframes;

[0017] Window reassembly module: Directly transmits the short control frame and performs self-healing reassembly of the fragmented microframes through a forward error correction decoding matrix to restore the original message.

[0018] Preferably, as a preferred embodiment of the edge-cloud collaborative smart warehousing heterogeneous data communication system of the present invention, wherein:

[0019] Through a multi-threaded concurrency mechanism, the system performs real-time status polling and interrupt event monitoring of various physical communication interfaces deployed in the warehouse, and receives raw byte streams from heterogeneous devices in the warehouse.

[0020] The raw byte stream is loaded into the access ring buffer. Based on the frame structure characteristics and timing attributes of different underlying communication protocols, physical frame boundaries are trimmed and the original message is extracted. Specifically, this includes:

[0021] For Controller Area Network (CAN) protocol: Read the CAN frame structure in the access ring buffer, parse the identifier field and data length code, extract the payload of the corresponding number of bytes, and generate CAN frame message;

[0022] For asynchronous serial bus data transmission protocol: determine frame boundaries based on bus silent time interval, match device address bits and function codes, perform integrity verification through cyclic redundancy check code, and extract bus data transmission protocol messages;

[0023] For the Unified Process Control Architecture Protocol (UCAP) and Message Queuing Telemetry Transmission Protocol (MQTP): the length identifier field of the application layer protocol header in the UCTP payload is extracted using the sliding window algorithm, and the complete UCAP and MQTP messages are then truncated.

[0024] The complete, trimmed original message undergoes protocol feature parsing, and communication and service attributes are extracted. Data type pre-identification and labeling are then performed, specifically including:

[0025] If the predicted control command is related to equipment emergency stop, collision warning, or arrival confirmation, it is marked as a data type of extremely short control frame;

[0026] If the data is pre-determined to involve visual barcode images, device operation logs, or high-frequency sensor status, it is marked as a long message data type;

[0027] The payload data of the original message is kept unaltered, and the checksum and link layer encapsulation information related to the underlying physical link are stripped away to generate a standardized message to be processed, which is then pushed into the buffer queue.

[0028] Preferably, as a preferred embodiment of the edge-cloud collaborative smart warehousing heterogeneous data communication system of the present invention, wherein:

[0029] Read standardized pending messages in the buffer queue in real time and extract the original payload, protocol type identifier and original receive timestamp;

[0030] Based on the protocol type identifier and the service attribute code in the original payload, calculate the maximum tolerable latency threshold and the disruption tolerance, specifically including:

[0031] If the standardized message to be processed is marked as a data type of extremely short control frame, then its maximum tolerable delay time threshold is set to 1 millisecond;

[0032] If the standardized message to be processed is marked as a long message data type, then its maximum tolerable delay time threshold is set to 500 milliseconds;

[0033] If the standardized message to be processed is marked as a data type of very short control frame, then the destruction tolerance flag is set to an indestructible state.

[0034] If the standardized message to be processed is marked as a data type of long message data, then the destruction tolerance flag is set to a destructible state;

[0035] The clock source and incrementing counter of the edge data access gateway are invoked to generate a global sequence reassembly identifier, and a semantic tag header of fixed size of four bytes is assembled according to a preset bit format.

[0036] The semantic tag header specifically includes:

[0037] The first byte to the second byte: stores the global sequence reassembly identifier;

[0038] The third byte stores the maximum tolerable latency threshold;

[0039] The fourth byte stores a combination of the violation tolerance flag and the protocol type flag.

[0040] The semantic tag header is encapsulated at the very beginning of the original payload of the standardized message to be processed, generating a tagged message, which is then pushed into the processing queue.

[0041] Preferably, as a preferred embodiment of the edge-cloud collaborative smart warehousing heterogeneous data communication system of the present invention, wherein:

[0042] Real-time reception of tagged packets, parsing of the semantic tag header, reading of the maximum tolerable delay time threshold and the violation tolerance flag, and execution of dual-channel routing and traffic splitting, specifically including:

[0043] Extremely short control frame diversion: If the maximum tolerable delay time threshold is lower than the preset low latency threshold, and the damage tolerance is marked as non-destructible, the marked message is determined to be an extremely short control frame and diverted to the high-speed transparent transmission channel.

[0044] Long message data diversion: If the maximum tolerable delay time threshold is higher than the preset low delay threshold, or the damage tolerance is marked as destructible, the marked message is determined to be long message data and diverted to the slice encoding channel.

[0045] The high-speed transparent transmission channel will push tagged messages that are determined to be extremely short control frames into the priority transmission queue.

[0046] The slice coding channel obtains the coherence time measurement value of the current physical wireless channel, and calculates the maximum number of bytes transmitted in a single frame of the physical wireless channel based on the ratio of the current physical wireless channel transmission rate to the preset safety margin of the coherence time measurement value.

[0047] The maximum number of bytes is determined by subtracting the size of the four-byte semantic tag header and the forward error correction verification information from the maximum number of bytes.

[0048] The original payload of the long message data is continuously cut based on the maximum safe payload size of a single frame to generate K ordered original slices;

[0049] The original slice is subjected to forward error correction and erase code redundancy encoding to generate N encoded slices, and a four-byte semantic tag header corresponding to the long message data is copied and embedded in each of the encoded slices.

[0050] Forward error correction verification information is appended to the end of each encoded slice to construct N fragmented microframes, which are then pushed into the normal transmission queue in sequence.

[0051] Preferably, as a preferred embodiment of the edge-cloud collaborative smart warehousing heterogeneous data communication system of the present invention, wherein:

[0052] It monitors the preemptive priority transmission queue and the normal transmission queue in real time, and simultaneously senses the channel status and signal strength of multiple available independent physical wireless channels.

[0053] When a very short control frame is stored in the preemptive priority transmission queue, a hardware-level preemption interrupt is triggered, specifically including:

[0054] Link physical interruption: If any physical wireless channel is currently transmitting fragmented microframes, immediately send an abort signal to the corresponding physical layer controller to interrupt the transmission;

[0055] Discard interrupted microframes: Discard interrupted microframe fragments without re-entering the queue, and use forward error correction erasure codes at the receiving end to decode and fill in the gaps;

[0056] Dequeueing of extremely short control frames: The extremely short control frames are extracted and dequeued from the priority transmission queue;

[0057] The outgoing very short control frame is copied into multiple copies and transmitted in parallel simultaneously through multiple independent physical wireless channels;

[0058] When the priority transmission queue is empty and the normal transmission queue is not empty, the scheduling process for fragmented microframes is executed, specifically including:

[0059] The fragmented microframes corresponding to the same long message data are dynamically allocated to multiple independent physical wireless channels according to the polling strategy, and adjacent fragmented microframes are transmitted through different physical channels.

[0060] The fragmented micro-frames corresponding to different long message data are time-domain interleaved and rearranged, and then sent concurrently through the corresponding physical wireless channels.

[0061] After each fragment of a microframe is sent, the normal transmission queue pointer is updated until the normal transmission queue is empty or is preempted and interrupted.

[0062] After the very short control frame is sent, the preemption interruption flag is cleared and control of the physical wireless channel is released. The scheduling and processing of fragmented microframes is automatically resumed, and the fragmented microframes in the normal transmission queue are scattered, scheduled, and sent concurrently.

[0063] Preferably, as a preferred embodiment of the edge-cloud collaborative smart warehouse heterogeneous data communication system of the present invention, wherein:

[0064] It can monitor multiple independent physical wireless channels in real time, receive wireless signals and unpack and extract data frames;

[0065] Perform low-level physical verification on the data frame, calculate the cyclic redundancy check (CRC) code of the data frame, and compare it with the check code carried at the end of the frame:

[0066] If the verification fails, the data frame is discarded.

[0067] If the verification is successful, the data frame is determined to be a valid data frame.

[0068] Parse the semantic marker header of the first four bytes of the valid data frame, read the maximum tolerable latency threshold and the violation tolerance flag, and perform dual-channel routing splitting, specifically including:

[0069] Extremely short control frame offloading: If the maximum tolerable delay time threshold is lower than the preset low latency threshold and the destruction tolerance is marked as non-destructible, the valid data frame is determined to be an extremely short control frame, and the global sequence reassembly identifier in its semantic tag header is read.

[0070] If the very short control frame is the first frame to arrive in a multi-channel transmission, push it into the very short control frame fast buffer;

[0071] If the extremely short control frame is a subsequently arriving duplicate frame, it is filtered and discarded.

[0072] Fragmented microframe routing: If the maximum tolerable latency threshold is higher than the preset low latency threshold, or the destruction tolerance is marked as destructible, the valid data frame is determined to be a fragmented microframe. Then, the global sequence reassembly identifier in its semantic tag header is read, classified according to the corresponding long message data, and pushed into the corresponding storage unit in the fragmented microframe circular buffer.

[0073] Preferably, as a preferred embodiment of the edge-cloud collaborative smart warehouse heterogeneous data communication system of the present invention, wherein:

[0074] Initialize the lock-free sliding window buffer pool based on atomic operations, and monitor the status of the ultra-short control frame high-speed buffer and the fragmented micro-frame circular buffer in real time.

[0075] Among them, the left boundary pointer of the lockless sliding window buffer pool points to the global sequence reassembly identifier of the earliest incomplete long message data currently waiting to be reassembled, and the right boundary pointer is dynamically maintained according to the preset window span;

[0076] When an extremely short control frame is stored in the extremely short control frame fast buffer, the extremely short control frame is dequeued using a lock-free queue read instruction, and the following processing is performed:

[0077] Remove the four-byte semantic tag header at the very beginning of the extremely short control frame to recover the original payload;

[0078] The original payload is submitted transparently to the cloud scheduling and control engine;

[0079] When a fragmented microframe is stored in the fragmented microframe circular buffer, the global sequence reassembly identifier and slice number in the semantic tag header of the fragmented microframe are read, and it is determined whether the global sequence reassembly identifier falls within the left and right boundaries of the current unlocked sliding window buffer pool.

[0080] If the fragmented microframe falls within the window boundary, the atomic swap instruction is used to store the fragmented microframe into the data matrix slot of the corresponding global sequence recombination identifier in the unlocked sliding window buffer pool;

[0081] If it falls behind the left edge of the window, the fragmented microframe is discarded;

[0082] For long message data corresponding to a global sequence reassembly identifier in the unlocked sliding window buffer pool, the number of received valid fragmented microframes is counted in real time, and the number of valid fragmented microframes is compared with the preset forward error correction decoding threshold.

[0083] When the number of effective fragmented microframes is greater than the preset forward error correction decoding threshold, the forward error correction decoding matrix is ​​invoked to perform algebraic operations on the received fragmented microframes and inverse decode to recover the original slice.

[0084] Perform reassembly and memory release on the original slices recovered from the inversion decoding, specifically including:

[0085] The original slices are assembled sequentially according to their slice numbers, and the forward error correction and verification information appended at the end is removed to reconstruct and restore the original long message data.

[0086] The restored original long message data is submitted to the cloud, and the left boundary pointer of the lockless sliding window buffer pool is smoothly moved forward by the atomic instructions of comparison and exchange, releasing the memory space of the data matrix slot that has been reorganized.

[0087] On the other hand, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements the steps of an edge-cloud collaborative smart warehouse heterogeneous data communication system as described above.

[0088] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements the steps of an edge-cloud collaborative smart warehouse heterogeneous data communication system as described above.

[0089] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0090] (1) Through hardware-level preemption interruption and physical link real-time interruption mechanism, extremely short control frames can be stored in the priority transmission queue, the transmitted fragmented micro-frames can be physically interrupted immediately, and they can be copied into multiple copies and sent in parallel on multiple independent physical wireless channels to ensure that key control commands can still arrive reliably when a single channel is blocked, thus achieving sub-millisecond deterministic ultra-low latency transmission.

[0091] (2) The long message is divided into original slices with a slice duration shorter than the channel coherence time and encoded into fragmented microframes. Fragmented microframes that are interrupted or lost due to channel interference are directly discarded and do not need to be retransmitted. The cloud front-end receiving node only needs to receive any valid microframe to decode the original long message through algebraic operation self-healing inversion, thus completely getting rid of the vicious consumption of channel bandwidth by the end-to-end retransmission mechanism.

[0092] (3) By using the spatial scattering and temporal staggered scheduling strategy, the fragmented microframes corresponding to the same long message are dynamically allocated to multiple independent physical wireless channels according to the polling strategy, and adjacent fragmented microframes and microframes of different long messages are staggered and resent in the temporal domain. Through the spatial and temporal scattering on the physical channel, the risk of sudden fading and multipath interference of the physical channel is greatly reduced, and high-reliability data transmission is guaranteed in high congestion and high packet loss environments.

[0093] (4) The lock-free sliding window buffer pool based on the comparison and exchange atomic instructions enables extremely short control frames to be submitted directly after stripping the tag header, and long message microframes to be loaded without lock through the atomic exchange instructions. After reaching the decoding threshold, it self-heals and reassembles, smoothly pushing the window pointer to move, completely eliminating the overhead of lock contention, and ensuring that the cloud still maintains extremely high reassembly throughput and system response speed when facing high-concurrency heterogeneous data streams. Attached Figure Description

[0094] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0095] Figure 1 This is a flowchart of a method for a smart warehouse heterogeneous data communication system based on edge-cloud collaboration according to the present invention.

[0096] Figure 2 This is a system architecture block diagram of a smart warehouse heterogeneous data communication system based on edge-cloud collaboration according to the present invention. Detailed Implementation

[0097] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0098] Example 1, referring to Figure 1 This is the first embodiment of the present invention. This embodiment provides a smart warehouse heterogeneous data communication system based on edge-cloud collaboration, including an edge data access gateway located at the warehouse site and a cloud front-end receiving node located at the upper-level control center. Specifically, it includes: a data access module, a feature extraction module, a classification and encoding module, a preemptive sending module, a receiving buffer module, and a window reassembly module.

[0099] The edge data access gateway includes:

[0100] Data access module: Real-time collection of raw messages sent by heterogeneous devices in the warehouse;

[0101] Through a multi-threaded concurrency mechanism, the system performs real-time status polling and interrupt event monitoring of various physical communication interfaces deployed in the warehouse, and receives raw byte streams from heterogeneous devices in the warehouse.

[0102] It should be further noted that the raw byte stream refers to the continuous byte sequence without boundaries received by various physical communication interfaces, while the raw message refers to the data unit with a complete protocol physical frame structure extracted or cut from the raw byte stream according to the frame structure characteristics, timing attributes or byte matching algorithms corresponding to each underlying protocol.

[0103] The various physical communication interfaces include, but are not limited to, controller area network bus interface, asynchronous serial communication interface, Ethernet physical interface and wireless LAN antenna interface;

[0104] The heterogeneous equipment in the warehouse includes, but is not limited to, automated guided vehicles, stacker cranes, automated shuttles, programmable logic controllers, and image acquisition equipment.

[0105] The raw byte stream is loaded into the access ring buffer. Based on the frame structure characteristics and timing attributes of different underlying communication protocols, physical frame boundaries are trimmed and the original message is extracted. Specifically, this includes:

[0106] For Controller Area Network (CAN) protocol: Read the CAN frame structure in the access ring buffer, parse the identifier field and data length code, extract the payload of the corresponding number of bytes, and generate CAN frame message;

[0107] For asynchronous serial bus data transmission protocol: determine frame boundaries based on bus silent time interval, match device address bits and function codes, perform integrity verification through cyclic redundancy check code, and extract bus data transmission protocol messages;

[0108] For the Unified Process Control Architecture Protocol (UCAP) and Message Queuing Telemetry Transmission Protocol (MQTP): the length identifier field of the application layer protocol header in the UCTP payload is extracted using the sliding window algorithm, and the complete UCAP and MQTP messages are then truncated.

[0109] It should also be noted that for the process control unified architecture protocol, the message total size field of the fixed header is read, and for the message queue telemetry transmission protocol, the variable length remaining length field of the fixed header is read and the total message length is calculated.

[0110] The complete, trimmed original message undergoes protocol feature parsing, and communication and service attributes are extracted. Data type pre-identification and labeling are then performed, specifically including:

[0111] If the predicted control command is related to equipment emergency stop, collision warning, or arrival confirmation, it is marked as a data type of extremely short control frame;

[0112] If the data is pre-determined to involve visual barcode images, device operation logs, or high-frequency sensor status, it is marked as a long message data type;

[0113] The payload data of the original message is kept unaltered, and the checksum and link layer encapsulation information related to the underlying physical link are stripped away to generate a standardized message to be processed, which is then pushed into the buffer queue.

[0114] Feature extraction module: performs protocol-independent stripping and analysis on the original message, and generates and embeds a time-series semantic tag header;

[0115] Read standardized pending messages in the buffer queue in real time and extract the original payload, protocol type identifier and original receive timestamp;

[0116] Based on the protocol type identifier and the service attribute code in the original payload, calculate the maximum tolerable latency threshold and the disruption tolerance, specifically including:

[0117] If the standardized message to be processed is marked as a data type of extremely short control frame, then its maximum tolerable delay time threshold is set to 1 millisecond;

[0118] If the standardized message to be processed is marked as a long message data type, then its maximum tolerable delay time threshold is set to 500 milliseconds;

[0119] If the standardized message to be processed is marked as a data type of very short control frame, then the destruction tolerance flag is set to an indestructible state.

[0120] If the standardized message to be processed is marked as a data type of long message data, then the destruction tolerance flag is set to a destructible state;

[0121] The clock source and incrementing counter of the edge data access gateway are invoked to generate a global sequence reassembly identifier, and a semantic tag header of fixed size of four bytes is assembled according to a preset bit format.

[0122] The semantic tag header specifically includes:

[0123] The first byte to the second byte: stores the global sequence reassembly identifier;

[0124] The third byte stores the maximum tolerable latency threshold;

[0125] The fourth byte stores a combination of the violation tolerance flag and the protocol type flag.

[0126] The semantic tag header is encapsulated at the very beginning of the original payload of the standardized message to be processed, generating a tagged message, which is then pushed into the processing queue.

[0127] Classification and coding module: Divides the original message into very short control frames and long message data, dynamically slices the long message data based on the coherence time of the wireless channel, and applies forward error correction erasure code for redundant coding to generate fragmented micro-frames;

[0128] Real-time reception of tagged packets, parsing of the semantic tag header, reading of the maximum tolerable delay time threshold and the violation tolerance flag, and execution of dual-channel routing and traffic splitting, specifically including:

[0129] Extremely short control frame diversion: If the maximum tolerable delay time threshold is lower than the preset low latency threshold, and the damage tolerance is marked as non-destructible, the marked message is determined to be an extremely short control frame and diverted to the high-speed transparent transmission channel.

[0130] Long message data diversion: If the maximum tolerable delay time threshold is higher than the preset low delay threshold, or the damage tolerance is marked as destructible, the marked message is determined to be long message data and diverted to the slice encoding channel.

[0131] The high-speed transparent transmission channel will push tagged messages that are determined to be extremely short control frames into the priority transmission queue.

[0132] The slice coding channel obtains the coherence time measurement value of the current physical wireless channel, and calculates the maximum number of bytes transmitted in a single frame of the physical wireless channel based on the ratio of the current physical wireless channel transmission rate to the preset safety margin of the coherence time measurement value.

[0133] The maximum number of bytes is determined by subtracting the size of the four-byte semantic tag header and the forward error correction verification information from the maximum number of bytes.

[0134] The original payload of the long message data is continuously cut based on the maximum safe payload size of a single frame to generate K ordered original slices;

[0135] It should also be noted that K is an integer greater than or equal to 1, and the transmission time of each original slice transmitted individually is less than the coherence time measurement of the physical wireless channel.

[0136] The original slice is subjected to forward error correction and erase code redundancy encoding to generate N encoded slices, where N>K, and includes K data slices and M=NK check slices. For each of the encoded slices, a four-byte semantic tag header corresponding to the long message data is copied and embedded.

[0137] Forward error correction verification information is appended to the end of each encoded slice to construct N fragmented microframes, which are then pushed into the normal transmission queue in sequence.

[0138] Preemptive transmission module: The extremely short control frame and fragmented microframes are transmitted through multiple physical channels with temporal interleaving and spatial fragmentation, and a preemptive mechanism is triggered when the extremely short control frame arrives.

[0139] It monitors the preemptive priority transmission queue and the normal transmission queue in real time, and simultaneously senses the channel status and signal strength of multiple available independent physical wireless channels.

[0140] When a very short control frame is stored in the preemptive priority transmission queue, a hardware-level preemption interrupt is triggered, specifically including:

[0141] Link physical interruption: If any physical wireless channel is currently transmitting fragmented microframes, immediately send an abort signal to the corresponding physical layer controller to interrupt the transmission;

[0142] Discard interrupted microframes: Discard interrupted microframe fragments without re-entering the queue, and use forward error correction erasure codes at the receiving end to decode and fill in the gaps;

[0143] Dequeueing of extremely short control frames: The extremely short control frames are extracted and dequeued from the priority transmission queue;

[0144] The outgoing very short control frame is copied into multiple copies and transmitted in parallel through multiple independent physical wireless channels to ensure that if a single channel is blocked, the other physical channels will transmit the very short control frame to the receiving end.

[0145] When the priority transmission queue is empty and the normal transmission queue is not empty, the scheduling process for fragmented microframes is executed, specifically including:

[0146] The fragmented microframes corresponding to the same long message data are dynamically allocated to multiple independent physical wireless channels according to the polling strategy, and adjacent fragmented microframes are transmitted through different physical channels.

[0147] The fragmented microframes corresponding to different long message data are interleaved and rearranged in the time domain to avoid the concentrated transmission of consecutive fragmented microframes of a single long message data on the same physical channel, and are sent concurrently through the corresponding physical wireless channels.

[0148] After each fragment of a microframe is sent, the normal transmission queue pointer is updated until the normal transmission queue is empty or is preempted and interrupted.

[0149] After the very short control frame is sent, the preemption interruption flag is cleared and control of the physical wireless channel is released. The scheduling and processing of fragmented microframes is automatically resumed, and the fragmented microframes in the normal transmission queue are scattered, scheduled, and sent concurrently.

[0150] The cloud-based front-end receiving node includes:

[0151] Receive buffer module: Receives the short control frame and fragmented microframes;

[0152] It can monitor multiple independent physical wireless channels in real time, receive wireless signals and unpack and extract data frames;

[0153] Perform low-level physical verification on the data frame, calculate the cyclic redundancy check (CRC) code of the data frame, and compare it with the check code carried at the end of the frame:

[0154] If the verification fails, the data frame is discarded.

[0155] If the verification is successful, the data frame is determined to be a valid data frame.

[0156] Parse the semantic marker header of the first four bytes of the valid data frame, read the maximum tolerable latency threshold and the violation tolerance flag, and perform dual-channel routing splitting, specifically including:

[0157] Extremely short control frame offloading: If the maximum tolerable delay time threshold is lower than the preset low latency threshold and the destruction tolerance is marked as non-destructible, the valid data frame is determined to be an extremely short control frame, and the global sequence reassembly identifier in its semantic tag header is read.

[0158] If the very short control frame is the first frame to arrive in a multi-channel transmission, push it into the very short control frame fast buffer;

[0159] If the extremely short control frame is a subsequently arriving duplicate frame, it is filtered and discarded.

[0160] Fragmented microframe routing: If the maximum tolerable latency threshold is higher than the preset low latency threshold, or the destruction tolerance is marked as destructible, the valid data frame is determined to be a fragmented microframe. Then, the global sequence reassembly identifier in its semantic tag header is read, classified according to the corresponding long message data, and pushed into the corresponding storage unit in the fragmented microframe circular buffer.

[0161] Window reassembly module: Directly transmits the short control frame and performs self-healing reassembly of the fragmented microframes through a forward error correction decoding matrix to restore the original message;

[0162] Initialize the lock-free sliding window buffer pool based on atomic operations, and monitor the status of the ultra-short control frame high-speed buffer and the fragmented micro-frame circular buffer in real time.

[0163] Among them, the left boundary pointer of the lockless sliding window buffer pool points to the global sequence reassembly identifier of the earliest incomplete long message data currently waiting to be reassembled, and the right boundary pointer is dynamically maintained according to the preset window span;

[0164] When an extremely short control frame is stored in the extremely short control frame fast buffer, the extremely short control frame is dequeued using a lock-free queue read instruction, and the following processing is performed:

[0165] Remove the four-byte semantic tag header at the very beginning of the extremely short control frame to recover the original payload;

[0166] The original payload is submitted transparently to the cloud scheduling and control engine;

[0167] When a fragmented microframe is stored in the fragmented microframe circular buffer, the global sequence reassembly identifier and slice number in the semantic tag header of the fragmented microframe are read, and it is determined whether the global sequence reassembly identifier falls within the left and right boundaries of the current unlocked sliding window buffer pool.

[0168] If the fragmented microframe falls within the window boundary, the atomic swap instruction is used to store the fragmented microframe into the data matrix slot of the corresponding global sequence recombination identifier in the unlocked sliding window buffer pool;

[0169] If it falls behind the left edge of the window, the fragmented microframe is discarded;

[0170] For long message data corresponding to a global sequence reassembly identifier in the unlocked sliding window buffer pool, the number of received valid fragmented microframes is counted in real time, and the number of valid fragmented microframes is compared with the preset forward error correction decoding threshold.

[0171] When the number of effective fragmented microframes is greater than the preset forward error correction decoding threshold, the forward error correction decoding matrix is ​​invoked to perform algebraic operations on the received fragmented microframes and inverse decode to recover the original slice.

[0172] Perform reassembly and memory release on the original slices recovered from the inversion decoding, specifically including:

[0173] The original slices are assembled sequentially according to their slice numbers, and the forward error correction and verification information appended at the end is removed to reconstruct and restore the original long message data.

[0174] The restored original long message data is submitted to the cloud, and the left boundary pointer of the lockless sliding window buffer pool is smoothly moved forward by the atomic instructions of comparison and exchange, releasing the memory space of the data matrix slot that has been reorganized.

[0175] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0176] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the steps of implementing a smart warehouse heterogeneous data communication system based on edge-cloud collaboration as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0177] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A smart warehousing heterogeneous data communication system based on edge-cloud collaboration, characterized in that, It includes an edge data access gateway located at the warehouse site and a cloud-based front-end receiving node located at the upper-level control center, specifically including: a data access module, a feature extraction module, a classification and coding module, a preemptive transmission module, a receiving buffer module, and a window reassembly module; The edge data access gateway includes: Data access module: Real-time collection of raw messages sent by heterogeneous devices in the warehouse; Feature extraction module: performs protocol-independent stripping and analysis on the original message, and generates and embeds a time-series semantic tag header; Classification and coding module: Divides the original message into very short control frames and long message data, dynamically slices the long message data based on the coherence time of the wireless channel, and applies forward error correction erasure code for redundant coding to generate fragmented micro-frames; Preemptive transmission module: The extremely short control frame and fragmented microframes are transmitted through multiple physical channels with temporal interleaving and spatial fragmentation, and a preemptive mechanism is triggered when the extremely short control frame arrives. The cloud-based front-end receiving node includes: Receive buffer module: Receives the short control frame and fragmented microframes; Window reassembly module: Directly transmits the short control frame and performs self-healing reassembly of the fragmented microframes through a forward error correction decoding matrix to restore the original message.

2. The smart warehousing heterogeneous data communication system based on edge-cloud collaboration according to claim 1, characterized in that: In the data access module, a multi-threaded concurrency mechanism is used to perform real-time status polling and interrupt event monitoring of various physical communication interfaces deployed in the warehouse, and to receive raw byte streams sent by heterogeneous devices in the warehouse. The original byte stream is loaded into the access ring buffer, and the physical frame boundaries are clipped and the original message is extracted according to the frame structure characteristics and timing attributes of different underlying communication protocols. The complete, trimmed original message undergoes protocol feature parsing, and communication and service attributes are extracted. Data type pre-identification and labeling are then performed, specifically including: If the predicted control command is related to equipment emergency stop, collision warning, or arrival confirmation, it is marked as a data type of extremely short control frame; If the data is pre-determined to involve visual barcode images, device operation logs, or high-frequency sensor status, it is marked as a long message data type; The payload data of the original message is kept unaltered, and the checksum and link layer encapsulation information related to the underlying physical link are stripped away to generate a standardized message to be processed, which is then pushed into the buffer queue.

3. The smart warehousing heterogeneous data communication system based on edge-cloud collaboration according to claim 1, characterized in that: In the feature extraction module, standardized messages to be processed in the buffer queue are read in real time, and the original payload, protocol type identifier and original reception timestamp are extracted. Based on the protocol type identifier and the service attribute code in the original payload, calculate the maximum tolerable latency threshold and the disruption tolerance. The clock source and incrementing counter of the edge data access gateway are invoked to generate a global sequence reassembly identifier, and a semantic tag header of fixed size of four bytes is assembled according to a preset bit format. The semantic tag header is encapsulated at the very beginning of the original payload of the standardized message to be processed, generating a tagged message, which is then pushed into the processing queue.

4. The smart warehousing heterogeneous data communication system based on edge-cloud collaboration according to claim 1, characterized in that: The classification and coding module receives tagged messages in real time, parses the semantic tag header, reads the maximum tolerable delay time threshold and the violation tolerance flag, and performs dual-channel routing, specifically including: Extremely short control frame diversion: If the maximum tolerable delay time threshold is lower than the preset low latency threshold, and the damage tolerance is marked as non-destructible, the marked message is determined to be an extremely short control frame and diverted to the high-speed transparent transmission channel. Long message data diversion: If the maximum tolerable delay time threshold is higher than the preset low delay threshold, or the damage tolerance is marked as destructible, the marked message is determined to be long message data and diverted to the slice encoding channel. The high-speed transparent transmission channel will push tagged messages that are determined to be extremely short control frames into the priority transmission queue. The slice coding channel obtains the coherence time measurement value of the current physical wireless channel, and calculates the maximum number of bytes transmitted in a single frame of the physical wireless channel based on the ratio of the current physical wireless channel transmission rate to the preset safety margin of the coherence time measurement value. The maximum number of bytes is determined by subtracting the size of the four-byte semantic tag header and the forward error correction verification information from the maximum number of bytes. The original payload of the long message data is continuously cut based on the maximum safe payload size of a single frame to generate K ordered original slices; The original slice is subjected to forward error correction and erase code redundancy encoding to generate N encoded slices, and a four-byte semantic tag header corresponding to the long message data is copied and embedded in each of the encoded slices. Forward error correction verification information is appended to the end of each encoded slice to construct N fragmented microframes, which are then pushed into the normal transmission queue in sequence.

5. The smart warehouse heterogeneous data communication system based on edge-cloud collaboration according to claim 1, characterized in that: In the preemptive transmission module, the preemptive priority transmission queue and the normal transmission queue are monitored in real time, and the channel status and signal strength of multiple available independent physical wireless channels are simultaneously sensed. When a very short control frame is stored in the preemptive priority transmission queue, a hardware-level preemption interrupt is triggered, specifically including: Link physical interruption: If any physical wireless channel is currently transmitting fragmented microframes, immediately send an abort signal to the corresponding physical layer controller to interrupt the transmission; Discard interrupted microframes: Discard interrupted microframe fragments without re-entering the queue, and use forward error correction erasure codes at the receiving end to decode and fill in the gaps; Dequeueing of extremely short control frames: The extremely short control frames are extracted and dequeued from the priority transmission queue; The outgoing very short control frame is copied into multiple copies and transmitted in parallel simultaneously through multiple independent physical wireless channels; When the priority transmission queue is empty and the normal transmission queue is not empty, the scheduling process for fragmented microframes is executed, specifically including: The fragmented microframes corresponding to the same long message data are dynamically allocated to multiple independent physical wireless channels according to the polling strategy, and adjacent fragmented microframes are transmitted through different physical channels. The fragmented micro-frames corresponding to different long message data are time-domain interleaved and rearranged, and then sent concurrently through the corresponding physical wireless channels. After each fragment of a microframe is sent, the normal transmission queue pointer is updated until the normal transmission queue is empty or is preempted and interrupted. After the very short control frame is sent, the preemption interruption flag is cleared and control of the physical wireless channel is released. The scheduling and processing of fragmented microframes is automatically resumed, and the fragmented microframes in the normal transmission queue are scattered, scheduled, and sent concurrently.

6. The smart warehousing heterogeneous data communication system based on edge-cloud collaboration according to claim 1, characterized in that: The receiving buffer module concurrently monitors multiple independent physical wireless channels in real time, receives wireless signals, and unpacks and extracts data frames. Perform low-level physical verification on the data frame, calculate the cyclic redundancy check (CRC) code of the data frame, and compare it with the check code carried at the end of the frame: If the verification fails, the data frame is discarded. If the verification is successful, the data frame is determined to be a valid data frame. Parse the semantic marker header of the first four bytes of the valid data frame, read the maximum tolerable latency threshold and the violation tolerance flag, and perform dual-channel routing splitting, specifically including: Extremely short control frame offloading: If the maximum tolerable delay time threshold is lower than the preset low latency threshold and the destruction tolerance is marked as non-destructible, the valid data frame is determined to be an extremely short control frame, and the global sequence reassembly identifier in its semantic tag header is read. If the very short control frame is the first frame to arrive in a multi-channel transmission, push it into the very short control frame fast buffer; If the extremely short control frame is a subsequently arriving duplicate frame, it is filtered and discarded. Fragmented microframe routing: If the maximum tolerable latency threshold is higher than the preset low latency threshold, or the destruction tolerance is marked as destructible, the valid data frame is determined to be a fragmented microframe. Then, the global sequence reassembly identifier in its semantic tag header is read, classified according to the corresponding long message data, and pushed into the corresponding storage unit in the fragmented microframe circular buffer.

7. The smart warehousing heterogeneous data communication system based on edge-cloud collaboration according to claim 1, characterized in that: In the window reorganization module, an unlocked sliding window buffer pool based on atomic operations is initialized, and the status of the ultra-short control frame high-speed buffer and the fragmented micro-frame circular buffer is monitored in real time. When an extremely short control frame is stored in the extremely short control frame fast buffer, the extremely short control frame is dequeued using a lock-free queue read instruction, and the following processing is performed: Remove the four-byte semantic tag header at the very beginning of the extremely short control frame to recover the original payload; The original payload is submitted transparently to the cloud scheduling and control engine; When a fragmented microframe is stored in the fragmented microframe circular buffer, the global sequence reassembly identifier and slice number in the semantic tag header of the fragmented microframe are read, and it is determined whether the global sequence reassembly identifier falls within the left and right boundaries of the current unlocked sliding window buffer pool. If the fragmented microframe falls within the window boundary, the atomic swap instruction is used to store the fragmented microframe into the data matrix slot of the corresponding global sequence recombination identifier in the unlocked sliding window buffer pool; If it falls behind the left edge of the window, the fragmented microframe is discarded; For long message data corresponding to a global sequence reassembly identifier in the unlocked sliding window buffer pool, the number of received valid fragmented microframes is counted in real time, and the number of valid fragmented microframes is compared with the preset forward error correction decoding threshold. When the number of effective fragmented microframes is greater than the preset forward error correction decoding threshold, the forward error correction decoding matrix is ​​invoked to perform algebraic operations on the received fragmented microframes and inverse decode to recover the original slice. Perform reassembly and memory release on the original slices recovered from the inversion decoding, specifically including: The original slices are assembled sequentially according to their slice numbers, and the forward error correction and verification information appended at the end is removed to reconstruct and restore the original long message data. The restored original long message data is submitted to the cloud, and the left boundary pointer of the lockless sliding window buffer pool is smoothly moved forward by the atomic instructions of comparison and exchange, releasing the memory space of the data matrix slot that has been reorganized.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the smart warehouse heterogeneous data communication system based on edge-cloud collaboration as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the smart warehouse heterogeneous data communication system based on edge-cloud collaboration as described in any one of claims 1 to 7.

Images