A wireless data recording transmission method based on the Internet of Things

Abstract

The application relates to the technical field of Internet of Things communication and data processing, and particularly discloses a wireless data recording transmission method based on the Internet of Things. A bionic routing module is arranged in a node, a lymphoid hibernation mode is triggered when data redundancy or cache overload is detected, normal uploading is suspended, and a low-power cleaning pulse packet is sent; adjacent nodes evaluate the effectiveness of cache data according to the low-power cleaning pulse packet, compress or remove invalid redundant data, dynamically disconnect inefficient links, and enhance the bandwidth weight of high-frequency stable links to complete synapse remodeling; and data transmission is restored after the hibernation ends. The application realizes network self-cleaning and link self-optimization through the above mechanism, prolongs the service life of the node without additional hardware, reduces transmission delay after waking up, and is suitable for large-scale low-power high-reliability Internet of Things scenes.

Description

Technical Field

[0001] This invention belongs to the field of Internet of Things (IoT) communication and data processing, and specifically relates to a wireless data recording and transmission method based on IoT. Background Technology

[0002] With the rapid development of IoT technology, the wireless interconnection and data interaction of large-scale sensor nodes has become a core support for smart cities, industrial monitoring, and environmental monitoring. In complex monitoring environments, IoT nodes, through distributed sensing and wireless communication technologies, enable real-time collection, recording, and remote transmission of massive amounts of environmental information. Since nodes are typically deployed in energy-constrained and environmentally variable scenarios, balancing data throughput, energy efficiency, and network robustness through efficient transmission protocols has become a key direction for the continued evolution of IoT technology.

[0003] Resource scheduling and management during wireless data recording and transmission typically involves the dynamic adjustment of node sleep mechanisms and caching strategies. This technical solution aims to reduce hardware power consumption during non-communication periods and achieve centralized uplink transmission of cached data within the communication window through preset scheduling logic. However, with the exponential growth of data volume, a single sleep mechanism and static cache management logic are insufficient to cope with sudden traffic pressures and network congestion. This poses a more stringent challenge to the node's system self-healing capabilities and end-to-end real-time transmission performance.

[0004] Existing data transmission methods often rely on fixed-slot sleep strategies or simple linear buffer management. During network congestion, this can easily lead to node buffer overflows and system deadlocks, resulting in the loss of critical data packets and network communication paralysis. Traditional protocols lack intelligent identification and proactive cleanup capabilities for redundant data, causing a large amount of invalid data to occupy valuable communication bandwidth and storage space. This not only shortens the overall lifespan of the network but also causes severe transmission delays for nodes after they are woken up, making it difficult to resolve the performance contradiction between sleep mode and efficient response. Therefore, a wireless data recording and transmission method based on the Internet of Things (IoT) is desired. Summary of the Invention

[0005] The purpose of this invention is to provide a wireless data recording and transmission method based on the Internet of Things, which can effectively solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a wireless data recording and transmission method based on the Internet of Things, comprising the following specific steps: Step 1: Deploy a bionic routing module in the IoT node. This bionic routing module continuously monitors network traffic status and cache utilization. When it detects that the data redundancy exceeds the preset threshold or the cache utilization reaches the predetermined upper limit, it triggers a lymphatic hibernation mode. Step 2: After entering the lymphatic-like dormancy mode, the node suspends regular data upload tasks and only periodically generates and sends low-power cleaning pulse packets. The cleaning pulse packets carry fluid dynamics guidance information to guide adjacent nodes to coordinately adjust the data flow diffusion path. Step 3: Based on the guidance information transmitted by the cleaning pulse packet, each node evaluates the validity of the data packets in its local cache, and compresses or selectively removes data packets that are determined to be invalid or redundant, thereby freeing up storage space; Step 4: After the cleaning process is completed, based on the usage frequency and transmission success rate of historical communication paths, inefficient communication links are dynamically disconnected, and higher bandwidth weights are allocated to high-frequency stable links to complete the synaptic remodeling operation. Step 5: Exit the lymphatic hibernation mode, restore the normal data recording and uploading functions, and efficiently transmit the cleaned and optimized data stream through the rebuilt communication link.

[0007] Preferably, the bionic routing module in step 1 simulates the regulation mechanism of glial cells on neural synapses. It is equipped with a congestion sensing unit and a redundancy identification unit. The congestion sensing unit is used to collect the channel occupancy rate and queue length change trend in real time, while the redundancy identification unit determines whether there is duplicate or expired data by comparing the timestamp and content similarity of data packets.

[0008] Preferably, the cleaning pulse packet in step 2 adopts a simplified protocol format, containing only source node identifier, target area identifier and fluid guidance vector information, and its transmission power is controlled at the minimum level to maintain basic connectivity, so as to minimize energy consumption.

[0009] Preferably, the validity assessment process in step 3 is based on multi-dimensional criteria, including the timeliness of the data packet, the uniqueness of the content, and the confirmation status of the receiving end. If a data packet is not confirmed within a predetermined time and its content highly overlaps with other data to be transmitted, it is marked as a removable object.

[0010] Preferably, the compression process in step 3 uses lossless dictionary encoding to merge and store data packets with the same prefix or structural characteristics, while the targeted removal selects edge nodes far from the backbone path to perform the cleanup operation first based on the network topology, so as to avoid affecting the core data channel.

[0011] Preferably, the inefficient communication link in step 4 refers to a link whose average transmission success rate is lower than a preset threshold and whose bandwidth utilization rate remains low over the past few communication cycles. The system dynamically updates the link status by maintaining a link performance log table and makes a disconnection decision accordingly.

[0012] Preferably, the bandwidth weight enhancement in step 4 is achieved by fine-tuning the physical layer modulation parameters or the medium access control layer time slot allocation strategy, so that the high-frequency link can carry a larger proportion of data traffic in subsequent communications.

[0013] Preferably, the recovery process of the regular data recording and uploading function in step 5 includes a gradual wake-up phase, in which the node gradually increases the sampling frequency and transmission power until it reaches a normal working state, so as to prevent the buffer overflow from being triggered again due to instantaneous high load.

[0014] Preferably, the duration of the lymphatic hibernation mode is adaptively adjusted according to the overall network congestion level. The more severe the congestion, the longer the cleaning cycle, but it does not exceed the maximum hibernation time limit set by the system, so as to ensure that the response capability for critical events is not excessively weakened.

[0015] Preferably, during the cleaning and synaptic remodeling operations, the IoT node retains the ability to transmit high-priority emergency data. This data can bypass the cache and be directly uploaded through the reserved emergency channel, ensuring the reliability of the system under abnormal operating conditions.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention introduces a lymphatic system-like mechanism to construct a wireless data recording and transmission method with active cleaning and self-optimization capabilities, which solves the problem of data loss and system deadlock caused by buffer overflow in traditional IoT nodes under high load.

[0017] 2. This invention simultaneously cleans up invalid data and optimizes the link structure during network hibernation, which not only extends the lifespan of nodes and the entire network, but also unexpectedly reduces transmission latency after wake-up, breaking the traditional understanding that hibernation inevitably leads to response delays. This method does not require additional hardware support and can achieve efficient resource utilization through intelligent scheduling at the protocol level, making it suitable for large-scale, low-power, and high-reliability IoT application scenarios. Attached Figure Description

[0018] Figure 1 This is a flowchart according to the present invention.

[0019] Figure 2 This is a schematic diagram of data flow according to the present invention.

[0020] Figure 3 The present invention provides a flowchart for cleaning lymphatic dormant data.

[0021] Figure 4 The present invention provides a flowchart for link synaptic remodeling.

[0022] Figure 5The present invention provides a flowchart for guiding the cleaning pulse path. Detailed Implementation

[0023] Example 1: Please refer to the appendix Figure 1 To be continued Figure 5 To make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to specific embodiments.

[0024] In a wireless data recording and transmission method based on the Internet of Things (IoT), this invention constructs a communication protocol framework with self-cleaning and link optimization capabilities by integrating lymphatic system mechanisms from neuroscience. The core of this communication protocol framework lies in transforming the dormant phase of IoT nodes into an active cache cleaning and topology reshaping process.

[0025] In step 1, a biomimetic routing module is deployed in the IoT node. This biomimetic routing module continuously monitors the network traffic status and cache utilization. When it detects that the data redundancy exceeds a preset threshold or the cache utilization reaches a predetermined upper limit, a lymphatic hibernation mode is triggered.

[0026] During the execution of step 1, the bionic routing module internally includes a congestion sensing unit and a redundancy identification unit. The congestion sensing unit collects real-time data on channel occupancy and queue length trends. Specifically, the congestion sensing unit performs high-frequency sampling of the received signal strength indication and channel idle assessment signal provided by the physical layer to calculate the proportion of time the channel is occupied per unit time. When this proportion exceeds a preset threshold of 0.75 for five consecutive sampling periods, and the number of pending data packets in the node's transmission queue reaches 80% of the total buffer capacity, the congestion sensing unit generates a congestion trigger signal.

[0027] The redundancy identification unit determines whether duplicate or expired data exists by comparing the timestamps and content similarity of data packets. In specific implementation, the redundancy identification unit extracts a feature fingerprint for each data packet entering the cache and performs a hash comparison between the extracted feature fingerprint and the historical packet fingerprints stored in the scrolling window.

[0028] The extracted fingerprint fields include the media access control layer source address field, network layer protocol type field, the first 64 bytes of valid data in the data payload segment, and the data packet generation timestamp field. These fields are concatenated byte-order and used as input for hash calculation. The hash function uses the MD5 hash algorithm to output a 128-bit fixed-length binary fingerprint for fixed-length fast comparison. The size of the scrolling window is set to 20% of the maximum number of data packets that the node cache can hold. The window adopts a first-in-first-out update mechanism: when a new data packet enters the cache, its corresponding fingerprint is added to the tail of the window, while the earliest historical fingerprint stored at the head of the window is removed. When the number of fingerprints stored in the window has not reached the upper limit, only the addition operation is performed, and no removal is performed. All fingerprints in the window are arranged in the order of the data packet entering the cache and stored in the node's on-chip random access memory. The traversal range of a single hash comparison is limited to all historical fingerprints in the current window.

[0029] If two data packets have a content similarity exceeding 95% and their timestamp interval is less than a preset lifespan, they are considered redundant data. The redundancy identification unit calculates data redundancy by accumulating the number of redundant packets per unit time. When the data redundancy exceeds a preset threshold of 30%, a redundancy trigger signal is generated even if the cache occupancy rate has not yet reached its limit.

[0030] When either the congestion trigger signal or the redundancy trigger signal is activated, the bionic routing module will send a mode switching request to the main controller, causing the node to switch from the normal transmission mode to a lymphatic dormancy mode. During the triggering process, the node first completes the transmission or reception of the last complete data packet in progress to ensure the integrity of the data frame, and then locks the application layer data generation interface, ceasing the transmission of new service payloads to the network layer.

[0031] In step 2, after entering the lymphatic-like dormancy mode, the node suspends regular data upload tasks and only periodically generates and sends low-power cleaning pulse packets. The cleaning pulse packets carry fluid dynamics guidance information to guide adjacent nodes to coordinately adjust the data flow diffusion path.

[0032] The cleaning pulse packet adopts a simplified protocol format, whose structure includes a 1-byte source node identifier, a 1-byte target area identifier, and 2 bytes of fluid guidance vector information. The transmission power is precisely controlled at the minimum level that can maintain basic connectivity with adjacent nodes within a one-hop range, for example, set to -20 dBmW, to minimize energy consumption.

[0033] The generation of fluid guidance vector information is based on a fluid dynamics model. In this model, nodes analogize the data flow in the network to an incompressible fluid, the communication bandwidth between nodes to the cross-sectional area of ​​a pipe, and the buffer capacity to local pressure. Based on its own pressure distribution and that of its perceived neighboring nodes, each node calculates a gradient vector pointing towards a low-pressure area.

[0034] The specific calculation logic of the gradient vector is as follows: First, define the local pressure value of the node. The calculation formula is: ,in This represents the total capacity of the node cache. This represents the current cache capacity used by the node. The pressure value ranges from 0 to 1; a smaller value indicates higher cache usage and greater local data pressure. A two-dimensional coordinate system is established with the node's own position as the origin. The pressure value of each neighboring node within a one-hop range and the orientation angle of that neighboring node relative to the node are obtained. The pressure value of each neighboring node is used as a weight, multiplied by the corresponding unit vector, and then summed. The direction of the resulting vector sum is the direction of the gradient vector. The magnitude of the gradient vector is the difference between the node's own pressure value and the lowest pressure value among all its neighbors.

[0035] The acquisition period of neighbor node pressure data is consistent with the sending period of cleaning pulse packets, which is set to 10 milliseconds. The pressure data is carried through the reserved field of the cleaning pulse packet and is directly transmitted only between adjacent nodes in one hop, without global flooding across multiple hops, thereby reducing the communication overhead of the synchronization process.

[0036] To avoid data loop transmission, each data packet to be forwarded carries a hop count field, and each node maintains a fingerprint record table of forwarded data packets with a length of 50 entries. When the fingerprint of a data packet received by a node already exists in the local record table, or when the hop count of the data packet exceeds the preset maximum threshold of 8 hops, the node refuses to forward the data packet, thereby eliminating the problem of data looping back within the local network.

[0037] The gradient vector is encapsulated in a cleaning pulse packet and diffused outwards. Upon receiving the pulse packet, neighboring nodes parse the guidance information within and, combined with their own pressure status, update their local routing tables. This coordinated adjustment process simulates the physical properties of the brain's lymphatic system using cerebrospinal fluid flow to remove metabolic waste, enabling redundant data in the network to diffuse to edge nodes or nodes with idle storage, thus alleviating congestion on the main path.

[0038] In step 3, based on the guidance information transmitted by the cleaning pulse packet, each node evaluates the validity of the data packets in its local cache, and compresses or selectively removes data packets that are determined to be invalid or redundant, thereby freeing up storage space.

[0039] The validity assessment process is based on multi-dimensional criteria, including the timeliness of the data packet, the uniqueness of its content, and the receiver's acknowledgment status. The system maintains a time-to-live counter for each data packet in the cache. If a data packet does not receive an acknowledgment from the receiver within a predetermined time, and its content highly overlaps with subsequently arriving high-priority data packets, or if the service flow to which the data packet belongs has expired, then the data packet is marked as a removable object.

[0040] For packets marked as removable, the system performs compression or targeted removal. Compression employs lossless dictionary encoding, merging packets with the same prefix or structural characteristics. Specifically, nodes maintain a local dynamic dictionary, replacing frequently occurring protocol headers or fixed data fields with short index codes. For example, if packets reported by multiple sensors contain the same device ID and status mask, the compression engine extracts this duplicate information and stores it in the dictionary header, retaining only the changing measurements and indices in the body, thus reducing cache usage.

[0041] Targeted removal prioritizes edge nodes far from the backbone path for cleaning operations based on the network topology. The system pre-identifies each node's centripetal strength (the probability that the node is on the shortest path from the source to the aggregation point) using a topology discovery protocol. Edge nodes with lower centripetal strength, upon receiving a cleaning command, will prioritize discarding data packets already backed up by other redundant paths, preventing the cleaning operation from affecting the connectivity of core data channels.

[0042] In step 4, after the cleaning process is completed, inefficient communication links are dynamically disconnected based on the usage frequency and transmission success rate of historical communication paths, and higher bandwidth weights are allocated to high-frequency stable links to complete the synaptic remodeling operation.

[0043] An inefficient communication link is defined as one with an average transmission success rate below 60% over the past 20 communication cycles and consistently low bandwidth utilization. The system dynamically updates the link status by maintaining a link performance log table. The link performance log table records the retransmission count, average round-trip time, and signal-to-noise ratio fluctuations for each active link. When the overall score of a link falls below a preset safety threshold, the bionic routing module makes a disconnection decision, removing it from the current route candidate list, similar to the synaptic pruning process in the nervous system.

[0044] The system enhances the bandwidth weight of high-frequency stable links. This enhancement is achieved by fine-tuning the physical layer modulation parameters or the medium access control layer time slot allocation strategy. The system increases the priority parameters of stable links when contending for the channel, or allocates more consecutive time slots to the high-frequency stable link in the time-division multiple access-based scheduling table. The instantaneous throughput of the physical layer can also be improved by increasing the order of the modulation and coding scheme, such as switching from binary phase shift keying to 16th-order quadrature amplitude modulation. This dynamic weight allocation allows high-frequency links to carry a larger proportion of data traffic in subsequent communications, achieving on-demand resource reconfiguration.

[0045] In step 5, the lymphatic hibernation mode is exited, the normal data recording and uploading functions are restored, and the cleaned and optimized data stream is efficiently transmitted through the reshaped communication link.

[0046] The recovery process for regular data logging and uploading includes a gradual wake-up phase. During this phase, nodes do not instantly return to full capacity; instead, they gradually increase the sampling frequency and transmit power in three sub-phases. In the first sub-phase, nodes only restore 20% of the sensor sampling frequency and probe link activity with the lowest transmit power. In the second sub-phase, the sampling frequency is increased while the power control loop is adjusted based on link feedback. In the third sub-phase, normal operation is fully restored. This gradual mechanism effectively prevents buffer overflows or channel collisions caused by a large number of nodes waking up simultaneously due to a sudden surge in load.

[0047] Furthermore, the duration of the lymphatic dormancy mode is adaptively adjusted based on the overall network congestion level. The system defines a global congestion factor, which is equal to the weighted sum of the average buffer occupancy rate and the average channel collision probability of all nodes.

[0048] The weighting coefficient for average buffer occupancy is 0.6, and the weighting coefficient for average channel collision probability is 0.4. The two parameters are multiplied by their respective weighting coefficients and then summed to obtain a global congestion factor with a value ranging from 0 to 1.

[0049] Global state information is collected using a distributed, hierarchical aggregation method: the network is divided into multiple regional clusters according to its topology, and each regional cluster elects a cluster head node; ordinary nodes within a region report their own buffer occupancy rate and channel conflict probability data to their respective cluster head node. The cluster head node calculates the regional average buffer occupancy rate and regional average channel conflict probability, and then aggregates them hop-by-hop to the network's root aggregation node; the root aggregation node summarizes the parameters of all regions, calculates the global congestion factor for the entire network, and then distributes the global congestion factor value hop-by-hop to each terminal node via the downlink signaling link. The global state collection period is set to 500 milliseconds, and the parameter collection process is only initiated in regions where nodes trigger a lymphatic dormancy mode; global state collection is not performed in non-congested regions, thereby reducing the overall communication overhead of the network.

[0050] The higher the congestion factor, the longer the cleaning cycle should be set to ensure sufficient time for deep cleaning. However, to guarantee real-time performance, the maximum cleaning cycle is limited to within 200 milliseconds.

[0051] During the cleaning and synaptic remodeling operations, IoT nodes retain the ability to transmit high-priority emergency data. This type of data is assigned the highest scheduling level. When this data arrives, the node immediately pauses the current cleaning or compression task and uploads it directly through the reserved channel using the reserved emergency deterministic time slots. The emergency channel uses an independent hardware buffer and priority queue, and is not subject to the logical limitations of the lymphatic hibernation mode, ensuring the extreme reliability of the system under abnormal operating conditions.

[0052] Example 2: In Example 2, we will elaborate on the specific implementation details of this method in a large-scale industrial sensor network environment, especially the application logic when dealing with high-concurrency burst traffic.

[0053] In step 1, pressure, vibration, and temperature sensor nodes deployed at the industrial site monitor the working environment in real time through their internal biomimetic routing modules. Due to the complex electromagnetic interference at industrial sites, the congestion sensing unit employs an energy detection algorithm based on a sliding window. This energy detection algorithm divides the past 100 milliseconds into 10 observation time slots and counts the proportion of unexpected signal energy exceeding the noise floor by 6 dB in each time slot. If this proportion exceeds 70% for four consecutive time slots, it is determined to be logical congestion caused by deterioration of the communication environment.

[0054] The redundancy identification unit plays a crucial role here. In industrial monitoring, many sensors report highly repetitive data under normal operating conditions. The redundancy identification unit adapts to different production cycles by setting dynamically changing content comparison thresholds. When the production line is in a stable operating period, this content comparison threshold is set to 90%, meaning that even minor fluctuations will be considered redundancy and trigger the cleaning logic.

[0055] In step 2, when a node enters a lymphatic-like dormancy mode, its physical layer circuitry enters a low-power listening state. The cleaning pulse packets sent at this time not only contain spatial location information but also embed a scalar value representing "node health." This health status integrates remaining battery power, internal memory wear leveling, and current operating temperature. Hydrodynamic guidance information is adjusted based on this health status, causing data flow to tend to converge towards nodes with higher health and richer resources. This logically simulates the process of nutrients flowing to metabolically active areas in biological tissue.

[0056] In the specific implementation of step 3, the validity evaluation algorithm introduces the parameter of "expected data lifecycle value." For data packets with high instantaneous value, such as those experiencing sudden pressure changes, their expected lifecycle value is set to infinity, prohibiting their removal during the cleaning process. For regular heartbeat packets or status query response packets, if they remain in the cache for more than 500 milliseconds and are not successfully sent, their validity score drops below the threshold for triggering removal. During compression, the lossless dictionary encoder performs deep optimization based on the specific frame structure of the industrial protocol. By stripping redundant transaction identifiers and protocol control fields, cache utilization can be improved by more than 3 times.

[0057] In step 4, the synaptic remodeling operation involves a deep adjustment of the network logical topology. In industrial environments, equipment locations are relatively fixed, but the quality of the wireless channel fluctuates periodically due to the movement of robotic arms or metal shielding. The system identifies these periodically failing links using a link performance log table. For links whose transmission success rate drops sharply during specific periods, the synaptic pruning mechanism temporarily marks them as "dormant synapses," excluding them from bandwidth allocation during remodeling. Conversely, for paths with strong penetration and weak multipath effects, the system fine-tunes their minimum contention window size at the medium access control layer, increasing their probability of gaining transmission rights in channel contention.

[0058] In step 5, the progressive wake-up phase employs a feedback-based power increment strategy. Before resuming uplink, the node first sends a miniature synchronization frame. If an acknowledgment is received from the sink node within the predetermined round-trip time, it indicates that the rebuilt link is fully available, and only then is the transmit power increased from the low-power level to the nominal operating level. If no acknowledgment is received, the node re-triggers the cleaning pulse in step 2 to fine-tune the steering vector again.

[0059] In this embodiment, the adaptive time control logic of the lymphatic hibernation mode is as follows: a basic cleaning duration value is set, and this basic cleaning duration value is multiplied by the average retransmission count of the current node. If the retransmission count increases, it indicates a severe conflict, and the cleaning time is increased accordingly. Simultaneously, to prevent nodes from losing critical industrial instructions due to prolonged hibernation, the system inserts a 50-microsecond "instruction listening window" in the middle of each cleaning cycle, specifically for capturing downlink control messages from the control center.

[0060] Example 3: In Example 3, the present invention is applied to a highly mobile urban smart transportation IoT system. In this scenario, the topological relationships between nodes change extremely rapidly, and traditional routing maintenance mechanisms often struggle to respond in real time.

[0061] In step 1, the vehicle-mounted node or road test unit monitors not only its own buffer rate through the bionic routing module, but also the density of surrounding neighboring nodes. The congestion sensing unit estimates node density by listening to the frequency of beacon frames sent by surrounding nodes. When the node density exceeds 500 per square kilometer and a single node has more than 20 concurrent connection requests, the system determines that logical congestion has occurred. The redundancy identification unit then focuses on checking the repetition of geographical location information. For a large number of duplicate state packets generated near the same coordinate point, they are identified as highly redundant data.

[0062] In step 2, the cleaning pulse packets sent by the mobile node entering the lymphatic dormancy mode contain fluid guidance vector information that includes the node's velocity and direction of motion. The fluid dynamics model is here extended to "relative fluid dynamics," meaning that the data flow not only flows towards low-pressure areas but also tends to diffuse towards nodes whose motion aligns with its own. This effectively utilizes the physical connectivity opportunities provided by vehicle movement and reduces the waste of channel resources caused by reverse transmission.

[0063] In step 3, a multi-dimensional criterion evaluation is performed on data packets in a mobile environment. The receiver's acknowledgment status is assigned the highest weight. If a data packet, although still timely, has been successfully received by multiple base stations or roadside units using spatial diversity technology during movement, it is determined to be redundant and can be removed from the local cache. During compression, the system uses differential coding technology to record only the changes in vehicle coordinates, speed, and acceleration relative to the previous moment, compressing the originally massive positioning information into an extremely short incremental sequence.

[0064] In step 4, the synaptic remodeling mechanism manifests as a rapid switching of motion-aware links. Inefficient links are defined here as links with drastic signal-to-noise ratio fluctuations caused by rapid node movement. The link performance log records Doppler frequency shift parameters. When the Doppler frequency shift exceeds a preset frequency offset threshold, the system immediately performs synaptic pruning, disconnecting unstable links. For node pairs predicted to maintain a stable relative position within the next 3 seconds, extremely high bandwidth weights are assigned, and their signal gain is enhanced through physical layer beamforming technology.

[0065] In step 5, the gradual wake-up process is tightly coupled with the node's movement state. When a node detects that it has entered a signal coverage blind spot or a high-interference area, it delays the wake-up process and remains in a hibernation cleaning mode to conserve power and clear the cache. Only when it enters a region with good signal strength does it quickly execute the three-stage wake-up procedure.

[0066] In such highly dynamic scenarios, the ability to transmit high-priority emergency data is particularly important. For example, collision avoidance warning information is encapsulated in specific microsecond-level ultra-short frames. These frames have absolute preemption rights at the media access control layer, capable of interrupting any ongoing cleaning pulse transmission or data compression operations. This design ensures that the real-time performance of traffic safety is 100% guaranteed even when the system is undergoing in-depth maintenance and optimization.

[0067] Example 4: In Example 4, the present invention is applied to a long-range, low-power agricultural environmental monitoring Internet of Things (IoT). Such systems are typically battery-powered, extremely sensitive to energy consumption, and their communication links are significantly affected by vegetation growth and weather.

[0068] In step 1, sensor nodes deployed in the farmland use the congestion sensing unit of the biomimetic routing module to monitor battery power and communication success rate. When the battery voltage drops, or when the retransmission rate increases due to rain or fog, the system automatically triggers a lymphatic dormancy mode. The redundancy identification unit, taking into account the slow changes in temperature and humidity data over long periods, identifies and marks redundant historical data within the range of numerical fluctuations by setting a large timestamp comparison window.

[0069] In step 2, due to the large node spacing, an adaptive enhancement strategy is adopted for the transmission power of the cleaning pulse packets. If no fluid feedback is received from neighboring nodes after sending two consecutive pulse packets, the node will gradually increase its power in 3-dB increments until a basic fluid guidance path is established. The guidance vector information also includes a "solar intensity" parameter, guiding the data flow towards nodes with solar energy replenishment and abundant energy, thereby achieving energy load balancing across the entire network.

[0070] In step 3, the data cleaning process focuses on preserving trends over very long periods. Validity assessment criteria check whether the data packet was triggered by an anomaly. For example, if soil moisture is below a critical value, the data packet will be considered core data and retained; while periodically reported data within the normal range, if its content is completely identical to the previous sampling point, will be selectively discarded. During compression, the dictionary encoder preloads a set of data model dictionaries commonly used in agricultural environments to further improve the compression ratio.

[0071] In step 4, the synaptic remodeling operation is achieved by fine-tuning the listening interval of the medium access control layer. For high-frequency and stable links, the corresponding listening period is shortened from 2 seconds to 500 milliseconds to reduce transmission latency; while for inefficient links with harsh environments and high packet loss rates, the listening period is lengthened to 10 seconds or even longer to enter a deep energy-saving state. This achieves dynamic sparsity processing of the network topology.

[0072] In step 5, the exit mechanism employs a combination of timed triggering and event triggering. Under normal circumstances, the node exits sleep mode and performs data upload according to a preset ultra-long period; however, upon detecting drastic environmental changes such as rainfall or strong light, a gradual wake-up procedure is immediately initiated, rapidly transmitting the cleaned and optimized cached data to reflect the latest environmental dynamics.

[0073] Through the detailed description of the above embodiments, it is clear that the wireless data recording and transmission method based on the Internet of Things described in this invention achieves proactive defense against network congestion and continuous optimization of the topology at the protocol level by simulating the biological mechanism of a lymphatic system. To quantitatively verify the technical effectiveness of this method, a paired control experiment was set up: a conventional wireless data recording and transmission method with fixed-period upload, ordinary local queues, and static link selection was used as the control group; the method described above, which enables cache effectiveness assessment, redundancy compression, link governance, and progressive wake-up mechanisms, was used as the experimental group. The two groups maintained complete consistency in node size, operating batches, and load conditions. A total of 720 paired records were collected, covering 120 operating batches and 6 node units. The overall comparison results of the three core technical effects are shown in Table 1. Table 1. Overall Comparison of the Effects of the Three Core Technologies

[0074] To assess the improvement in cache utilization, statistics were broken down into six independent node units to verify the repeatability of this effect across different hardware nodes, thus preventing the overall mean from being artificially inflated by a few abnormal nodes. The statistical results are shown in Table 2. Table 2 Comparison of Cache Effectiveness by Node Unit

[0075] To assess the reduction in transmission latency after wake-up, statistics were collected over 12 consecutive time windows to verify the temporal stability of this effect during continuous operation. The results are shown in Table 3. Table 3 Comparison of Transmission Delay After Wake-up by Time Window

[0076] To verify the significant reduction in data collision rate, aggregate statistics were performed across four different network region groups to validate the spatial consistency of this effect across different topological regions. The results are shown in Table 4. Table 4 Comparison of Data Collision Rates by Region Group

[0077] At the technical level, the present invention achieves its core function through the following logic: Congestion perception and pattern triggering logic: The system does not rely on a single buffer full-load signal, but instead predicts system pressure in advance by comprehensively considering channel utilization, queue growth rate, and the inherent redundancy of data content. By calculating the ratio of newly added data entropy to the channel's physical capacity per unit time, when this ratio consistently exceeds a preset parameter of 0.8, a preventative lymphatic-like sleep mechanism is initiated. This predictive mechanism avoids the passive situation of trying to remedy a deadlock after a node has entered a deadlock state.

[0078] The fluid dynamics-driven collaborative cleaning logic: The cleaning process is not an isolated node behavior, but rather a collective sensing achieved through low-power pulses. Fluid guidance vectors between nodes constitute a virtual potential energy field. Guided by the potential energy gradient, data packets spontaneously migrate from high-pressure nodes to low-pressure nodes. This migration is not disordered diffusion, but rather an efficient movement based on lossless dictionary encoding compression. During the migration, invalid data is discarded, while valid data is aggregated and rearranged.

[0079] The link self-healing logic based on synaptic remodeling treats communication links as resilient biological synapses. Each physical link is scored across multiple dimensions using long-term statistical data accumulated in a link performance log table. These scoring dimensions include, but are not limited to, average power spectral density fluctuation, maximum consecutive retransmissions, round-trip delay variance, and survivability under different weather / interference conditions. Based on the score, the system performs either "pruning" or "strengthening." This remodeling not only optimizes current transmission efficiency but also provides the system with predictive adaptability to future channel changes.

[0080] Gradual wake-up and emergency backup logic: The system addresses the issue of sudden network impact after large-scale hibernation through a phased recovery mechanism. In the wake-up sequence, control plane signaling takes precedence over business plane data, ensuring that network management logic is established before data flow. The preservation of the emergency channel ensures that the system does not lose its responsiveness to critical risks while pursuing efficient data cleaning. This dual-path design is logically isolated, ensuring that the emergency reporting function always maintains deterministic real-time performance regardless of how the cleaning mode evolves.

[0081] At the physical implementation level, this method can be embedded as a software module and deployed in the protocol stack of various IoT chips without rewriting existing hardware radio frequency units. The bionic routing module obtains channel quality parameters by calling the chip's underlying register interface and performs cache cleanup and compression through a standard memory management unit. This makes the solution highly versatile and low-cost to deploy.

[0082] In the data processing flow, each raw bitstream entering a node undergoes the following refined processing: First, it enters the ingress filter of the bionic routing module for initial redundancy identification; second, if sleep mode is triggered, it is sent to the compression engine for prefix merging based on the dynamic dictionary; third, based on the vector guidance parsed from the cleaning pulse, the data stream is redirected to the target address; fourth, in the synaptic reshaping unit, transmit time slots and modulation parameters are allocated according to the latest weight table. This series of processes ensures that the network can maintain a high effective data throughput even in extremely harsh communication environments.

[0083] In summary, this invention, through cross-disciplinary borrowing from glial cell regulation and lymphoid cleansing mechanisms, breaks away from the traditional "dormant is waiting" mindset in IoT protocols. By introducing proactive computing and communication operations during the dormant period, it achieves secondary development and deep recycling of network resources. This not only extends the lifespan of devices, but more importantly, it creates a dynamically evolving, self-healing intelligent network ecosystem. In practical applications, this transmission protocol with "metabolic" capabilities can reduce data collision rates in large-scale networks, improve system stability under extreme loads, and provide crucial technical support for the development of the IoT towards large-scale, highly reliable systems.

[0084] The following further explanation concerns the various parameters and their value logic involved in this invention. In all judgment and calculation logic, the system employs textual logical judgments rather than purely mathematical reliance. For example, regarding cache occupancy rate judgment, the system sets three levels: "Safe," "Warning," and "Dangerous." When the occupancy rate is at the "Safe" level, the node focuses on efficient data transmission; when it enters the "Warning" level, the redundancy identification unit begins high-frequency operation to search for compression space; when it enters the "Dangerous" level, a lymphatic dormancy mode is forcibly triggered. This hierarchical state machine control logic ensures the smoothness and stability of system switching.

[0085] The system prioritizes stability and applies dynamic fine-tuning when allocating link weights. If a link's transmission success rate fluctuates within 5% over 10 consecutive monitoring periods and has an average of over 90%, the system increases its weight by two levels. If the fluctuation exceeds 20%, even if the average success rate is acceptable, the weight is reduced by one level to mitigate potential link failure risks. These subtle adjustments accumulate to form the macroscopic manifestation of synaptic remodeling.

[0086] In the gradual wake-up phase after exiting sleep mode, the time ratio of the three sub-phases can be configured according to the specific application scenario. In industrial scenarios with high real-time requirements, the ratio is set to 1:1:2, i.e., rapid detection and rapid recovery; in agricultural scenarios where energy efficiency is prioritized, the ratio is set to 2:3:5, i.e., cautious detection and slow recovery. This flexibility ensures that the algorithm can adapt to various heterogeneous IoT application environments.

[0087] In the execution of the method of this invention, all data transformation, routing calculation, and link evaluation are completed in the local microcontroller unit of the node, giving full play to the advantages of edge computing and reducing dependence on the central server. This not only reduces the communication pressure on the core network but also improves the system's ability to respond quickly to local network congestion.

[0088] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A wireless data recording and transmission method based on the Internet of Things, characterized in that, Includes the following steps: Deploy a biomimetic routing module in an IoT node to continuously monitor network traffic status and cache utilization. When the data redundancy reaches the preset redundancy threshold or the cache occupancy rate reaches the predetermined upper limit, the bionic routing module sends a mode switching request to the main controller, triggering the IoT node to enter a lymphatic hibernation mode. After entering the lymphatic dormancy mode, the IoT node stops the data upload task of the regular business load and periodically generates and sends low-power cleaning pulse packets. The cleaning pulse packets carry fluid dynamics guidance information to guide adjacent nodes to coordinately adjust the data flow diffusion path. Based on the fluid dynamics guidance information transmitted by the cleaning pulse packet, each node performs a multi-dimensional validity evaluation of the data packets in its local cache, and compresses or selectively removes data packets that are determined to be invalid or redundant in order to free up cache storage space. After the cleaning process is completed, a synaptic remodeling operation is performed based on the usage frequency and transmission success rate of the historical communication path. The synaptic remodeling operation includes dynamically disconnecting inefficient communication links and allocating higher bandwidth weights to high-frequency stable links. Exit the lymphatic dormancy mode and enter the gradual wake-up phase to restore normal data recording and uploading functions. The cleaned and optimized data stream is then transmitted through the redesigned communication link.

2. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: The biomimetic routing module is equipped with a congestion sensing unit and a redundancy identification unit. The congestion sensing unit samples the received signal strength indication and channel idle assessment signal provided by the physical layer to calculate the proportion of time the channel is occupied per unit time. When the ratio exceeds a preset threshold of 0.75 for five consecutive sampling periods, and the number of data packets to be transmitted in the sending queue of the IoT node reaches 80% of the total buffer capacity, the congestion sensing unit generates a congestion trigger signal. During the triggering process, after the IoT node completes the transmission operation of the last complete data frame currently in progress, it locks the data generation interface of the application layer.

3. The wireless data recording and transmission method based on the Internet of Things according to claim 2, characterized in that: The redundancy identification unit determines whether there is duplicate or expired data by comparing the timestamps and content similarity of data packets. The redundancy identification unit extracts a feature fingerprint for each data packet entering the cache and performs a hash comparison between the feature fingerprint and the historical packet fingerprint stored in the scrolling window; When the content similarity of two data packets exceeds 95% and the timestamp interval between the two data packets is less than the preset lifespan, they are determined to be redundant data. The redundancy identification unit calculates the data redundancy by accumulating the number of redundant packets per unit time. When the data redundancy exceeds a preset threshold of 30%, a redundancy trigger signal is generated, causing the IoT node to enter the lymphatic dormancy mode.

4. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: The cleaning pulse packet adopts a simplified protocol format, which includes a source node identifier of bytes, a target area identifier of bytes, and two bytes of fluid guidance vector information; The IoT node controls its transmit power to the lowest level required to maintain connectivity with adjacent nodes within a one-hop range; The generation of the fluid guidance vector information is based on a fluid dynamics model, in which the data flow in the network is abstracted as an incompressible fluid, the communication bandwidth between nodes is abstracted as the cross-sectional area of ​​a pipe, and the buffer capacity is abstracted as local pressure. The IoT node calculates a gradient vector pointing to the high remaining storage space area based on its own and the pressure distribution of its neighboring nodes, and encapsulates the gradient vector in the cleaning pulse packet to diffuse it outward.

5. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: The validity assessment process is based on a criterion system consisting of the timeliness of the data packet, the uniqueness of its content, and the confirmation status of the receiving end. The IoT node maintains a time-to-live counter for each data packet in the cache; if the data packet does not receive an acknowledgment character from the receiver within a predetermined time, and the content of the data packet overlaps with the content of subsequent high-priority data packets by more than a preset ratio, or the service flow to which the data packet belongs has expired, then the data packet is marked as a removable object. The compression process employs a lossless dictionary encoding method. The IoT node maintains a local dynamic dictionary, replacing frequently occurring protocol headers or fixed data fields with index codes, and merging and storing data packets with the same prefix or structural characteristics.

6. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: The targeted removal process prioritizes cleaning up edge nodes that are far from the main path, based on the network topology. The IoT node identifies its centripetal degree in the network through a topology discovery protocol. The centripetal degree is the probability that the IoT node is on the shortest path from the data source to the aggregation point. When the centripetal force is lower than a preset probability threshold, the IoT node is determined to be an edge node. Upon receiving a cleaning instruction, the edge node prioritizes discarding data packets that have already been backed up by other redundant paths; During the cleaning process performed by the IoT nodes, the system defines a global congestion factor, which is obtained by weighted summation of the average buffer occupancy rate and the average channel collision probability of all nodes. The duration of the lymphatic dormancy mode increases with the increase of the global congestion factor, and the maximum value does not exceed 200 milliseconds.

7. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: The inefficient communication link refers to a link with an average transmission success rate of less than 60% over the past 20 communication cycles and a bandwidth utilization rate lower than a preset level. The IoT node dynamically updates the link status by maintaining a link performance log table, which records the retransmission count, average round-trip time, and signal-to-noise ratio fluctuation of each active link. When the overall score of a link is lower than a preset security threshold, the bionic routing module removes the link from the current route candidate list and performs synapse pruning.

8. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: The allocation of bandwidth weights is achieved by fine-tuning the physical layer modulation parameters or the medium access control layer time slot allocation strategy. The IoT node increases the priority parameter of the stable link when competing for the channel, allocates more continuous time slots for the stable link in the time division multiple access-based scheduling table, and increases the order of the modulation and coding scheme, switching the modulation mode from binary phase shift keying to sixteenth-order quadrature amplitude modulation to improve the instantaneous throughput of the physical layer. For links that fail due to periodic environmental interference, they are marked as dormant synapses and no bandwidth resources are allocated during reconstruction.

9. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: The progressive wake-up phase includes three sub-phases; In the first sub-phase, the IoT node restores 20% of the sensor sampling frequency and probes link activity with the lowest transmit power; In the second sub-phase, the IoT node increases the sampling frequency to 50% and adjusts the power control loop based on link feedback. In the third sub-phase, the IoT node fully recovers to normal operating status; During the lymphatic dormancy mode, the IoT node inserts a 50-microsecond instruction listening window in the middle of each cleaning cycle to capture downlink control messages from the control center.

10. The wireless data recording and transmission method based on the Internet of Things according to claim 1, characterized in that: During the cleaning and synaptic remodeling operations, the IoT node retains the ability to transmit high-priority urgent data. The high-priority emergency data is assigned the highest scheduling level and has preemption rights at the media access control layer. When the high-priority emergency data arrives, the IoT node immediately suspends the current cleaning or compression task and uploads the high-priority emergency data through the reserved channel using the reserved emergency deterministic time slot. The reserved channel uses an independent hardware buffer and priority queue, and its logical processing is not limited by the lymphatic dormancy mode.

Images