Industrial wireless network trusted scheduling method and device based on dynamic block chain

Abstract

The invention discloses an industrial wireless network credible scheduling method and device based on a dynamic block chain, and relates to the technical field of wireless network scheduling, and the method comprises the following steps: constructing a cross-node cache pressure real-time observation layer, carrying out the millisecond-level continuous collection of a cache queue fluctuation trajectory of each node in an industrial wireless network, and obtaining a cache queue fluctuation trajectory of each node in the industrial wireless network; a traceable cache pressure evolution baseline is generated in the observation layer; according to the method, causal coherent decomposition operation is performed based on a cache pressure evolution baseline, pseudo-random interference components generated by wireless link disturbance in observation data are separated, key factors causing cache pressure inversion are extracted, and a dynamic pressure mode graph reflecting a node cache state change rule is generated according to the key factors. According to the method, real-time identification and correction of cache pressure reversal in the industrial wireless network are realized through the dynamic block chain, and time continuity and space stability of scheduling are guaranteed; the task priority self-adaptive adjustment is realized under the energy and time joint optimization, and the stability, credibility and robustness of the system are improved.

Description

Technical Field

[0001] This invention relates to the field of wireless network scheduling technology, and specifically to a trusted scheduling method and apparatus for industrial wireless networks based on dynamic blockchain. Background Technology

[0002] Trusted scheduling of industrial wireless networks based on dynamic blockchain refers to introducing the immutability, decentralization, and traceability of blockchain into the resource allocation and scheduling process of wireless networks in complex industrial scenarios. Its core idea is to record key states such as node buffer pressure, link latency, and energy consumption through a dynamic blockchain mechanism, and to achieve task priority determination, spectrum and time slot allocation, and cross-node collaborative scheduling under the drive of smart contracts. Since link fluctuations and interference are unavoidable, the blockchain can provide trusted timestamps and consistency verification for the scheduling process, ensuring that scheduling decisions are not forged or tampered with. Simultaneously, with the help of traceable consensus results on the chain, scheduling paths can be quickly reconstructed under conditions of interference, latency, or node failure, achieving continuity and security in task execution. This method enables industrial wireless networks to possess a verifiable, fair, and robust scheduling mechanism under high dynamic and high reliability requirements, thereby improving the overall stability and trustworthiness of the system.

[0003] The existing technology has the following shortcomings: In dynamic industrial wireless network scenarios with frequent node switching, node buffer pressure may experience transient reversals within millisecond time slots. This means that a node initially under low load may suddenly switch to high load due to link disturbances, while a node under high load may abnormally switch to low load within the same time slot. Such transient reversals directly cause errors in the priority determination process of the blockchain scheduling contract, incorrectly allocating critical tasks to suboptimal nodes. This leads to disordered task execution, out-of-order writes to the state database, and ultimately disrupts the continuity and reliability of the scheduling link, severely impacting the real-time performance and security of the industrial wireless network.

[0004] 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

[0005] The purpose of this invention is to provide a trusted scheduling method and apparatus for industrial wireless networks based on dynamic blockchain, so as to solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a trusted scheduling method for industrial wireless networks based on dynamic blockchain, comprising the following steps: A cross-node cache pressure real-time observation layer is constructed to continuously collect the cache queue fluctuation trajectory of each node in the industrial wireless network at the millisecond level, and a traceable cache pressure evolution baseline is generated in the observation layer to characterize the dynamic trend of node cache pressure changes over time and identify potential triggering conditions for transient reversals. Based on the baseline of cache pressure evolution, causal coherence decomposition is performed to separate the pseudo-random interference components caused by wireless link disturbances in the observation data, extract the key factors that lead to the reversal of cache pressure, and generate a dynamic pressure pattern diagram that reflects the changing pattern of node cache state, providing a structured reference for subsequent predictive analysis. Based on the dynamic pressure pattern diagram, a shadow prediction sequence is constructed. Nodes that may experience cache pressure reversal within the prediction window are marked in advance. The shadow prediction sequence is then input into the blockchain scheduling process to form a redundant verification path for comparison and verification with the main scheduling path, which can be used for subsequent time-series consistency correction. Based on the shadow prediction sequence, double mirror time-scale anchor points are set up on the cache pressure change trajectory. The time series across nodes are uniformly rearranged, and the observed transient cache pressure is reversed and corrected into a continuous monotonic evolution trajectory, thereby ensuring the consistency of the reference data used for blockchain scheduling in the time dimension. A topological energy allocation landscape is constructed based on the monotonic evolution trajectory after unified rearrangement. The cache pressure state is mapped to quantified energy parameters. During the energy allocation process, the weight decay of the labeled inversion nodes is set to reduce the priority impact of risk nodes in the scheduling process and ensure the topological stability of energy allocation. Under the constraints of topological energy allocation landscape, trusted scheduling operations are performed. By comprehensively utilizing the uniformly rearranged time series and shadow prediction sequences, the task priorities in blockchain smart contracts are dynamically corrected, thereby achieving stable execution of the task chain and ensuring the trustworthiness of the scheduling process in an industrial wireless network environment with frequent switching of multiple nodes.

[0007] Preferably, the steps for constructing a cross-node cache pressure real-time observation layer include: Within the topology of the industrial wireless network, a data acquisition path is established for the buffer queues of all participating nodes. By embedding high-precision time synchronization acquisition endpoints at the buffer interfaces of each node and adopting a clock synchronization algorithm, the sampling time of each node is unified to the same millisecond-level reference frame, thereby obtaining a multi-dimensional state vector of node buffer pressure. The original data of the cache pressure of each node is subjected to time-series normalization and noise reduction. The sliding time window segmentation algorithm and weighted exponential smoothing function are used to reduce instantaneous jitter. The time consistency of cross-node pressure sampling points is achieved on the global timeline through a two-way time calibration strategy. Feature extraction and trend modeling are performed based on time-consistent cache pressure fluctuation trajectories. Gradient time series fitting algorithm is used to identify key inflection points of cache pressure and generate a cache pressure evolution baseline marked with on-chain timestamps. Set time differential monitoring points on the cache pressure evolution baseline, mark the reversal risk of intervals where the pressure change rate changes abruptly and is accompanied by link delay or data rate anomalies, and record them in the blockchain state log with timestamps to form a traceable cache pressure evolution record.

[0008] Preferably, in the step of setting time differential monitoring points on the cache pressure evolution baseline, for sudden changes in the cache pressure change rate from positive to negative or from negative to positive, a reversal risk range is determined by combining the comprehensive criteria of sudden increase in link latency and sudden decrease in data transmission rate, and this range is written into the blockchain state log in the form of encrypted timestamps.

[0009] Preferably, the steps for performing causal coherence decomposition operations based on the cache pressure evolution baseline include: After obtaining the cache pressure evolution baseline after time synchronization and denoising, causal relationship pre-modeling is performed on the cache pressure time series of each node, the conditional mutual information matrix between nodes is calculated, and an initial causal graph containing causal pointing relationships between nodes is constructed through delay Granger causality analysis. Under causal constraints, coherent feature decomposition is performed, and the multidimensional time series data of the cache pressure evolution baseline is mapped to the frequency domain space through short-time Fourier transform. The weighted principal component projection algorithm is used to separate the high-frequency interference components and low-frequency trend changes in a hierarchical manner. The decomposed cache pressure sequence is input into the causal weighted regression framework to calculate the relative contributions of cache input rate, output rate, link bandwidth fluctuation rate, energy consumption rate, and task queue burst rate to pressure changes, and to generate a node-level factor influence spectrum to identify key reversal factors. Key factors are uniformly expressed according to time and space dimensions. A three-dimensional dynamic stress pattern diagram is constructed through time sliding window clustering algorithm and causal intensity mapping. Combined with blockchain timestamp labeling, a traceable data entity is formed.

[0010] Preferably, in the process of constructing the three-dimensional dynamic pressure pattern diagram, a time sliding window clustering algorithm is used to aggregate nodes with similar causal relationships into coherent subgroups, and a pressure propagation manifold is generated within each coherent subgroup based on the causal strength and time phase difference, and synchronized and calibrated by a blockchain timestamp.

[0011] Preferably, the step of constructing a shadow prediction sequence based on the dynamic stress pattern map includes: A time-domain mapping of the future prediction window is established based on the temporal characteristics and causal correlation of the dynamic pressure pattern diagram. The prediction window length is determined by an adaptive time step algorithm, and a set of temporal evolution equations for the propagation of node buffer pressure is constructed based on the causal intensity matrix. Using the cached pressure trajectory in the dynamic pressure pattern diagram as the initial input, the future pressure state of the node is iteratively solved by combining the time-recursive autoregressive extended model with the time evolution equation set, and the shadow prediction sequence is obtained through the cooperative perturbation balance mechanism. The difference between the rate of change of node cache pressure and the historical evolution baseline is calculated based on the predicted sequence. For nodes whose predicted trends change abruptly and whose rate of change exceeds the threshold, the reversal risk index is calculated and labeled in advance. At the same time, the labeling confidence is improved through the cross-validation mechanism of the predicted trends of adjacent nodes. The shadow prediction sequence and annotation results are written to the blockchain verification chain in the form of timestamps and dynamically compared with the main scheduling path. When a deviation is detected between the two, a verification mechanism is triggered to correct the task priority.

[0012] Preferably, the step of deploying dual mirror time-scale anchor points on the cache pressure change trajectory based on the shadow prediction sequence includes: The time characteristics of the cache pressure change trajectory are analyzed based on the shadow prediction sequence. The sliding differential algorithm is used to calculate the cache pressure change rate and its second derivative. When the change rate changes abruptly and the second derivative exceeds the threshold, the reversal boundary point is determined. Forward and reverse mirror anchor points are established before and after the boundary point to form a double mirror anchoring structure. Using the double mirror timescale anchor point as a reference, the time offset between each node is calculated and a global time offset matrix is ​​constructed. Time slip correction is performed by combining linear interpolation and nonlinear phase fitting, so that the cache pressure change curve is synchronously aligned in the global time dimension. The time series is segmented using double mirror anchor points as boundaries. When a transient reversal segment is detected, the curve is corrected into a continuous monotonic evolution trajectory through mirror mapping and resampling, while maintaining the minimum mean square error between the corrected pressure change rate and the predicted rate. The uniformly rearranged cache pressure time series is mapped to the on-chain time domain according to the blockchain timestamp, a global time series index table of cache pressure evolution is established, and recorded as immutable time-series consistent data through the on-chain hash verification mechanism.

[0013] Preferably, the steps for constructing a topological energy distribution landscape based on the uniformly rearranged monotonic evolutionary trajectory include: After obtaining the monotonic evolution trajectory after unified rearrangement, the cache pressure state of each node is integrally transformed to calculate the energy density of cache load change per unit time, and the cache energy value is mapped to the standardized interval through energy normalization to form an energy state matrix. The correlation coefficient of buffered energy between nodes is calculated based on the energy state matrix, and a weighted directed graph is established. The directed graph is then mapped into a continuous energy flow field using the topological flow field modeling method, so that the buffered energy of each node is dynamically updated according to the energy inflow and outflow of adjacent nodes. In the energy flow field equation, a dynamic weight attenuation coefficient is set for the marked reversal nodes, and the attenuation magnitude is determined by combining the reversal risk index and the historical reversal frequency. At the same time, energy conservation constraints are introduced to maintain the total energy balance. The gradient flow convergence algorithm is used to iteratively optimize the energy distribution, calculate the gradient variance of the energy potential surface to verify the topological equilibrium state, and map the equilibrium energy distribution result to the blockchain scheduling parameter space and record it in the form of hash signature.

[0014] Preferably, the steps for performing trusted scheduling operations under the constraints of the topological energy allocation landscape include: The equilibrium results of the topological energy allocation landscape are coupled and mapped with the time series after unified rearrangement to establish a time-energy bidirectional binding mechanism. The energy weights of high-risk nodes are dynamically decayed according to the shadow prediction sequence to form the initial matrix of task priorities. The initial task priority matrix is ​​input into the blockchain smart contract as a parameter. The contract executes three stages in sequence: priority evaluation, node matching and task allocation. Based on the connection strength function of the topological energy landscape, the contract selects a target node with stable energy flow for the task. The scheduling results output by the blockchain smart contract are compared with the shadow prediction sequence on a task-by-task basis. When a timing deviation or energy distribution difference is detected, the scheduling correction factor is calculated and fed back into the contract logic to achieve dynamic priority correction. The scheduling calculation results, correction factors, and topological energy balance status are written into the blockchain ledger in the form of timestamps and node signatures, forming a traceable task execution time sequence.

[0015] The industrial wireless network trusted scheduling device based on dynamic blockchain includes a buffer pressure real-time observation module, a causal coherence decomposition and analysis module, a shadow prediction sequence generation module, a time-scaled anchor point rearrangement and correction module, a topology energy allocation construction module, and a trusted scheduling execution module. The real-time cache pressure observation module constructs a cross-node real-time cache pressure observation layer, continuously collects the cache queue fluctuation trajectory of each node in the industrial wireless network at the millisecond level, and generates a traceable cache pressure evolution baseline in the observation layer. The causal coherence decomposition analysis module performs causal coherence decomposition operations based on the buffer pressure evolution baseline. It separates the pseudo-random interference components caused by wireless link disturbances in the observation data, extracts the key factors that lead to buffer pressure reversal, and generates a dynamic pressure pattern diagram that reflects the changing pattern of node buffer state. The shadow prediction sequence generation module constructs a shadow prediction sequence based on the dynamic pressure pattern diagram, marks the nodes that may have cache pressure reversal within the prediction window in advance, and inputs the shadow prediction sequence into the blockchain scheduling process to form a redundant verification path for comparison and verification with the main scheduling path. The time-stamped anchor rearrangement correction module deploys double mirrored time-stamped anchors on the cache pressure change trajectory based on the shadow prediction sequence, performs unified rearrangement of the cross-node time series, and reverses and corrects the observed transient cache pressure into a continuous monotonic evolution trajectory. The topology energy allocation construction module constructs a topology energy allocation landscape based on the monotonic evolution trajectory after unified rearrangement, maps the cache pressure state to quantified energy parameters, and sets weight decay for labeled inversion nodes during the energy allocation process to reduce the priority impact of risk nodes in the scheduling process. The trusted scheduling execution module performs trusted scheduling operations under the constraints of the topological energy allocation landscape. It comprehensively utilizes the uniformly rearranged time series and shadow prediction series to dynamically correct the task priorities in the blockchain smart contract.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention introduces a dynamic blockchain-driven multi-level trusted scheduling mechanism to achieve real-time identification, prediction, and correction of transient cache pressure reversal phenomena in industrial wireless network environments, significantly improving the temporal consistency and spatial stability of task scheduling. By constructing a cross-node cache pressure real-time observation layer at the millisecond scale and combining a dual verification mechanism of causal coherence decomposition and shadow prediction sequences, this invention can predict node state reversal trends before task allocation, pre-marking and weighting high-risk nodes to avoid critical tasks being incorrectly assigned to nodes with unstable loads. It effectively suppresses state misordering and task disorder during node switching, ensuring that blockchain scheduling results maintain temporal continuity and execution chain integrity even under high-speed dynamic conditions, significantly improving the real-time performance and scheduling reliability of industrial wireless networks in complex environments.

[0017] This invention achieves joint reliable optimization of the energy and time domains under the constraints of topological energy allocation landscape. This allows for dynamic adjustment of task priorities to move beyond relying on a single cached state and instead make adaptive decisions based on a comprehensive feedback mechanism of unified rearranged time series, shadow prediction sequences, and topological energy distribution. Through energy mapping and weight decay mechanisms, the influence of risky nodes is automatically weakened, while the energy contribution of stable nodes is enhanced. This enables dynamic balancing and self-correction of the scheduling path under the constraints of blockchain smart contracts. Ultimately, even under conditions of frequent multi-node switching, link fluctuations, and task overlap, the system maintains the stability and reliability of the global scheduling topology, ensuring that the task execution order is verifiable, traceable, and robust. Attached Figure Description

[0018] 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.

[0019] Figure 1 This is a flowchart of the trusted scheduling method for industrial wireless networks based on dynamic blockchain according to the present invention.

[0020] Figure 2 This is a schematic diagram of the modules of the industrial wireless network trusted scheduling device based on dynamic blockchain of the present invention. Detailed Implementation

[0021] 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.

[0022] This invention provides, for example Figure 1 The trusted scheduling method for industrial wireless networks based on dynamic blockchain, as shown, includes the following steps: A cross-node cache pressure real-time observation layer is constructed to continuously collect the cache queue fluctuation trajectory of each node in the industrial wireless network at the millisecond level, and a traceable cache pressure evolution baseline is generated in the observation layer to characterize the dynamic trend of node cache pressure changes over time and identify potential triggering conditions for transient reversals. To address the transient reversal of cache pressure caused by frequent switching between multiple nodes, a cross-node cache pressure real-time observation layer needs to be built before scheduling execution. This layer enables millisecond-level continuous acquisition and characterization of the fluctuation trajectory of each node's cache queue, thereby providing an accurate, continuous, and traceable state baseline for subsequent dynamic blockchain trusted scheduling. This implementation specifically includes the following steps: Within the topology of the industrial wireless network, data acquisition paths are configured for the buffer queues of all participating nodes. By embedding high-precision time-synchronized acquisition endpoints at the buffer interface of each node, a clock synchronization algorithm is used to unify the sampling time of all nodes to the same millisecond-level reference frame, thus avoiding data distortion caused by sampling drift. During the establishment of the acquisition path, the buffer pressure of each node is defined as the ratio of the current buffer queue length to the upper limit of the buffer capacity, and combined with the node's input rate, output rate, and instantaneous link bandwidth changes to form a multi-dimensional buffer state vector. The observation layer uses this state vector to quantify the instantaneous buffer pressure of the nodes on a millisecond-level time scale, thereby providing an accurate pressure scale for subsequent continuous acquisition.

[0023] After synchronously collecting the cache pressure data from each node, the observation layer needs to perform time-series normalization and denoising on the raw cache pressure data from each node. To this end, the observation layer applies a sliding time window segmentation algorithm to the cache pressure sequence of each node, dividing continuous millisecond-level data into overlapping time window intervals, and employing a weighted exponential smoothing function within each interval to reduce the impact of instantaneous jitter. Next, to address asynchronous sampling errors between nodes, the observation layer uses a bidirectional time calibration strategy to relocate the pressure sampling points of each node on the global timeline, thereby ensuring strict consistency in cache pressure changes across different nodes within the same time reference frame. This process ensures that the fluctuation trajectory of cache pressure is comparable and concatenable in both spatial and temporal dimensions.

[0024] After obtaining time-consistent and denoised cache pressure fluctuation trajectories, the observation layer needs to perform continuous feature extraction and trend modeling on these trajectories to generate a traceable cache pressure evolution baseline. Specifically, the observation layer employs a multidimensional autoregressive sliding motion analysis method to jointly analyze the cache pressure change rate, fluctuation amplitude, stable duration, and abrupt inflection points for each node within continuous time slots. Based on this, the observation layer uses a gradient time series fitting algorithm to identify key inflection points where cache pressure transitions from an increasing to a decreasing range or vice versa, and constructs a dynamic evolution curve of cache pressure based on the distribution of these inflection points along the time axis. This evolution curve exhibits the macroscopic evolution trend of node cache states at different time scales and is calibrated using on-chain timestamps, thus forming a traceable cache pressure evolution baseline. This baseline not only reflects the long-term trend of cache load for each node but also preserves microscopic instantaneous fluctuation information, enabling the system to identify and predict abnormal cache pressure reversal behavior in subsequent steps.

[0025] After establishing the cache pressure evolution baseline, the observation layer needs to identify and label the potential triggering conditions for transient reversals based on this baseline. To this end, the observation layer sets continuous time-differential monitoring points on the evolution baseline and dynamically monitors the cache pressure change rate of each node. When the observation layer detects that the pressure change rate rapidly changes from positive to negative or vice versa within a short time window, and this change is accompanied by sudden anomalies in link latency or data transmission rate, the observation layer determines that there is a potential risk of transient cache pressure reversal within that period. Subsequently, the observation layer records this risk interval in the form of a timestamp and writes it to the on-chain state log using the immutability of blockchain, thus ensuring that each transient reversal event has a verifiable record and a traceable historical trajectory. This process not only enables subsequent causal coherence decomposition operations to be accurately calculated based on clear reversal trigger boundaries but also provides a true and reliable state reference for the trusted scheduling of the entire industrial wireless network. Through the above four consecutive steps, the observation layer achieves dynamic synchronous observation of cross-node cache pressure, continuous feature extraction, trend modeling, and potential anomaly identification on a millisecond-level time scale. Ultimately, it forms a cache pressure evolution baseline that is continuous in time, verifiable in data, and interpretable in semantics, providing accurate basic data support for subsequent dynamic blockchain trusted scheduling.

[0026] The above implementation details the construction of the cross-node buffer pressure real-time observation layer and its specific operation in industrial wireless networks. This method not only solves the problem that traditional monitoring methods cannot capture rapid fluctuations in buffer pressure with millisecond-level accuracy, but also ensures the consistency and traceability of cross-node data through time synchronization and noise reduction calibration, fundamentally improving the accuracy and reliability of status observation in industrial wireless networks.

[0027] Based on the baseline of cache pressure evolution, causal coherence decomposition is performed to separate the pseudo-random interference components caused by wireless link disturbances in the observation data, extract the key factors that lead to the reversal of cache pressure, and generate a dynamic pressure pattern diagram that reflects the changing pattern of node cache state, providing a structured reference for subsequent predictive analysis. To extract the essential factors leading to transient reversals from the complex process of cache pressure evolution, after constructing the cache pressure evolution baseline, a causal coherence decomposition operation is performed on the multi-node observation data contained in the baseline. This operation effectively separates pseudo-random noise components caused by wireless link disturbances, inter-node interference, or environmental fluctuations, thereby extracting the decisive cache pressure evolution factors and constructing a dynamic pressure pattern diagram reflecting the changing patterns of cache state. This implementation includes the following sub-steps: After obtaining the buffer pressure evolution baseline after time synchronization and denoising, causal pre-modeling is performed on it at the observation layer to clarify the dependency paths of buffer pressure changes. To this end, firstly, based on the time series of buffer pressure at each node on the evolution baseline, the conditional mutual information matrix between nodes is calculated to quantify the temporal dependency strength between nodes. Subsequently, the delay Granger causality analysis method is used to explicitly model the sequential dependencies between the buffer pressure sequences of each node, thereby constructing an initial causal graph containing the causal pointing relationships between nodes. This causal graph not only describes the direction of the impact of a node's buffer pressure change on the buffer states of other nodes, but also reveals the propagation path of wireless link disturbances in the network topology. Through this causal pre-modeling step, the system provides structured causal constraints for subsequent coherent decomposition operations, ensuring that the decomposition process can identify the true relationship between interference and inversion under the premise of causal consistency.

[0028] Based on causal constraints, coherent feature decomposition and interference identification operations are performed to address the numerous pseudo-random interference components present in the cache pressure evolution baseline. Specifically, the multi-dimensional time-series data of the cache pressure evolution baseline is mapped to the frequency domain representation space, and the energy distribution of pressure fluctuations in different frequency bands is extracted using short-time Fourier transform. Subsequently, based on the causal constraint matrix obtained in the previous sub-step, a weighted principal component projection algorithm is used to group the frequency domain energy distribution according to causal paths, performing hierarchical separation of high-frequency transient fluctuations and low-frequency trend changes. Through this hierarchical coherent decomposition process, high-frequency pseudo-random components caused by external link interference or random data congestion can be removed from the evolution trajectory, retaining only the low-frequency stable changes directly related to the node's internal operating state. This separation process not only significantly improves the signal-to-noise ratio of the cache pressure curve but also makes subsequent factor extraction more physically interpretable.

[0029] After obtaining the purified evolution signal after interference separation, causal weight extraction is performed on the signal to identify the key factors that truly lead to cache pressure reversal. Specifically, the node cache pressure sequence after coherent decomposition is input into the causal weight regression framework, and the relative contribution of each influencing factor (including cache input rate, output rate, link bandwidth fluctuation rate, node energy consumption rate, and task queue burst rate) to the cache pressure change is calculated. This regression process is based on a joint estimation mechanism of weighted least squares and Bayesian confidence intervals to ensure that the obtained causal weights reflect both statistical significance and maintain causal coherence over time. Subsequently, the causal weight results of all nodes are summarized, and a node-level factor influence spectrum is generated by time-weighted superposition, thereby identifying the core influencing factors that dominate cache pressure reversal within a specific time period. For example, when the cache input rate fluctuation and link delay change of a node show phase reversal within a highly correlated interval, the system can determine that the node is in a potential reversal state. Through this causal weight extraction mechanism, the root causes of reversal events can be captured on a millisecond time scale, providing both physical and statistical evidence for subsequent predictions.

[0030] After identifying the key factors for cache pressure reversal at each node, these factors are uniformly expressed in temporal and spatial dimensions to generate a dynamic pressure pattern diagram that can be used for subsequent predictive analysis. To this end, the causal weight spectra of each node are first smoothly overlapped along the time axis, and a time-sliding window clustering algorithm is introduced to aggregate nodes with similar causal relationships within consecutive time periods into coherent subgroups. Then, within each coherent subgroup, a pressure propagation manifold is constructed based on the causal influence intensity and temporal phase difference, mapping the influence direction, intensity, and lag relationship between nodes to a visualized three-dimensional parameter space. The coordinate axes of this three-dimensional space represent time, causal intensity, and cache pressure amplitude, respectively, and the trajectory curves within the space constitute the core structure of the dynamic pressure pattern diagram. This pattern diagram reveals the evolution and propagation law of cache pressure among different nodes through multi-layer mapping, and can intuitively present the pressure reversal path driven by key factors. Furthermore, this dynamic pressure pattern diagram is synchronized with the blockchain's on-chain timestamps to form a traceable data entity, providing a structured reference and physical constraints for subsequent shadow prediction sequence generation. Through this step, the buffer pressure evolution baseline is transformed from the original time series form into a dynamic pressure pattern with causal correlation, spatial coherence and predictability, so that the buffer state change pattern of industrial wireless networks can be identified, understood and utilized in a structured form.

[0031] This implementation method achieves the transformation from the buffer pressure evolution baseline to a dynamic pressure pattern diagram through causal coherence decomposition. This process not only overcomes the technical bottleneck of traditional noise filtering in distinguishing between random interference and genuine inversion signals, but also, by introducing causal constraints and weight extraction mechanisms, makes the identification of inversion causes more accurate and interpretable. The resulting dynamic pressure pattern diagram comprehensively expresses the node buffer state change patterns in three dimensions: time, frequency, and causality, providing a data foundation for the subsequent establishment of shadow prediction sequences and the dynamic correction of reliable scheduling decisions.

[0032] Based on the dynamic pressure pattern diagram, a shadow prediction sequence is constructed. Nodes that may experience cache pressure reversal within the prediction window are marked in advance. The shadow prediction sequence is then input into the blockchain scheduling process to form a redundant verification path for comparison and verification with the main scheduling path, which can be used for subsequent time-series consistency correction. To identify and correct potential cache pressure reversal risks before task scheduling, after constructing the dynamic pressure pattern graph, a shadow prediction sequence parallel to the main scheduling path is generated based on this graph. This shadow prediction sequence proactively extrapolates cache pressure changes within future time windows, pre-marks nodes prone to reversal, and incorporates the prediction results into the blockchain scheduling process to form redundant verification paths, thus providing a comparable reference trajectory for subsequent time-series consistency correction. This implementation includes the following sub-steps: After obtaining the dynamic stress pattern diagram, a time-domain mapping for the future prediction window is established based on the temporal characteristics and causal relationships of the diagram. Specifically, the stress change trajectory of each node in the dynamic stress pattern diagram is unfolded along the time axis, and the length of the future prediction window is determined by an adaptive time step algorithm. The boundary of the prediction window is jointly determined by the average cache update cycle of the node and the stable link state duration, ensuring that the prediction time range covers critical periods where reversals may occur while avoiding redundant calculations. Subsequently, based on the causal intensity matrix in the dynamic stress pattern diagram, a propagation equation for cache stress changes is established, representing the influence path between nodes in temporal differential form, thus forming a set of time evolution equations for generating shadow prediction sequences. In this set of equations, the future state of node cache stress is not only related to its current stress value but also coupled with the key reversal factors identified in the previous stage (including input rate fluctuations, link delay changes, and energy consumption differences), enabling the prediction process to maintain consistency with the actual physical process.

[0033] After the prediction window is determined, forward evolutionary inference is performed on the node cache trajectories in the dynamic pressure pattern graph to generate initial estimates of the shadow prediction sequence. Specifically, the cache pressure evolution trajectory in the dynamic pressure pattern graph is used as the initial input, and a time-recursive autoregressive extended model is used to predict and solve it. This prediction and solution process combines the time evolution equations established in the previous stage to iteratively calculate the pressure changes of each node in future time steps. To improve the stability and robustness of the prediction, a cooperative perturbation balancing mechanism is introduced. That is, in each prediction iteration, the differential gradients of the main pressure trajectory and the reverse perturbation trajectory are calculated simultaneously. By balancing the offset directions of the two, the error amplification effect caused by link noise or temporary interference is suppressed. The final shadow prediction sequence maintains a continuous correspondence with the original pressure evolution baseline on the time axis, enabling the system to obtain trend estimates of cache pressure changes within the future window and providing a quantitative basis for subsequent node labeling.

[0034] After obtaining a preliminary estimate of the shadow prediction sequence, risk assessment and node pre-labeling are performed on the prediction results to identify which nodes may experience cache pressure reversal within the prediction window. To this end, a difference analysis is performed between the rate of change of cache pressure in the prediction sequence and the rate of change of the historical evolution baseline, and the phase shift of the prediction sequence in the time dimension is calculated. When a node's predicted pressure curve shows a sudden change from a downward trend to an upward trend within a short period, or its rate of change exceeds a preset threshold, the node is determined to have potential reversal risk. Subsequently, based on the intensity of the change in the prediction curve, the duration of the reversal, and the degree of causal coupling, the reversal risk index of the node is calculated, and nodes with a risk index higher than the threshold are pre-labeled. To prevent misjudgment, a confidence backtracking mechanism is introduced during the labeling process. This involves cross-validating the target node using the prediction states of adjacent nodes in the dynamic pressure pattern graph. If the prediction trends of adjacent nodes are consistent and the temporal phase correlation is strong, the labeling confidence is increased; otherwise, the confidence is decreased. This method enables accurate identification and labeling of potential cache pressure reversal nodes during the prediction stage, thus providing early warning information for subsequent scheduling verification.

[0035] After pre-labeling nodes, the shadow prediction sequence is integrated with the blockchain scheduling process to form a redundant verification path for comparison and verification with the main scheduling path. Specifically, the state and labeling results of each node in the shadow prediction sequence are written to the parallel verification chain of the blockchain in the form of timestamps to ensure the immutability and traceability of the prediction data. During the blockchain consensus execution phase, the system simultaneously loads the main scheduling path and the shadow prediction path, and dynamically compares them in terms of node priority, task allocation order, and cache state changes. When a significant deviation is detected between the task allocation results of the main scheduling path and the shadow prediction path, the system triggers a verification mechanism to recalculate the task priorities of the relevant nodes in the main path to avoid scheduling errors caused by transient reversals. In addition, to ensure consistency in the time dimension, each time step of the shadow prediction sequence corresponds to a specific block number in the blockchain. Through the block timestamp synchronization mechanism, the time alignment between the main path and the shadow path is maintained, thus providing a verifiable comparison basis in the subsequent time-series consistency correction phase. Through this redundant verification mechanism deeply coupled with blockchain, the system can detect potential anomalies before scheduling and execution, correct task priorities in advance, and effectively prevent task reordering, link disorder and trustworthiness damage caused by cache pressure reversal.

[0036] This implementation method achieves early identification and verification of potential buffer pressure reversal risks in industrial wireless networks by generating shadow prediction sequences based on dynamic pressure pattern graphs and constructing blockchain redundancy verification paths. The method achieves millisecond-level time precision in predicting future states within the prediction window. Combined with a causal propagation model and confidence labeling mechanism, the shadow prediction results are both forward-looking and verifiable.

[0037] Based on the shadow prediction sequence, double mirror time-scale anchor points are set up on the cache pressure change trajectory. The time series across nodes are uniformly rearranged, and the observed transient cache pressure is reversed and corrected into a continuous monotonic evolution trajectory, thereby ensuring the consistency of the reference data used for blockchain scheduling in the time dimension. To eliminate the time-series disorder caused by transient cache pressure reversals between nodes, a time-stamping anchoring mechanism needs to be constructed based on the shadow prediction sequence. This mechanism enables unified alignment and reversal correction across nodes in the time dimension. By deploying dual-mirror time-stamping anchors along the cache pressure change trajectory and uniformly rearranging the time series of multiple nodes based on the anchor information, the transient reversal phenomenon in the original observation data can be transformed into a continuous and monotonic evolution trajectory, thereby ensuring that the input data used for trusted blockchain scheduling maintains strict consistency in the time dimension. This implementation includes the following sub-steps: Based on the shadow prediction sequence generated in the previous stage, the temporal characteristics of the cache pressure change trajectory are precisely analyzed to determine the placement of time-stamped anchor points. Specifically, the cache pressure change rate of each node in the shadow prediction sequence is used as the main analysis variable, and its first and second time derivatives over continuous time periods are calculated using a moving differential algorithm. When a sudden change in the sign of the cache pressure change rate is detected and the second derivative exceeds a threshold, the time position can be identified as a potential reversal boundary point. To ensure the physical consistency of the anchor points, forward mirror anchor points and reverse mirror anchor points are established before and after each reversal boundary point, forming a dual-mirror anchoring structure on the time axis. The forward mirror anchor point is used to record the initial state of cache pressure changing from low load to high load, and the reverse mirror anchor point is used to record the recovery state of cache pressure changing from high load to low load. The dual mirror anchor points are bound to the node identifier of the shadow prediction sequence through timestamp indexes, thereby establishing a mappable temporal reference framework in multi-node scenarios, providing an alignment basis for subsequent time reordering.

[0038] After the dual-mirror time-stamp anchor points are deployed, the time series of each node are synchronized using these anchor points as references to achieve initial alignment across the time domain of each node. To this end, the time offset between anchor points of different nodes is first calculated, and a global time offset matrix is ​​constructed based on the order in which the anchor points appear. This matrix, with nodes as rows and time anchor points as columns, records the time phase difference of each node in each reversal event. Subsequently, based on this time offset matrix, a time-slip correction algorithm is executed, using a combination of linear interpolation and nonlinear phase fitting to reproject the time series of all nodes onto a unified global reference timeline. To avoid distortion introduced by time interpolation, a local curvature preservation constraint is applied to the pressure change curve of each node during the correction process, ensuring that the curve retains the smoothness and continuity of its original trend after correction. After this step, the time misalignment caused by asynchronous sampling and link delay is effectively eliminated, and the buffer pressure change trajectories of all nodes are synchronized in the global time dimension.

[0039] After time alignment, transient cache pressure reversals are identified and corrected using dual-mirror time-stamp anchors, transforming discontinuous pressure changes into monotonic evolution trajectories. Specifically, the cache pressure time series of each node is segmented, with the dual-mirror anchors serving as boundaries, dividing the series into multiple time segments. Within each segment, the monotonicity index of pressure changes is calculated. When a reverse jump in pressure values ​​is detected within a segment, it is identified as a transient reversal. At this point, through the mirror correspondence between anchors, the time series of the reversal segment is mirrored and resampled. That is, starting from the forward mirror anchor and ending at the reverse mirror anchor, the time series of the reversal interval is folded along the mirror axis and then re-unfolded, so that the curve with the original abrupt rise-fall is smoothly corrected into a continuous monotonic rise or fall curve. To ensure the physical consistency of the correction results, a constraint function is introduced during the mirror resampling process, ensuring that the corrected pressure change rate maintains the minimum mean square error difference with the prediction rate of the shadow prediction sequence, thus ensuring that the corrected curve is not only smooth and continuous but also truly reflects the node's operating trend. Through this mirror correction mechanism, transient reversals are transformed into a continuous evolutionary process, effectively eliminating the scheduling misdirection risk caused by reversals.

[0040] After mirroring and correcting the cache pressure trajectories of all nodes, these rearranged time series are globally integrated to form temporally consistent reference data that can be directly used for blockchain scheduling. Specifically, the uniformly rearranged cache pressure time series of each node are remapped to the on-chain time domain according to the blockchain timestamp, ensuring that each pressure state change point corresponds one-to-one with the block time. During this mapping process, a dynamic time warping algorithm based on anchor intervals is used to uniformly adjust the time step across nodes, ensuring that the number of data points for all nodes within the same on-chain time period remains consistent. Subsequently, based on the temporal dependencies between nodes, a global time series index table of cache pressure evolution is established, and this index table is immutably recorded through an on-chain hash verification mechanism. This index table serves as the input benchmark for blockchain scheduling, ensuring that the cache pressure data upon which all scheduling decisions are based has complete consistency and traceability in the time dimension. Finally, the temporal disorder caused by transient reversals in the observed data is completely corrected, forming a continuous, monotonic, and globally unified evolution trajectory, providing a reliable temporal basis for subsequent energy allocation and task priority adjustment.

[0041] This implementation achieves unified alignment and transient reversal correction of cross-node time series through a dual-mirror time-stamp anchor point deployment and time rearrangement mechanism based on shadow prediction sequences. This method not only solves the time misordering problem caused by asynchronous node sampling and transient fluctuations in traditional industrial wireless networks, but also transforms the complex reversal behavior into an interpretable monotonic evolution process through mirror mapping and time warping algorithms, thereby achieving global consistency in the time dimension during blockchain scheduling. By precisely mapping the rearranged time series to the blockchain timestamps, this method provides a highly synchronized reference data foundation for trusted scheduling, ensuring that scheduling decisions maintain strict order and reliability even in environments with frequent multi-node switching.

[0042] A topological energy allocation landscape is constructed based on the monotonic evolution trajectory after unified rearrangement. The cache pressure state is mapped to quantified energy parameters. During the energy allocation process, the weight decay of the labeled inversion nodes is set to reduce the priority impact of risk nodes in the scheduling process and ensure the topological stability of energy allocation. To achieve stable allocation of task resources under conditions of frequent multi-node switching and dynamic changes in cache state, after completing unified rearrangement and transient reversal correction of cross-node time series, a topological energy allocation landscape that reflects the coupling relationship of cache pressure between nodes and the energy distribution law is further constructed based on the monotonic evolution trajectory after unified rearrangement. This topological energy allocation landscape achieves dynamic balance and global stability of the energy scheduling topology of the industrial wireless network by mapping cache pressure state to quantified energy parameters and combining a weight decay mechanism for labeled reversal nodes. This implementation includes the following sub-steps: After obtaining the monotonic evolution trajectory after unified rearrangement, the buffer pressure state of each node is expressed in terms of energy to form the basic measurement system of the topological energy landscape. Specifically, the buffer pressure change curve of each node in continuous time slots is integrally transformed to calculate its buffer load change energy density per unit time. This energy density reflects the combined effect of buffer potential energy and transmission kinetic energy of the node in the data flow process. Then, to ensure the comparability of energy parameters between different nodes, energy normalization is required. The buffer energy value of each node is mapped to the standardized interval [0, 1], and the link delay and signal quality between nodes are used as weighting coefficients to correct the difference in energy distribution between nodes, so that the difference in energy distribution between nodes can accurately reflect their communication reliability in the physical topology. Through the above process, an energy state matrix is ​​formed with node buffer pressure as the dominant variable and energy density as the quantitative feature. This matrix serves as the underlying expression structure of the topological energy distribution landscape and provides a mathematical basis for subsequent energy correlation modeling.

[0043] Based on the energy state matrix, topological relationships between nodes are established, thereby constructing a topological structure of the energy distribution landscape in space. Specifically, firstly, the cached energy correlation coefficient between any two nodes is calculated. This correlation coefficient is determined by the similarity of the energy change rates of the two nodes and the spatial distance decay function. Subsequently, a weighted directed graph is constructed based on the cached energy correlation coefficient, treating nodes as energy sources or sinks, with the weights of directed edges between nodes representing the energy transfer intensity. To characterize the energy evolution trend of the entire network, a topological flow field modeling method is adopted, mapping the weighted directed graph into a continuous energy flow field, allowing the cached energy of each node to be dynamically updated through the energy inflow and outflow of adjacent nodes. This topological flow field visually presents a multidimensional potential energy surface of energy distribution, with its local gradient changes corresponding to the rate of change of node cache pressure. At this point, the overall energy landscape of the network forms an analyzable spatial topological structure, which can reflect the energy interactions between nodes and express the dynamic evolution path of cache pressure, providing spatial constraints for the next step of weight adjustment.

[0044] After establishing the topological energy allocation landscape, a weight decay mechanism for labeled inverted nodes is introduced during the energy allocation process to prevent risky nodes from causing instability in the scheduling process. Specifically, inverted nodes pre-labeled in the previous stage's shadow prediction sequence are introduced into the current energy landscape model, and their energy output weights are dynamically decayed in the energy flow field equations. The decay coefficient is determined jointly based on the risk index of the inverted node and its historical inversion frequency. When a node has a high inversion risk index or a large number of historical inversions, the decay coefficient increases accordingly, thereby reducing the node's weight contribution in the energy transmission network. To ensure the overall balance of energy distribution, an energy conservation constraint is also introduced, i.e., adaptive compensation is applied to the energy output weights of other non-inverted nodes, keeping the total energy of the entire topological energy flow field constant. Through this process, the system can actively isolate potential interference from risky nodes at the energy level, thereby preventing incorrect task priority allocation due to local energy mutations in subsequent blockchain scheduling calculations. This mechanism makes the topological energy allocation not only self-stable but also fault-tolerant to abnormal nodes.

[0045] After achieving energy weight decay and compensation balance, the topological energy distribution landscape undergoes global optimization and dynamic stability verification to ensure topological continuity and scheduling reliability. To this end, a gradient flow convergence algorithm is used to iteratively evolve the energy flow field multiple times, causing the energy distribution to tend towards a smooth and stable state. In each iteration, the overall gradient variance of the energy potential surface is calculated. When the variance converges to below a preset threshold, the energy distribution landscape is considered to have reached topological equilibrium. Subsequently, the energy distribution in equilibrium is mapped to the blockchain's scheduling parameter space, using energy parameters as task priority correction factors. This allows the scheduling decision to automatically adjust the task allocation order based on energy weights during blockchain smart contract execution. Simultaneously, to ensure the verifiability of topological stability, the final equilibrium energy distribution result is written to the blockchain ledger in the form of a hash signature, ensuring traceability and immutability. Through these processes, the constructed topological energy distribution landscape not only achieves balanced energy transmission between nodes but also ensures the transparency and trustworthiness of the energy distribution process through an on-chain notarization mechanism. The resulting topological energy landscape becomes the energy constraint benchmark for blockchain-based trusted scheduling, ensuring that the scheduling results are stable in spatial structure, continuous in time dimension, and verifiable in terms of credibility.

[0046] This implementation transforms the monotonic evolution trajectory, after unified rearrangement, into a topological energy allocation landscape, thereby realizing the quantitative energy expression and topological allocation management of buffer pressure states in industrial wireless networks. This method utilizes innovative techniques such as energy density mapping, topological flow field modeling, weight decay, and energy conservation balance to construct an energy scheduling structure that combines self-stability and traceability. By weakening the influence of inverted nodes and strengthening the contribution of steady-state nodes in the energy dimension, the system can maintain the stability of the global scheduling topology when facing dynamic load fluctuations and node switching, providing an energy-constrained basis for subsequent reliable scheduling calculations.

[0047] Under the constraints of topological energy distribution landscape, a trusted scheduling operation is performed. By comprehensively utilizing the uniformly rearranged time series and shadow prediction series, the task priority in the blockchain smart contract is dynamically corrected, thereby achieving stable execution of the task chain and ensuring the trustworthiness of the scheduling process in an industrial wireless network environment with frequent switching of multiple nodes. To ensure the continuity and reliability of task allocation and execution in environments with frequent multi-node switching, a reliable scheduling operation is performed under the constraints of the topological energy allocation landscape. This scheduling operation uses a uniformly rearranged time series as the time consistency benchmark and shadow prediction sequences as the redundancy verification reference. Dynamic adjustments to task priorities are implemented through blockchain smart contracts, thereby achieving global stable control of the scheduling chain within the three-dimensional space of energy, time, and prediction. This implementation includes the following sub-steps: After the topological energy allocation landscape is established, the balance results of the energy landscape are coupled and mapped with the temporal rearrangement sequence to form the constraint boundary for scheduling computation. Specifically, the energy weights of nodes in the topological energy landscape are synchronously matched with the monotonic evolution time series after unified rearrangement, so that each time step corresponds to an energy allocation state. Through this time-energy bidirectional binding mechanism, the scheduling computation process is ensured to have strict temporal sequence consistency and spatial energy topological constraints. Subsequently, based on the inverted node information pre-labeled in the shadow prediction sequence, its weight factor in the energy landscape is dynamically decayed, thereby reducing the scheduling participation of high-risk nodes before scheduling begins. At this point, the task scheduling priority of each node is no longer determined by a single cache load or task arrival rate, but by a comprehensive weighted average of temporal evolution trend, energy stability, and predicted risk, constructing a multi-dimensional collaborative initial task priority matrix. This matrix serves as the input benchmark for scheduling computation, providing structured constraints for subsequent reliable computation.

[0048] After establishing the scheduling input matrix, task scheduling operations based on blockchain smart contracts are executed. Specifically, the initial task priority matrix is ​​written into the blockchain smart contract body as a contract trigger parameter, and the scheduling calculation process is executed simultaneously through consensus nodes. The core logic of the smart contract includes three stages: priority evaluation, node matching, and task allocation. In the priority evaluation stage, the contract calculates the urgency of each task and the available energy of the nodes in real time based on the dynamic changes of energy weights and time series, forming a priority correction coefficient. In the node matching stage, the contract selects the target node with the most stable energy flow for the task based on the connection strength function of the topological energy landscape. In the task allocation stage, the system automatically avoids risky nodes based on the labeling information of high-risk nodes in the predicted sequence, and allocates key tasks to stable nodes or multi-node collaborative node groups to enhance the execution continuity of the task chain. This blockchain-driven scheduling operation has decentralized and tamper-proof characteristics, ensuring that every task allocation decision is jointly confirmed by the on-chain consensus mechanism, thereby making the scheduling process credible and unforgeable at the algorithm execution level.

[0049] After completing the real-time scheduling of blockchain smart contracts, the scheduling results are compared with and dynamically corrected using shadow prediction sequences to achieve continuous adaptive optimization of scheduling priorities. Specifically, the task execution order of the current scheduling operation is compared task-by-task with the predicted order within the same time window in the shadow prediction sequence, calculating the temporal deviation and energy distribution difference between the two. When a significant deviation is detected between the actual execution order of a task and the prediction result of the shadow prediction sequence, the system initiates a dynamic correction mechanism. This mechanism calculates a scheduling correction factor based on the time series offset and energy weight difference, and feeds it back to the blockchain contract logic, automatically adjusting the task priority weights during the next round of consensus execution. This process is equivalent to forming a "dynamic closed-loop correction circuit" on the scheduling chain, enabling the system to continuously correct the priority model based on real-time execution feedback. To prevent the correction process from introducing new uncertainties, the system performs an on-chain stability verification after each correction operation, ensuring that the corrected scheduling result remains in a topological equilibrium state through joint convergence detection of energy potential variance and time offset mean. Through this comparison and correction mechanism, the system achieves continuous self-learning and self-calibration of the task scheduling process, enabling the scheduling priority to be dynamically optimized in real time with the network.

[0050] After dynamic correction and verification, the entire scheduling process is solidified and recorded on-chain to ensure the traceability and verifiability of task execution. Specifically, the scheduling calculation results, priority correction factors, shadow prediction deviation values, and topological energy balance state are packaged into a set of scheduling credential data, and timestamps and node signature information are added and written into the blockchain ledger. Each scheduling credential corresponds to a unique block number, thus forming a time-series mapping of task execution on the blockchain. This mapping can be used for subsequent time-series consistency checks and task backtracking verification, ensuring that the system can restore the true scheduling state based on on-chain records in the event of any abnormal interruption or malicious tampering at any node. Simultaneously, to ensure the continuous execution of the task chain, the system immediately triggers the prediction update process for the next round of task allocation after the scheduling credential is generated, using the latest scheduling results as the correction input for the shadow prediction sequence, achieving dynamic coupling between the prediction chain and the execution chain. Through this reliable solidification and self-updating mechanism, the scheduling system achieves continuity in time, energy stability in space, and a reliable closed loop in logic, making the task scheduling process of the entire industrial wireless network verifiable, traceable, and self-healing.

[0051] This implementation achieves a three-dimensional fusion of time consistency, energy stability, and predictive redundancy by performing blockchain-based trusted scheduling operations under the constraints of the topological energy allocation landscape. The method ensures temporal consistency through unified rearrangement of time series, achieves spatial energy balance through topological energy landscape constraints, and constructs a trusted scheduling closed loop through continuous comparison and correction of shadow prediction sequences. Scheduling priority is no longer a static parameter but a verifiable variable that evolves dynamically in real-time based on on-chain smart contracts, enabling industrial wireless networks to maintain stable task chain execution and ensure the reliability of the scheduling process even under frequent node switching and high dynamic disturbances.

[0052] This invention introduces a dynamic blockchain-driven multi-level trusted scheduling mechanism to achieve real-time identification, prediction, and correction of transient cache pressure reversal phenomena in industrial wireless network environments, significantly improving the temporal consistency and spatial stability of task scheduling. By constructing a cross-node cache pressure real-time observation layer at the millisecond scale and combining a dual verification mechanism of causal coherence decomposition and shadow prediction sequences, this invention can predict node state reversal trends before task allocation, pre-marking and weighting high-risk nodes to avoid critical tasks being incorrectly assigned to nodes with unstable loads. It effectively suppresses state misordering and task disorder during node switching, ensuring that blockchain scheduling results maintain temporal continuity and execution chain integrity even under high-speed dynamic conditions, significantly improving the real-time performance and scheduling reliability of industrial wireless networks in complex environments.

[0053] This invention achieves joint reliable optimization of the energy and time domains under the constraints of topological energy allocation landscape. This allows for dynamic adjustment of task priorities, moving away from reliance on a single cached state and instead enabling adaptive decision-making based on a comprehensive feedback mechanism of unified rearranged time series, shadow prediction sequences, and topological energy distribution. Through energy mapping and weight decay mechanisms, the influence of risky nodes is automatically weakened, while the energy contribution of stable nodes is enhanced. This achieves dynamic balancing and self-correction of the scheduling path under the constraints of blockchain smart contracts. Ultimately, even under conditions of frequent multi-node switching, link fluctuations, and task overlap, the system maintains the stability and reliability of the global scheduling topology, ensuring that task execution order is verifiable, traceable, and robust, significantly improving the operational security and system-level self-consistency capabilities of industrial wireless networks.

[0054] This invention provides, for example Figure 2 The industrial wireless network trusted scheduling device based on dynamic blockchain shown includes a cache pressure real-time observation module, a causal coherence decomposition and analysis module, a shadow prediction sequence generation module, a time-scaled anchor point rearrangement and correction module, a topology energy allocation construction module, and a trusted scheduling execution module. The real-time cache pressure observation module constructs a cross-node real-time cache pressure observation layer, continuously collects the cache queue fluctuation trajectory of each node in the industrial wireless network at the millisecond level, and generates a traceable cache pressure evolution baseline in the observation layer. The causal coherence decomposition analysis module performs causal coherence decomposition operations based on the buffer pressure evolution baseline. It separates the pseudo-random interference components caused by wireless link disturbances in the observation data, extracts the key factors that lead to buffer pressure reversal, and generates a dynamic pressure pattern diagram that reflects the changing pattern of node buffer state. The shadow prediction sequence generation module constructs a shadow prediction sequence based on the dynamic pressure pattern diagram, marks the nodes that may have cache pressure reversal within the prediction window in advance, and inputs the shadow prediction sequence into the blockchain scheduling process to form a redundant verification path for comparison and verification with the main scheduling path. The time-stamped anchor rearrangement correction module deploys double mirrored time-stamped anchors on the cache pressure change trajectory based on the shadow prediction sequence, performs unified rearrangement of the cross-node time series, and reverses and corrects the observed transient cache pressure into a continuous monotonic evolution trajectory. The topology energy allocation construction module constructs a topology energy allocation landscape based on the monotonic evolution trajectory after unified rearrangement, maps the cache pressure state to quantified energy parameters, and sets weight decay for labeled inversion nodes during the energy allocation process to reduce the priority impact of risk nodes in the scheduling process. The trusted scheduling execution module performs trusted scheduling operations under the constraints of the topological energy allocation landscape. It comprehensively utilizes the uniformly rearranged time series and shadow prediction series to dynamically correct the task priorities in the blockchain smart contract.

[0055] The trusted scheduling method for industrial wireless networks based on dynamic blockchain provided in this embodiment of the invention is implemented by the aforementioned trusted scheduling device for industrial wireless networks based on dynamic blockchain. For details of the specific methods and processes of the trusted scheduling device for industrial wireless networks based on dynamic blockchain, please refer to the embodiments of the trusted scheduling method for industrial wireless networks based on dynamic blockchain, which will not be repeated here.

[0056] 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 trusted scheduling method for industrial wireless networks based on dynamic blockchain, characterized in that, Includes the following steps: A cross-node cache pressure real-time observation layer is constructed to continuously collect the cache queue fluctuation trajectory of each node in the industrial wireless network at the millisecond level, and generate a traceable cache pressure evolution baseline in the observation layer. Based on the baseline of cache pressure evolution, causal coherence decomposition is performed to separate the pseudo-random interference components caused by wireless link disturbances in the observation data, extract the key factors that lead to the reversal of cache pressure, and generate a dynamic pressure pattern diagram that reflects the changing pattern of node cache state. Based on the dynamic pressure pattern diagram, a shadow prediction sequence is constructed. Nodes that may experience cache pressure reversal within the prediction window are marked in advance. The shadow prediction sequence is then input into the blockchain scheduling process to form a redundant verification path for comparison and verification with the main scheduling path. Based on the shadow prediction sequence, double mirror time-scale anchor points are set up on the cache pressure change trajectory, and the time series across nodes are uniformly rearranged to reverse and correct the observed transient cache pressure into a continuous monotonic evolution trajectory. A topological energy allocation landscape is constructed based on the monotonic evolution trajectory after unified rearrangement. The cache pressure state is mapped to quantified energy parameters. During the energy allocation process, the weight decay of the labeled inversion nodes is set to reduce the priority impact of risk nodes in the scheduling process. Trusted scheduling operations are performed under the constraints of the topological energy distribution landscape. By comprehensively utilizing the uniformly rearranged time series and shadow prediction series, the task priorities in blockchain smart contracts are dynamically corrected.

2. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 1, characterized in that, The steps to build a real-time observation layer for cross-node cache pressure include: Within the topology of the industrial wireless network, a data acquisition path is established for the buffer queues of all participating nodes. By embedding high-precision time synchronization acquisition endpoints at the buffer interfaces of each node and adopting a clock synchronization algorithm, the sampling time of each node is unified to the same millisecond-level reference frame, thereby obtaining a multi-dimensional state vector of node buffer pressure. The original data of the cache pressure of each node is subjected to time-series normalization and noise reduction. The sliding time window segmentation algorithm and weighted exponential smoothing function are used to reduce instantaneous jitter. The time consistency of cross-node pressure sampling points is achieved on the global timeline through a two-way time calibration strategy. Feature extraction and trend modeling are performed based on time-consistent cache pressure fluctuation trajectories. Gradient time series fitting algorithm is used to identify key inflection points of cache pressure and generate a cache pressure evolution baseline marked with on-chain timestamps. Set time differential monitoring points on the cache pressure evolution baseline, mark the reversal risk of intervals where the pressure change rate changes abruptly and is accompanied by link delay or data rate anomalies, and record them in the blockchain state log with timestamps to form a traceable cache pressure evolution record.

3. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 2, characterized in that, In the step of setting time differential monitoring points on the cache pressure evolution baseline, for sudden changes in the rate of change of cache pressure from positive to negative or from negative to positive, a reversal risk range is determined by combining the comprehensive criteria of sudden increase in link latency and sudden drop in data transmission rate, and this range is written into the blockchain state log in the form of encrypted timestamps.

4. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 1, characterized in that, The steps for performing causal coherence decomposition based on the cache pressure evolution baseline include: After obtaining the cache pressure evolution baseline after time synchronization and denoising, causal relationship pre-modeling is performed on the cache pressure time series of each node, the conditional mutual information matrix between nodes is calculated, and an initial causal graph containing causal pointing relationships between nodes is constructed through delay Granger causality analysis. Under causal constraints, coherent feature decomposition is performed, and the multidimensional time series data of the cache pressure evolution baseline is mapped to the frequency domain space through short-time Fourier transform. The weighted principal component projection algorithm is used to separate the high-frequency interference components and low-frequency trend changes in a hierarchical manner. The decomposed cache pressure sequence is input into the causal weighted regression framework to calculate the relative contributions of cache input rate, output rate, link bandwidth fluctuation rate, energy consumption rate, and task queue burst rate to pressure changes, and to generate a node-level factor influence spectrum to identify key reversal factors. The key reversal factors are uniformly expressed according to the time and space dimensions. A three-dimensional dynamic stress pattern diagram is constructed by time sliding window clustering algorithm and causal intensity mapping, and traceable data entities are formed by combining blockchain timestamp labeling.

5. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 4, characterized in that, In the process of constructing the three-dimensional dynamic pressure pattern diagram, a time sliding window clustering algorithm is used to aggregate nodes with similar causal relationships into coherent subgroups. Within each coherent subgroup, a pressure propagation manifold is generated based on the causal strength and time phase difference, and synchronized and calibrated using a blockchain timestamp.

6. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 1, characterized in that, The steps for constructing a shadow prediction sequence based on a dynamic stress pattern diagram include: A time-domain mapping of the future prediction window is established based on the temporal characteristics and causal correlation of the dynamic pressure pattern diagram. The prediction window length is determined by an adaptive time step algorithm, and a set of temporal evolution equations for the propagation of node buffer pressure is constructed based on the causal intensity matrix. Using the cached pressure trajectory in the dynamic pressure pattern diagram as the initial input, the future pressure state of the node is iteratively solved by combining the time-recursive autoregressive extended model with the time evolution equation set, and the shadow prediction sequence is obtained through the cooperative perturbation balance mechanism. The difference between the rate of change of node cache pressure and the historical evolution baseline is calculated based on the predicted sequence. For nodes whose predicted trends change abruptly and whose rate of change exceeds the threshold, the reversal risk index is calculated and labeled in advance. At the same time, the labeling confidence is improved through the cross-validation mechanism of the predicted trends of adjacent nodes. The shadow prediction sequence and annotation results are written to the blockchain verification chain in the form of timestamps and dynamically compared with the main scheduling path. When a deviation is detected between the two, a verification mechanism is triggered to correct the task priority.

7. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 6, characterized in that, The steps for deploying dual mirror time-scaled anchors on the cache pressure change trajectory based on shadow prediction sequences include: The time characteristics of the cache pressure change trajectory are analyzed based on the shadow prediction sequence. The sliding differential algorithm is used to calculate the cache pressure change rate and its second derivative. When the change rate changes abruptly and the second derivative exceeds the threshold, the reversal boundary point is determined. Forward and reverse mirror anchor points are established before and after the boundary point to form a double mirror anchoring structure. Using the double mirror timescale anchor point as a reference, the time offset between each node is calculated and a global time offset matrix is ​​constructed. Time slip correction is performed by combining linear interpolation and nonlinear phase fitting, so that the cache pressure change curve is synchronously aligned in the global time dimension. The time series is segmented using double mirror anchor points as boundaries. When a transient reversal segment is detected, the curve is corrected into a continuous monotonic evolution trajectory through mirror mapping and resampling, while maintaining the minimum mean square error between the corrected pressure change rate and the predicted rate. The uniformly rearranged cache pressure time series is mapped to the on-chain time domain according to the blockchain timestamp, a global time series index table of cache pressure evolution is established, and recorded as immutable time-series consistent data through the on-chain hash verification mechanism.

8. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 7, characterized in that, The steps for constructing a topological energy distribution landscape based on the uniformly rearranged monotonic evolutionary trajectory include: After obtaining the monotonic evolution trajectory after unified rearrangement, the cache pressure state of each node is integrally transformed to calculate the energy density of cache load change per unit time, and the cache energy value is mapped to the standardized interval through energy normalization to form an energy state matrix. The correlation coefficient of buffered energy between nodes is calculated based on the energy state matrix, and a weighted directed graph is established. The directed graph is then mapped into a continuous energy flow field using the topological flow field modeling method, so that the buffered energy of each node is dynamically updated according to the energy inflow and outflow of adjacent nodes. In the energy flow field equation, a dynamic weight attenuation coefficient is set for the marked reversal nodes, and the attenuation magnitude is determined by combining the reversal risk index and the historical reversal frequency. At the same time, energy conservation constraints are introduced to maintain the total energy balance. The gradient flow convergence algorithm is used to iteratively optimize the energy distribution, calculate the gradient variance of the energy potential surface to verify the topological equilibrium state, and map the equilibrium energy distribution result to the blockchain scheduling parameter space and record it in the form of hash signature.

9. The trusted scheduling method for industrial wireless networks based on dynamic blockchain according to claim 8, characterized in that, The steps for performing trusted scheduling operations under the constraints of the topological energy allocation landscape include: The equilibrium results of the topological energy allocation landscape are coupled and mapped with the time series after unified rearrangement to establish a time-energy bidirectional binding mechanism. The energy weights of high-risk nodes are dynamically decayed according to the shadow prediction sequence to form the initial matrix of task priorities. The initial task priority matrix is ​​input into the blockchain smart contract as a parameter. The contract executes three stages in sequence: priority evaluation, node matching and task allocation. Based on the connection strength function of the topological energy landscape, the contract selects a target node with stable energy flow for the task. The scheduling results output by the blockchain smart contract are compared with the shadow prediction sequence on a task-by-task basis. When a timing deviation or energy distribution difference is detected, the scheduling correction factor is calculated and fed back into the contract logic to achieve dynamic priority correction. The scheduling calculation results, correction factors, and topological energy balance status are written into the blockchain ledger in the form of timestamps and node signatures, forming a traceable task execution time sequence.

10. A trusted scheduling device for industrial wireless networks based on dynamic blockchain, used to implement the trusted scheduling method for industrial wireless networks based on dynamic blockchain as described in any one of claims 1-9, characterized in that, It includes a real-time cache pressure monitoring module, a causal coherence decomposition and analysis module, a shadow prediction sequence generation module, a time-scaled anchor rearrangement and correction module, a topology energy allocation construction module, and a trusted scheduling execution module; The real-time cache pressure observation module constructs a cross-node real-time cache pressure observation layer, continuously collects the cache queue fluctuation trajectory of each node in the industrial wireless network at the millisecond level, and generates a traceable cache pressure evolution baseline in the observation layer. The causal coherence decomposition analysis module performs causal coherence decomposition operations based on the buffer pressure evolution baseline. It separates the pseudo-random interference components caused by wireless link disturbances in the observation data, extracts the key factors that lead to buffer pressure reversal, and generates a dynamic pressure pattern diagram that reflects the changing pattern of node buffer state. The shadow prediction sequence generation module constructs a shadow prediction sequence based on the dynamic pressure pattern diagram, marks the nodes that may have cache pressure reversal within the prediction window in advance, and inputs the shadow prediction sequence into the blockchain scheduling process to form a redundant verification path for comparison and verification with the main scheduling path. The time-stamped anchor rearrangement correction module deploys double mirrored time-stamped anchors on the cache pressure change trajectory based on the shadow prediction sequence, performs unified rearrangement of the cross-node time series, and reverses and corrects the observed transient cache pressure into a continuous monotonic evolution trajectory. The topology energy allocation construction module constructs a topology energy allocation landscape based on the monotonic evolution trajectory after unified rearrangement, maps the cache pressure state to quantified energy parameters, and sets weight decay for labeled inversion nodes during the energy allocation process to reduce the priority impact of risk nodes in the scheduling process. The trusted scheduling execution module performs trusted scheduling operations under the constraints of the topological energy allocation landscape. It comprehensively utilizes the uniformly rearranged time series and shadow prediction series to dynamically correct the task priorities in the blockchain smart contract.

Images