Elastic node scheduling method and system for massive IoT terminal data access

Abstract

This invention relates to the fields of network security and distributed computing technology, and discloses a node elastic scheduling method and system for massive IoT terminal data access. The method includes: real-time analysis of the instantaneous arrival rate of the data stream to be accessed to extract nonlinear fluctuation characteristics; calculation of the geometric volume expansion rate of the data stream during physical memory allocation; construction of a two-dimensional memory topology map mapping the resident core protection processes; and projection of a virtual expanded geometry generated based on the geometric volume expansion rate onto the map; and Boolean interference analysis of the dynamic outer boundary and the core protection processes under coordinate system constraints, generating secure scheduling instructions or memory isolation instructions based on whether the physical addressing space overlaps or conflicts. This invention materializes the burst characteristics of network traffic into the geometric expansion behavior of node physical memory allocation, proactively avoiding the structural collapse risk of addressing overlap in the underlying hardware shared cache area, and constructing a physical immune isolation barrier for edge nodes.

Description

Technical Field

[0001] This invention relates to a node elastic scheduling method and system for massive IoT terminal data access, belonging to the field of network security and distributed computing technology. Background Technology

[0002] Currently, IoT edge data centers typically employ dynamic matching strategies based on hardware computing power consumption when handling massive concurrent access from terminal devices. The central scheduler statistically analyzes the rate and volume of incoming data streams and calculates the overall load status of the edge node's central processing unit and physical memory. It then allocates network bandwidth and preprocessing computing power according to a binning algorithm to maintain the throughput of the basic system. However, with the concurrent access of heterogeneous IoT terminal devices, edge nodes simultaneously carry high-frequency sensor acquisition streams in plaintext and low-frequency critical control streams in encrypted form. When the scheduling system only directs heterogeneous data to the same edge node based on load statistics, data streams with different security levels share the same buffer and system within the node. Physical addressing overlap occurs on the bus, and node processors allocate a large number of discrete physical memory pages for bursts of heterogeneous data in a very short period of time, causing a dramatic geometric expansion of the data addressing boundary. For example, Chinese invention patent with authorization announcement number CN114077485B discloses a service scheduling and deployment method for IoT edge computing node resources. It analyzes computing resource requirements in the cloud, generates a list of optional deployment node services based on network latency information, and selects the optimal node service deployment through an iterative approach. This technology is essentially a macro-load sharing mechanism based on the total system load assessment. The scheduling logic focuses on network topology level balance and ignores the physical memory addressing security inside the node at the surface level.

[0003] The industry typically uses methods such as expanding the physical memory capacity of a single node or introducing time-series prediction models to alleviate the aforementioned address space compression pressure. However, these conventional methods, which rely on total capacity expansion or quantitative prediction, cannot eliminate the interference phenomenon in the underlying data space. The disorderly spread of high-frequency plaintext data payloads in the memory address space crosses the basic physical isolation distance and penetrates into the high-density control component area, causing cache-side channel leakage. Analysis shows that the existing technology mainly has the following shortcomings: 1. It lacks spatial topology planning for the internal processing links of the node in the pre-scheduling stage, and the concurrent parsing of heterogeneous data streams is very likely to cause underlying semantic conflicts and bus competition; 2. Simply relying on quantitative prediction mechanisms cannot capture the risk of underlying physical address fragmentation and overlap caused by instantaneous concurrent access, and the highly sensitive control data with mixed residence faces the threat of cross-protocol security pollution.

[0004] Therefore, the technical problem to be solved by this invention is how to break through the traditional computing power transfer and scheduling logic, accurately calculate the probability of underlying semantic interference before the actual access of the data stream, construct a dynamic generation mechanism for node computing topology with physical isolation characteristics, block node structure failure caused by concurrent processing of heterogeneous protocols, and construct a complete IoT edge node security protection protocol to ensure the topological integrity and logical security of service nodes under the impact of massive heterogeneous data. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of this invention is as follows: A node elastic scheduling method for massive IoT terminal data access, comprising the following steps: The system analyzes the transport layer load characteristics and instantaneous arrival rate of the data stream to be accessed in real time, extracts the memory allocation granularity parameters contained in the transport layer load characteristics, and extracts nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate. The geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node is calculated based on the nonlinear fluctuation characteristic parameters, and a virtual expanded geometry corresponding to the data stream to be accessed is generated based on the memory allocation granularity parameters and the geometric volume expansion rate. Call the underlying memory addressing status data of the distributed network access node to construct a two-dimensional memory topology map that maps the physical address range of the resident core protection process, and project the virtual expansion geometry into the coordinate system of the two-dimensional memory topology map; Under the coordinate system constraints of the two-dimensional memory topology map, Boolean interference analysis is performed on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process, and a conflict determination signal is output to characterize the overlapping state of the physical addressing space. When the collision determination signal is a non-overlapping collision, an addressing release command is issued for the data stream to be accessed, so as to allow it to access the distributed network access node. When the conflict determination signal indicates that there is an overlapping conflict, a memory isolation instruction is issued for the data stream to be accessed to freeze the port mapping of the data stream to be accessed.

[0006] Preferably, the step of performing Boolean interference analysis on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process specifically includes: extracting the centroid coordinates of the physical address of the resident core protection process in the two-dimensional memory topology map, and delineating a rigid protection area centered on the centroid coordinates of the physical address according to a preset access permission level protocol; projecting the virtual expansion geometry along the time axis as a dynamic trajectory band, and calculating the shortest Euclidean distance between the outer envelope of the dynamic trajectory band and the boundary line of the rigid protection area; if the shortest Euclidean distance is less than a preset physical isolation distance threshold, the conflict determination signal is set to indicate an overlapping conflict, and the topology coordinates of the collision are recorded; if the shortest Euclidean distance is greater than or equal to the physical isolation distance threshold, the conflict determination signal is set to indicate a non-overlapping conflict, and the allowed addressing range of the data stream to be accessed in the distributed network access node is locked.

[0007] Preferably, the step of calculating the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on nonlinear fluctuation characteristic parameters determines the dynamic expansion radius of the virtual expansion geometry based on the following expansion model: , where R(t) is the dynamic expansion radius of the virtual expanding geometry at time t; This is the real-time value of the instantaneous arrival rate; The effective payload size of the data stream to be accessed; α is the preset addressing period constant for distributed network access nodes; α and β are weighted correction coefficients with the dimension of rate-to-space conversion. The weighted correction coefficients are dynamically calculated based on the dispersion index of the currently available memory fragments of the distributed network access nodes.

[0008] Preferably, the step of constructing a two-dimensional memory topology map mapping the physical address range of the resident core protection processes specifically includes: obtaining the physical page frame distribution data of all currently active processes by reading the page table of the memory management unit of the distributed network access node; converting the discrete address segments in the physical page frame distribution data into rectangular placeholders in a two-dimensional Cartesian coordinate system; identifying the key processes performing encryption operations among the active processes and marking their corresponding rectangular placeholders as highly sensitive geometric regions; in the two-dimensional Cartesian coordinate system, extending a preset address width outward along the outer contour of the highly sensitive geometric region to generate and apply a logical no-fly zone boundary that prohibits ordinary data streams from writing; and generating a two-dimensional memory topology map based on the closed range of the rectangular placeholders and the logical no-fly zone boundary.

[0009] Preferably, the step of extracting nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate specifically includes: statistically analyzing the number of bit stream arrivals per unit time using a sliding window algorithm to generate a time-series waveform reflecting the instantaneous arrival rate; identifying spike pulses in the time-series waveform whose slope exceeds a preset abrupt change threshold, and extracting the actual rising slope of the spike pulse as a nonlinear fluctuation characteristic parameter; and substituting the nonlinear fluctuation characteristic parameter into a preset function mapping relationship to update the benchmark weighting factor for calculating the geometric volume expansion rate.

[0010] Preferably, after issuing the memory isolation instruction for the data stream to be accessed, the method further includes: reporting alarm information containing conflict coordinates and conflict intensity to the central control plane, wherein the conflict intensity is determined by the ratio of the overlapping area between the virtual expansion geometry and the rigid protection area; querying the preset backup node resource pool, obtaining the memory free fragmentation ratio of each backup node, and selecting a backup node whose memory free fragmentation ratio is higher than the preset available level; routing the access request of the data stream to be accessed to the backup node, and reallocating physical page frames in the two-dimensional memory topology map of the backup node.

[0011] Preferably, after issuing the addressing and release command for the data stream to be accessed, the method further includes: tracking the actual memory occupancy trajectory of the accessed data stream in the distributed network access node in real time; calculating the area deviation value between the actual coverage area of ​​the actual memory occupancy trajectory and the nominal coverage area of ​​the virtual expansion geometry predicted before scheduling; updating the algorithm feature weight parameters for calculating the geometric volume expansion rate when the area deviation value exceeds the preset fault tolerance deviation threshold; and refreshing the placeholder status in the two-dimensional memory topology map based on the actual memory occupancy trajectory as the underlying addressing status data for subsequent Boolean interferometry analysis.

[0012] Preferably, before the step of calling the underlying memory addressing status data of the distributed network access node, the method further includes: establishing a mapping rule base from physical memory addresses to a unified virtual coordinate system, using the mapping rule base to standardize fragmented physical memory space into a unified two-dimensional planar layer; continuously monitoring the number of concurrent threads of each processing core inside the distributed network access node, marking the logical addressing area corresponding to the processor core whose number of concurrent threads continuously exceeds a preset full load threshold as the underlying resource dissipation layer; superimposing the underlying resource dissipation layer onto the two-dimensional memory topology map, and setting the area covered by the underlying resource dissipation layer as an unallocatable static obstacle when performing Boolean interference analysis.

[0013] Preferably, the step of calculating the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on the nonlinear fluctuation characteristic parameters further includes: calculating the load entropy value of the data stream to be accessed, and confirming that the data stream to be accessed has padding attack characteristics when the load entropy value is lower than a preset normal information flow benchmark judgment threshold; for the data stream to be accessed with padding attack characteristics, extracting a penalty gain coefficient with a value greater than 1, and multiplying the penalty gain coefficient by the initially calculated geometric volume expansion rate to obtain the amplified final geometric volume expansion rate; generating an extended dynamic outer boundary based on the final geometric volume expansion rate to achieve the desired effect in Boolean dynamics. To amplify the probability of collisions during analysis, the method also includes: periodically extracting all data stream access records and their resulting physical addressing space overlap conflict event logs within a historical time period; analyzing the conflict event logs using clustering algorithms to count the interference frequency of each coordinate block in the two-dimensional memory topology map; identifying coordinate blocks whose interference frequency exceeds a preset frequency threshold and defining them as high interference-sensitive topology regions; issuing a reserved isolation command to permanently lock the high interference-sensitive topology regions in the two-dimensional memory topology map, and forcibly resetting the physical address of ordinary security level data streams to a safe and available space outside the high interference-sensitive topology regions when issuing subsequent addressing release commands.

[0014] A node elastic scheduling system for massive IoT terminal data access, the system includes: The data stream parsing module is used to parse the transport layer payload characteristics and instantaneous arrival rate of the data stream to be accessed in real time, extract the memory allocation granularity parameters contained in the transport layer payload characteristics, and extract the nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate. The geometric expansion calculation module is used to calculate the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on the nonlinear fluctuation characteristic parameters, and to generate the virtual expansion geometry corresponding to the data stream to be accessed based on the memory allocation granularity parameters and the geometric volume expansion rate. The topology graph construction module is used to call the underlying memory addressing status data of the distributed network access nodes, construct a two-dimensional memory topology graph that maps the physical address range of the resident core protection processes, and project the virtual expanded geometry into the coordinate system of the two-dimensional memory topology graph. The interference analysis module is used to perform Boolean interference analysis on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process under the coordinate system constraints of the two-dimensional memory topology map, and outputs a conflict determination signal to characterize the overlapping state of the physical address space. The security scheduling execution module is used to issue an addressing and allowing instruction for the data stream to be accessed when the conflict determination signal indicates no overlapping conflict, so as to allow it to access the distributed network access node; when the conflict determination signal indicates overlapping conflict, it issues a memory isolation instruction for the data stream to be accessed, so as to freeze the port mapping of the data stream to be accessed.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. In the massive IoT terminal data access, a low-level isolation scheduling mechanism based on orthogonal computation of feature space is established. The central scheduler extracts the protocol type parameters and encryption depth parameters of the access data stream to construct a security feature vector, calculates the dot product coefficient value of the data stream to be accessed and the data stream residing in the node, and analyzes the low-level semantic conflict probability between data streams at the hardware physical level based on the spatial orthogonal relationship represented by the dot product coefficient value. When it is determined that there is a risk of spatial interference, the system manages the backup edge infrastructure to create a resource slice with exclusive network port allocation in an independent memory address range. This mechanism changes the traditional edge computing technical route of simply relying on computing power load indicators to schedule heterogeneous data. It cuts off the cache side channel pollution path of high-frequency plaintext data to the encrypted control command stream from the initial stage of data access routing, and builds a low-level defense barrier for edge computing nodes from memory space to system bus transmission layer. By introducing spatial topology security verification in the network service access control process, the protocol stability and system protection strength of edge nodes in complex network environments are effectively improved.

[0016] 2. A dynamic topology routing mechanism oriented towards the physical memory addressing range of nodes is constructed. The central scheduler maps the available memory addressing range of the target edge node to the network port into a two-dimensional topology design map. According to the preset mapping rules, the multi-source data stream is transformed into virtual processing components carrying characteristic occupancy area and security isolation radius. The system calls the layout algorithm to deduce the addressing center for each component in the above map and calculates the absolute isolation distance value of the physical address of any component. By setting the preset physical isolation distance as a rigid precondition for the issuance of orchestration instructions, the system converts the unpredictable network layer concurrent access risk into geometric topology interference prevention in the design stage. This structural layout avoids the structural collapse risk of addressing overlap of heterogeneous high-density data streams in the same underlying hardware shared buffer area, and constructs a physical immune isolation barrier for edge nodes, thereby establishing a node security access and topology protection protocol for access terminal behavior at the network service level. Attached Figure Description

[0017] Figure 1 This is a timing flowchart of the node elastic scheduling backbone for massive IoT data access in this invention. Figure 2 This is a logical diagram of the core branch modules and parameter architecture of the node elastic scheduling method of the present invention.

[0018] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] A node elastic scheduling method for massive IoT terminal data access includes the following steps: The system analyzes the transport layer load characteristics and instantaneous arrival rate of the data stream to be accessed in real time, extracts the memory allocation granularity parameters contained in the transport layer load characteristics, and extracts nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate. The geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node is calculated based on the nonlinear fluctuation characteristic parameters, and a virtual expanded geometry corresponding to the data stream to be accessed is generated based on the memory allocation granularity parameters and the geometric volume expansion rate. Call the underlying memory addressing status data of the distributed network access node to construct a two-dimensional memory topology map that maps the physical address range of the resident core protection process, and project the virtual expansion geometry into the coordinate system of the two-dimensional memory topology map; Under the coordinate system constraints of the two-dimensional memory topology map, Boolean interference analysis is performed on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process, and a conflict determination signal is output to characterize the overlapping state of the physical addressing space. When the collision determination signal is a non-overlapping collision, an addressing release command is issued for the data stream to be accessed, so as to allow it to access the distributed network access node. When the conflict determination signal indicates that there is an overlapping conflict, a memory isolation instruction is issued for the data stream to be accessed to freeze the port mapping of the data stream to be accessed.

[0021] Preferably, the step of performing Boolean interference analysis on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process specifically includes: extracting the centroid coordinates of the physical address of the resident core protection process in the two-dimensional memory topology map, and delineating a rigid protection area centered on the centroid coordinates of the physical address according to a preset access permission level protocol; projecting the virtual expansion geometry along the time axis as a dynamic trajectory band, and calculating the shortest Euclidean distance between the outer envelope of the dynamic trajectory band and the boundary line of the rigid protection area; if the shortest Euclidean distance is less than a preset physical isolation distance threshold, the conflict determination signal is set to indicate an overlapping conflict, and the topology coordinates of the collision are recorded; if the shortest Euclidean distance is greater than or equal to the physical isolation distance threshold, the conflict determination signal is set to indicate a non-overlapping conflict, and the allowed addressing range of the data stream to be accessed in the distributed network access node is locked.

[0022] Preferably, the step of calculating the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on nonlinear fluctuation characteristic parameters determines the dynamic expansion radius of the virtual expansion geometry based on the following expansion model: , where R(t) is the dynamic expansion radius of the virtual expanding geometry at time t; This is the real-time value of the instantaneous arrival rate; The effective payload size of the data stream to be accessed; α is the preset addressing period constant for distributed network access nodes; α and β are weighted correction coefficients with the dimension of rate-to-space conversion. The weighted correction coefficients are dynamically calculated based on the dispersion index of the currently available memory fragments of the distributed network access nodes.

[0023] Preferably, the step of constructing a two-dimensional memory topology map mapping the physical address range of the resident core protection processes specifically includes: obtaining the physical page frame distribution data of all currently active processes by reading the page table of the memory management unit of the distributed network access node; converting the discrete address segments in the physical page frame distribution data into rectangular placeholders in a two-dimensional Cartesian coordinate system; identifying the key processes performing encryption operations among the active processes and marking their corresponding rectangular placeholders as highly sensitive geometric regions; in the two-dimensional Cartesian coordinate system, extending a preset address width outward along the outer contour of the highly sensitive geometric region to generate and apply a logical no-fly zone boundary that prohibits ordinary data streams from writing; and generating a two-dimensional memory topology map based on the closed range of the rectangular placeholders and the logical no-fly zone boundary.

[0024] Preferably, the step of extracting nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate specifically includes: statistically analyzing the number of bit stream arrivals per unit time using a sliding window algorithm to generate a time-series waveform reflecting the instantaneous arrival rate; identifying spike pulses in the time-series waveform whose slope exceeds a preset abrupt change threshold, and extracting the actual rising slope of the spike pulse as a nonlinear fluctuation characteristic parameter; and substituting the nonlinear fluctuation characteristic parameter into a preset function mapping relationship to update the benchmark weighting factor for calculating the geometric volume expansion rate.

[0025] Preferably, after issuing the memory isolation instruction for the data stream to be accessed, the method further includes: reporting alarm information containing conflict coordinates and conflict intensity to the central control plane, wherein the conflict intensity is determined by the ratio of the overlapping area between the virtual expansion geometry and the rigid protection area; querying the preset backup node resource pool, obtaining the memory free fragmentation ratio of each backup node, and selecting a backup node whose memory free fragmentation ratio is higher than the preset available level; routing the access request of the data stream to be accessed to the backup node, and reallocating physical page frames in the two-dimensional memory topology map of the backup node.

[0026] Preferably, after issuing the addressing and release command for the data stream to be accessed, the method further includes: tracking the actual memory occupancy trajectory of the accessed data stream in the distributed network access node in real time; calculating the area deviation value between the actual coverage area of ​​the actual memory occupancy trajectory and the nominal coverage area of ​​the virtual expansion geometry predicted before scheduling; updating the algorithm feature weight parameters for calculating the geometric volume expansion rate when the area deviation value exceeds the preset fault tolerance deviation threshold; and refreshing the placeholder status in the two-dimensional memory topology map based on the actual memory occupancy trajectory as the underlying addressing status data for subsequent Boolean interferometry analysis.

[0027] Preferably, before the step of calling the underlying memory addressing status data of the distributed network access node, the method further includes: establishing a mapping rule base from physical memory addresses to a unified virtual coordinate system, using the mapping rule base to standardize fragmented physical memory space into a unified two-dimensional planar layer; continuously monitoring the number of concurrent threads of each processing core inside the distributed network access node, marking the logical addressing area corresponding to the processor core whose number of concurrent threads continuously exceeds a preset full load threshold as the underlying resource dissipation layer; superimposing the underlying resource dissipation layer onto the two-dimensional memory topology map, and setting the area covered by the underlying resource dissipation layer as an unallocatable static obstacle when performing Boolean interference analysis.

[0028] Preferably, the step of calculating the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on the nonlinear fluctuation characteristic parameters further includes: calculating the load entropy value of the data stream to be accessed, and confirming that the data stream to be accessed has padding attack characteristics when the load entropy value is lower than a preset normal information flow benchmark judgment threshold; for the data stream to be accessed with padding attack characteristics, extracting a penalty gain coefficient with a value greater than 1, and multiplying the penalty gain coefficient by the initially calculated geometric volume expansion rate to obtain the amplified final geometric volume expansion rate; generating an extended dynamic outer boundary based on the final geometric volume expansion rate to achieve the desired effect in Boolean dynamics. To amplify the probability of collisions during analysis, the method also includes: periodically extracting all data stream access records and their resulting physical addressing space overlap conflict event logs within a historical time period; analyzing the conflict event logs using clustering algorithms to count the interference frequency of each coordinate block in the two-dimensional memory topology map; identifying coordinate blocks whose interference frequency exceeds a preset frequency threshold and defining them as high interference-sensitive topology regions; issuing a reserved isolation command to permanently lock the high interference-sensitive topology regions in the two-dimensional memory topology map, and forcibly resetting the physical address of ordinary security level data streams to a safe and available space outside the high interference-sensitive topology regions when issuing subsequent addressing release commands.

[0029] A node elastic scheduling system for massive IoT terminal data access, the system includes: The data stream parsing module is used to parse the transport layer payload characteristics and instantaneous arrival rate of the data stream to be accessed in real time, extract the memory allocation granularity parameters contained in the transport layer payload characteristics, and extract the nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate. The geometric expansion calculation module is used to calculate the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on the nonlinear fluctuation characteristic parameters, and to generate the virtual expansion geometry corresponding to the data stream to be accessed based on the memory allocation granularity parameters and the geometric volume expansion rate. The topology graph construction module is used to call the underlying memory addressing status data of the distributed network access nodes, construct a two-dimensional memory topology graph that maps the physical address range of the resident core protection processes, and project the virtual expanded geometry into the coordinate system of the two-dimensional memory topology graph. The interference analysis module is used to perform Boolean interference analysis on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process under the coordinate system constraints of the two-dimensional memory topology map, and outputs a conflict determination signal to characterize the overlapping state of the physical address space. The security scheduling execution module is used to issue an addressing and allowing instruction for the data stream to be accessed when the conflict determination signal indicates no overlapping conflict, so as to allow it to access the distributed network access node; when the conflict determination signal indicates overlapping conflict, it issues a memory isolation instruction for the data stream to be accessed, so as to freeze the port mapping of the data stream to be accessed.

[0030] Example 1: When 100,000 heterogeneous IoT terminals concurrently access the same distributed network access node within a 1ms to 5ms time window, this node simultaneously receives plaintext environmental monitoring sensor data streams and encrypted industrial control streams. If, based on computing power and network bandwidth load statistics, these data streams are directed to the same physical memory shared buffer, the addressing boundary of the plaintext data undergoes geometric expansion, crossing the basic physical isolation distance and penetrating into the control component area where the encrypted industrial control stream resides, causing physical addressing overlap. To avoid this physical addressing overlap, the data stream parsing module analyzes the transport layer load characteristics and instantaneous arrival rate of the data stream to be accessed in real time. The memory allocation granularity parameters included in the transport layer payload features are extracted. Based on the instantaneous arrival rate, the bit stream arrival volume per unit time is statistically analyzed using a sliding window algorithm. A time-series waveform reflecting the instantaneous arrival rate is generated. Spikes with slopes greater than the 50 Mbps / s abrupt change threshold in the time-series waveform are identified, and the actual rising slope of the spikes is extracted as a nonlinear fluctuation characteristic parameter. The geometric expansion calculation module calculates the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on the nonlinear fluctuation characteristic parameter. The dynamic expansion radius R(t) of the virtual expansion geometry is determined using an expansion model. The specific calculation formula is as follows: Where t is the time variable, and R(t) is the virtual expanding geometry in... The dynamic expansion radius at any given time. This is the real-time value of the instantaneous arrival rate. The payload size of the data stream to be accessed. α and β are the preset addressing period constants for distributed network access nodes. α and β are weighted correction coefficients with a conversion dimension from rate to spatial dimension. The weighted correction coefficients are dynamically calculated based on the dispersion index of currently available memory fragments in the distributed network access nodes. A virtual expanded geometry corresponding to the data stream to be accessed is generated based on memory allocation granularity parameters and geometric volume expansion rate. When determining the weighted correction coefficients α and β, the access load is simulated using 64-byte standardized data packets during the initialization phase. The increment of physical page frame address span allocated by the memory management unit under different instantaneous arrival rates is monitored. The slope of the address span increment relative to the instantaneous arrival rate is obtained through linear fitting and set as α. The ratio of isolated free physical pages to the total number of free physical pages is statistically analyzed within the sampling period to monitor the dispersion index of available memory fragments. According to the compensation coefficient table containing different dispersion ranges, when the dispersion index exceeds 30%, β is increased to 1.2 times to compensate for the geometric volume expansion deviation caused by the discontinuity of physical memory allocation.

[0031] The topology map construction module calls the underlying memory addressing status data of the distributed network access nodes. It obtains the physical page frame distribution data of all currently active processes by reading the page table of the memory management unit. The discrete address segments in the physical page frame distribution data are converted into rectangular placeholders in a two-dimensional Cartesian coordinate system. Key processes running encryption operations are identified among the active processes, and their corresponding rectangular placeholders are marked as highly sensitive geometric regions. A 4KB address width is extended outward along the outer contour of the highly sensitive geometric region in the two-dimensional Cartesian coordinate system to generate a logical no-fly zone boundary that prohibits ordinary data streams from writing. Based on the closed range of the rectangular placeholders and the logical no-fly zone boundary, a two-dimensional memory topology map is constructed, mapping the physical address range of the resident core protected processes. The virtual expanded geometry is projected onto the coordinate system of the two-dimensional memory topology map. When constructing the two-dimensional memory topology map, a preset address span constant is used. The values ​​are set to correspond to the L2 cache line or 4KB physical page boundary of the distributed network access node processor, enabling the two-dimensional memory topology map to simulate the underlying hardware cache conflict correlation characteristics. Before performing Boolean interferometry, an axis-aligned bounding box is established for the virtual expanded geometry for rapid overlap pre-check. When there is an overlap between the bounding box and the physical address range of the core protection process, the geometry calculation module is called to calculate the shortest Euclidean distance, ensuring physical isolation of the address space and reducing the computational load of the algorithm under large-scale concurrent access. Under the coordinate system constraints of the two-dimensional memory topology map, the interferometry analysis module extracts the core protection processes already resident in the two-dimensional memory topology map. The physical address centroid coordinates of the protection process are used to delineate a rigid protection zone centered on the physical address centroid coordinates according to a preset access permission level protocol. The virtual expansion geometry is projected along the time axis as a dynamic trajectory band. The shortest Euclidean distance between the outer envelope of the dynamic trajectory band and the boundary line of the rigid protection zone is calculated. Boolean interference results are generated by comparing the dynamic outer boundary of the virtual expansion geometry with the physical address range of the resident core protection process. Through the linkage constraints between the above features, the instantaneous arrival rate and nonlinear fluctuation characteristic parameters provide quantitative inputs characterizing the transient burst intensity for the calculation of the geometric volume expansion rate. The dynamic trajectory band generated by the product dilation rate defines the interference boundary in the two-dimensional memory topology map, transforming data volume conflicts in the communication protocol into geometric topological repulsion quantities in the two-dimensional coordinate system. When the shortest Euclidean distance is greater than or equal to the 4KB physical isolation distance threshold, a conflict determination signal representing non-overlapping conflicts is output, and the security scheduling module issues an addressing release instruction for the data stream to be accessed, locking the permitted addressing interval to allow it to access the distributed network access node. When the shortest Euclidean distance is less than the 4KB physical isolation distance threshold, a conflict determination signal representing overlapping conflicts is output, and the security scheduling module issues an addressing release instruction for the data stream to be accessed. The system uses memory isolation instructions to freeze the port mapping of the data stream to be accessed, reports alarm information including conflict coordinates and conflict intensity to the central control plane, obtains the memory free fragmentation ratio of each standby node, selects standby nodes with a memory free fragmentation ratio greater than or equal to 60% of the available water level, routes the access request of the data stream to be accessed to the standby node, and reallocates physical page frames in the two-dimensional memory topology map of the standby node. In this way, the physical isolation boundary blocks the spread of the underlying addressing boundary caused by the concurrent access of heterogeneous data streams, and the core control data storage structure within the edge node maintains a non-overlapping state of physical addressing space.

[0032] Example 2: A hardware-in-the-loop test platform was built using a network simulation tester and an edge computing server. The edge computing server was configured with a 64-core processor with a main frequency of 3.0 GHz and 256 GB of physical memory. The network simulation tester generated concurrent access requests from IoT terminals according to the Poisson distribution model and output a baseline data stream containing plaintext sensor data and encrypted control data. The baseline data stream was superimposed with channel random noise with a bit error rate of 10% and contained low-entropy data packets accounting for 15% of the total data stream with a load entropy value lower than the baseline judgment threshold for normal information streams. The experiment established control group 1, control group 2, and the sample group of this invention. Control group 1 adopted a method based on arrival rate and remaining memory capacity. The temporal capacity threshold scheduling algorithm, in the control group 2, extracts instantaneous arrival rate and nonlinear fluctuation characteristic parameters to calculate the geometric volume expansion rate. It compares the shortest Euclidean distance between the outer envelope of the dynamic trajectory band and the boundary line of the rigid protection area without introducing a penalty gain coefficient. The sample group of this invention calculates the load entropy value of the data stream to be accessed. When the load entropy value is lower than the preset normal information flow benchmark judgment threshold, a penalty gain coefficient with a value greater than 1 is obtained. The penalty gain coefficient is multiplied by the initially calculated geometric volume expansion rate to obtain the final geometric volume expansion rate. Based on the final geometric volume expansion rate, a virtual expanded geometry is generated and the shortest Euclidean distance is calculated in the two-dimensional memory topology map.

[0033] The number of concurrent access terminals in the test scenario was set at gradients of 50,000, 100,000, and 150,000. Monitoring indicators included the frequency of addressing overlap events where the shortest Euclidean distance was less than the 4KB physical isolation distance threshold and the core process latency. At a concurrency scale of 50,000, control group 1 recorded 12 addressing overlap events, with an average core process latency of 45.2ms. Control group 2 and the sample group of this invention had 0 addressing overlap events, and their core process latency ranged from 1.5ms to 2.1ms. When the concurrency scale was set to 100,000, control group 1... The addressing overlap event frequency was 145 times, and the core process latency was greater than 300ms. In control group two, when processing a baseline data stream containing low-entropy data packets, the shortest Euclidean distance output was greater than the 4KB physical isolation distance threshold, resulting in 27 addressing overlap events and an average core process latency of 58.6ms. In this invention's sample group, a virtual expanded geometry was generated based on the final geometric volume expansion rate amplified by a penalty gain coefficient. The calculated shortest Euclidean distance was less than the 4KB physical isolation distance threshold. The security scheduling module outputs a conflict determination signal indicating overlapping conflicts and issues a memory isolation instruction. The frequency of addressing overlap events in the invention sample group was 0, and the core process latency was 2.4ms. When the concurrency scale was set to 150,000, the available memory fragmentation dispersion index of the edge computing server dropped below the preset threshold, while the frequency of addressing overlap events in the control group 2 increased to 412. Based on the conflict judgment signal that the shortest Euclidean distance is less than the 4KB physical isolation distance threshold, the invention sample group obtained the memory free fragmentation ratio of each backup node, selected backup nodes with a memory free fragmentation ratio greater than or equal to 60% of the available water level, routed access requests to the backup nodes, and used the two-dimensional memory topology map of the backup nodes. In the spectrum, physical page frames are reallocated. The frequency of addressing overlap events in the sample group of this invention is 0 at a concurrency scale of 150,000, and the core process latency is 12.5ms. The sample group of this invention extracts the load entropy value to determine the penalty gain coefficient to correct the geometric volume expansion rate calculated from the instantaneous arrival rate and nonlinear fluctuation characteristic parameters. Combined with the two-dimensional memory topology map and the shortest Euclidean distance, the conflict judgment signal is output. Under the concurrency scale where the available memory fragmentation dispersion index decreases, memory isolation instructions are issued or routes are routed to the backup node. The physical addressing space of the core control process in the distributed network access node remains in a non-overlapping state.

[0034] Example 3: When a distributed network access node operates under continuous concurrent conditions and there is fragmented free physical memory, and the data stream to be accessed contains low-information-density padding attack packets, the static mapping parameters deviate from the physical addressing boundary and spread. The data stream parsing module scans the memory management unit page table of the distributed network access node, counts the total number of currently free physical pages and the number of isolated free physical pages that cannot be merged with adjacent free physical pages, obtains the dispersion index of available memory fragments by calculating the ratio of the number of isolated free physical pages to the total number of free physical pages, adds the constant 1 to the dispersion index to obtain the correction multiplier, and multiplies the preset base constant with the correction multiplier to obtain the weighted correction coefficient with the dimension of rate to spatial dimension conversion. The data stream parsing module extracts the effective payload sequence of the data stream to be accessed, counts the occurrence probability of various byte values ​​in the effective payload sequence, calculates the product of each occurrence probability and its corresponding base-2 logarithm, calculates the cumulative value of the negative sum of all products to generate the payload entropy value, and compares the payload entropy value with the preset normal information flow benchmark judgment threshold.

[0035] When the load entropy value is less than the normal information flow benchmark judgment threshold, it is determined that the data flow to be accessed carries padding attack characteristics. The security scheduling module calculates the difference between the normal information flow benchmark judgment threshold and the load entropy value, divides the difference by the normal information flow benchmark judgment threshold to obtain the deviation ratio, and adds the constant 1 to the deviation ratio to calculate the penalty gain coefficient. The geometric expansion calculation module extracts the instantaneous arrival rate and nonlinear fluctuation characteristic parameters, and calculates the initial geometric volume expansion rate by combining the weighted correction coefficient. The penalty gain coefficient is multiplied by the initial geometric volume expansion rate to obtain the final geometric volume expansion rate. The topology map construction module generates a two-dimensional memory topology map based on the final geometric volume expansion rate. The corresponding virtual expansion geometry is formed. The interference analysis module calculates the shortest Euclidean distance between the outer envelope of the dynamic trajectory band formed by the projection of the virtual expansion geometry along the time axis and the boundary line of the rigid protection area. By extracting the dispersion index and load entropy value and combining the instantaneous arrival rate and nonlinear fluctuation characteristic parameters, the final geometric volume expansion rate is calculated. The fragmentation degree of physical memory and the low entropy characteristics of attack packets are transformed into the extension increment of the dynamic trajectory band in the two-dimensional Cartesian coordinate system. The security scheduling module issues memory isolation instructions based on the shortest Euclidean distance of less than 4KB physical isolation distance threshold to block the physical address space overlap conflict induced by physical memory fragmentation and malicious filling.

[0036] Example 4: When the system faces a physical environment composed of heterogeneous hardware components and no data security baseline has been established, the distributed network access node is connected to an independent and isolated test network. The data injection device inputs known, anomaly-free standard industrial sensor data streams and control command streams into the distributed network access node. The data stream parsing module collects the baseline load sequence of the standard industrial sensor data streams and control command streams within a continuous time period, counts the frequency of occurrence of various byte values ​​in each independent sampling window, calculates the baseline load entropy value of each sampling window based on the negative sum of the product of the frequency and its corresponding base-2 logarithmic value, aggregates the baseline load entropy values ​​of all sampling windows to construct a baseline entropy value statistical histogram, extracts the expected value and standard deviation from the baseline entropy value statistical histogram, and calculates the normal information stream baseline judgment threshold. The specific mathematical model is as follows: ,in The threshold for judging the normal information flow is μ, which is the expected value of the benchmark load entropy value in the benchmark entropy value statistical histogram, σ is the corresponding standard deviation, and k is the dimensionless confidence interval adjustment coefficient. This calculation step extracts the distribution characteristics of standard business data into a one-dimensional scalar benchmark and writes it into the local security policy storage domain of the distributed network access node.

[0037] For the physical memory components inside the distributed network access node, the test process continuously writes fixed-size discrete probe data blocks to the physical memory address space under system idle conditions, records the critical dispersion index that triggers the page table reorganization action of the memory management unit, divides the constant 1 by the critical dispersion index to obtain the hardware sensitivity factor, and sets the ratio of the hardware sensitivity factor to the memory bus limit transmission rate as a preset basic constant with a conversion dimension from rate to space. In the actual data access state, the data flow parsing module dynamically updates the weighted correction coefficient based on the product of the preset basic constant and the correction factor. The parameters obtained by the aforementioned calibration procedure limit the calculation of the final geometric volume expansion rate to the physical architecture constraints of the bearer node and the statistical characteristics of the standard service flow, ensuring that the distributed network access node maintains an interference-free isolation state of the physical address space of the core control command flow when dealing with traffic surges caused by concurrent access of heterogeneous terminals.

[0038] Example 5: When a distributed network access node faces concurrent data streams and its internal processing cores experience uneven hardware load hotspots, the one-dimensional linear physical memory allocation mechanism conceals the actual computing power dissipation state of the processing cores. The system directs the virtual expansion geometry to physical page frames in a thread overload state, causing addressing latency. Before accessing the underlying memory addressing state data of the distributed network access node, the topology map construction module reads the hardware configuration parameters of the memory management unit, establishes a mapping rule base from physical memory addresses to a unified virtual coordinate system, and calculates the coordinate positions of discrete address segments in the two-dimensional plane layer based on the mapping rule base. The specific mathematical formula used for coordinate calculation is as follows: as well as Where x and y are the abscissa and ordinate in a two-dimensional Cartesian coordinate system. This is a decimal integer index of the underlying physical memory address. To characterize the preset address span constant of the two-dimensional transformation grid width, this procedure uses a mapping rule base to transform fragmented physical memory space into a two-dimensional planar layer of a unified format. The topology map construction module monitors the real-time concurrent thread count of each processing core inside the distributed network access node, compares the real-time concurrent thread count with the preset full load threshold, extracts the time period during which the real-time concurrent thread count is continuously greater than the preset full load threshold, locks the corresponding high-load processing core, and extracts the logical addressing region corresponding to the current hardware mount of the high-load processing core. The projection contour of the logical addressing region in the two-dimensional planar layer is marked as the underlying resource dissipation layer.

[0039] The topology map construction module overlays the underlying resource dissipation layer onto a two-dimensional memory topology map that maps the physical address range of the resident core protection processes. The interference analysis module sets the area covered by the underlying resource dissipation layer as an unallocable static obstacle when calculating the Boolean interference results. To handle high-frequency concurrency from massive IoT terminals, the Boolean interference analysis algorithm is offloaded to the hardware acceleration unit for execution. The system monitors the core load status in real time. When the number of concurrent threads on the core continuously exceeds a preset full-load threshold, the underlying resource dissipation layer is overlaid onto the two-dimensional memory topology map and marked as an unallocable static obstacle. The system then guides the addressing release instruction to reset the physical address of subsequent data streams to a load level below 60%. Using storage blocks, local hardware overheating or instruction execution delays are avoided, maintaining the security of concurrent access logic for heterogeneous terminals in node processing. If the dynamic outer boundary of the virtual expansion geometry extends along the time axis and touches the edge coordinates of static obstacles, the interference analysis module outputs a conflict determination signal representing overlapping conflicts. Based on this, the security scheduling execution module issues memory isolation instructions or triggers cross-node routing mechanisms. The mapping and conversion procedure of the one-dimensional physical address space is synchronized with the processor load status monitoring logic, transforming the abnormal computing power attenuation of the underlying hardware into obstacle boundaries in the two-dimensional memory topology map. Distributed network access nodes block address boundary propagation behavior under boundary conditions where local computing resources are fully loaded.

[0040] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A node elastic scheduling method for massive IoT terminal data access, characterized in that, Includes the following steps: The system analyzes the transport layer load characteristics and instantaneous arrival rate of the data stream to be accessed in real time, extracts the memory allocation granularity parameters contained in the transport layer load characteristics, and extracts nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate. The geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node is calculated based on the nonlinear fluctuation characteristic parameters, and a virtual expanded geometry corresponding to the data stream to be accessed is generated based on the memory allocation granularity parameters and the geometric volume expansion rate. Call the underlying memory addressing status data of the distributed network access node to construct a two-dimensional memory topology map that maps the physical address range of the resident core protection process, and project the virtual expansion geometry into the coordinate system of the two-dimensional memory topology map; Under the coordinate system constraints of the two-dimensional memory topology map, Boolean interference analysis is performed on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process, and a conflict determination signal is output to characterize the overlapping state of the physical addressing space. When the collision determination signal is a non-overlapping collision, an addressing release command is issued for the data stream to be accessed, so as to allow it to access the distributed network access node. When the conflict determination signal indicates that there is an overlapping conflict, a memory isolation instruction is issued for the data stream to be accessed to freeze the port mapping of the data stream to be accessed.

2. The node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, The steps of performing Boolean interferometry analysis on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process specifically include: extracting the centroid coordinates of the physical addresses of the resident core protection process in the two-dimensional memory topology map, and delineating a rigid protection area centered on the centroid coordinates of the physical addresses according to the preset access permission level protocol; projecting the virtual expansion geometry along the time axis as a dynamic trajectory band, and calculating the shortest Euclidean distance between the outer envelope of the dynamic trajectory band and the boundary line of the rigid protection area; if the shortest Euclidean distance is less than the preset physical isolation distance threshold, the conflict determination signal is set to overlapping conflict, and the topology coordinates of the collision are recorded; if the shortest Euclidean distance is greater than or equal to the physical isolation distance threshold, the conflict determination signal is set to non-overlapping conflict, and the allowed addressing range of the data stream to be accessed in the distributed network access node is locked.

3. The node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, The steps for calculating the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on nonlinear fluctuation characteristic parameters, and determining the dynamic expansion radius of the virtual expansion geometry based on the following expansion model: , where R(t) is the dynamic expansion radius of the virtual expanding geometry at time t; This is the real-time value of the instantaneous arrival rate; The effective payload size of the data stream to be accessed; α is the preset addressing period constant for distributed network access nodes; α and β are weighted correction coefficients with the dimension of rate-to-space conversion. The weighted correction coefficients are dynamically calculated based on the dispersion index of the currently available memory fragments of the distributed network access nodes.

4. The node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, The steps for constructing a two-dimensional memory topology map mapping the physical address range of resident core protected processes specifically include: obtaining the physical page frame distribution data of all currently active processes by reading the page tables of the memory management units of distributed network access nodes; converting the discrete address segments in the physical page frame distribution data into rectangular placeholders in a two-dimensional Cartesian coordinate system; identifying key processes performing encryption operations among the active processes and marking their corresponding rectangular placeholders as highly sensitive geometric regions; extending a preset address width outward along the outer contour of the highly sensitive geometric region in the two-dimensional Cartesian coordinate system to generate and apply a logical no-fly zone boundary that prohibits ordinary data streams from writing; and generating a two-dimensional memory topology map based on the closed range of the rectangular placeholders and the logical no-fly zone boundary.

5. A node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, The steps for extracting nonlinear fluctuation characteristic parameters based on instantaneous arrival rate specifically include: statistically analyzing the amount of bit stream arrivals per unit time using a sliding window algorithm to generate a time-series waveform reflecting the instantaneous arrival rate; identifying spike pulses in the time-series waveform whose slope exceeds a preset abrupt change threshold, and extracting the actual rising slope of the spike pulse as a nonlinear fluctuation characteristic parameter; and substituting the nonlinear fluctuation characteristic parameter into a preset function mapping relationship to update the benchmark weighting factor for calculating the geometric volume expansion rate.

6. The node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, After issuing the memory isolation command for the data stream to be accessed, the process also includes: reporting alarm information containing conflict coordinates and conflict intensity to the central control plane, where the conflict intensity is determined by the ratio of the overlapping area between the virtual expansion geometry and the rigid protection area; querying the preset backup node resource pool, obtaining the memory free fragmentation ratio of each backup node, and selecting a backup node with a memory free fragmentation ratio higher than the preset available level; routing the access request of the data stream to be accessed to the backup node, and reallocating physical page frames in the two-dimensional memory topology map of the backup node.

7. A node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, After issuing the addressing and release command for the data stream to be accessed, the process also includes: real-time tracking of the actual memory occupancy trajectory of the accessed data stream in the distributed network access nodes; calculating the area deviation between the actual coverage area of ​​the actual memory occupancy trajectory and the nominal coverage area of ​​the virtual expansion geometry predicted before scheduling; updating the algorithm feature weight parameters for calculating the geometric volume expansion rate when the area deviation exceeds the preset fault tolerance deviation threshold; and refreshing the placeholder status in the two-dimensional memory topology map based on the actual memory occupancy trajectory as the underlying addressing status data for subsequent Boolean interferometry analysis.

8. A node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, Before calling the underlying memory addressing status data of the distributed network access node, the process also includes: establishing a mapping rule base from physical memory addresses to a unified virtual coordinate system, using the mapping rule base to standardize fragmented physical memory space into a unified two-dimensional planar layer; continuously monitoring the number of concurrent threads of each processing core inside the distributed network access node, marking the logical addressing area corresponding to the processor core whose number of concurrent threads continuously exceeds a preset full load threshold as the underlying resource dissipation layer; overlaying the underlying resource dissipation layer onto the two-dimensional memory topology map, and setting the area covered by the underlying resource dissipation layer as an unallocatable static obstacle when performing Boolean interference analysis.

9. A node elastic scheduling method for massive IoT terminal data access according to claim 1, characterized in that, The step of calculating the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on nonlinear fluctuation characteristic parameters further includes: calculating the load entropy value of the data stream to be accessed, and confirming that the data stream to be accessed has filling attack characteristics when the load entropy value is lower than a preset normal information flow benchmark judgment threshold; for the data stream to be accessed with filling attack characteristics, extracting a penalty gain coefficient with a value greater than 1, and multiplying the penalty gain coefficient by the initially calculated geometric volume expansion rate to obtain the amplified final geometric volume expansion rate; generating an extended dynamic outer boundary based on the final geometric volume expansion rate to perform Boolean interference analysis. To amplify the probability of collisions, the method also includes: periodically extracting all data stream access records and their resulting physical addressing space overlap conflict event logs within a historical time period; using clustering algorithms to analyze the conflict event logs and statistically analyze the interference frequency of each coordinate block in the two-dimensional memory topology map; identifying coordinate blocks whose interference frequency exceeds a preset frequency threshold and defining them as high interference-sensitive topology regions; issuing a reserved isolation command to permanently lock the high interference-sensitive topology regions in the two-dimensional memory topology map, and forcibly resetting the physical address of ordinary security level data streams to a safe and available space outside the high interference-sensitive topology regions when issuing subsequent addressing release commands.

10. A node elastic scheduling system for massive IoT terminal data access, used to implement the node elastic scheduling method for massive IoT terminal data access as described in claim 1, characterized in that the system... include: The data stream parsing module is used to parse the transport layer payload characteristics and instantaneous arrival rate of the data stream to be accessed in real time, extract the memory allocation granularity parameters contained in the transport layer payload characteristics, and extract the nonlinear fluctuation characteristic parameters based on the instantaneous arrival rate. The geometric expansion calculation module is used to calculate the geometric volume expansion rate of the data stream to be accessed during the physical memory allocation process of the distributed network access node based on the nonlinear fluctuation characteristic parameters, and to generate the virtual expansion geometry corresponding to the data stream to be accessed based on the memory allocation granularity parameters and the geometric volume expansion rate. The topology graph construction module is used to call the underlying memory addressing status data of the distributed network access nodes, construct a two-dimensional memory topology graph that maps the physical address range of the resident core protection processes, and project the virtual expanded geometry into the coordinate system of the two-dimensional memory topology graph. The interference analysis module is used to perform Boolean interference analysis on the dynamic outer boundary of the virtual expansion geometry and the physical address range of the resident core protection process under the coordinate system constraints of the two-dimensional memory topology map, and outputs a conflict determination signal to characterize the overlapping state of the physical address space. The security scheduling execution module is used to issue an addressing and allowing instruction for the data stream to be accessed when the conflict determination signal indicates no overlapping conflict, so as to allow it to access the distributed network access node; when the conflict determination signal indicates overlapping conflict, it issues a memory isolation instruction for the data stream to be accessed, so as to freeze the port mapping of the data stream to be accessed.

Images