Gateway management system and method based on multiple internet of things protocols

Abstract

The invention relates to the technical field of Internet of Things communication, and discloses a gateway management system and method based on multiple Internet of Things protocols, and the system comprises a processor architecture, a protocol conversion module, a collaborative management module, a unified security framework module and an edge computing module. The method is applied to the system. According to the application, efficient interconnection and dynamic collaborative management of multi-protocol equipment are realized through FPGA module hardware acceleration protocol conversion, a genetic algorithm and a Petr network hybrid task scheduling framework and a unified security mechanism irrelevant to a protocol, and various different mainstream Internet of Things protocols are compatible through a modular processor architecture and a dynamic expansion interface, so that the implementation is convenient, and the implementation is easy. According to the method, the hardware cost is reduced, rapid integration of a new protocol is supported, local and cloud resource allocation is further optimized through an intelligent switching mechanism of edge computing and cloud collaboration, the system stability and the data processing efficiency are improved, and the method is widely applied to scenes such as smart home, industrial Internet of Things and smart cities.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things communication, in particular to a gateway management system and method based on multiple Internet of Things protocols. BACKGROUND

[0002] With the rapid development of Internet of Things technology, various intelligent devices widely use multiple communication protocols (such as Zigbee, LoRa, Modbus, MQTT, CoAP, Matter, etc.) for data interaction. However, the differences in physical layer, data link layer and application layer of different protocols lead to interoperability problems, and existing Internet of Things gateway systems face significant challenges in multi-protocol compatibility, dynamic resource allocation, real-time guarantee and security.

[0003] Specifically, traditional Internet of Things gateways rely on software parsing to achieve multi-protocol conversion, and the layer-by-layer parsing mechanism leads to high latency, especially when converting Zigbee, Modbus, etc. and MQTT, HTTP, etc. application layer protocols, real-time performance cannot meet the high requirements of industrial automation, smart home, etc. For example, in Zigbee to MQTT conversion, the average delay of software parsing can reach more than 50ms, which cannot support real-time control requirements sensitive to latency. In addition, the software parsing method is difficult to handle high-concurrency protocol conversion tasks, resulting in low system resource utilization and limited processing capacity. Secondly, existing gateways lack efficient dynamic resource allocation and load balancing mechanisms when facing frequent device additions and deletions, network fluctuations or burst tasks. Traditional load balancing strategies are mostly for server clusters and do not solve the real-time resource conflict problem of cross-protocol devices. For example, in the industrial Internet of Things scenario, when Modbus and OPCUA protocol devices are connected at the same time, if the resources of ARM processors and FPGA modules are not reasonably allocated, it may cause key tasks (such as production line control) to be delayed due to resource competition, affecting production efficiency. In addition, the scheduling algorithms of existing gateways are mostly static rules, which cannot dynamically optimize the execution path according to task priority, resource state and protocol characteristics.

[0004] Therefore, the existing Internet of Things gateway has the problems of insufficient multi-protocol compatibility and weak dynamic resource scheduling capability. Therefore, there is an urgent need for a new multi-protocol gateway management technology to promote the interconnection and efficient collaboration of Internet of Things devices. SUMMARY

[0005] The purpose of the present application is to provide a gateway management system and method based on multiple Internet of Things protocols to solve the technical problems raised in the background art.

[0006] To achieve the above purpose, the present application discloses the following technical solutions:

[0007] In a first aspect, the application discloses a gateway management system based on multiple Internet of Things protocols, which comprises:

[0008] A processor architecture, comprising an ARM processor and an FPGA module, the ARM processor being used for performing system management, edge computing task scheduling and northbound interface docking, and the FPGA module being used for performing protocol conversion and real-time data stream processing;

[0009] A protocol conversion module configured in the FPGA module, which performs efficient conversion of different Internet of Things protocols through parallel processing;

[0010] A collaborative management module, which is used for dynamically allocating cross-protocol task resources and optimizing task execution paths based on a hybrid task scheduling algorithm of genetic algorithm and Petri net;

[0011] A unified security framework module, which is used for protocol-independent key management, integration of cross-protocol security mechanisms and zero-trust architecture, and unified management of symmetric encryption and asymmetric encryption;

[0012] An edge computing module, which is used for local data analysis and dynamic adjustment of task priority according to the analysis result.

[0013] Preferably, the hybrid task scheduling algorithm comprises:

[0014] Genetic algorithm optimization task allocation: multiple protocol tasks are encoded into chromosomes, task priority, resource utilization and time delay are evaluated through an adaptive function, and an optimal scheduling path is generated based on the fitness value of the chromosome calculated by the adaptive function;

[0015] Petri net real-time monitoring and adjustment: a Petri net service model is established, resource conflicts are processed through fuzzy Petri net, and task dependency relationships are processed through a directed acyclic graph model.

[0016] Preferably, the multiple protocol tasks are encoded into chromosomes in the form of a three-dimensional matrix, wherein the first dimension corresponds to the hardware resource units of the ARM processor and the FPGA module, the second dimension corresponds to the task instances of different Internet of Things protocols, and the third dimension represents the task priority level.

[0017] The adaptive function comprises a weighted sum of the relationship between task priority and deadline, system resource utilization and system time delay to obtain the fitness value.

[0018] Preferably, the generation of the optimal scheduling path based on the fitness value of the chromosome calculated by the adaptive function comprises:

[0019] Initialization of population: an initial population is formed based on the generated chromosomes;

[0020] Computing fitness: calculating the fitness value of each chromosome based on the fitness function;

[0021] Selection operation: performing selection operation based on the fitness value of each chromosome;

[0022] Crossover operation: pairing the selected chromosomes two by two and exchanging part of the gene fragments according to the crossover probability;

[0023] Mutation operation: randomly changing part of the genes of the chromosomes with a mutation probability that meets a preset mutation probability threshold condition;

[0024] Iterative optimization: repeating the calculation of fitness, selection operation, crossover operation and mutation operation for multiple generations of evolution, and stopping iteration when the iteration termination condition is met;

[0025] Determining the optimal scheduling path: after iteration, selecting the chromosome with the highest fitness value in the last generation population, and taking the scheduling scheme represented by the chromosome as the optimal scheduling path.

[0026] Preferably, the Petri net service model is established by processing resource conflicts through fuzzy Petri nets and processing task dependency relationships through a directed acyclic graph model, comprising:

[0027] Place design: the places include task queue places, CPU resource places and FPGA module resource places, which are represented by different color tokens to represent protocol types and resource occupation states;

[0028] Transition trigger condition: including task dependency label matching and resource place token threshold, wherein the task dependency label matching is to verify the completion state of the pre-task through the topological sorting of the directed acyclic graph model, and if the dependent task is not completed, the token flow of the subsequent task is blocked; the resource place token threshold is that when the CPU resource place token exceeds 70%-80% of the task demand and the FPGA module resource place token is greater than or equal to 100% of the task demand, the task execution is triggered;

[0029] The degree of membership function is used to quantify the urgency of the task, and when multiple tasks compete for the same resource, the resource is preferentially allocated to the task with higher urgency.

[0030] Preferably, the unified security framework module comprises:

[0031] Protocol-independent key management unit, for supporting key full life cycle management, the full life cycle management includes key generation, key storage, key distribution, key rotation and key destruction;

[0032] Cross-protocol security middleware, for uniformly processing the compatibility policy of TLS and AES, and implementing role-based access control;

[0033] A zero-trust architecture for verifying the legitimacy of device identity through a ternary peer-to-peer digital certificate authentication and for attack protection through vulnerability scanning and denial of service.

[0034] Preferably, the key generation includes generating a 128-bit key as a symmetric key based on the SM4 algorithm, generating a 512-bit key pair as an asymmetric key based on the SM2 algorithm, and using a hardware true random number generator to make the key entropy value not less than 256 bits during the key generation process.

[0035] The key storage includes splitting the master key into three parts and storing them in the secure area of the ARM processor, the non-volatile storage unit of the FPGA module, and the cloud encryption database.

[0036] The key distribution includes transmitting the key using a secure channel, carrying the device fingerprint in the data packet header, and using AES-256-GCM encryption during the key transmission process.

[0037] Preferably, the ternary peer-to-peer digital certificate includes:

[0038] The device certificate includes a device unique identifier, a device fingerprint, and a certificate validity period.

[0039] The gateway certificate includes a gateway public key, a CA agency digital signature, and a supported protocol list.

[0040] The cloud certificate includes a cloud server endpoint URL, a TLS server name indication, and a CRL update address.

[0041] Preferably, the local data analysis and dynamic adjustment of task priority based on the analysis results include:

[0042] The local data analysis is performed through a lightweight AI framework, and a collaborative mode of edge computing and cloud processing is dynamically selected based on the network state and task requirements; the selection mechanism of the collaborative mode is:

[0043] Local processing mode trigger condition: when the network delay is ≥ 50 ms and the task data volume is ≥ 10 MB, the local processing mode is automatically triggered, and the local processing preferentially uses a lightweight AI framework for data compression;

[0044] Cloud collaborative mode trigger condition: when new protocol data is detected or global data aggregation across gateways is needed, the cloud collaborative processing mode is switched to, and a Kafka message queue is enabled for multi-gateway data stream processing; the dynamic selection algorithm is implemented based on a decision tree model, the input parameters of the decision tree model include network packet loss rate, task real-time level, and data privacy sensitivity, and the output of the decision tree model is 0 representing the local processing mode or 1 representing the cloud collaborative mode.

[0045] In a second aspect, the application discloses a gateway management method based on multiple Internet of Things protocols, applied to the gateway management system based on multiple Internet of Things protocols as described above, and characterized in that the method comprises the following steps:

[0046] The processor architecture composed of the ARM processor and the FPGA module is adopted, the ARM processor is used to perform system management, edge computing task scheduling and northbound interface docking, and the FPGA module is used to perform protocol conversion and real-time data stream processing;

[0047] In the protocol conversion of the FPGA module, different Internet of Things protocols are efficiently converted through parallel processing;

[0048] The hybrid task scheduling algorithm based on the genetic algorithm and the Petri net is adopted to dynamically allocate cross-protocol task resources and optimize the task execution path;

[0049] The unified security framework module integrating the protocol-independent key management, the cross-protocol security mechanism integration and the zero-trust architecture is used to perform unified management of symmetric encryption and asymmetric encryption;

[0050] Based on the edge computing module, local data analysis is performed, and task priority is dynamically adjusted according to the analysis result.

[0051] Beneficial effects: The gateway management system and method based on multiple Internet of Things protocols of the application realize efficient interconnection and dynamic collaborative management of multiple protocol devices through FPGA module hardware acceleration protocol conversion, genetic algorithm and Petri net hybrid task scheduling framework and protocol-independent unified security mechanism. Through the modular processor architecture and the dynamic expansion interface, multiple different mainstream Internet of Things protocols are compatible, the hardware cost is reduced, and new protocols are quickly integrated. Through the intelligent switching mechanism of edge computing and cloud cooperation, the local and cloud resource allocation is further optimized, the system stability and data processing efficiency are improved, and the system is widely applicable to smart home, industrial Internet of Things and smart city scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description only some embodiments of the application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0053] Figure 1 The structural block diagram of the gateway management system based on multiple Internet of Things protocols provided by the embodiments of the application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0055] In this document, the term "comprising" is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such a process, method, article or device. Without more limitations, the elements defined by the statement "comprising" do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0056] The embodiments provide a gateway management system based on multiple Internet of Things protocols in a first aspect. Figure 1 The system includes:

[0057] Structure one: processor architecture, including an ARM processor and an FPGA module. The ARM processor is used for system management, edge computing task scheduling and northbound interface docking, and the FPGA module is used for protocol conversion and real-time data stream processing, and is interconnected with the ARM processor through an interface.

[0058] Structure two: a protocol conversion module configured in the FPGA module, which efficiently converts different Internet of Things protocols through parallel processing, and the Internet of Things protocols include several of Zigbee, LoRa, Modbus, MQTT, CoAP and Matter protocols.

[0059] Structure three: a cooperative management module, which is used for dynamically allocating cross-protocol task resources and optimizing task execution paths based on a hybrid task scheduling algorithm of genetic algorithm and Petri net. The Petri net is a mathematical model suitable for describing asynchronous and concurrent computer system models, and is a mathematical representation of discrete parallel systems.

[0060] Specifically, the hybrid task scheduling algorithm used in the embodiments includes:

[0061] Genetic algorithm optimization task allocation: encode multiple protocol tasks into chromosomes, evaluate task priority, resource utilization and time delay through an adaptability function, and generate an optimal scheduling path based on the adaptability value of the chromosome calculated by the adaptability function;

[0062] Real-time monitoring and adjustment of Petri net: a Petri net service model is established, resource conflicts are handled through fuzzy Petri net, and task dependency relationship is handled through directed acyclic graph model.

[0063] Further, the encoding of the multi-protocol task into the chromosome includes encoding the multi-protocol task into the chromosome in a three-dimensional matrix form, wherein the first dimension corresponds to the hardware resource units of the ARM processor and the FPGA module, the second dimension corresponds to the task instances of different Internet of Things protocols, and the third dimension represents the task priority level (urgent / regular / low priority).

[0064] The fitness function includes a weighted sum of the relationship between task priority and deadline, system resource utilization, and system delay condition to obtain a fitness value.

[0065] Feasibly, the fitness function is specifically:

[0066]

[0067] Wherein, α, β and γ are weight coefficients, and satisfy α+β+γ=1;TaskPriority i TaskPriority is the task priority, which is a 0-1 normalized value;Deadline i Deadline is the task deadline;CerrentTime is the current timestamp;ResourceUtilization CPU ResourceUtilization is the resource utilization of the CPU in the ARM processor, which is a 0-1 normalized value;ResourceUtilization FPGA ResourceUtilization is the resource utilization of the FPGA module, which is a 0-1 normalized value;MaxDelay is the maximum delay allowed by the system;CurrentDelay is the current delay;max(·,1) is used to avoid zero denominator. The calculation of the relationship between task priority and deadline tends to prioritize tasks with high priority and urgent deadlines. If a solution can better complete such tasks, the calculation result of the first item will be higher, thereby improving the overall fitness. For the calculation of system resource utilization, if the solution can balance the use of hardware resources and avoid overloading or idling of a module, the calculation result of this part will be higher, which helps to improve the fitness. The calculation of the system delay condition reflects the consideration of task processing delay. The smaller the current delay, the larger the calculation result of this part, and the higher the fitness, that is, low-delay scheduling solutions are encouraged.

[0068] Feasibly, the fitness value of the chromosome calculated based on the fitness function generates an optimal scheduling path, including:

[0069] Initialization of population: based on the generated chromosomes to form the initial population, each chromosome represents a scheme to allocate different Internet of Things protocol tasks (Zigbee, LoRa, etc.) to ARM processor and FPGA module resource units;

[0070] Calculation of fitness: based on the fitness function to calculate the fitness value of each chromosome;

[0071] Selection operation: based on the fitness value of each chromosome to select the better chromosomes from the current population into the next generation, common selection methods such as roulette selection, tournament selection, etc. Take the roulette selection method as an example, the probability of each chromosome being selected is proportional to its fitness value, the higher the fitness value, the larger the "sector area" it occupies on the roulette, the higher the probability of being selected. Through selection operation, the excellent scheduling scheme is retained and the poor scheme is eliminated;

[0072] Cross operation: pair the selected chromosomes two by two, exchange part of the gene fragments according to the crossover probability; for example, for two two-dimensional matrix encoded chromosomes, randomly select part of the rows or columns in the matrix to exchange, thus generating a new scheduling scheme. Cross operation helps to inherit the excellent characteristics of parent chromosomes and generate new combinations;

[0073] Mutation operation: with a preset mutation probability threshold condition (such as 0.01-0.05), the mutation probability is used to randomly change part of the genes of the chromosome, such as changing the allocation position of a task in the matrix, mutation operation can increase the diversity of the population and prevent the algorithm from falling into local optimal solution;

[0074] Iterative optimization: repeat the calculation of fitness, selection operation, cross operation and mutation operation for multiple generations of evolution, after multiple generations of evolution, the chromosomes in the population will gradually approach the optimal solution (i.e. the optimal scheduling path); when the iteration termination condition (such as the continuous 10 generations of fitness value fluctuation ≤5%) is met, stop iteration;

[0075] Determination of optimal scheduling path: after the iteration is completed, in the last generation population, select the chromosome with the highest fitness value, and take the scheduling scheme represented by the chromosome as the optimal scheduling path, which achieves a good balance in task priority processing, resource utilization and delay control, etc.

[0076] Through the above process, the genetic algorithm uses the calculation results of the fitness function to screen and optimize the scheduling scheme in the continuous evolution process, and finally determines the optimal scheduling path for multiple Internet of Things protocol tasks, realizing the efficient allocation of resources and the optimal execution of tasks in the gateway management system.

[0077] It is feasible that the Petri net service model is established, resource conflicts are handled through the fuzzy Petri net, and task dependency relationships are handled through the directed acyclic graph model, including:

[0078] The library design includes a task queue library, a CPU resource library, and an FPGA module resource library, which are represented by tokens of different colors to represent protocol types and resource occupation states.

[0079] The transition trigger condition includes task dependency label matching and resource library token threshold value, wherein the task dependency label matching is that the completion state of a preceding task is verified through topological sorting of the directed acyclic graph model, and if a dependent task is not completed, the token flow of a subsequent task is blocked; and the resource library token threshold value is that when the CPU resource library token exceeds 70%-80% of the task demand and the FPGA module resource library token is greater than or equal to 100% of the task demand, the task execution is triggered; the resource library can be understood as a container for storing system resources, and the token represents the number or state of resources, and when the number of tokens in the resource library reaches or exceeds the number of resources required for the current task (i.e., the task demand threshold), this condition is met. For example, a task requires 5 units of computing resources, and only when the number of tokens in the resource library reaches 5 or more, the condition meets the transition trigger requirement. Only when the two conditions are met at the same time, the transition is triggered, thereby promoting the change of the system state;

[0080] The task urgency is quantified through the membership function, when multiple tasks compete for the same resource, the resource is preferentially allocated to the task with higher task urgency, i.e., the task with more urgent remaining execution time, larger data quantity, and larger ratio of hardware processing capacity, so as to dynamically balance resource use, reduce task delay, and improve overall system operation efficiency. Secondly, if the resource is insufficient, the task queue buffer is triggered.

[0081] In this embodiment, the calculation expression of the membership function is:

[0082]

[0083] wherein T low and T high are threshold values of task urgency, and x is the remaining execution time of the task.

[0084] Structure four: a unified security framework module for protocol-independent key management, cross-protocol security mechanism integration, and zero-trust architecture, which uniformly manages symmetric encryption and asymmetric encryption.

[0085] Specifically, the unified security framework module includes:

[0086] A protocol-independent key management unit is configured to support key life cycle management, including key generation, key storage, key distribution, key rotation, and key destruction.

[0087] A cross-protocol security middleware is configured to uniformly process compatibility policies of TLS and AES, and implement role-based access control; TLS refers to a transport layer security protocol configured to provide confidentiality, data integrity, and authenticity between two communicating applications; AES refers to an advanced encryption standard in cryptography, which is a block encryption standard.

[0088] A zero-trust architecture is configured to verify the legitimacy of device identity through a ternary peer-to-peer digital certificate authentication, and to perform attack protection through vulnerability scanning and denial of service. The vulnerability scanning module supports real-time updating of the CVE vulnerability library, and the scanning period can be configured to be 5 minutes to 24 hours. The scanning strategies include protocol stack version detection (such as Modbus TCP 1.1b), default credential scanning, and open port scanning. The denial of service attack protection module detects abnormal traffic based on entropy value. The calculation formula for detecting abnormal traffic is: Where p i is the proportion of the ith traffic. When the traffic entropy value suddenly increases and the traffic rate is greater than or equal to 150% of the throughput, the traffic cleaning strategy (such as discarding non-TLS traffic) is triggered.

[0089] Possibly:

[0090] The key generation includes: generating a 128-bit key as a symmetric key based on the SM4 algorithm, and generating a 512-bit key pair as an asymmetric key based on the SM2 algorithm. The key entropy value is not less than 256 bits through a hardware true random number generator during the key generation process.

[0091] The key storage includes: splitting the master key into three parts and storing them in the secure area of the ARM processor, the non-volatile storage unit of the FPGA module, and the cloud encryption database.

[0092] The key distribution includes: transmitting the key using a secure channel (such as TLS 1.3), and the transmission data packet header carries the device fingerprint (such as the SHA-256 hash value of the MAC address + CPU serial number). The key transmission process uses AES-256-GCM encryption.

[0093] The ternary peer-to-peer digital certificate includes:

[0094] The device certificate includes a device unique identifier, a device fingerprint, and a certificate validity period.

[0095] The gateway certificate includes a gateway public key, a CA agency digital signature, and a supported protocol list.

[0096] Cloud certificates, including cloud service endpoint URLs, TLS server name indication, and CRL update addresses.

[0097] Structure five: edge computing module, for local data analysis and dynamic adjustment of task priority according to analysis results.

[0098] Specifically, the local data analysis and dynamic adjustment of task priority according to analysis results include:

[0099] Perform local data analysis through a lightweight AI framework, and dynamically select a collaborative mode of edge computing and cloud processing according to network status and task requirements; wherein the selection mechanism of the collaborative mode is:

[0100] Local processing mode trigger condition: when network delay ≥ 50 ms and task data volume ≥ 10 MB, automatically trigger local processing mode, and local processing preferentially uses a lightweight AI framework for data compression;

[0101] Cloud collaboration mode trigger condition: When a new protocol data (such as an unregistered protocol type or a new version of a registered protocol) is detected or global data aggregation across gateways is required, switch to cloud collaboration processing mode, enable Kafka message queue to realize multi-gateway data stream processing; the dynamic selection algorithm is realized based on a decision tree model, and the input parameters of the decision tree model include: network packet loss rate, task real-time level (divided into emergency / regular / asynchronous three levels) and data privacy sensitivity (according to the GDPR data classification standard), and the output of the decision tree model is 0 representing local processing mode or 1 representing cloud collaboration mode. Kafka platform is a distributed stream processing platform that supports processing millions of messages per second, suitable for real-time processing of large-scale device data streams in the Internet of Things, messages can be persisted to disk, supporting data backtracking and fault recovery, combined with KafkaStreams or Flink tools, supporting real-time data analysis and processing. The main functions of the Kafka message queue include: when multiple gateways need to collaborate on tasks (such as cross-device linkage, global data analysis), the Kafka platform acts as a centralized message queue to unify and converge data streams from different gateways, that is: each gateway publishes local processed data (such as sensor data, device status) to a specific topic of the Kafka platform, and cloud consumers subscribe to these topics to aggregate and analyze cross-gateway data streams in real time; when a new protocol (such as Zigbee3.1) is detected, the gateway cannot be locally parsed and needs to send data to the cloud for unified processing, that is: the gateway encapsulates the original protocol data into Kafka messages and sends them to the unknown protocol topic in the cloud, and the cloud service dynamically loads the new protocol parsing module to process the results and return them to the gateway through the Kafka platform; when the network delay is low, the task is preferentially processed at the edge, and when the cloud demand (such as global data aggregation) is detected, it is dynamically switched to the cloud mode, that is: the gateway decides whether to send data to the Kafka platform according to the network state (such as RTT≥50ms) and task demand, and the cloud consumer pulls data from the Kafka platform, executes complex analysis (such as machine learning model training), and sends the results to the gateway.

[0102] In summary, the gateway management system based on multiple Internet of Things protocols in the embodiment has a protocol conversion efficiency that is greatly improved compared with traditional software analysis methods through the FPGA module parallel processing mechanism (such as a Zigbee physical layer demodulator and an MQTT message parser), meets the needs of scenarios with high real-time requirements, reduces the power consumption of the FPGA module and solves the clock competition problem through dynamic clock enablement and independent clock domain design, and prolongs the service life of the device; the optimal scheduling path is generated through a genetic algorithm, the resource allocation is adjusted in real time through a Petri net, the combination of global optimization and real-time monitoring is achieved, the response time of cross-protocol devices and the utilization rate of task devices are shortened, the local processing or cloud collaboration mode is dynamically selected based on a decision tree model, the network load is reduced, and the real-time performance of critical tasks is ensured; the uniform security framework that is independent of protocols ensures the security of keys and implements fine-grained permission management, and the three-way peer-to-peer digital certificate authentication blocks illegal access, and vulnerability scanning and denial-of-service attack protection improve the robustness of the system.

[0103] Therefore, the gateway management system based on multiple Internet of Things protocols in the embodiment solves the industry pain points of multi-protocol Internet of Things gateways in terms of compatibility, real-time performance, collaborative management, and security through three core technologies of FPGA hardware acceleration, hybrid scheduling algorithms, and a uniform security framework. The following comprehensive advantages are achieved:

[0104] Technical advancement: the FPGA gated clock design, the genetic algorithm, and the hybrid scheduling framework of the Petri net all have high technical advancement;

[0105] Industry applicability: covers multiple scenario needs such as smart home, industrial Internet of Things, and smart city;

[0106] Commercial value: reduces system costs, improves operation and maintenance efficiency, and supports the rapid integration of emerging protocols (such as Matter).

[0107] In a second aspect, the embodiment provides a gateway management method based on multiple Internet of Things protocols, applied to the gateway management system based on multiple Internet of Things protocols as described above, and characterized in that the method comprises the following steps:

[0108] A processor architecture composed of an ARM processor and an FPGA module is used, the ARM processor performs system management, edge computing task scheduling, and northbound interface docking, and the FPGA module performs protocol conversion and real-time data stream processing;

[0109] In the protocol conversion of the FPGA module, different Internet of Things protocols are efficiently converted through parallel processing, and the Internet of Things protocols include several of Zigbee, LoRa, Modbus, MQTT, CoAP, and Matter protocols;

[0110] The hybrid task scheduling algorithm based on the genetic algorithm and the Petri net is adopted to dynamically allocate cross-protocol task resources and optimize a task execution path.

[0111] The unified security framework module of the integrated protocol-independent key management, cross-protocol security mechanism integration and zero-trust architecture is used to perform unified management of symmetric encryption and asymmetric encryption.

[0112] Based on the edge computing module, local data analysis is performed, and task priority is dynamically adjusted according to the analysis result.

[0113] It should be noted that the gateway management method based on the multiple Internet of Things protocols in the embodiment corresponds to the gateway management system based on the multiple Internet of Things protocols described above, and therefore, parts (including but not limited to specific implementation technical means and technical effects) not specifically described in the gateway management method based on the multiple Internet of Things protocols in the embodiment can be correspondingly referred to the related description in the gateway management system based on the multiple Internet of Things protocols described above, which will not be repeated here.

[0114] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be realized in hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor can be realized in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA module), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be instructed by a computer program to relevant hardware. In implementation, the above program can be stored in a computer readable storage medium or transmitted as one or more instructions or codes on a computer readable storage medium. The computer readable storage medium includes a computer storage medium and a communication medium, wherein the communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that can be accessed by a computer. The computer readable storage medium can include but is not limited to RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program codes in the form of instructions or data structures and can be accessed by a computer.

[0115] Finally, it should be noted that the above only describes the preferred embodiments of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, modifications or equivalent replacements of the technical solutions described in the foregoing embodiments can still be made by those skilled in the art, or some technical features can be replaced by equivalent features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A gateway management system based on multiple Internet of Things protocols, characterized in that: The system includes: Processor architecture, including an ARM processor and an FPGA module. The ARM processor is used to perform system management, edge computing task scheduling, and northbound interface docking, and the FPGA module is used to perform protocol conversion and real-time data stream processing; A protocol conversion module, configured in the FPGA module, efficiently converts different IoT protocols through parallel processing; Collaborative management module, which is used to dynamically allocate cross-protocol task resources and optimize task execution paths based on a hybrid task scheduling algorithm based on genetic algorithms and Petri nets; Unified security framework module for protocol-independent key management, cross-protocol security mechanism integration and zero-trust architecture, and unified management of symmetric and asymmetric encryption; Edge computing module, used for local data analysis and dynamic adjustment of task priorities based on analysis results.

2. The gateway management system based on multiple Internet of Things protocols according to claim 1 is characterized in that: The hybrid task scheduling algorithm includes: Genetic algorithm optimizes task allocation: Multi-protocol tasks are encoded as chromosomes, and task priority, resource utilization, and latency are evaluated using a fitness function. The optimal scheduling path is generated based on the chromosome fitness value calculated by the fitness function. Petri net real-time monitoring and adjustment: Establish a Petri net service model, handle resource conflicts through fuzzy Petri nets, and handle task dependencies through a directed acyclic graph model.

3. The gateway management system based on multiple Internet of Things protocols according to claim 2 is characterized in that: Encoding the multi-protocol task into a chromosome includes encoding the multi-protocol task into a chromosome in a three-dimensional matrix form, wherein a first dimension corresponds to a hardware resource unit of the ARM processor and the FPGA module, a second dimension corresponds to a task instance of different IoT protocols, and a third dimension represents a task priority level; The fitness function includes weighted summing of the relationship between task priority and deadline, system resource utilization and system delay to obtain a fitness value.

4. The gateway management system based on multiple Internet of Things protocols according to claim 3 is characterized in that: The step of generating an optimal scheduling path based on the fitness value of the chromosome calculated based on the fitness function includes: Initialize the population: Construct the initial population based on the generated chromosomes; Calculate fitness: Calculate the fitness value of each chromosome based on the fitness function; Selection operation: selection operation is performed based on the fitness value of each chromosome; Crossover operation: pair the selected chromosomes and exchange some gene fragments according to the crossover probability; Mutation operation: Randomly change some genes of the chromosome with a mutation probability that meets the preset mutation probability threshold condition; Iterative optimization: Repeatedly calculate fitness, select operations, crossover operations, and mutation operations for multiple generations of evolution. When the iteration termination condition is met, the iteration stops. Determine the optimal scheduling path: After the iteration, in the last generation of the population, select the chromosome with the highest fitness value, and use the scheduling solution represented by the chromosome as the optimal scheduling path.

5. The gateway management system based on multiple Internet of Things protocols according to claim 2 is characterized in that: The Petri net service model is established, resource conflicts are handled by using fuzzy Petri nets, and task dependencies are handled by using a directed acyclic graph model, including: Place design: Places include task queue places, CPU resource places, and FPGA module resource places. Different color tokens are used to indicate protocol type and resource occupancy status. Transition trigger conditions: including task dependency tag matching and resource library token threshold, wherein the task dependency tag matching is: verifying the completion status of the predecessor task through the topological sorting of the directed acyclic graph model, and blocking the token flow of the subsequent task if the dependent task is not completed; the resource library token threshold is when the CPU resource library token exceeds 70% to 80% of the task requirement and the FPGA module resource library token is greater than or equal to 100% of the task requirement, triggering task execution The urgency of tasks is quantified by the membership function. When multiple tasks compete for the same resource, resources are allocated preferentially to tasks with higher urgency.

6. The gateway management system based on multiple Internet of Things protocols according to claim 1, characterized in that: The unified security framework module includes: A protocol-independent key management unit that supports full key lifecycle management, including key generation, key storage, key distribution, key rotation, and key destruction. Cross-protocol security middleware for unified processing of TLS and AES compatibility policies and implementation of role-based access control; Zero-trust architecture is used to verify the legitimacy of device identity through three-way peer-to-peer digital certificate authentication and to protect against attacks through vulnerability scanning and denial of service.

7. The gateway management system based on multiple Internet of Things protocols according to claim 6, characterized in that: The key generation includes: generating a 128-bit key as a symmetric key based on the SM4 algorithm, generating a 512-bit key pair as an asymmetric key based on the SM2 algorithm, and during the key generation process, using a hardware true random number generator to ensure that the key entropy value is not less than 256 bits; The key storage includes: splitting the master key into three parts, storing them in the secure area of ​​the ARM processor, the non-volatile storage unit of the FPGA module and the cloud encryption database respectively; The key distribution includes: using a secure channel to transmit the key, the transmission data packet header carries the device fingerprint, and the key transmission process is encrypted using AES-256-GCM.

8. The gateway management system based on multiple Internet of Things protocols according to claim 6, characterized in that: The three-element digital certificate includes: Device certificate, including the device's unique identifier, device fingerprint, and certificate validity period; Gateway certificate, including gateway public key, CA digital signature, and supported protocols list; Cloud certificate, including the cloud service endpoint URL, TLS server name indicator, and CRL update address.

9. The gateway management system based on multiple Internet of Things protocols according to claim 1, characterized in that: The local data analysis and dynamic adjustment of task priorities based on the analysis results include: A lightweight AI framework is used to perform local data analysis, and the collaborative mode between edge computing and cloud processing is dynamically selected based on network status and task requirements. The collaborative mode selection mechanism is as follows: Local processing mode triggering conditions: When the network delay is ≥50ms and the task data volume is ≥10MB, local processing mode is automatically triggered. Local processing prioritizes the use of a lightweight AI framework for data compression. Cloud collaboration mode triggering conditions: When new protocol data is detected or cross-gateway global data aggregation is required, it switches to cloud collaboration processing mode and enables Kafka message queues to implement multi-gateway data streaming processing; the dynamic selection algorithm is implemented based on a decision tree model. The input parameters of the decision tree model include: network packet loss rate, task real-time level and data privacy sensitivity. The output of the decision tree model is 0 for local processing mode or 1 for cloud collaboration mode.

10. A gateway management method based on multiple Internet of Things protocols, applied to the gateway management system based on multiple Internet of Things protocols according to any one of claims 1 to 9, characterized in that: The method comprises the following steps: Through the processor architecture composed of ARM processor and FPGA module, the ARM processor performs system management, edge computing task scheduling and northbound interface docking, while the FPGA module performs protocol conversion and real-time data stream processing; In the protocol conversion of the FPGA module, different IoT protocols are efficiently converted through parallel processing; A hybrid task scheduling algorithm based on genetic algorithm and Petri net is used to dynamically allocate cross-protocol task resources and optimize task execution paths; Use a unified security framework module that integrates protocol-independent key management, cross-protocol security mechanism integration, and zero-trust architecture to unify the management of symmetric and asymmetric encryption; Based on the edge computing module, local data analysis is performed and task priorities are dynamically adjusted based on the analysis results.

Images