Soft bus power terminal heterogeneous interconnection method and system based on Ethernet protocol stack

Abstract

The embodiment of the invention discloses a soft bus power terminal heterogeneous interconnection method and system based on an Ethernet protocol stack. The method comprises the following steps: dynamically calculating fragment parameters according to the difference of different protocol stacks in heterogeneous network interconnection; based on a dynamic programming algorithm, generating a self-adaptive fragmentation scheme according to the input parameters, and forming an integrated fragmentation recombination framework; based on an intelligent decision-making mechanism, an optimal fragmentation strategy and a recombination mechanism are generated, a simulation platform is established, different network scenes and service flows are simulated, a simulation result of the simulation platform is combined with a result of the intelligent decision-making mechanism, and the effect of an intelligent decision-making algorithm is optimized. And transmitting and recombining the data packet by adopting an encryption and authentication mechanism. According to the scheme, the efficiency and the reliability of data transmission in the heterogeneous network are remarkably improved, the compatibility problem among different network protocol stacks is effectively solved, the fragmentation frequency during cross-network transmission is reduced, the time sequence and the integrity of data packets are ensured, and the processing overhead and the network delay are reduced.

Description

Technical Field

[0001] The present application relates to the technical field of data transmission, and in particular to a method and system for heterogeneous interconnection of soft bus power terminals based on an Ethernet protocol stack. Background Art

[0002] In the scenario of connecting across heterogeneous networks, the data fragmentation and reassembly strategies between different network protocol stacks have encountered a series of technical difficulties and challenges. First, due to the significant differences in the maximum transmission unit (MTU) of different networks, data packets must undergo the process of fragmentation and reassembly when traversing these networks, but the fragmentation and reassembly methods of each protocol stack are different, and there is a lack of unified specifications and standards. Secondly, the fragmentation and reassembly process must ensure the timing and integrity of the data packets to ensure that the data packets can reach the destination in the correct order and be accurately reassembled, which puts forward high technical requirements for the design and implementation of the protocol stack. In addition, the fragmentation and reassembly operations bring additional processing overhead and network delay, which affects the transmission efficiency and quality of the network. Summary of the invention

[0003] The embodiments of the present application provide a method and system for heterogeneous interconnection of soft bus power terminals based on the Ethernet protocol stack, so as to improve the efficiency and reliability of data transmission in heterogeneous networks, effectively solve the compatibility issues between different network protocol stacks, reduce the number of fragmentation times during cross-network transmission, ensure the timing and integrity of data packets, and reduce processing overhead and network latency.

[0004] According to one aspect of the present application, a method for heterogeneous interconnection of soft bus power terminals based on an Ethernet protocol stack is provided, the method comprising:

[0005] Dynamically calculate fragmentation parameters for fragmenting data packets according to differences in different protocol stacks in heterogeneous network interconnection, and define a common data packet format and reassembly rules to form a unified reassembly protocol, so that the unified reassembly protocol is deployed in the network layer, the transport layer and the application layer, so that the network layer, the transport layer and the application layer process the data packet based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field;

[0006] In the fragmentation reassembly process, based on the dynamic programming algorithm, an adaptive fragmentation scheme is generated according to the input parameters, and a unified reassembly protocol is generated according to the adaptive fragmentation scheme, and the encapsulation and decapsulation algorithm of the data packet is optimized. According to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm, an integrated fragmentation reassembly framework is formed; wherein the input parameters include the maximum transmission unit of each protocol layer and the data packet format information;

[0007] In the integrated sharding and reorganization scheme, based on the intelligent decision-making mechanism, the optimal sharding strategy and reorganization mechanism are generated according to the predicted network status data, business type data and service quality requirement data, and a sharding and reorganization simulation platform is established to simulate different network scenarios and business traffic, evaluate the feasibility and effectiveness of the sharding and reorganization scheme, and combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism to verify and optimize the effect of the intelligent decision-making algorithm;

[0008] In the process of fragmentation and reassembly, encryption and authentication mechanisms are adopted, modular design and standardized interfaces are adopted, the data exchange format and communication protocol between modules are defined, and data packets are transmitted and reassembled.

[0009] According to one aspect of the present application, a soft bus power terminal heterogeneous interconnection system based on an Ethernet protocol stack is provided, the system comprising:

[0010] A fragmentation module is used to dynamically calculate fragmentation parameters for fragmenting data packets according to the differences of different protocol stacks in heterogeneous network interconnection, and define a common data packet format and reassembly rules to form a unified reassembly protocol, so as to deploy the unified reassembly protocol in the network layer, the transport layer and the application layer, so that the network layer, the transport layer and the application layer process the data packet based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field;

[0011] The adaptive sharding module is used to generate an adaptive sharding scheme according to input parameters based on a dynamic programming algorithm during the sharding reorganization process;

[0012] A unified reassembly module, used to generate a unified reassembly protocol according to the adaptive fragmentation scheme, and a cross-layer optimization module, used to optimize the encapsulation and decapsulation algorithms of data packets;

[0013] An integrated generation module, used to form an integrated fragmentation reassembly framework according to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm; wherein the input parameters include the maximum transmission unit and data packet format information of each protocol layer;

[0014] The intelligent decision-making module is used in the integrated sharding and reorganization solution. Based on the intelligent decision-making mechanism, it generates the optimal sharding strategy and reorganization mechanism according to the predicted network status data, business type data and service quality requirement data;

[0015] The simulation module is used to establish a simulation platform for sharding and reassembly, simulate different network scenarios and service flows, and evaluate the feasibility and effectiveness of sharding and reassembly solutions.

[0016] The optimization module is used to combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism to verify and optimize the effect of the intelligent decision-making algorithm;

[0017] The encrypted transmission module is used to adopt encryption and authentication mechanisms in the process of fragmentation and reorganization, adopt modular design and standardized interfaces, define the data exchange format and communication protocol between modules, and transmit and reorganize data packets.

[0018] The technical solution of the embodiment of the present application dynamically calculates the fragmentation parameters for fragmenting data packets according to the differences of different protocol stacks in heterogeneous network interconnection, and defines a common data packet format and reassembly rules to form a unified reassembly protocol, so that the unified reassembly protocol is deployed in the network layer, the transport layer and the application layer, so that the network layer, the transport layer and the application layer process the data packet based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field; in the fragmentation reassembly process, based on a dynamic programming algorithm, an adaptive fragmentation scheme is generated according to the input parameters, and a unified reassembly protocol is generated according to the adaptive fragmentation scheme, the encapsulation and decapsulation algorithm of the data packet is optimized, and the data packet is processed according to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet. Encapsulation and decapsulation algorithms form an integrated fragmentation and reassembly framework; wherein, the input parameters include the maximum transmission unit and data packet format information of each protocol layer; in the integrated fragmentation and reassembly scheme, based on the intelligent decision-making mechanism, according to the predicted network status data, business type data and service quality requirement data, the optimal fragmentation strategy and reassembly mechanism are generated, and a simulation platform for fragmentation and reassembly is established to simulate different network scenarios and business traffic, evaluate the feasibility and effectiveness of the fragmentation and reassembly scheme, combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism, and verify and optimize the effect of the intelligent decision-making algorithm; in the process of fragmentation and reassembly, encryption and authentication mechanisms are adopted, modular design and standardized interfaces are adopted, the data exchange format and communication protocol between modules are defined, and data packets are transmitted and reassembled. The above scheme significantly improves the efficiency and reliability of data transmission in heterogeneous networks, effectively solves the compatibility issues between different network protocol stacks, reduces the number of fragments during cross-network transmission, ensures the timing and integrity of data packets, and reduces processing overhead and network delay.

[0019] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 A flowchart of a method for heterogeneous interconnection of soft bus power terminals based on an Ethernet protocol stack provided in an embodiment of the present application;

[0022] Figure 2 A schematic diagram of the structure of a soft bus power terminal heterogeneous interconnection system based on an Ethernet protocol stack provided in an embodiment of the present application;

[0023] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0025] It should be noted that the terms "first", "second", "third", "fourth", "actual", "preset", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0026] Figure 1 A flowchart of a method for heterogeneous interconnection of soft bus power terminals based on Ethernet protocol stack is provided in an embodiment of the present application. The embodiment of the present application can be applied to the case of determining the operating status of power transmission equipment based on heterogeneous interconnection of soft bus power terminals based on Ethernet protocol stack. The method can be executed by a heterogeneous interconnection device of soft bus power terminals based on Ethernet protocol stack. The heterogeneous interconnection device of soft bus power terminals based on Ethernet protocol stack can be implemented in the form of hardware and / or software. The heterogeneous interconnection device of soft bus power terminals based on Ethernet protocol stack can be configured in electronic equipment. Figure 1 As shown, the method includes:

[0027] S110. Dynamically calculate fragmentation parameters for fragmenting data packets according to the differences in different protocol stacks in heterogeneous network interconnection, and define a common data packet format and reassembly rules to form a unified reassembly protocol, so that the unified reassembly protocol can be deployed in the network layer, transport layer and application layer, so that the network layer, the transport layer and the application layer can process the data packets based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field.

[0028] Exemplarily, the optimal fragmentation size is dynamically calculated based on the MTU of different protocol stacks in the heterogeneous network to form an adaptive fragmentation parameter, which is then used to fragment the data packet, and sequence numbers and check information are added to form fragmented data packets in a unified format, which are transmitted to the receiving end through the network.

[0029] The receiving end reassembles the fragments according to the sequence number and verification information. If the verification fails, it requests retransmission. At the same time, it uses the PID control algorithm and decision tree algorithm to adjust the fragmentation parameters in real time according to the packet loss rate and delay of the network status. The decision tree is trained through historical data to predict the optimal fragmentation parameters, and new data packets are continuously fragmented, transmitted and reassembled.

[0030] S120. During the fragmentation reassembly process, based on a dynamic programming algorithm, an adaptive fragmentation scheme is generated according to input parameters, and a unified reassembly protocol is generated according to the adaptive fragmentation scheme, and the encapsulation and decapsulation algorithm of the data packet is optimized. According to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm, an integrated fragmentation reassembly framework is formed; wherein the input parameters include the maximum transmission unit of each protocol layer and the data packet format information.

[0031] Exemplarily, the MTU and data packet format information of each protocol layer in the heterogeneous network is obtained as input to the adaptive slicing module, and the adaptive slicing module calculates the optimal slicing size using a dynamic programming algorithm to generate an adaptive slicing solution.

[0032] The unified reassembly module designs a unified data packet format according to the scheme to form a reassembly protocol. At the same time, the cross-layer optimization module uses a heuristic algorithm to optimize the packet encapsulation and decapsulation of the network layer, transport layer and application layer.

[0033] The above two modules are integrated to form an integrated shard reorganization framework, which dynamically adjusts the shard size, unifies the reorganization format and optimizes cross-layers. The reinforcement learning algorithm is used to optimize the shard reorganization parameters in real time according to the network conditions.

[0034] S130. In the integrated sharding reorganization scheme, based on the intelligent decision-making mechanism, the optimal sharding strategy and reorganization mechanism are generated according to the predicted network status data, business type data and service quality requirement data, and a sharding reorganization simulation platform is established to simulate different network scenarios and business traffic, evaluate the feasibility and effectiveness of the sharding reorganization scheme, combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism, and verify and optimize the effect of the intelligent decision-making algorithm.

[0035] Exemplarily, after obtaining the input parameters of network status, business type and service quality requirements, the data is preprocessed, including cleaning and normalization, and then input into the intelligent decision module. The intelligent decision module uses a decision tree algorithm to analyze the data and determine the optimal sharding strategy and reorganization mechanism.

[0036] The historical data is trained by the support vector machine algorithm to establish a prediction model for network status and business needs. Based on the prediction model, the network status, business type and service quality requirements for the next time period are obtained, and the intelligent decision-making module generates sharding parameters based on the network status, business type and service quality requirements for the next time period.

[0037] Based on the predicted sharding parameters, the sharding and reorganization parameters are dynamically adjusted to obtain the sharding strategy and reorganization mechanism, which are input into the sharding and reorganization module for adaptive processing. For heterogeneous networks, the network topology is analyzed to extract key parameters.

[0038] A simulation model was built in NS-3 to simulate data transmission and set evaluation indicators. The comprehensive evaluation indicators were calculated by weighted summation, and the fragmentation and reassembly parameters were optimized using the particle swarm optimization algorithm. Wireshark was used to monitor the actual network performance, compare it with the simulation results, calculate the deviation, and feed it back to the intelligent decision-making and prediction module for iterative optimization.

[0039] S140. In the process of fragmentation and reassembly, encryption and authentication mechanisms are adopted, modular design and standardized interfaces are adopted, data exchange formats and communication protocols between modules are defined, and data packets are transmitted and reassembled.

[0040] The data packet will be divided into multiple fragments with unique identifiers according to the fragment size and the number of fragments, and encrypted by an encryption algorithm to ensure confidentiality; the integrity of the fragments and the credibility of the source are verified through an authentication mechanism; the encrypted fragments are transmitted to the target node through a standardized interface, carrying metadata information; after the target node receives it, the validity of the fragments is verified through decryption and authentication, and the fragments are reassembled according to the identifiers and metadata to restore the original data; and a machine learning algorithm is used to perform an integrity check on the reassembled data.

[0041] The technical solution of the embodiment of the present application dynamically calculates the fragmentation parameters for fragmenting data packets according to the differences of different protocol stacks in heterogeneous network interconnection, and defines a common data packet format and reassembly rules to form a unified reassembly protocol, so that the unified reassembly protocol is deployed in the network layer, the transport layer and the application layer, so that the network layer, the transport layer and the application layer process the data packet based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field; in the fragmentation reassembly process, based on a dynamic programming algorithm, an adaptive fragmentation scheme is generated according to the input parameters, and a unified reassembly protocol is generated according to the adaptive fragmentation scheme, the encapsulation and decapsulation algorithm of the data packet is optimized, and the data packet is processed according to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet. Encapsulation and decapsulation algorithms form an integrated fragmentation and reassembly framework; wherein, the input parameters include the maximum transmission unit and data packet format information of each protocol layer; in the integrated fragmentation and reassembly scheme, based on the intelligent decision-making mechanism, according to the predicted network status data, business type data and service quality requirement data, the optimal fragmentation strategy and reassembly mechanism are generated, and a simulation platform for fragmentation and reassembly is established to simulate different network scenarios and business traffic, evaluate the feasibility and effectiveness of the fragmentation and reassembly scheme, combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism, and verify and optimize the effect of the intelligent decision-making algorithm; in the process of fragmentation and reassembly, encryption and authentication mechanisms are adopted, modular design and standardized interfaces are adopted, the data exchange format and communication protocol between modules are defined, and data packets are transmitted and reassembled. The above scheme significantly improves the efficiency and reliability of data transmission in heterogeneous networks, effectively solves the compatibility issues between different network protocol stacks, reduces the number of fragments during cross-network transmission, ensures the timing and integrity of data packets, and reduces processing overhead and network delay.

[0042] As a non-limiting implementation method, the fragmentation parameters for fragmenting the data packet are dynamically calculated according to the differences of different protocol stacks in the heterogeneous network interconnection, including:

[0043] Dynamically calculate the optimal fragmentation size according to the MTU of different protocol stacks in the heterogeneous network to form fragmentation parameters, fragment the data packet based on the fragmentation parameters, and add sequence numbers and check information to form fragmented data packets in a unified format, which are transmitted to the receiving end through the network;

[0044] The receiving end reassembles the fragments according to the sequence number and verification information, and requests retransmission if the verification fails;

[0045] According to the packet loss rate, delay status and transmission quality of the network status in the subsequent time, the packet loss rate, delay status and transmission quality of the network status in the subsequent time are predicted; based on the decision tree algorithm, the optimal fragmentation parameters are predicted; and based on the optimal fragmentation parameters, new data packets are fragmented, transmitted and reassembled.

[0046] As a non-limiting implementation method, the packet loss rate, delay status and transmission quality of the network state in the subsequent time are predicted according to the packet loss rate, delay status and transmission quality of the network state, the optimal fragmentation parameters are predicted based on the packet loss rate, delay status and transmission quality of the network state in the subsequent time of the decision tree algorithm, and the new data packet is fragmented, transmitted and reassembled based on the optimal fragmentation parameters, including:

[0047] According to the network's packet loss rate, delay status and transmission quality, the PID control algorithm is used to dynamically adjust the adaptive fragmentation parameters;

[0048] Using the decision tree algorithm, a classification model is constructed based on historical transmission data and network parameters including bandwidth, delay, and packet loss rate to predict the optimal adaptive sharding parameters under the current network conditions;

[0049] By learning and training historical data, a decision tree is generated, and classification prediction is performed based on current network parameters to obtain the optimal sharding parameter value;

[0050] Use the predicted adaptive fragmentation parameters to fragment new data packets and repeat the above-mentioned transmission, reorganization, and dynamic adjustment process to continuously optimize transmission efficiency and quality and adapt to the dynamic changes of heterogeneous networks.

[0051] Exemplarily, the above scheme is further explained. The maximum transmission unit of different protocol stacks in heterogeneous networks dynamically calculates the optimal fragmentation size to obtain adaptive fragmentation parameters; the adaptive fragmentation parameters are used to fragment the data packet, and sequence numbers and check information are added to the fragmented data packets to form fragmented data packets in a unified format; the fragmented data packets in a unified format are transmitted through the network; at the receiving end, they are reassembled according to the sequence numbers and check information; if the received fragmented data packets pass the check, they are assembled in sequence to obtain a complete original data packet; if they fail the check, the sending end is requested to retransmit the fragmented data packet until it passes the check; according to the network's packet loss rate, delay status and transmission quality, a PID control algorithm is used to dynamically adjust The adaptive slicing parameters are adjusted to optimize the slicing size and transmission efficiency. By introducing the three parameters of proportion, integration and differentiation, the optimal slicing size is calculated in real time according to the changes in the network status. The decision tree algorithm is used to build a classification model based on historical transmission data and network parameters including bandwidth, delay, and packet loss rate to predict the optimal adaptive slicing parameters under the current network conditions and improve transmission performance. Through learning and training of historical data, a decision tree is generated, and classification prediction is performed according to the current network parameters to obtain the optimal slicing parameter value. The predicted adaptive slicing parameters are used to slice new data packets, and the above-mentioned transmission, reorganization, and dynamic adjustment process are repeated to continuously optimize transmission efficiency and quality and adapt to the dynamic changes of heterogeneous networks.

[0052] Specifically, in order to obtain the maximum transmission unit of the network link, the path MTU discovery technology is used. By sending a series of probe data packets of different sizes, including starting from 68 bytes and increasing by a power of 2 until 1500 bytes, the MTU value in the returned ICMP error message is recorded, and the minimum value is taken as the path MTU.

[0053] When calculating the adaptive fragment size based on the path MTU, it is necessary to consider the 20-byte overhead of the IP header and the transport layer protocol header overhead, including 20 bytes for TCP and 8 bytes for UDP. The preset application layer data length is 1024 bytes, so the actual fragment size is 1492-20-20=1452 bytes. To improve transmission efficiency, the fragment size is appropriately adjusted to 1024 bytes, and each data packet is split into two fragments.

[0054] In the fragmented IP header, the identification field is used to identify the original data packet, the sequence number field indicates the position of the current fragment in the original data packet, the DF flag is set to 0 to allow fragmentation, the MF flag is set to 1 except for the last fragment, and the offset field indicates the position of the fragment in the original datagram in units of 8 bytes.

[0055] The fragments are transmitted through the network layer IP protocol. The maximum segment size is specified in the MSS option of the transport layer TCP protocol as the path MTU minus the IP and TCP header length to ensure that the transport layer segmentation is consistent with the network layer fragmentation.

[0056] The receiving end caches and reassembles the fragments according to the identification, sequence number, and offset fields in the IP header until the last fragment with the MF flag bit set to 0 is received.

[0057] After reorganization, the receiving end calculates the CRC32 checksum of the data part and compares it with the checksum calculated by the sending end. If they are consistent, the data is submitted to the upper layer protocol for processing; otherwise, it is discarded and the sending end is notified to retransmit.

[0058] In summary, adaptive fragmentation technology detects the path MTU, dynamically adjusts the fragment size, and uses the IP fragmentation and reassembly mechanism to ensure efficient and reliable data transmission in network environments with different MTUs.

[0059] Specifically, it also includes the calculation process of the path MTU discovery technology:

[0060] Send a sequence of probe data packets: 68 bytes, 136 bytes, 272 bytes, 544 bytes, 1092 bytes, 2184 bytes, etc. until 1500 bytes. Record the MTU value in the ICMP error message returned by each probe packet. Calculation process example: If the MTU indicated by the received ICMP error message is 1492 bytes when the probe packet size is 544 bytes, and the MTU indicated by the received ICMP error message is 2984 bytes when the probe packet size is 1092 bytes, then take the minimum of these values ​​as the path MTU;

[0061] Calculation result: Path MTU = min (1492, 2984) = 1492 bytes (this is the minimum MTU value received by default).

[0062] The process of calculating the adaptive shard size:

[0063] Consider IP header overhead: 20 bytes;

[0064] Consider the transport layer protocol header overhead: 20 bytes for TCP and 8 bytes for UDP;

[0065] Calculation process: Path MTU - IP header overhead - transport layer header overhead = 1492 - 20 - 20 = 1452 bytes (if TCP is used);

[0066] If UDP is used, the calculation process is: 1492-20-8=1464 bytes;

[0067] The calculation process for adjusting the shard size:

[0068] To improve transmission efficiency, the default fragment size is 1024 bytes;

[0069] Calculate the number of fragments: Application layer data length / fragment size = 1464 / 1024 rounded up to 2, so each data packet can be split into two fragments. If the calculated application layer data length is 1024 bytes, then 1024 / 1024 = 1, and no fragmentation is required.

[0070] The calculation process of the fragmentation-related fields in the IP header:

[0071] Identification field: used to identify the original data packet, usually a randomly or sequentially generated value;

[0072] Sequence number field: indicates the position of the current fragment in the original data packet, in bytes;

[0073] Offset field: Indicates the position of the fragment in the original datagram in units of 8 bytes;

[0074] Example of calculating the offset: If the offset of the first shard is 0, the offset of the second shard is the size of the first shard divided by 8, that is, (1024 / 8)=128.

[0075] The calculation process of the transport layer MSS option is:

[0076] Path MTU-IP header overhead-TCP header overhead = 1492-20-20 = 1452 bytes,

[0077] Therefore, the MSS option is set to 1452 bytes.

[0078] The calculation process of the receiving end reorganization and verification is as follows:

[0079] Cache the fragments according to the identification, sequence number, and offset fields in the IP header;

[0080] Calculate the CRC32 checksum of the reassembled data portion;

[0081] Comparison process: If the CRC32 checksum calculated by the receiving end is consistent with the checksum sent by the sending end, the data is submitted to the upper layer protocol for processing; otherwise, the data is discarded and the sending end is notified to retransmit;

[0082] Through these calculation processes, adaptive sharding technology can ensure efficient and reliable data transmission in network environments with different MTUs.

[0083] As a non-limiting implementation method, based on a dynamic programming algorithm, an adaptive fragmentation scheme is generated according to input parameters, and a unified reassembly protocol is generated according to the adaptive fragmentation scheme, and the encapsulation and decapsulation algorithm of the data packet is optimized. According to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm, an integrated fragmentation reassembly framework is formed, including:

[0084] Obtain the maximum transmission unit and data packet format information of each protocol layer in the heterogeneous network as input information, calculate the optimal fragmentation size based on the dynamic programming algorithm, and generate an adaptive fragmentation solution;

[0085] Design a unified data packet format according to the adaptive fragmentation scheme to form a unified reassembly protocol;

[0086] Use heuristic algorithms to optimize packet encapsulation and decapsulation at the network layer, transport layer, and application layer;

[0087] According to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm, an integrated fragmentation reassembly framework is formed.

[0088] As a non-limiting implementation method, the optimal shard size is calculated based on a dynamic programming algorithm to generate an adaptive sharding solution, including:

[0089] Construct a state transition equation based on the following formula:

[0090] dp[i] = min(dp[i], dp[j] + 1), where 0 ≤ j < i, j + s ≤ maximum transmission unit, and s represents the fragment size; assume the data packet length is 1024 bytes, and the state dp[i] represents the minimum number of fragments for dividing the first i bytes into fragments;

[0091] Obtain the minimum number of fragments for dividing the first i bytes into fragments and the optimal fragment size according to the algorithm, and generate an adaptive fragmentation scheme.

[0092] Exemplarily, further expand and explain the above scheme. Obtain the maximum transmission unit and data packet format information of each protocol layer in the heterogeneous network as input parameters for the adaptive fragmentation module; the adaptive fragmentation module uses the dynamic programming algorithm according to the input parameters to calculate the optimal fragment size and generate an adaptive fragmentation scheme; the unified recombination module designs a unified data packet format according to the adaptive fragmentation scheme, including fields such as fragment sequence number and checksum, to form a unified recombination protocol; the cross-layer optimization module uses a heuristic algorithm for the data packets of the network layer, transport layer, and application layer to optimize the encapsulation and decapsulation processes, reducing redundant fields and computational overhead; integrate the adaptive fragmentation scheme, unified recombination protocol, and cross-layer optimization strategy into an integrated fragmentation and recombination framework to form a complete solution; use this integrated solution to fragment and recombine the data transmitted in the heterogeneous network, and improve the transmission efficiency and reliability by dynamically adjusting the fragment size, unified recombination format, and cross-layer optimization of encapsulation; according to the changes in the network conditions, use the reinforcement learning algorithm to optimize the parameters of the integrated fragmentation and recombination solution in real time, including the fragment size and recombination delay, to achieve adaptive coordination.

[0093] Specifically, to obtain the maximum transmission unit and data packet format information of each protocol layer in the heterogeneous network, a combination of active probing and passive listening is used, including actively sending ICMP probe messages of different lengths and extracting the MTU value according to the returned error information. Passively listen to the data frames on the network interface and count the data link layer protocol type and data packet length distribution. It is preset that after detection, the MTU of Ethernet is 1500 bytes, the MTU of the MPLS network is 4470 bytes, and the MTU of the IPv4 network is 1480 bytes. The data packet format information is obtained by parsing the protocol specifications, including the Ethernet frame format, MPLS label format, and IP datagram format. Input the detected MTU value and the parsed data packet format parameters into the adaptive fragmentation module, and use the dynamic programming algorithm to calculate the optimal fragment size.

[0094] This algorithm constructs a state transition equation to solve the fragment size corresponding to the minimum number of fragments;

[0095] Assume the data packet length is 1024 bytes, and the state dp[i] represents the minimum number of fragments when dividing the first i bytes into several fragments;

[0096] The state transition equation is: dp[i] = min(dp[i], dp[j] + 1), where 0 ≤ j < i, j + s ≤ MTU, and s represents the fragment size;

[0097] According to the algorithm, the optimal fragment size is 1024 bytes, and an adaptive fragmentation scheme is generated;

[0098] The unified recombination module designs a unified data packet format including fields such as fragment sequence number, data length, and checksum according to the adaptive fragmentation scheme to ensure the parsing and recombination of data packets between different protocol networks.

[0099] The cross-layer optimization module adopts a heuristic algorithm to optimize the encapsulation and decapsulation processes, including: for IPv4 data packets, the checksum field can be omitted to reduce computational overhead; for MPLS data packets, the label stack can be merged to reduce the header length; through cross-layer mapping, the conversion of data packet formats between different protocol layers is achieved.

[0100] Finally, the adaptive fragmentation scheme, unified recombination protocol, and cross-layer optimization strategy are integrated into an integrated fragmentation and recombination framework. This framework fragments the application layer data transmitted in heterogeneous networks, then encapsulates it in a unified format, and uses cross-layer optimization for transmission. The receiving end decapsulates and recombines it in a unified format to restore the original data.

[0101] To further improve the transmission efficiency, a reinforcement learning algorithm is used to adjust the fragment size and recombination delay parameters in real time; by defining the state space, action space, and reward function, the Q-learning algorithm is used to continuously try and learn to obtain the optimal parameter combination, including defining the fragment size and recombination delay as actions, and the throughput and delay as rewards, continuously trying different action combinations, collecting feedback rewards, updating the Q-value table, and finally obtaining the optimal strategy; through adaptive collaborative optimization, the integrated fragmentation and recombination scheme can be dynamically adjusted according to changes in network conditions to ensure efficient transmission in a heterogeneous networking environment.

[0102] Specifically, it also includes the calculation process of actively detecting the MTU value:

[0103] Send ICMP probe messages of different lengths, for example, starting from 64 bytes and increasing by 64 bytes each time until the message length exceeds 1500 bytes; record the MTU value in the ICMP error information returned after each send.

[0104] Example of the calculation process: If the ICMP error information indicates that the MTU is 1500 bytes when sending a 128-byte probe message, then stop the detection.

[0105] Passive monitoring of the calculation process of the data link layer protocol type and data packet length distribution: Count the protocol type and length of the statistical frame, including capturing the data frame through the packet capture tool; calculate the average and standard deviation of the packet length of each protocol type.

[0106] The calculation process of the adaptive sharding module:

[0107] Input parameters: Ethernet MTU = 1500 bytes, MPLS network MTU = 4470 bytes, IPv4 network MTU = 1480 bytes. Use dynamic programming algorithm to calculate the optimal fragment size.

[0108] Calculation process of state transfer equation:

[0109] Initialize the dp array, dp[0] = 0 (indicates that no fragmentation is required when there is no data);

[0110] Iterate over each possible fragment size s, for each i (1 to packet length 1024 bytes):

[0111] dp[i] = infinity (indicates the initial state);

[0112] For each j (0 to i-1):

[0113] If j+s<=MTU, update dp[i]=min(dp[i],dp[j]+1);

[0114] Calculation example: Preset MTU = 1500 bytes, data packet length = 1024 bytes, then dp

[1024] represents the minimum number of fragments into which 1024 bytes are divided.

[0115] The calculation process of the cross-layer optimization module:

[0116] For IPv4 packets, the checksum field is omitted to reduce the computational overhead. The time complexity of calculating the checksum in the original IPv4 header is O(n), and the complexity is reduced to O(1) after omitting it.

[0117] For MPLS packets, the calculation process of the combined label stack is as follows:

[0118] Assume that the original label stack has m labels, each of which is 4 bytes long. After merging, the number of labels may be reduced to n. The saved header space is (mn)*4 bytes.

[0119] The reinforcement learning algorithm adjusts the fragment size and reorganizes the calculation process of the delay parameters in real time:

[0120] Define the state space: current network conditions, including bandwidth, latency, and packet loss rate;

[0121] Define the action space: possible values ​​of fragment size and reassembly delay;

[0122] Define a reward function: a performance indicator based on throughput and latency, for example, reward = throughput / (1+latency).

[0123] The calculation process of the Q-learning algorithm is:

[0124] Initialize the Q value table;

[0125] At each time step, choose an action, perform the action, and observe the reward and next state;

[0126] Update Q value: Q(s,a)=Q(s,a)+α(R+γmax_a'Q(s',a')-Q(s,a)), where α is the learning rate and γ is the discount factor;

[0127] Repeat the above process until the Q value converges and the optimal strategy is obtained;

[0128] Through these calculation processes, the integrated sharding reorganization scheme can be dynamically adjusted according to changes in network conditions to ensure efficient transmission in a heterogeneous networking environment;

[0129] As mentioned above, in the calculation process of the Q-learning algorithm, the characters involved and their physical meanings are as follows: Q(s,a): This is the Q function, which represents the expected reward (or value) of taking action a in state s; s: The current state, which is a description of the network status, such as the current bandwidth, delay, and packet loss rate; a: The action taken, in this case, refers to the selection of a specific shard size and reassembly delay; R: The immediate reward obtained from the environment after performing action a, which reflects the quality of the action, such as a function of throughput and delay; s': The next state reached after performing action a; a': Any action that may be taken in state s'; max_a'Q(s',a'): The maximum value of the Q value among all possible actions in state s', representing the maximum expected reward that can be obtained in state s'; α: Learning rate, which determines the degree of influence of new information on the old Q value, 0<α≤1; γ: Discount factor, which determines the importance of future rewards relative to immediate rewards, 0≤γ≤1.

[0130] The update formula of the Q-learning algorithm is:

[0131] Q(s,a)=Q(s,a)+α(R+γmax_a'Q(s',a')-Q(s,a));

[0132] The physical meaning of this formula is:

[0133] Q(s,a): What we need to update is the Q value of taking action a in the current state s;

[0134] α(R+γmax_a'Q(s',a')-Q(s,a)): This is the amount of Q value update, which consists of three parts.

[0135] R: is the actual instant reward; γmax_a'Q(s',a'): is the maximum expected return considering future rewards, that is, the maximum return that can be obtained by taking the optimal strategy starting from the next state s'; Q(s,a): is the old Q value of taking action a in the current state s. Combining these three parts, we get the update amount of the Q value, which takes into account the combined impact of the instant reward and future rewards; α: this factor controls the amplitude of the update, so that the Q value will not change drastically due to a single experience, but will learn gradually.

[0136] As a non-restrictive implementation method, based on the intelligent decision-making mechanism, according to the predicted network status data, business type data and service quality requirement data, the optimal sharding strategy and reorganization mechanism are generated, and a simulation platform for sharding and reorganization is established to simulate different network scenarios and business traffic, evaluate the feasibility and effectiveness of the sharding and reorganization plan, and combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism to verify and optimize the effect of the intelligent decision-making algorithm, including:

[0137] Obtain network status data, business type data, and service quality requirement data, and perform prediction based on the prediction model to obtain network status data, business type data, and service quality requirement data in the subsequent period;

[0138] Use the decision tree algorithm to determine the optimal sharding strategy and reorganization mechanism based on the network status data, business type data, and service quality requirement data in the subsequent period;

[0139] Based on the prediction results, the fragmentation and reassembly parameters are dynamically adjusted and input into the fragmentation and reassembly module for adaptive processing;

[0140] For heterogeneous networks, analyze network topology to extract key parameters, build simulation models in NS-3, simulate data transmission, and set evaluation indicators;

[0141] The comprehensive evaluation index is calculated by weighted summation, and the fragmentation and reorganization parameters are optimized by particle swarm optimization algorithm;

[0142] Use Wireshark to monitor the actual network performance, compare the calculated deviation with the simulation results, and feed the deviation back to the decision tree algorithm and prediction algorithm for iterative optimization.

[0143] As a non-limiting implementation method, Wireshark is used to monitor the actual network performance, compare the calculated deviation with the simulation results, and feed the deviation back to the decision tree algorithm and prediction algorithm for iterative optimization, including:

[0144] For heterogeneous network interconnection scenarios, obtain the network topology and protocol type information of each subnet, analyze the network topology, and extract key feature parameters;

[0145] A heterogeneous network interconnection model was constructed in the network simulation platform NS-3, and the extracted key characteristic parameters were input into the simulation model to simulate the data transmission process in different network scenarios.

[0146] In the simulation platform, the evaluation indicators of fragment reassembly are set based on the service priority and service quality level for different service traffic and service quality requirements; the evaluation indicators include latency, throughput, and packet loss rate;

[0147] The weighted summation method is used to calculate the weight of each evaluation index and obtain the comprehensive evaluation index;

[0148] Through the particle swarm optimization algorithm, the shard size and reorganization timing parameters are optimized with comprehensive evaluation indicators as the optimization target, and the feedback is given to the intelligent decision-making process and the reorganization process;

[0149] Use network performance monitoring tools to obtain performance monitoring data in the actual network, compare and analyze the simulation evaluation results and performance monitoring data, and calculate the evaluation index deviation;

[0150] The decision tree algorithm and prediction algorithm are iteratively optimized based on the evaluation indicator deviation through parameter adjustment and model retraining.

[0151] As a non-limiting implementation, the method further includes:

[0152] The fragment size, reassembly delay, and cache size parameters are used as chromosomes, and the throughput, delay, and reassembly success performance indicators are used as fitness functions. Genetic algorithms are used to adaptively tune the parameters of the integrated fragment reassembly framework.

[0153] Through selection, crossover, and mutation genetic operations, the parameter combination is continuously optimized to obtain the optimal sharding and reorganization configuration.

[0154] Furthermore, the above scheme is further explained. After obtaining the input parameters of network status, service type and service quality requirement, the obtained data is pre-processed, including data cleaning and normalization, and the processed data is input into the intelligent decision module;

[0155] The intelligent decision-making module uses a decision tree algorithm to analyze input data according to preset decision rules and determine the optimal sharding strategy and reorganization mechanism;

[0156] The historical network status and service demand data are trained through the support vector machine algorithm to obtain the prediction model of network status and service demand;

[0157] Input current input parameters into the prediction model to predict changes in network conditions and business demand trends in the future;

[0158] According to the prediction results, the fragmentation size and reorganization strategy parameters are dynamically adjusted, and the adjusted parameters are input into the fragmentation and reorganization module to perform adaptive fragmentation and reorganization on the data.

[0159] For heterogeneous network interconnection scenarios, after obtaining the network topology and protocol type information of each subnet, the network topology is analyzed to extract key feature parameters;

[0160] A heterogeneous network interconnection model was constructed in the network simulation platform NS-3, and the extracted key characteristic parameters were input into the simulation model to simulate the data transmission process in different network scenarios.

[0161] In the simulation platform, for different service flows and service quality requirements, based on service priority and service quality level, the evaluation indicators for fragment reassembly are set, including latency, throughput, and packet loss rate;

[0162] The weighted summation method is used to calculate the weight of each evaluation index and obtain a comprehensive evaluation index to evaluate the feasibility and effectiveness of the sharding reorganization plan;

[0163] Through the particle swarm optimization algorithm, the shard size and reorganization timing parameters are optimized with the comprehensive evaluation index as the optimization target, the optimized parameter settings are obtained, and then fed back to the intelligent decision-making module and the shard reorganization module;

[0164] Use the network performance monitoring tool Wireshark to obtain performance monitoring data in the actual network, compare and analyze the simulation evaluation results and performance monitoring data, and calculate the evaluation index deviation;

[0165] The indicator deviation is input into the intelligent decision-making module and prediction module, and the decision tree algorithm and support vector machine prediction model are iteratively optimized. Through parameter adjustment and model retraining, the performance and adaptability of sharding reorganization are continuously improved.

[0166] Specifically, the above process is further expanded to obtain the input parameters of network status, service type and service quality requirements, and network monitoring tools and traffic analysis systems are used, including using Wireshark to capture and parse data packets in heterogeneous networks, and extract network delay, packet loss rate, and bandwidth utilization performance indicators. NetFlow is used to collect and analyze network traffic to identify the traffic distribution and proportion of different service types. Through QoS policy configuration, the service quality requirements of different voice, video, and data services are obtained, including delay, jitter, and packet loss rate thresholds. The data obtained from network monitoring and traffic analysis are input into the intelligent decision-making mechanism, and the decision tree algorithm C4.5 is used to train the data.

[0167] The C4.5 algorithm calculates the information gain ratio of each attribute, selects the optimal partitioning attribute, and recursively constructs a decision tree, including dividing the fragmentation strategy into three categories: fixed fragmentation, dynamic fragmentation, and adaptive fragmentation according to network delay and packet loss rate parameters;

[0168] According to the business type and service quality requirements, the reorganization mechanism is divided into three categories: immediate reorganization, delayed reorganization and priority reorganization;

[0169] For the selected fragmentation strategy and reorganization mechanism, the SVM algorithm is used to optimize the parameters;

[0170] The network status, service type and service quality requirements in the historical data are used as feature vectors, and the corresponding shard size and reorganization strategy are used as labels to train the SVM model;

[0171] Through cross-validation and grid search, the kernel function type, penalty factor, and kernel function parameters of SVM are optimized to obtain the optimal sharding size and reorganization strategy parameters.

[0172] On the sharding and reassembly simulation platform, the data transmission process is simulated for different network scenarios and service flows, including using OPNET to establish a heterogeneous network simulation model and configure network topology, link bandwidth, and transmission delay parameters;

[0173] Generate actual business traffic, including HTTP, FTP, VoIP, and set traffic size, burst characteristics, and priority parameters;

[0174] Run the integrated sharding and reassembly framework on the simulation platform to record the end-to-end latency, throughput, and reassembly success performance indicators;

[0175] Use BP neural network algorithm to train the intelligent decision-making mechanism;

[0176] The input samples include network conditions, service types, and service quality requirement characteristics, and the output samples are the selection of fragmentation strategy and reassembly mechanism;

[0177] By setting the number of hidden layer nodes, activation function, and learning rate parameters, the convergence speed and generalization ability of the BP neural network are optimized;

[0178] Continuous iterative training enables the intelligent decision-making mechanism to adaptively select the optimal sharding strategy and reorganization mechanism according to changes in network conditions;

[0179] Finally, a genetic algorithm is used to adaptively tune the parameters of the sharding reorganization framework;

[0180] The fragment size, reassembly delay, and cache size parameters are used as chromosomes, and the throughput, delay, and reassembly success performance indicators are used as fitness functions;

[0181] Through selection, crossover, and mutation genetic operations, the parameter combination is continuously optimized to obtain the optimal fragmentation and recombination configuration, including setting the population size to 50, the crossover probability to 0.8, the mutation probability to 0.1, and the number of iterations to 100;

[0182] After genetic algorithm optimization, the optimal fragment size is 1024 bytes, the reassembly delay is 50 milliseconds, and the cache size is 10MB, which increases the throughput of heterogeneous networks by 20%, reduces the delay by 15%, and achieves a reassembly success rate of more than 99%.

[0183] Through intelligent decision-making, simulation optimization and adaptive parameter tuning, the integrated sharding reorganization solution can dynamically select the optimal strategy and parameters according to different network scenarios and business needs, and achieve efficient data transmission in heterogeneous networks.

[0184] Specifically, network monitoring and traffic analysis data are input into the intelligent decision-making mechanism:

[0185] Extraction of network delay: measure the round-trip time of the data packet from the source to the destination, and calculate the average value, including that if the measured round-trip time of the data packet is T, the network delay D = T / 2;

[0186] Calculation of packet loss rate: Assume the total number of sent packets is N, the number of lost packets is L, then the packet loss rate P = L / N;

[0187] Calculation of bandwidth utilization: Assume that the actual bandwidth used is B_used and the total bandwidth is B_total, then the bandwidth utilization U = B_used / B_total;

[0188] C4.5 algorithm training process:

[0189] Calculation of information gain ratio: For each attribute A, calculate the information gain Gain(A) and the split information SplitInfo(A), and then calculate the information gain ratio GainRatio(A) = Gain(A) / SplitInfo(A);

[0190] SVM algorithm parameter optimization:

[0191] Cross-validation: Divide the data set into k subsets, perform k training and validation, and use a different subset as the validation set each time, and the rest as the training set;

[0192] Grid search: For each parameter combination (C, γ), where C is the penalty factor and γ is the kernel function parameter, use cross-validation to evaluate the model performance and select the parameter combination with the best performance;

[0193] Data transmission process on the simulation platform:

[0194] Network topology configuration: determine the number of nodes N, the number of links M, the bandwidth B and delay D of each link;

[0195] Business traffic generation: For each business traffic (HTTP, FTP, VoIP), set the traffic size V, burst characteristic Burst and priority Priority;

[0196] BP neural network algorithm training:

[0197] Input sample processing: converting network conditions, service types and service quality requirements into feature vector X;

[0198] Output sample processing: convert the selection results of the fragmentation strategy and reorganization mechanism into a label vector YX;

[0199] Network training: Calculate outputs through forward propagation and adjust weights through back propagation until the network error drops to an acceptable range;

[0200] Genetic algorithm parameter tuning:

[0201] Chromosome encoding: Encode the fragment size, reassembly delay, and cache size into chromosomes;

[0202] Fitness function: Define the fitness function F = f (throughput, delay, reorganization success), where f is a weighted combination function;

[0203] Genetic operations: The selection operation selects excellent individuals based on fitness, the crossover operation randomly exchanges some genes of two individuals, and the mutation operation randomly changes certain bits of individual genes;

[0204] The physical meanings of the characters in the above content are:

[0205] T: Round trip time, which refers to the total time it takes for a data packet to be sent from the sender to the receiver and back to the sender;

[0206] D: Network latency refers to the time it takes for a data packet to travel from the sender to the receiver;

[0207] N: total number of packets sent, which refers to the total number of packets that the sender attempts to send within a period of time;

[0208] L: The number of lost packets, which refers to the number of packets sent by the sender but not confirmed received by the receiver within a period of time;

[0209] P: Packet loss rate, which refers to the ratio of the number of lost data packets to the total number of sent data packets;

[0210] B_used: actual bandwidth used, which refers to the bandwidth actually used to transmit data in the network over a period of time;

[0211] B_total: total bandwidth, which refers to the maximum available bandwidth of the network link;

[0212] Gain(A): Information gain of attribute A, which refers to the reduction in information entropy after introducing attribute A into the decision tree;

[0213] SplitInfo(A): the split information of attribute A, which refers to the entropy of data set D after the introduction of attribute A;

[0214] GainRatio(A): Information gain ratio of attribute A, which is the ratio of information gain to segmentation information;

[0215] C: penalty factor in SVM, used to control the model's tolerance to classification errors;

[0216] γ: kernel function parameter in SVM, used to adjust the influence range of the kernel function;

[0217] k: The number of subsets in cross-validation, which means dividing the data set into k subsets for validation;

[0218] V: Traffic size, which refers to the data volume of the business traffic generated in the simulation;

[0219] Burst: refers to the characteristic that the business traffic suddenly increases in a short period of time;

[0220] Priority: refers to the priority of business traffic in network transmission;

[0221] X: Feature vector, which refers to the set of data features input into the neural network;

[0222] YX: label vector, which refers to the target output set in neural network training;

[0223] F: fitness function, used to evaluate the quality of each individual in the genetic algorithm;

[0224] f: Weighted combination function, used to calculate the weighted sum in the fitness function.

[0225] As a non-restrictive implementation method, in the process of fragmentation and reassembly, encryption and authentication mechanisms are adopted, modular design and standardized interfaces are adopted, data exchange formats and communication protocols between modules are defined, and data packets are transmitted and reassembled, including:

[0226] If the sharding and reorganization technology is used, the data is divided into multiple shards according to the preset shard size and number, and a unique identifier is generated for each shard;

[0227] For each shard, an encryption algorithm is used for encryption processing;

[0228] The encrypted fragments are transmitted to the receiving end through a standardized interface, carrying metadata information;

[0229] If the receiving end receives the fragment, the validity of the fragment is verified through a decryption and authentication process;

[0230] According to the unique identifier and metadata of the shards, the shards are reassembled according to the predetermined order and rules to obtain the complete original data;

[0231] A machine learning algorithm is used to verify the integrity of the reorganized data.

[0232] Further, to further explain the above scheme, if the sharding and reorganization technology is adopted, the data is divided into multiple shards according to the preset shard size and number, and a unique identifier is generated for each shard;

[0233] For each shard, an encryption algorithm is used to encrypt it to ensure the confidentiality of the shard data;

[0234] Verify the integrity and source credibility of the shards through the authentication mechanism to prevent the shards from being tampered with or forged;

[0235] The encrypted shards are transmitted to the target node through a standardized interface, carrying the necessary metadata information;

[0236] After receiving the shard, the target node verifies the validity of the shard through the decryption and authentication process;

[0237] According to the unique identifier and metadata of the shards, the shards are reassembled according to the predetermined order and rules to obtain the complete original data;

[0238] Machine learning algorithms are used to perform integrity checks on the reorganized data to ensure data accuracy and reliability.

[0239] Specifically, in the shard reorganization technology, the original data is first divided into multiple shards using a hash algorithm (SHA-256) according to the preset shard size (1MB) and number (100), and a unique hash value is generated for each shard as an identifier;

[0240] Then, for each shard, a symmetric encryption algorithm (AES-256) and a randomly generated key are used for encryption to ensure the confidentiality of the shard data;

[0241] The hash value of the shard is signed through a digital signature mechanism (RSA signature) to verify the integrity and source credibility of the shard and prevent the shard from being tampered with or forged;

[0242] The encrypted shards are transmitted to the target node through the RESTful API, along with the shards’ hash value, sequence number, and total number metadata information;

[0243] After receiving the shard, the target node confirms the validity of the shard through the process of decryption and signature verification;

[0244] According to the hash value and metadata of the shards, the merge sort algorithm is used to reorganize the shards in the original order to obtain the complete original data;

[0245] Finally, the support vector machine (SVM) algorithm is used to check the integrity of the reorganized data. The accuracy and reliability of the data are judged by calculating the similarity between the feature vector and the support vector of the data. The similarity threshold is set to 95% to ensure the consistency of the data.

[0246] Specifically, the sharding process also includes:

[0247] The default original data size is D MB, the default fragment size is F MB (1 MB), then the number of fragments N = ceil (D / F), where ceil means rounding up;

[0248] For each shard, calculate its hash value: use the SHA-256 algorithm to calculate the hash of the shard data to obtain a 256-bit hash value H;

[0249] Encryption process:

[0250] Generate key: randomly generate a 256-bit key K for AES-256 encryption;

[0251] Encrypted shards: Use the AES-256 algorithm and key K to encrypt each shard to obtain the encrypted shard data E;

[0252] Digital Signature Process:

[0253] Calculate the hash value: Use the private key to perform RSA signature on the hash value H of each shard to obtain the signature S;

[0254] Verification process: The receiver uses the public key to verify the signature S to ensure that the hash value H has not been tampered with;

[0255] Transmission process:

[0256] Send the encrypted shard E through the RESTful API, along with metadata information, including the shard’s hash value H, sequence number i (1 to N), and total number N;

[0257] Decryption and verification process:

[0258] The receiver uses the same key K to decrypt the encrypted slice E and obtain the original slice data;

[0259] Use the public key to verify the signature S of the shard to ensure the integrity and source credibility of the shard;

[0260] Reorganization process:

[0261] The receiver sorts the shards using the merge sort algorithm based on the shard’s hash value H and sequence number i.

[0262] Merge the shards in the sorted order to restore the original data D;

[0263] Integrity verification process:

[0264] Calculate the eigenvector: Calculate the eigenvector X for the reorganized data D;

[0265] Training SVM model: Use known correct data sets to train the SVM model and obtain model parameters;

[0266] Verify data: Input the feature vector X of the reorganized data D into the SVM model and calculate the prediction result;

[0267] Calculate similarity: compare the similarity between the predicted results and the actual data. If the similarity is greater than or equal to 95%, the data consistency is considered to be high and the data is accurate and reliable.

[0268] Through these calculations and digital processing, we can more clearly understand the various steps in the fragmentation reorganization technology;

[0269] The physical meanings of the characters in the above content are:

[0270] The following is a supplementary explanation of the characters used in the above content and their meanings:

[0271] D: size of the original data (in MB);

[0272] F: The preset fragment size (in MB), for example 1MB;

[0273] N: the total number of shards, calculated as N = ceil (D / F), where ceil means rounding up;

[0274] H: The hash value of the shard, calculated using the SHA-256 algorithm;

[0275] K: randomly generated key for AES-256 encryption, length is 256 bits;

[0276] E: Encrypted shard data, encrypted using the AES-256 algorithm;

[0277] S: Digital signature of the shard hash value, generated using the private key through the RSA signature algorithm;

[0278] i: the serial number of the shard, indicating the position of the shard in the original data, ranging from 1 to N;

[0279] X: Feature vector of reorganized data, used for integrity check by support vector machine (SVM) algorithm;

[0280] Similarity threshold: A set similarity threshold, such as 95%, is used to judge the accuracy and reliability of the data.

[0281] Figure 2 The present invention provides a schematic diagram of a structure of a soft bus power terminal heterogeneous interconnection system based on an Ethernet protocol stack, which can execute the soft bus power terminal heterogeneous interconnection method based on an Ethernet protocol stack provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method. Figure 2 As shown, the system comprises:

[0282] The fragmentation module 210 is used to dynamically calculate fragmentation parameters for fragmenting data packets according to the differences of different protocol stacks in heterogeneous network interconnection, and define a common data packet format and reassembly rules to form a unified reassembly protocol, so as to deploy the unified reassembly protocol in the network layer, the transport layer and the application layer, so that the network layer, the transport layer and the application layer process the data packets based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field;

[0283] The adaptive sharding module 220 is used to generate an adaptive sharding scheme according to input parameters based on a dynamic programming algorithm during the sharding reorganization process;

[0284] A unified reassembly module 230, for generating a unified reassembly protocol according to the adaptive fragmentation scheme, and a cross-layer optimization module for optimizing the encapsulation and decapsulation algorithms of data packets;

[0285] An integrated generation module 240 is used to form an integrated fragmentation reassembly framework according to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm; wherein the input parameters include the maximum transmission unit of each protocol layer and the data packet format information;

[0286] The intelligent decision module 250 is used in the integrated sharding reorganization solution to generate the optimal sharding strategy and reorganization mechanism based on the intelligent decision mechanism according to the predicted network status data, service type data and service quality requirement data;

[0287] The simulation module 260 is used to establish a simulation platform for shard reassembly, simulate different network scenarios and service flows, and evaluate the feasibility and effectiveness of the shard reassembly solution;

[0288] The optimization module 270 is used to combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism to verify and optimize the effect of the intelligent decision-making algorithm;

[0289] The encrypted transmission module 280 is used to adopt encryption and authentication mechanisms, modular design and standardized interfaces, define data exchange formats and communication protocols between modules, and transmit and reassemble data packets during the fragment reassembly process.

[0290] The fragmentation module 210 dynamically calculates fragmentation parameters for fragmenting data packets according to the differences of different protocol stacks in heterogeneous network interconnection, including:

[0291] In the embodiment of the present application, the optimal fragmentation size is dynamically calculated according to the MTU of different protocol stacks in the heterogeneous network to form fragmentation parameters, the data packet is fragmented based on the fragmentation parameters, and sequence numbers and check information are added to form a fragmented data packet in a unified format, which is transmitted to the receiving end through the network;

[0292] The receiving end reassembles the fragments according to the sequence number and verification information, and requests retransmission if the verification fails;

[0293] According to the packet loss rate, delay status and transmission quality of the network status in the subsequent time, the packet loss rate, delay status and transmission quality of the network status in the subsequent time are predicted; based on the decision tree algorithm, the optimal fragmentation parameters are predicted; and based on the optimal fragmentation parameters, new data packets are fragmented, transmitted and reassembled.

[0294] In the embodiment of the present application, the fragmentation module 210 predicts the packet loss rate, delay status and transmission quality of the network state in the subsequent time according to the packet loss rate, delay status and transmission quality of the network state, predicts the optimal fragmentation parameters based on the packet loss rate, delay status and transmission quality of the network state in the subsequent time of the decision tree algorithm, and fragments, transmits and reassembles the new data packet based on the optimal fragmentation parameters, including:

[0295] According to the network's packet loss rate, delay status and transmission quality, the PID control algorithm is used to dynamically adjust the adaptive fragmentation parameters;

[0296] Using the decision tree algorithm, a classification model is constructed based on historical transmission data and network parameters including bandwidth, delay, and packet loss rate to predict the optimal adaptive sharding parameters under the current network conditions;

[0297] By learning and training historical data, a decision tree is generated, and classification prediction is performed based on current network parameters to obtain the optimal sharding parameter value;

[0298] Use the predicted adaptive fragmentation parameters to fragment new data packets and repeat the above-mentioned transmission, reorganization, and dynamic adjustment process to continuously optimize transmission efficiency and quality and adapt to the dynamic changes of heterogeneous networks.

[0299] In the embodiment of the present application, the integrated generation module 240 generates an adaptive fragmentation scheme based on the dynamic programming algorithm according to the input parameters, generates a unified reassembly protocol according to the adaptive fragmentation scheme, optimizes the encapsulation and decapsulation algorithm of the data packet, and forms an integrated fragmentation reassembly framework according to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm, including:

[0300] Obtain the maximum transmission unit and data packet format information of each protocol layer in the heterogeneous network as input information, calculate the optimal fragmentation size based on the dynamic programming algorithm, and generate an adaptive fragmentation solution;

[0301] Design a unified data packet format according to the adaptive fragmentation scheme to form a unified reassembly protocol;

[0302] Use heuristic algorithms to optimize packet encapsulation and decapsulation at the network layer, transport layer, and application layer;

[0303] According to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm, an integrated fragmentation reassembly framework is formed.

[0304] In the embodiment of the present application, the integrated generation module 240 calculates the optimal shard size based on a dynamic programming algorithm and generates an adaptive sharding solution, including:

[0305] Construct a state transition equation based on the following formula:

[0306] dp[i] = min(dp[i], dp[j] + 1), where 0 ≤ j < i, j + s ≤ the maximum transmission unit, and s represents the fragmentation size; assuming the data packet length is 1024 bytes, the state dp[i] represents the minimum number of fragments for dividing the first i bytes into fragments;

[0307] According to the algorithm, obtain the minimum number of fragments for dividing the first i bytes into fragments and the optimal fragmentation size, and generate an adaptive fragmentation scheme.

[0308] In the embodiment of the present application, the intelligent decision-making module 250, based on the intelligent decision-making mechanism, generates an optimal fragmentation strategy and a recombination mechanism according to the predicted network condition data, service type data, and quality of service requirement data, and establishes a simulation platform for fragmentation and recombination, simulates different network scenarios and traffic flows, evaluates the feasibility and effectiveness of the fragmentation and recombination scheme, combines the simulation results of the simulation platform with the results of the intelligent decision-making mechanism, and verifies and optimizes the effect of the intelligent decision-making algorithm, including:

[0309] Obtain the network condition data, service type data, and quality of service requirement data, and based on the prediction model, predict the network condition data, service type data, and quality of service requirement data in the subsequent time;

[0310] Use the decision tree algorithm to determine the optimal fragmentation strategy and recombination mechanism according to the network condition data, service type data, and quality of service requirement data in the subsequent time;

[0311] Based on the prediction results, dynamically adjust the fragmentation and recombination parameters, and input them into the fragmentation and recombination module for adaptive processing;

[0312] For heterogeneous networks, analyze the network topology to extract key parameters, construct a simulation model in NS-3, simulate data transmission, and set evaluation metrics;

[0313] Calculate the comprehensive evaluation metric through weighted summation, and optimize the fragmentation and recombination parameters with the particle swarm optimization algorithm;

[0314] Use Wireshark to monitor the actual network performance, calculate the deviation compared with the simulation results, and feedback the deviation to the decision tree algorithm and the prediction algorithm for iterative optimization.

[0315] In the embodiment of the present application, the simulation module 260 uses Wireshark to monitor the actual network performance, calculates the deviation compared with the simulation results, and feedbacks the deviation to the decision tree algorithm and the prediction algorithm for iterative optimization, including:

[0316] For heterogeneous network interconnection scenarios, obtain the network topology and protocol type information of each subnet, analyze the network topology, and extract key feature parameters;

[0317] A heterogeneous network interconnection model was constructed in the network simulation platform NS-3, and the extracted key characteristic parameters were input into the simulation model to simulate the data transmission process in different network scenarios.

[0318] In the simulation platform, the evaluation indicators of fragment reassembly are set based on the service priority and service quality level for different service traffic and service quality requirements; the evaluation indicators include latency, throughput, and packet loss rate;

[0319] The weighted summation method is used to calculate the weight of each evaluation index and obtain the comprehensive evaluation index;

[0320] Through the particle swarm optimization algorithm, the shard size and reorganization timing parameters are optimized with comprehensive evaluation indicators as the optimization target, and the feedback is given to the intelligent decision-making process and the reorganization process;

[0321] Use network performance monitoring tools to obtain performance monitoring data in the actual network, compare and analyze the simulation evaluation results and performance monitoring data, and calculate the evaluation index deviation;

[0322] The decision tree algorithm and prediction algorithm are iteratively optimized based on the evaluation indicator deviation through parameter adjustment and model retraining.

[0323] In the embodiment of the present application, the optimization module 270 is specifically used for:

[0324] The fragment size, reassembly delay, and cache size parameters are used as chromosomes, and the throughput, delay, and reassembly success performance indicators are used as fitness functions. Genetic algorithms are used to adaptively tune the parameters of the integrated fragment reassembly framework.

[0325] Through selection, crossover, and mutation genetic operations, the parameter combination is continuously optimized to obtain the optimal sharding and reorganization configuration.

[0326] In the embodiment of the present application, the encrypted transmission module 280 adopts encryption and authentication mechanisms, modular design and standardized interfaces, defines the data exchange format and communication protocol between modules, and transmits and reassembles data packets during the fragmentation and reassembly process, including:

[0327] If the sharding and reorganization technology is used, the data is divided into multiple shards according to the preset shard size and number, and a unique identifier is generated for each shard;

[0328] For each shard, an encryption algorithm is used for encryption processing;

[0329] The encrypted fragments are transmitted to the receiving end through a standardized interface, carrying metadata information;

[0330] If the receiving end receives the fragment, the validity of the fragment is verified through a decryption and authentication process;

[0331] According to the unique identifier and metadata of the shards, the shards are reassembled according to the predetermined order and rules to obtain the complete original data;

[0332] A machine learning algorithm is used to verify the integrity of the reorganized data.

[0333] Figure 3 A schematic diagram of an electronic device 10 that can be used to implement an embodiment of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0334] like Figure 3 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11 based on communication, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0335] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0336] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 executes the various methods and processes described above, such as a soft bus power terminal heterogeneous interconnection method based on an Ethernet protocol stack.

[0337] In some embodiments, the heterogeneous interconnection method of soft bus power terminals based on the Ethernet protocol stack can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the heterogeneous interconnection method of soft bus power terminals based on the Ethernet protocol stack described above can be executed. Alternatively, in other embodiments, the processor 11 can be configured to execute the heterogeneous interconnection method of soft bus power terminals based on the Ethernet protocol stack in any other appropriate manner (for example, by means of firmware).

[0338] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0339] Computer programs for implementing the methods of the present application may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable communication system, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0340] In the context of the present application, a computer readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device or equipment. A computer readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer readable storage medium may be a machine readable signal medium. A more specific example of a machine readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0341] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0342] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0343] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0344] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the information expected by the technical solution of this application can be achieved, and this document is not limited here.

[0345] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for heterogeneous interconnection of soft bus power terminals based on Ethernet protocol stack, characterized in that: The method comprises: Dynamically calculate fragmentation parameters for fragmenting data packets according to differences in different protocol stacks in heterogeneous network interconnection, and define a common data packet format and reassembly rules to form a unified reassembly protocol, so that the unified reassembly protocol is deployed in the network layer, the transport layer and the application layer, so that the network layer, the transport layer and the application layer process the data packet based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field; In the fragmentation reassembly process, based on the dynamic programming algorithm, an adaptive fragmentation scheme is generated according to the input parameters, and a unified reassembly protocol is generated according to the adaptive fragmentation scheme, and the encapsulation and decapsulation algorithm of the data packet is optimized. According to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm, an integrated fragmentation reassembly framework is formed; wherein the input parameters include the maximum transmission unit of each protocol layer and the data packet format information; In the integrated sharding and reorganization scheme, based on the intelligent decision-making mechanism, the optimal sharding strategy and reorganization mechanism are generated according to the predicted network status data, business type data and service quality requirement data, and a sharding and reorganization simulation platform is established to simulate different network scenarios and business traffic, evaluate the feasibility and effectiveness of the sharding and reorganization scheme, and combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism to verify and optimize the effect of the intelligent decision-making algorithm; In the process of fragmentation and reassembly, encryption and authentication mechanisms are adopted, modular design and standardized interfaces are used, and the data exchange format and communication protocol between modules are defined to transmit and reassemble data packets.

2. The method according to claim 1, characterized in that The fragmentation parameters for fragmenting data packets are dynamically calculated based on the differences in different protocol stacks in heterogeneous network interconnection, including: Dynamically calculate the optimal fragmentation size according to the MTU of different protocol stacks in the heterogeneous network to form fragmentation parameters, fragment the data packet based on the fragmentation parameters, and add sequence numbers and check information to form fragmented data packets in a unified format, which are transmitted to the receiving end through the network; The receiving end reassembles the fragments according to the sequence number and verification information, and requests retransmission if the verification fails; According to the packet loss rate, delay status and transmission quality of the network status in the subsequent time, the packet loss rate, delay status and transmission quality of the network status in the subsequent time are predicted; based on the decision tree algorithm, the optimal fragmentation parameters are predicted; and based on the optimal fragmentation parameters, new data packets are fragmented, transmitted and reassembled.

3. The method according to claim 2, characterized in that Predict the packet loss rate, delay status and transmission quality of the network status in the subsequent time according to the packet loss rate, delay status and transmission quality of the network status in the subsequent time based on the decision tree algorithm. Predict the optimal fragmentation parameters, and fragment, transmit and reassemble new data packets based on the optimal fragmentation parameters, including: According to the network's packet loss rate, delay status and transmission quality, the PID control algorithm is used to dynamically adjust the adaptive fragmentation parameters; Adopt the decision tree algorithm to construct a classification model based on historical transmission data and network parameters including bandwidth, latency, and packet loss rate, and predict the optimal adaptive fragmentation parameters under the current network conditions; Generate a decision tree through learning and training of historical data, perform classification prediction according to the current network parameters, and obtain the optimal fragmentation parameter values; Use the predicted adaptive fragmentation parameters to fragment new data packets, repeat the above processes of transmission, recombination, and dynamic adjustment, continuously optimize the transmission efficiency and quality, and adapt to the dynamic changes of heterogeneous networks.

4. The method according to claim 1, characterized in that Based on the dynamic programming algorithm, generate an adaptive fragmentation scheme according to the input parameters, and generate a unified recombination protocol according to the adaptive fragmentation scheme. Optimize the packet encapsulation and decapsulation algorithms, and form an integrated fragmentation and recombination framework according to the adaptive fragmentation scheme, the unified recombination protocol, and the optimized packet encapsulation and decapsulation algorithms, including: Obtain the maximum transmission unit and packet format information of each protocol layer in the heterogeneous network as input information, calculate the optimal fragmentation size based on the dynamic programming algorithm, and generate an adaptive fragmentation scheme; Design a unified packet format according to the adaptive fragmentation scheme to form a unified recombination protocol; Adopt a heuristic algorithm to optimize the packet encapsulation and decapsulation of the network layer, transport layer, and application layer; Form an integrated fragmentation and recombination framework according to the adaptive fragmentation scheme, the unified recombination protocol, and the optimized packet encapsulation and decapsulation algorithms.

5. The method according to claim 4, characterized in that Calculate the optimal fragmentation size based on the dynamic programming algorithm to generate an adaptive fragmentation scheme, including: Construct a state transition equation based on the following formula: dp[i] = min(dp[i], dp[j] + 1), where 0 ≤ j < i, j + s ≤ the maximum transmission unit, and s represents the fragmentation size; assume the packet length is 1024 bytes, and the state dp[i] represents the minimum number of fragments for dividing the first i bytes into fragments; Generate an adaptive fragmentation scheme according to the minimum number of fragments for dividing the first i bytes into fragments and the optimal fragmentation size obtained by the algorithm.

6. The method according to claim 1, characterized in that Based on the intelligent decision-making mechanism, generate the optimal fragmentation strategy and recombination mechanism according to the predicted network condition data, service type data, and quality of service requirement data, and establish a simulation platform for fragmentation and recombination to simulate different network scenarios and traffic flows, evaluate the feasibility and effectiveness of the fragmentation and recombination scheme, and combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism to verify and optimize the effect of the intelligent decision-making algorithm, including: Obtain network condition data, service type data, and quality of service requirement data, and predict the network condition data, service type data, and quality of service requirement data in the subsequent time based on the prediction model; Use the decision tree algorithm to determine the optimal fragmentation strategy and recombination mechanism according to the network condition data, service type data, and quality of service requirement data in the subsequent time; Based on the prediction results, dynamically adjust the fragmentation and recombination parameters, and input them into the fragmentation and recombination module for adaptive processing; For heterogeneous networks, analyze the network topology to extract key parameters, construct a simulation model in NS-3 to simulate data transmission, and set evaluation indicators; The comprehensive evaluation index is calculated by weighted summation, and the fragmentation and reorganization parameters are optimized by particle swarm optimization algorithm; Use Wireshark to monitor the actual network performance, compare the calculated deviation with the simulation results, and feed the deviation back to the decision tree algorithm and prediction algorithm for iterative optimization.

7. The method according to claim 1, characterized in that Use Wireshark to monitor actual network performance, compare it with simulation results, and calculate deviations. Feedback the deviations to the decision tree algorithm and prediction algorithm for iterative optimization, including: For heterogeneous network interconnection scenarios, obtain the network topology and protocol type information of each subnet, analyze the network topology, and extract key feature parameters; A heterogeneous network interconnection model was constructed in the network simulation platform NS-3, and the extracted key characteristic parameters were input into the simulation model to simulate the data transmission process in different network scenarios. In the simulation platform, the evaluation indicators of fragment reassembly are set based on the service priority and service quality level for different service traffic and service quality requirements; the evaluation indicators include latency, throughput, and packet loss rate; The weighted summation method is used to calculate the weight of each evaluation index and obtain the comprehensive evaluation index; Through the particle swarm optimization algorithm, the shard size and reorganization timing parameters are optimized with comprehensive evaluation indicators as the optimization target, and the feedback is given to the intelligent decision-making process and the reorganization process; Use network performance monitoring tools to obtain performance monitoring data in the actual network, compare and analyze the simulation evaluation results and performance monitoring data, and calculate the evaluation index deviation; The decision tree algorithm and prediction algorithm are iteratively optimized based on the evaluation indicator deviation through parameter adjustment and model retraining.

8. The method according to claim 7, characterized in that The method further comprises: The fragment size, reassembly delay, and cache size parameters are used as chromosomes, and the throughput, delay, and reassembly success performance indicators are used as fitness functions. Genetic algorithms are used to adaptively tune the parameters of the integrated fragment reassembly framework. Through selection, crossover, and mutation genetic operations, the parameter combination is continuously optimized to obtain the optimal sharding and reorganization configuration.

9. The method according to claim 1, characterized in that: In the process of fragmentation and reassembly, encryption and authentication mechanisms are used, modular design and standardized interfaces are adopted, data exchange formats and communication protocols between modules are defined, and data packets are transmitted and reassembled, including: If the sharding and reorganization technology is used, the data is divided into multiple shards according to the preset shard size and number, and a unique identifier is generated for each shard; For each shard, an encryption algorithm is used for encryption processing; The encrypted fragments are transmitted to the receiving end through a standardized interface, carrying metadata information; If the receiving end receives the fragment, the validity of the fragment is verified through a decryption and authentication process; According to the unique identifier and metadata of the shards, the shards are reassembled according to the predetermined order and rules to obtain the complete original data; A machine learning algorithm is used to verify the integrity of the reorganized data.

10. A soft bus power terminal heterogeneous interconnection system based on Ethernet protocol stack, characterized in that: The system comprises: A fragmentation module is used to dynamically calculate fragmentation parameters for fragmenting data packets according to the differences of different protocol stacks in heterogeneous network interconnection, and define a common data packet format and reassembly rules to form a unified reassembly protocol, so as to deploy the unified reassembly protocol in the network layer, the transport layer and the application layer, so that the network layer, the transport layer and the application layer process the data packet based on the unified reassembly protocol; the reassembly protocol includes a data packet identifier, a fragment sequence number, and a checksum field; The adaptive sharding module is used to generate an adaptive sharding scheme according to input parameters based on a dynamic programming algorithm during the sharding reorganization process; A unified reassembly module, used to generate a unified reassembly protocol according to the adaptive fragmentation scheme, and a cross-layer optimization module, used to optimize the encapsulation and decapsulation algorithms of data packets; An integrated generation module, used to form an integrated fragmentation reassembly framework according to the adaptive fragmentation scheme, the unified reassembly protocol and the optimized data packet encapsulation and decapsulation algorithm; wherein the input parameters include the maximum transmission unit and data packet format information of each protocol layer; The intelligent decision-making module is used in the integrated sharding and reorganization solution. Based on the intelligent decision-making mechanism, it generates the optimal sharding strategy and reorganization mechanism according to the predicted network status data, business type data and service quality requirement data; The simulation module is used to establish a simulation platform for sharding and reassembly, simulate different network scenarios and service flows, and evaluate the feasibility and effectiveness of sharding and reassembly solutions. The optimization module is used to combine the simulation results of the simulation platform with the results of the intelligent decision-making mechanism to verify and optimize the effect of the intelligent decision-making algorithm; The encrypted transmission module is used to adopt encryption and authentication mechanisms in the process of fragmentation and reorganization, adopt modular design and standardized interfaces, define the data exchange format and communication protocol between modules, and transmit and reorganize data packets.

Images