Miniature data transactioner and its intelligent transaction system

Abstract

The application relates to the field of edge computing and Internet of Things, and discloses a micro data transaction device, which comprises an edge computing unit used for executing a centralized or decentralized matching algorithm and a transaction protocol, a multi-dimensional detection module comprising a compliance detection unit, a security detection unit, a quality evaluation unit and a scene applicability evaluation unit, a hardware accelerator used for accelerating detection tasks and encryption operations, a communication module supporting multi-protocol data transmission, a storage unit used for buffering transaction data and smart contract codes, and a power management unit controlling total power consumption of the device, wherein the edge computing unit comprises a hierarchical trusted execution environment architecture composed of a secure enclave and a general-purpose computing unit. The application adopts a hardware isolation architecture of the secure enclave and the general-purpose computing unit, and combines a dynamic key encryption shared memory technology, so that the effects of private computing and efficient matching parallel execution are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of edge computing and Internet of Things, in particular to a micro data transaction device and an intelligent transaction system thereof. BACKGROUND

[0002] The micro data transaction device is a special device deployed in an edge computing node or an Internet of Things terminal, which is used to realize decentralized transaction and intelligent matching of data resources in a distributed environment. The intelligent transaction system thereof supports data ownership verification, privacy protection transaction and multi-chain contract collaborative execution by integrating security computing, dynamic pricing and cross-chain communication modules, and is suitable for energy transaction, medical data sharing and other scenarios. The core feature of the system is the miniaturized hardware architecture design (such as RISC-V SoC integration) and low power optimization, which can realize high throughput transaction processing in a resource-constrained environment.

[0003] Existing data transaction technologies are mostly based on cloud computing platforms or general blockchain nodes, and use centralized matching engines and fixed security strategies. For example, the trusted execution environment (TEE) technology relies on a single enclave (such as Intel SGX) to realize data isolation, and guarantees transaction privacy through software encryption; the data pricing adopts a static double auction model to determine the transaction price based on the historical average price or a fixed premium coefficient; the hardware platform uses a general-purpose processor (such as x86 CPU) to run full-node software, and manages power consumption through software-defined strategies.

[0004] The core problem of the existing technology is that the trusted execution environment architecture does not match the computing resources of the micro device. The traditional solution adopts a single TEE design (such as SGX), which needs to exclusively use computing resources for encryption operations, resulting in a conflict between transaction matching and privacy computing tasks for hardware resources. For example, zero-knowledge proof verification and transaction matching need to be executed in series within the enclave, causing transaction delay to exceed 100 ms, which cannot meet the millisecond-level response requirement of scenarios such as smart grids. The root cause of the problem is that the hardware isolation mechanism is not optimized for the computing power characteristics of micro devices, resulting in a mutual exclusion between security and efficiency. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a micro data transaction device and an intelligent transaction system thereof, which solves the problem that the single trusted execution environment architecture of the prior art cannot adapt to the resource characteristics of micro devices, resulting in difficulty in achieving both security isolation and computing efficiency.

[0006] To achieve the above purpose, the present application realizes the following technical scheme: a micro data transaction device, comprising:

[0007] an edge computing unit for executing centralized or decentralized matching algorithms and transaction protocols;

[0008] Multi-dimensional detection module, including compliance detection unit, security detection unit, quality assessment unit and scenario applicability assessment unit;

[0009] Hardware accelerator, for accelerating detection tasks and encryption operations;

[0010] Communication module, supporting multi-protocol data transmission;

[0011] Storage unit, for caching transaction data and smart contract code;

[0012] Power management unit, controlling total power consumption of the device.

[0013] Preferably, the edge computing unit comprises:

[0014] Hierarchical trusted execution environment architecture, composed of a secure enclave and a general computing unit;

[0015] The secure enclave runs independently based on the RISC-V instruction set, for performing zero-knowledge proof verification, post-quantum key agreement and data privacy calculation tasks;

[0016] The general computing unit runs a decentralized matching algorithm and a Gossip protocol, and interacts with sensitive data of the secure enclave through a memory isolation wall;

[0017] Decentralized matching algorithm module, integrating an improved double auction model and a dynamic pricing engine, with a response time ≤ 500ms;

[0018] Cross-chain smart contract conversion middleware, converting source chain contract logic into target chain executable code through a semantic adaptation layer;

[0019] Hardware acceleration interface, directly connected with the hardware accelerator through a PCIe4.0 bus, offloading compliance detection, hash calculation and encryption tasks.

[0020] Preferably, the dynamic pricing engine calculates the equilibrium price based on the improved double auction model, with the formula being:

[0021]

[0022] Wherein α is the supply and demand elasticity coefficient, is the buyer's offer, is the seller's asking price;

[0023] Distributed broadcast unit, broadcasting transaction demand in a P2P network using the Gossip protocol, with a coverage radius ≥ 1km, and a node selection strategy being:

[0024] N eighbor(u) ={v|distance(u,v)≤R∧deg(v)≤deg max};

[0025] where deg(v) is the number of neighbors of node v, deg max is the maximum connection threshold.

[0026] Preferably, the decentralized matching algorithm module further comprises:

[0027] Cross-chain smart contract conversion middleware realizes contract logic mapping by the following steps:

[0028] Parsing the syntax structure of the source chain contract to generate an intermediate language;

[0029] Reconstructing the IL into equivalent contract logic according to the syntax rules of the target chain;

[0030] Optimizing the contract execution order based on the Gas cost model of the target chain to minimize transaction fees;

[0031] A real-time state feedback unit integrates a Q-learning reinforcement learning model to dynamically adjust the cross-chain routing strategy, and its scoring formula is:

[0032] Score = w1·T h +w2·(1-C)+w3·(1-L);

[0033] where T h , C, L are the throughput, fee, and delay of the target chain, and the weights w1, w2, w3 are dynamically updated according to the transaction success rate.

[0034] Preferably, the multi-dimensional detection module comprises:

[0035] A compliance detection unit for verifying whether the data transaction complies with the data privacy regulations and data sovereignty requirements of the target region, which includes:

[0036] A dynamic legal knowledge graph that integrates version tracking and conflict detection of data regulations in 50+ countries and generates a compliance report through semantic matching;

[0037] A cross-language clause alignment module based on the LaBSE multilingual embedding model and the FAISS index library to realize real-time mapping of multilingual legal clauses.

[0038] Preferably, the security detection unit further comprises:

[0039] An integrated national encryption algorithm acceleration module and a post-quantum encryption module, including:

[0040] An SM2 / SM4 national encryption algorithm hardware accelerator for data signature and encrypted transmission;

[0041] CRYSTALS-Kyber post-quantum key agreement module, which realizes the anti-quantum attack protection of NIST PQC standard through ASIC chip.

[0042] Preferably, the quality assessment unit verifies the data quality by the following steps:

[0043] Hash check: Calculate the BLAKE3 hash value of the data stream and compare it with the hash value recorded by the evidence platform;

[0044] Timestamp deviation detection: verify that the data generation timestamp and the standard time of the time service center deviate ≤0.5 seconds.

[0045] Preferably, the scene applicability assessment unit uses a natural language processing model to analyze the metadata tags, including:

[0046] BERT semantic matching engine, generating vector representation of data description text and calculating cosine similarity with buyer scene label;

[0047] Authorization chain verification submodule, querying the buyer's data usage permission evidence for the target scene through blockchain.

[0048] Preferably, the hardware accelerator includes:

[0049] Scene assessment accelerator, integrating low-precision quantized AI inference engine, supporting INT8 data format parallel computing, used for real-time processing of semantic matching tasks of data scene labels;

[0050] Security compliance accelerator, based on ASIC chip design, supporting the following functions:

[0051] Hardware-level acceleration of national secret algorithm, including key generation, data signature and encryption;

[0052] Key agreement and packaging of NIST post-quantum cryptography algorithm CRYSTALS-Kyber;

[0053] Parallel computing of BLAKE3 hash algorithm, optimizing throughput through SIMD instruction set;

[0054] Hardware interface, directly connected with edge computing unit through PCIe4.0 bus, realizing the following collaborative mechanisms:

[0055] Offload hash calculation and signature tasks in compliance detection to security compliance accelerator;

[0056] Offload NLP inference tasks of scene applicability assessment to scene assessment accelerator;

[0057] Dynamic power management unit, using voltage frequency regulation technology, dynamically switching the following modes according to task load:

[0058] High power consumption mode: all acceleration modules are enabled, voltage 1.2V, frequency 200MHz;

[0059] Low power consumption mode: only SM4 encryption module is enabled, voltage 0.8V, frequency 50MHz.

[0060] The application also provides an intelligent transaction system of a micro data transaction device, comprising a plurality of micro data transaction devices, and the following collaborative components:

[0061] National-level computing power network nodes for storing and verifying distributed transaction data;

[0062] Cross-link routing center for dynamically selecting the optimal blockchain path through a reinforcement learning model;

[0063] Edge-cloud collaborative management platform for monitoring device status and dynamically updating compliance rule library.

[0064] The application provides a micro data transaction device and an intelligent transaction system thereof.

[0065] 1. The application adopts a hardware isolation architecture of a secure enclave and a general-purpose computing unit, combines a dynamic key encryption shared memory technology, and achieves the effects of privacy computing and efficient matching parallel execution. Compared with the performance bottleneck caused by a single trusted execution environment in the prior art, the problem that safety isolation and high-throughput data processing cannot be considered together is solved, and the application is particularly suitable for resource-limited scenarios of micro data transaction devices.

[0066] 2. The application integrates four-dimensional detection modules of compliance, safety, quality and scene applicability, combines an improved double-auction dynamic pricing formula, and realizes full-dimensional trusted verification and real-time price adaptation of transaction data. Compared with the isolated detection or static pricing model in the traditional scheme, the problem of single evaluation dimension and market response lag is solved, and the rapid decision-making demand of micro devices for complex transaction scenarios is met.

[0067] 3. The application significantly improves the efficiency of data tag matching, encryption operation and cross-chain synchronization through a scene evaluation accelerator, a safety compliance accelerator and a multi-link aggregation transmission protocol. Compared with the high delay caused by the dependence of the existing software scheme on a general-purpose processor, the defect that the micro device is difficult to support high-concurrency real-time transactions is solved, and the balance between transaction throughput and energy efficiency is ensured.

[0068] 4. The application realizes device endurance extension and cross-chain data reliable synchronization through DVFS load prediction, multi-power switching and CRDT consistency protocol. Compared with the fixed power consumption strategy or single-chain storage architecture in the traditional scheme, the pain points of uncontrollable energy consumption and frequent cross-chain data conflicts are solved, and the long-time stable operation of the micro data transaction device in a distributed environment is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0069] Fig. 1 is a main framework diagram of the present application;

[0070] Fig. 2 is a dynamic data fingerprint generation flowchart of the present application. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the specification of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0072] Please refer to the drawings in the specification of the present application Figs. 1-2 The embodiment of the present application provides a micro data transaction device, which comprises:

[0073] An edge computing unit is used to execute a centralized or decentralized matching algorithm and a transaction protocol;

[0074] A multi-dimensional detection module comprises a compliance detection unit, a security detection unit, a quality evaluation unit and a scene applicability evaluation unit;

[0075] A hardware accelerator is used to accelerate detection tasks and encryption operations;

[0076] A communication module supports multi-protocol data transmission;

[0077] A storage unit is used to cache transaction data and smart contract code;

[0078] A power management unit controls the total power consumption of the device.

[0079] The edge computing unit comprises:

[0080] A layered trusted execution environment architecture is composed of a secure enclave and a general-purpose computing unit;

[0081] The secure enclave independently runs based on the RISC-V instruction set and is used to execute zero-knowledge proof verification, post-quantum key agreement and data privacy calculation tasks;

[0082] The general-purpose computing unit runs a decentralized matching algorithm and a Gossip protocol and interacts with sensitive data of the secure enclave through a memory isolation wall;

[0083] A decentralized matching algorithm module integrates an improved double-auction model and a dynamic pricing engine, and the response time is ≤500 ms;

[0084] Cross-chain smart contract conversion middleware, which converts source chain contract logic into target chain executable code through a semantic adaptation layer;

[0085] Hardware acceleration interface, directly connected with hardware accelerator through PCIe4.0 bus, offloading compliance detection, hash calculation and encryption tasks.

[0086] Dynamic pricing engine, which calculates the equilibrium price based on the improved double auction model, whose formula is:

[0087]

[0088] Where α is the supply and demand elasticity coefficient, is the buyer's offer, is the seller's asking price;

[0089] Distributed broadcast unit, which broadcasts transaction demand in P2P network using Gossip protocol, coverage radius ≥1km, and node selection strategy is:

[0090] N eighbor(u) ={v|distance(u,v)≤R∧deg(v)≤deg max};

[0091] Where deg(v) is the number of neighbors of node v, and deg max is the maximum connection threshold.

[0092] The decentralized matching algorithm module also includes:

[0093] Cross-chain smart contract conversion middleware, which realizes contract logic mapping through the following steps:

[0094] Parse the syntax structure of the source chain contract to generate an intermediate language;

[0095] Reconstruct IL into equivalent contract logic according to the syntax rules of the target chain;

[0096] Optimize contract execution order based on the Gas cost model of the target chain to minimize transaction fees;

[0097] Real-time state feedback unit, integrating Q-learning reinforcement learning model, dynamically adjusts cross-chain routing strategy, whose scoring formula is:

[0098] Score=w1·T h +w2·(1-C)+w3·(1-L);

[0099] Where T h , C, L are the throughput, fee and delay of the target chain, and weights w1, w2, w3 are dynamically updated according to the transaction success rate.

[0100] The multi-dimensional detection module comprises:

[0101] A compliance detection unit for verifying whether the data transaction complies with data privacy regulations and data sovereignty requirements in the target region, comprising:

[0102] A dynamic legal knowledge graph integrating version tracking and conflict detection of data regulations in 50+ countries, generating a compliance report through semantic matching;

[0103] A cross-language clause alignment module based on LaBSE multilingual embedding model and FAISS index library, realizing real-time mapping of multilingual legal clauses.

[0104] The security detection unit further comprises:

[0105] An integrated national encryption algorithm acceleration module and post-quantum encryption module, comprising:

[0106] An SM2 / SM4 national encryption algorithm hardware accelerator for data signature and encrypted transmission;

[0107] A CRYSTALS-Kyber post-quantum key agreement module that realizes NISTPQC standard anti-quantum attack protection through ASIC chips.

[0108] The quality assessment unit verifies data quality through the following steps:

[0109] Hash check: Calculate the BLAKE3 hash value of the data stream and compare it with the hash value recorded by the evidence platform;

[0110] Timestamp deviation detection: Verify that the data generation timestamp and the standard time deviation of the time center are ≤0.5 seconds.

[0111] The scene applicability evaluation unit uses natural language processing models to analyze metadata tags, including:

[0112] BERT semantic matching engine, generating vector representation of data description text and calculating cosine similarity with buyer scenario label;

[0113] An authorized chain verification submodule queries the buyer's data usage permission for the target scenario through the blockchain.

[0114] The hardware accelerator comprises:

[0115] A scene evaluation accelerator integrating a low-precision quantized AI inference engine, supporting INT8 data format parallel computing, for real-time processing of semantic matching tasks of data scene labels;

[0116] A security compliance accelerator based on ASIC chip design, supporting the following functions:

[0117] Hardware-level acceleration of SM4, including key generation, data signing, and encryption.

[0118] Key agreement and encapsulation of NIST post-quantum cryptography algorithm CRYSTALS-Kyber.

[0119] Parallel computing of BLAKE3 hash algorithm, optimized throughput through SIMD instruction set.

[0120] Hardware interface, directly connected to the edge computing unit through the PCIe 4.0 bus, to realize the following collaborative mechanisms:

[0121] Offload hash computation and signature tasks in compliance detection to the security compliance accelerator.

[0122] Offload NLP inference tasks for scenario applicability evaluation to the scenario evaluation accelerator.

[0123] Dynamic power management unit, using voltage frequency regulation technology, dynamically switching the following modes according to task load:

[0124] High power consumption mode: enable all acceleration modules, voltage 1.2V, frequency 200MHz.

[0125] Low power consumption mode: only SM4 encryption module is enabled, voltage 0.8V, frequency 50MHz.

[0126] In this embodiment, the edge computing unit adopts a layered trusted execution environment architecture to realize secure isolation and efficient computation of data transactions. Specifically, the edge computing unit includes a secure enclave and a general-purpose computing unit, which realize parallel processing of sensitive tasks and general tasks through a hardware-level isolation mechanism, ensuring that privacy computing tasks are executed in a protected environment, while general computing tasks maintain high throughput.

[0127] As an option, the secure enclave is implemented based on a reduced instruction set architecture. For example, the secure enclave uses the RISC-V instruction set to build an independent running environment, and is configured with a special secure storage area and a cryptographic coprocessor. The secure storage area is used to store key materials and intermediate results of privacy computation, and the cryptographic coprocessor is used to perform high-security demand tasks such as zero-knowledge proof verification and post-quantum key agreement. It should be noted that the secure enclave divides access permissions through a memory management unit. If an unauthorized process attempts to access the memory area of the secure enclave, a hardware exception interrupt will be triggered and sensitive data will be cleared, thereby avoiding the risk of side channel attacks.

[0128] In one possible implementation, the general computing unit runs a lightweight operating system kernel for processing the decentralized matching algorithm, transaction broadcast protocol and smart contract parsing tasks. It can be understood that the data interaction between the general computing unit and the secure enclave is realized through a shared memory area, and preferably, the shared memory area adopts a dynamic encryption strategy, and the data is encrypted by the secure enclave to generate a temporary session key when writing, and needs to be decrypted after the permission verification of the secure enclave when reading.

[0129] Specifically, the decentralized matching algorithm module integrates an improved double auction model and a dynamic pricing engine. The dynamic pricing engine calculates the equilibrium price according to the real-time supply and demand relationship, and its formula is defined as:

[0130]

[0131] Wherein: Pj is the bid of the jth buyer, representing the highest bid of the buyer for the target data or service;

[0132] Pk is the asking price of the kth seller, representing the minimum acceptable price of the seller for providing data or service;

[0133] D supply is the current total supply, which is obtained by accumulating the total amount of data to be traded by all sellers;

[0134] D demand is the current total demand, which is obtained by accumulating the total demand of all buyers;

[0135] α is the supply and demand elasticity coefficient, used to adjust the sensitivity of the price to the supply and demand imbalance, and the value range is 0 < α ≤ 1, and preferably, it is dynamically adjusted according to the sliding window analysis of historical transaction data.

[0136] The formula running process is: bid aggregation: collect the buyer bid set and the seller asking price set Sort by price to generate a supply and demand curve;

[0137] Equilibrium calculation: calculate the weighted average price base value, and adjust the influence of supply and demand imbalance on the price through the normalization factor max(D supply ,D demand );

[0138] Transaction matching: cross-match the highest bid of the buyer with the lowest asking price of the seller at the equilibrium price P match to generate the final transaction pair.

[0139] Normalization factor: eliminate the dimensional difference of supply and demand through max(D supply ,D demand ), so that the price fluctuation amplitude is proportional to the market imbalance degree;

[0140] Elasticity dynamic update: When the market supply and demand fluctuate frequently, increase alpha to improve price sensitivity; when the market is stable, reduce alpha to maintain price stability.

[0141] The distributed broadcast unit uses the Gossip protocol to broadcast transaction demand in the P2P network, and the node selection strategy is defined as:

[0142] N eighbor(u) ={v|distance(u,v)≤R∧deg(v)≤deg max};

[0143] distance(u,v): the physical distance between node u and node v, estimated by GPS coordinates or network delay;

[0144] R: communication coverage radius, for example, set to 1 km to meet the regional transaction demand;

[0145] deg(v): the current number of connections of node v, indicating the number of neighbor nodes;

[0146] deg max : maximum connection threshold, used to avoid network congestion.

[0147] Implementation process:

[0148] Neighbor selection: only select nodes with distance ≤ R and connection number ≤ deg max as broadcast targets;

[0149] Load balancing: when the number of connections of node v exceeds deg max x, automatically stop receiving new requests;

[0150] Priority broadcast: high-priority transaction requests (such as real-time energy data) are given priority in spreading through the Gossip protocol.

[0151] Physical distance limit: limit the communication radius R to ensure that transaction demand spreads within a reasonable geographic range, suitable for regional energy transaction scenarios;

[0152] Connection threshold: limit single-node load through deg max to prevent network topology overload

[0153] The middleware realizes the compatibility conversion of multi-chain contract logic through the semantic adaptation layer, which includes the following steps:

[0154] Syntax analysis: parse the abstract syntax tree (AST) of the source chain smart contract, extract function logic, state variables, and event listeners;

[0155] Intermediate language generation: convert the parsing results into an intermediate representation (IR) that is chain-independent, preferably in the LLVM IR format;

[0156] Target chain reconstruction: map the IR to equivalent contract code according to the characteristics of the target chain virtual machine instruction set (such as EVM bytecode, WASM instruction set);

[0157] Gas optimization: based on the Gas cost model of the target chain, adjust the contract execution order through static code analysis, such as splitting loops into parallel batches to reduce Gas consumption.

[0158] Intermediate language layer: abstract expression of multi-chain contract logic through IR, solving the interaction barriers of heterogeneous chains such as Ethereum and Fabric;

[0159] Gas cost optimization: optimize contract code structure to adapt to the economic model of the target chain, reducing user transaction costs.

[0160] As an option, the compliance detection unit integrates a dynamic legal knowledge graph and a cross-language clause alignment module. Specifically, the dynamic legal knowledge graph extracts key entities (such as "personal data" and "cross-border transmission restrictions") from data privacy regulations by crawling official websites of multiple national legislative bodies and regulatory documents, and constructs a graph structure with law, region, and data type as nodes.

[0161] It should be noted that the graph realizes the correlation analysis of clauses and transaction data through a semantic matching engine. For example, when it is detected that the data recipient is located in the European Union, the clause on data subject consent in the GDPR is automatically matched, and a compliance report is generated.

[0162] In one possible implementation, the cross-language clause alignment module encodes multilingual legal clauses into a unified vector space based on the LaBSE (Language-agnostic BERT Sentence Embedding) model.

[0163] Preferably, the FAISS (Facebook AISimilarity Search) index library is used to realize real-time semantic similarity retrieval. It can be understood that this module solves the alignment problem of Chinese and English legal clauses, such as mapping the 35th article of the "People's Republic of China Data Security Law" to the GDPR Article 44, ensuring that cross-border transactions meet the requirements of both jurisdictions.

[0164] The security detection unit integrates a national encryption algorithm acceleration module and a post-quantum encryption module. Specifically, the national encryption algorithm acceleration module supports SM2 digital signature and SM4 symmetric encryption algorithm. For example, the SM2 signature private key is generated in a secure enclave, and the signature process is implemented through a hardware accelerator. Preferably, the signature speed can meet the high-concurrency transaction demand; the SM4 encryption adopts the CBC mode to block encrypt the data stream, ensuring the data confidentiality in the transmission process.

[0165] As an option, the post-quantum encryption module implements quantum attack-resistant key agreement based on the CRYSTALS-Kyber algorithm. It should be noted that the module accelerates polynomial multiplication and modulus reduction operations through an ASIC chip. Preferably, the key packaging and unpackaging delay is controlled within milliseconds to meet the real-time transaction scenario requirements.

[0166] The quality assessment unit verifies data integrity and timeliness through hash check and timestamp deviation detection. Specifically, the hash check includes the following steps:

[0167] Data slice hash calculation: slice the input data stream and calculate the BLAKE3 hash value of each slice;

[0168] Evidence comparison: aggregate the slice hash values into a Merkle tree structure and compare them with the root hash value recorded by the blockchain evidence platform to ensure that the data has not been tampered with.

[0169] It can be understood that the timestamp deviation detection module synchronizes with the national time service center through the NTP protocol to verify whether the deviation of the data generation timestamp and the standard time is within the allowed range. For example, when the deviation exceeds the threshold, the data is automatically marked as invalid and an alarm is triggered.

[0170] The scenario applicability assessment unit uses a natural language processing model to parse the metadata label and match it with the buyer's scenario demand. Specifically:

[0171] BERT semantic matching engine: encode the data description text (such as "lung cancer CT image data set") and the buyer's scenario label (such as "AI drug research and development") into 768-dimensional vectors, and calculate the cosine similarity;

[0172] Authorization chain verification: query the buyer's data usage authorization evidence for the target scenario through the blockchain, and for example, verify whether the medical institution has the "cancer research" scenario authorization certificate.

[0173] In one possible implementation, the semantic matching engine integrates low-precision quantization technology, converting the model weight from FP32 to INT8 format, and preferably, uses a hardware accelerator to improve the inference speed to meet real-time requirements.

[0174] As an option, the scene evaluation accelerator integrates a low-precision quantized AI inference engine for real-time processing of semantic matching tasks of data scene labels. Specifically, the accelerator supports parallel computing of INT8 data format, significantly reducing the computational resource consumption by quantizing the weights of a pre-trained model (such as BERT) from FP32 to INT8. It can be understood that the quantization process adopts a dynamic range calibration technique, preferably by calibrating the activation value distribution through a data set statistics to dynamically adjust the quantization parameters to minimize the precision loss.

[0175] Exemplarily, in a medical data transaction scenario, the scene evaluation accelerator parses the metadata labels (such as "lung cancer" and "high resolution") of CT image data, calculates the semantic similarity between the data and the buyer's scene demand (such as "AI drug research and development") through an INT8 quantized model, and outputs a matching score to assist transaction decision-making.

[0176] The security compliance accelerator is based on an ASIC chip design and supports hardware-level acceleration of national cryptographic algorithms, post-quantum cryptographic algorithms, and hash algorithms. Specifically:

[0177] National cryptographic algorithm acceleration module:

[0178] SM2 signature: Key generation and signature verification are implemented through a dedicated elliptic curve operation circuit, preferably, private key generation is completed in a secure enclave, and the signature process is processed through a hardware pipeline in parallel;

[0179] SM4 encryption: The data stream is divided into blocks for encryption using the CBC (Cipher Block Chaining) mode, and high-throughput encryption transmission is achieved through a SPN (Substitution-Permutation Network) structure hardware circuit.

[0180] Post-quantum cryptographic module:

[0181] Key agreement and encapsulation resistant to quantum attacks are implemented based on the CRYSTALS-Kyber algorithm. It should be noted that the module integrates a polynomial multiplier and a modulus reduction circuit through an ASIC chip, and preferably, the key encapsulation process includes the following steps:

[0182] Generate a Kyber parameter set (such as Kyber-768) to construct a lattice problem;

[0183] Accelerate polynomial multiplication operations through NTT (Fast Fourier Transform);

[0184] Encapsulate the shared key and bind it with the receiver's public key to generate a quantum attack-resistant ciphertext.

[0185] Data integrity check is implemented by BLAKE3 algorithm, and the parallelism of hash calculation is optimized by SIMD (Single Instruction Multiple Data) instruction set. For example, the module divides the input data stream into 512-bit data blocks, and processes multiple data blocks in parallel through a multi-stage pipeline, significantly improving the hash throughput.

[0186] The hardware accelerator is directly connected to the edge computing unit through the PCIe 4.0 bus, realizing low-latency task offloading. Specifically:

[0187] Task offloading queue: the edge computing unit encapsulates tasks such as hash calculation and signature verification in compliance detection as instruction packages and submits them to the task queue of the accelerator;

[0188] Interrupt notification mechanism: after the accelerator completes task processing, it sends an interrupt signal to the edge computing unit through MSI-X (Message Signal Interrupt Extension), triggering the result to return;

[0189] Data consistency protocol: DMA (Direct Memory Access) technology is used to realize zero-copy data transmission between the accelerator and the main memory, avoiding performance loss caused by CPU intervention.

[0190] The dynamic power management unit dynamically adjusts the voltage and frequency of the accelerator according to the task load. Specifically:

[0191] High power mode: when the number of concurrent tasks is detected to be ≥3 or the hash calculation request is intensive, all acceleration modules are activated, the voltage is raised to 1.2V, and the frequency is adjusted to 200MHz;

[0192] Low power mode: when the device is idle for more than a threshold value or the power is lower than a preset value, only the SM4 encryption module is enabled, the voltage is reduced to 0.8V, the frequency is adjusted to 50MHz, and the remaining modules enter sleep state.

[0193] It can be understood that power management dynamically switches between different modes by monitoring the task queue depth and system energy consumption in real time, to balance performance and energy efficiency.

[0194] As an option, the dynamic protocol switching engine dynamically selects the transmission protocol according to the network state, device power and data priority. Specifically, the engine triggers protocol switching through the following parameters:

[0195] Network signal strength: when the received signal strength indication (RSSI) is lower than a threshold value, switch to a high penetration protocol (such as LoRa);

[0196] Data urgency: high-priority transaction data (such as real-time energy transaction instructions) is switched to a low-latency protocol (such as 5GNR);

[0197] Device remaining power: switch to low-power protocol (e.g. NB-IoT) when power is below a pre-set value.

[0198] It can be understood that the switching logic is implemented through a state machine model. By way of example, the state transition function is defined as:

[0199] Protocol next = f(RSSI, Power remain , Priority data );

[0200] wherein Protocol next is the next state protocol type, and f is a switching function based on a weighted decision tree.

[0201] The adaptive coding and modulation module dynamically adjusts the coding rate and modulation mode according to the channel quality. Specifically, the module optimizes the transmission efficiency through the following steps:

[0202] Channel state estimation: the receiving end periodically feeds back the channel quality indication (CQI), including the signal-to-noise ratio (SNR) and the bit error rate (BER);

[0203] Modulation and coding scheme (MCS) selection: select the optimal MCS level based on the CQI table, and the mapping relationship is:

[0204]

[0205] wherein γ min is the minimum signal-to-noise ratio threshold, and Δγ is the signal-to-noise ratio interval;

[0206] Dynamic adjustment: if the number of consecutive transmission failures exceeds the threshold, reduce the MCS level to improve robustness.

[0207] It should be noted that the module supports two encoding modes, LDPC (low-density parity-check code) and Polar code. Preferably, the encoding scheme is selected according to the protocol type (e.g. 5G NR prefers to use Polar code).

[0208] The communication module integrates the Q-learning algorithm to optimize the multi-hop transmission path. Specifically:

[0209] State space: defined as {node remaining power, link delay, node load};

[0210] Action space: select the next hop node or switch the transmission protocol;

[0211] Reward function:

[0212]

[0213] where w1, w2, w3 are weight coefficients, Delay is the end-to-end delay, Power is the energy consumption, and PacketLoss is the packet loss rate. saving For energy saving gain, PacketLoss is the packet loss rate.

[0214] For example, in the energy data transaction scenario, the Q-learning model selects a low-delay and high-reliability transmission path through an exploration-exploitation strategy while balancing node energy consumption.

[0215] The multi-link aggregation transmission module improves throughput by binding multiple physical links. Specifically:

[0216] Link offloading: split data streams and transmit them concurrently through different protocols (such as 5G NR and LoRa dual channels);

[0217] Load balancing: allocate data piece sizes based on real-time bandwidth of links, with allocation weights as:

[0218]

[0219] where B i is the available bandwidth of the ith link;

[0220] Reassembly and verification: the receiving end reassembles data pieces according to sequence numbers and performs BLAKE3 hash verification for integrity.

[0221] Low-power wake-up mechanism reduces standby power consumption by combining periodic listening with event triggering. Specifically:

[0222] Periodic listening: the device wakes up to receive beacon frames every fixed time window (e.g., 60 seconds) to detect whether there is data to be received;

[0223] Event triggering: when high-priority data arrives, the sending end wakes up the receiving device through a long preamble. Preferably, the preamble length is positively correlated with the transmission distance.

[0224] As an option, the communication module uniquely identifies and integrity checks data packets through a dynamic data fingerprint generation algorithm before data transmission. Specifically, the algorithm process includes the following steps:

[0225] Feature extraction: generate a feature vector based on data content, with dimension compression formula as:

[0226]

[0227] where D raw is the original data, LSH is the local sensitive hash function, k is the number of hash bits (preferably, k = 128), denotes the XOR operation;

[0228] Dynamic fingerprint generation: according to the current timestamp t and the device unique identifier ID dev Generate dynamic fingerprint:

[0229] F fingerprint = SM3(V feature ||t||ID dev );

[0230] Wherein, SM3 is a national secret hash algorithm, and || represents data splicing;

[0231] Fingerprint embedding and verification: embed F fingerprint in the data packet header, and the receiving end verifies the data integrity and source legality through the same algorithm.

[0232] It should be noted that the algorithm is adapted to the computing power of the micro device through lightweight design, and the single fingerprint generation time is less than 5ms (based on Cortex-M7 core test).

[0233] In this embodiment, the storage unit is used to cache transaction data and smart contract code, and high-efficiency access and secure persistence are realized through a hierarchical storage architecture and a dynamic data management strategy. Specifically, the storage unit includes a transaction cache area, a contract code library and a distributed storage interface, supporting multi-mode data reading and writing and cross-chain data indexing.

[0234] As an option, the storage unit uses the LRU-K (Least Recently Used-K times) algorithm to manage the transaction cache area. Specifically, the cache eviction strategy dynamically adjusts the priority based on the access frequency and time locality of the data block, and the weight calculation formula is defined as:

[0235]

[0236] Wherein, f i is the access frequency of data block i, t last is the recent access time, w1 and w2 are weight coefficients, and T max is the maximum storage duration threshold. It should be noted that when the cache space is insufficient, the data block with the lowest score is preferentially evicted.

[0237] For example, in the energy transaction scenario, real-time electricity price data with high frequency access is retained in the cache area, and low-frequency historical data is stored in the distributed storage node.

[0238] The storage unit isolates sensitive data and non-sensitive data through hardware-level security partitioning. Specifically:

[0239] Sensitive data area: store key materials, user privacy data, etc., and use a dynamic key encryption strategy. Preferably, the encryption key is generated by a secure enclave and stored by an SM4 algorithm;

[0240] Non-sensitive data area: stores transaction logs, contract codes, etc., uses transparent encryption (TDE) technology, and the encryption key is managed by the hardware security module (HSM).

[0241] It is understood that the dynamic key generation process includes:

[0242] Securely generate a temporary session key K temp ;

[0243] Use K temp After encrypting the data, the ciphertext and key are encapsulated into a secure object;

[0244] The security object is written into the storage unit and the integrity check value is recorded through the hash chain.

[0245] The storage unit achieves data persistence and cross-node synchronization through distributed storage protocols (such as IPFS and Arweave). Specifically:

[0246] Data sharding and redundant coding: Transaction data is sharded into fixed-size data blocks, and redundant check blocks are generated through erasure coding. Preferably, the ratio of data blocks to check blocks is set to k:n.

[0247] Cross-chain indexing service: Builds a B+ tree index structure with data hash as the primary key, supporting fast retrieval of cross-chain transaction records. For example, index metadata is stored in a lightweight blockchain (such as MerklePatriciaTrie) to ensure immutability.

[0248] It should be noted that the redundant coding parameters k and n are dynamically adjusted according to the availability of storage nodes and network latency. When the node online rate is lower than the threshold, the check block ratio is automatically increased to improve fault tolerance.

[0249] The storage unit optimizes storage resources through hot and cold data classification strategies. Specifically:

[0250] Hot data: Frequently accessed transaction status data (such as unconfirmed transaction queues) is retained in the NVMe SSD, supporting low-latency read and write;

[0251] Warm data: Periodically accessed contract codes and historical transaction logs are transferred to SATA SSD;

[0252] Cold data: Low-frequency archived data (such as transaction records that have not been accessed for more than 6 months) is migrated to distributed storage nodes, and compression algorithms (such as Zstandard) are triggered to reduce storage costs.

[0253] In one possible implementation, the data migration strategy is based on an access pattern prediction model, the input features of which include data age, access frequency and associated transaction scenarios. By way of example, an LSTM (Long Short-Term Memory) model is used to predict the access probability in the next 7 days, and the storage hierarchy is dynamically adjusted.

[0254] The storage unit implements cross-chain data eventual consistency through an improved CRDT (Conflict-Free Replicated Data Type) protocol. Specifically:

[0255] Operation log synchronization: data update operations are encoded as incremental logs, which are broadcast to other chain nodes through the Gossip protocol;

[0256] Conflict resolution: when a version conflict is detected, the operation records are merged based on timestamps and transaction priorities. Preferably, a VectorClock is used to mark the operation timing, and the resolution rule is defined as:

[0257]

[0258] where Priority(O) is the priority weight of transaction operation O, which is calculated based on factors such as transaction amount and participant credit score.

[0259] The storage unit dynamically switches storage media according to real-time I / O pressure. Specifically:

[0260] High load mode: when the number of concurrent read and write requests exceeds a threshold, NVMe SSD and memory cache are used in cooperation to process, and DMA (Direct Memory Access) technology is used to bypass the CPU to achieve zero-copy data transmission;

[0261] Low load mode: when the system is idle, data is migrated to high-density QLC SSD to reduce power consumption and prolong hardware life.

[0262] It can be understood that the switching strategy dynamically adjusts the activation state of the storage medium by monitoring the I / O queue depth, delay and error rate in real time.

[0263] In this embodiment, the power management unit is used to control the total power consumption of the device, and through dynamic voltage and frequency adjustment (DVFS), task scheduling strategy optimization and multi-power input switching technology, the balance between energy consumption and performance is achieved. Specifically, the unit dynamically adjusts the power supply strategy by monitoring the system load, remaining power and external power supply state in real time, to ensure reliable operation of the device under various working conditions.

[0264] As an option, the DVFS module dynamically adjusts the voltage and frequency of the processor core based on system load prediction. Specifically, the module achieves this by the following steps:

[0265] Load coefficient calculation:

[0266]

[0267] wherein Q active is the current active task number, Q max is the maximum concurrent task capacity of the processor;

[0268] Voltage-frequency mapping: select the optimal voltage-frequency pair (V, g) according to the table lookup of β, which satisfies the mapping relationship:

[0269] f = f base min(1, β + δ);

[0270] wherein f base is the reference frequency, and δ is the load margin coefficient, used to prevent frequency mutation;

[0271] Smooth transition: use a ramp control algorithm to gradually adjust the voltage and frequency to avoid instantaneous current shock. It should be noted that the load margin coefficient δ is dynamically calibrated by historical load fluctuation data, and preferably, an exponential weighted moving average (EWMA) algorithm is used for updating.

[0272] It should be noted that the load margin coefficient δ is dynamically calibrated by historical load fluctuation data, and preferably, an exponential weighted moving average (EWMA) algorithm is used for updating.

[0273] The intelligent task scheduling strategy allocates computing resources according to task priority and energy cost. Specifically, the task priority weight W i is calculated by the following formula:

[0274]

[0275] wherein P task is the task urgency (determined by transaction type), E saving is the estimated energy saving benefit, and α is the weight adjustment factor. It can be understood that high priority tasks (such as real-time transaction matching) are allocated to high-performance cores, and low priority tasks (such as log archiving) are allocated to low-power coordination processors.

[0276] For example, in the medical data transaction scenario, patient privacy computing tasks are assigned the highest priority and are executed on high-performance cores; non-real-time data backup tasks are delayed until the device is idle.

[0277] The multi-power input switching module supports seamless switching of external DC power, lithium batteries and energy harvesting units (such as solar energy). Specifically:

[0278] Priority strategy: prefer to use external power supply, and automatically switch to lithium battery when external power is detected to be disconnected;

[0279] Hybrid power supply mode: when the external power supply is insufficient, the parallel power supply of the external power supply and the lithium battery is coordinated through the maximum power point tracking (MPPT) technology, and the output current distribution formula is:

[0280]

[0281] where P req is the total power consumption demand of the system, P ext is the available power of the external power supply, η is the conversion efficiency, V bat is the battery voltage;

[0282] Energy harvesting optimization: when the solar panel is connected, the MPPT algorithm dynamically adjusts the load impedance to make the energy harvesting unit work at the maximum power point.

[0283] It should be noted that a voltage slow rise / slow fall circuit is used during power switching to avoid device reset caused by power interruption or voltage sudden change.

[0284] The energy prediction module predicts the future energy consumption trend of the device through the LSTM (Long Short Term Memory) model. Specifically:

[0285] Input features: historical power consumption sequence, current task queue depth, environmental temperature;

[0286] Output result: power consumption curve E pred (t) in the future T time window;

[0287] State machine switching: according to

[0288] If , enter power saving mode and turn off non-critical peripherals (such as display screen, redundant communication module);

[0289] If , restore full function mode.

[0290] It can be understood that the prediction model is combined with offline training and online fine-tuning, and preferably, the federated learning framework is used to aggregate energy consumption mode data of multiple devices to improve prediction accuracy.

[0291] It can be understood that the prediction model is combined with offline training and online fine-tuning, and preferably, the federated learning framework is used to aggregate energy consumption mode data of multiple devices to improve prediction accuracy.

[0292] The power consumption monitoring module collects current, voltage and temperature data of each functional unit in real time, and realizes abnormal protection through the following mechanism:

[0293] Overcurrent protection: if the detected current instantaneous value exceeds the threshold I max, immediately cut off the power supply and trigger the fuse mechanism;

[0294] Thermal management: when the chip temperature T junc > 80°C, dynamic frequency reduction and start the cooling fan; safe

[0295] Battery health management: estimate the battery cycle life by coulomb counting method, when the health H bat <80%, limit the maximum discharge current to 70% of the rated value.

[0296] The intelligent transaction system of a micro data transactioner described below can be referred to in conjunction with the micro data transactioner described above.

[0297] An intelligent transaction system of a micro data transactioner includes a plurality of micro data transactioners and the following collaborative components:

[0298] National-level computing power network nodes for distributed transaction data storage and verification;

[0299] Cross-link routing center, dynamically selects the optimal blockchain path through reinforcement learning model;

[0300] Edge-cloud collaborative management platform, monitors device status and dynamically updates compliance rule base

[0301] The system of the embodiment can be used to execute the above-mentioned device embodiments, and the principles and technical effects are similar, which will not be repeated here.

[0302] Embodiment 1:

[0303] In the cross-institutional medical data sharing scenario, the micro data transactioner executes the following processes:

[0304] Privacy data verification: patient electronic medical record data is processed by layered trusted execution environment, sensitive fields (such as name, ID number) are encrypted in secure enclave, and non-sensitive fields (such as diagnosis conclusion) are matched by general computing unit;

[0305] Dynamic fingerprint generation: the communication module generates a dynamic fingerprint F fingerprint from the encrypted medical record data, and transmits it to the cloud medical data platform through the 5G NR protocol;

[0306] Cross-chain storage: the distributed storage interface stores the transaction record into the medical alliance chain and the IPFS node through the CRDT protocol, and synchronizes the on-chain index of each institution;

[0307] Compliance detection: the security compliance accelerator verifies the access rights of the data user in real time, and if an unauthorized institution requests decryption, it immediately triggers transaction termination and records the violation log. ​

[0308] Example 2:

[0309] In the industrial IoT device status data transaction, the micro data transactioner runs as follows:

[0310] Multi-protocol switching: The communication module automatically switches to the LoRa protocol according to the network signal strength (RSSI = -90 dBm) in the factory, and uploads the device vibration sensor data to the edge server;

[0311] Low power consumption optimization: The power management unit detects that the device power is lower than 20%, triggers the DVFS module to reduce the processor frequency to 800MHz, and closes unnecessary peripherals (such as redundant communication links);

[0312] Dynamic pricing: The intelligent transaction system adjusts the data price based on the real-time device health score (predicted by the LSTM model), and the device data pricing is floated by 30% when the health score is lower than 70%;

[0313] Conflict resolution: The distributed storage interface detects version conflicts of device status data between multiple edge nodes, and merges data records based on timestamps and device priorities (such as production line critical device priority).

[0314] Although the embodiments of the present application have been shown and described, it can be understood by those skilled in the art that various changes, modifications, replacements and variations can be made to these embodiments without departing from the principles and spirits of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A micro data transactioner, characterized by, Comprise: Edge computing unit for executing centralized or decentralized matching algorithm and transaction protocol; Multi-dimensional detection module including compliance detection unit, security detection unit, quality assessment unit and scenario applicability assessment unit; Hardware accelerator for accelerating detection tasks and encryption operations; Communication module supporting multi-protocol data transmission; Storage unit for caching transaction data and smart contract code; Power management unit to control the total power consumption of the device; The edge computing unit comprises: Hierarchical trusted execution environment architecture composed of secure enclave and general computing unit; The secure enclave runs independently based on the RISC-V instruction set, and is used to perform zero-knowledge proof verification, post-quantum key agreement and data privacy calculation tasks; The general computing unit runs the decentralized matching algorithm and Gossip protocol, and interacts with sensitive data through the memory isolation wall and the secure enclave; Decentralized matching algorithm module, integrating improved double auction model and dynamic pricing engine, response time ≤ 500ms; Cross-chain smart contract conversion middleware converts source chain contract logic into executable code of target chain through semantic adaptation layer; Hardware acceleration interface directly connected with hardware accelerator through PCIe4.0 bus, offloading compliance detection, hash calculation and encryption tasks; The decentralized matching algorithm module further comprises: Cross-chain smart contract conversion middleware realizes contract logic mapping through the following steps: Parse the syntax structure of the source chain contract to generate an intermediate language; Reconstruct the IL into equivalent contract logic according to the syntax rules of the target chain; Optimize the contract execution order based on the Gas cost model of the target chain to minimize transaction fees; Real-time state feedback unit integrates Q-learning reinforcement learning model to dynamically adjust cross-chain routing strategy, whose score formula is: Score ; wherein are the throughput, commission, latency of the target chain, respectively, weights are updated dynamically according to the transaction success rate.

2. A micro data transactioner according to claim 1, wherein, The dynamic pricing engine in the decentralized matching algorithm module calculates the equilibrium price based on the improved double auction model, whose formula is: ； wherein, : final matching price, which is a uniform transaction price calculated for a batch of buyers and sellers in a transaction matching; : Number of buyers matched with this trade; : Number of sellers participating in the match; for supply elasticity coefficient, for buyer offer, for seller asking price; : total supply of data in the current transaction pool; : total data demand in the current transaction pool; Distributed broadcast unit adopts Gossip protocol to broadcast transaction demand in P2P network, with coverage radius ≥ 1km, and node selection strategy is: ； wherein is the number of neighbors of a node is the number of neighbors of a node is the maximum number of connections threshold : a set of neighbor nodes of the node ; : the current node that initiates the broadcast; : a candidate neighbor node in the network; : for computing nodes and nodes between the geographic or logical distance; : a preset maximum broadcast radius; : degree of a node ​ : maximum degree threshold set by the system.

3. A micro data transactioner according to claim 1, wherein, The multi-dimensional detection module comprises: Compliance detection unit for verifying whether data transaction complies with data privacy regulations and data sovereignty requirements in target region, which includes: Dynamic legal knowledge graph integrates version tracking and conflict detection of data regulations in 50+ countries, and generates compliance report through semantic matching; Cross-language clause alignment module realizes real-time mapping of multi-lingual legal clauses based on LaBSE multi-language embedding model and FAISS index library.

4. A micro data transactioner according to claim 1, wherein, The security detection unit further comprises: Integrates national encryption algorithm acceleration module and post-quantum encryption module, including: SM2 / SM4 national encryption algorithm hardware accelerator for data signature and encrypted transmission; CRYSTALS-Kyber post-quantum key agreement module realizes NISTPQC standard anti-quantum attack protection through ASIC chip.

5. A micro data transactioner according to claim 1, wherein, The quality assessment unit verifies data quality through the following steps: Hash check: Calculate the BLAKE3 hash value of the data stream and compare it with the hash value recorded in the evidence platform; Timestamp deviation detection: verify the deviation between data generation timestamp and standard time of time service center ≤0.5 seconds.

6. A micro data transactioner according to claim 1, wherein, The scene suitability evaluation unit parses the metadata label using a natural language processing model, including: The BERT semantic matching engine generates a vector representation of the data description text and calculates the cosine similarity with the buyer's scene label. The authorized chain verification submodule queries the buyer's data usage rights record for the target scene through the blockchain.

7. A micro data transactioner according to claim 1, wherein, The hardware accelerator includes: The scene evaluation accelerator integrates a low-precision quantized AI inference engine and supports INT8 data format parallel computing for real-time processing of semantic matching tasks for data scene labels. The security compliance accelerator is based on an ASIC chip design and supports the following functions: Hardware-level acceleration of national encryption algorithms, including key generation, data signing, and encryption; Key negotiation and packaging of the NIST post-quantum cryptography algorithm CRYSTALS-Kyber; Parallel computing of the BLAKE3 hash algorithm to optimize throughput through the SIMD instruction set; Hardware interface: direct connection with the edge computing unit through the PCIe4.0 bus to implement the following collaboration mechanisms: Offload hash computation and signature tasks in compliance detection to the security compliance accelerator; Offload NLP inference tasks for scene suitability evaluation to the scene evaluation accelerator; Dynamic power management unit: use voltage frequency regulation technology to dynamically switch between the following modes based on task load: High power consumption mode: enable all acceleration modules, voltage 1.2V, frequency 200MHz; Low power consumption mode: only enable the SM4 encryption module, voltage 0.8V, frequency 50MHz.

8. An intelligent trading system of a micro data trader according to any one of claims 1 to 7, characterized in that, Multiple micro data transactioners and the following collaborative components: National-level computing power network nodes for distributed transaction data storage and verification; Cross-chain routing center: dynamically selects the optimal blockchain path through a reinforcement learning model; Edge-cloud collaborative management platform: monitors device status and dynamically updates the compliance rule library.

Images