Data processing system and processing method of hybrid cloud in wireless communication system

Abstract

The invention discloses a data processing system and a data processing method of a hybrid cloud in a wireless communication system. The method comprises the steps that wireless communication equipment receives and sends data through a multi-mode communication interface, and a hybrid cloud platform receives the data and distributes the data to a public cloud or a private cloud for processing and storage according to a preset strategy. And then, the data processing module performs real-time cleaning, decoding and analysis to generate a processing result. Based on a deep reinforcement learning algorithm, the network optimization unit optimizes and adjusts communication parameters. And the processing result and the optimization instruction are fed back to the wireless communication equipment through the feedback channel, so that the performance and efficiency of the equipment in the network are improved. The invention further shows the characteristics of guaranteeing data security, a data cleaning mechanism, optimizing strategies, improving the resource utilization rate and the like, and the system efficiency and the data processing efficiency are enhanced.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication, and more specifically, relates to a data processing system and method of a hybrid cloud in a wireless communication system. Background Art

[0002] With the rapid development of information technology, wireless communication systems are increasingly widely used in modern society, covering multiple fields such as mobile communication, Internet of Things (IoT), smart home, and smart city. Wireless communication systems need to process a large amount of data traffic, ensuring real-time data transmission and efficient processing to meet users' requirements for high bandwidth, low latency, and high reliability. However, traditional wireless communication data processing methods often rely on centralized cloud computing architectures and face several challenges and limitations.

[0003] Firstly, there are resource bottlenecks in centralized cloud computing for data processing and storage. With the rapid increase in the number of wireless devices and the accelerating data generation speed, a single public cloud platform is difficult to efficiently handle the real-time processing requirements of a large amount of data. This not only leads to an increase in data processing latency but also may affect the overall performance and user experience of the communication system. In addition, the centralized architecture lacks flexibility in dealing with sudden traffic peaks and is prone to resource overload and service interruption.

[0004] Secondly, data security and privacy protection have become important issues faced by wireless communication systems. Wireless communication involves a large amount of sensitive information and privacy data of users. During the data transmission and storage processes in a centralized cloud platform, it is easily targeted by network attacks and data leaks. Traditional security protection measures are often difficult to cope with complex and ever-changing network threats. Especially in a multi-tenant environment, the data isolation and access control mechanisms of public cloud platforms have certain deficiencies and cannot fully guarantee the security and privacy of user data.

[0005] In addition, with the development of edge computing and fog computing technologies, distributed data processing has gradually become an important means to improve the performance of wireless communication systems. Edge computing can significantly reduce data transmission latency and improve the system's response speed by sinking data processing tasks to network edge nodes. However, relying solely on edge computing also has problems such as limited resources, insufficient computing power, and complex management, and it is difficult to meet the large-scale data processing requirements in complex application scenarios. Therefore, how to effectively combine the advantages of cloud computing and edge computing to build a flexible, efficient, and secure hybrid cloud architecture has become a hot and difficult issue in current research.

[0006] In this context, the hybrid cloud, as an architecture that combines the advantages of public clouds and private clouds, has gradually attracted attention. The hybrid cloud can intelligently allocate data to public and private clouds for processing and storage according to different data types and processing requirements, making full use of the elasticity and large-scale computing power of public clouds while ensuring data security and privacy protection in private clouds. In addition, the hybrid cloud architecture can also achieve efficient management of data traffic through dynamic resource scheduling and optimization, improving the overall performance and scalability of the system.

[0007] Currently, although the hybrid cloud has achieved certain application results in enterprise-level applications, its specific application research in wireless communication systems is still relatively limited. Existing solutions mostly focus on the basic management and scheduling of cloud resources, lacking in-depth optimization methods for the characteristics of wireless communication data. For example, how to dynamically adjust the data allocation strategy between public and private clouds according to the real-time requirements of wireless communication; how to optimize network parameters using advanced machine learning algorithms to improve communication efficiency; how to ensure the transparency and immutability of the data processing process through blockchain technology are all urgent problems to be solved.

[0008] In addition, the multi-modal data processing requirements in wireless communication systems also pose higher demands on the hybrid cloud architecture. Wireless communication devices usually have multiple communication interfaces, such as cellular communication, short-range wireless transmission, and satellite positioning. The data generated by these interfaces is diverse in type and complex in processing requirements. Traditional data processing methods are difficult to efficiently integrate and process this multi-source heterogeneous data, resulting in low data processing efficiency and affecting the overall performance and user experience of the system. Summary of the Invention

[0009] The present invention proposes a data processing method in a wireless communication system based on a hybrid cloud, aiming to optimize the data processing flow of the wireless communication system and improve the performance and security of the system through intelligent data allocation and processing strategies. This method comprehensively utilizes the advantages of public and private clouds, combines advanced data processing technologies and optimization algorithms, and realizes the efficient, real-time, and secure processing of massive data in wireless communication systems, with broad application prospects and significant technical advantages.

[0010] To achieve the above objectives, the present invention is implemented by the following technical solutions: The described processing method includes: The wireless communication device receives and sends data through a multi-modal communication interface, which includes a cellular communication module, a short-range wireless transmission unit, and a satellite positioning module. The communication interface is connected to the baseband processing chip through a PCIe 4.0 bus to achieve multi-mode parallel transmission; The hybrid cloud platform receives data from the wireless communication device and allocates the data to a public cloud or a private cloud for processing and storage according to a preset strategy; The data processing module performs real-time cleaning, decoding, and analysis on the data allocated to the public cloud or private cloud, generating a processing result; The network optimization unit optimizes and adjusts the communication parameters of the wireless communication device based on the deep reinforcement learning algorithm; The processing result and optimization instructions are fed back to the wireless communication device through a multimodal feedback channel to enhance the performance and efficiency of the device in the network.

[0011] In one solution, the data processing unit of the wireless communication device includes a digital signal processor for data preprocessing, and the digital signal processor is used to perform format standardization, dynamic compression, and encryption processing on the received data.

[0012] In one solution, the hybrid cloud platform is based on a zero-trust architecture and adopts dynamic identity authentication technology through a secure tunnel interface. The dynamic identity authentication technology includes the HMAC-SHA3-512 algorithm to ensure the security of data transmission.

[0013] In one solution, the data processing module adopts a multi-level data cleaning mechanism, combining a dual-modal architecture of rule-based cleaning and deep anomaly detection to remove invalid or incorrect data.

[0014] In one solution, the network optimization unit adopts a federated learning architecture, realizes distributed optimization of policies through edge computing nodes, and ensures the privacy of device data during the optimization process.

[0015] In one solution, the feedback mechanism includes recording all optimization operations in a blockchain ledger, and the blockchain ledger adopts the PBFT consensus algorithm to ensure the auditability of operations and the non-repudiation of data.

[0016] In one solution, the hybrid cloud platform enables a federated learning-driven emergency policy library, extracts the optimal parameter combination based on the historical operation records of the device group, and performs hot deployment after policy preprocessing through edge computing nodes.

[0017] In one solution, the hybrid cloud platform uses an improved quantum annealing algorithm for data flow routing optimization, and adjusts the allocation of data between public cloud and private cloud nodes in real time to achieve multi-objective optimization and improve resource utilization.

[0018] In one solution, the wireless communication device further includes an intelligent cache management module, and the intelligent cache management module dynamically adjusts the cache size using an improved LRU-K algorithm to optimize the storage efficiency of real-time communication data, non-real-time data, and metadata.

[0019] A data processing system and method in a wireless communication system using a hybrid cloud according to claim 1, characterized in that: the data processing module further includes a real-time fault prediction unit, which predicts the health status of the device based on a Bayesian spatio-temporal network and triggers an early warning and maintenance mechanism when a potential fault is detected. Advantages of the present invention: The data processing method for a wireless communication system based on a hybrid cloud provided by the present invention significantly improves the performance of the wireless communication system in terms of data processing efficiency, system performance, and security through an intelligent data allocation and processing strategy, as specifically reflected in the following aspects: The present invention adopts a hybrid cloud architecture, organically combines the advantages of public clouds and private clouds, and optimizes the data processing process. The high elasticity and large-scale computing power of public clouds can efficiently handle the real-time processing requirements of massive data, while private clouds ensure the security and privacy of sensitive data. Through dynamic resource scheduling and intelligent data allocation, the system can flexibly select the most suitable cloud resources for processing according to different data types and processing requirements, thereby significantly improving the overall efficiency of data processing and the response speed of the system.

[0020] The present invention has significant advantages in data security and privacy protection. The hybrid cloud platform is based on a zero-trust architecture, adopts dynamic identity authentication technology through a secure tunnel interface, and combines the HMAC-SHA3-512 algorithm to ensure the security and integrity of data transmission. At the same time, blockchain technology is used in the feedback mechanism to record all optimization operations, and the PBFT consensus algorithm is used to ensure the auditability of operations and the non-repudiation of data, effectively preventing data tampering and unauthorized access, and greatly enhancing the security and reliability of the system.

[0021] When the present invention is faced with drastic fluctuations in network conditions, it can enable an emergency policy library driven by federated learning, extract the optimal parameter combination based on the historical operation records of the device group, and perform hot deployment after preprocessing the policy through edge computing nodes. This mechanism improves the adaptability and recovery speed of the system in extreme network environments, ensuring the high availability of wireless communication services and the continuous optimization of the user experience.

[0022] In summary, the present invention significantly improves the data processing efficiency, system performance, and security of wireless communication systems through an innovative combination of a hybrid cloud architecture, advanced optimization algorithms, and multi-level data security mechanisms, and has broad application prospects and significant technical advantages. Description of the Drawings

[0023] Figure 1 It is a flowchart of the method of the present invention; Figure 2 It is a block diagram of the system of the present invention. Detailed implementation manners

[0024] For the convenience of understanding the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Typical embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive.

[0025] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as understood by those skilled in the technical field to which the present invention belongs. The terms used in the description of the present invention in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. For the convenience of understanding the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Typical embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive A data processing system and method in a wireless communication system for a hybrid cloud. With the continuous development of wireless communication technology and cloud computing technology, building an efficient and secure hybrid cloud data processing architecture has become the key to improving the performance of wireless communication systems. Based on the existing technology, the present invention proposes an innovative hybrid cloud data processing method for the data processing requirements of wireless communication systems, which has important theoretical value and application significance.

[0026] The system includes one or more wireless communication devices, a hybrid cloud platform, and a data processing module. The data sources may be various types of wireless transmissions, including communications between wireless devices, communications between wireless devices and base stations, etc.

[0027] The hybrid cloud platform is between the public cloud and the private cloud, can receive data from wireless communication devices, and allocate the data to the public cloud and the private cloud for processing and storage according to preset policies. We can perform real-time cleaning, decoding, and analysis on the data content and traffic of wireless communications.

[0028] The data processing module is connected to the hybrid cloud platform, responsible for processing and analyzing the data received from wireless communication devices, and feeding back the results to the wireless communication devices. This can provide important data support to optimize the performance and efficiency of wireless communication devices in the network, and can help improve the overall performance of the wireless communication network.

[0029] In addition, according to actual needs, the system may also include a user interface for displaying processing results, and a security module for protecting the security of the system and data.

[0030] Overall, the hybrid-cloud-based wireless communication data processing system and method greatly improve the efficiency and security of wireless communication data processing, while making the processing more flexible and convenient.

[0031] As Figure 1 and Figure 2 shown, S1: The wireless communication device receives and sends data through a multimodal communication interface, which includes a cellular communication module, a short-range wireless transmission unit, and a satellite positioning module. The communication interface is connected to the baseband processing chip through a PCIe 4.0 bus to achieve multi-mode parallel transmission.

[0032] The data transceiver implementation process of a wireless communication device includes a multi-modal communication interface, a data processing unit, and an intelligent cache management module. The device body integrates a cellular communication module (supporting 5G NR, LTE-A Pro), a short-range wireless transmission unit (Wi-Fi 6E / 802.11ax, Bluetooth 5.3), and a satellite positioning module (GNSS). Each communication interface is connected to the baseband processing chip through a PCIe 4.0 bus to achieve multi-mode parallel transmission capabilities. The data acquisition end includes a real-time communication data stream interface and a local storage medium access interface. Among them, the real-time communication data source is the wireless signal received by the built-in millimeter-wave phased array antenna array of the device. After being processed by a low-noise amplifier (LNA) and an analog-to-digital converter (ADC), it forms a digital baseband signal; the stored data accesses the local 3D NAND flash memory array of the device through the NVMe protocol, and the erasure code (EC) technology is used to ensure data integrity. The data processing unit is equipped with a dedicated DSP coprocessor and uses a two-stage pipeline architecture to implement data preprocessing: the first-stage pipeline performs data format standardization, uniformly encapsulating data from different sources into standard data frames containing timestamps, source addresses, and QoS levels; the second-stage pipeline implements dynamic compression and encryption, automatically selecting the LZ4 or Zstandard compression algorithm according to the data type, and implementing per-packet encryption using a hybrid key system based on elliptic curve cryptography (ECC). The intelligent cache management module dynamically adjusts the cache size through an improved LRU-K algorithm, and divides three cache partitions with different priorities in the device memory: the real-time communication data uses DDR5 memory to build a circular buffer to ensure low-latency transmission; the non-real-time data is stored in the 3D XPoint persistent memory area to support burst traffic peak shaving; the metadata cache uses SRAM to achieve nanosecond-level access. The device working mode controller dynamically selects the optimal transmission path according to the wireless channel quality indicator (CQI), received signal strength (RSSI), and base station load status obtained by the network status perception module: when detecting cellular network congestion, it automatically enables the multi-path TCP (MPTCP) protocol to transmit critical data in parallel through Wi-Fi and cellular links; in weak signal scenarios, it starts the store-and-forward mode, temporarily storing the data in the local encrypted storage area waiting for the network to recover. The data integrity verification unit performs a three-level verification mechanism before sending: first, it performs a basic verification using CRC-32, then generates a data fingerprint through SHA-256, and finally uses forward error correction coding based on Reed-Solomon to enhance transmission reliability. The built-in QoS management engine of the device classifies service flows according to the 3GPP TS 23.203 standard, assigns the highest priority to URLLC-class data, and implements a transmission preemption mechanism through a dedicated hardware queue to ensure that the end-to-end delay is less than 1 ms. The entire transceiver process supports adaptive switching between TDD and FDD duplex modes, and is equipped with an intelligent impedance matching circuit to keep the antenna array at the optimal voltage standing wave ratio (VSWR < 1.5) at different frequency bands.After the device completes data encapsulation, the transmission protocol stack verified by the secure boot chain sends the data packet to the hybrid cloud platform, and at the same time retains the digitally signed timestamp log locally to form a complete data traceability chain.

[0033] S2. The hybrid cloud platform receives the data from the wireless communication device and distributes the data to the public cloud or the private cloud for processing and storage according to the preset policy.

[0034] After the data is transmitted to the hybrid cloud platform, it first undergoes access verification through the secure tunnel interface module. This module implements dynamic identity authentication based on the zero-trust architecture, uses the improved HMAC-SHA3-512 algorithm to generate the data packet fingerprint, and the verified data stream enters the multi-protocol parsing engine. The parsing engine extracts the data feature vector through deep packet inspection technology , where represents the data sensitivity score. After the pre-trained BERT-NER model identifies the entity types in the data, according to the formula: ; is calculated, represents the preset entity sensitivity weight, is the frequency of occurrence of the corresponding entity; represents the real-time requirement index. According to the QoS label in the data packet header and the service level agreement (SLA) parameters, through the formula: ; dynamically calculates the timeliness coefficient, where is the maximum allowable delay, is the lower limit of the physical transmission delay, is the business criticality attenuation factor. The feature vector is synchronously input into the load perception unit of the intelligent decision engine, which obtains the resource status matrix of the public cloud and the private cloud in real time through the cloud platform API , and the matrix element represents the availability of the i-th type of resource at the j-th node, where is the current utilization rate, is the resource elasticity coefficient, is the resource reallocation delay. The decision engine constructs a multi-objective optimization function: ; where is the decision variable vector, is the security weight matrix, is the processing delay matrix, , is the weighting coefficient and satisfies , is the safety threshold of resource utilization. This optimization problem is solved by an improved quantum annealing algorithm, and the Hamiltonian is constructed as follows: ; where the coupling strength , and the bias field is linearly related to the current cloud load status. After obtaining the optimal decision , the data stream is dynamically routed to the target cloud node, and the routing policy is implemented by the software-defined network (SDN) controller to issue a flow table.

[0035] During the decision-making process, an elastic compensation mechanism runs synchronously. When it is detected that the deviation between the actual processing delay and the expected value exceeds the threshold , a dynamic adjustment based on Lyapunov optimization is triggered: ; where represents the queue backlog of the i-th type of task, is the drift boundary constant, and represent the task arrival rate and processing rate respectively. This mechanism realizes the real-time adaptation of resource supply by adjusting the step size of virtual machine scaling, where is the learning rate parameter. The event logs generated during the entire decision-making process are stored by the blockchain smart contract, and an improved Practical Byzantine Fault Tolerance (PBFT) consensus algorithm is used to ensure the traceability of the decision. Its consensus verification function is defined as: ; where is the maximum number of fault-tolerant nodes, is the digital signature of the participating nodes. The final data allocation decision and resource scheduling strategy are encapsulated as metadata tags and stored together with the data block in the distributed object storage system to form a complete data life cycle management chain.

[0036] S3. The data processing module performs real-time cleaning, decoding, and analysis on the data allocated to the public cloud or private cloud to generate processing results.

[0037] After the data is transmitted to the target cloud node, the data processing module starts a multi-level pipeline operation. First, heterogeneous data purification is implemented through a distributed data cleaning engine. This engine adopts a dual-mode architecture of rule-driven and model-driven: the rule layer builds a regular expression verification library based on the RFC specification to perform format compliance checks on structured data fields; the model layer deploys a pre-trained deep anomaly detection network (DAN). The input layer of this network receives the feature vector composed of device status parameters, where is the bit error rate (BER), is the variance of radio frequency power fluctuation. A time series feature extractor is constructed through a three-layer gated recurrent unit (GRU), and an anomaly probability value is output , where is the GRU hidden state vector, is the Sigmoid activation function. When , the data discard mechanism is triggered ( is the dynamic threshold, which is adaptively adjusted by kernel density estimation according to the historical data distribution). The cleaned data enters the multi-protocol decoding cluster, which adopts a microservices architecture to achieve parallel processing: for binary stream data, a deserialization template generated by the Protocol Buffers compiler is used for structured conversion; for encrypted data packets, the decryption key bound to the device hardware fingerprint is obtained through the key management center, and 256-bit parallel AES-GCM decryption is implemented on the FPGA acceleration card; the multi-modal data fusion unit uses the dynamic time warping (DTW) algorithm to align the timestamps of data from different sources, satisfying the formula: , where is the time axis mapping function, achieving nanosecond-level synchronization accuracy.

[0038] The real-time fault prediction engine constructs a spatio-temporal joint prediction model based on the device operation data stream. Its core algorithm consists of an improved Bayesian spatio-temporal network (BSTN): Define the device health state as a binary latent variable (0 represents normal, 1 represents fault), and the observation variable set includes radio frequency parameters, baseband metrics, and environmental sensor data. The posterior probability is solved through variational inference: ; where the state transition probability is obtained by statistical analysis of the device historical maintenance records, and the observation likelihood function is fitted using a Gaussian mixture model. The model online learning module uses stochastic gradient variational inference (SGVB) to achieve real-time parameter update, and the update rule is: ; where is the variational parameter, is the adaptive learning rate. The prediction engine outputs a fault risk index: ; where is the weight coefficient, characterizes the parameter mutation intensity. When exceeds the threshold , the early warning system issues a level 4 response mechanism: 1) Send pre-maintenance instructions to the device through the control plane interface; 2) Dynamically adjust the device transmit power ; 3) Reserve the execution status image of the hot backup resources in the cloud; 4) Generate an immutable maintenance work order based on the blockchain.

[0039] During the data processing, a fault tolerance compensation mechanism runs synchronously. The triple redundant computing pipeline is implemented using N-version programming design: the main channel executes the regular cleaning and decoding process, the secondary channel runs in the degraded mode (using a low-dimensional feature space to simplify the calculation), and the emergency channel retains the original data snapshot. The outputs of the three channels reach a consensus through the Byzantine protocol: ; This mechanism ensures the continuity of data processing in case of a single point of failure. All intermediate results are stored in the time series database cluster, and an improved TSBS compression encoding is used to reduce the storage overhead. The compression algorithm satisfies: ; where is the coding length, is the information entropy. The finally processed data stream is injected into the distributed message queue for consumption by the upper-layer business system. At the same time, the generated data quality report is sent back to the device side through a secure channel to form a closed-loop optimization loop.

[0040] S4. The network optimization unit optimizes and adjusts the communication parameters of the wireless communication device based on the deep reinforcement learning algorithm.

[0041] After completing the data processing and fault prediction, the system activates the network optimization unit based on the Multi-Agent Deep Deterministic Policy Gradient (MADDPG). This engine constructs a distributed cooperative control framework and models each wireless communication device as an independent agent. The agent's observation space contains multi-dimensional environmental states: where is the instantaneous signal-to-noise ratio, is the remaining useful life estimate value provided by the fault prediction engine in the S3 stage, Δf t-1 is the frequency offset in the previous time slot. The action space A i = [P min , P max × [f min , f max is defined as the continuous power-frequency joint regulation domain. An optimization strategy π θ : O i → A i is generated through the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm, and its objective function is: where the immediate reward function comprehensively considers the spectral efficiency Energy consumption and interference coordination index γ is the discount factor, D KL terms are used to constrain the magnitude of policy updates. The network parameter updates adopt the centralized training and distributed execution (CTDE) paradigm, and the critic network Q φ receives the global state Its Bellman equation is expressed as: where `o~clip(N(0,σ),-c,c) is the exploration noise, and the target network parameters φ′ and θ′ are updated softly as φ′←τφ+(1-τ)φ′. During actual deployment, a lightweight policy network is embedded at the device side which extracts features from the complex network π in the cloud through knowledge distillation technology θ to satisfy: The dynamic spectrum allocation module adopts an improved federated learning architecture, and each base station periodically uploads the local policy gradient to the edge aggregator, and the gradient fusion is realized through secure multi-party computation: where the weight is proportional to the number of user devices served, and ξ is the gradient clipping threshold. For the burst interference scenario, the system introduces a fast response mechanism based on Lyapunov optimization, and defines the virtual queue Q(t + 1) = max[Q(t)+a(t)-b(t),0], where a(t) is the interference intensity measurement value and b(t) is the amount of coordinable resources. By minimizing the drift-plus-penalty term: min P,f E[Q(t)(a(t)-b(t))]+V·E[EC(t)] the sub-second adjustment of power-frequency parameters is realized (V is the control parameter). The optimization result is configured for the device physical layer in real time through the software-defined radio (SDR) interface, and the pre-distortion compensation algorithm is adopted to eliminate the non-linear effect: where the pre-distortion coefficient a k ,φ k ,θ k are dynamically generated by the deep neural network. All optimization operations are recorded in the blockchain ledger, and the threshold signature scheme is adopted to ensure the auditability of the operations. The signature verification satisfies: Finally, a closed-loop network optimization system is formed to realize the collaborative enhancement of device-level parameter adjustment and network-level performance optimization.

[0042] S5. Feed the processing result and the optimization instruction back to the wireless communication device through a multi-modal feedback channel to improve the performance and efficiency of the device in the network.

[0043] After calculating the network optimization parameters, the system pushes the optimization instruction and the analysis result to the target wireless communication device through a multi-modal feedback channel. This process is implemented using a priority-based hierarchical transmission protocol: high-priority control instructions (such as emergency power adjustment and frequency band switching) are transmitted through a dedicated low-latency channel and encapsulated as adaptive protocol data units (APDUs) compliant with the IEEE 1905.1 standard. The data payload part uses the LZ4 real-time compression algorithm to reduce the transmission overhead; the general-priority parameter configuration information is sent through the device management plane channel and encrypted end-to-end using the TLS 1.3 session key pre-set during device registration. After receiving the instruction, the intelligent agent execution engine deployed on the device side first verifies the legality of the instruction through an embedded security sandbox. This sandbox integrates a runtime integrity verification module, which ensures that the received configuration parameters are consistent with the hash value of the cloud signature certificate through hash tree comparison to prevent man-in-the-middle attacks. The verified instruction enters the dynamic parameter mapping stage, where the policy converter in the device firmware converts the abstract optimization instruction into a control signal recognizable by the underlying hardware driver. This conversion process relies on the parameter mapping table provided by the device manufacturer and supports dynamic matching based on the device hardware fingerprint. For example, it automatically selects the corresponding power-frequency curve calibration parameters for different models of radio frequency front-end modules. The actual parameter adjustment adopts a progressive incremental update strategy, and the PID controller is used to smoothly transition the device state to avoid communication interruption caused by parameter mutation. Specifically, it performs a rate limit on the transmit power adjustment amount ΔP , and the frequency switching executes a phase synchronization compensation algorithm to maintain carrier continuity. After the adjustment is completed, the device actively reports the status change confirmation information through the control plane and simultaneously triggers a local closed-loop verification mechanism - real-time simulating the network performance indicators under the current parameters through the digital twin system. If the deviation between the simulation result and the cloud prediction value exceeds the preset threshold (such as throughput difference > 15%), it will automatically roll back to the previous stable configuration version. All feedback operations generate a non-repudiable evidence chain, using the digital signature technology based on the national cryptography SM2 algorithm to store the operation log on the blockchain, and constructing a three-dimensional traceability matrix through the device unique identifier and the geospatial information. For extreme scenarios with severe fluctuations in network conditions, the system enables an emergency policy library driven by federated learning, extracts the optimal parameter combination in similar scenarios from the historical operation records of the device group, and performs hot deployment after preprocessing the policy through the edge computing node. The finally formed device behavior data, network optimization effect, and abnormal event records are synchronized to the cloud analysis system through the data return pipeline, used to update the fault prediction model parameters in the S3 stage and the deep reinforcement learning policy library in the S4 stage, and complete the global closed-loop optimization from decision generation to effect verification.

[0044] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), or the like.

[0045] It should be understood that the detailed description of the technical solutions of the present invention with the aid of the preferred embodiments is illustrative rather than restrictive. Those of ordinary skill in the art can modify the technical solutions described in each embodiment on the basis of reading the specification of the present invention, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data processing system and method for a hybrid cloud in a wireless communication system, characterized in that: The processing method comprises: The wireless communication device receives and sends data through a multimodal communication interface, which includes a cellular communication module, a short-range wireless transmission unit, and a satellite positioning module. The multimodal communication interface is connected to a baseband processing chip through a PCIe 4.0 bus to achieve multi-standard parallel transmission. The hybrid cloud platform receives data from the wireless communication device and distributes the data to a public cloud or a private cloud for processing and storage according to a preset strategy; The data processing module cleans, decodes and analyzes the data allocated to the public cloud or private cloud in real time to generate processing results; The network optimization unit optimizes and adjusts the communication parameters of the wireless communication device based on the deep reinforcement learning algorithm; The processing results and optimization instructions are fed back to the wireless communication device through a multimodal feedback channel to improve the performance and efficiency of the device in the network.

2. A hybrid cloud data processing system and method in a wireless communication system according to claim 1, characterized in that: The data processing unit of the wireless communication device includes a digital signal processor for data preprocessing, and the digital signal processor is used to perform format standardization, dynamic compression and encryption processing on the received data.

3. The data processing system and method of a hybrid cloud in a wireless communication system according to claim 1, characterized in that: The hybrid cloud platform is based on a zero-trust architecture and adopts dynamic identity authentication technology through a secure tunnel interface. The dynamic identity authentication technology includes the HMAC-SHA3-512 algorithm to ensure the security of data transmission.

4. The data processing system and method of a hybrid cloud in a wireless communication system according to claim 1, characterized in that: The data processing module adopts a multi-level data cleaning mechanism, combining a bimodal architecture of rule-based cleaning and deep anomaly detection to remove invalid or erroneous data.

5. The data processing system and method of a hybrid cloud in a wireless communication system according to claim 1, characterized in that: The network optimization unit adopts a federated learning architecture to achieve distributed optimization of strategies through edge computing nodes and ensure the privacy of device data during the optimization process.

6. A hybrid cloud data processing system and method in a wireless communication system according to claim 1, characterized in that: The feedback mechanism of the multimodal feedback channel includes: recording all optimization operations in a blockchain ledger, and the blockchain ledger adopts a PBFT consensus algorithm to ensure the auditability of operations and the non-repudiation of data.

7. A hybrid cloud data processing system and method in a wireless communication system according to claim 1, characterized in that: The hybrid cloud platform enables a federated learning-driven emergency policy library, extracts the optimal parameter combination based on the historical operation records of the device group, and implements hot deployment after policy preprocessing through edge computing nodes.

8. The data processing system and method of a hybrid cloud in a wireless communication system according to claim 1, characterized in that: The hybrid cloud platform uses an improved quantum annealing algorithm to optimize data flow routing and adjusts the distribution of data between public cloud and private cloud nodes in real time to achieve multi-objective optimization and improve resource utilization.

9. A hybrid cloud data processing system and method in a wireless communication system according to claim 1, characterized in that: The wireless communication device also includes an intelligent cache management module, which uses an improved LRU-K algorithm to dynamically adjust the cache size to optimize the storage efficiency of real-time communication data, non-real-time data and metadata.

10. The data processing system and method of a hybrid cloud in a wireless communication system according to claim 1, characterized in that: The data processing module further includes a real-time fault prediction unit, which predicts the health status of the equipment based on the Bayesian spatiotemporal network and triggers an early warning and maintenance mechanism when a potential fault is detected.

Images