Real-time fragmented transmission method of heterogeneous encrypted multimodal multimedia messages for 5G networks

Abstract

The present invention discloses a real-time sharding transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks, which relates to the technical field of multimedia transmission optimization. This real-time sharding transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks obtains the sensitivity label and value density of multimedia data, combines the 5G slice resource status, and optimizes the sharding granularity and priority through a reinforcement learning algorithm. The encryption algorithm cluster is dynamically selected according to the sharding rules, and the alliance chain is used to realize cross-slice key distribution and update. Based on the spatiotemporal correlation of multimodal data, a fountain code-driven redundant coding strategy is constructed, and the sharding transmission path is optimized in combination with a game theory model. Finally, the transmission traffic characteristics and key distribution logs are analyzed, and the encryption strength threshold and sharding retransmission strategy are dynamically adjusted. This method improves the security, reliability and transmission efficiency of multimedia data in 5G networks, and is suitable for low-latency, high-reliability multimedia communication scenarios.

Description

Technical Field

[0001] The present invention relates to the technical field of multimedia transmission optimization, and specifically to a real-time fragmented transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks. Background Art

[0002] With the commercialization of 5G networks, the demand for high-bandwidth, low-latency transmission for multimodal multimedia applications is surging. At the same time, data security threats are becoming increasingly complex, and the differentiated requirements for encryption strength and transmission reliability in different business scenarios have significantly increased. Traditional single encryption and static slicing mechanisms are difficult to adapt to the heterogeneous resource characteristics and dynamic business requirements of 5G network slicing. There is an urgent need to build an intelligent slicing transmission system for multi-slice collaboration.

[0003] Existing methods mostly use fixed sharding granularity and static encryption strategies, and are unable to dynamically adjust sharding priority and encryption strength according to the network slice resource status, resulting in high-value data shards potentially causing transmission congestion or security vulnerabilities due to resource competition; traditional centralized key distribution mechanisms are difficult to support multi-slice collaboration scenarios, and have problems such as high key synchronization delay and low credibility of cross-domain distribution paths; existing redundancy strategies lack intelligent perception of the spatiotemporal correlation of shards, and fixed redundancy factors lead to bandwidth waste or insufficient error correction capabilities, and are unable to adapt to dynamic network packet loss rates; traditional routing algorithms only consider latency or bandwidth indicators, ignoring the multi-objective game relationship between security level, energy consumption cost and slice resource competition, making it difficult to achieve global optimal transmission. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a real-time fragmented transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks, which solves the problems of the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a real-time sharding transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks, comprising the following steps: S1. obtaining the sensitivity label and value density of multimedia data, and monitoring the resource status of 5G network slices in real time, dynamically generating sharding granularity and priority mapping rules through a reinforcement learning algorithm, extracting spatiotemporal key features in combination with a convolutional network, and outputting a sharding sequence adapted to 5G slice resources and a corresponding encryption strength label; S2. receiving the sharding sequence and encryption strength label, dynamically selecting an encryption algorithm cluster based on the sharding granularity and priority mapping rules, implementing cross-slice key distribution and update through a consortium chain, and generating encrypted sharding data blocks and key distribution logs; S3. constructing a fountain code-driven redundant encoding strategy based on the spatiotemporal correlation of the encrypted sharding data blocks and multimodal data, optimizing the sharding transmission path in combination with a game theory model, and outputting a mapping table of associated sharding groups and transmission paths; S4. analyzing the transmission traffic characteristics and key distribution logs based on the mapping table of associated sharding groups and transmission paths, and dynamically adjusting the encryption strength threshold and sharding retransmission strategy.

[0006] Furthermore, the specific process of obtaining the sensitivity label and value density of multimedia data and monitoring the resource status of 5G network slices in real time is as follows: based on the distribution density of sensitive words in text data annotated by natural language processing, the key frame motion trajectory characteristics of video data are extracted, the frequency domain burst frequency of audio data is analyzed, and the value density level of multimedia data is dynamically calculated in combination with the mutation amplitude of the sensor waveform; the remaining bandwidth of the uRLLC slice, the delay jitter of the eMBB slice, and the packet loss rate of the mMTC slice are collected in real time through the 5G core network interface to generate a 5G network resource state vector.

[0007] Furthermore, the specific process of dynamically generating sharding granularity and priority mapping rules through reinforcement learning algorithm is as follows: constructing a state space with multimedia data value density feature vector and 5G network slicing resource state vector as input, defining sharding granularity level and transmission priority weight as action space, designing a multi-objective reward function that comprehensively considers encryption strength matching, delay deviation and terminal energy consumption, and generating a dynamic mapping matrix between sharding granularity and priority through strategy iteration.

[0008] Furthermore, the convolutional network is combined to extract spatiotemporal key features, and the specific process of outputting a slice sequence adapted to 5G slice resources and the corresponding encryption strength label is as follows: the spatiotemporal features of the video stream are extracted through a lightweight convolutional network, the motion correlation between key frames and the sudden silence segments in the audio stream are identified, and the frequency domain mutation points of the sensor waveform are integrated to generate spatiotemporal key feature vectors; according to the multimedia data value density feature vector and the 5G slice resource status, the slice boundaries are dynamically divided and the encryption strength labels are marked, and a slice sequence adapted to the uRLLC, eMBB, and mMTC slice characteristics is output.

[0009] Furthermore, the specific process of dynamically selecting the encryption algorithm cluster according to the sharding granularity and priority mapping rules is as follows: construct a heterogeneous encryption algorithm pool, generate a multi-dimensional matching index according to the sharding priority weight and encryption strength label, and use the sliding time window to count the network jitter coefficient and computing load fluctuation characteristics of the historical sharding; according to the current slice resource remaining bandwidth and terminal hardware acceleration capabilities, dynamically combine symmetric encryption algorithms and lightweight asymmetric encryption algorithms to form an algorithm cluster with sharding granularity and encryption strength linkage, including: fine-grained sharding: binding attribute-based homomorphic encryption algorithm, multimodal data spatiotemporal The feature is to generate dynamic keys based on attribute parameters. The spatiotemporal features of multimodal data include the hash value of video key frames and the peak value of sensor waveforms. Medium-grained sharding: data blocks are cross-encrypted through a layered hybrid encryption algorithm, and keys are dynamically distributed through an edge node consensus protocol. Coarse-grained sharding: dynamic mask perturbation encryption is used, and keys are issued on demand through 5G control plane signaling to adapt to low-energy transmission requirements. According to the spatiotemporal key feature vectors of the sharding sequence, the key length adaptive adjustment mechanism and encryption mode switching events are triggered to achieve real-time coupling of algorithm cluster parameters and network slice QoS indicators.

[0010] Furthermore, cross-slice key distribution and update are achieved through the alliance chain. The specific process of generating encrypted shard data blocks and key distribution logs is as follows: the shard key is split into multiple sub-key components according to the slice domain, and the threshold signature rules and consensus verification conditions for cross-domain key distribution are defined; through the dynamic authorization channel between alliance chain nodes, the sub-key components are pushed to the edge computing nodes corresponding to the uRLLC, eMBB, and mMTC slices, and the key version number and distribution path topology are recorded in the tamper-proof block log; combined with the encryption timestamp and lifecycle label of the shard data block, periodic key update events are triggered, and the key copies of the heterogeneous blockchain network are synchronously updated using the cross-chain protocol.

[0011] Furthermore, the specific process of constructing a fountain code-driven redundant coding strategy is as follows: dynamically generate redundancy factors based on the encryption strength labels and spatiotemporal correlations of the sharded data blocks, and adjust the coding degree distribution function of the fountain code according to the spatiotemporal feature distribution density of the multimodal data; arrange the encrypted sharded data blocks and redundant check blocks in a two-dimensional matrix in the frequency domain and time domain, dynamically detect the shard group correlation strength through a sliding window and inject differentiated redundancy; combine the real-time packet loss rate of the slice and the transmission reliability index to establish a dynamic adaptation relationship between the redundant coding parameters and QoS requirements, and generate a multi-slice collaborative redundancy strategy.

[0012] Furthermore, the shard transmission path is optimized by combining the game theory model, and the specific process of outputting the mapping table of associated shard groups and transmission paths is as follows: the available bandwidth of the transmission node, the slice resource competition coefficient and the slice security level are defined as the strategy space, and a composite utility function including delay cost, security risk loss and energy consumption cost is designed; the Nash equilibrium point is solved by the evolutionary game algorithm to generate the initial matching scheme of the shard group and the transmission path, and the path selection weight is dynamically corrected according to the spatiotemporal key feature vector of the shard sequence; the credibility assessment mechanism is constructed in combination with the blockchain key distribution topology, and a dynamic mapping table of the shard group and slice resources is output.

[0013] Furthermore, according to the mapping table of associated shard groups and transmission paths, the transmission traffic characteristics and key distribution logs are analyzed, and the specific process of dynamically adjusting the encryption strength threshold and shard retransmission strategy is as follows: real-time monitoring of the packet loss rate, key distribution delay and blockchain consensus verification success rate of the shard transmission path, and construction of a joint assessment matrix of network status and security risks; a dynamic adjustment model of the encryption strength threshold is established based on the reinforcement learning algorithm, and the shard encryption mode switching event is triggered according to the risk level, and a shard retransmission priority queue is generated at the same time.

[0014] The specific process is as follows: construct a multi-objective optimization space including encryption strength gradient, end-to-end delay constraint and energy consumption threshold, define the dynamic approximation conditions of the Pareto front, solve the optimal solution set for security and efficiency, and dynamically adjust the objective function weight coefficient based on the fluctuation characteristics of network slice resources; according to the shard value density level, generate a hierarchical response instruction set, including encryption algorithm cluster reorganization instructions, shard path switching instructions and redundant coding reconstruction instructions.

[0015] The present invention has the following beneficial effects:

[0016] (1) A real-time fragmented transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks obtains the sensitivity labels and value density of multimedia data, monitors the resource status of 5G network slices in real time, and dynamically generates fragmentation granularity and priority mapping rules based on reinforcement learning algorithms. This method can flexibly adjust the fragmentation granularity and encryption strength of data according to changes in network resources. This dynamic algorithm-based adjustment can effectively ensure that highly sensitive and high-value data is transmitted first when the network load is high, reducing the data packet loss rate and improving transmission efficiency. At the same time, combined with the spatiotemporal key features extracted by the convolutional network, the generation of fragmentation sequences and encryption strength labels is more accurate, thereby improving data security and transmission reliability.

[0017] (2) A real-time fragmented transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks. Through a redundant coding strategy based on fountain codes and a fragmented transmission path design optimized by game theory, it can ensure the reliability and stability of multimodal data transmission in complex network environments. By combining a multi-objective optimization algorithm to dynamically adjust the encryption strength threshold and fragment retransmission strategy, it optimizes the balance between security and transmission efficiency during the transmission process. In the event of network fluctuations or congestion during transmission, the system can also dynamically adjust the strategy to ensure rapid data recovery and secure transmission, ultimately ensuring an efficient and stable data transmission experience.

[0018] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of the real-time fragmented transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks of the present invention.

[0020] Figure 2 This is a flowchart of the multimedia message transmission steps of the present invention.

[0021] Figure 3 This is the flow chart of step S1 of the present invention DETAILED DESCRIPTION

[0022] The embodiment of the present application solves the problems of low resource utilization efficiency, insufficient credibility of cross-slice key distribution, mismatch between redundant coding and network status, and imbalance between security and efficiency caused by single-dimensional optimization of transmission paths caused by the static coupling of sharding strategy and encryption mechanism in the existing technology through a real-time sharding transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks. Specifically, it includes: breaking through the traditional fixed sharding mode, dynamically generating sharding granularity and priority mapping rules through reinforcement learning, and realizing real-time adaptation of network slice resource status and multimodal data value density; solving the problem of cross-domain key management, building a key distribution mechanism combining threshold signatures and dynamic authorization channels based on the alliance chain, and ensuring the credibility and timeliness of key distribution in multi-slice scenarios; overcoming the rigid defects of redundant coding, and proposing a dynamic redundancy strategy of fountain codes driven by spatiotemporal correlation to realize intelligent matching of coding degree distribution function with network packet loss rate and shard priority; eliminating the limitation of single dimension of path optimization, building a multi-objective game model that integrates security level, delay cost and slice resource competition coefficient, and generating an optimal path mapping that takes into account both security and transmission efficiency.

[0023] The overall idea of the solution in the embodiments of this application is as follows:

[0024] Obtain the sensitivity labels and value density of multimedia data, and monitor the resource status of 5G network slices in real time. Dynamically generate slicing granularity and priority mapping rules through reinforcement learning algorithms, combine convolutional networks to extract key spatiotemporal features, and output slicing sequences adapted to 5G slice resources and corresponding encryption strength labels.

[0025] Receive the shard sequence and encryption strength label, dynamically select the encryption algorithm cluster according to the shard granularity and priority mapping rules, implement cross-shard key distribution and update through the alliance chain, and generate encrypted shard data blocks and key distribution logs.

[0026] According to the spatiotemporal correlation of encrypted shard data blocks and multimodal data, a fountain code-driven redundant encoding strategy is constructed. The shard transmission path is optimized in combination with the game theory model, and a mapping table of associated shard groups and transmission paths is output.

[0027] According to the mapping table of associated fragment groups and transmission paths, the transmission traffic characteristics and key distribution logs are analyzed to dynamically adjust the encryption strength threshold and fragment retransmission strategy.

[0028] See also Figure 1 and Figure 2 , an embodiment of the present invention provides a technical solution: a real-time sharding transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks, comprising the following steps: S1. obtaining the sensitivity label and value density of multimedia data, and monitoring the resource status of 5G network slices in real time, dynamically generating sharding granularity and priority mapping rules through a reinforcement learning algorithm, extracting spatiotemporal key features in combination with a convolutional network, and outputting a sharding sequence adapted to 5G slice resources and a corresponding encryption strength label; S2. receiving the sharding sequence and encryption strength label, dynamically selecting an encryption algorithm cluster according to the sharding granularity and priority mapping rules, realizing cross-slice key distribution and update through a consortium chain, and generating encrypted sharding data blocks and key distribution logs; S3. constructing a fountain code-driven redundant encoding strategy based on the spatiotemporal correlation of the encrypted sharding data blocks and multimodal data, optimizing the sharding transmission path in combination with a game theory model, and outputting a mapping table of associated sharding groups and transmission paths; S4. analyzing the transmission traffic characteristics and key distribution logs based on the mapping table of associated sharding groups and transmission paths, and dynamically adjusting the encryption strength threshold and sharding retransmission strategy.

[0029] In this implementation, step S1: The system first obtains the sensitivity label and value density of multimedia data. This is used to determine the data's security requirements and transmission priority. Specifically, the sensitivity label represents the importance of the data (such as private data or financial data), while the value density represents the value of the data, enabling prioritization during transmission. The system monitors the resource status of 5G network slices in real time to assess the current network resource load. This information is input into a reinforcement learning algorithm to dynamically generate a mapping rule between sharding granularity and priority. Combined with the temporal and spatial key features of the data extracted by a convolutional neural network, it further generates a sharding sequence and corresponding encryption strength label tailored to the current 5G network resource situation. This allows the system to optimize data sharding and encryption strategies based on different data characteristics and network resource conditions. Step S2: Select an encryption algorithm and perform cross-slice key distribution. In this step, after receiving the sharding sequence and its corresponding encryption strength label, the system dynamically selects an appropriate encryption algorithm cluster based on the sharding granularity and priority mapping rule generated in step 1. Here, an "encryption algorithm cluster" refers to a collection of multiple encryption algorithms, with the appropriate algorithm selected based on different data requirements. After selecting the encryption algorithm, the system uses consortium chain technology (a form of distributed ledger technology) to securely distribute and update keys across different 5G network slices. This ensures synchronized and secure key management and updates across slices, ensuring that data transmission within each slice is properly encrypted and protected. Finally, the system generates encrypted shard data blocks and records key distribution logs for subsequent traceability and management. Step S3: Constructing a redundant encoding strategy and optimizing transmission paths. This step constructs a redundant encoding strategy driven by fountain codes based on the spatiotemporal correlations between the encrypted shard data blocks and multimodal data. Fountain codes are an efficient encoding method that ensures that even if some data is lost during transmission, the original information can still be recovered from the remaining data. During transmission, the shard transmission path is optimized based on a game theory model. This model simulates the game between multiple transmission nodes to determine the optimal transmission path, maximizing data transmission success rate and minimizing latency. Finally, the system outputs a mapping table of associated shard groups and transmission paths, indicating the paths along which each shard data is transmitted within the network and the relationships between them. Step S4: Based on the mapping table of associated shard groups and transmission paths obtained in Step S3, the system analyzes the network's transmission traffic characteristics and key distribution logs. By monitoring transmission conditions in real time, it dynamically adjusts the encryption strength threshold and shard retransmission strategy. For example, if a network path is congested or has a high packet loss rate, the system will automatically reduce encryption strength or select a more reliable transmission path to ensure the smoothest possible data transmission. During this adjustment process, the system, incorporating a multi-objective optimization algorithm, balances security and transmission efficiency, generating risk response instructions to address varying network conditions.In this way, the system can effectively deal with unstable factors in the 5G network and ensure the security and efficiency of the data transmission process.

[0030] See also Figure 3 Specifically, the process of obtaining the sensitivity label and value density of multimedia data and monitoring the resource status of 5G network slices in real time is as follows: the distribution density of sensitive words in text data is annotated by natural language processing, the key frame motion trajectory characteristics of video data are extracted, the frequency domain burst frequency of audio data is analyzed, and the value density level of multimedia data is dynamically calculated in combination with the mutation amplitude of the sensor waveform; the remaining bandwidth of the uRLLC slice, the delay jitter of the eMBB slice, and the packet loss rate of the mMTC slice are collected in real time through the 5G core network interface to generate a 5G network resource state vector.

[0031] In this implementation, during multimedia data transmission, natural language processing (NLP) is used to annotate the distribution density of sensitive words in text data. Natural language processing (NLP) is used to identify sensitive words (e.g., ID card numbers) within text data (e.g., messages and documents). The distribution density of sensitive words reflects the privacy level of the text data. If a text contains a large number of sensitive words, its sensitivity label is assessed as high; otherwise, it is assessed as low. Keyframe motion trajectory feature extraction for video data: Video data is processed to extract keyframes and analyze their motion trajectories. Regions of a video with rapid movement or change (e.g., rapidly changing scenes or actions) may contain more sensitive information and therefore receive a higher sensitivity label. Motion trajectory features reflect the motion patterns of objects in the video and can help identify areas requiring special protection. Frequency domain burst frequency analysis for audio data: Frequency domain analysis is performed on audio data to identify burst frequencies (e.g., prominent sound peaks in the audio). If the audio contains sensitive content (e.g., specific speech signals or noise), the burst frequency will vary, affecting the generation of the sensitivity label. Analysis of the mutation amplitude of sensor waveforms: Perform waveform analysis on data collected by sensors (such as temperature sensors, motion sensors, etc.) to identify the mutation amplitude. A mutation in sensor data often means the occurrence of an important event (such as a security-related alarm or abnormal situation), so the sensitivity label of this data is higher. Dynamic calculation of the value density level of multimedia data: Dynamically calculate the "value density" of each type of data through comprehensive analysis of multiple data features (such as text sensitive words, video motion features, audio frequency domain features, sensor data, etc.). Value density indicates the value of the data. The higher the value density, the more important the data (such as critical business data, important decision support data, etc.). Dynamic calculation of value density level: ; Parameter description: : Sensitive word distribution density, : The intensity of the motion trajectory of adjacent key frames, : frequency domain burst frequency, : waveform mutation amplitude, , , , : Weight coefficient (must satisfy V: Value density level. This dynamic calculation can update the value level of multimedia data in real time based on changes in multimedia data. Real-time monitoring of 5G network slice resource status. 5G network slicing is a key mechanism in 5G technology that enables network resources to be divided according to different requirements (such as low latency and high bandwidth). When monitoring network resource status in real time, the system collects and analyzes 5G network status information through the following methods: Remaining bandwidth of uRLLC slices: uRLLC (Ultra-Reliable Low-Latency Communication) slices provide ultra-high reliability and low-latency communication services. The system monitors the remaining bandwidth of uRLLC slices in real time, which represents the available bandwidth capacity within the slice. Bandwidth is a key factor affecting transmission efficiency. The greater the remaining bandwidth, the stronger the network transmission capacity; conversely, the lower the bandwidth, the higher the network load. Latency jitter of eMBB slices: eMBB (Enhanced Mobile Broadband) slices support high-speed, high-capacity broadband communications, suitable for scenarios such as high-definition video and large-scale data transmission. Latency jitter refers to the variation in latency during data transmission. Real-time monitoring of latency and jitter can help assess network stability. Excessive latency and jitter can degrade transmission quality, impacting the user experience for applications such as video streaming and online gaming. Packet loss rate of mMTC slices: mMTC (Massive Machine Type Communications) slices are suitable for connecting large numbers of IoT devices, featuring high device density and low energy consumption. The packet loss rate refers to the proportion of data packets lost during data transmission. Monitoring the packet loss rate helps assess network reliability. A high packet loss rate indicates poor network quality, potentially impacting normal communication of IoT devices. Based on the above monitoring data, the system generates a 5G network resource state vector (NRSV). This vector includes key network performance indicators such as the remaining bandwidth of the uRLLC slice, the latency and jitter of the eMBB slice, and the packet loss rate of the mMTC slice. Using this data, the system can assess network status in real time and provide a basis for subsequent slice transmission decisions.

[0032] Specifically, the specific process of dynamically generating sharding granularity and priority mapping rules through reinforcement learning algorithm is as follows: construct a state space with multimedia data value density feature vector and 5G network slice resource state vector as input, define sharding granularity level and transmission priority weight as action space, design a multi-objective reward function that comprehensively considers encryption strength matching, delay deviation and terminal energy consumption, and generate a dynamic mapping matrix between sharding granularity and priority through strategy iteration.

[0033] In this implementation, the reinforcement learning algorithm dynamically generates detailed fragmentation rules and constructs the input vector in the state space: the multimedia data value density feature vector: ; : density of sensitive words in text; M: intensity of motion trajectory of key frames in video; B: burst frequency in audio and video domain; A: amplitude of sensor waveform mutation (physical unit). 5G network slice resource state vector: : uRLCC slice remaining bandwidth (Mbps); :eMBB slice delay jitter; :mMTC slice packet loss rate (percentage). State space: Operator ⊕: represents vector concatenation, forming a 7-dimensional state vector. Action space definition, slicing granularity level (discrete action): fine granularity: slicing duration ; Medium particle size: ; Coarse-grained: Transmission priority weight (continuous action): High priority: Weight ; Medium priority: weight ; Low priority: weight Action space: Z={{(T,w)}}, ; Multi-objective reward function design, sub-objective quantification: encryption strength matching degree: ; : Actual encryption strength (e.g., 1.0 for AES-256 and 0.8 for ECC); : Encryption strength required by the sensitivity label (high sensitivity = 1.0, medium sensitivity = 0.8, low sensitivity = 0.5). Delay deviation: ; : actual transmission delay (ms); : Target latency (10ms for uRLLC slicing, 50ms for eMBB, and 100ms for mMTC). Terminal energy consumption: ; : Terminal encryption and transmission energy consumption; : Maximum allowed energy consumption of the terminal. Comprehensive reward function (weighted harmonic mean): ; Weight coefficient: :Represents the priority of security, real-time performance and energy efficiency (default ). Policy iteration generates a dynamic mapping matrix and a policy function (deep deterministic policy gradient, DDPG): Actor network: input state S, output action Z; Critic network: input S and Z, output Q value estimation. Policy update rule: ; θ: Actor network parameters; μ(S): policy function; Q(S,Z): Q value estimation of the Critic network. Dynamic mapping matrix: ; Matrix elements: Each row represents an action combination (sharding granularity + priority) and its corresponding expected reward R.

[0034] Specifically, the process of extracting spatiotemporal key features by combining a convolutional network and outputting a slice sequence adapted to 5G slice resources and the corresponding encryption strength label is as follows: Spatiotemporal feature extraction of the video stream is performed through a lightweight convolutional network, motion correlation between key frames and sudden silence segments in the audio stream are identified, and the frequency domain mutation points of the sensor waveform are integrated to generate spatiotemporal key feature vectors; based on the multimedia data value density feature vector and the 5G slice resource status, the slice boundaries are dynamically divided and the encryption strength labels are marked, and a slice sequence adapted to the uRLLC, eMBB, and mMTC slice characteristics is output.

[0035] In this implementation, the spatiotemporal feature extraction process of the video stream is as follows: a lightweight convolutional network (such as MobileNetV3) is used to extract the spatiotemporal features of the video stream to capture the motion correlation between key frames. ;Parameter description: : Input video frame sequence (dimension: time × space × channel); : Convolution kernel weights (initialized by pre-training model); : Output spatiotemporal feature vector. Audio stream burst silence detection process: Detect the burst silence segment in the audio stream through short-term energy analysis and extract the burst frequency feature. Formula: ; ;Parameter description: : The first sampling points; : Energy threshold of silent segment (dynamically adjusted according to ambient noise); : The proportion of silent segments (unitless ratio). Sensor waveform frequency domain mutation point analysis, process: Perform fast Fourier transform (FFT) on the sensor waveform (such as vibration, temperature) to detect frequency domain energy mutation. Formula: ; ;Parameter description: : sensor waveform time series; : frequency domain energy distribution vector; : Frequency domain mutation amplitude (physical unit). Multimodal spatiotemporal key feature fusion process: Weighted fusion of video, audio, and sensor features to generate a unified spatiotemporal key feature vector. Formula: ;Parameter description: : Weight coefficient, dynamically adjusted according to data value density, to meet ; The fused spatiotemporal key feature vector. Dynamic shard boundary division and encryption label marking process: Combine the value density feature vector and 5G slice resource status to dynamically divide the shard boundary and assign encryption strength labels. Shard duration calculation: ;Parameter description: : value density level (0~1); : 5G slice resource state vector; : Normalization coefficient (to ensure within a reasonable range); : Sharding duration.

[0036] Specifically, the specific process of dynamically selecting an encryption algorithm cluster based on the sharding granularity and priority mapping rules is as follows: construct a heterogeneous encryption algorithm pool, generate a multi-dimensional matching index based on the sharding priority weight and encryption strength label, and use a sliding time window to count the network jitter coefficient and computing load fluctuation characteristics of the historical sharding; according to the current slice resource remaining bandwidth and terminal hardware acceleration capabilities, dynamically combine symmetric encryption algorithms and lightweight asymmetric encryption algorithms to form an algorithm cluster with sharding granularity and encryption strength linkage, including: fine-grained sharding: binding attribute-based homomorphic encryption algorithm, using multimodal data spatiotemporal The feature is to generate dynamic keys based on attribute parameters. The spatiotemporal features of multimodal data include the hash value of video key frames and the peak value of sensor waveforms. Medium-grained sharding: data blocks are cross-encrypted through a layered hybrid encryption algorithm, and keys are dynamically distributed through an edge node consensus protocol. Coarse-grained sharding: dynamic mask perturbation encryption is used, and keys are issued on demand through 5G control plane signaling to adapt to low-energy transmission requirements. According to the spatiotemporal key feature vectors of the sharding sequence, the key length adaptive adjustment mechanism and encryption mode switching events are triggered to achieve real-time coupling of algorithm cluster parameters and network slice QoS indicators.

[0037] This implementation builds a heterogeneous encryption algorithm pool: This aggregates multiple encryption algorithms, including symmetric encryption and lightweight asymmetric encryption, to form an algorithm pool. Each algorithm is categorized based on different performance requirements and resource constraints to accommodate different granularity and encryption strength requirements. Multi-dimensional matching indexes are generated: Shard priority weights and encryption strength labels are used to generate multi-dimensional matching indexes. These indexes determine the appropriate encryption algorithm based on the spatiotemporal characteristics of the shard, the encryption strength label, and network status characteristics (such as bandwidth and load). Historical network jitter and computational load statistics are collected using a sliding time window to analyze the network jitter coefficient and computational load fluctuation characteristics of historical shards. This process helps optimize the current encryption algorithm selection to adapt to changing network conditions and load fluctuations. Dynamic encryption algorithm cluster selection: Based on the remaining bandwidth of the current slice and the acceleration capabilities of the terminal hardware, a suitable encryption algorithm is dynamically combined and selected. For shards of different granularities (fine, medium, and coarse), the corresponding encryption algorithm is selected: For fine-grained shards, the attribute-based homomorphic encryption algorithm is bound to generate dynamic keys based on the spatiotemporal characteristics of multimodal data. Medium-grained sharding uses features such as video keyframe hash values and sensor waveform peaks to cross-encrypt data blocks using a layered hybrid encryption algorithm. Keys are dynamically distributed through an edge node consensus protocol. Coarse-grained sharding utilizes dynamic mask perturbation encryption, with keys distributed on demand via 5G control plane signaling to meet low energy consumption requirements. Key length adaptive adjustment and encryption mode switching are triggered based on the spatiotemporal key feature vectors of the sharding sequence. This mechanism adjusts encryption strength and efficiency based on real-time network conditions and data requirements to ensure that the network slice's QoS (Quality of Service) metrics are met.

[0038] Specifically, the specific process of cross-slice key distribution and update through the alliance chain to generate encrypted shard data blocks and key distribution logs is as follows: the shard key is split into multiple sub-key components according to the slice domain, and the threshold signature rules and consensus verification conditions for cross-domain key distribution are defined; through the dynamic authorization channel between the alliance chain nodes, the sub-key components are pushed to the edge computing nodes corresponding to the uRLLC, eMBB, and mMTC slices, and the key version number and distribution path topology are recorded in the tamper-proof block log; combined with the encryption timestamp and lifecycle label of the shard data block, periodic key update events are triggered, and the key copies of the heterogeneous blockchain network are synchronously updated using the cross-chain protocol.

[0039] In this implementation, the shard key is split into multiple subkey components based on the slice domain: The key is split into multiple subkey components based on different slice domains (such as uRLLC, eMBB, and mMTC). This approach ensures that different slices have independent and dedicated key components based on their specific quality of service requirements (such as latency, bandwidth, and number of connections), thereby improving security and flexibility. A threshold signature rule and consensus verification conditions are defined for cross-domain key distribution: A threshold signature rule is designed that requires verification by multiple nodes to complete key distribution and update. This strategy ensures that the key distribution process is consistent across multiple consortium chain nodes, thereby improving the security of the key update process. The consensus verification conditions ensure that key updates can only be performed when specific conditions (such as the number of nodes and signature requirements) are met during cross-domain key distribution, thereby avoiding the risk of single points of failure or malicious operations. Subkey components are pushed to slice edge computing nodes via a dynamic authorization channel between consortium chain nodes: Utilizing the dynamic authorization channel between consortium chain nodes, each subkey component is pushed to the edge computing nodes of the uRLLC, eMBB, and mMTC slices. This step ensures the security of key transmission through encrypted transmission and access control, while also ensuring that keys are distributed only to appropriate nodes. Recording the key version number and distribution path topology in an immutable block log: Each key distribution is recorded in the consortium chain's block log. This makes the key distribution process completely transparent and traceable, and prevents tampering or retroactive modification, enhancing the system's trustworthiness and auditability. Triggering periodic key update events based on the cryptographic timestamps and lifecycle tags of shard data blocks: Based on the cryptographic timestamps and lifecycle tags of shard data blocks, the system can automatically trigger periodic key update events. This mechanism ensures regular key updates throughout the lifecycle, thereby ensuring long-term data security and preventing security risks caused by key leakage or expiration. Synchronizing key copies across heterogeneous blockchain networks using a cross-chain protocol: Through the cross-chain protocol, key updates are not only performed within a single consortium chain but are also synchronized to relevant nodes across heterogeneous blockchain networks. This step ensures the consistency and synchronization of key copies across different blockchain networks, contributing to the security and consistency of the cross-chain system.

[0040] Specifically, the specific process of constructing a redundant coding strategy driven by fountain codes is as follows: a redundancy factor is dynamically generated based on the encryption strength label and spatiotemporal correlation of the sharded data blocks, and the coding degree distribution function of the fountain code is adjusted according to the spatiotemporal feature distribution density of the multimodal data; the encrypted sharded data blocks and redundant check blocks are arranged in a two-dimensional matrix in the frequency domain and time domain, and the shard group correlation strength is dynamically detected through a sliding window and differentiated redundancy is injected; combined with the real-time packet loss rate of the slice and the transmission reliability index, a dynamic adaptation relationship between the redundant coding parameters and the QoS requirements is established to generate a multi-slice collaborative redundancy strategy.

[0041] In this implementation, redundancy factors are dynamically generated based on the encryption strength labels and spatiotemporal correlation of the data chunks. The redundancy factor is dynamically adjusted based on each data chunk's encryption strength label (e.g., high, medium, or low encryption strength) and the spatiotemporal correlation of the data (e.g., the temporal relationship between video frames, the temporal characteristics of sensor data, etc.). This means that chunks with higher encryption strengths may require more redundancy to mitigate the computational and transmission overhead associated with high-strength encryption. For chunks with strong spatiotemporal correlation, redundancy factors are generated to ensure data integrity and redundancy are maintained during transmission, preventing packet loss or data corruption. The coding degree distribution function of the fountain code is adjusted based on the spatiotemporal feature density of multimodal data. The spatiotemporal feature density of multimodal data (e.g., video, audio, sensor data) is analyzed to adjust the coding degree distribution function of the fountain code. The spatiotemporal feature density refers to the distribution of data in the spatiotemporal domain, and the coding degree distribution function determines how the redundancy code is allocated. This adjustment ensures that an appropriate amount of redundant data is generated between different data chunks, optimizing the storage and transmission efficiency of the redundancy. Arrange the encrypted data blocks and redundancy check blocks in a two-dimensional matrix in the frequency and time domains: The encrypted data blocks and redundancy check blocks are combined into a two-dimensional matrix in the frequency and time domains and arranged. This process leverages the characteristics of the frequency and time domains, taking into account the distribution characteristics of both time and frequency during the encoding process, allowing for more effective adaptation to the transmission requirements of diverse network environments. For example, the frequency and time domain arrangement helps reduce packet loss under varying transmission conditions and improves transmission stability. Dynamically detect the correlation strength of shard groups and inject differentiated redundancy using a sliding window: Utilizing a sliding window technique to dynamically detect the correlation strength between different shard groups, this detection method assesses the correlation and redundancy between data. If certain shards are highly correlated, differentiated redundancy can be injected as needed. For example, shard groups with strong correlations may require less redundancy, while shard groups with weaker correlations may require more redundancy. This strategy intelligently adjusts the amount and distribution of redundancy based on data characteristics. By combining the real-time packet loss rate of slices and transmission reliability indicators, a dynamic adaptation relationship between redundancy coding parameters and QoS requirements is established: Redundancy coding parameters are dynamically adjusted based on the real-time packet loss rate and transmission reliability indicators. Network environments with high packet loss rates may require more redundancy to ensure data integrity and reliability, while networks with higher transmission reliability can reduce the generation of redundant data. By establishing dynamic adaptation relationships, we can ensure that the redundant coding strategy can meet the required Quality of Service (QoS) requirements under various network conditions, optimizing transmission efficiency and data security. Generating a multi-slice collaborative redundancy strategy: In a multi-slice environment, the collaborative redundancy strategy is generated by combining information such as the network status, packet loss rate, and transmission reliability of each slice.This means that different slices (such as uRLLC, eMBB, and mMTC) will jointly optimize the configuration of redundant coding based on their respective QoS requirements, network conditions, and redundancy requirements to ensure collaborative work between multiple slices and avoid network overload or waste of redundant resources.

[0042] Specifically, the game theory model is combined to optimize the shard transmission path and the specific process of outputting the mapping table of associated shard groups and transmission paths is as follows: define the available bandwidth of the transmission node, the slice resource competition coefficient and the slice security level as the strategy space, and design a composite utility function that includes delay cost, security risk loss and energy consumption cost; solve the Nash equilibrium point through the evolutionary game algorithm, generate the initial matching scheme of shard groups and transmission paths, and dynamically correct the path selection weight according to the spatiotemporal key feature vectors of the shard sequence; combine the blockchain key distribution topology to build a credibility assessment mechanism and output the dynamic mapping table of shard groups and slice resources.

[0043] In this implementation plan, the policy space is defined: Available bandwidth of transmission node: represents the bandwidth capability of each transmission node (such as base station, router, etc.), measured in bits per second (bps). This is an important factor affecting the data transmission rate. Slice resource competition coefficient: represents the degree of competition for each slice resource in a multi-slice network environment. This coefficient measures the difficulty of resource sharing between slices. The larger the value, the more intense the competition. Slice security level: represents the security requirements of each data slicing, usually based on factors such as encryption strength and data sensitivity. This level affects the security policy of the slicing path selection. These parameters constitute the policy space in the game model, that is, the policy space that game participants (transmission paths, slicing, etc.) can choose. Design a composite utility function: Design a composite utility function: The composite utility function combines the delay cost, security risk loss, and energy consumption cost to evaluate the pros and cons of different path selections. The composite utility function can be expressed as: ;in: : represents the total utility of path P. : represents the delay cost of path P, which is usually calculated as the sum of the delays of each node in the path, in seconds. : Represents the security risk loss of path P, which is usually related to the security level and encryption strength of the path. A higher risk loss means a poorer path security. : Indicates the energy consumption cost of path P, which is usually related to factors such as the number of nodes on the path and bandwidth requirements, and is expressed in watts. Parameter Description: 、 、 : Represent the importance coefficients of latency, risk, and energy consumption, respectively, reflecting the degree of influence of different factors on path selection. Solving the Nash equilibrium using an evolutionary game algorithm: The evolutionary game algorithm is used to simulate the game between shards and transmission paths, finding the strategic balance point between each participant (shard group and transmission path), that is, the Nash equilibrium. Each shard group adjusts its strategy based on the utility function, and the transmission path also makes its choice based on its own resource status. The strategy update in the evolutionary game can be expressed as: ;in: : Indicates the strategy of the i-th shard group in the t-th round (i.e., the selected transmission path). : represents the average utility of all current paths. : Indicates the step size of the policy update and controls the update rate. The evolutionary game is continuously iterated to eventually obtain a stable Nash equilibrium point, which represents the optimal policy allocation between each shard group and the transmission path. Dynamically correct the path selection weight: Dynamically adjust the path selection weight according to the spatiotemporal key feature vectors of the shard sequence (for example, factors such as latency, packet loss rate, and network congestion). This means that as the network status changes, the weights of certain paths may increase or decrease. For example, when the network is congested, it may be more preferable to choose a low-latency path. The update formula for weight adjustment can be expressed as: in: represents the path selection weight of the i-th packet and the j-th transmission path at time t. Represents the spatiotemporal characteristics of the packet corresponding to path j, including delay, network load, and packet loss. Represents the average spatiotemporal characteristics of all paths. Indicates the step size of weight update. Through this dynamic correction mechanism, the path selection weight can reflect the changes in network status in real time.

[0044] Specifically, according to the mapping table of associated shard groups and transmission paths, the transmission traffic characteristics and key distribution logs are analyzed, and the specific process of dynamically adjusting the encryption strength threshold and shard retransmission strategy is as follows: real-time monitoring of the packet loss rate, key distribution delay and blockchain consensus verification success rate of the shard transmission path, and construction of a joint assessment matrix of network status and security risks; a dynamic adjustment model of the encryption strength threshold is established based on the reinforcement learning algorithm, and the shard encryption mode switching event is triggered according to the risk level, and a shard retransmission priority queue is generated at the same time.

[0045] In this implementation, real-time monitoring of transmission path performance indicators includes: Packet loss rate: The packet loss rate is the ratio of data packets that fail to reach the receiving end during transmission to the total number of transmitted packets. Real-time monitoring of the packet loss rate helps assess network reliability, especially under high load or congestion. Key distribution latency: Key distribution latency represents the delay from key generation to actual key distribution to the target node. High key distribution latency can delay encryption operations and affect the timeliness of data transmission. Blockchain consensus verification success rate: The blockchain consensus verification success rate measures the effectiveness of the consensus mechanism between nodes and whether it can quickly and reliably verify the legitimacy of key distribution and transmission paths. These performance indicators are collected and analyzed in real time through monitoring and recording. Constructing a joint assessment matrix for network status and security risks: An assessment matrix is constructed by comprehensively considering network performance (such as packet loss rate and latency) and security risks (such as the security of key transmission and the efficiency of blockchain verification). This matrix dynamically adjusts transmission strategies based on different network status and security risks. The joint assessment matrix typically includes the following dimensions: Network performance: This includes packet loss rate, latency, bandwidth utilization, etc. Security risk dimension: including the security of key distribution, encryption algorithm strength, security of transmission path, etc. Dynamically adjust the encryption strength threshold based on reinforcement learning algorithm: Reinforcement learning algorithm is used to dynamically adjust the encryption strength threshold according to network status and security risks. Reinforcement learning continuously updates strategies to maximize a certain goal (such as maximizing the security or reliability of the network) through interaction with the environment. The core of the reinforcement learning model is the state-action-reward mechanism. In this process: State: the current network status (such as packet loss rate, latency, etc.) and security risk level. Action: the strategy for adjusting the encryption strength threshold. For example, the strength of the encryption algorithm can be adjusted (such as switching from symmetric encryption to stronger asymmetric encryption), or the key length of the encryption algorithm can be changed. Reward: rewards are given based on a joint evaluation of network performance and security risks. For example, lower packet loss rates and lower latency can result in higher rewards. Update formula for reinforcement learning model: ;in: : represents the expected benefit of taking action a in state s. : Learning rate, which controls the influence of new experience. r: The reward value of the current step. : Discount factor, measuring the importance of future rewards. s': New state. a': New action. Based on the results of the joint evaluation matrix, if the current network status and security risk level exceed the set threshold, an encryption mode switch event will be triggered. For example, if the packet loss rate is very high or the key distribution security is insufficient, a more secure encryption mode can be automatically switched. Switching events can be implemented through a dynamic triggering mechanism, for example, using a time window-based triggering mechanism.

[0046] In summary, this application has at least the following effects:

[0047] This real-time, segmented transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks uses a multi-objective optimization algorithm to balance encryption strength, transmission latency, and energy consumption, achieving an optimal compromise between data security and transmission efficiency in diverse network environments. A dynamic adjustment mechanism ensures full utilization of network resources while guaranteeing high data security. Based on fluctuations in network slice resources and actual transmission conditions, it automatically adjusts encryption algorithms, transmission paths, and redundant encoding strategies, flexibly adapting to various network environment changes and achieving real-time optimization. By precisely controlling encryption strength, path selection, and redundant encoding strategies, it optimizes network load, reduces unnecessary latency and energy consumption, improves overall transmission efficiency, and adapts to resource constraints in diverse scenarios. Game-theory-based path optimization and redundant encoding strategies effectively prevent data loss during transmission, ensuring the integrity and stable transmission of critical data, thereby improving system transmission reliability. Through risk assessment and dynamic response instruction generation, it identifies security risks and performance bottlenecks in network transmission in real time, and automatically adjusts encryption modes, path selection, and retransmission strategies to address diverse security requirements and transmission conditions.

[0048] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0049] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0050] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0051] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0052] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0053] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A real-time fragmented transmission method for heterogeneous encrypted multimodal multimedia messages for 5G networks, characterized by: The following steps are involved: S1. Obtain the sensitivity labels and value density of multimedia data, monitor the resource status of 5G network slices in real time, dynamically generate slice granularity and priority mapping rules through reinforcement learning algorithms, and combine convolutional networks to extract key spatiotemporal features. Output a slice sequence adapted to 5G slice resources and the corresponding encryption strength label. S2. Receive the shard sequence and encryption strength label, dynamically select the encryption algorithm cluster based on the shard granularity and priority mapping rules, implement cross-shard key distribution and update through the consortium chain, and generate encrypted shard data blocks and key distribution logs; S3. Based on the spatiotemporal correlation of encrypted sharded data blocks and multimodal data, a fountain code-driven redundant encoding strategy is constructed. This strategy is combined with a game theory model to optimize shard transmission paths and output a mapping table of associated shard groups and transmission paths. S4. Analyze the transmission traffic characteristics and key distribution logs based on the associated fragment group and transmission path mapping table, and dynamically adjust the encryption strength threshold and fragment retransmission strategy; The specific process of constructing a fountain code-driven redundant encoding strategy is as follows: Dynamically generate redundancy factors based on the encryption strength labels and spatiotemporal correlation of sharded data blocks, and adjust the encoding degree distribution function of the fountain code based on the spatiotemporal feature distribution density of multimodal data; Arrange the encrypted shard data blocks and redundancy check blocks in a two-dimensional matrix in the frequency and time domains, dynamically detect the shard group correlation strength through a sliding window, and inject differentiated redundancy; Combined with the real-time packet loss rate and transmission reliability indicators of the slice, a dynamic adaptation relationship between redundant coding parameters and QoS requirements is established to generate a multi-slice collaborative redundancy strategy.

2. The method for real-time fragmented transmission of heterogeneous encrypted multimodal multimedia messages for 5G networks according to claim 1 is characterized in that: The specific process of obtaining the sensitivity label and value density of multimedia data and monitoring the resource status of 5G network slices in real time is as follows: Based on the distribution density of sensitive words in text data annotated by natural language processing, the key frame motion trajectory features of video data are extracted, the frequency domain burst frequency of audio data is analyzed, and the value density level of multimedia data is dynamically calculated by combining the mutation amplitude of sensor waveforms. The remaining bandwidth of the uRLLC slice, the delay jitter of the eMBB slice, and the packet loss rate of the mMTC slice are collected in real time through the 5G core network interface to generate a 5G network resource state vector.

3. The method for real-time fragmented transmission of heterogeneous encrypted multimodal multimedia messages for 5G networks according to claim 2 is characterized in that: The specific process of dynamically generating sharding granularity and priority mapping rules through the reinforcement learning algorithm is as follows: A state space is constructed with the multimedia data value density feature vector and the 5G network slice resource state vector as input, the slicing granularity level and transmission priority weight are defined as the action space, a multi-objective reward function is designed to comprehensively consider encryption strength matching, delay deviation and terminal energy consumption, and a dynamic mapping matrix between slicing granularity and priority is generated through strategy iteration.

4. The method for real-time fragmented transmission of heterogeneous encrypted multimodal multimedia messages for 5G networks according to claim 3 is characterized in that: The specific process of extracting spatiotemporal key features using a convolutional network and outputting a slice sequence adapted to 5G slice resources and the corresponding encryption strength label is as follows: A lightweight convolutional network is used to extract spatiotemporal features from video streams, identify motion correlations between key frames and sudden silences in audio streams, and fuse frequency domain mutation points of sensor waveforms to generate spatiotemporal key feature vectors. According to the multimedia data value density feature vector and 5G slice resource status, the slice boundaries are dynamically divided and marked with encryption strength tags, and a slice sequence adapted to the uRLLC, eMBB, and mMTC slice characteristics is output.

5. The method for real-time fragmented transmission of heterogeneous encrypted multimodal multimedia messages for 5G networks according to claim 4 is characterized in that: The specific process of dynamically selecting an encryption algorithm cluster based on sharding granularity and priority mapping rules is as follows: Build a heterogeneous encryption algorithm pool, generate a multi-dimensional matching index based on shard priority weights and encryption strength labels, and use a sliding time window to calculate the network jitter coefficient and computing load fluctuation characteristics of historical shards; Based on the remaining bandwidth of the current slice resources and the terminal hardware acceleration capabilities, symmetric encryption algorithms and lightweight asymmetric encryption algorithms are dynamically combined to form an algorithm cluster that links slice granularity and encryption strength, including: Fine-grained sharding: Binding attribute-based homomorphic encryption algorithms generates dynamic keys using the spatiotemporal characteristics of multimodal data as attribute parameters. The spatiotemporal characteristics of multimodal data include hash values of video key frames and sensor waveform peaks. Medium-granularity sharding: Data blocks are cross-encrypted using a layered hybrid encryption algorithm, and keys are dynamically distributed through edge node consensus protocols. Coarse-grained sharding: Encryption is performed through dynamic mask perturbation, and keys are delivered on demand through 5G control plane signaling to adapt to low-energy transmission requirements; According to the spatiotemporal key feature vectors of the slicing sequence, the key length adaptive adjustment mechanism and encryption mode switching events are triggered to achieve real-time coupling of algorithm cluster parameters and network slice QoS indicators.

6. The method for real-time fragmented transmission of heterogeneous encrypted multimodal multimedia messages for 5G networks according to claim 5, characterized in that: The specific process of cross-shard key distribution and update through the consortium chain to generate encrypted shard data blocks and key distribution logs is as follows: Split the shard key into multiple sub-key components according to the shard domain, and define the threshold signature rules and consensus verification conditions for cross-domain key distribution; Through the dynamic authorization channel between the alliance chain nodes, the sub-key components are pushed to the edge computing nodes corresponding to the uRLLC, eMBB, and mMTC slices, and the key version number and distribution path topology are recorded in the tamper-proof block log; Combining the encrypted timestamp and lifecycle tag of the sharded data block, periodic key update events are triggered, and the key copies of heterogeneous blockchain networks are synchronously updated using the cross-chain protocol.

7. The method for real-time fragmented transmission of heterogeneous encrypted multimodal multimedia messages for 5G networks according to claim 6, characterized in that: The specific process of optimizing the shard transmission path by combining the game theory model and outputting the mapping table of associated shard groups and transmission paths is as follows: Define the available bandwidth of transmission nodes, the slice resource competition coefficient, and the slice security level as the policy space, and design a composite utility function that includes latency cost, security risk loss, and energy consumption cost; The Nash equilibrium point is solved through the evolutionary game algorithm to generate the initial matching scheme between the shard group and the transmission path, and the path selection weight is dynamically modified according to the spatiotemporal key feature vectors of the shard sequence; Combined with the blockchain key distribution topology, a credibility assessment mechanism is constructed to output a dynamic mapping table between shard groups and slice resources.

8. The method for real-time fragmented transmission of heterogeneous encrypted multimodal multimedia messages for 5G networks according to claim 7, characterized in that: The specific process of dynamically adjusting the encryption strength threshold and fragment retransmission strategy based on the associated fragment group and transmission path mapping table, analyzing the transmission traffic characteristics and key distribution logs, is as follows: Real-time monitoring of the packet loss rate, key distribution delay, and blockchain consensus verification success rate of the shard transmission path, and the establishment of a joint assessment matrix of network status and security risks; A dynamic adjustment model for encryption strength threshold is established based on the reinforcement learning algorithm, which triggers the fragment encryption mode switching event according to the risk level and generates a fragment retransmission priority queue.

Images