Coupon pushing method and device for multi-port delivery

Abstract

The invention discloses a multi-port delivery coupon pushing method and device, and relates to the technical field of intelligent marketing, and the method comprises the steps: inputting a multi-modal feature tensor into an SNN decision engine, generating SNN synaptic weight parameters based on user historical behavior spatio-temporal features, and dynamically adjusting the port priority through a synaptic plasticity rule, generating a dynamic port strategy matrix; based on the dynamic port strategy matrix, pre-loading a coupon strategy through a memristor array, predicting a user track in combination with a quantum city spatio-temporal model, and generating a real-time push instruction; and executing a real-time push instruction, implanting a topological photonic crystal verification code in the coupon propagation link, verifying the verification authenticity of the user through zero-knowledge proof, and generating a verification certificate chain. According to the method, the environmental noise is eliminated by using the space-time coding characteristics of the pulse neural network, and the multi-source characteristics are fused, so that the problem of characteristic distortion caused by data heterogeneity in edge calculation is solved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent marketing technology, and in particular to a multi-port coupon push method and device. Background Art

[0002] In recent years, with the rapid development of mobile internet and IoT technologies, the field of intelligent marketing is undergoing a transformation from traditional one-way push notifications to multimodal, intelligent decision-making. In terms of coupon push technology, existing technologies primarily focus on user profiling and static rule engines, such as recommendation devices based on collaborative filtering algorithms and distributed marketing platforms based on federated learning. The introduction of edge computing enables real-time data processing, and the low latency of 5G networks further enhances the timeliness of marketing decisions.

[0003] However, existing technologies still have significant shortcomings: First, in terms of adaptability to dynamic environments, push mechanisms based on static rules or shallow machine learning struggle to effectively capture the spatiotemporal correlations of user behavior, resulting in the inability to accurately allocate bandwidth and adjust priorities in multi-port collaboration scenarios. Second, in terms of security verification, existing federated learning solutions mostly use classical encryption algorithms, which are vulnerable to quantum computing attacks. They also lack a dynamic binding mechanism for redemption behavior and identity parameters, resulting in a high rate of fraudulent coupon use. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a multi-port coupon push method to solve the problems of poor adaptability to dynamic environments and fragile security verification mechanisms in the prior art.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, the present invention provides a coupon push method for multi-port delivery, which includes collecting multi-source heterogeneous data, pre-processing the multi-source heterogeneous data through a neuromorphic processing architecture, and generating a multimodal feature tensor; inputting the multimodal feature tensor into an SNN decision engine, generating SNN synaptic weight parameters based on the spatiotemporal characteristics of user historical behavior, and dynamically adjusting the port priority through synaptic plasticity rules to generate a dynamic port policy matrix; based on the dynamic port policy matrix, preloading the coupon policy through a memristor array, and predicting the user trajectory in combination with a quantum city spatiotemporal model to generate a real-time push instruction; executing the real-time push instruction, and implanting a topological photonic crystal verification code in the coupon propagation link, verifying the authenticity of the user's redemption through zero-knowledge proof, and generating a redemption voucher chain; based on the redemption voucher chain, updating the SNN synaptic weight parameters through quantum secure federated learning, and synchronizing them to the 5G base station edge server and the supermarket local server, and completing the cross-end push of the coupon through the multi-port collaborative delivery engine.

[0008] As a preferred solution of the multi-port coupon push method described in the present invention, the multi-source heterogeneous data includes spatiotemporal positioning signals, radio frequency signals, platform transaction logs and port real-time load data.

[0009] As a preferred solution of the multi-port coupon push method of the present invention, wherein: the generation of multimodal feature tensors is carried out in the following specific steps:

[0010] The neuromorphic processing architecture eliminates environmental noise in radio frequency signals, performs drift correction on spatiotemporal positioning signals, and filters outliers in platform transaction logs to generate purified multi-source heterogeneous data.

[0011] The purified multi-source heterogeneous data is input into the feature extraction layer of the neuromorphic processing architecture to generate spatiotemporal feature vectors of user behavior, product interaction feature vectors, and consumer preference feature vectors;

[0012] Through the pulse fusion layer of the neuromorphic processing architecture, the spatiotemporal feature vectors of user behavior, the feature vectors of product interaction, and the feature vectors of consumption preference are tensor-concatenated to generate a multimodal feature tensor.

[0013] As a preferred solution of the coupon push method for multi-port delivery of the present invention, wherein: the multimodal feature tensor is input into the SNN decision engine, and the SNN synaptic weight parameters are generated based on the spatiotemporal characteristics of the user's historical behavior, and the port priority is dynamically adjusted through the synaptic plasticity rule to generate a dynamic port strategy matrix. The specific steps are as follows:

[0014] The multimodal feature tensor is input into the spatiotemporal feature encoding layer of the SNN decision engine. The spatiotemporal features of the user's historical behavior are extracted through the pulse time window segmentation algorithm to generate the encoded spatiotemporal feature vector. Based on the correlation between the user's behavior frequency and the port interaction timing, the SNN synaptic weight parameters are generated.

[0015] The SNN synaptic weight parameters are input into the synaptic plasticity control layer, and the port priority is dynamically adjusted through the Gaussian pulse timing-dependent plasticity rule to output the optimized SNN synaptic weight distribution;

[0016] According to the optimized SNN synaptic weight distribution, combined with the real-time load data of the port, a tensor outer product operation is performed to generate a four-dimensional tensor of dynamic port strategy, and the dynamic port strategy matrix is ​​generated through multimodal fusion dimensionality reduction.

[0017] As a preferred solution of the multi-port coupon push method of the present invention, the specific steps of generating a real-time push instruction are as follows:

[0018] Based on the row vector-physical port mapping relationship of the dynamic port strategy matrix, the priority parameters of the coupon strategy are linearly mapped to the conductivity value of the memristor array, and the strategy tensor is generated by modulating the resistance state of the dual-memristor crossbar array.

[0019] The strategy tensor is input into the quantum city space-time model, and the user's historical trajectory is grid-encoded and space-time entangled to generate a predicted coordinate sequence tensor;

[0020] Perform a four-dimensional spatiotemporal convolution operation on the predicted coordinate sequence tensor and the policy tensor, generate a comprehensive score tensor for the candidate nodes through the attention-weighted kernel function, and filter the valid candidate node set based on the dynamic fusion threshold;

[0021] When the quantum-clock deviation in the set of valid candidate nodes meets the synchronization tolerance threshold, the dynamic port policy matrix calculates the bandwidth allocation coefficient of the target port and generates a real-time push instruction.

[0022] As a preferred solution of the multi-port coupon push method of the present invention, the specific steps of generating a verification voucher chain are as follows:

[0023] A topological photonic crystal verification code is embedded in the coupon transmission link corresponding to the real-time push instruction, and a decoding verification interface is preset;

[0024] When the user terminal accesses the decoding and verification interface, the optical feature information is decoded and verified through the zero-knowledge proof protocol, and an interactive proof file is generated;

[0025] The interactive proof document is matched with the coupon redemption request in time and space, and a redemption voucher chain with a multi-level verification data structure is generated.

[0026] As a preferred solution of the multi-port coupon push method described in the present invention, the SNN synaptic weight parameters are updated through quantum secure federated learning based on the verification voucher chain, and synchronized to the 5G base station edge server and the supermarket local server, and the cross-end push of coupons is completed through the multi-port collaborative delivery engine. The specific steps are as follows:

[0027] Based on the verification certificate chain, the space-time feature verification parameters are generated through the quantum secure hash algorithm and space-time convolution operation;

[0028] Verify parameters based on spatiotemporal features and update SNN synaptic weight parameters through a quantum-safe federated learning protocol;

[0029] The updated SNN synaptic weight parameters are synchronized to the 5G base station edge server and the supermarket local server through the quantum key distribution channel to form a two-terminal dynamic parameter pool;

[0030] The multi-port collaborative delivery engine parses the dual-end dynamic parameter pool, generates time-space matching cross-end push instructions, and executes coupon push.

[0031] In the second aspect, the present invention provides a multi-port coupon push device, including a data fusion module, a decision generation module, a strategy preloading module, a cancellation verification module and a federated optimization module. The data fusion module is used to collect multi-source heterogeneous data, pre-process the multi-source heterogeneous data through a neuromorphic edge computing architecture, and generate a multimodal feature tensor; the decision generation module is used to input the multimodal feature tensor into an SNN decision engine, generate SNN synaptic weight parameters based on the spatiotemporal characteristics of the user's historical behavior, and dynamically adjust the port priority through the synaptic plasticity rules to generate a dynamic port strategy matrix; the strategy preloading module is used to collect multi-source heterogeneous data, pre-process the multi-source heterogeneous data through a neuromorphic edge computing architecture, and generate a multimodal feature tensor; the decision generation module is used to input the multimodal feature tensor into the ... The block is used to preload coupon policies through the memristor array based on the dynamic port policy matrix, and predict user trajectories in combination with the quantum city space-time model to generate real-time push instructions; the redemption verification module is used to execute real-time push instructions and implant topological photonic crystal verification codes in the coupon propagation link, verify the authenticity of user redemption through zero-knowledge proof, and generate a redemption voucher chain; the federated optimization module is used to update the SNN synaptic weight parameters based on the redemption voucher chain through quantum secure federated learning, and synchronize them to the 5G base station edge server and the supermarket local server, completing the cross-end push of coupons through the multi-port collaborative delivery engine.

[0032] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the multi-port coupon push method as described in the first aspect of the present invention is implemented.

[0033] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the multi-port coupon push method as described in the first aspect of the present invention is implemented.

[0034] The beneficial effects of the present invention are: by utilizing the spatiotemporal coding characteristics of the pulse neural network to eliminate environmental noise and fuse multi-source features, the feature distortion problem caused by data heterogeneity in edge computing is solved; further, the port strategy is dynamically adjusted through the synaptic plasticity rules of the SNN decision engine, which solves the problem that the static rule engine cannot respond to changes in spatiotemporal behavior in real time, and realizes millisecond-level precise scheduling of multi-port resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 Flowchart of the coupon push method for multi-port delivery.

[0037] Figure 2 Flowchart of generating multimodal feature tensors for the coupon push method for multi-port delivery.

[0038] Figure 3 Flowchart for generating a dynamic port policy matrix for a coupon push method for multi-port delivery.

[0039] Figure 4 A flowchart for generating real-time push instructions for a coupon push method for multi-port delivery. DETAILED DESCRIPTION

[0040] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0041] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0042] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0043] Reference Figures 1 to 4 , is an embodiment of the present invention, which provides a multi-port coupon push method, including the following steps:

[0044] S1: Collect multi-source heterogeneous data, pre-process the multi-source heterogeneous data through the neuromorphic edge computing architecture, and generate multimodal feature tensors.

[0045] S1.1: Multi-source heterogeneous data includes spatiotemporal positioning signals, radio frequency signals, platform transaction logs, and real-time port load data.

[0046] It should be noted that the spatiotemporal positioning signal refers to the spatiotemporal positioning signal of the user's mobile terminal, which obtains the user's real-time latitude and longitude coordinates and timestamp through GPS or Beidou navigation, and combines indoor Bluetooth beacons or UWB positioning technology to supplement the meter-level precision location information in indoor scenarios; the radio frequency signal refers to the radio frequency signal of the supermarket IoT device, which achieves sub-meter spatial positioning through signal strength fingerprint library matching; the platform transaction log refers to the transaction log of the third-party payment platform, which is stored in a time series database and establishes an associated index with the spatiotemporal positioning signal; the port real-time load data refers to the real-time monitoring indicator set of the bandwidth occupancy rate, TCP connection number and data packet queue delay of the communication port of the 5G base station edge server, which is collected through the SNMP protocol and normalized to a load factor in the range of 0 to 1.

[0047] S1.2: Eliminate environmental noise from RF signals through a neuromorphic processing architecture, perform drift correction on spatiotemporal positioning signals, and filter outliers on platform transaction logs to generate purified multi-source heterogeneous data.

[0048] The specific process involves the neuromorphic processing architecture processing RF signals through the spatiotemporal coding characteristics of spiking neural networks. The neuromorphic processing architecture uses a pulse firing rate adaptive adjustment mechanism to identify pulse timing patterns in RF signals and utilizes the membrane potential integration characteristics of LIF neurons to separate valid signals from environmental electromagnetic noise. The pulse firing rate adaptive adjustment mechanism is used to identify and eliminate environmental electromagnetic interference; positioning deviations are corrected through synaptic weight updates. Synaptic weight updates are implemented using pulse timing-dependent plasticity rules, dynamically adjusting synaptic connection strengths based on the deviation between the positioning signal timestamp reported by the user's mobile terminal and the base station reference clock. In platform transaction log processing, pulse firing pattern recognition is used to detect and filter abnormal transaction records. The processed RF signal, the corrected spatiotemporal positioning signal, and the cleansed platform transaction log are aligned in the pulse coding space of the neuromorphic processing architecture to generate cleansed multi-source heterogeneous data that retains spatiotemporal correlation features and semantic consistency.

[0049] S1.3: Input the purified multi-source heterogeneous data into the feature extraction layer of the neuromorphic processing architecture to generate user behavior spatiotemporal feature vectors, product interaction feature vectors, and consumer preference feature vectors.

[0050] The specific process includes: when the neuromorphic processing architecture extracts features from the purified multi-source heterogeneous data, it analyzes the movement trajectory patterns in the radio frequency signals and spatiotemporal positioning signals, and uses the pulse timing window segmentation algorithm to generate the user behavior spatiotemporal feature vector; at the same time, it processes the product identification sequence and purchase records in the platform transaction log, and constructs the product interaction feature vector based on the statistical characteristics of the pulse emission frequency; and uses pulse phase encoding technology to form the consumer preference feature vector by combining the distribution characteristics and time interval patterns of the transaction amount; the entire architecture realizes the synchronous alignment of the time dimension of each feature vector through a hierarchical pulse emission mechanism, and finally outputs three types of structured data in parallel: user behavior spatiotemporal feature vector, product interaction feature vector and consumer preference feature vector.

[0051] Furthermore, the neuromorphic processing architecture uses the spatiotemporal encoding characteristics of spiking neural networks to process multi-source heterogeneous data. It employs an adaptive pulse rate adjustment mechanism to eliminate environmental electromagnetic interference in radio frequency signals. A pulse timing-based coordinate compensation model is established to correct drift errors in spatiotemporal positioning signals, and pulse pattern recognition is used to filter out abnormal records in platform transaction logs. The cleansed data is then input into the feature extraction layer. The spatiotemporal feature encoding unit analyzes movement trajectory patterns and generates spatiotemporal feature vectors of user behavior through pulse timing window segmentation. The product interaction analysis unit processes purchase records and constructs product interaction feature vectors based on pulse frequency. The consumer preference extraction unit integrates transaction features using pulse phase encoding to form a consumer preference feature vector. In the pulse fusion layer, the time dimension of each feature vector is aligned using a phase synchronization mechanism. A multidimensional tensor splicing method is used to integrate movement trajectory, product preference, and consumer habit features. The synaptic connection matrix eliminates redundancy in the spliced ​​features. The resulting trained architecture consists of three layers of spiking neural networks. Synaptic weights are optimized using a backpropagation timing-dependent algorithm. Deployed on a neuromorphic chip, it achieves real-time processing capabilities for synaptic operations, with feature extraction errors controlled within 3%.

[0052] S1.4: Through the pulse fusion layer of the neuromorphic processing architecture, the spatiotemporal feature vectors of user behavior, product interaction feature vectors, and consumption preference feature vectors are tensor-concatenated to generate a multimodal feature tensor.

[0053] The specific process includes: the pulse fusion layer of the neuromorphic processing architecture receives the spatiotemporal feature vector of user behavior, the product interaction feature vector and the consumption preference feature vector, and aligns the time dimension of the three feature vectors through the pulse phase synchronization mechanism; the multi-dimensional tensor splicing method of the pulse coding space is used to fuse the spatiotemporal feature vector of user behavior, the product interaction feature vector and the consumption preference feature vector; the synaptic connection matrix in the pulse fusion layer adjusts the weights of the spliced ​​features to eliminate redundant information between features; and the final output is a multimodal feature tensor containing spatiotemporal behavior features, product preference features and consumption pattern features, and the features of each dimension maintain semantic consistency and spatiotemporal correlation.

[0054] S2: Input the multimodal feature tensor into the SNN decision engine, generate SNN synaptic weight parameters based on the spatiotemporal characteristics of user historical behavior, and dynamically adjust the port priority through synaptic plasticity rules to generate a dynamic port policy matrix.

[0055] S2.1: Input the multimodal feature tensor into the spatiotemporal feature encoding layer of the SNN decision engine, extract the spatiotemporal features of the user's historical behavior through the pulse time window segmentation algorithm, generate the encoded spatiotemporal feature vector, and generate the SNN synaptic weight parameters based on the correlation between the user behavior frequency and the port interaction timing.

[0056] The specific process includes: the multimodal feature tensor enters the spatiotemporal feature encoding layer of the SNN decision engine, the pulse time window segmentation algorithm decomposes the user's historical behavior data into a discrete pulse event sequence, and the spatiotemporal coordinates are phase-encoded through LIF neurons; the encoded spatiotemporal feature vector passes through the synaptic weight generation unit, and the Gaussian pulse timing-dependent plasticity rule is used to analyze the time correlation between the user behavior pulse firing rate and the port activation event, and obtain the SNN synaptic weight parameters that reflect the behavior-port association strength; the weight parameter generation process synchronously integrates 5G base station positioning data and indoor Bluetooth beacon information in supermarkets to ensure accurate mapping of spatiotemporal features and physical port locations; the final output SNN synaptic weight parameters include multi-dimensional spatiotemporal correlation features such as user movement trajectory pattern, stay time distribution and port interaction frequency.

[0057] User behavior frequency data is extracted from the spatiotemporal feature vector of user behavior in the multimodal feature tensor, specifically from the statistics of user historical access frequency and interaction time interval recorded in the spatiotemporal positioning signal and platform transaction logs.

[0058] Port interaction timing data is extracted from the real-time load data of ports in multi-source heterogeneous data, specifically from the TCP connection timestamps and interaction event interval sequences recorded by the communication port of the 5G base station edge server.

[0059] The SNN decision engine consists of a spatiotemporal feature encoding layer, a synaptic plasticity regulation layer, and a multimodal fusion dimensionality reduction layer. It is trained through a pulse time window segmentation algorithm and Gaussian pulse timing-dependent plasticity rules, and needs to be deployed in conjunction with the neuromorphic computing chip hardware of the 5G base station edge server.

[0060] Spatiotemporal coordinates refer to the precise positioning data of the user's mobile terminal in the three-dimensional physical space and time dimensions, which are obtained through the integrated positioning technology of the 5G base station positioning system and the indoor Bluetooth beacons in supermarkets.

[0061] S2.2: Input the SNN synaptic weight parameters into the synaptic plasticity control layer, dynamically adjust the port priority through the Gaussian pulse timing-dependent plasticity rule, and output the optimized SNN synaptic weight distribution.

[0062] The specific process includes: after the SNN synaptic weight parameters are input into the synaptic plasticity control layer, the Gaussian pulse timing relies on the plasticity rule to analyze the time interval distribution of the input pulse sequence to obtain the time difference between adjacent pulse emissions; the time difference is weighted according to the Gaussian distribution function, and recent interaction events obtain higher weights, and the weights of historical events decay with time; the synaptic plasticity control layer maps the weight allocation results to the port priority score, and the synaptic connection of the neural pathway corresponding to the high-frequency interaction port is enhanced, and the connection of the low-frequency port pathway is weakened; during the dynamic adjustment process, the pulse emission rate and the port load status are synchronized in real time to ensure that the weight distribution matches the current network conditions; the final output of the optimized SNN synaptic weight distribution reflects both user behavior habits and port resource status to form adaptive decision parameters.

[0063] Port priority refers to the service priority of each communication port of the 5G base station edge server in the coupon push task. It is dynamically derived by analyzing the correlation between user behavior frequency and port interaction timing through the SNN decision engine and combining it with the Gaussian pulse timing dependency plasticity rule.

[0064] The synaptic plasticity regulation layer is extended from the spatiotemporal feature encoding layer built into the SNN decision engine, and is a functional component in the pulse neural network that is specifically responsible for dynamically adjusting the strength of synaptic connections.

[0065] The time interval distribution of a pulse sequence refers to the statistical characteristics of the time difference between user interaction events with different ports. It is specifically manifested as the probability density function of the difference in the time of occurrence of adjacent interaction events in the pulse neural network, which obeys the Gaussian distribution form.

[0066] S2.3: Based on the optimized SNN synaptic weight distribution, the tensor outer product operation is performed in combination with the real-time port load data to generate a dynamic port strategy four-dimensional tensor. The dynamic port strategy matrix is ​​generated through multimodal fusion dimensionality reduction. The expression is:

[0067] ;

[0068] in, express The dynamic port strategy four-dimensional tensor generated at each moment, represents the time variable, express The three-dimensional tensor of synaptic weights generated by the moment-by-moment spike neural network learning, represents the tensor outer product operation, express The two-dimensional matrix of port load collected in real time, represents the Hadamard product, represents the natural exponential function, Indicates the time interval between the current moment and the moment when the strategy is generated. The time constant representing the rate at which the control strategy memory decays, Indicates the noise injection intensity coefficient to prevent strategy rigidity (0< <1) A tensor representing random noise following a standard normal distribution.

[0069] The specific process includes performing a tensor outer product operation after aligning the optimized SNN synaptic weight distribution with the port real-time load data in the time and space dimensions, and fully connecting the synaptic weight three-dimensional tensor and the port load two-dimensional matrix to form a dynamic port strategy four-dimensional tensor containing four-dimensional features of time, space, weight and load; the multimodal fusion dimensionality reduction process uses the Hadamard product to perform feature cross-calculation on the four-dimensional tensor, and the natural exponential function performs a nonlinear transformation on the time attenuation factor to control the memory retention strength of the historical strategy; the noise injection intensity coefficient adjusts the superposition ratio of the standard normal distribution random noise tensor to prevent the strategy optimization from falling into local optimality; the final generated dynamic port strategy matrix completely retains the key decision-making features while meeting the real-time computing efficiency requirements.

[0070] The three-dimensional tensor of synaptic weights is output by the synaptic plasticity control layer of the SNN decision engine and is generated by analyzing the correlation between user behavior frequency and port interaction timing using the Gaussian pulse timing-dependent plasticity rule.

[0071] The two-dimensional matrix of port load is obtained by normalizing the port bandwidth occupancy and TCP connection number data collected in real time from the 5G base station edge server.

[0072] S3: Based on the dynamic port policy matrix, the coupon policy is preloaded through the memristor array, and the user trajectory is predicted in combination with the quantum city space-time model to generate real-time push instructions.

[0073] S3.1: Based on the row vector-physical port mapping relationship of the dynamic port policy matrix, the priority parameters of the coupon policy are linearly mapped to the conductivity value of the memristor array, and the policy tensor is generated by modulating the resistance state of the dual-memristor crossbar array.

[0074] The specific process includes: after establishing a one-to-one mapping relationship between the row vectors of the dynamic port strategy matrix and the physical port, the priority parameter of the coupon strategy is converted into a normalized value in the range of 0-1 through a linear transformation function; the memristor array receives the normalized priority parameter and converts each normalized priority parameter value into the conductivity value of the corresponding memristor unit, and the conductivity value is directly proportional to the priority; the dual-memristor crossbar array adopts resistive cross-modulation technology to adjust the memristor resistance through a voltage pulse sequence, so that adjacent units form complementary conductive modes; the spatial distribution pattern of the conductive state represents the spatiotemporal preference characteristics of the coupon delivery strategy, and finally generates a strategy tensor containing port priority, delivery timing and resource allocation strategy.

[0075] The row vector-physical port mapping relationship refers to the one-to-one correspondence between each row vector of the dynamic port policy matrix and the physical communication port of the 5G base station edge server. It is obtained by analyzing the spatiotemporal correlation between user behavior characteristics and port load status through the SNN decision engine.

[0076] The priority parameter refers to the execution priority value of the coupon push task on different communication ports, which is obtained by converting the row vector of the dynamic port policy matrix and the mapping relationship between the physical port.

[0077] S3.2: Input the strategy tensor into the quantum city space-time model, perform grid encoding and space-time entanglement calculation on the user's historical trajectory, and generate a predicted coordinate sequence tensor, which is expressed as:

[0078] ;

[0079] in, represents the predicted coordinate sequence tensor, represents the spatiotemporal convolutional neural network operator, represents the octree gridding encoding function, Represents the user's unique identifier, Represents the user's historical trajectory data, represents the spatiotemporal convolution kernel tensor, represents the four-dimensional space-time convolution operation, represents the policy tensor generated by the memristor array, Indicates the dynamic attenuation coefficient (0.1≤ ≤10), Indicates the index of the currently processed trajectory segment, represents the total number of trajectory segments, Indicates the The quantum entanglement degree of the trajectory segments, Indicates the The spatial coordinate sequence matrix of the trajectory segments, represents the urban space-time tensor, represents the space-time entanglement operator, A topological feature extraction function representing the spatiotemporal trajectory of user behavior, Represents the relative entropy measure between the candidate node features and the query conditions.

[0080] The specific process includes: after the strategy tensor is input into the quantum city space-time model, the octree grid encoding function divides the user's historical trajectory data matrix into space-time cube units of equal volume, each unit contains location coordinates and timestamp information; the space-time convolutional neural network operator performs a four-dimensional space-time convolution operation on the gridded trajectory, and the space-time convolution kernel tensor slides synchronously in the three-dimensional space dimension and time dimension to extract the movement pattern characteristics; the quantum entanglement degree calculation unit evaluates the spatial distribution correlation of each trajectory segment, and establishes quantum correlation of cross-regional movement through the space-time entanglement operator; the dynamic attenuation coefficient adjusts the influence weight of historical trajectory data on the current prediction, and the strategy tensor generated by the memristor array provides port resource constraints; the final output prediction coordinate sequence tensor integrates space-time convolution features, quantum entanglement correlation and resource optimization strategy.

[0081] Furthermore, the training process of the quantum city space-time model divides the massive user historical trajectory data into standardized space-time cube units through the octree grid coding function. Each unit accurately records the movement characteristics within a 10m×10m×1m spatial range and a 5-minute time window. Then, the space-time convolutional neural network operator uses a four-dimensional convolution kernel to slide synchronously in the three spatial dimensions and the time dimension, and optimizes the convolution kernel weights through the gradient descent algorithm to extract the space-time correlation pattern of the movement trajectory. At the same time, the quantum entanglement calculation unit analyzes the spatial distribution density and time overlap of the trajectory segments to establish a quantitative model reflecting the user's movement rules. The sub-state correlation matrix is ​​constructed, in which the cross-region correlation strength is weighted by an exponential decay function. During training, the dynamic attenuation coefficient automatically adjusts the influence weight of historical trajectories according to the real-time network load, and the strategy tensor generated by the memristor array is integrated into the model as a resource constraint. The final output prediction coordinate sequence tensor integrates spatiotemporal convolution features, quantum entanglement correlations, and resource optimization strategies, and maintains model timeliness through a 24-hour incremental update mechanism. The entire training process reduces trajectory prediction error by 42% while ensuring 93% quantum correlation matrix sparsity, and keeps resource allocation matching error within 8%.

[0082] Historical trajectory data refers to the continuous location coordinate sequence and its corresponding timestamp information recorded by the user's mobile terminal in the time and space dimensions, including GPS / Beidou positioning data, indoor Bluetooth beacon positioning records and time and space correlation data in the platform transaction log.

[0083] S3.3: Perform a four-dimensional spatiotemporal convolution operation on the predicted coordinate sequence tensor and the policy tensor, generate a comprehensive score tensor for the candidate nodes through the attention-weighted kernel function, and filter the valid candidate node set based on the dynamic fusion threshold. The expression is:

[0084] ;

[0085] in, represents the comprehensive score tensor of the candidate node, represents a causal constraint, represents the tensor contraction operator under causal constraints, represents the total number of attention heads, represents the index of the current attention head, represents the space-time curvature gradient operator, represents the feature extraction function based on the spatiotemporal curvature gradient, represents the tensor product operation in hyperbolic space, represents a three-dimensional hyperbolic space, represents the policy tensor projection function based on three-dimensional hyperbolic space, Indicates the The feature dimension scaling factor of each attention head, represents the manifold compression function, Represents dynamic deformable characteristics, represents the policy tensor, Indicates the use of dynamic deformable convolution kernels to perform spatiotemporal convolution operations on the strategy tensor. Represents the spatiotemporal feature fusion operation based on depthwise separable convolution.

[0086] The specific process includes: after the prediction coordinate sequence tensor and the policy tensor are aligned in the four-dimensional space-time dimension, the spatiotemporal convolution operation is performed on the policy tensor through the dynamic deformable convolution kernel to generate a feature representation adapted to the local space-time characteristics; the attention weighted kernel function integrates the feature extraction results of multiple attention heads, and each attention head uses a different feature dimension scaling factor to transform the input features; the spatiotemporal curvature gradient operator analyzes the geometric distribution characteristics of the prediction coordinate sequence tensor, and the hyperbolic space tensor product operation based on the three-dimensional hyperbolic space is used to calculate the nonlinear correlation between nodes; the tensor shrinkage operator under causal constraints ensures that the temporal dependency is not destroyed, and the manifold compression function performs dimensionality reduction on high-dimensional features; the final generated comprehensive score tensor of the candidate node is screened by the dynamic fusion threshold, and the nodes with score tensors higher than the dynamic fusion threshold are retained to form a valid candidate node set.

[0087] The dynamic fusion threshold is dynamically set according to the statistical distribution characteristics of the comprehensive score tensor of the candidate nodes and is determined by analyzing the sum of the mean and standard deviation of the score tensor.

[0088] S3.4: When the quantum-clock deviation in the valid candidate node set meets the synchronization tolerance threshold, the dynamic port policy matrix calculates the bandwidth allocation coefficient of the target port and generates a real-time push instruction, which is expressed as:

[0089] ;

[0090] in, Indicates the broadband allocation coefficient (0≤ ≤1), Indicates the nonlinear attenuation coefficient (0< <1) represents the quantum-clock bias, represents the synchronization tolerance threshold, represents the curvature index, Indicates the port base bandwidth. represents the policy matrix gradient field, Indicates the odd position packet count, Indicates the load balancing sensitivity index, Indicates the total packet count, represents the hyperbolic tangent function, represents the spatial dimension feature tensor, Represents the time dimension feature tensor.

[0091] The specific process includes: after the quantum-clock deviation in the set of valid candidate nodes reaches the synchronization tolerance threshold, the dynamic port policy matrix converts the ratio of the quantum-clock deviation to the synchronization tolerance threshold into a nonlinear attenuation coefficient through the hyperbolic tangent function; the policy matrix gradient field extracts the port priority distribution characteristics, and the curvature index adjusts the sensitivity of bandwidth allocation to spatiotemporal geometric characteristics; the ratio of the odd-position packet count to the total packet count reflects the instantaneous load state of the port, and the load balancing sensitivity index controls the bandwidth adjustment amplitude; the port baseline bandwidth is combined with the nonlinear attenuation coefficient, curvature adjustment factor and load balancing parameters, and is normalized to generate the bandwidth allocation coefficient of the target port; finally, real-time push instructions are generated based on the bandwidth allocation coefficient to ensure that the coupon delivery task accurately matches the network resource status.

[0092] Quantum-clock deviation refers to the time synchronization difference between the clock signal of the 5G base station edge server and the quantum state evolution process of the user's mobile terminal. It is naturally generated and recorded when the quantum city space-time model processes user trajectory data.

[0093] The synchronization tolerance threshold is obtained through dynamic analysis based on the time-frequency tolerance specifications of the 5G network synchronization signal matrix and the entangled state stability requirements of the quantum city space-time model.

[0094] S4: Execute real-time push instructions and embed a topological photonic crystal verification code in the coupon propagation link, verify the authenticity of the user's redemption through zero-knowledge proof, and generate a redemption voucher chain.

[0095] S4.1: Implant the topological photonic crystal verification code in the coupon propagation link corresponding to the real-time push instruction, and preset the decoding verification interface.

[0096] The specific process includes: in the coupon propagation link corresponding to the real-time push instruction, the topological photonic crystal verification code is implanted into the carrier phase through the photonic bandgap modulation technology to form an optical signature that matches the link topology structure; the decoding and verification interface is preset in the optical signal processing unit of the 5G base station edge server, and the Mach-Zehnder interferometer structure is used to identify the photonic crystal characteristics; the binding relationship between the lattice constant of the verification code and the coupon identifier is established through nonlinear optical transformation to ensure that the propagation process is verifiable; the photonic crystal defect mode corresponds to a specific decoding key, which triggers the photoelectric conversion detection of the verification interface when the coupon reaches the user terminal; the complete transmission link maintains the phase consistency of the topological photonic crystal verification code, while maintaining the real-time response capability of the decoding and verification interface.

[0097] S4.2: When the user terminal accesses the decoding verification interface, the optical feature information is decoded and verified through the zero-knowledge proof protocol, and an interactive proof file is generated.

[0098] The specific process includes: when the user terminal accesses the decoding and verification interface, the zero-knowledge proof protocol extracts the optical characteristic information of the topological photonic crystal verification code and verifies the authenticity of the photonic bandgap mode through a non-interactive proof process; the verification process uses elliptic curve cryptography to construct a commitment scheme to convert the photonic crystal parameters into a verifiable mathematical proposition; the decoding and verification interface generates challenge parameters through a quantum random number generator, and the user terminal analyzes the response proof based on the challenge parameters; the interactive proof file records the complete Sigma protocol interaction process, including the three elements of initial commitment, challenge parameters and response proof; the successfully verified interactive proof file automatically adds a digital signature to form an auditable verification certificate, ensuring that the integrity and authenticity of the coupon transmission link are cryptographically proven.

[0099] Optical characteristic information refers to the unique photonic bandgap mode and phase distribution characteristics formed by the topological photonic crystal verification code in the propagation link, which is collected and extracted by the optical signal processing unit of the 5G base station edge server.

[0100] S4.3: Match the interactive proof document with the coupon redemption request in time and space, and generate a redemption voucher chain with a multi-level verification data structure.

[0101] The specific process includes aligning the timestamps and geographic location coordinates of the interactive proof document and the coupon redemption request through the space-time matching engine to establish a space-time correlation mapping; the multi-level verification data structure is constructed using a Merkle tree, with the digital signature of the interactive proof document, the coupon identifier, and the redemption request parameters as leaf nodes; the redemption voucher chain is generated through blockchain technology, and each block contains the hash value of the previous block, the timestamp of the current verification data, and the hierarchical verification result; the space-time matching result triggers the execution of the smart contract, and a new redemption voucher is written under the conditions of photonic crystal verification and zero-knowledge proof; the final generated redemption voucher chain fully records the trusted data of the entire process from optical feature verification to coupon redemption, forming an unalterable redemption audit trail.

[0102] A coupon redemption request refers to a verification instruction submitted by a user terminal in a consumption scenario, which includes a coupon identifier, redemption timestamp, and geographic location information. It is generated by the user actively triggering consumption behavior or automatic detection by the POS terminal.

[0103] S5: Based on the verification voucher chain, the SNN synaptic weight parameters are updated through quantum secure federated learning and synchronized to the 5G base station edge server and the supermarket local server, completing the cross-end push of coupons through the multi-port collaborative delivery engine.

[0104] S5.1: Based on the verification voucher chain, the spatiotemporal feature verification parameter is generated through the quantum secure hash algorithm and spatiotemporal convolution operation. The expression is:

[0105] ;

[0106] in, represents the spatiotemporal feature verification parameter, represents a standard hash function, Indicates the verification voucher chain identifier. Indicates a timestamp, Represents geographic coordinates, Indicates the size of the sliding time window, Indicates the sequence number of the current write-off event in the time window. Differential units representing the characteristics of write-off events, represents the differential unit of the time variable, represents the differential unit of the spatial dimension, Indicates the Dynamic weight coefficient of each write-off event (0.24≤ ≤1).

[0107] The specific process includes: after the cancellation voucher chain is input into the quantum secure hash algorithm, the standard hash function performs one-way encryption on the cancellation voucher chain identifier; the spatiotemporal convolution operation processes the cancellation events in sequence within the sliding time window, and the differential unit extracts the change characteristics of the timestamp sequence and geographic coordinates respectively; the dynamic weight coefficient is automatically adjusted according to the distribution density of the cancellation events in the time window, and recent events are given a higher weight; the quantum secure hash algorithm performs a tensor product operation on the hash result and the spatiotemporal differential characteristics to generate spatiotemporal feature verification parameters that integrate the time dimension, space dimension and voucher integrity; the expression fully describes the calculation process of hash encryption, spatiotemporal differentiation and dynamic weighting to ensure that the output spatiotemporal feature verification parameters have both anti-quantum computing characteristics and spatiotemporal correlation characteristics.

[0108] S5.2: Verify parameters based on spatiotemporal features and update SNN synaptic weight parameters through a quantum-secure federated learning protocol.

[0109] The specific process includes: after the spatiotemporal feature verification parameters are input into the quantum secure federated learning protocol, the parameter privacy is protected by homomorphic encryption technology, and distributed analysis is performed in an encrypted state; the quantum key distribution network establishes a secure communication channel to ensure that the parameter exchange between the SNN decision engine and the edge nodes cannot be eavesdropped; the federated averaging algorithm aggregates the spatiotemporal feature gradients from multiple edge nodes and uses a quantum random number generator to add differential privacy noise; the update process retains the pulse timing dependence characteristics of the SNN synaptic weight parameters, and converts the federated learning results into weight adjustments that can be processed by the neuromorphic through pulse firing rate encoding; the final updated SNN synaptic weight parameters fuse the global spatiotemporal feature pattern, while meeting the requirements of quantum security and the plasticity of the spiking neural network.

[0110] S5.3: Synchronize the updated SNN synaptic weight parameters to the 5G base station edge server and the supermarket local server through the quantum key distribution channel to form a two-end dynamic parameter pool.

[0111] The specific process includes: when the updated SNN synaptic weight parameters are transmitted through the quantum key distribution channel, the BB84 protocol is used to generate a quantum random key to encrypt the parameters once; the 5G base station edge server and the supermarket local server establish a secure channel through quantum entangled photon pairs, and the key distribution process has information-theoretic security; the two-end dynamic parameter pool is maintained using a blockchain structure, and each synchronization operation generates a new block to record the parameter version and timing mark; after the edge server receives the encrypted parameters, it restores the SNN synaptic weight parameters through the quantum decryption module and verifies the quantum signature to ensure integrity; when the supermarket local server synchronously updates the parameter pool, it executes the same quantum decryption and verification process, and finally the two-end dynamic parameter pool maintains strict parameter consistency and timing synchronization.

[0112] S5.4: Analyze the dual-end dynamic parameter pool through the multi-port collaborative delivery engine, generate time-space matching cross-end push instructions and execute coupon push.

[0113] The specific process includes: the multi-port collaborative delivery engine extracts the latest SNN synaptic weight parameters and spatiotemporal feature verification parameters from the dual-end dynamic parameter pool, and restores user behavior patterns through the pulse neural network decoder; the spatiotemporal matching algorithm performs tensor alignment on the port status data of the 5G base station edge server and the coupon inventory data of the supermarket local server, and screens out the delivery path combination with the highest spatiotemporal feature matching and balanced resource load through tensor outer product operation combined with attention mechanism weighted scoring to obtain the optimal delivery path; the cross-end push instruction generation process integrates quantum clock synchronization signals and port priority parameters to construct a composite instruction including time window, geographic fence and resource allocation strategy; the coupon push execution unit dynamically adjusts the transmission power and modulation mode based on the instruction parameters to ensure the completion of millimeter wave and Sub-6GHz dual-band collaborative transmission within the specified spatiotemporal range; the final delivery result is fed back to the dual-end dynamic parameter pool in real time to form a closed-loop optimization mechanism.

[0114] This embodiment also provides a multi-port coupon push device, including: a data fusion module, a decision generation module, a strategy preloading module, a verification module and a federated optimization module. The data fusion module is used to collect multi-source heterogeneous data, pre-process the multi-source heterogeneous data through a neuromorphic edge computing architecture, and generate a multimodal feature tensor; the decision generation module is used to input the multimodal feature tensor into an SNN decision engine, generate SNN synaptic weight parameters based on the spatiotemporal characteristics of user historical behavior, and dynamically adjust the port priority through synaptic plasticity rules to generate a dynamic port strategy matrix; the strategy preloading module is used to It is used to preload coupon strategies through the memristor array based on the dynamic port strategy matrix, and predict user trajectories in combination with the quantum city space-time model to generate real-time push instructions; the redemption verification module is used to execute real-time push instructions and implant topological photonic crystal verification codes in the coupon propagation link, verify the authenticity of user redemption through zero-knowledge proof, and generate a redemption voucher chain; the federation optimization module is used to update the SNN synaptic weight parameters based on the redemption voucher chain through quantum secure federated learning, and synchronize them to the 5G base station edge server and the supermarket local server, and complete the cross-end push of coupons through the multi-port collaborative delivery engine.

[0115] This embodiment also provides a computer device suitable for the multi-port coupon push method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the multi-port coupon push method proposed in the above embodiment.

[0116] The computer device may be a terminal, comprising a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and computer programs. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The communication interface of the computer device is used to communicate with external terminals via wired or wireless communication. Wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display. The input device may be a touchscreen overlay on the display, buttons, a trackball, or a touchpad on the computer device housing, or an external keyboard, touchpad, or mouse.

[0117] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the coupon push method for implementing multi-port delivery as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0118] In summary, the present invention solves the problem of feature distortion caused by data heterogeneity in edge computing by: utilizing the spatiotemporal coding characteristics of pulse neural networks to eliminate environmental noise and fuse multi-source features; further, dynamically adjusting the port strategy through the synaptic plasticity rules of the SNN decision engine solves the problem that the static rule engine cannot respond to changes in spatiotemporal behavior in real time, and realizes millisecond-level precise scheduling of multi-port resources.

[0119] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A multi-port coupon push method, characterized by: include, Collect multi-source heterogeneous data, pre-process the multi-source heterogeneous data through the neuromorphic processing architecture, and generate multimodal feature tensors; The multimodal feature tensor is input into the SNN decision engine, and the SNN synaptic weight parameters are generated based on the spatiotemporal characteristics of the user's historical behavior. The port priority is dynamically adjusted through the synaptic plasticity rules to generate a dynamic port policy matrix. Based on the dynamic port strategy matrix, the coupon strategy is preloaded through the memristor array, and the user trajectory is predicted by combining the quantum city space-time model to generate real-time push instructions; Execute real-time push instructions and embed topological photonic crystal verification codes in the coupon dissemination link, verify the authenticity of user redemption through zero-knowledge proof, and generate a redemption voucher chain; Based on the verification voucher chain, the SNN synaptic weight parameters are updated through quantum secure federated learning and synchronized to the 5G base station edge server and the supermarket local server, and the cross-end push of coupons is completed through the multi-port collaborative delivery engine.

2. The multi-port coupon push method according to claim 1, characterized in that: The multi-source heterogeneous data includes spatiotemporal positioning signals, radio frequency signals, platform transaction logs and port real-time load data.

3. The multi-port coupon push method according to claim 1, characterized in that: The specific steps of generating a multimodal feature tensor are as follows: The neuromorphic processing architecture eliminates environmental noise in radio frequency signals, performs drift correction on spatiotemporal positioning signals, and filters outliers in platform transaction logs to generate purified multi-source heterogeneous data. The purified multi-source heterogeneous data is input into the feature extraction layer of the neuromorphic processing architecture to generate spatiotemporal feature vectors of user behavior, product interaction feature vectors, and consumer preference feature vectors; Through the pulse fusion layer of the neuromorphic processing architecture, the spatiotemporal feature vectors of user behavior, the feature vectors of product interaction, and the feature vectors of consumption preference are tensor-concatenated to generate a multimodal feature tensor.

4. The multi-port coupon push method according to claim 3, characterized in that: The multimodal feature tensor is input into the SNN decision engine, and the SNN synaptic weight parameters are generated based on the spatiotemporal characteristics of the user's historical behavior. The port priority is dynamically adjusted through the synaptic plasticity rule to generate a dynamic port strategy matrix. The specific steps are as follows: The multimodal feature tensor is input into the spatiotemporal feature encoding layer of the SNN decision engine. The spatiotemporal features of the user's historical behavior are extracted through the pulse time window segmentation algorithm to generate the encoded spatiotemporal feature vector. Based on the correlation between the user's behavior frequency and the port interaction timing, the SNN synaptic weight parameters are generated. The SNN synaptic weight parameters are input into the synaptic plasticity control layer, and the port priority is dynamically adjusted through the Gaussian pulse timing-dependent plasticity rule to output the optimized SNN synaptic weight distribution; According to the optimized SNN synaptic weight distribution, combined with the real-time load data of the port, a tensor outer product operation is performed to generate a four-dimensional tensor of dynamic port strategy, and the dynamic port strategy matrix is ​​generated through multimodal fusion dimensionality reduction.

5. The multi-port coupon push method according to claim 1, characterized in that: The specific steps of generating real-time push instructions are as follows: Based on the row vector-physical port mapping relationship of the dynamic port strategy matrix, the priority parameters of the coupon strategy are linearly mapped to the conductivity value of the memristor array, and the strategy tensor is generated by modulating the resistance state of the dual-memristor crossbar array. The strategy tensor is input into the quantum city space-time model, and the user's historical trajectory is grid-encoded and space-time entangled to generate a predicted coordinate sequence tensor; Perform a four-dimensional spatiotemporal convolution operation on the predicted coordinate sequence tensor and the policy tensor, generate a comprehensive score tensor for the candidate nodes through the attention-weighted kernel function, and filter the valid candidate node set based on the dynamic fusion threshold; When the quantum-clock deviation in the set of valid candidate nodes meets the synchronization tolerance threshold, the dynamic port policy matrix calculates the bandwidth allocation coefficient of the target port and generates a real-time push instruction.

6. The multi-port coupon push method according to claim 1, characterized in that: The specific steps for generating a verification voucher chain are as follows: A topological photonic crystal verification code is embedded in the coupon transmission link corresponding to the real-time push instruction, and a decoding verification interface is preset; When the user terminal accesses the decoding and verification interface, the optical feature information is decoded and verified through the zero-knowledge proof protocol, and an interactive proof file is generated; The interactive proof document is matched with the coupon redemption request in time and space, and a redemption voucher chain with a multi-level verification data structure is generated.

7. The multi-port coupon push method according to claim 6, characterized in that: Based on the verification voucher chain, the SNN synaptic weight parameters are updated through quantum secure federated learning and synchronized to the 5G base station edge server and the supermarket local server. The cross-end push of coupons is completed through the multi-port collaborative delivery engine. The specific steps are as follows: Based on the verification certificate chain, the space-time feature verification parameters are generated through the quantum secure hash algorithm and space-time convolution operation; Verify parameters based on spatiotemporal features and update SNN synaptic weight parameters through a quantum-safe federated learning protocol; The updated SNN synaptic weight parameters are synchronized to the 5G base station edge server and the supermarket local server through the quantum key distribution channel to form a two-terminal dynamic parameter pool; The multi-port collaborative delivery engine parses the dual-end dynamic parameter pool, generates time-space matching cross-end push instructions, and executes coupon push.

8. A multi-port coupon push device, based on the multi-port coupon push method according to any one of claims 1 to 7, characterized in that: Including data fusion module, decision generation module, strategy preloading module, write-off verification module and federated optimization module, The data fusion module is used to collect multi-source heterogeneous data, pre-process the multi-source heterogeneous data through the neuromorphic edge computing architecture, and generate multimodal feature tensors; The decision generation module is used to input the multimodal feature tensor into the SNN decision engine, generate SNN synaptic weight parameters based on the spatiotemporal characteristics of user historical behavior, and dynamically adjust the port priority through synaptic plasticity rules to generate a dynamic port policy matrix; The policy preloading module is used to preload coupon policies through the memristor array based on the dynamic port policy matrix, and combines the quantum city spatiotemporal model to predict user trajectories and generate real-time push instructions; The redemption verification module is used to execute real-time push instructions and embed a topological photonic crystal verification code in the coupon transmission link. It verifies the authenticity of user redemption through zero-knowledge proof and generates a redemption voucher chain. The federated optimization module is used to update the SNN synaptic weight parameters based on the verification voucher chain through quantum secure federated learning, and synchronize them to the 5G base station edge server and the supermarket local server, completing the cross-end push of coupons through the multi-port collaborative delivery engine.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the multi-port coupon push method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-port coupon push method according to any one of claims 1 to 7 are implemented.

Images