A mobile payment encryption system based on data random algorithm

Abstract

The application discloses a kind of mobile payment encryption systems based on data random algorithm, comprising: acquisition module is used to acquire multimodal data, and constructs cross-modal entropy source pool;Random seed generation module is used to extract feature vector from cross-modal entropy source pool, generates initial random seed;Timing adaptive random chain module is used to generate random factor sequence based on initial random seed recursion, and form random chain structure;Encryption module is used to jointly encode random factor sequence and the transaction data to be encrypted, generates dynamic encryption parameter set, and encrypts transaction data;Chain updating module is used to collect network environment parameters, and update timing adaptive random chain, generate new dynamic encryption parameter set to form chain evolution encryption data sequence.The application utilizes cross-modal entropy source and timing random chain, realizes mobile payment data dynamic encryption, with the advantages of high-entropy randomness, strong attack resistance and high transmission security.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of encryption technology, and in particular to a mobile payment encryption system based on data random algorithm. BACKGROUND

[0002] With the popularity of mobile payment, traditional encryption methods generally rely on a single pseudo-random number generator to generate keys to encrypt transaction data. Although such methods can meet the basic security needs of payment to some extent, the randomness source is single and the entropy value is limited, which is easy to be speculated or reconstructed under high-intensity attacks. In addition, existing technologies tend to be static in the encryption process, making it difficult to dynamically adjust according to real-time changes in user operation characteristics and network environment parameters in the payment scenario, resulting in insufficient key update frequency and limited attack resistance.

[0003] In the prior art, cross-modal data has not been fully utilized, and the generation of pseudo-random numbers cannot reflect user behavior characteristics and sensor environment information, lacking multi-dimensional entropy source support. At the same time, existing methods lack a time-series chain recursion mechanism, and cannot construct a continuously changing encryption parameter set through a dynamically evolving random factor sequence, which is easy to be speculated by attackers through historical data analysis, thereby reducing the security of payment data. Therefore, how to introduce cross-modal entropy sources and time-series adaptive random chains to improve the entropy level and unpredictability of encryption parameters has become a technical problem to be solved. SUMMARY

[0004] One object of the present application is to provide a mobile payment encryption system based on data random algorithm. The present application uses cross-modal entropy sources and time-series random chains to achieve dynamic encryption of mobile payment data, with the advantages of high-entropy randomness, strong attack resistance, and high transmission security.

[0005] According to the mobile payment encryption system based on data random algorithm of the present application, it comprises:

[0006] The acquisition module is used to acquire multi-modal data during the mobile terminal payment process and construct a cross-modal entropy source pool.

[0007] The random seed generation module is used to extract a feature vector from the cross-modal entropy source pool and call a data random algorithm to generate an initial random seed.

[0008] The time-series adaptive random chain module is used to recursively generate a random factor sequence based on the initial random seed to form a random chain structure that dynamically evolves over time.

[0009] The encryption module is used to jointly encode the random factor sequence and the transaction data to be encrypted to generate a dynamic encryption parameter set, and encrypt the transaction data accordingly to output encrypted transaction data.

[0010] A chain update module is configured to collect network environment parameters during payment data transmission and update a time-adaptive random chain to generate a new dynamic encryption parameter set to form a chain-evolving encrypted data sequence.

[0011] Optionally, the modules are connected through the following method:

[0012] Multi-modal data generated by the mobile terminal during the payment process is collected, time alignment and normalization processing are performed on the multi-modal data, and a disturbance factor is introduced to construct a cross-modal entropy source pool.

[0013] A feature vector is extracted from the cross-modal entropy source pool, a data random algorithm is called to perform disturbance mapping on the feature vector to generate an initial random seed, and the initial random seed is used as a chain starting point of the time-adaptive random chain.

[0014] A random factor sequence is generated based on the initial random seed, and a time-adaptive random chain is constructed.

[0015] The random factor sequence and the transaction data to be encrypted are jointly encoded to generate a dynamic encryption parameter set, and the dynamic encryption parameter set is called to encrypt the transaction data to obtain encrypted transaction data.

[0016] During the payment data transmission process, network environment parameters are collected in real time, the network environment parameters are input into the time-adaptive random chain to update the chain node state, a new dynamic encryption parameter set is generated based on the updated random factor sequence, and the new dynamic encryption parameter set is called to encrypt subsequent transaction data to form a chain-evolving encrypted data sequence, and the encryption is completed.

[0017] Optionally, the construction of the cross-modal entropy source pool includes the following specific steps:

[0018] Multi-modal data generated by the mobile terminal during the payment process is collected, the multi-modal data includes touch track feature parameters, operation force, input speed and acceleration data, and an original input vector is constructed based on a unified time index.

[0019] The original input vector is time-aligned and normalized to obtain a normalized input vector.

[0020] Instantaneous information entropy is calculated for each feature component in the normalized input vector.

[0021] The instantaneous information entropy is weighted and fused, the instantaneous information entropy of each feature component is multiplied by the corresponding weight and accumulated to obtain an intermediate output of the cross-modal entropy source pool.

[0022] A disturbance factor is introduced in the intermediate output of the cross-modal entropy source pool, the disturbance control coefficient and the disturbance factor are multiplied, and the multiplication result is added to the intermediate output of the cross-modal entropy source pool to obtain the cross-modal entropy source pool.

[0023] Optionally, the generation of the initial random seed comprises the following specific steps.

[0024] A nine-dimensional entropy value set is extracted from the cross-modal entropy source pool and constructed as a feature vector.

[0025] The feature vector is input into a data random algorithm, and each component of the feature vector is sequentially subjected to power function transformation, logarithmic function transformation and trigonometric function transformation, and the three types of transformation results are combined according to a preset weight to form an output vector after perturbation mapping.

[0026] The nine components of the output vector after perturbation mapping are weighted and accumulated through a weight coefficient, and the accumulated result is normalized to obtain the initial random seed.

[0027] Optionally, the construction of the time-series adaptive random chain comprises the following specific steps.

[0028] The initial random seed is taken as the starting point of the time-series adaptive random chain, and the initial random seed is taken as the random factor corresponding to the first chain node.

[0029] At the moment of the time series, a nine-dimensional entropy value set is extracted from the cross-modal entropy source pool.

[0030] A random factor recursion relationship is established, and the random factor of the current chain node is expressed as the joint result of the random factor of the previous chain node and the cross-modal entropy source pool at the current moment:

[0031] ;

[0032] Wherein, represents the random factor at time , represents the random factor at time , represents the time-series inheritance coefficient, the value range of which is between zero and one, represents the weight coefficient of the component, represents the component entropy value of the entropy source pool at time ;

[0033] The random factor of the current chain node is taken as the random factor of the chain node, and its value range is constrained by normalization to be between zero and one.

[0034] Sequentially recursively until time , a random factor sequence is formed, and a complete time-series adaptive random chain is constructed with the random factor sequence.

[0035] Optionally, the generation of the encrypted transaction data comprises the following specific steps:

[0036] Preprocessing the transaction data to be encrypted, dividing the complete transaction data into a plurality of transaction data units, normalizing and format-encoding each transaction data unit to obtain standardized transaction data units;

[0037] At the tth moment in time, extracting a random factor at the moment from the time-series adaptive random chain, combining the random factor with each standardized transaction data unit, performing a product operation on each standardized transaction data unit and the random factor, adding a corresponding offset factor, and performing a modulo operation on the obtained value with a preset modulo operation parameter to obtain a coding result;

[0038] Combining all coding results into a coding vector and combining with nine entropy value components output by the cross-modal entropy source pool at the moment, performing a bitwise XOR operation on each coding result and the value after nonlinear function transformation of the corresponding entropy value component to obtain a dynamic encryption parameter;

[0039] Combining all dynamic encryption parameters into a dynamic encryption parameter set, and performing an encryption operation on the standardized transaction data units based on the dynamic encryption parameter set, specifically, performing a product operation on each standardized transaction data unit and the corresponding dynamic encryption parameter, and taking the remainder of the preset modulo operation parameter to obtain an encrypted transaction data unit;

[0040] Combining all encrypted transaction data units into encrypted transaction data.

[0041] Optionally, the generation of the chain-evolved encrypted data sequence comprises the following specific steps:

[0042] Collecting network environment parameters during the payment data transmission process, the network environment parameters including transmission delay, bandwidth and packet loss rate, and concatenating into a network state vector, and setting a current time window;

[0043] Normalizing the network state vector using a minimum and maximum normalization method to obtain a standardized network state vector;

[0044] Introducing the standardized network state vector into the random factor recursive relationship of the time-series adaptive random chain, updating the recursive relationship, and the updated recursive relationship being the original recursive relationship plus the weighted sum of the three components of the standardized network state vector and the corresponding adjustment coefficient;

[0045] Performing normalization constraint on the calculation result of the updated recursive relationship to limit the value of the random factor to between zero and one as the random factor of the current moment chain node;

[0046] The updated random factor sequence is obtained by recursively deriving in chronological order until the end of the current time window, and the updated timing adaptive random chain is formed by the updated random factor sequence;

[0047] A new dynamic encryption parameter set is generated based on the updated timing adaptive random chain;

[0048] The new dynamic encryption parameter set is called to perform encryption processing on subsequent transaction data, generate a set of encryption transaction data dynamically evolving over time, and form a chain-evolving encryption data sequence.

[0049] The beneficial effects of the present application are:

[0050] The present application effectively solves the problems of single randomness source, insufficient entropy, low key update frequency and limited attack resistance in the prior art by introducing a combination mechanism of cross-modal entropy source pool and timing adaptive random chain. In the mobile payment process, the system not only collects touch trajectory feature parameters, operation force, input speed and acceleration data, but also further combines environmental noise to form a disturbance factor, so that the cross-modal entropy source pool has a feature vector output with multi-dimensional, high complexity and high entropy characteristics. In this way, even if the attacker masters the encryption algorithm logic, he cannot synchronously obtain the user behavior characteristics and dynamic environmental noise, thereby improving the unpredictability and anti-cracking ability of the key.

[0051] Secondly, in the generation process of the random seed, the present application realizes nonlinear disturbance mapping by using the combination of power function transformation, logarithmic function transformation and trigonometric function transformation, so that the output result has multiple nonlinear characteristics in numerical distribution. Combined with the dynamic adjustment of the weight coefficient, the generated initial random seed not only has a numerical range controlled between zero and one, but also maintains high entropy and high unpredictability, providing a solid foundation for subsequent chain recursion. Compared with the prior art, the present application introduces multi-dimensional entropy source and nonlinear disturbance in the random seed generation stage, so that the starting point of the encryption chain has very high randomness and anti-guessing ability.

[0052] In the chain evolution process, the timing adaptive random chain constructed by the present application dynamically updates the random factor through the recursive formula. Each chain node not only inherits the random characteristics of the previous moment, but also introduces the output of the cross-modal entropy source pool at the current moment, realizing adaptive evolution in the time dimension. Meanwhile, transmission delay, bandwidth and packet loss rate network environment parameters are introduced in the payment data transmission stage, and the chain updating module further drives the dynamic change of the random factor sequence. In this way, the encryption parameter set is no longer a static key, but a dynamic result continuously evolving over time and network state. The attacker must crack the entire chain, and the difficulty increases exponentially.

[0053] Compared with the existing encryption method relying on a single pseudo-random number, the application simultaneously introduces a double random protection mechanism in the spatial dimension and the time dimension, so that the mobile payment system has higher security, stronger attack resistance and better adaptability, and realizes the balance between high-entropy randomness and real-time encryption efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0054] The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application, and do not limit the application. In the drawings:

[0055] Fig. 1 A flow chart of a mobile payment encryption system based on a data random algorithm is provided for the application;

[0056] Fig. 2 A recursive structure schematic diagram of a time sequence adaptive random chain of a mobile payment encryption system based on a data random algorithm is provided for the application. DETAILED DESCRIPTION

[0057] The application will now be described in further detail with reference to the drawings. These drawings are simplified schematic diagrams, and only illustrate the basic structure of the application in a schematic manner, and therefore only show the components related to the application.

[0058] REFERENCE Figs. 1-2 A mobile payment encryption system based on a data random algorithm comprises:

[0059] A collection module is configured to collect multi-modal data during a mobile terminal payment process and construct a cross-modal entropy source pool;

[0060] A random seed generation module is configured to extract a feature vector from the cross-modal entropy source pool and call a data random algorithm to generate an initial random seed;

[0061] A time sequence adaptive random chain module is configured to recursively generate a random factor sequence based on the initial random seed and form a random chain structure dynamically evolving with time;

[0062] An encryption module is configured to jointly encode the random factor sequence and transaction data to be encrypted, generate a dynamic encryption parameter set, and encrypt the transaction data according to the dynamic encryption parameter set to output encrypted transaction data;

[0063] A chain updating module is configured to collect network environment parameters during a payment data transmission process and update the time sequence adaptive random chain to generate a new dynamic encryption parameter set to form a chain-evolving encrypted data sequence.

[0064] In this embodiment, the modules are realized through the following methods:

[0065] The multi-modal data generated by the mobile terminal in the payment process is collected, time alignment and normalization processing are performed on the multi-modal data, and a disturbance factor is introduced to construct a cross-modal entropy source pool;

[0066] A feature vector is extracted from the cross-modal entropy source pool, a data random algorithm is called to perform disturbance mapping on the feature vector, an initial random seed is generated, and serves as a chain starting point of a time-adaptive random chain;

[0067] A random factor sequence is generated based on the initial random seed, and a time-adaptive random chain is constructed;

[0068] The random factor sequence and the transaction data to be encrypted are jointly encoded to generate a dynamic encryption parameter set, and the transaction data is encrypted by calling the dynamic encryption parameter set to obtain encrypted transaction data;

[0069] In the payment data transmission process, network environment parameters are collected in real time, the network environment parameters are input into the time-adaptive random chain to update the chain node state, a new dynamic encryption parameter set is generated based on the updated random factor sequence, and the subsequent transaction data is encrypted by calling the new dynamic encryption parameter set, forming a chain-evolved encryption data sequence, and completing the encryption.

[0070] In the embodiment, the construction of the cross-modal entropy source pool includes the following specific steps:

[0071] The multi-modal data generated by the mobile terminal in the payment process is collected, the multi-modal data includes touch track feature parameters, operation force, input speed and acceleration data, and an original input vector is constructed based on a unified time index, the touch track feature parameters include track path length, average sliding speed, track curvature change rate and track direction angle fluctuation degree;

[0072] The trajectory path length refers to the total distance of the finger movement on the screen surface in one touch operation, which is accumulated by the coordinate difference of continuous trajectory points, the average sliding speed refers to the average speed of the finger sliding on the screen, which is obtained by dividing the trajectory path length by the touch duration, the trajectory curvature change rate refers to the change of the trajectory bending degree along the path length, which reflects the bending complexity of the user sliding trajectory, which is obtained by calculating the ratio of the radian value of the direction angle difference of adjacent points in the touch trajectory point set to the path length, and taking the average value of the calculation results of all trajectory points, the trajectory direction angle fluctuation degree refers to the change amplitude of the direction angle of the trajectory in the whole sliding process, which reflects the stability and irregularity of the user operation trajectory, which is obtained by calculating the sequence of the tangent direction angle of adjacent points in the touch trajectory point set, and performing statistical analysis on the variance of the sequence of the direction angle, the operation force refers to the pressure value generated by the finger on the screen during the touch operation, the input speed refers to the ratio of the finger displacement to the time during the input or sliding process on the screen, which reflects the operation speed and input rhythm of the user, and the acceleration data represents the linear acceleration value of the mobile terminal in the three-dimensional coordinate system.

[0073] The original input vector is time-aligned and normalized, and the normalization is to subtract the mean value of each feature component in the original input vector and divide it by the standard deviation to obtain a normalized input vector.

[0074] The instantaneous information entropy of each feature component in the normalized input vector is calculated.

[0075] The instantaneous information entropy is obtained by sampling each feature component of the normalized input vector within a preset time window, constructing a probability distribution based on the frequency of the sampling values, and summing the product of the probability of each value in the probability distribution and the logarithm of the value and taking the opposite number.

[0076] The instantaneous information entropy is weighted and fused, and the instantaneous information entropy of each feature component is multiplied by the corresponding weight and accumulated to obtain the intermediate output of the cross-modal entropy source pool.

[0077] The generation of the weight of the weighted fusion specifically comprises: for each feature component, collecting multi-modal data samples, and statistically processing the instantaneous information entropy of the feature component in all samples to obtain the average information entropy value and variance of the feature component; the average information entropy values and variances of all feature components are normalized respectively; the normalized average information entropy values and variances are weighted and combined according to a preset proportion coefficient to form an initial fusion weight of the feature component, the selection of the proportion coefficient is set according to system design requirements, and is used to balance the contribution degrees of long-term uncertainty and instantaneous fluctuation; the initial fusion weights of all feature components are subjected to summation constraint normalization, that is, each initial fusion weight is divided by the cumulative value of all initial fusion weights, so that a fusion weight set with a weight sum of one is obtained; the instantaneous information entropy of each feature component is multiplied by the corresponding fusion weight and accumulated, and the obtained result is the intermediate output of the cross-modal entropy source pool;

[0078] A disturbance factor is introduced into the intermediate output of the cross-modal entropy source pool, and a disturbance control coefficient and the disturbance factor are multiplied, and the multiplication result is added to the intermediate output of the cross-modal entropy source pool to obtain the cross-modal entropy source pool.

[0079] The disturbance control coefficient is a proportion parameter between 0 and 1, which can be set autonomously as needed, and is used to adjust the weight of the disturbance factor in the final output of the cross-modal entropy source pool, so as to ensure that there is enough randomness and not excessive fluctuation.

[0080] The generation of the disturbance factor specifically comprises: sampling an environmental noise segment through a microphone of a mobile terminal, and intercepting a plurality of frames of original audio signals within a preset time window; performing fast Fourier transform on the original audio signals to convert the time domain noise signals into frequency domain components, and extracting high-frequency energy distribution features to represent the uncertainty of the noise; performing normalization processing on the high-frequency energy distribution features to obtain a standardized noise vector; superimposing a random offset generated by a pseudo-random number generator in the standardized noise vector to avoid the correlation of the environmental noise in different time segments; taking the superimposed result as the disturbance factor, and performing dimension unification processing on the intermediate output of the cross-modal entropy source pool and the disturbance factor.

[0081] The cross-modal entropy source pool is obtained by calculating the instantaneous information entropy of the touch screen trajectory feature parameters, the operation force, the input speed and the acceleration data, and performing weighted fusion, and combining the disturbance factor to obtain an entropy value set for generating a high-entropy random seed.

[0082] In the embodiment, the generation of the initial random seed comprises the following specific steps:

[0083] A nine-dimensional entropy value set is extracted from the cross-modal entropy source pool and constructed as a feature vector.

[0084] The nine-dimensional entropy value set refers to a vector set composed of nine entropy value results output by the cross-modal entropy source pool in the processing process, each entropy value result corresponding to the instantaneous information entropy of a feature component in a specific time window, including: an entropy value result of trajectory path length, an entropy value result of average sliding speed, an entropy value result of trajectory curvature change rate, an entropy value result of trajectory direction angle fluctuation degree, an entropy value result of input speed, an entropy value result of operation force, and entropy value results of acceleration information in three-axis directions. The above nine entropy values are collectively subjected to unified time alignment, normalization processing, weighted fusion and disturbance enhancement to form the nine-dimensional entropy value set.

[0085] The feature vector is input into the data random algorithm, and power function transformation, logarithmic function transformation and trigonometric function transformation are sequentially performed on each component of the feature vector. The three types of transformation results are combined according to the preset weight to form an output vector after disturbance mapping. Specifically, for each component, the power function transformation value, the logarithmic function transformation value and the trigonometric function transformation value are sequentially calculated to obtain three different intermediate results. According to the preset weight parameter, the three types of transformation results are linearly combined in a weighted manner, i.e., the power function transformation value is multiplied by the corresponding weight, the logarithmic function transformation value is multiplied by the corresponding weight, the trigonometric function transformation value is multiplied by the corresponding weight, and then the three types of weighted results are summed to form the disturbance mapping result of the component. The disturbance mapping results of all components are sequentially arranged to form the output vector after disturbance mapping. The output vector has the nonlinear stretching characteristics of the power function, the dynamic compression characteristics of the logarithmic function and the periodic disturbance characteristics of the trigonometric function in the numerical distribution, thereby ensuring higher entropy level and unpredictability in the initial random seed generation process.

[0086] The nine components of the output vector after disturbance mapping are weighted and accumulated through weight coefficients, and the accumulated results are normalized to obtain the initial random seed. The initial random seed is a single numerical output, the value range of which is limited between zero and one, and has high entropy characteristics and high unpredictability.

[0087] The generation of the weight coefficient is specifically as follows: a stability index of each component in the cross-modal entropy source pool is calculated according to the entropy value fluctuation range of the component, the stability index being defined as the variance value of the component in a preset time window; the stability indexes of all components are normalized by dividing the variance value of each component by the sum of all variance values to obtain the relative fluctuation proportion corresponding to each component; the reciprocal of the relative fluctuation proportion is taken as the initial weight of the component, so that the weight of the component with larger fluctuation is reduced and the weight of the component with smaller fluctuation is increased; all initial weights are normalized again so that the sum of the nine weight coefficients is equal to one to form the final weight coefficient set.

[0088] In the embodiment, the construction of the time sequence adaptive random chain includes the following specific steps:

[0089] Taking an initial random seed as a starting point of a time-adaptive random chain, and taking the initial random seed as a random factor corresponding to a first chain node, the time-adaptive random chain is composed of a plurality of chain nodes, and each chain node corresponds to a time-encryption state;

[0090] At a time point of the time sequence, a nine-dimensional entropy value set is extracted from the cross-modal entropy source pool, each component of the nine-dimensional entropy value set representing an entropy source pool component entropy value at the time point

[0091] A random factor recursive relationship is established, and the random factor of the current chain node is expressed as a joint result of the random factor of the previous chain node and the cross-modal entropy source pool at the current time point:

[0092]

[0093] wherein, represents the random factor at the time point represents the random factor at the time point represents a time inheritance coefficient, and the value range is between zero and one, represents a weight coefficient of the i-th component, represents the i-th entropy source pool component entropy value at the time point The generation of the time inheritance coefficient is specifically that, within a preset time window, the autocorrelation coefficient of each component of the nine-dimensional entropy value set in the cross-modal entropy source pool between adjacent time points is calculated, the autocorrelation coefficients of the nine components are averaged to obtain a time correlation index of the cross-modal entropy source pool as a whole, the time correlation index is taken as an initial value of the time inheritance coefficient, and the initial value is normalized to form the time inheritance coefficient used in the random factor recursive formula; The random factor of the current chain node is taken as the random factor of the i-th chain node, and the value range thereof is limited to be constrained between zero and one through normalization;

[0094]

[0095] The time inheritance coefficient is taken as the initial value of the time inheritance coefficient, and the initial value is normalized to form the time inheritance coefficient used in the random factor recursive formula;

[0096] The time inheritance coefficient is taken as the initial value of the time inheritance coefficient, and the initial value is normalized to form the time inheritance coefficient used in the random factor recursive formula; ​​​​​​​, form a random factor sequence, and construct a complete time-adaptive random chain with the random factor sequence, wherein the time-adaptive random chain refers to taking an initial random seed as a chain starting point, continuously generating chain nodes according to a recursive formula in a time sequence, each chain node corresponding to a random factor, and being adjusted through a time inheritance coefficient and a weighting coefficient, thereby forming a random factor sequence dynamically evolving over time, and the random factor sequence forms a chain structure in time sequence.

[0097] In the embodiment, the generation of the encrypted transaction data includes the following specific steps:

[0098] The transaction data to be encrypted is preprocessed, the complete transaction data is divided into a plurality of transaction data units, each transaction data unit is normalized and format-coded to obtain a standardized transaction data unit;

[0099] At the tth moment in the time sequence, a random factor at the moment is extracted from the time-adaptive random chain, the random factor is combined with each standardized transaction data unit, a product operation is performed on each standardized transaction data unit and the random factor, an offset factor is added, the obtained value is subjected to a modulo operation with a preset modulo operation parameter, the modulo operation result is limited in a range of zero to the preset modulo operation parameter minus one, and is taken as an encoding result;

[0100] The generation of the offset factor is specifically calculating a mean value of nine entropy value components output by the cross-modal entropy source pool within a preset time window; measuring the difference of each transaction data unit from the mean value, and obtaining the absolute difference value between the transaction data unit value and the mean value; normalizing the difference measurement results of all transaction data units; mapping the normalized results to an offset factor set to ensure that each transaction data unit corresponds to a unique offset factor, thereby introducing a disturbance related to the noise level of the cross-modal entropy source pool in the joint encoding operation;

[0101] All encoding results are combined into an encoding vector, and are combined with the nine entropy value components output by the cross-modal entropy source pool at the moment, a bitwise XOR operation is performed on the value of each encoding result and the corresponding entropy value component after nonlinear function transformation, and a dynamic encryption parameter is obtained;

[0102] All dynamic encryption parameters are combined into a dynamic encryption parameter set, and an encryption operation is performed on the standardized transaction data units based on the dynamic encryption parameter set, specifically, a product operation is performed on each standardized transaction data unit and the corresponding dynamic encryption parameter, and a modulo operation is performed on the preset modulo operation parameter, to obtain an encrypted transaction data unit;

[0103] All encrypted transaction data units are combined into encrypted transaction data.

[0104] In the embodiment, the generation of the chain-evolved encrypted data sequence comprises the following specific steps:

[0105] Network environment parameters are collected in the payment data transmission process, the network environment parameters include transmission delay, bandwidth and packet loss rate, and are spliced into a network state vector, a current time window is set, the current time window is used to limit the time range of the random factor recursion, and the length of the current time window corresponds to the transmission period of a transaction data or a group of transaction data;

[0106] The network state vector is normalized by using the minimum and maximum normalization method to obtain a standardized network state vector;

[0107] The standardized network state vector is introduced into the random factor recursion relationship of the time-adaptive random chain, and the recursion relationship is updated, and the updated recursion relationship is the original recursion relationship plus the weighted sum of the three components of the standardized network state vector and the corresponding adjustment coefficients, the adjustment coefficients respectively represent the weights of the transmission delay, the bandwidth and the packet loss rate, and the specific generation steps include: for the transmission delay, the bandwidth and the packet loss rate, the variance value of each type of network parameter is calculated in the preset time window; the variance values of the three network parameters are normalized to obtain the relative fluctuation proportion; the reciprocal of the relative fluctuation proportion is taken; the three adjustment coefficients obtained are normalized again, so that the sum of the three adjustment coefficients is equal to one, to obtain the final adjustment coefficients, and ensure that the influence weights of the three types of network parameters on the random factor generation in the recursion formula are reasonably distributed;

[0108] The calculation result of the updated recursion relationship is normalized to limit the value of the random factor to between zero and one, as the random factor of the chain node at the current time;

[0109] The recursion is performed in time sequence until the end of the current time window, to obtain an updated random factor sequence, and the updated time-adaptive random chain is formed by the updated random factor sequence;

[0110] A new set of dynamic encryption parameters is generated based on the updated time-adaptive random chain;

[0111] The new set of dynamic encryption parameters is called to perform encryption processing on subsequent transaction data, to generate a set of encrypted transaction data dynamically evolved with time, and to form a chain-evolved encrypted data sequence;

[0112] After the mobile terminal completes the generation of the encrypted transaction data, the formed encrypted data sequence is transmitted to the payment server through a communication link, a cross-modal entropy source pool constructed synchronously with the mobile terminal is called at the payment server side, a set of entropy value components corresponding to the current time window is output, a time-adaptive random chain recursively constructed synchronously with the mobile terminal is called, the time-adaptive random chain is composed of multiple chain nodes, each chain node corresponds to a random factor, and a set of dynamic encryption parameters consistent with the mobile terminal is generated based on the set of entropy value components and the random factor, in the decryption process, each dynamic encryption parameter unit corresponding to each encrypted transaction data unit in the received encrypted data sequence is called to perform reverse operation, the reverse operation includes reverse mapping of the encrypted transaction data unit under the modulo operation parameter and performing range constraint to restore its numerical value to the pre-encryption state, so as to obtain the decrypted transaction data unit, all decrypted transaction data units are combined in time sequence into a decrypted transaction data set, and format restoration processing is performed on the set to restore to the original transaction data, so as to complete the encrypted data decryption and secure recovery based on the joint driving of the cross-modal entropy source pool and the time-adaptive random chain.

[0113] Embodiment 1:

[0114] In order to verify the feasibility of the application in implementation, the application is applied to the mobile payment platform of a branch of a commercial bank in East China, the bank processes a large number of payment transactions every day, covering multiple scenarios such as shopping, transportation, catering, medical treatment and the like. However, under the existing technical conditions, the traditional encryption method usually relies on a single pseudo-random number generator to generate a key to encrypt transaction data, the randomness of this method is single, the entropy value is limited, and attackers can capture transaction data for a long time, use statistical rules to try to guess the encryption parameter or reconstruct the key, so that the security of payment data is difficult to meet the requirements of high-risk network environment, and the application is proposed to overcome the shortcomings of the existing method in randomness, dynamics and attack resistance.

[0115] In the specific application process, when the user initiates a payment operation through the mobile terminal, the acquisition module first acquires multi-modal data, performs time alignment and normalization processing, and inputs a disturbance factor generated by the environmental noise segment collected by the terminal microphone to form a cross-modal entropy source pool, providing multi-dimensional high-entropy input for subsequent random seed generation.

[0116] After the cross-modal entropy source pool is constructed, the system extracts a feature vector therefrom, calls a data random algorithm to perform a nonlinear disturbance mapping, obtains an initial random seed, takes the random seed as a starting point of a time-adaptive random chain, and transmits and evolves the random seed in multiple chain nodes through a recursive formula. In the process of generating a random factor, each chain node not only inherits a random feature of a previous node, but also combines an entropy value component of the cross-modal entropy source pool at a current time, so that the random factor sequence has a dynamic change characteristic over time. Since the output results of the cross-modal entropy source pool at different times are highly related to user operations and environmental states, an attacker cannot calculate the random factor of a future chain node through historical data, and thus the key unpredictability is improved.

[0117] In the payment data encryption phase, the system performs joint encoding on the random factor sequence and the transaction data to be encrypted, generates a dynamic encryption parameter set, and encrypts the transaction data. The transaction data is divided into multiple transaction data units in the preprocessing link, each transaction data unit is combined with a random factor, a modulo operation and a nonlinear function transformation are performed, a dynamic encryption parameter is obtained, and the encryption is completed by using the parameter. The encrypted transaction data is continuously driven by the chain update module in the transmission process. The system collects network environment parameters including transmission delay, bandwidth fluctuation and packet loss rate in real time, and updates the chain node state by introducing these parameters into the time-adaptive random chain. New random factors generated as the network environment changes will further act on the encryption of subsequent data, forming a chain-evolved encrypted data sequence.

[0118] When the encrypted data sequence is transmitted to the payment server, the server side synchronously constructs the cross-modal entropy source pool and the time-adaptive random chain. Since the terminal and the server use the same data acquisition method and recursive logic, both parties can generate consistent random factor sequences and dynamic encryption parameter sets in the same time window. After receiving the encrypted data, the server calls the corresponding dynamic encryption parameter set to perform reverse decryption, restores the encrypted transaction data unit to the original transaction data, and performs format restoration processing on the restored data to ensure that the final transaction data is consistent with the data submitted by the user. In the whole process, there is no independent key distribution link, which fundamentally avoids the risk of key leakage.

[0119] In order to verify the performance of the application in implementation, the traditional method is compared.

[0120] Table 1 Comparison of experimental data of the method of the application and the traditional encryption method

[0121]

[0122] As can be seen from Table 1, in terms of randomness entropy value index, the result of the method of the application is 7.9, which is significantly improved compared with 5.8 of the traditional method, which shows that the multi-dimensional disturbance introduced by the cross-modal entropy source pool and the time sequence adaptive random chain effectively increases the entropy value, so that the encryption parameter is more unpredictable. This high-entropy characteristic is difficult to achieve by the traditional single pseudo-random number generation mechanism.

[0123] In terms of encryption delay, the average encryption delay of the method of the application is 25 milliseconds, which is slightly increased compared with 22 milliseconds of the traditional method, but this difference is within an acceptable range and will not affect the user experience in actual payment scenarios. In other words, while improving security, the application maintains high real-time performance.

[0124] In terms of anti-replay attack capability, the success rate of the method of the application reaches 91%, which is much higher than 62% of the traditional method. This is because the dynamic chain evolution of the random factor sequence makes it difficult for attackers to bypass the encryption verification by repeating old data, thereby enhancing the system's ability to resist replay attacks.

[0125] In terms of anti-pattern speculation, the application performs particularly outstandingly, with a success rate of 93%, which is significantly improved compared with 57% of the traditional method. This is due to the introduction of user behavior and environmental data by the cross-modal entropy source pool, which makes each encryption parameter related to the operation state and external noise. Even if the attacker masters the historical data, it is difficult to make regular speculation.

[0126] The above describes only the preferred specific embodiments of the application, but the protection scope of the application is not limited thereto. Any person skilled in the art, within the technical scope disclosed by the application, according to the technical solution and inventive concept of the application, can make equivalent replacement or change, which should be covered within the protection scope of the application.

Claims

1. A mobile payment encryption system based on a data randomization algorithm, characterized in that, include: The data acquisition module is used to collect multimodal data during mobile terminal payment and to build a cross-modal entropy source pool. The random seed generation module is used to extract feature vectors from the cross-modal entropy source pool and call the data randomization algorithm to generate the initial random seed; The time-adaptive random chain module is used to recursively generate a random factor sequence based on an initial random seed, forming a random chain structure that evolves dynamically over time. The encryption module is used to jointly encode the random factor sequence with the transaction data to be encrypted, generate a dynamic encryption parameter set, encrypt the transaction data accordingly, and output encrypted transaction data. The chain update module is used to collect network environment parameters during payment data transmission and update the time-adaptive random chain to generate a new set of dynamic encryption parameters to form a chain-evolving encrypted data sequence. The modules are connected in the following way: Collect multimodal data generated by mobile terminals during the payment process, perform time alignment and normalization processing on the multimodal data, introduce a perturbation factor, and construct a cross-modal entropy source pool; Feature vectors are extracted from the cross-modal entropy source pool, and a data randomization algorithm is called to perform perturbation mapping on the feature vectors to generate an initial random seed, which serves as the chain starting point of the time-series adaptive random chain. A random factor sequence is generated based on an initial random seed, and a time-adaptive random chain is constructed. The random factor sequence is jointly encoded with the transaction data to be encrypted to generate a dynamic encryption parameter set. The transaction data is then encrypted by calling the dynamic encryption parameter set to obtain encrypted transaction data. During the payment data transmission process, network environment parameters are collected in real time and input into the time-adaptive random chain to update the state of the chain nodes. A new set of dynamic encryption parameters is generated based on the updated random factor sequence, and the new set of dynamic encryption parameters is called to encrypt subsequent transaction data, forming a chain-evolving encrypted data sequence to complete the encryption. The construction of the time-adaptive random chain includes the following specific steps: The initial random seed is used as the starting point of the time-adaptive random chain, and the initial random seed is used as the random factor corresponding to the first chain node. In the time series At any given moment, a nine-dimensional entropy value set is extracted from the cross-modal entropy source pool; Establish a recursive relationship for random factors, representing the random factor of the current chain node as the joint result of the random factor of the previous chain node and the cross-modal entropy source pool at the current time: ； in, Indicates at time random factors, Indicates at time random factors, This represents the temporal inheritance coefficient, with a value ranging from zero to one. Indicates the first The weighting coefficients of each component, Indicates at time The Entropy values ​​of each entropy source pool component; Use the random factor of the current chain node as the first... The random factor of each chain node is constrained to a range of zero to one by normalization. Recursively calculate until time... A random factor sequence is formed, and a complete time-adaptive random chain is constructed using the random factor sequence.

2. The mobile payment encryption system based on a data randomization algorithm according to claim 1, characterized in that, The construction of the cross-modal entropy source pool includes the following specific steps: Multimodal data generated by mobile terminals during the payment process is collected. The multimodal data includes touch screen trajectory feature parameters, operation force, input speed and acceleration data, and an original input vector is constructed based on a unified time index. The original input vector is time-aligned and normalized to obtain a normalized input vector; Calculate the instantaneous information entropy for each feature component in the normalized input vector; The instantaneous information entropy is weighted and fused, and the instantaneous information entropy of each feature component is multiplied by the corresponding weight and accumulated to obtain the intermediate output of the cross-modal entropy source pool. A perturbation factor is introduced into the intermediate output of the cross-modal entropy source pool, and the perturbation control coefficient and the perturbation factor are multiplied together. The result of the multiplication is added to the intermediate output of the cross-modal entropy source pool to obtain the cross-modal entropy source pool.

3. The mobile payment encryption system based on a data randomization algorithm according to claim 1, characterized in that, The generation of the initial random seed includes the following specific steps; Extract a nine-dimensional entropy value set from the cross-modal entropy source pool and construct it as a feature vector; The feature vector is input into the data randomization algorithm, and each component of the feature vector is transformed by a power function, a logarithmic function, and a trigonometric function in sequence. The three transformation results are combined according to preset weights to form the output vector after perturbation mapping. The nine components of the output vector after perturbation mapping are weighted and accumulated using weight coefficients, and the accumulated result is normalized to obtain the initial random seed.

4. A mobile payment encryption system based on a data randomization algorithm according to claim 1, characterized in that, The generation of the encrypted transaction data includes the following specific steps: The encrypted transaction data is preprocessed by dividing the complete transaction data into several transaction data units, normalizing and formatting each transaction data unit to obtain standardized transaction data units. At time t of the time series, the random factor at that time is extracted from the time-adaptive random chain. The random factor is combined with each standardized transaction data unit. The product operation is performed on each standardized transaction data unit and the random factor. The corresponding offset factor is added. The resulting value is then subjected to a modulo operation with the preset modulo operation parameters. The modulo operation result is used as the encoding result. All encoding results are combined into an encoding vector, and then combined with the nine entropy components output by the cross-modal entropy source pool at that moment. A bitwise XOR operation is performed on the value of each encoding result and the corresponding entropy component after nonlinear function transformation to obtain the dynamic encryption parameters. All dynamic encryption parameters are combined into a dynamic encryption parameter set, and encryption operations are performed on standardized transaction data units based on the dynamic encryption parameter set. Specifically, a product operation is performed on each standardized transaction data unit and its corresponding dynamic encryption parameter, and the remainder is taken with respect to a preset modulo operation parameter to obtain the encrypted transaction data unit. Combine all encrypted transaction data units into encrypted transaction data.

5. A mobile payment encryption system based on a data randomization algorithm according to claim 1, characterized in that, The generation of the chain-evolution encrypted data sequence includes the following specific steps: During the payment data transmission process, network environment parameters are collected, including transmission latency, bandwidth, and packet loss rate, and are concatenated into a network state vector to set the current time window; The network state vector is normalized by using the minimum and maximum value normalization method to obtain a standardized network state vector; In the recursive relation of the random factors of the time-adaptive random chain, a standardized network state vector is introduced to update the recursive relation. The updated recursive relation is the original recursive relation plus the weighted sum of the three components of the standardized network state vector and the corresponding adjustment coefficients. The calculation results of the updated recursive relation are subjected to normalization constraints, limiting the value of the random factor to between zero and one, which is used as the random factor of the chain node at the current time. The process is repeated chronologically until the current time window ends, resulting in an updated random factor sequence. This updated random factor sequence is then used to construct an updated time-adaptive random chain. A new set of dynamic encryption parameters is generated based on the updated time-adaptive random chain; A new set of dynamic encryption parameters is invoked to perform encryption processing on subsequent transaction data, generating a set of encrypted transaction data that evolves dynamically over time, and forming a chain-evolving sequence of encrypted data.

Images