Data protection method based on genetic algorithm

Abstract

The invention discloses a data protection method based on a genetic algorithm. The method comprises the steps of predefining, uniform sampling set initialization, global non-dominated sorting, crossover operation design, mutation operation design, iteration strategy and data protection. The invention belongs to the field of data protection, and particularly relates to a data protection method based on a genetic algorithm, and the method comprises the steps: introducing a uniform sampling set initialization strategy, and carrying out the unbiased coverage in a whole feasible space; a small-world network-based self-adaptive pairing mechanism is introduced to perform pairing selection, so that low search efficiency caused by later excessive random is avoided; the data protection effect is improved; a dynamic retention probability and disturbance factor design interlace operation is introduced, an excellent data protection strategy is adaptively retained, and smooth transition of exploration-utilization tradeoff is provided in different iteration stages; and polynomial variation and local perturbation design variation operation are introduced to avoid local optimum, so that the data protection strategy is finer and more accurate, and the data protection security is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data protection, and in particular to a data protection method based on a genetic algorithm. Background Art

[0002] Data protection methods refer to a series of technical means used to ensure the confidentiality, integrity and availability of data, prevent data from unauthorized access, tampering or loss, and enable reliable recovery when needed. However, general data protection methods have problems such as imbalanced initial data protection strategies, serious undersampling of some strategy spaces, easy omission of potential high-quality solutions, limited neighborhood search and information exchange, and poor data protection effects; general data protection methods lack effective guidance, leading to the loss of excellent data protection strategies, insufficient search accuracy for data protection strategies, and low data protection security. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a data protection method based on genetic algorithm. In view of the problem that the initial data protection strategy of the general data protection method is unbalanced, some strategy spaces are seriously undersampled, potential high-quality solutions are easily missed, and neighborhood search and information exchange are limited, which leads to poor data protection effect, the scheme introduces a uniform sampling set initialization strategy to cover the entire feasible space unbiasedly, and provide rich and diverse inspirations for the subsequent search of data protection strategies; introduces an adaptive pairing mechanism based on a small-world network for mate selection, maintains high randomness in the early stage and strong local stability in the later stage, avoids the low search efficiency caused by excessive randomness in the later stage, and maintains the necessary neighborhood reshaping ability; thereby improving the data protection effect; in view of the problem that the general data protection method lacks effective guidance, resulting in the loss of excellent data protection strategies, and the search accuracy of data protection strategies is insufficient, which leads to low data protection security, the scheme introduces dynamic retention probability and perturbation factor to design crossover operations, adaptively retains excellent data protection strategies, and provides a smooth transition of exploration-utilization trade-offs at different iteration stages; introduces polynomial mutation and local perturbation to design mutation operations to avoid local optimality, thereby making the data protection strategy more precise and accurate, and improving data protection security.

[0004] The technical solution adopted by the present invention is as follows: The data protection method based on genetic algorithm provided by the present invention comprises the following steps:

[0005] Step S1: pre-definition;

[0006] Step S2: Initialize uniform sampling set;

[0007] Step S3: global non-dominated sorting;

[0008] Step S4: crossover operation design;

[0009] Step S5: mutation operation design;

[0010] Step S6: Iteration strategy;

[0011] Step S7: Data protection.

[0012] Further, in step S1, the pre-definition is to obtain the data to be protected, and each data protection decision individual x represents a data protection strategy, including encryption level, redundant backup ratio, storage node distribution, access control granularity and protection parameter settings; based on the encryption level, redundant backup ratio, storage node distribution, access control granularity and protection parameter settings, a parameter optimization space is established; an optimization target is defined, and the optimization target It is expressed as: ;in, is the data loss rate; It is resource consumption; is the access delay time.

[0013] Further, in step S2, the uniform sampling set initialization is to use uniform sampling set generation to initialize the data protection strategy; the uniform sampling set generation is expressed as: ; The data protection policy initialization is expressed as: ;in, is the initial good point value of the jth dimension of the i-th individual; i is the individual index, and j is the dimension index of the parameter optimization space; is the initial value of the jth dimension of the i-th data protection decision-making individual; and are the minimum and maximum values ​​of the jth dimension of the parameter optimization space, respectively.

[0014] Further, in step S3, the global non-dominated sorting is to sort all data protection strategies; construct a small-world network, define the number of nodes N, each data protection decision individual corresponds to a node in the network, and the total number of nodes is equal to the size of the data protection decision population; define the degree K, each node in the initial ring graph is connected to K / 2 neighbors on the left and right, with a total of K edges; define the adaptive reconnection probability p, expressed as: , where gen is the current iteration number; maxgen is the maximum iteration number; is the basic reconnection probability; γ is the exponential control factor; for each edge on the ring, after being disconnected with probability p, it is randomly connected to any other node in the network to obtain a small-world network; for the i-th data protection decision individual , take out its neighbor set in the small-world network, perform non-dominated sorting on the neighbor set, and select the top individual as the spouse of the data protection decision-making individual ; Define the domination relationship. If , and there is at least one u such that ; then is superior to .

[0015] Further, in step S4, the crossover operation design preferentially selects the parents with a higher ranking in the global non-dominated sorting, and obtains a new generation of data protection strategy individuals through the crossover operation; in the early stage of gen < maxgen / 3, linear crossover is used. The linear crossover is expressed as: ; ; where and are the values of the j-th dimension of the two new offspring data protection strategy individuals generated by the i-th data protection decision individual through the crossover operation in the early stage; is the value of the j-th dimension; is the value of the j-th dimension; in the later stage, optimal crossover is used. The optimal crossover is expressed as: ; ; where and are the values of the j-th dimension of the two new offspring data protection strategy individuals generated by the i-th data protection decision individual through the crossover operation in the later stage; is the value of the j-th dimension of a random data protection strategy individual; is the value of the j-th dimension of the data protection strategy individual with the highest ranking in the non-dominated sorting; is the perturbation factor; ; Introduce the dynamic retention probability , which is expressed as: ; where is the crossover distribution index; c is a value randomly selected from a uniform distribution; each pair of parents generates two offspring, calculates the non-dominated ranks of the offspring respectively. If , only retain the better offspring; otherwise retain both; rand is a random number between 0 and 1.

[0016] Further, in step S5, the mutation operation design uses the Lévy mutation step size in the first half of gen < maxgen / 2. The step size is defined as: ; The update is expressed as: ; where RL is a random number obeying the Lévy distribution; is the value of the j-th dimension of the data protection strategy individual obtained by the mutation operation in the first half; is a random perturbation factor of the uniform distribution; in the second half, polynomial mutation is used for fine adjustment. The update is expressed as: ; ;in, is the value of the j-th dimension of the data protection strategy individual obtained by the second half of the mutation operation; is the polynomial variation value; is the multinomial distribution index; is a random number between 0 and 1, independent of rand; for each individual after mutation, a dimension is randomly selected to perform local perturbation, and the formula used is: ; ;in, and is the result after local perturbation; is the local perturbation factor, .

[0017] Furthermore, in step S6, the iteration strategy is to select the next generation of parent and child data protection strategy individuals according to non-dominated sorting in each generation, keep the population size unchanged, and repeat the crossover and mutation operations until the maximum number of iterations is reached or the optimization target converges.

[0018] Furthermore, in step S7, the data protection is to output the optimal data protection strategy individual after the iteration is completed, obtain the data protection configuration based on the individual position, and finally realize data protection.

[0019] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0020] (1) In view of the problem that the initial data protection strategy of general data protection methods is unbalanced, some strategy spaces are seriously undersampled, potential high-quality solutions are easily missed, and neighborhood search and information exchange are limited, which leads to poor data protection effect. This scheme introduces a uniform sampling set initialization strategy to cover the entire feasible space unbiasedly, providing rich and diverse inspiration for the subsequent search of data protection strategies; introduces an adaptive pairing mechanism based on a small-world network for mate selection, maintains high randomness in the early stage and strong local stability in the later stage, avoids inefficient search caused by excessive randomness in the later stage, and maintains the necessary neighborhood reshaping ability; thereby improving the data protection effect.

[0021] (2) In view of the problem that general data protection methods lack effective guidance, resulting in the loss of excellent data protection strategies, insufficient search accuracy for data protection strategies, and thus low data protection security, this scheme introduces dynamic retention probability and perturbation factor to design cross operations, adaptively retains excellent data protection strategies, and provides a smooth transition of exploration-utilization trade-offs at different iteration stages; polynomial mutation and local perturbation are introduced to design mutation operations to avoid local optimality, thereby making the data protection strategy more precise and accurate, and improving data protection security. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1A schematic diagram of the flow of the data protection method based on genetic algorithm provided by the present invention.

[0023] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0025] In the description of the present invention, it should be understood that terms such as "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred system or element must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as a limitation on the present invention.

[0026] Example 1, see Figure 1 The data protection method based on genetic algorithm provided by the present invention comprises the following steps:

[0027] Step S1: pre-definition; obtaining the data to be protected and defining the data protection decision-making individuals;

[0028] Step S2: uniform sampling set initialization; using uniform sampling set generation to initialize data protection strategy;

[0029] Step S3: global non-dominated sorting; introducing adaptive reconnection probability and using network topology to implement non-dominated sorting of data protection strategy;

[0030] Step S4: crossover operation design; introducing a two-stage crossover strategy of early linear crossover and late optimal crossover;

[0031] Step S5: mutation operation design; introduce local perturbation to design a dual-mode mutation mechanism;

[0032] Step S6: Iteration strategy; reshaping the population based on non-dominated sorting;

[0033] Step S7: Data protection: individual data protection is implemented based on the optimal data protection strategy after the iteration.

[0034] Example 2, see Figure 1, this embodiment is based on the above embodiment. In step S1, the predefined is to obtain the data to be protected. It is assumed that each data protection decision individual x represents a data protection strategy, including encryption level, redundant backup ratio, storage node distribution, access control granularity and protection parameter settings; a parameter optimization space is established based on encryption level, redundant backup ratio, storage node distribution, access control granularity and protection parameter settings; an optimization target is defined, and the optimization target It is expressed as: ;in, It is the data loss rate, which measures the degree of data loss to its original state under the data protection strategy; is resource consumption, which measures the computing and storage resources consumed during the data protection process; It is the access delay time, which measures the delay time of data access under the data protection policy.

[0035] Example 3, see Figure 1 This embodiment is based on the above embodiment. In step S2, the uniform sampling set initialization is to ensure that the initial data protection strategy combination uniformly covers various possibilities and prevents the initial omission of potential high-quality solutions. The uniform sampling set generation is used to initialize the data protection strategy; the uniform sampling set generation is expressed as: ; The data protection policy initialization is expressed as: ;in, is the initial good point value of the jth dimension of the i-th individual; i is the individual index, and j is the dimension index of the parameter optimization space; is the initial value of the jth dimension of the i-th data protection decision-making individual; and are the minimum and maximum values ​​of the jth dimension of the parameter optimization space, respectively.

[0036] Example 4, see Figure 1 This embodiment is based on the above embodiment. In step S3, the global non-dominated sorting is to sort all data protection strategies; a small-world network is constructed, the number of nodes N is defined, each data protection decision individual corresponds to a node in the network, and the total number of nodes is equal to the size of the data protection decision population; the degree K is defined, and each node in the initial ring graph is connected to K / 2 neighbors on the left and right, with a total of K edges; the adaptive reconnection probability p is defined, which is expressed as: , to avoid large fluctuations in the network structure in the later stage, where gen is the current iteration number; maxgen is the maximum iteration number; is the basic reconnection probability; γ is the exponential control factor; for each edge on the ring, after being disconnected with probability p, it is randomly connected to any other node in the network to obtain a small-world network; for the i-th data protection decision individual , take out its neighbor set in the small-world network, perform non-dominated sorting on the neighbor set, and select the individual with the highest ranking as the spouse of the data protection decision individual ; Define the dominance relationship. If , and there is at least one u such that ; then is better than ; and are two data protection decision individuals defining the dominance relationship; is for any , all satisfy , corresponding to the optimization goal of step S1.

[0037] By performing the above operations, aiming at the problems existing in the general data protection method, such as unbalanced initial data protection strategies, serious undersampling in some strategy spaces, easy omission of potential high-quality solutions, limited neighborhood search and information exchange, and thus poor data protection effects, this solution introduces a uniform sampling set initialization strategy to perform unbiased coverage in the entire feasible space, providing rich and diverse inspirations for the subsequent search of data protection strategies; introduces an adaptive pairing mechanism based on the small-world network for spouse selection, maintaining high randomness in the early stage and strong local stability in the later stage, avoiding low search efficiency caused by excessive randomness in the later stage, and at the same time maintaining the necessary neighborhood reshaping ability; thereby improving the data protection effect.

[0038] Example Five, refer to Figure 1 , based on the above example, in step S4, the crossover operation design uses existing high-quality data protection configurations to guide the new generation of data protection strategy combinations and improve the protection quality; preferentially select the parent with a higher ranking in the global non-dominated sorting, and obtain the new generation of data protection strategy individuals through the crossover operation; use linear crossover in the early exploration of gen < maxgen / 3, and the linear crossover is expressed as: ; ; where, and are the values of the jth dimension of the two new offspring data protection strategy individuals generated by the crossover operation of the ith data protection decision individual in the early stage; is the value of the jth dimension; is the value of the jth dimension; in the later stage, use the preferred crossover, and the preferred crossover is expressed as: ; ; where, and are the values of the jth dimension of the two new offspring data protection strategy individuals generated by the crossover operation of the ith data protection decision individual in the later stage; is the value of the jth dimension of the random data protection strategy individual; It is the value of the j-th dimension of the data protection strategy individual with the highest ranking in non-dominated sorting; It is the perturbation factor; ; Introduce a dynamic retention probability , expressed as: ; Among them, is the crossover distribution index; c is a value randomly selected from a uniform distribution; two offspring are generated for each pair of parents, and the non-dominated ranks of the offspring are calculated respectively. If , only the better offspring is retained; otherwise, both are retained; rand is a random number between 0 and 1.

[0039] Example 6, refer to Figure 1 , this example is based on the above example. In step S5, the mutation operation design is to design a dual-mode mutation mechanism, perform large random jumps in the data protection parameter space, explore data protection strategies that have not been covered, and prevent local convergence; use the Lévy mutation step size in the first half of gen < maxgen / 2, and the step size is defined as: ; The update is expressed as: ; Among them, RL is a random number obeying the Lévy distribution; is the value of the j-th dimension of the data protection strategy individual obtained by the mutation operation in the first half; is a random perturbation factor of the uniform distribution; in the second half, polynomial mutation is used for fine adjustment, and the update is expressed as: ; ; Among them, is the value of the j-th dimension of the data protection strategy individual obtained by the mutation operation in the second half; is the polynomial mutation value; is the polynomial distribution index; is a random number between 0 and 1, independent of rand; for each mutated individual, a dimension is randomly selected to perform local perturbation to add random details to the individual, and the formula used is: ; ; Among them, and are the results after local perturbation; is the local perturbation factor, .

[0040] By performing the above operations, in order to address the problem that general data protection methods lack effective guidance, resulting in the loss of excellent data protection strategies, and insufficient search accuracy for data protection strategies, which in turn leads to low data protection security, this scheme introduces dynamic retention probability and perturbation factor to design cross operations, adaptively retains excellent data protection strategies, and provides a smooth transition of exploration-utilization trade-offs at different iteration stages; polynomial mutation and local perturbation are introduced to design mutation operations to avoid local optimality, thereby making the data protection strategy more precise and accurate, and improving data protection security.

[0041] Embodiment 7, see Figure 1 ,This embodiment is based on the above embodiment. In step S6, the iteration strategy is to select the next generation of the parent and child data protection strategy individuals according to non-dominated sorting in each generation, keep the population size unchanged, and repeat the crossover and mutation operations until the maximum number of iterations is reached or

[0042] Embodiment 8, see Figure 1 This embodiment is based on the above embodiment. In step S7, data protection is to output the optimal data protection strategy individual after the iteration is completed, obtain the data protection configuration based on the individual position, and finally realize data protection.

[0043] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0044] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0045] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A data protection method based on a genetic algorithm, characterized in that: The method comprises the following steps: Step S1: pre-definition; obtaining the data to be protected and defining the data protection decision-making individuals; Step S2: uniform sampling set initialization; using uniform sampling set generation to initialize data protection strategy; Step S3: global non-dominated sorting; introducing adaptive reconnection probability and using network topology to implement non-dominated sorting of data protection strategy; Step S4: crossover operation design; introducing a two-stage crossover strategy of early linear crossover and late optimal crossover; Step S5: mutation operation design; introduce local perturbation to design a dual-mode mutation mechanism; Step S6: Iteration strategy; reshaping the population based on non-dominated sorting; Step S7: Data protection: individual data protection is implemented based on the optimal data protection strategy after the iteration.

2. The data protection method based on genetic algorithm according to claim 1, characterized in that: In step S2, the uniform sampling set initialization is to use uniform sampling set generation to initialize the data protection strategy; the uniform sampling set generation is expressed as: ; The data protection policy initialization is expressed as: ;in, is the initial good point value of the jth dimension of the i-th individual; i is the individual index, and j is the dimension index of the parameter optimization space; is the initial value of the jth dimension of the i-th data protection decision-making individual; and are the minimum and maximum values ​​of the jth dimension of the parameter optimization space, respectively.

3. The data protection method based on genetic algorithm according to claim 2 is characterized in that: In step S3, the global non-dominated sorting is to sort all data protection policies; Construct a small-world network and define the number of nodes N. Each data protection decision-making individual corresponds to a node in the network, and the total number of nodes is equal to the size of the data protection decision-making population. Define the degree K. In the initial ring graph, each node is connected to K / 2 neighbors on the left and right, with a total of K edges. Define the adaptive reconnection probability p, expressed as: , where gen is the current iteration number; maxgen is the maximum iteration number; is the basic reconnection probability; γ is the exponential control factor; for each edge on the ring, after being disconnected with probability p, it is randomly connected to any other node in the network to obtain a small-world network; for the i-th data protection decision individual , take out its neighbor set in the small-world network, perform non-dominated sorting on the neighbor set, and select the top individual as the spouse of the data protection decision-making individual ; Define the dominance relationship, if , and there exists at least one u such that ;but Compare Better; and These are the two data protection decision-makers that define the dominance relationship.

4. The data protection method based on genetic algorithm according to claim 3 is characterized in that: In step S4, the crossover operation design preferentially selects the parents with a higher ranking in the global non-dominated sorting, and obtains a new generation of data protection strategy individuals through the crossover operation. In the early exploration when gen < maxgen / 3, linear crossover is used, and the linear crossover is expressed as: ; ; where and are the values of the j-th dimension of two new offspring data protection strategy individuals generated by the crossover operation for the i-th data protection decision individual in the early stage; is the value of the j-th dimension; is the value of the j-th dimension; In the later stage, optimal crossover is used, and the optimal crossover is expressed as: ; ; where and are the values of the j-th dimension of two new offspring data protection strategy individuals generated by the crossover operation for the i-th data protection decision individual in the later stage; is the value of the j-th dimension of a random data protection strategy individual; is the value of the j-th dimension of the data protection strategy individual with the highest ranking in the non-dominated sorting; is the perturbation factor; ; Introduce the dynamic retention probability , which is expressed as: ; where is the crossover distribution index; c is a value randomly selected from a uniform distribution; For each pair of parents, two offspring are generated, and the non-dominated ranks of the offspring are calculated. If , only the better offspring is retained; otherwise, both are retained; rand is a random number between 0 and 1.

5. The data protection method based on genetic algorithm according to claim 4 is characterized in that: In step S5, the mutation operation design uses the Lévy mutation step size in the first half of gen < maxgen / 2, and the step size is defined as: ; The update is expressed as: ; where RL is a random number obeying the Lévy distribution; is the value of the j-th dimension of the data protection policy individual obtained by the mutation operation in the first half; is a random perturbation factor of the uniform distribution; in the second half, polynomial mutation is used for fine adjustment, and the update is expressed as: ; ; where is the value of the j-th dimension of the data protection policy individual obtained by the mutation operation in the second half; is the polynomial mutation value; is the polynomial distribution exponent; is a random number between 0 and 1, independent of rand; for each mutated individual, a dimension is randomly selected to perform local perturbation, and the formula used is: ; ; where and are the results after local perturbation; is the local perturbation factor, .

6. The data protection method based on genetic algorithm according to claim 1, characterized in that: In step S1, the pre-definition is to obtain the data to be protected, and each data protection decision individual x represents a data protection strategy, including encryption level, redundant backup ratio, storage node distribution, access control granularity and protection parameter settings; based on encryption level, redundant backup ratio, storage node distribution, access control granularity and protection parameter settings, a parameter optimization space is established; an optimization target is defined, and the optimization target It is expressed as: ;in, is the data loss rate; It is resource consumption; is the access delay time.

7. The data protection method based on genetic algorithm according to claim 1, characterized in that: In step S6, the iteration strategy is to select the next generation of parent and child data protection strategy individuals according to non-dominated sorting in each generation, keep the population size unchanged, and repeat the crossover and mutation operations until the maximum number of iterations is reached or the optimization target converges.

8. The data protection method based on genetic algorithm according to claim 1, characterized in that: In step S7, the data protection is to output the optimal data protection strategy individual after the iteration is completed, obtain the data protection configuration based on the individual position, and finally realize data protection.

Images