Optimization Algorithm Of S-box Based On Computational Intelligence Optimization

Block cipher is an important branch in modern cryptography research,the birth anddevelopment of which has great value of theory and wide application. In block cipheralgorithm,S-box is the only nonlinear components,it provides confounding effect whichis necessary for block cipher algorithm. Construction of safe and effective S-box is theimportant and difficult part of the design block cipher.In recent years, through the simulation natural ecological system mechanism tosolve complex optimization problem of intelligent bionic algorithm for traditionaloptimization techniques have appeared to handle the combinatorial optimizationproblem of provide the feasible solution. This paper are mainly of S-box related theoryand intelligent bionic algorithm theory to the further research of analysis, the basicalgorithm is designed on the basis of some optimization algorithm, mixed algorithm andfusion algorithm, and based on the algorithm of structure and S-box experimentalanalysis. The main contents include:Firstly, a kind of improved algorithm for S-box based on algebraic method.ForS-box,the period of affine transformed pair is short,the period of iterative-output isshort and the algebraic expression is less.Based on these characteristics,an improvedS-box was constructed. The simulation results show that kind of improved algorithm iseffective in cryptographic properties. In addition,a number of8′8S-boxes is providedbased on replacement scheme. And its feasibility be confirmed through the analysis.Secondly, a kind of optimization algorithm of S-boxes based on genetic algorithmwas discussed. By analysising the principle,conception and concrete steps of basalgenetic algorithm, we can improve the algorithm from primary group and geneticmethod. Then constitute application based on the improved genetic algorithm. Resultsof the experiments indicate that the S-box created by the improved algorithm has greatraise on the performance of cryptology such as degree of nonlinearity, differenceuniformity and alavanching. Meanwhile, rate of convergence and fitness value are alsoimproved.Thirdly,a kind of hybrid optimization algorithm based on genetic tabu search forS-box was discussed. In view of the fact that it’s rough for genetic algorithm to usefeedback in the system and other shortcomings,such as premature and lack of theability of hill-climbing.we bring in the tabu search algorithm which has special functionof memory,and come up with this hybrid algorithm,applying for S-Box structure. Aftertesting and verifying by experiment, the S-Box has great code features in provided bythis algorithm.Some people compare this S-Box with other algorithms,finding thatthis algorithm is good at searching the best method efficiently. Finally, a kind of fusion optimization algorithm based on genetic tabu ant colonyare discussed. Throgh the analysis and comparison for basic genetic algorithm、tabusearch algorithm and ant colony algorithm，for their respective advantages anddisadvantages，put forward genetic tabu ant colony fusion algorithm.Throgh thesimulation experiments of NP-hard30problem, show that the algorithm has a very goodsearching capability and efficiency of the optimization. The structural design of theS-box based on Integration algorithms, shows the algorithm generated S-box in thenonlinearity, difference evenness, avalanche of cryptography performance have verygood improvement through this experiment.