Research On Quantum Genetic Algorithm And Its Application In Intrusion Detection

Quantum Genetic Algorithm (QGA) is a new optimum method that combines quantum computation with Genetic Algorithm. It has significant value in research and application. Quantum genetic algorithm has good global search capability, but its local search capability is bad. To solve this problem, in the dissertation, quantum genetic algorithm has been further studied and improved with clonal selection theory. The improved quantum genetic algorithm has better search capability. Meanwhile, in this dissertation, the improvement and application in intrusion detection of quantum genetic algorithm have been studied. The main contents of the dissertation are as follows:1. The research and development of the quantum genetic algorithm are analyzed, some basic concepts and principles of quantum computing are also discussed, and the basic principles of quantum genetic algorithm, such as quantum bits, quantum chromosome, the process and the specific implementation steps of quantum genetic algorithm are elaborated. And some common quantum genetic operator is studied.2. To enhance the local search capability of quantum genetic algorithm,a novel quantum genetic algorithm based on Immune memory clonal operator is proposed. This method introduces immune memory clonal strategy to enhance the local search capacity of the algorithm, and further improves the search efficiency of the algorithm by using dynamic step length in adjustment of rotation angle of quantum gates mechanism and quantum cross operator. The experiment takes 0-1 knapsack and TSP problem of combinatorial optimization as example, the experimental results show that the improved quantum genetic algorithm is better than the traditional genetic algorithms and quantum genetic algorithm, both in convergence speed and optimization results.3. To solve the problems of the traditional clustering algorithm in intrusion detection, in the dissertation, we apply quantum genetic algorithm to intrusion detection, and propose a new detection algorithm based on quantum genetic clustering. This method first uses clustering algorithm to establish the set of original clusters, then optimizes the set of original clusters with quantum genetic algorithm, and obtains the optimal clustering results to detect intrusion. The experimental result which uses the data sets of KDDCUP99 shows that this approach has better detection rate and false positive rate, and can detect unknown intrusions efficiently.