Ma Yuanyuan

Simulation Research on Optimal Detection of Intrusion Node Information in Network

The optimal detection of intrusion node information in the network can guarantee the safe and stable operation of the network. When the intrusion node information is detected, it is necessary to obtain the optimal parameters of SVM according to the optimal acquisition path of the node to complete the detection of intrusion node information. The traditional method uses the ant colony to find the network node path, get the support vector machine parameters, but ignores the optimization of the parameters, resulting in the information detection results are not accurate. An improved detection method of intrusion node information based on attribute attack graph is proposed. The intrusion signal with intrusion node information is decomposed into IMF single frequency intrusion signal, and the state transition equation of the network intrusion detection system is obtained. The ant colony theory and the support vector machine parameters are merged, and the network intrusion check rate is used as the objective function, and the ant is changed to the ant, and the nodes on the optimal path are connected to get the SVM optimal parameters. Based on this parameter, the intrusion node information detection is completed. The experimental results show that the proposed method has high accuracy and can improve the accuracy of embedded computer network intrusion detection.

Simulation of Attack Signal Path Identification in Radio Network Information

It is possible to improve the safety performance of the radio network by accurately identifying the attack signal path in the radio network information. When the attack signal path identification is carried out, it is necessary to calculate the vulnerability of the attack signal path, obtain the time series sample set of the network attack path, train the sample set for different vulnerabilities, and complete the path recognition. The traditional method is to establish the attack graph model. It cannot calculate the vulnerability of the attack signal path and cannot carry out the training, which leads to the limitation of the path recognition and the low efficiency. An improved identification method of attack signal path in radio network information with improved attack graph is proposed. Firstly, the attack map of the traditional method is transformed according to the theory of graph theory, and the concept of vulnerability factor is introduced into the improved attack graph. The vulnerability of the attack route is improved, and then the sliding time window is used to construct the network attack path recognition Time series sample set, Ada-Boosting method is used to train the sample sets with different vulnerabilities, and the regression matrix is obtained by using the training results. Finally, the attack signal path identification in the radio network information is completed. The simulation results show that the improved method has high accuracy and can guarantee the smooth operation of the radio network.

Distributed GEP function mining on consistency merger in grid environment

Distributed function mining is an important field of distributed data mining. In order to solve local model merger of function mining in grid environments, this paper presents consistency merger of local function model (CMLFM). On the basis of CMLFM, distributed GEP function mining on consistency merger (DGEPFM-CM) is proposed which combines with grid service. Simulated experiments show that the time-consuming of DGEPFM-CM is less than traditional GEP. With the increasing of grid nodes, the global fitting error of DGEPFM-CM apparently decreases.