Abstract. First, this article analyzes the Application layer Distributed Denial of Service(DDoS)'s attack principle and characteristic. According to the difference between normal users' browsing patterns and abnormal ones, user sessions are extracted from the web logs of normal users and similarities between different sessions are calculated .Because traditional K-mean Clustering algorithm is easy to fail into local optimal, the Particle Swarm Optimization K-mean Clustering algorithm is used to generate a detecting model. This model can been used to detect whether the undetermined sessions are DDoS attacks or not. The experiment show that this method can detect attacks effectively and have a good performance in adaptability.

Numerical calculations were made for the film pressure and carried load of the hydrodynamic journal bearing with very large eccentricity ratios, that is, very low clearances when the combined effects of surface roughness and physically adsorbed layer were considered. The shaft was rotating with perfectly smooth surface, and the sleeve was stationary with the nanoscale sinusoidal surface roughness. The fluid piezo‐viscous effect was also incorporated. Owing to the coexistence of the nanoscale adsorbed layer and the macroscopic intermediate continuum fluid film, the multiscale approach was used to simultaneously solve this sandwich film lubrication problem. It was found that owing to low bearing clearances the effect of the adsorbed layer normally very significantly increases the lubricant film pressure, depending on the fluid‐bearing surface interaction; While at the same time both the surface roughness effect and the fluid piezo‐viscous effect are particularly significant compared to those classically calculated, owing to the local more severe film squeezing and the resulting local higher pressures. By considering the effect of the adsorbed layer, the surface roughness more significantly increases the load‐carrying capacity of the journal bearing for very large eccentricity ratios, especially for a strong fluid‐bearing surface interaction. It is suggested that in modeling the hydrodynamic journal bearing for large eccentricity ratios, the combined effects of the surface roughness, physically adsorbed layer and fluid piezo‐viscous property should be considered.

In "Internet+" era, how to real-time monitoring and build effective pre-warning systems of the network public opinion crisis has become a required courses for government departments and enterprises. This paper sufficiently considers the development，changes in laws and characteristics of the network public opinion crisis, and establishes a pre-warning index system with 7 indexes of network public opinion. Because of the final index redundancy and the lack of data, building the public opinion of Pre-Warning model based on PCA and SVM, and using PSO to optimize the parameters of SVM. Experiment shows the pre-warning index system of network public opinion and the model of PCA-PSO-SVM is effective and feasible.

No abstract