It is important to improve the real-time of online intrusion detection system in the early stage of network intrusion. Aiming at the early detection on network intrusion that detects the anomaly traffic at beginning phase of network attack, feature is extracted to describe the behavior of network invasion, and the algorithm of extraction is designed. An online intrusion detection system is represented based on the algorithm of GHSOM. Experimental result proves that most attacks’ early detected ratio is above 80% used by this method, and early detection optimizes speed and efficiency of online intrusion detection system.
In order to improve the performance of multi-core processor, based on traditional hardware accelerator, this paper presents a novel computing array design scheme. The communication ports between the computing array and the processor are mapped in the address space of extended register file, which makes the computing array and the processor tightly coupled. The computing units are connected by Network-on-Chip(NoC) which enables the computing array be flexibly configured and highly shared by the multi-core processor. A 1 024-point Fast Fourier Transform(FFT) and an H.264 decoder are implemented on the experimental platform, and results show that the scheme can improve the performance and power consumption significantly compared to pure software solution.
Device-to-Device(D2D) communication attracts much research attention due to its ability to increase the efficiency of cellular networks. However, performance of D2D communication in multi-cell cellular systems is not fully addressed. Aiming at this problem, this paper analyzes the uplink capacity when D2D communication is enabled in a multi-cell Code Division Multiple Access(CDMA) system. System level simulation is conducted to evaluate the uplink capacity gain of D2D link and its affecting factors. Simulation results show that D2D communication provides significant capacity gain. In addition, the performance gain decreases with D2D link distance and increases with the position of the link in system, but stays almost constant over different system loads.
An efficient pipelined architecture is presented in this paper for solving the problem of high hardware cost of R-L modular exponentiation algorithm, which is formed of Montgomery modular multiplication built by using pipelining technique. The parallel calculation of algorithm can be executed and the hardware cost can be also reduced in the new architecture. Besides, two extra pre-processing and post-processing for converting an integer to its N-residue format in the conventional modular exponentiation algorithm are avoided to reduce the iteration time. The result shows that the new architecture can achieve high data throughput rate of more than 14 Mb/s on Xilinx Field Programmable Gata Array(FPGA) of Virtex-2 series when performs modular exponentiation, while occupy only about half hardware resources when compared with the conventional parallel architecture.
The research on information collection of real-time rode traffic based on floating car is not perfect currently, with the problems of the complexity of collection content, inaccuracy of collection information. This paper presents a method for information collection of real-time road condition for the simplified road network model based on the GPS terminal, which includes information collection method for simplified network model considering road segment travel time and stopping time on the road. GPS terminals are in charge of the calculation of collection content and map matching rather than the server. Experimental results show that the method simplifies collection content and makes collected information more accurate, which can not only reduce the high load of the server-side, but also provide convenience for the transmission of collection information.
Based on moving object probabilistic nearest neighbor query in free space, the concept of constrained network moving object Probabilistic Nearest Neighbor query(CNPNN) is put forward, and the CNPNN algorithm based on network probabilistic Voronoi diagram is proposed. The probabilistic measure based on the network distance is used to derive the network probabilistic Voronoi cells of the uncertain objects, and the network probabilistic Voronoi diagram is built to cover the constrained network. R+ tree is used to index network probabilistic Voronoi cells for decreasing search time. Network probabilistic Voronoi cell containing query object is located to acquire query object’s most likely Nearest Neighbor(NN). Experimental results show that the time complexity of algorithm is O(n2+mlogmn), has a better performance under certain conditions.
Aiming at the problem of the holes caused by exposing regions in virtual viewpoint rendering, this paper proposes a virtual viewpoint rendering method based on image pyramid inpainting. The viewpoint warping equation is utilized to generate a virtual viewpoint. The image pyramid method is employed to restore holes inside the warped view. Depth information is incorporated into the cost function for Gaussian plus zero-canceling filter and down-sampling, low resolution image can be up-sampling, the expanded image information can be used to fill the holes. Experimental results show the proposed algorithm has superiority in both subjective details and the results of Peak Signal to Noise Ratio(PSNR), and virtual viewpoint of image rendering has not distortion, it can restrain the generation of false image in the edge.