Accurate traffic flow forecasting is crucial to the development of intelligent transportation systems and advanced traveler information systems. Since Support Vector Machine (SVM)have better generalization performance and can guarantee global minima for given training data, it is believed that SVR is an effective method in traffic flow forecasting. But with the sharp increment of traffic data, traditional serial SVM can not meet the real-time requirements of traffic flow forecasting. Parallel processing has been proved to be a good method to reduce training time. In this paper, we adopt a parallel sequential minimal optimization (Parallel SMO) method to train SVM in multiple processors. Our experimental and analytical results demonstrate this model can reduce training time, enhance speed-up ratio and efficiency and better satisfy the real-time demands of traffic flow forecasting.

in face of urban traffic congestion, the existing trafficcontrol system available today pursues rapid evacuation of vehicles at singleintersection and fails to take into account of the coordination between intersections,resulting in the worsening of urban traffic congestion. In order to solve theurban traffic congestion problem, an adaptive control based on the priority ofthe city path is proposed based on the PageRank priority model, a bettersolution to the traffic congestion of the city. In this strategy, the aim is tominimize the queue length of vehicles in urban road network. The releasepriority of each section is calculated based on the PageRank model. For asingle intersection point of control, the vehicle traffic volume is optimizedby the Maximum Clique Problem theory. The experiment takes 19 intersections inDalian as examples and is carried out on the VISSIM traffic simulation platform.The simulation results show that the adaptive control based on the pathpriority has good effect on reducing urban traffic congestion and evacuating vehicles.

As a type of Mobile Ad-Hoc Network (MANET), Vehicle Ad-hoc Network (VANET) has its own features such as high speed vehicles, large number of vehicles and highly dynamic driving environment. These features bring new requirements to VANET. Transmission speed, respond time and reliability level of communication have very strict instantaneity and Quality of service (QoS) demands. There are many optimized designs to satisfy the demands on each layer of communication in VANET. But there are lack of optimized works on safety messages’ communication sub-layer. In this paper, we introduce queuing theory to model and design the communication sub-layer. The results of simulation and field test show our design has achieved the requirements on both performance and function. The average improvements on performance are one times than before.

No abstract

Security of authentication protocol is an important factor to network security. To ensure protocol security, attacks on authentication protocol were classified from the aspect of security theory and implement. On this basis, an authentication protocol security assessment was discussed. The framework determines security risk of authentication from security of password system, vulnerability of protocol itself and environment threat possibility. It can be used to assess security and capability to resist risks objectively