The complex traffic environment of the intersection region makes traffic accidents occur frequently in this section. Intelligent network vehicles can give early warning of possible dangers through communication between vehicles and other traffic participants, so as to reduce the occurrence of traffic accidents. At present, many automobile enterprises are studying the landing application of vehicle to everything (V2X). This paper focuses on the application requirements of intersection collision warning (ICW) to V2X. Testing is an integral part of vehicle internet and can ensure safe and effective use of vehicle internet.In this paper, we describe in detail the testing process and results of communications performance of vehicle-to-vehicle (V2V) at an obstructed intersection in a closed test field where there are a lot of cars as a background. In addition, we have proposed a message forwarding mechanism based on vehicle to vehicle to infrastructure to vehicle (V2I2V), and the test results show that V2I2V performs better than V2V at an intersection with obstacles.

Today's data center networks have already pushed many functions to hosts. A fundamental question is how to divide functions between network and software. We present DumbNet, a new data center network architecture with no state in switches. DumbNet switches have no forwarding tables, no state, and thus require no configurations. Almost all control plane functions are pushed to hosts: they determine the entire path of a packet and then write the path as tags in the packet header. Switches only need to examine the tags to forward packets and monitor the port state. We design a set of host-based mechanisms to make the new architecture viable, from network bootstrapping and topology maintenance to network routing and failure handling. We build a prototype with 7 switches and 27 servers, as well as an FPGA-based switch. Extensive evaluations show that DumbNet achieves performance comparable to traditional networks, supports application-specific extensions like flowlet-based traffic engineering, and stays extremely simple and easy-to-manage.

Bulk transfer on the wide-area network (WAN) is a fundamental service to many globally-distributed applications. It is challenging to efficiently utilize expensive WAN bandwidth to achieve short transfer completion time and meet mission-critical deadlines. Advancements in software-defined networking (SDN) and optical hardware make it feasible and beneficial to quickly reconfigure optical devices in the optical layer, which brings a new opportunity for traffic management on the WAN.

In order to effectively determine whether a vehicle is turning or not, we proposed a method to map arbitrary consecutive GPS heading information to 2 dimensional feature space. Then we applied K-means clustering algorithm to divide the feature space into 2 classes: going straight and turning. After that, we used supervised learning algorithm to analyze these labeled data and build a model to recognize vehicle moving state. The experimental results showed that the model built in this way has good generalization. Based on the above research achievement, we designed and implemented a vehicle moving state recognition learning system for dynamic in-car navigation systems and applied this learning system to the map-matching field. The improved map-matching algorithm was tested on a complex urban road network and the result showed that the new algorithm can significantly improve the performance of the junction match.

This paper presents a heterogeneous network interconnection model based on a kind of translucent routing gateway and Linux virtual device to tackle with the communication issue about heterogeneous communication between various devices and Internet in the vehicular networks. In this model, an in-vehicle CAN node is regarded as a usual node in the Internet and will be assigned an IPv6 address and a MAC address as well. This mechanism will make the external terminals connect to these nodes without the perception of the in-vehicle gateway, which will implement the remote condition monitoring and fault diagnosis on a specified unit in the in-vehicle networks. Experiment result shows that this model may establish the communication between the nodes and various devices in the heterogeneous networks.

Order-k Markov Model can be used in many fields such as Natural Language Understanding, Coding, Mobile Path Prediction and so on to make prediction and then control. But the model has to face the problem of state space expansion. Taking the mobile path prediction as the research background, the paper firstly proposes a Step-k Markov model and validates its feasibility. Secondly, a hybrid Markov predictor model is put forward based on the Step-k Markov model. The complexity of the Hybrid Markov Model is O(N) while the Order-k Markov model is O(N2). And the memory demand of the hybrid Markov model is O(N2) while Order-k Markov model is O(N3). Finally, it is proved that the hybrid Markov predictor can get close performance with Order-k Markov Predictor at much lower expense by conditional entropy analysis and user mobility data analysis. Also, it can alleviate the zero probability problem in Order-k Markov model to some extent. The hybrid Markov predictor is more practical than Order-k Markov predictor under WLAN.

Path prediction is an important issue in QoS of wireless networks. The paper points out problems in some existed path prediction schemes, especially the state space expansion problem in order-k Markov predictor. And it firstly proposes a step-k Markov model and validates its feasibility. Secondly, a hybrid Markov predictor model and its improved models are put forward based on the step-k Markov model. Because of the order-2 Markov model's best performance in order-k Markov models, the Hybrid Markov model takes the order-2 Markov model as its target. The state space's complexity of the Hybrid Markov Model is 0(N) while the order-2 Markov model is O(N2). And the memory demand of the hybrid Markov model is O(N2) while Order-2 Markov model is O(N3). Finally, it is proved that the hybrid Markov predictor can get close performance with order-2 Markov predictor at much lower expense by conditional entropy analysis and user mobility data analysis. Also it can alleviate the zero probability problem in order-k Markov model to some extent. The hybrid Markov predictor is more practical than order-k Markov predictors under WLAN.