Software Defined Vehicular Networks Research Articles

Building SDN-Based Agricultural Vehicular Sensor Networks Based on Extended Open vSwitch

Software-defined vehicular sensor networks in agriculture, such as autonomous vehicle navigation based on wireless multi-sensor networks, can lead to more efficient precision agriculture. In SDN-based vehicle sensor networks, the data plane is simplified and becomes more efficient by introducing a centralized controller. However, in a wireless environment, the main controller node may leave the sensor network due to the dynamic topology change or the unstable wireless signal, leaving the rest of network devices without control, e.g., a sensor node as a switch may forward packets according to stale rules until the controller updates the flow table entries. To solve this problem, this paper proposes a novel SDN-based vehicular sensor networks architecture which can minimize the performance penalty of controller connection loss. We achieve this by designing a connection state detection and self-learning mechanism. We build prototypes based on extended Open vSwitch and Ryu. The experimental results show that the recovery time from controller connection loss is under 100 ms and it keeps rule updating in real time with a stable throughput. This architecture enhances the survivability and stability of SDN-based vehicular sensor networks in precision agriculture.

The Characterizes of Communication Contacts Between Vehicles and Intersections for Software-Defined Vehicular Networks

In software-defined vehicular networks, a centralized controller in the network through the road-side unit will greatly promote the efficiency of the network efficiency by temporarily storing the packets, improving the delivery ratio and providing highly reliable transmission. The inter-contact time between moving vehicles and intersections, and the duration of each contact are two key metrics in the designing and evaluating of this kind of software-defined vehicular networks, because they decide the frequency of contacts between vehicles and the intersections and how much data can be sent during the contact. In this paper, we analyze traces of thousands of operational taxies in Shanghai. By studying the inter-contact time and the duration of each contact between taxi-intersection pairs, we find that the distribution of inter-contact time follows a power-law up to a small value and decays exponentially afterwards. We find that a power-law with exponential decay model can approximate the distribution of inter-contact time better than exponential model, power law model and piecewise exponential decay model. The duration of contacts, on the other hand, mainly exhibits an exponential decay. Our findings make way for the future study of and the deployment of software-defined vehicular networks.