Ing-Chau Chang

Efficient VANET Unicast Routing Using Historical and Real-Time Traffic Information

In this paper, we propose an intersection graph-based vehicular ad hoc network (VANET) architecture. Using the available electronic MAP and historical traffic statistics from public traffic databases, we create an intersection graph (IG) consisting of all connected road segments, which have shorter average inter-vehicle distances than the wireless transmission range, as its edges and intersections of these road segments as its vertices. We then calculate the least cost routing path in the IG. Hence, the source vehicle leverages the proposed IG and IG bypass routing protocols to greedily forward unicast packets to the destination vehicle via each intermediate intersection on the least cost IG path. Further, we also propose the IG routing path recovery process to handle the broken IG path in real-time. Finally, we execute NS2 simulations to exhibit that the IG and IG bypass routing protocols significantly outperform four well-known VANET ones in terms of the average packet delivery ratio, end-to-end delay and hop count.

A Secure Ambulance Communication Protocol for VANET

Vehicular ad hoc networks (VANETs) have been a research focus in recent years. VANET’s main applications are enhancing road safety and reducing traffic accidents. Moreover, the VANET system can also reduce the time it takes for emergency vehicles to arrive at the accident location. The security of the transmission messages is of utmost importance, and to protect the transmission messages we propose a secure ambulance communication protocol for VANET to ensure that messages will not be revealed or stolen. The proposed scheme combines symmetric encryption, message authentication codes and digital signature mechanisms, and thereby achieves non-repudiation, availability, integrity, confidentiality, mutual authentication, session key security, known-key security and the ability to prevent known attacks. Finally, with NS2 simulation results that are based on realistic vehicle density statistics and the Taipei city road map, we argue that our secure ambulance communication protocol is effective in real VANET scenarios.

The IG-Ferry Protocol to Support Efficient Controlled Replication in Vehicular Delay Tolerant Network

In the past, we have proposed an intersection graph (IG)-based vehicular ad hoc network (VANET) architecture for efficient unicast communication in the city urban. Thus, the source vehicle can leverage the proposed IG and IG bypass routing protocols to greedily forward unicast packets to the destination vehicle via each intermediate intersection in the calculated least cost IG path. However, when the traffic density is not high enough, there may be several small and disjoint IG graphs in the city. If the source and destination vehicles are not located at the same IG graph, the IG routing protocol fails to find the least cost path between them to forward unicast packets. In this paper, we extend the concept of controlled replication in the Vehicular Delay Tolerant Network (VDTN) to efficiently spray a limited number of packet copies, which is denoted by the value of packet tokens, to three different kinds of relay vehicles, especially the public bus which has a fixed itinerary between the source and destination, with our proposed IG-Ferry protocol and the delay evaluation function. Finally, we conduct NS2 simulations to show that this IG-Ferry protocol outperforms three well-known controlled replication protocols, in turns of average packet delivery ratio and average end-to-end transmission delay of delivered packets.

On design of a dynamic carpooling system based on vehicle information shared through the VANET

Because both drivers and passengers may have different time constraints and preferences for carpooling, it is hard to manually find the best carpooling match, i.e., the lowest fuel-cost route, between them in the traditional carpooling web site. With the help of real-time traffic information exchanged among nearby vehicles through the vehicular ad-hoc network (VANET), this paper proposes the dynamic Carpooling System with real-time Vehicular Information (CSVI) system architecture, algorithms and message flows to calculate the route with the lowest fuel cost, according to dynamic route planning results of the VBA* algorithm that we previously proposed. These algorithms significantly reduce the time complexity under a huge number of possible carpooling combinations. Finally, simulation results exhibit excellent performance results of this VANET-based carpooling system over traditional ones, with the help of VBA* and proposed CSVI algorithms.

An Efficient Two-Mode Routing Protocol in Dense and Sparse Vehicular Ad Hoc Networks

This paper adopts a two-mode intersection graph-based routing protocol to support efficient packet forwarding for both dense and sparse vehicular ad hoc networks (VANET). We first create an intersection graph (IG) consisting of all connected road segments, which densities are high enough. Hence, the source vehicle leverages the proposed IG/IG bypass mode to greedily forward unicast packets to the boundary intersection via the least cost path of current IG. We then perform the IG-Ferry mode to spray a limited number of packet copies via relay vehicles to reach the boundary intersection of another IG where the destination vehicle resides. NS2 simulations are conducted to show that the two-mode IG/IG-Ferry outperforms well-known VANET routing protocols, in terms of average packet delivery ratios and end-to-end transmission delays.

A Credit-Based Incentive Protocol for Stimulating Network-Coded Cooperative Content Distribution in VANET

As the high-speed vehicular node passes the coverage area of a road side unit in the vehicular ad hoc network, it may not be able to receive all content information broadcast by the RSU during the short contact duration with the RSU. Hence, when the vehicular node leaves the coverage area of the RSU, it has to adopt the cooperative content distribution approach to share its received content pieces, which are encoded by the network coding technology, with all neighboring nodes. However, these approaches do not consider that vehicular nodes in reality may have different degrees of selfishness and interests on various types of content information, which in turn hinders those selfish nodes from sharing their contents with others. In this paper, we will propose the Social Code scheme, adopting the content bitmap, for each node to calculate the number of different network-coded content pieces, its corresponding re-broadcast priority and the correct number of re-encoded NC content pieces with all neighbor nodes first. Social Code then classifies vehicular nodes into four types and further integrates corresponding credit-based incentive protocols to stimulate nodes willingness to re-broadcast their content pieces. Finally simulation results show that Social Code outperforms over traditional NC-based cooperative content distribution schemes, if they adopt the credit-based incentive protocols without considering node selfishness and interests.