In this project, we introduce a new Cellular Automata (CA) approach to construct an urban traffic mobility model. The cellular automata concept has been shown and is observed in our work that it is capable of capturing and reproducing realistic features of traffic flow. Due to its discrete nature, the CA model allows very fast implementation and can simulate a very large network microscopically in real time. Based on the developed model, characteristics of global traffic patterns in urban areas are studied. Our results show that how different control mechanisms used at intersections such as cycle duration, green split, and coordination of traffic lights could have a significant bearing on inter-vehicle spacing distribution and traffic dynamics. Several interesting observations are made and these findings provide important insights into the network connectivity behavior of urban traffic, which are essential for designing appropriate routing protocols for vehicular ad hoc networks in urban scenarios.

Vehicular ad hoc networks (VANETs) have emerged as a serious and promising candidate for providing ubiquitous communications both in urban and highway scenarios. Consequently, nowadays it is widely believed that VANETs will be able to support both safety and non-safety applications. For both classes of applications, since a zero-infrastructure is the typical premise assumed, it is crucial to understand the dynamics of network connectivity when one operates without relying on any telecommunications infrastructure. In this project, we provide a comprehensive framework for network connectivity of urban VANETs using the key metrics of interest (such as link duration, connection duration, and re-healing time). Our study, in addition to extensive simulations based on a new Cellular Automata Model for mobility developed earlier, also provides a comprehensive analytical framework. This analytical framework leads to closed form results which facilitate physical insight into the impact of key system parameters on network connectivity. The predictions of our analytical framework also shed light on which type of safety and non-safety applications can be supported by urban VANETs.

Several vehicular communication applications will involve multicast/broadcast communications where all vehicles in a certain region of interest are the intended recipients of particular messages. While there are several existing broadcast routing protocols for highway VANETs, very few solutions exist for urban VANETs in cities like New York City or Chicago. This project attempts to fill this gap by proposing a new broadcast routing protocol, namely UV-CAST, a completely distributed broadcast protocol that can be implemented by using only the local information available to each vehicle in an urban VANET. The protocol is designed by taking into account the two-dimensional road topology in urban settings. In contrast to one-dimensional highway scenarios, routing protocol design in urban areas is a much more challenging task for many reasons: i) direction of vehicles in urban areas may change at intersections while direction of vehicles on highways do not change until they leave the highway; ii) while a message in highway scenarios is disseminated in only one direction, message dissemination direction in urban areas may encompass 360 degrees. The performance of the UV-CAST protocol has been evaluated in terms of reachability, received distance, and network overhead in a regular Manhattan Street scenario as well as in a real city. Overall, the results show that the performance of the new UV-CAST protocol is excellent. While the proposed UV-CAST protocol assumes no infrastructure support, it can also utilize infrastructure support whenever it exists. Such infrastructure support could further enhance the performance of the UV-CAST protocol.

As the previous projects are mainly focused on safety applications in VANETs, we envision that non-safety applications such as such as Internet Access, Content, Map, or Database Download (CMDD) will be the first ones to emerge and be realized in VANETs. These applications require end-to-end (e.g., connection-oriented) transmission control in order to guarantee a desired level of performance. Nonetheless, designing a Transport Control Protocol (TCP) in VANETs is a very challenging task as the end-to-end data transmission involves multi-hop wireless transmission. The mobility and wireless channel conditions (such as fading, shadowing, multipath, etc.) exacerbate the well-known problems of using TCP over wireless links. In this project, a new VANET TCP is proposed which can detect packet losses, identify the different causes of such losses (such as congestion, disconnection due to mobility, wireless channel conditions), and resolve such problems using different and suitable mechanisms for each case. It is shown that casting the end-to-end transport control as a routing problem whereby the stability of each link is ensured in a dynamic fashion is a powerful approach to guaranteeing the stability of a multi-hop end-to-end route. Finally, unlike the existing bottom-up approaches, a cross-layer design paradigm whereby the network layer utilizes the transport layer information in making routing decisions is shown to be beneficial. The proposed new approach seems very promising as it can support end-to-end control for up to 5 hops.

Urban transportation is one of the grand challenges of our times. This is a chronic problem which is only getting worse as the migration to urban areas is accelerating. In its present form, the use of traffic lights to control and regulate urban traffic has proven to be very costly and inadequate in terms of scalability. Leveraging the new vehicle-to-vehicle communication capability of modern cars, we propose the use of in-vehicle traffic lights for informing the drivers to cross or stop at an intersection, replacing roadside traffic lights by temporary virtual infrastructures that are created and operated by elected vehicles at each intersection. We provide compelling evidence that this self-organizing traffic paradigm can increase the average flow rates substantially (by up to 60%) in addition to rendering traffic control and management ubiquitous. It is shown that the proposed in-vehicle traffic lights scheme is a scalable and cost-effective solution. More information could be found at VTL website