Using Utility and Microutility for Information Dissemination in Vehicle Ad Hoc Networks Juan Liu, Daniel Greene, Marc Mosko, Jim Reich, Yukio Hirokawa, Tatsuo Mikami, and Tomoyoshi Takebayashi Abstract— We describe an approach to propagating streams of information in Vehicle Ad-Hoc Networks (VANETs) based on sources of information anticipating where their information will be useful. In this paper we describe how sources can model the potential usefulness of their information using utility functions. These utility functions are converted to more compact “microutilities” that travel with the individual data packets. The microutilities allow the information forwarding protocols to operate distributedly and independently on individual data packets, while achieving good overall coordination and delivery for entire data streams. We describe the algorithms used to convert utility functions to microutilities. Our algorithms insure that both proactive planning and reactive dropping of information in-transit are done consistent with the needs of the different applications. In this way data streams from both high priority (safety) and lower priority (traffic and commercial) applications can be propagated in the same network. We show experimental results that demonstrate the advantage of such a utility–microutility approach in serving the needs of diverse intelligent transportation system applications. I. I NTRODUCTION Vehicle Ad-Hoc Networks (VANETs) are especially chal- lenging environments for spreading information between vehicles and other transportation participants. Network topol- ogy changes rapidly, the density of the network shows extreme differences between rural and urban areas, and trucks and buildings block and interfere intermittently with transmissions. VANETs require highly robust information dissemination protocols. At the same time, VANETs are expected to support diverse information dissemination ap- plications, including, at one extreme, safety applications requiring frequent and reliable information delivery at short ranges, and at other extremes, traffic and entertainment applications requiring less-frequent, less-reliable delivery at longer distances. In this paper we develop quality of service (QoS) mechanisms that allow multiple diverse applications to share the same constrained communication infrastructure. We use a novel utility-driven approach, where at the data sources each application uses a utility function to specify its desired spatial and temporal data delivery characteristics. The information is propagated through the network based on its expected utility to recipients. The use of utility function allows the overall system to allocate communication This work is joint work of Palo Alto Research Center and Fujitsu Limited. Juan Liu, Daniel Greene, Marc Mosko, and Jim Reich are with Palo Alto Research Center, USA. Email: {jjliu, greene, mmosko, jreich}@parc.com Yukio Hirokawa, Tatsuo Mikami, and Tomoyoshi Takebayashi are with Fujitsu Limited, Japan. Email: {y.hirokawa@jp.fujitsu.com, nra24256@nifty.com, take@labs.fujitsu.com} resources optimally across data sources, and enables indi- vidual data sources to fine-tune communication expenditure tailored to their respective information delivery needs. At the same time, it avoids routing, subscription, or connection infrastructure that may be cumbersome and less reliable in the harsh VANET environment. The applications can adapt their utility functions to local traffic patterns to better serve their recipients. Here we focus on streams of data samples, where the utility is not for single data samples, but rather for rates of delivery of samples from the same source. Many vehicle information dissemination applications fit this pattern. For situational awareness applications, regular samples of loca- tion and velocity data will be useful to recipients, and the utility will increase with higher rates of data (since they will increase the fidelity of the filtering and tracking algorithms). For entertainment applications, video streams can be encoded with progressive or error-correcting schemes so that image fidelity (and hence utility) increases with higher data rates. Our approach derives simple “microutilities” from the utility functions supplied at the source. The microutilities are considerable simpler and more compact than the util- ity functions; they travel with the individual data samples and enable quick in-transit decision making. Microutilities work in two ways. They proactively implement Variable Resolution Information Dissemination (VRID), which sends a different amount of information to nodes at different distances. Secondly, reactive mechanism in the microutilities lets the forwarding nodes of the network drop data in a way that is consistent with the overall utilities of the data types flowing in the network. The actions of many microutilities in aggregate approximates utility. A. Related work The use of utility and microutility extends our earlier work on variable-resolution information dissemination (VRID) [1]. A companion paper [2] describes an entire information dissemination system that is based on the concepts of utility and microutility explored in this paper. We have made use of several economic concepts, espe- cially the concept of utility. Others have proposed to use economic mechanisms (e.g. markets and auctions) to allocate computational and networking resources (e.g. [3], [4]). We have not adopted such a full market approach because of the complexity it would impose on individual elements of the system. In our system, costs are set algorithmically, utility is supplied by application writers on behalf of all expected