In English

Implementing a Cooperative Driving Scenario using Chaos - Evaluating Chaos in the Context of Inter-vehicle Communication

Aron Manucheri
Göteborg : Chalmers tekniska högskola, 2016. 86 s.
[Examensarbete på avancerad nivå]

This thesis investigates the use of the data-sharing primitive Chaos in the domain of cooperative driving. Chaos is a primitive which natively supports all-to-all data sharing in low-power wireless networks. Chaos enables programmable in-networking processing in conjunction with synchronous transmissions, and this enables efficient sharing of data, in contrast to traditional approaches.

The thesis investigates if Chaos can be used for a cooperative driving scenario consisting of vehicles driving through an intersection. It evaluates if Chaos’ unique properties, with regard to all-to-all data sharing, can be used successfully for the scenario. Doing this the thesis evaluates Chaos in the context of inter-vehicle communication, seeing if the efficient data sharing capabilities of Chaos can lay ground for more efficient cooperative driving solutions.

By building a complete simulation system, in which the simulation representing the cooperative-driving scenario is simulated, and which uses Chaos to handle all the data sharing, we can obtain results of how Chaos performs for the scenario, and the feasibility of Chaos in such contexts.

We found in this thesis that Chaos can be successfully used for the scenario of vehicles driving through an intersection, in which the vehicles decide their actions solely by inter-vehicle communication, using Chaos.

Chaos provides high reliability, with successful sharing of data with around 30 participating vehicles. It also provides good scalability, with data sharing delays in the area up to a few hundred milliseconds even for scenarios with around 30 vehicles. Much of this delay was found to be implementation dependent, based on the maximum clock speed of the underlying hardware in which Chaos runs on, and thus not a constraint of Chaos. While not tested experimentally, nothing speaks against that this scalability and reliability should hold for even more vehicles, with a linear increase in data sharing latency. From the evaluation Chaos is found to be a good fit for the domain.

Nyckelord: Chaos, cooperative driving, inter-vehicle communication, wireless communication, data sharing, vehicular ad hoc network, autonomous vehicles, all-to-all data sharing, network primitive, vehicle communication

Publikationen registrerades 2016-04-27. Den ändrades senast 2016-04-27

CPL ID: 235325

Detta är en tjänst från Chalmers bibliotek