Even though the introduction of ICT in transportation systems leads to several advantages in terms of efficiency of transport, mobility, traffic management, and in improved interfaces with other modes of transport, it also brings some drawbacks in terms of security also dependent on the human behavior. For this reason, in the last decades, attempts to characterize drivers' behavior have been mostly targeted towards risk assessment and, more recently, to the training of machine learning soft- ware for autonomous driving. In this paper, we propose for the first time to use driver behavioral characterization to build a general reputation profile, that can be used to realize inno- vative, reputation-aware automotive services. As a first step towards realizing this vision, we present guidelines for the design of a secure vehicular infrastructure that is capable of collecting information generated from vehicles sensors and the environment, and to compose the collected information into driver reputation profiles. In turn, these profiles are securely exchanged in the infrastructure to realize reputation-aware automotive services, a sample of which are described in the paper. As a fundamental component of the infrastructure, we show that: iq multi-dimensional reputation profiles can be formed building upon the recently introduced notion of driver DNA; and iiq multi-dimensional comparison of profiles can be achieved by means of a reputational lattice rooted in the notion of algebraic c-semiring.
A privacy preserving infrastructure for driver's reputation aware automotive services
G Costantino;I Matteucci;F Martinelli;P Santi
2019
Abstract
Even though the introduction of ICT in transportation systems leads to several advantages in terms of efficiency of transport, mobility, traffic management, and in improved interfaces with other modes of transport, it also brings some drawbacks in terms of security also dependent on the human behavior. For this reason, in the last decades, attempts to characterize drivers' behavior have been mostly targeted towards risk assessment and, more recently, to the training of machine learning soft- ware for autonomous driving. In this paper, we propose for the first time to use driver behavioral characterization to build a general reputation profile, that can be used to realize inno- vative, reputation-aware automotive services. As a first step towards realizing this vision, we present guidelines for the design of a secure vehicular infrastructure that is capable of collecting information generated from vehicles sensors and the environment, and to compose the collected information into driver reputation profiles. In turn, these profiles are securely exchanged in the infrastructure to realize reputation-aware automotive services, a sample of which are described in the paper. As a fundamental component of the infrastructure, we show that: iq multi-dimensional reputation profiles can be formed building upon the recently introduced notion of driver DNA; and iiq multi-dimensional comparison of profiles can be achieved by means of a reputational lattice rooted in the notion of algebraic c-semiring.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.